IMAGE EVALUATION TEST TARGET (MT-3) // .^'' V^;"* < '^.M ,^. fe z i5> 1.0 I.I 1^12^ 12.5 1.8 11-25 il.4 IIIIII.6 Va e /f v: ?> / •^^' A V Photographic Sdences Corporation 23 WEST MAIN STREFT WEBSTER, NY. 14580 (716) 872-4503 CIHM/ICMH Microfiche Series. CIHM/ICMH Collection de microfiches. Canadian Institute foi Historical Microreproductions / Institut Canadian de microreproductions historiques Technical and Bibliographic Notes/Notns techniques et bibliographiques The Institute has attempted to obtain the best original copy available for filming. Features of this copy which may bo bibliographically unique, which may alter any of the images in the reproduction, or which may significantly change the usual method of filming, are checked below. D D D D D D D D □ D Coloured covers/ Couverture de couleur Covers damaged/ Couverture endommag6e Covers restored and/or laminated/ Couverture restaurie et/ou pellicul6e Cover title missing/ Le titre de couverture manque Coloured maps/ Cartes giographiques en couleur Coloured ink (i.e. other than blue or black)/ Encre de couleur (i.e. autre que bleue ou noire) Coloured plates and/or illustrations/ Planches et/ou illustrations en couleur Bound with other material/ Reli6 avec d'autres documents Tight binding may cause shadows or distortion along interior margin/ La reliure serrde peut causer de I'ombre ou de la distortion le long de la marge intirieure Blank leaves added during restoration may appear within the text. Whenever possible, these have been omitted from filming/ II se peut que certaines pages blanches ajouties lors d'une restauration apparaissent dans le texte, mais, lorsque cela 6tait possible, ces pages n'ont pas 6t6 filmies. Additional comments:/ Commentaires supplimentaires; L'Institut a microfilm^ le meilleur exemplaire qu'il lui a 6ti possible de se procurer. Les details de cet exemplaire qui sont peut-Atre uniques du point de vue bibliographique, qui peuvent modifier une image reproduite, ou qui peuvent exiger une modification dans la m6thode normale de filmage sont indiquis ci-dessous. □ Coloured pages/ Pages de couleur □ Pages damaged/ Pages endommag6es □ Pages restored and/or laminated/ Pages restaurdes et/ou pelliculies □ Pages discoloured, stained or foxed/ Pages dicolordes, tache'.ies ou piqudes □ Pages detached/ Pages d(6tach6es Showthroughy Transparence Quality of prir Qualiti indgale de I'impression Includes supplementary materit Comprend du materiel suppl^mentaire Only edition evailable/ Seule Edition disponible I I Showthrough/ I I Quality of print varies/ r~n Includes supplementary material/ I I Only edition evailable/ D Pages wholly or partially obscured by errata slips, tissues, etc., have been refilmed to ensure the best possible image/ Les pages totalement ou partiellement obscurcies par un feuillet d'errata, une pelure, etc., ont 6t6 filmies A nouveau de fapon A obtenir la meilleure image possible. This item is filmed at the reduction ratio checked below/ Ce document est filmi au taux de reduction indiquA ci-dessous. 10X 14X 18X 22X 26X 30X ] 12X 16X 20X 24X 28X 32X The copy filmed here has heen reproduced thanks to the generosity of: Legislative Library of British Columbia L'exemplaire filmd fut reproduit grdce d la g6n6rosit6 de: Legislative Library of British Columbia The images appearing here are the best quality possible considering the condition and legibility of the original copy and in keeping with the filming contract specifications. Les images suivantes ont 6t6 reproduites avec le plus grand soin, compte tenu de la condition et de la nettet^ de l'exemplaire film6, et en conformity avec les conditions du contrat de filmage. Original copies in printed paper covers are filmed beginning with the front cover and ending on the last page with a printed or illustrated impres- sion, or the back cover when appropriate. All other original copies are filmed beginning on the first page with a printed or illustrated impres- sion, and ending on the last page with a printed or illustrated impression. The last recorded frame on each microfiche shall contain the symbol — ^ (meaning "CON- TINUED"), or the symbol V (meaning "END"), whichever applies. Les exemplaires originaux dont la couverture en papier est imprimde sont film6s en commencant par le premier plat et en terminant soit par la dernidre page qui comporte une empreinte d'impression ou d'illustration, soit par le second plat, selon le cas. Tous les actres exemplaires originaux sont film^s en commenpant par la premidre page qui comporte une empreinte d'impression ou d'illustration et en terminant par la dernidre page qui comporte une telle empreinte. Un des symboles suivants apparaltra sur la dernidre image de cheque microfiche, selon le cas: le symbole — ♦► signifie "A SUIVRE", le symbole V signifie "FIN". Maps, plates, charts, etc., may be filmed at different reduction ratios. Those too large to be entirely included in one exposure are filmed beginning in the uppei- left hand corner, left to right and top to bottom, as many frames as required. The following diagrams illustrate the method: Les cartes, planches, tableaux, etc., peuvent 6tre filmds d des taux de reduction diffdrents. Lorsque le document est trop grand pour dtre reproduit en un seul clichd, il est film6 d partir de Tangle sup6rieur gauche, de gauche d droite, et de haut en bas, en prenant le nombre d'images n^cessaire. Les diagrammes suivants illustrent la m^thode. 1 2 3 1 2 3 4 5 6 DE PONTIBUS: A POCKET-BOOK FOK BRIDGE ENGINEERS. BY J. A. L. WADDELL, C.E., U.A.So., Ma.E.; Knhiht Comummler of the Jopnnese Order of the Kisiug Sun: Con- sulting Engineer, Kanma City, M".; Memher of the American Society of Civu iLiigineers; of La Societe den Ingenieurs Civils Paris: of the Rensselaer Society of Engin- eers and the Society for the Promotion of En- gineering Education ; and Honorary Member of the Kogaku Kyokai (Japanese Engineer- ing Society.) FIRST EDITION. FlUST THOUSAND. NEW YORK: JOHN WILEY & SONS. London CHAPMAN & HALL, Limited. 1898. Copyright, 1898, BY J. A. L. WADDELL. BonitRT nnnMMONP. klkctuottpkr A^p phintkr, ^f.w vorb. To iUc(5iU Umocrsitn, tlie representative Institution of learning of the Dominion of Canada {the mitfior's native country), this little treatise is DEDICATED as a mark of the author's grateful appreciation of the distinction accorded him by that university in 188?. in conferring upon him ttco engineering degrees. 1;)L074 PREFACE. In presenting; to the public a new teclinicul work, it is the rtistoiii to offer some sort of apology for its appearance ; hence tile aullior of lliis treatise, in order not to be C(»ii8itlere(i pecu- liar, feels it incunil)ent upon liiui to do liltewise. Moreover, tliurtj is in this case a good reason for beginning his book with an apology, l)ccau8e of his audacity in imposing upon the good nature of the engineering ))rofessiou by asking its mein- 1)6 IS to read still another work iipon tlie already overwritten subject of bridges. If he cmild do so, the author would here plead primuin tnnpns; but this is l)y no means liis first ollente. Perhaps, liioiigh, the fact liiat it is twelve years since tlie appe.-irance of his last book (p.implilets, of couise, ex- cepted) will be considered by his critics as'a "mitigating cir- cumstance" in this case. But, to speak seriously, if this work were a mere rehash of otjjer boolcs, or if It dealt witlj tlie same old, worn-;)Ut sub- jects, the author would not presume to present it to the engineering profession ; but, on the contrary, in writing it he lias endeavored to nuike the contents as original as possible, and to treat essentially of the finidamental principles of bridge-designing and their application. It will be noticed throughout the book that quotations from other works on bridges are "conspicuous by their absence," and that the autlior has drawn almost entirely upon liis own ])ri>fessioual practice for examples to ilhistrale Ids text. For the latter no a|>ology is recjuired, because his own designs (as far as the process of development lias permitted) have naturally been made in conformity with the principles which he herein offers as a guide to bridge-designing ; and tljey are therefore more appropriate as illustrations than the designs of others. VI pkkfa(;r. The author desires it lo br disliiiclly iiiKU'istood jii the out- set tliiit lif by no iiu-titis claiiuH that Uk.* nietlioils of 8 Hs tliosc of Profs. Kurr, Dii BoIh, and Jolinson, and lliul, if he were lu nttL'iii|)l to illustrulc tUe principles by ac'lmil exiiniplcs of designing, his book would never be tinislied. As staled in Clmpters XI and XIX, the second edition of tlie author's "General S|)eclttctitions for Highway Bridges of Iron and Steel" and the Mist edition of his "Coniproniise Stniidiird System of Live Loads for Hallway Hi idgcs and the Kquivalenls for Saini; " are now exhausted, and will not be reprinted, as this treatise will replace them. In writing Chapters XV. XVI, XVII. and XVIII it was found necessary to copy cerlain portions of Chapter XIV in order to make the various specilications complete; but the amount of repetition was made as small as possible by referring, wherever no changes were introduced, whole sec- tions of one set of speciiicalions to the corresponding sections in a preceding set. The subject of suspension bridges is not dealt with in this work, partly because until lately the author has not paid much attention to this class of struct uies, and partly because they are so dilTerent from other bridges, being suitable for very long spans only, that each suspension bridge requires special specltications of its own. The author has the presumption to hope that there will be considerable demand for this book, for he considers that it will be useful to the following classes of readers : first, to praiaising bridge-engineers, because of many little suggestions that will help them to effect improvements and to avoid mis- takes ; second, to young engineers in oflices of bridge s\)e- cialists and of bridge-manufacturing companies, for perfecting them in their work ; third, to professors of civil engineering, to show them the pmctical side of bridge-designing and building, and to aid them in giving their lectures on bridges ; fourth, to students of civil engineering, as a supplementary text-l)ook that will enable them to understand the application of what they have learned during their course in bridges; fifth, to railroad engineers, because of the bridge specltications contained, and to instil into their minds tbe importance of VIII PKKFACK. having their bridges properly dosignod, tnnniifiictured, in- spected, shipped, and erected ; and, sixth, to ti few county coininissioiiers, wlio may (icHire to obtuin througli the spccili- cations good liighway bridges at mininiiiin Icgititniite cost. The autlior lias endeavored to nial(e the various speciflcn- tions in this boolc tliorough, correct, and (;oinpIclc. If he lias failed to do so in any particular, lie woidd feel deeply in- debted to r.iiy one wlio will point out to him how and wliero ; and he would be grateful to any reader who will ill^llul lum of any typographical or other errors that he may discover , for all errors found in the tirst edition will be corrected in liie second, provided tlie work l»e well enough received i)y the profession to warrant the issue of another edition. In conclusion the author desires to acknowledge with many thanks his indebtedness to bis assistant engineers, Ira O. Hedrick, Assoc. M. Am. Soc. C E. ; I^ee Treadwell, M. Am. Soc. C. E. ; and John L. Harrington, Jun. Am. 8oc. C. E., for valuable aid rendered him in the preparation and checking of the MS, of this work. Kansas City, Mo„ Oct. 18, 1897. CONTENTS. r-HAPTKR PAQB I. Introduction . . ... 1 II. First Principles of r)<'Hi>riiiiiK Vi III Ti-ii« Kconomy in I)esij{n . .30 IV. .Kstlietics in Design 30 V. ('untile ver Hri(iK<'8. 5,5 VI. Ardies 79 VII. Tre.stlesai;.; Viutliicts 80 VIII. Klovateii Riiilroad.s 91 IX. M()Vftl)U' Bridtfes In (lenernl UW X. FU'volvinj? DravvhridKe* 119 XI. lliKliwiiy Bridges 130 XII. Combined Bridges . 133 XIII. DetailinK ]:18 XIV. Oeneral Specifications Ooverninj; the Designing of Steel Railroad Hridpcs and Viaducts anil the Siiperstnictuie of Elevated Railroads HI XV. Specifications for Railroad Draw Spans . . ItW XVI. General Specifications {;overidng the Designing of Steel Higliway Uridtces iind Viaducts. .. . . iit-i XVII. Specifications for HiKhway Draw-Sptns . •^.l' XVIII Ueueral Sjwcillcations (Jovernin>r tlie jVlaniifacture, Ship- ment, and Kre«;tion of Steel Bn Iges, Trehtles, Viaducts, and Elevated Railroads 34.) XIX. The Compromise Standard System of Live Loads for Rail way Bridges and the Kqui valenta for Same '.JOS XX. Timber Trestles in XXI. Inspection of Materials and Workmanship 'i6\ XXII. Designing of Piers 301 XXIII. Triangulatlon 317 XXIV. Office Practice ...328 TABUts 349 Index 373 u LIST OF TABLES. TAll TABLE PAGE 400 I. Coefficients of Impact for Railway Bridges, / = , . ,^ . . 351 L + nOO 100 II. Coefficients of Impact for Highway Bridges, '=,.,,„• -352 80/ III. Intensities for Inclined End Posts, P - 18000 - - SM r IV. Intensities for Top-Cliord Compression members, P ^ 18000 - — 354 r V. Intensities for Intermediate Posts and Subdiagonals, us)} P= 10000 - - 8.5.') r Intensities for Columns of Viaducts and Elevated itailroads, 00/ and for all Lateral Struts, /' ^ 16000 - - :«6 Centrifugal Force in iVrcenlages of Live Load . . , ',Vu Sizes and Weights of Stay-I'lates and Lacing-Bars for Ordi nary I'osts ,. 3.58 Bending Moments on Pins. 361 Bearing on Pins ... 362 Intensitie.s for Forlted Ends and Extension-Plates of Ci-m- pression-Members, P = 10000 - ^p' 363 Shearing and Bearing Values of Rivets 364 Coefficients of U'tantf for both Compression and Tension Stresses in Bottom Chords of Through-Bridges and Top Chords of D»'ok-Bridges, due to Wind Loads applied to s^aid Chords, when the Lateral System is of Double Cancella- tion 36R IV sec 9 XrV. Coefficients of • (where 71 = Number of Panels in n Span) for Wind-Load Stresses in the Diagonals of Lateral Systems of Single Cancellation. These Coetnclents apply X VI. VII. VIIL IX. X, XI. XII XIll. LIST OF TAI5LKS. xi TABLE XV. PAOK tfi Lateral Systems C()iim><>s«'(1 of Intersecting Diagonal UodsorofSiiiKi ■ I'iagoniil Struts a66 Coetllcients of U' tan fl foi- » '.unpression-Stresses iu Wind- ward Bottom Chords uf Throiigii Bridges, und Windward Top Chords of I)eck-Brid^'e», one to Wind Loads apphed directly to said Chords, when the Lateral System is of Single CaneelUitiun. The Tensile Stresses in Leeward Chords are iiuiuerically Kqual to the Compression Stresses given in Table for One Panel nearer End of Span 367 Intensities of Working Stresses foi Vnfit-ns Materials 368 Maximnni Stresses imder Dead and Live Loads in Pratt Trusses 'J'j' XVUl. Superelevations of Outer Rail on Curves... 374 XVI. XVII. ''V LIST OF PLATES OF CURVES AND DIAGKAMS. PLATE 1. Axle Concentrations for tlie Conipronuse Sta.ulard System of Live Loads for Railway Biitlges. 11, DiaKrani of Live Load End Shears for Raihviiy Bridges 111. DiaKiain of E<|iiivaleiil Live Loads for Railway Plate Girders. IV Diagram of Equivalent Live Loads for Trusses of Railway Bridges V. Diasnuu of Live Loads per Square Foot of Floor for Highway Bridges. VI. Diagram of Equivalent Live Loads from Electric Cars on Highway Bridges. VH. Diagram of Wind Loads for Railway Bridges, Vni Diagram of Wind Loads for Highway Bridges. IX. Diagram of Reactions for Balanced Loads on Draw Spans. X. Diagram of Weights of Metal in Trusses of Cantilever Bridges. xii DE PONTIBUS. ClIAPTKR I INTRODUCTION. Winr.E it is true tlmt the devulopment of bridge-building ill America owes much to the syst«'m so long in vog»« of bidding on ronipetilive plans, in tliat sucii competition has tended to sharpen the wits of the engineers of tlie competing companies, it is equally true that the sjiid competition lias done all the good it can for the sciouce of bridge-designing, and now acts as a clog to prevent its further tidvancemcnt. The correctness of this assertion .scarcely needs any demon.stratioii, but it may be well, notwitlistanding, to give here a few rea-sons therefor. As human nature is the .same the world over, and as men in general are working for the almighty dollar, it stands to rea- son that when a bridge-company's engineer is preparing upon fi.xed specifications a design to be used as a basis for a competi- tive bid, and when he knows that in nineteen cases out of twenty the contract will be awarded to the lowest bidder whose design conforms to the letter of the specitications, although it may not be up to the requirements of good engineering practice, he will take advantage of every weak point and omission in said specitications, even if his engineer's conscience proclaim the design he .submits to be worse liiaii faulty. As it is enlirely pnicticjible to take advantage of any set of railroad-bridge specilications yet published, it is evident, l)K I'OXTIBUS. that as long as conipftitivo, lump-sum bids arc the f.ishion, just so long will railroatl bridges be badly designed. As for highway bridges, their letting, ilesigning, and construction are so often left in tlie hands of such incompetent and unseru pulous parties that, until some fundamental cliauge in existing conditions be eltecled, notiiing can be done to improve the present imscieiititic, wretciied, and even crimiual methods of highway-bridge building. Concerning llie prejudicial eirect of competitive designing upon tlie development of the science of bridgc-engiru'cring, the author can spctak autlioritatively, l)ecause for about six years he acted as engineer to a bridge company During that time he lost many contracts for small bridges because lie insisted on incorporating in his designs certain features re (pnring extra metal, which features he considered essential, although they were not called for in the specitications. On the other hand, he once earned a conunission of more than ten thousaiul dollars upon a single piece of work by knowing how to lake the greatest advaniage of the specitications up>)n which bids were recpiested. In defence of this action, iiow- ever, it must be mentioned that it was undersio .d at the outset that, after tlie seleciiou of the successful competitor, the contract was to be adjusted upou the basis of a pound pri(;e for the metal-work. liy reason o'' this feature, the autiior was able to correct later on all tiie weak points of his prelimi- nary design to such an extent that the structure until within " few years was by far the best of its kiLi. .,uilt. Tins case is given merely as an illustration of how great are the possibilities for trimming a design wiiich is based upou onlinary standard spe<;iiications, and how great is the tempta- tion to take advantage thereof. Anotiur way to illustrate this point is to compare the weights of the structures manu- factured and built by any bridge eoinpauy by the lump sum with the weights of similar structures manufactured and built by the same company for a pound i)rice. 'J'iu! ditterenct^ in weight often runs as high as fifteen or twenty per cent or even higher, if there be no supervising engineer to hold the bridge company in check. ^^ |i I NT UO DUCT ION. 3 iisbion. As for ruction miscru xisliiig ovu the ukIs of i)g that For a raihviiy conipany, the most satisfactory method of building bridges is cither to have a pcrniaueut, competent bridge engineer in its employ, or to relaiti some specialist of established reputation to i)rep:ire specifications and complete detiiilcd plans (not working drawings, however) for all ils bridges, and to provide competent inspectors to see that during manufacture, shipment, and erectiou the plans and specifications are strictly followed. Tlio necessity for the specialist to stand between the pur- chaser and the manufacturer of structural steel is, as a general rule, not appreciated l)y the purchaser, \inles3 ho has already had some experience in letting contracts for and in the building of steel structures without engineering aid in the designing and supervision. When the purchaser puts himself in the hunds of the manufacturer, he is pretty sure to get tlie worst of it ; for if the contract be let by schedule prices the struc- tuie is lial)le to be loaded down with useless metal, wldle if the contrac't be let ft)r a lump sum the structure will probably he ruined by having the metal " skinned " out of it, especially in the most important parts, viz., tlie details. Moreover, the manufacturer is seldom capable of evolving a truly lirst-class design, for the reason that his training has always been in the line of his own pecuniary interests, which are to obtain the ma.vimum of pay for the minimum of structure ; so that, even when given all the metal and money that ho could ask for when preparing a design, he would not succeed in making a really good one, simpl}- becauNO of not knowing h(>w. On the other haiul, the specialist should stand between the contractor and th(! purchaser, so as to see that the latter does not take any undue advantage of the former by means of a harsh or unjust interpretation of the si)ecitications, especially when the contractor has suffered loss or delay on account of causes beyond his control. Occasionally a .nanufaeturer offers to prepare the plans for certain portions (if the work on the plea that he has had so much more e-xperionce iti such matters than the engineer. The accei)li.nce of this o!!er by either the purchaser or theengineer is a nusiake ; for the engineer, if he have sufficient ability to DK 1'0NTIBU8. warrant liis being rotaincd on ihe work, can by careful study almost always evolve a belier design than can tlie contractor, even if it be the lirst experience wliidi the former lias had in connection witii the portion of the work under consiileration. On two or three occasions only, and several years ago, the author was either induced or conipelleil by the purchaser to defer to the greater experience of the contracting engineer ; aud in each case he lias had reason to regret the concession ; so he has conchidcd that in future he will receive with thanks any suggestions which the manufacturer may ofl'er, give them due cousideraliou, and then make the design as he liimself sees fit. Considerable opposition to the methods of design advanced in this work and to the specilications given is anticipateil, on the plea that the requirements are too exacting and llial the class of work called fcr is unnecessarily refined and conse- quently expensive. To such opposition the author would reply as follows : First, the designing and ])uilding of bridges and similar structures cannot be too well or carefully done ; and, second, that within the last three years, upon some fifty or sixty thousand tons of the author's work designed in accord- ance with the said methods and built in accordance witli the said specifications, the prices quoted by the competing man- ufacturing companies were extraordinarily low, and tluit no complaint of any account has since been raised by the manu- facturers in respect to tlie expense involved by either the de- signs or the specifications. The principles of design given in the succeeding chapter should be adhered to in all structural metal-work ; and any violation of any one of them is a mistake that will be regretted sooner or later by the parties owning the structure. Many of these principles are violated constantly by shop draftsmen, even when the engineer's drawings s!)0W the details correctly. This is due partly lo custom in designing certain details in certain ways, aud partly to the ignorance of the draftsmen. The author would urge upon young engineers who are work- ing on plans for structural metal-work to adhere to the prin- ciples herein given whenever it is practicable for them to (Jp INTKODUCTIO.V. D so. Had more attention been paid to first principles of de- sign when tlie plans for most of the New Yorii and Brooklyn elevated railroads were beitii; jirepnred, millions of dollars would have been saved. This statement can be verilied ijy u perusal of the author's paper on Elevated liiiilroads, referred to and quoted from in Chapter VIII, more especially the resume of the discussions and Mr, Iledrick's report on tiie said New York and Brooklyn elevated railroads. In spile of all tlie care that the most expert desiguer can give to his work, errors of greater or less magnitude will creep in occasionally; and one can always improve somewhat upon any tinished design. Such being the case, it follows that the designing of steel structures should be intrusted to expert and disinterested designers only, instead of, as is generally the case, to the cheap draftsmen employed by the manufacturing companies. Tiiere are a few features of the specifications given in Chap- ters XIV. and XVIII. vvliich will require a little explanation or comment. This will be given in this chapter, the various items being treated, as nearly as may be, in the order in which they occur in the said specifications. "A" Truss Bridges. This style of structure, originated by the author and covered by letters patent, is a four-panel truss-bridge having eye-bars in bottom chords and centre verticals, and rigid members for all the other portions of the trusses and for the entire lateral system. It was evolved in this way; P'or a number of years the author was dissatisfied with all railroad bridges for spans between the superior limit of the plate-girder and a length of about one luindred and fifty feet, ordinary piu-connectcd, through, Pratt truss(!S being too light and vibratory, and the riveted bridges as then built being clumsy, unscientific, and imeconomical. On this account he tried for some time to find an opportunity to experiment upon a design of his own to till a portion of the gap, but the opportunity did not occur until April 1893, when he was retained by the General Manager of 6 DK I'ONriHifg. the Kansiis City, Piltsbiiru;, jiiid Quif Ruilrotul Company to design some bridges. Afier ti little porsuasiou the General jNranuger was induced to agree to hiiilil a 100-ft. "A" truss span as au experiiueut; but when Ik; sjiw the completed i»lans he ordered at once four bridges to be built therefrom, and tins style of structure was soon afterwards adopted as the standard 100- ft. span for the road. Tliese bridges have shown such rigidity under tralllc that they have been used on the St. Louis Southwestern Railway, and have been adopted as the standard for spans between si.\ly- live feet and one hundred and sixteen feet by the Nii)pon Kail- way Company of Japan. The advantages of this type of bridge are great rigidity in all directions, ease and clieai)ness of erection, and economy of metal when it is compared with structures of other types hav- ing equal stnength and i igidity. IMPACT. The uncertainty as to the magnitude of the ellcct of impact on bridges has for many years been a stiunl)ling-bl()ck in the path of systenuzation of bridge-designing, and wMl continue to be so until some one makes an exhaustive series of experi- ments upon the actual intensities of working stres.ses on ail main members of modern bridges of the various types. Tlie making of these experiments has long been a dream of tiie author's, and it now looks as if it would amount to more than a mere dream ; for the reason that the general manager of one of the principal Western railroads has agreed to join the author in the making of a number of such experiments on certain bridges of the author's designing, the railroad company to furnish the train and all facilities, and the general manager and the author to provide the ajjparalus and experimenters. It is only lack of time that has prevented these experiments from being made this year, and it is expected that they will b(! finished in 1898. It is hoped that tlifc result of the experi- ments will be either to determine a proper formida or curve of jierceiitages of impact for r.iilroad Itridges, or else to INTUODL'CTIO.V. 7 iniiugurate n soiit'S of fiiiUicr ox|H'riiiUMils lliat will (Ictenniiie it. Ml'!iii while the author lias adopted temporaiil}' the forimila givL'ii ill Ciiaptcr XIV., vi/, , / = 40000 L -f- oUO' in whi( h 1 U the percentage for impact to he added to the live I' iid, and L is the length in feet of span or portion of span iliat is covered hy the said load. This foiinula was estahlished to suit the average practice of lialf a dozen of the leading bridge engineers of the United States, as given in their standard specitlcations, and not because I he author considers tliat u will give truly correct percentages tor iinpiict. In spite of ail that has Iwjcn said to the contrary in the past or that may be said in the future, tlie impact method of proportioning bridges is the only rational and scientifically practical metliod of designing, even if the amounts of impact assumed be not aljsolulely correct ; for the method carries the effect of impact into every detail and group of rivets, instead of merely ailecting the sections of the main members, as do tlie other methods in common u.sc. The assumption madi; in some specifications that the live lo:id is always twice as important and destructive as the dead load, irrespective of whether tlie member considered be a panel suspender or a bottom chord-bai in a flve-hunciii,cl foot span, is absurd, and involves far greatei errors than those tliat would be caused by any incorrectness in tlie assumed impact formula. The author acknowledges that he anticipates finding the values given by tli' formula somewhat high ; but it must be remembered tliat the s.iid formula is intended to cover in a general way, also, the ellVcts of small variations from correct- ness in shop-work, or to provide for what the noted bridge engineer, tlie late C. Slialer Smilii, used to term tlie factor of 8 i)F i'()N"riM(r.s. Using n uniform tension intensity of 18,000 lbs. for eyc-lmrs niiefoii! tlic full strciigili of the coiiipression- luenibcr is diiveloped. PUINCIPLE V. There are no bridge specificationg yet written, and there probably never will be any, v/hich will enable an engineer to make a complete design for an important bridge without using his judgment to settle many points which the speci- fications do not properly cover; or as Mr. Theodore Cooper puts it : "The most perfect system of rules to insure success must be interpreted upon the broad grounds of pro- fessional intelligence and common sense." At first tliougiil ouc! iniylit conclude Ihut this speaks badly for modern standard bridge speiitications, and to a certain limited e.xteiit he would be right ; for while it is (juite true that no railway- bridge specilieatioiis yet publisiied begin to cover the entire ground of ordinary' bridge-designing at all adequately, or nearly as thoroughly as they niigiit readily be made to do, neverlhele.ss it is also true that the science of bridge-designing is such a profound and intricate one that it is absolutely ini|>os.slble in any specification to cover the entire field and niake rules to govern the scientific i)roportioning of all parts of all structures. The author ha.'i done his best in Chapters XIV. -XIX. of this little treatise to lender the last statemeut incorrect, but with what success time alone can prove. PlfIN< IIM.K VI. In every detail of bridge-designing the principles of true economy must bo applied by every one who desires to be a successful bridge engine^ir. Tins subject is such an important one (hat to its cousidera tion the whole of the ue.vt chapter will be devoted. PitiNfiiM.K vn. In bridge-designing rigidity is quite as important an ele- ment as is mere strength. IG I)K rONTIBUS. Ill fact each of thcst' properties is depenihmt upon the other, because if a structure be amply proportioned in its main members for the assunud loads, but improperly sway- braced, the actual dynamic stresses will be greatly in excess of the live-load stresses provided for, and the metal will be overslruined in consequence ; while, ercentages to add to live loads for impact, when such |)er- cent ages are established entirely by guesswork. An elabo- 'i FIKST PRINCIPLES OF DESIGNING. i: ■i rate system of tests of actual intensities of working stresses for all main members of modern steel bridges under live loads, applied wiih varying velocities, is probably more ur- gently needeil at the present time by the engineering profes- sion tluin is any other series of experiments. lu the specitications of this treatise the effect of impact is provided for, how correctly only such experiments as those just referred to can demonstrate. As pointed out in Chapter 1., the determination of the various amounts of impact was made solely by adopting a few lixed intensities of working stress and varying the percentages of impact so as to make tile structures designed thereby agree as nearly as may bo with the best general practice. If the impact formulic adopted are ever proved to be incorrect, it will be a simple matter to correct them in a later edition. Principlk X. In making the general layout of any structure, due atten- tion should be given to the architectural effect, even if the result be to increase the cost somewhat. There is no feature of bridge-designing which has been ig- nored in America to such an extent as has this ; and it is only of late years that even a few American engineers have pnid any attention whatsoever to lesthetics in that branch of engi- neering. The subject is such an important one that to its con- sideration Chapter IV. will be specially devoted. Principle XI. For the sake of uniformity, and to conform to the un- written laws of fitness, it is often necessary in bridge- designing to employ metal which is not really needed for either strength or rigidity. The designer who recognizes this fact will usually produce structures of liner appearance than the designer who ignores it because of false notions of economy, 1 18 UK I'ONTn4l.S. PniNcirLE XII. Before starting a design, one should obtain complete data for same. If he fails to do so, lie will genciiilly liiive to niiikr !ill( iniioii after ulteration as the woik progresses; and, as one cliange usuall}' entails several others, it will result that, by the lime the work is tinished, enough labor will have been expended thereon to complete two such designs, for which proper data were furnished at the outset. Principle XIII. A skew- bridge is a structure the building of which should always be avoided when it is practicable. It is generally possible to stpiare the crossing either by swinging the centre line, or by lengthening the spans and squaring the piers or abutments, tionielinies, however, it is not practicable to do either, ir; which case the engineer must make tlie best of a bad business. The objections to a skew- bridge are these ; First, it is fuliy twice as troublesome to de- sign as a s(piare structure; second, the liability to error in both shop and lield is greatly increased by the skew ; and, third, the resulting bridge is never so rigid, nor is it so satisfactory iu a number of particulars, as a bridge without this objection- able feature. PniNciPii, XIV. The best modern practice in bridge engineering does not countenraicc the building of structures having more than a single system of cancellation, except in lateral systems where the resulting ambiguity of stress distribution is of minor importance. Some engineer may question the coirc Itiess of this asser- tion ; but if 1h' will glance through the author's paper on " Some Disputed Points in Uailway-jiiidge Designing" pub- lished in the February and Mirch, 180,. numlier of tin; Trdns- (irf/'nns of the Anierjcivn !^'n;iciy of ("ivil IOiigiui'«rs, he wi'l see .. i FIRST riUKCirLES OF UKSIGNINU. 19 implete III. 'It; on cliaiigo le lime XTidod iv data ; lliiit, as a wjiole, the engineering profcssio.i indorses tlie state, nieut. The only ordinary cases wliere multiple systems are employed nowadays are those of the lattice girder and the Whipple truss. The former is conceded by the leading bridge- designers to be unscientific, clumsy, often unsightly, and always uneconomical; and as for ihe latter, there is no longer any excuse for its use, because it lias been ousted from ihe jiosltion k used to hold by the Petit truss, which excels it in every particular, including appearance, economy of material, and mathematical correctness. Principi.e XV. The employment of a redundant member in a truss or girder is never allowable under any circumstances, unless it be in the mid-panel of r- span having an odd number of panels, in which case, for the sake of appearance, two stiff diagonals can be used. The reason for this is ^'erfectly clear when one considers that it takes extra metal to I'uiid the said redundant member, aiil that its use upsets the calculations of stresses, rendering them in fact iusolvable. A lengthy treatise was published a few years ago in India upon a method of finding stresses in redundant members, in which much good mental energy was wasted, for the entire book might have been written in these four words: "Never use such members." It is nut often that an American engineer is found guilty of tniploying unneces- sary pieces in his designs, but one cannot say the same of his European brethren. PniNCIPLK XVI. The use of a curved strut or tie in bridge-designing for the sake of appearance (or for any other reason) is an abomina- tion that cannot for an instant be tolerated by a good de- signer. It is hardly necessary to make such a forcible remark as this to American engineers, although in travelling about the United Stales one occasionally runs across a violatjoa of the self- 20 I)K roNTIlU'S. evident iinderlying principle involved in tliis statement ; but the publislied records of some of tlie greatest bridges designed by Englisli engineers sliow the use of pieces of trusses so curved that actually tliere is compression on one extreme fibre and tension on the other. Archilecturnl effect is undoubteily livery commendable feature in ])ridg(!-designing ; but its adop- tion is no excuse for the violation of the fundamental principle that every compression or tension member of a truss or open- webbed girder should be absolutely siraighl from end to end. Tt seems almost unnecessary to state that the appearance; of curvature can be obtained i)y employing short panels and making each chord-length straight between panel points. PHINCII'LE XVII. In all structural metal-work, excepting only the machi- nery for operating movable parts, no torsion on any mem- ber should be allowed if it can possibly be avoided ; otherwise, the greatest care must be taken to provide ample strength and rigidity for every portion of the struc- ture affected by such torsion. It is not often that this question arises; nevertheless it is sometimes forced upon the consideration of llie engineer. It came up lately in the author's practice in the case where an elevated-railroad exit-stairway, having at mid-height a landing and a 180-degree turn, had to be supported by a single column in order to comply with the demands of adjacent property owners. PUIXCIPI.K XVIII. The gravity axes of all the main members of trusses and lateral systems coming together at any apex of a truss or girder should intersect in a point whenever such an ar- rangement is practicable ; otherwise the greatest care must be employed to insure that all the induced stresses and bending moments caused by the eccentricity be properly provided for. This is an important rule that is more often honored in the breach than in the observancaj ; in fact, as far as the author FIKST PRINCIPLES OF DESIGNING. 21 ; but gned es so fibre •felly idop- ciple opeii- C'licl. vo of uiid knows, tbere are only n. very few bridges iu which the desired intersection iu a single point of the axes of all niembors assembling at each apex is accomplished ; and iu most struc- tures where eccentricity exists for want of such intersection, its prejudicial effects are upt duly recognized and provided for. Principle XIX. Truss members and portions of truss members should always be arranged in pairs symmetrically about the plane of the truss, except in the case of single members, the axes of which lie in said plane of truss. One occasionally sees a violation of this principle even in important bridges ; but experience with structures in which it was iguorcd has l>een such as to show most clearly that this (;anuot be done with impunity, for the torsion resulting from eccentrically connected adjustable members is patent even to the uninitiated. PUINCIPLK XX. In proportioning main members of bridges, symmetry of section about tw^o principal planes at right angles to each other is a desideratum to be attained whenever practicable. Of course in top-chord nnd inclined end-post sections, wldcli should l)e designed with a cover-plate, symmetry about both principal planes is not attainable. The objectionable features caused by want of it, however, are provided against by the next axiom. Principle XXI. In both tension and compression members the centre line of applied stress must invariably coincide with the axial right line passing through the centres of gravity of all cross-sections of the member taken at right angles thereto. Until very lately this iniiporlanl i)riuciple nas been simply ignored, the effect being that the allowed intensities of work- 22 T>K PONTTHUS. iiig strt'sscs are often excicded l)y from fifty to one huudred per cent because of the eccentricity thus involved. Pk1N( II'LE XXII. The principle of symmetry in designing must be carried even iuio the riveting ; and groups of rivets must be made to balance about central lines and central planes to as great an extent as is practicable. The viohition of tliis i)rinciple was exceedingly common not very long ago ; and even to day, when checking the shop drawings (f some of the lending bridge-manufacturing compa- nies, the author's assistants have to correct occasional depart- ures therefrom. PurNCII'LE XXIII. In proportioning members of bridges to meet stresses and combinations of stresses it is important to consider duly the quality, frequency, and probability of the action of said stresses or combinations of stresses. As a rule, standard specilications lake care fairly well of this subject; nevertheless there will often occur iu one's practice casts which they do iu)t cover. The quality of stress should be considered in deternuning the sectional area of the member, because the greater the impact, other things remain- ing the same, the smaller should l;e the intensity of working stress. Tiie frecjuency of application of stress should lie consiiii'ied, because, if a certain stress or combination of stresses be of frequent occurrence, a small intensity of work- ing stress should be adopted, while for very iiifretpu-nt occur- rences the intensity can he taken considerably higher. Finally, the probability of the application of a certain load or loads should be considered ; because for inevitable loads or coml)iiiations of loads the mital sliould be strained fairly low, while for highly improbable loads or combinations of loads it is legitinuite to strain much higher. Just hcie the author wishes to slate most clearly and emphatically, that to indorse the points asserted under this heading one need not FIUST r-UKVClPLKH OF DESKiNlNO. 23 I)(! a bolievnr in tlio (ioctiines of Wolilcr niid Wcyrauch, and ill tiio tiu'ory of the fiitigiui of iii(;t;il.s, because one's conunon sense will lead him to jiroporlion seclion.s of bridge jneinl)er.s in uceordance with tiie foregoini,' views. In the spicitie.'ilions given in C'lnipteis XIV. and XVI. tlic inipiu;! fi)rn\ulie and the increased int( nsitie.s for eoinbiualioiis of stresses involving lliose due to wind loads take care of this feature of design for all structures excepting high railroad trestles, in wliich latter tlie designer's professional judguieut cannot well be eliminated. PllINCII'I.K XXIV. In all main members having an excess of section above that called for by the greatest combination of stresses, the entire detailing should be proportioned to correspond with the utmost working capacity of the member, and not merely for the greatest total stres.^ to which it may be subjected. In this connection, though, the reduced capac- ity of single angles connected by one leg only must not be forgotten. It is almost needless to state that most engineers, especially those connected with contracting companies, will disagree willi theautiioron the correctness of tins statement ; neverthe- less he has yet to see the lirsl case where adherence to the principle would involve improper, clumsy, or inappropriate construction. If it be right, for any reason, to use an e.xlra amount of metal in t lie section of a member, why is it not also right to design lh;il mend)er throughout so that, if tested to destruction, it would fail as a whole and not in a detail? It .seems to the autlior that the considerations which require extra section would demand either extra strength or extra rigidity, or both, in the details as well as in tlie section itself. PiUNrii'i.K XXV. In every bridge and trestle adequate provision must be made for the contraction and expansion of the metal. Neglecting to comply with this principle has often beeu the cause of failure and disaster. 24 1)K I'ONTIIU'S. Pkinciim.k XXVI. No matter how great its weight may be, every ordinary fixed span should be anchored effectively to itH aupports at each bearing on same. At one cud it sbould be ancliorctk ininiuvat)jy, and (it ibe other so as to providt; for loiii^itiidiniil expansion and contrac- tion. Such anciiorage prevents I lie dislodt;inj^ of the striic- tiire by \vind-p^e^^8ure or by an accidental blow from a mov- ing object. PuiNCIPhK XXVII. The bridge-designer should never forget that it is essen- tial throughout every design to provide adequate clearance for packing, and to leave ample room for assembling mem- bers in confined spaces. Tbere is no more fruitful source of profanity for bridge- ereclors than the neglect of this principle; and as nearly every (le.>^igner has to spend a year or two in learning toallow enough clearance, it follows that bridge-erec-tors shonld be given the benelil of " extenuating circumstances" when brought to judgment for their notorious addiction to the use of strong language. PiUNcrvLE XXVIII. Although for various reasons engineers are agreed that field-riveting should be reduced to a minimum, such an opinion should not be allowed to militate against the em- ploymsnt of rigid lateral systems. All designs should be arranged so that the field-rivets can be driven readily. One of the main reasons for the unsatisfactory condition of most of the elevated railroads of this country is that their de- signers endeavored in every possible way to avoid field-rivet- ing, so as to keep down the cost of erection, and in so doing fa'.led to develop the requisite amount of rigidity in the struc- tures. FIUST I'KINM'II'LES OF DKSKJN'ING. Ji5 PniNcu'LK XXIX. Rivets should not be used in direct tension. Ill the (lays of iron rivt'ts this was an important roqulrc- niont, for llic louson Ihut llic siiaiiks were often hooverstrainetl in I ooiing that the heads won d lly oil ; but tliis does not oc- cur witli steel rivets. Nevertlieless it is advisflblt; loadliere to the rule, except for very unimportant members where there is a great excess in the number of rivets above the theoretical requirements "' KINCIPLE XXX. For members of any importance two rivets do not m«ke an adequate connection. For such details us lattice burs, of course two rivets or even one rivet at each cud will sutUce ; but where a direct calcula- ble stress comes on the piece, and only two rivets at each end arc used, it will be found that they will work loose, while, if three are used, they will not, unless they be overstrained by the caic'jl«t(id stress on the piece. Principi.k XXXT. Designs must invariably be made so that all metal-work after erection shall be accessible to the paint-brush, except, of course, those surfaces which are in close contact either with each other or with the masonry. This clause very properly cuts out the use of closed col- umns, which are a fruitful source of condemnation of old bridges. PUINCIPI,E XXXII. In multiple-track structures, if any bracing-frames be used between panel points to connect the longitudinal gir- ders of adjoining tracks, they must be designed without diagonals, in order to prevent the transference of any ap- preciable portion ef the live load from one pair of girders to any other pair of same. 2G I)E POKTIHUS. Such a traiisfereuce would be doubly injurious ; because it would throw on some of the girders more live load than they were proportioned to carr}', and at the same time it would probably overstrain the diagonals and their connections, and would certainly tend to distort laterally the flange angles of the longitudinal girders. Principle XXXIII. In bridges, trestles, and elevated railroads, the thrust from braked trains and the traction should be carried from the stringers or longitudinal girders to the posts or col- umns writhout producing any horizontal bending moment on the cross-girders. This is a late requirement of the author's, that has been em- ployed in ids designs for a few years past. Its correclULSs was established in his before mentioned paper on Elevated Kail- roads. PlUNCIPLE XXXIV. In trestles and elevated railroads the columns should be carried up to the tops of the cross-girders or longitudinal girders and be effectively riveted thereto. The correctness of this proposition also was established in the said paper on Elevated Railroads. PiuNCirLB: XXXV. Every column that acts as a beam also should have solid webs at right angles to each other, as no .eliance can be placed on lacing to carry a transverse load down the column. The truth of this propo.sitlon is evident when one retiects that a single loose rivet or a single bent lacing-bar in the whole line of lacing will prevent the latter from carrying as a web a transverse load. Loose rivets and bent lacing-bars iire, unfor- tunately, not uncommon in structural metal-work. FIllST PRINCrPLKS OF DESIGNING. 2: le it licy Ul(i llU(i of PUINOIPLK XXXVI. In trestles and elevated railroads every column should be anchored so firmly to its pedestal that failure by over- turning or rupture could not occur in the neighborhood of the foot if the bent were tested to destruction. As long ago as 1891 the author designed pedestals which in- volved truly fixed ends for colunui feet; l)ut it is only wilhiii tlie last three years tliat such a detail lu.s begun to come into general use. The ordinary connection of columns lo pedestals by an unchor-l)()lt iit each of the four corners of the bedplate is extremely weak and ineirective. PUINCIPLK XXXVII. All pedestals for trestles, viaducts, and elevated railroads should be raised to such an elevation as to prevent the accumulation of dirt and moisture about the column feet, and all boxed spaces in the latter should be filled with extra-rich Portland-cement concrete. The neglect of these precautious causes the rapid deteriora- tion of tlic metal at bases of columns, and thus shortens the life of the structure. PUINCII'LE XXXVIII. In designing short members of open webbed, riveted work, it is better to increase the sectional area of the piece from ten to twenty five per cent than to try to develop the theoretical strength by using supplementary angles at the ends to connect to the plates. This principle is based upon the resvdts of some late tests of th(! author's on tlie strengtli of single angles and pairs of angles (•oiint'clod l)y one leg only, by which lie f "Uud tliat 6" X ''a" angl(!s thus connected dev('loi)ed idnety p< cent of the ulti- mate strenglli of a flat l)arof eipuil net section, and that3"X3" angles developed seventy-five per cent of same. 28 DE PONTIBUS. Pkinciple XXXIX. Star-struts formed of two angles with occasional short pieces of angle or plate for staying same do not make satis- factory members. Better results are obtained by placing the angles in the form of a T. Tlie truth of tliis slatemori .".< '-^ished by auolher series of e.vperlments of the aulhor .> niuat at the same time as were the liist-meutioned tests. The specimen cuhimiis did uot de- velop on the average more than seventy-five per cent of the ri'sistance they should have developed according to tlie usual straight-Hue formula lor metal of the same teu&iile strength. PUINCIPLE XL. In making estimates of weights of metal the computer should always be liberal in allowing for the weight of details. It is the author's experience that, in nenrlv every case, the weight of the tinished structure exceeds s;:i;i»tO tiie estimated weight, and mainly on account of the o i lOre metal for details than was figured upon. Ofcou;'. i >r sets out de- liberately to "skill " a bridge so as to save i i. tlitj metal he can, the actual weights of details may be made to underrun the estimate; but such a practice is most reprehensible. I Tl: the ( «Fo poin ferr« strui by ^ ticu che< whi agrj best B pult stru Principle XLI. In general details must always be proportioned to resist every direct and indirect stress tb,"'* may ever come upon them under any possible condition, -v^fchout subjecting any portion of their material to a stresu -^v ^ter than the legiti- mate corresponding working stress. This principle, w'li' u has been given before in several of the authov ^, |;n;vior. ■ works on bridges, involves the whole theory of oridge ue'i;iii..ig. FlliST rUINCIPLES OF DESIGNING. '^9 lort itia- ting Ties ere de- tbe sual h. Principle XLII. There is but one correct method of checking thoroughly the entire detailing of a finished design for a structure, viz.: " Follow each stress given on the stress diagram from its point of application on one main member until it is trans- ferred completely to either other main members or to the sub- structure, and see that each plate, pin, rivet, or other detail by which it travels has sufficient strength in every par- ticular to resist properly the stress that it thus carries; check also the sizes of such parts as stay-plates and lacing, which are not affected by the stresses given on the di- agram, e d see that said sizes are in conformity with the best modern practice." But to do all this as it sliould be done necessitates the coni- puter's being, in tiie highest sense of the term, an "expert on slriiclural metal- work." CHAPTER III. TRUE ECONOMY IN DESIGN. TnEATiSE after trcntiso has been written npon the subject of economy iu superstructure design, but unfortunately the result is simply a waste of good mental energy ; for the writers thereof invariably attack the problem by moans of complicated mathematical investigations, not recognizing the fact that the questions they endeavor to solve are altogether too intricate to be undertaken by mathematics. The object of each investigation ai)|iears to have been to establish an equation for the economic depth of truss, or that depth which corresponds to the minimum amount of metal required for said truss; and, to start the investigation, it seems to have been customary to make certain assumptions which are not even approximately correct. For instance, the principal as- sumption of several treatises in French and English is that the .sectional area and the weight of each member of a truss are directly proportional to its greatest stress; or, in other words, that in proportioning all members of trusses a constant intensity of working stress is to be used, while in reality for modern steel bridges the intensities vary from, .say, 6000 pounds up to 15,000 pounds, or, when impact is provided for, up to 18,000 pounds, and when both impact and wind stresses are included, up to nearly 34,000 pounds. Again, no distinc- tion is made between tension and compression members, and no account is taken of the greatly varying amounts of their percentages of weights of details. There is, however, oik; mathematical investigation concern- ing economic truss depths which, in the author's opinion, is approximately correct, and which is based on assumptions 30 THUE Kt'OJSOMY IX DESIGN. U tliat iiro very nearly true • but it liolds good only for parallel chords. It is this : Let A = weight of the chords, Ji — weight of ihc web, (J ~ weight of the truss, and I) = depth of truss. Then yl -^ /?. But the weight of the chords varies inversely as the depth, (I or A = jr, and the weight of the web varies ilireclly as the depth, or B= bl), where a and b are constants ; and therefore a C=-^ + bD. If O is to be made a minimum, we shall have, by dilleren- tiation, (f£ _ _ ^i dl) " D , + i = 0. or 4 Ji - ~ + ^ = 0. or A = B. As the second differential coetlicient, after substitution according to tiie usual method for maxima and minima, conies out ]>osilive, the; rcMilt obtained corresponds to a inininuiui. From this it is evident that, for trusses wiiii parallel chords, the greatest economy of material will jjrevail wlie.i the weight of the chords is equal to the weight of the web. The author has vcrilied this lonclusion by checkiuL^ tin; weights of chords and webs in a number of finished designs, limliug it to be absolutely reliable. However, it is not of much practical value, because the economic depths of trusses with piiralle' chords are pretty well known ; and, ag!un, when spans are in excess of 175 or 200 feet, the chords of through-bridges are seldom nuide parallel. Moreover, the best depth to u^e is not 32 I)K rONTIBUS. often the one which gives the least weight of metal in the trusses. Tlie iiuthor finds by experience tluU, for trusses with po- lygonal to[) chords, the erononiio depths, as fur as weiglit of metal is concerned, are generally much greater than certain iujportant condit'ons will permit to be used. For instance, especially in single-track bridges, after a certain truss depth is exceeded, tiie overturning effect of the wind-pressure is so great as to reduce the dead-load tension on the windward bot- tom chord to such an extent that the compression from the wind load carried by the lower lateral system causes reversion of stress, and such reversiou eye-bars are not adapted to with- stand. A very deep tru.ss recpiires an expensive traveller, and to decrease the theoretically economic dei)th increases the weight but slightly ; hence it is really economical to reduce the deptli of both truss and trav(;ller. Again, the total cost of a structure does not vary directly as the total weight of metal, for the reason that an increase in the sectional area of a piece adds nothing to the cost of its manufacture, and but little to the cost of erection ; so it is only for raw material and freight that the expense is really increased. Hence it is generally best to use truss depths con- siderably less than those which would re([uire the minimum amoimt of metal. For parallel chords, tlie theoretically eco- nomic truss depths vary from one fiflli of the span for spans of 100 feet to ai)0ut one sixth of tlie span for spans of '300 feet; but for modern railway through-bridges tlie lea.st allowable truss depth is about 28 teet, unless suspended tloor-beams be used, a detail whi(;h very properly has gone out of fashion. In two five-hundred -foot spans of a combined railway and highway bridge the aullior employed a truss depth of seventy- two feet; but this was determined by the reversal of stress in bottom chords through wind-pressure. A greater depth, if permissible, would have caused a saving in total weight of metal. In a design of the author's for a five-hundred-and-sixty- foot span a truss depth of ninety feet was adopted, but in this case the live load was very great, varying from tea TRUE ECONOMY IN DESIGN. 33 thousand pounds per lineal foot for short spjins to eight thousand pounds per lineal foot for long spans; and the bridge is twenty per cent wider than in the case of the two five- hundied-foot spans just mentioned. The greater the live loud and the wider the bridge, the greater can the truss depth be made advantageously. The little mathimalical investigation given in this chapter can be applied with advantage to plate-girder bridges and to the floor systems of truss bridges. If for ordinary cases, in designing plate girders, one will adopt s\jcli a depth as will make the total weight of the web with its splice-plates and stilfening angles about equal to the weight of the flanges, he will obtain an econoniicaily designed girder, and a deep and stiff one. For long spans, iiowever, this arrangement would make the girders so deep as to become clumsy and expensive to handle; consequently when a span exceeds, say, forty feet, the amount of metal in the flanges should be a little greater than that in the web; and the more the span exceeds forty feet the greater should be the relative amount of metal in the flanges. Concerning economic panel lengths, it is safe to make the following statement: "Within the limit set by good judgment and one's inherent sense of titness, the longer the panel the greater the economy of materia! in the superstructure." Of course, when one goes to such an extent as to use a thirty-fool panel in an ordinary single-track bridge he exceeds the limits referred to, because the lateral diagonals become too long, and their inclination to the chords becomes too flat for rigidity. Again, an extremely long panel would often cause the truss diagonals to have an unsightly appearance, because of their snuill inclination to the horizontal. There Is another mathematical investigation which is of practical value. It relates to the economic lengths of spans, and was first demonstrated, in print, by the author some six years ago in Indian Engineering, although the principle was announced three years before then in th'; first edition of his General Specifications for Highway Bridges of Iron and Steel. Strange to say, many engineers failed to see that there 34 DE I'ONTIHUS. is any diilerciicc between this principle and an old practice of forty years' standing. The principle is that " for any cross ing the greatest economy will be attained wheu the cost per lineal foot of the substructure is ecjual to the cost per lineal foot of the trusses and lateral systems." The old practice was to inake for economy the cost of a pier equal to the cost of the span that it supports, or, more proi)erly, equal to cue half of the cost of the two spans that it helps to support. Is not the difference between these two methods perfectly plain ? In one the cost of the pier is made etpiid to the cost of the trusses and laterals, and in the other it is made etpial to the cost of the trusses, laterals, and the floor system. When one considers that the cost of the floor system is sometimes almost as great as one half of the total cost of the super's* ructuic, lie will recognize how faidty the old nietliod was. The following is the demonstration of the principle, sim- plified to the greatest practicable extent. Let us assume a crossing of indefinite length, for which the depth of bed-rock is constant, and let S = cost per lineal foot of the substructure, T = cost per lineal foot of the trusses and laterals F = cost per lineal foot of the floor system, B = cost per lineal foot of the entire bridge, L = length of span; and then B = S+ 1'+ F. Now if we assume that slight changes in length of span will not affect materially the sizes of the piers, the cost per foot of the substructure will vary inversely as the span length, or -S = Again, the cost per foot of the trusses and laterals, for slight chaniics in length of span, may l)e assumed to vary nearly directly as the span length; hence we may write the equatiou T= iL. TKUE ECONOMY IN DESIGN. 36 le of loss per liu-al I was the If of The cost per foot of the tloor system is practically inde- pendent of the span length, being ii function of the panel length, which does not change materially with the span. We now have the eciualiou B = j^ + tL + F, In which B is to be made a minimum. Dlllereutiating, we have (as F is a constant) A further differentiation shows that the result corresponds to a minimum. Ill reality tho truss weight per foot increases more rapidly than the span length. If r is the ratio of span lengths, the truss weights, for small changes in span lengths, will vary almost exactly according to the ratio ?•' = \(r + ?•'). On the other hand, the weight per foot for the lateral system does not increase as rapidly as the span, unless the per] .id'cular distance between central planes of trus.ses also inci eases. Unfortunately, though, the gain in truss weight over that given by the assumed theory of variation is generally greater than the corresponding loss for the weight of lateral system, consecpienily the combined weights per foot of trusses and laterals generally increase a tritie faster than the span lengtii. This is i)artially olfset by the fact that the pound price of metal erected and painted will reduce a trifle as the weight per foot increases. Again, there is sometimes a small error in the assumption that the cost of the piers varies inversely as the span length because the size of each pier may have to be increased a little to accommodate the heavier spans. If the perpendicular dis lance between central planes of trusses is increased because o the greater span length, the cost of each pier will be increase( because of its greater length; but this will occur only occa sionally. 36 HE ro XT I BUS. Ignoring tlio lalttir coiillngiMicy, tlie two errors indicated, notwilhslnndiiig the fuel timt their effects :ne juhlilive, are so small us not to iiirect luateri.illy tlu; corri'(;tiiess of the results of this invesiigation concerning economic span lenglhs. Thisdeiuonstratioii proves tliiit, in any lavoiil of spans, with the conditions assumed, the greatest economy will hv. attained when the cost of the substructure per lineal foot of biidgc; is equal to the cost per lineal foot of the trusses and lateral systems Of course no such condition as a bridge of in- definite extent ever exists, nor is the bed-rock often level over the whole crossing ; nevertheless the principle c;in he applied to each |)ier and the si)ans that il helps to s\ipport l)y making the cost of each pier equal to one half of the total cost of the trusses and Ipterals of both spans. Since working out this demonstration some ten years ago, the author has made a practice of checking the correctness of the principle thereby established, by comparing the cost of substructure and nuper structure in a numoer of bridges which he has designed and built, with the result that he finds it to be exact. The principle will apply also to trestles and elevated roads, for in the latter, if we uuike the cost of the stringers or longitudinal girders of one span equal to the cost of the bent at one end of same, including its pedestals, we shall obtain the most economic layout. In an ordinary railroad trestle, con- sisting of alternate spans and towers, il will be necessary for greatest economy t(> have the cost of all the girders in two ^-pans (one span being over the tow^er) plus tiie cost of the longitudinal bracing of one tower, equal to the cost of the two bents of said tower, including their pedestals. On page 235 of the first edition of Prof. J. R. Johnson's " Tireory and Practice of Modern Fr:viicd S:riu.'tures,'' Mr Jkyun u.oes this method of the auth' ; in a slightly different form for determiinng the most ecoiu)mi(; ninuber of spans to adopt at any crossing, establishing the equation, y = A-\-B-\-(x l)C+l^^p, TUIK K(X)N().MY IN Id'.SKJK, a? iu which y (the total cost uf bridge) is a initiiiuum when G, where A = cost of two end ubutnu'iits in dollars ; B = cost of tlie floor luid tlmt part of tiie metal weight whicli reiimiiis constant, in dollars; C — cost of out! pitT in dollars, assumed as constant ; I — lenj^tli of bridge in feet ; X = number of spans ; p = price of nietiil per potiiid, in dollara; y = total cost of Itridge in dollara ; a = weiglit per foot of a span b feet in length, riuis fur all right ; but then he makes an assumption which will not be correct except for one live load, for one set of specific!'.' inns and for sincle-track railway bridges, viz., that for piu-conuectt'l spans aiK 1 a On iiccoint of lliis assum|)tioii his subsequent table of economic span lengths is not in any sense general, but is true only for single-track bridges designed for one standard live load and according to one standard set of specifications ; while his equations ii-dd good for bi 'ges of any kind .and loading, including higliway as well as railway structures. As a check on the correctness of Mr. Bryan's assumption that =r ^ for siugle-track bridges, the author lias looked up a k) some of his designs and has found the following : For a 375 ft. through-span, Class X, - = ^ ^- ; for a 362-ft. double-track through-span, Class Z, - = q ,7 ; »iiid for asimilur 490-ft. span, -— -— . For a *J80-ft. double-track deck-span, a y.o 1^8 DK I'ONTHU'M. • liissY, = —.1111(1 f(ir II Hiinilar 200.ft. ractical cfmsiderations. Unl, as I have already ol)served, tliis admiralde and prolific proKi'ess lias not cari'ied Willi it a correspomlinx pro>;ress in graci' and lieanty "f design. In fact. ''lese (pialities seem to appear in an invcrsi' [)roportion to the development of the structural scheme towards the jiractical idea of strength, stability, and eart of the system of creation. Is It reserved for man in his secondary creation to make things luilovely iti proportion to thility of the ideals wliicli have been estahlisiied by conditiouH long since outgrown in the i>rogi-ess of science. The attempts of the Ei\glish bridge-builders in iron in the early i)art of the century to meet thest* old ideas resulted in constructions which, though they may satisfy the eye of the artist, and combine more or less graeefullv with the laud scape, are uneconomical and unscientific. The principles of structure involved are incorrect, and unnecessary expense was incurred in forcing into the dtisign features conventionally acceptal)le, but which liad notli- ing to do witli the struv'tnre. und which in fact were a hindrance to it, concealing rather than illustrating it. The architect will not find it diUflcultto agree with Ids broiher tlie en- gineer, that a mask of ornamental cast iron, covering lliwexsential feat- nres ''f tlie structure in order to for(,'e upon it an efi. 't of grace, is illogical in the extreme. Indeeil, a great modern master of architecture has laid down the a.xi<>m: " A foi'in which admits of no explanation, or which is mere caprice, cannot be beautifid; and in architecture, cer- tainly, every form which is not inspired by the structure ought tliere- fore to bi^ rejeeteil." Tlie <'onscieniious modern architect aims to shape his design a rding to this reasotialile limitalioo, and be has been thereby enabled to proiluce occasioiuil etTects of beauty without impos- ing on his composition a single idea which is not suggested eitlier by the structure or by the use of th<- buildiii'..'. Kvimi a factory, ii gasometer, a railway shed, an elevator, need not challenge the architect in vain to piddntre effects of fitness not entirely iiu^onsistent with the re(|uirenients of art. Indeed, the engineer himself, with a.xioms or nia.xims nt art, has, in the evolution of the roof-truss, the locomotive, and many Indus trial niachines, succeeded in satisfying ideals of beauty in the very proc- ess of making them powerful, com|)ai'l, and economical of luaterial and space. The mudern steel-armored war-ship has already, in this early stage of its rapid development, 8U> ituted for the ideas of tnarifime beauty, speed, and strength which prevailed in the time of Xelson and the other great histu)ical admirals, ami which were celebrated in :he songs nf iijbdin and Ciinipbell, an entirely dilTereut ideal, baldly less imposing, Ihonghasyet without poetic recognition. Hut the evolntiun of the steel trussed liridge has aw' y(>t satisfied neither old ideals uf beauty, nor has it made new ideals. Its essential lines are drawn in apparent disregard or (*ontempt fin' grace of outline or elegance of de- tail. The difTlenlty seems to l)e inherent in the present approved struc- tural system of designing horizontal, straight, open-trussed girders ur cantilevers, resting on rigid vertical piers of masonry or iron, without regard to any other considerations excepting those of statics. The eye requires to be satisfied as well as the trained intelligence, and demandH not only grace of proportion, but a certain decorative eniphaKls expres- sive of 'special fuuctious. The primitive post and lintel structure of ;icsTilK'ri(;s is i)i<:st(m. 43 stone was as hopeless, apparently, as iia modern derivative, the steel- trussed biidge, until the (irecks, with iiner?-iiig instinct of art, C()riverte.s have been .sucombiu-'d in tension and compression as to produce a r>irticture u ipable of sustaining without change of form not oidy its own weight betvM-eii bearing [(oinis l.ir uj >irt, but that of moving trains, .•mil of bcni ing wii liout detriment ubrations and wind- pressures, and the expaiiiiou ami contraction of iir< material by changes of temperature. These compound lintels or tins.sen are in tli'-niselves triumphs of mind over niattei'. At this moment they e.\pre.ss a sta^'e of evolution which has l)een in process for a century, and which dor iiless w dl continue to develop in directions impossible to ant icipat«>. Tlioy are structures not deilicated to the immortal goils, like ' post and lintel in the Greek tem- ples, the decorative clniracter of wlucli was largely iiispirtHl by religious emotions, but devised tit meet secular and practical conditions of an ex- ceedingly unpoetic and luiimaginative cli.i cter. The mind of the architect appreciates the flue economy of these sensitive and com- plicated orgaidisms, but it also reogi that they are still in aclive process of development ; that they are on trial, and will not ivach linal results until tlirtj sitiill /nice nssittited thitsr CDiidition.i nf ijrace. and lii'iiuhj irhu'h are I'^m-ntial to comftli'tion. It is evident enoirgh that all the featiu'es of perfection in aiunnils have been very gradually evolved, by survival of the tittest and by adaptation to use, from the awkward and luonstrous shapes of the antediluvian period; that geological ero- sion and drift liave clothed the naked rocks with beauty ; and that the whole vegetablt^ creation has been improved by art. Nature herself is not contented with inelastic dogmas. In like manner, the locomotive, the steam-engine, the modern war-slop, have all l)econiu objects of awful 44 BE POKTIBITS. beauty, not becatise of tlie imposition of unnecessary features, but hecaus(^ of the natural and reasonable growth of their essential structiif"'. If, tlicreforc, llio ugly character of the present ateel-trussed liridge is in itself a proof of the iitimaturity of the seieiiee which has produced it, tlie remedy, of course, must resitle in the perfecting of the science, and this process of perfecting will be quickened, if beauty is rec&gnlze(i !■■ engineering as it is in architecture, as an aim and not as an acwident of growth. The architect, guides and hastens this progress towards the perfect type by fundamentally composing liis strnctmewith a view to an agreeal)le proportion of its parts ; in detail he studies to emphasize the special and important points of his structure by a decorative treat- ment which shall indicate conventionally the character of the work ac- complished at these points. It is true, perhaps, that the structural forms of materials with wliich the engineers have to work, especially in bridge-building, are hardly so elastic and manageable as those at the cernuiand of the architect even in liis simplest and most severely prac- tical problems; but it is none the less true that the training of the engineer leads him too often to an absolute disregard, if not contempt, for those reflnemeiits of proporticn and outline, and for all those delicate adaptations and adjustments of detail, which, though perhaps separately slight, and apparently of small importance, in combination tend to give distinction and a character of fitness and grace to works otherwise, from the point of view of art, rudely immature, basely mechanical, un- necessarily and insolently ugly. Mr. IleiM-y .lames says that the French talk of those who see en beau and those who see ini laiil. The performance of the modern steel-bridge designers would certaiidy seem to place them in the latter category. It is no less certain tliat this result comes not from temperament, which is natural, but from training, which is artificial. The severe and absolute conditions in which the bridge-builders work do not prevent them either from great differences in manner and method of design, or frotn frequent and utmecessary extravagances of expenditure ; but these ex- travagances are rarely, if ev('r, lavished in the services of beautj' ; be- caiisi' the cold and rarefied atmosphere of science and mechanicnl utility, in which they are accustomed to labor, has gradually frozen out (be tin T natural instinct which works for art and el«>gance in design. Ueauty of projiortion has often been proved by mathematics ; but mathematics, when it has lieen allowed to be the only elfirient in the development if a jiroblem of construction, has never accomplished beautiful results. Such results do not come by accident in any work of design, but liy the liberal and generous observance of natural laws. The education, therefore, which from the beginning does not give some recognitiiiii to grace, proportion, el(>gance, as essential parts of construc- tion, must be misleading and one-sided, and cannot lead to perfection. The recognition (f these ipialitles, I am entirely persuaded, does not necessarily imply any sacrifice of jiractical accuracy In design or of ^ESTHETICS IN DESIGN. 45 mechanical precision in worle folly to apoiid nny men! on its coiistniclioii than is culled for by the eiigiiiecriiig rc'iuireiiieiUs of the conditioiis, due ullowaiicc being made, of (our.so, for ii pu,s>ibie iieo[>ling of the forest or desert in the not veiy distant future. Tlie style of ornamentation for a bridge should always bo iu keeping with its general character; thus, in case of a light highway bridge, ornamental portals with filigree metal-work are appropriate, while iu large, massive railway bridges the ornamentation should be of a coarser and bolder character, commensurate with the size and use of the structure. The author is a Arm believer in the principle that true economy, engineering excellence of construction, and the best architectural effect will almost invariably be found to a( con> pany each other, and be inseparable in the designing of any biidgc. Moreover, any l)ri(lg(! built with due consideration for, tirst, ellicicnc}', second, appearance, and, third, economy, will be satisfactory and gratifying to not only the trained expert, but also to the general engineer and railroad man, and even to the public ; because when an observer notes that in such a stniclure all tlie engineering reciuirements are properly piovided for, that there is no evident waste of mate- rial, and that all due advantage has been taken of the coudi lions to render the bridge sightly and in harmony witli its surroundings, his eye will of necessity be pleased, and his inherent sense of fitness will cause hira to regard the structure with a feeling of pleasure. In suggesting that " if a steel trussed bridge, economically and wisely consirucn-d according to our present light, olTends our ideals of grace and beauty, the fault perhaps is not In the structure, but in the rigidity and immobility of the ideals which have been established by conditions long since out- grown in the progress of science," Air Van Brunt has prob- ably indicated tlie lines of convergence of engineering practice and architecturd ideals , for while, as before stated, much can be done with most bridge designs to improve them with- out increasing their cost or affecting their efhcieuc}', on the other hand it is often inipos ible for an engineer to modify a bridge design so as to meet fully the critical objections of a 4S DK PONTIHUS, good arcliitert without introducing fciiturcs boUi faulty ntid expensive. It is iiiglily probable liial if llie engineer will modify liis designs as uiucli as is legitimate to meet tlie fpsthetic re(iuirenieiits of the architect, the latter will gradually modify the rigitlity of his ideals, ami will see lines of grace, beauty, and fitness in the polygonal outlines of trussed bridges. Mr Van Brunt himself has already sliowu this to be true by giving his unqualilicd approval to the architectural elfei^l ot' the truss outlines in the draw-span of the author's bridge over the Missouri Hiver at Omaha, although these outlines were de- termined primarily for iitility and secondarily for appearance, and notwithstanding the fact that there is no attempt at even approximate curvature of chords in the entire span. To ret agnize and acknowledge the deficiencies of modern bridge designs from the artistic point of view is one thing, l)ul to show how they are to be remedied is another ; because, while it is easy to say that a certain stnu-lure does not (.'onu! up to one's ideal of grace and beauty, it is very difticull to show exactly where the defects are, and what should or can be done to remove them. Notwithstanding this, the author will now endeavor to establish a few fundamental rules which, if followed, ought to correct tlie most glaring sources of ugliness in bridge designs ; then, by entering more into detail, he will try to show Inw the structures may be decorated appropriately and inexpensively, The architectural treatment of bridge-designing may l»e divided into these four parts ; Ist. The layout of .spans, piers, and approaches. 3d. The oui lining of each span. 3d. The decoration of each spaa. 4th. The ornamentation of the entire structure by tho adop tion of elaborately artistic approaclies. In respect to the layout of spans, piers, and approaches for any bridge, there is one governing principle which should always be complied with, viz., that the entire structure, when ever possible, should be made perfectly symmetrical al)OMt a midc^le plane. ^STUETICrf IN DESIGN. 49 There is no featuie of a bridge so pleasing to tlie eyes of all observers, cultivated and ignorant alijje, as perfect symmetry in tlie layout of spans ; consequently it should be attained whenever practicable, even if some extra expense be involved thereby. Unfortunately the conditions are not always favorable to perfect symmetry of design, for tlie bed-rock will often dip rapidly, and thus necessitate the use of spans of different lengths, and the channel of the river often refuses to keep at midstream, persisting in hugi^iiig one sliore. In sucli cases it becomes -lecessary to do the best one can with the \ui favor- able conditions, and to make the structure sightly, if not sym- metrical. If there be a draw-span on one side of the river, it is best generally to niake all of the fixed spans alike. Should each span — because of the gradual shelving off of tlie bed- rock, and for the sake of economy — lie made longer as tlie bed- rock de('[)en8, the result will be unsightly, even if the incre- ment of span lengtli be regular, for the reason that to an observer there is no apparent motive for thus diversifying the spans. Any divergence from symmetry and regularity for which there is a self-evident reason produces no unfavorable impres- sion upon the beholder, although it maybe sufticient cause for failure to excite his admiration for the structure. If one can see at a glance the raison d'etre of all the principal parts and peculiar features of a bridge, his sense of fitness Avill be satisfied, and his general impression will be favorable , but the nearer the approach to perfect symmetry and the more artistic the outlines, the more thorough will be his apprecia- tion of the general effect of the structure. In making a study of the teslhetics of a bridge design, after determining what spans are applicable, it is well to make one or more layouts on a large .scale on the brown paper used in engineers' offices for pencil-drawings, indicating tlie circum- scribing lines of all main meinl)er8 to scale, anil tinting or filling between said linos with pencil-shading ; then tack the paper on a wall, and stand off at various distances to judge the effect. By doing th's one can form a very correct o])inion 50 PK roNTinrs. concerning tlie conipnmtive merits of several layouts, and can ascertain where and how any particular layout can be im- proved. A consiiltation with soveral members of one's office force upon the architectural features of the various designs will often result in an improved effect , for nothing else will bring out both the favorable and unfavorable characteristics of a plan like discussion. In the o>il lining of each span a great deal can be accom- plished towards beautifying a structure, aiul lliere is no l>etter way to study the geiuM-al effect of any proposed outline than the one just indicated, viz., laying out various trusses to scale, tacking the paper to a wall, and criticising them. It will surprise any one who tries this method to sec; how quickly he can detect the slightest variation from correctness in out- line, and what a difference in efTeel even a small change in u truss depth will produce. It wh- in this w.iy that the trusses of the Omaha druw-spiin were proportioned , aid it is doubt- ful if any improvement could be effected in their outlines when all factors involved in tlie (|Ueslion are duly considered. In this problem there were but three points to determine, viz., the depths of truss at the two hips and the depth at the tower, for the number of panels was settled by economic consideni- tions, and the straightness and section of the lop chords were necessitated by certain (piestions of efliciency. The depth at the outer hips was tinst determined by tiie requirements for clearance, rigidity, and appearance, then the depths at the intermediate hips and tower were settled by trial and discus- sion from the artistic point of view, due attention being paid to the engineering (jueslious involved by the various iDcliua- tions of top chords and incluied inner posts. In determining the out.ines of a span these few elementary principles are to be borne in mind; 1st. There is nothing so ugly in a bridge as parallel chords unless it be a skew. However, for spans between one hundred and twenty-tive feet and two hundred feet it is often best to use them, although in certain cases where the loads are great it is practicable to adopt polygonal top chords for spans con- siderably shorter than the superior 'imil just uientioned. T.STIIETICS IN DKblGN. 51 2d. While it is pencrally economiciil of mtiterinl to use very long imiiels, no such extreme length should be adopted ns would involve an awkward appearance due to flatness of diagonals. 3d. The curvature of the top cliord shoidd be made as great as is cousisteul, wilij a proper consideration of web stiffness and counterbracing. 4lh. When It is practicable in Petit trusses to curve tlie top chord to such an extent us to make too small the in(;lination of tlie eiul-posis to the horizontal, it is permissible to let the latter extend over one panel oidy and to make all the main diagonals extend over two panels. The effect is ungraceful, however, when the main diagonals occupy one panel each near llie ends of the span, and two panels each elsewhere. 5lh. When appearance alone is in question trusses very deep at mid-span arc desirable ; but an excessive truss depth is conducive to a reversion of bollom-chord stress — a condition which has either to be avoided or provided for by stiffening the bottom clionis. In extremely heavy bridges, especially where the dead load is unusually great, it is possible that au undue consideraiion for economy of metal might cau.se a designer to adopt a truss depth whidi would be actually too great for appearance . but this is not likely to occur very often because of other limiling conditions. Gth. Tljere are certain limiting relations between width of bridge, (le])th of truss, and length of span which lor the .sake of good effect ought not to be exceeded. I'sually the rules established on account of purely engineering questions will prevent these limits from being transgressed, thus proving a maxim which tin; author lias often maintained, viz., that in any design any violation of engineering principles is also a violation of good taste from an artistic point of view. 7th, A veiy graceful effect can be obtained by placing the lower liorizonlal struts of the overhead bracing in a cylindrical surface similar to that which contains the panel points of the top chords, but, of course, with different curvature. In respect to the decorutiou of each span of a bridge, it may be stated that a little oruameulation is generally much better 52 I)K r'ONTIBUS. than a great ileal, and that this lillle should be nppropiinte and in keeping with the general clmriicter of the slructuio. A prodigal use of cheap casl-iron trimmings at a portal of a steel bridge is not in good taste ; but it b itjifectly proper to decorate the intersections of the members of the portiil bracing by plates or rosettes, to surmount the upper horizontal portal strut by iin jpsthctically designed parnpet, to use orniimentnl corner brnckets beneath I lie lower portal strut, to employ fancy name-plates symmetrically arranged, and to place orna- mental figures of proper size and design at liie hips, pedestals, or middle of inclined end-posts. It is also permissible to ornament the intermediate transverse vc 'icul bracing to a slight degree b}' rosettes and knee-brace bi" such decoration should be applied sparingly. Again, in large bridges it is proper to be somewhat extravagant in the use of metal iit the portal for the sake of appearance, especially as such metal, if it does not add to the strength of the bridge, certainly increases its rigidity. The ornamentation of viaducts and elevated railways is something which has never received in America any attention worth mentioning, as is proved by the inherent ugliness of nearly all the elevated roads of our great cities, and the i)aiu ful plainness of our railwjiy trestles throughout the country. It is principally this neglect of aesthetics in design which has created such bitter ojiposition on the part of the property owners to the building of elevated roads in the heart of the city of Chicarfo. Electric lights and gas-fixtures of artistic pattern can be made great aids in securing a pleasing effect in designs for bridges and viaducts ; and at night a well-studied distribution of incandescent lights can be made to i)roduce a brilliant appearance at the portals of any large and important city bridge. Ornamental handrails are also of great service in decorating trestles and bridges, especially in deck structures, where these rails can be built in the form of a highly ornamental parapet. Architectural effect in bridge-building seldom derives much aid frv paint, for the reason that it is genenlly^ best, on ap /K8THKTIC8 IN DKSIGN. 58 count of l)olli convenience and good taste, to use but one color in pidnliiif< a bridge. A j)r()per choice of color, however, is u niiileriiil advantage ; and it is correct to vary the color in cer- tain accessory portions of tlie structure, such as machinery- liouses, the lettering on name-plates, etc. Some engineers have advocated painting tlu! tension and compression members of dillerent colors, but this would get one hito difficulties in spans where certain strictly tension-members ate made stiff. Ornamental figures should be painted of the same color ns the rest of the l)riilgc. In general, it may be stated that for ordi- nary conditions of landscape the heavier the structure the ligluer should be the color of the paint used, for the reason that if a bridge has an appearance inclining toward clumsi- ness this objectionable effect can be 'essened by reducing the prominence of its members; while, on the other hand, a bridge which is of such an extremely light and airy design as to pro- duce an ap learance of weakness can be made to look stronger by adopting a naint of dark color, and thus bringing its mem- bers into greater relief in respect to surrounding objects. With very dark backgrounds, however, it will often be advisable to use a light-colored paint even for slight structures, so as to give the bridge a definite outline. Ill regard to the ornamentation of bridges by the adoption of elaborately artistic approaches, but little has yet been done in America, the reason being that any money so expended has evidently no utilitarian purpose, and con3e(piently to the eye of the solely practical man appears to be entirely wasted. Ill Europe it is customary to ornament large and impoita-nt bridges in this way ; and the time is coming when it will be the practice in America also. A projier proportioning of jueis and abutments has a great deal to do with the obtaining of an artistically designed bridge; but, unfortunately, in these, even more than in the super- structure, the almighty dollar is generally the ruling influence in the design. In many bridges the piers do not seem to be massive enough for the spans ; and, as will be shown in Chap- ter XXII, too often they are not sufflciently large to meet cer- tain important engineering requirements, which are, as a rule, 54 DE PONTIHUS. iguored by the average designer, and occasionally even !)y some who consider themselves bridge experts. In the author's upinioM, if piers and al)utment3 be adequately designed from an engineering point of view, they will not fall far short of the ideal of artistic excellence. In concluding this chapter the author would advise each )f his readers to study carefully Chapter XXVI on "The Mi- thetic Design of Bridgtis," by David A. Molitor, Esq., C.E., in Prof. Johnson's work on the " Theory and Practice of Modern Framed Structures." Although most of Mr. Molitor'.s illus- trations are necessarily drawn from Europenn practice, there are many features thereof which it would l)c well for Ameri can bridge-designers to adopt ; notwithstanding the facts that European and American practice in bridge-building are fun- damentally and essentially dilleront, and that American engi- neers have little or nolh'.ig to learn from their brethren across the seas concern the science of bridge design. From an artis- tic point of view, however, it must be confessed that the average American bridge is inferior to the average European structure; so, while it is advisable thai American bridge-de- signers study carefully European pr.iclice in respect t.) aes- thetics, they should be caulicMis to avoid tlioiightless imitation; because decorative features which are appropriate to the heavy, nuissive, and costly bridge.s of Kurope would be out of jjlare when engrafted on the light, airy, and ecoiiomic structures that are cliaracteristic of American engineering practice. CHAPTER V. CANTILEVER BRIDGES. Theue seems to be a notion prevalent among the uninitiated (engineers tfX) often iachu'.ed) thai there is some inherent virtue in cantilever bridges which renders tbeni superior to ordinary structures, in what parliculars, however, the said uninitiated are ot often able lo state, although they generally claim that it is iu economy. This notion is entirely erroneous ; fov cantilever bridges are always inferior in rigidity to bridges of sinii)ie truss spans, and, exc'.p;ing for (cnaiu peculiar conditions, are also always more e.ercentage curves, by using which the total weight of metal in any canti- lever bridge of any ordinary type can be found very (juickly ami with considerable accuracy. Before proceeding to present tliese results, though, several other matters will receive consideration. In no work on l)ridges, that the author has ever seen, has there been given a statement of the various stresses for which 58 DE PONTIftUS. the several spans of a cautilever bridge should be figured ; 80 such a tabulation is herewith presented. Stresses in Susi'knded Span. First. Dead-load Stresses. Second. Live-load Stresses. Third. Impact-load Stresses. Fourth. Diiecl AViuil-load Stresses. Fifth. Transferred load Stresses. Sixth. Erection Stresses from Dead Load. Seventh, Erection Stresses from Wind Load. Stresses in Cantilevkr-arms. First. Stresses due to Dead Load on Suspended Span. Secon I. Stresses due to Live Load ou Suspended Span. Third. Stresses due to Lupact L )ad on Suspended Span. Fourth. Stresses due to Wind Load on Suspended Span. Fifth. Stresses due to Transferred Load on Suspended Span. Sixt^i. Stresses due to Erection of Suspended Span and caused by the Dead Load. Seventh. Stresses due to Erection of Suspended Span and cau.sed iiy tlie Wind Load, Eighth. Stresses due to Dead Lo.id on Cautilever-ann. Ninth. Stresses due to Live Loud on (,'anlileverarni. 2\'Hth. Slres.ses due to Liipact Load on (yantilever-arni. Eleventh. Stresses due to Wind Load on Cantilever arm Twelfth. Stresses due to Transferred Load on Cantilever- arm. This load affects oidy the niaiti inclined posts over piers. Stresses in Anchor-arms. First. Stres.ses due to Dead Load on Suspended Span. Second. Stresses due to Live Load on Suspended Spau. Third. Stresses due to Lupact Load on Suspended Span, Fourth. Stresses due to Wind Load on Suspended Span. Fifth. Stresses due to Transferred Load ou Suspended Spau. CANTILEVRU BIIIDGKS. 59 Sij;ih. Stresses due to Erection of Suspended Span and caused by tlie Dead Load. Seventk. Stresses due to Erection of Suspended Span and caused by the Wind Load. Eighth Stresses due lo Dead Load on Cantilever-arm. NitiUi. Stresses due to Livi; Load on Cantilever-arm. Tenth. Stresses due to Iiu|)aet Load on Cantilever-arm. Eleventh. Stresses due to Wind Load on Cantilever-arm. Twelfth. Stresses due to Dead Load on Anclior-arm. Thirteenth. Stresses due to Live Load on Anelior-arm. Fourteenth. Stresses due to Impact Load on Anchor-arm. Fifteenth Stresses due to Wind Load on Anchor-arm. Sixteenth. Stresses due to Transferred Load on Anchcrarin, STKE88K8 IN ]\LviN CeNTHAL SpANS, CUOIIU STUK88KS. Firftt. Stresses due to Dead Load from both Suspended Spans and Adjacent Cantilever-arms. Second. Stresses due to Live Load covering both Suspended Spans and Adjacent Cantilever-arms. Third. Stresses due to Lnpact for tlie latter case. Fourth. Sircsses due to Wind lioad on both Suspended Spans and both Adjacent Cantilever-arms. Fifth. Stresses due to Transfenepan. This loading produces a constant shear from end to end of Main Central Span. Fifth. Stresses due to Dead Load on Main Central Span. 8i.Hh. Stresses due to Advancing Live Load on Main Cen- tral Span. Seventh. Stresses due to Impact from last load. For certain conditions some of these stresses will not need to he considered, but in other cases they will, consequently it is necessary to insert them in *iie lists. For instance, in the cantilever nnd anchor arms the sixth and seventh items will generally be found to have no intluence on the sections of members, but in some cases they will, as in long-span high- way bridges with light live loads. In calculating erection stresses, the weight of the traveller must not be forgotten, as its intluence on such stresses is by no means inconsiderable. The combination of the various stresses retpiires both judg- ment and care, for some loads may or may not act together, and some produce tension while others produce compression in the .same member. Again, distinction must be made be- tween groups of stresses with and those without wind-stresses, so as to use the different intensities of working-stres.ses given in the speciticntions of Chapters XIV. and XVI. It would be loo tedious to give here t lie various combinations of stresses for each member of each span ; but it will suffice to 8;iy that the computer will have to fiiul for each main member in the entire bridge the greatest tension wIhmi wind-stresses are in- cluded, the greatest tension when they are excluded, the great- est compression when they are included, ami the greatest com- pression when they are ('.\(;luded, taking care not to group together any stresses that cannot e.\i>t simultaneously. The determination of the proper live load per lineal foot for any member of a cantilever l)ridge is oiu' nupiiriiig a little CANTILEVEK IJRIDGES. 6L care, the rule being that for the pie^c considered the length of span to be used in applying the live-load diagram is the total length of structure which must be covered by the moving load in order lo obtain the greatest stress in the said piece, ex- cepting only the suspended span and tlie main central span, for which the live loads actually imposed are to be treated exactly like those of simple spans. Of course, the impact is to l)e figured for llie length of structure that must be covered by the live load to produce the greatest stress in the piece un- der consideration. Some young engineers have an idea that the finding of stresses in cantilever bridges is a complicated matter. On the contrary, it is very simple, as every stress can be deter- mined by the ordinary prin(;iples of statics and very readily by the use of graphics. Although the work is simple, it is some- what long and tedious, as is evident from the preceding lists of stresses. The computer is ad viscvl, when finding the stresses, not to try to group the loadings any more than they arc grouped in tlie said lists, for, if lie docs, he will probably have to separate them while making his combinations. In respect to combinations of stresses during erection, there will be ao necessity for increasing the sections pro[)ortioned for other combinations, provided they arc as large as those required by the said erection-stress coml)inations with the in- tensities given in tiie specifications (Chapter XIV.) for com- binations that include Avind-stresses, viz., intensities thirty per cent higiicr than those for conUjinatious without wind stresses. Cantilever bridges may be made either through, deck, or half througli; but a combination of deck-spans for the anchor- arms and a tlirougii-span for the cantilever-arms and sus- pended span is awkward-looking and unsightly. There is a structure of this type across the St. Lawrence Itiver, near Montreal. It is no easy matter to give an artistic effect to a cantilever bridge; nevertheless it is generally wiihin the realms of pos- sibility to do so, althougli it must be confessed that most of the cxistinj^ structures of this type are uncomj)romisingly 6-' DE I'ONTIBUS. ugly. If a convex upward curvo can bo placcnl in the top chord of the su.speuded span, so as to reverse at (he ends into a concave iijiward curve on the cant Hover arm, a graceful effect will be obtained ; but the design generall}- will not be economical for erection on account of the large erection- stresses near the point of suspension. The author luis made a de.sign on these lines for a proposed MOO ft. span liiglivvay bridge to cross the Mississippi River al St. Louis ; and, as the suspeiuled span would be eroded on falsework, there is no want of economy involved. The layout with all the m:iin members dra,wn t(» true scale lias a very pleasing effect. In long spans like this it licconu's nec(s.sary to widen the cantilever and anchor arms uniformly fromeniis to main pi>.r.s, so as to obtain the requisite rigidity for resisting wind-pressure and .so as to keep the wind-stro.sses in bottom chords wliliin reasonable limits. It seldom |)ays, however, to build the trusses of these arms in planes inclined to the verlical, principally l)ecause of the complicated shop-work involved. The author has lately had occasion to design a number of large britlges for a proposed bran(di-line of the Nip|)on Kail- way of Japan. The line, which will be about one hundred miles long, is to follow the course of a mountain torrent that rises from twenty to twenty-live feet in two or ihree hours, and attains in places a depth of water exceeding one hundred feet with a total rise of sixty feet. Of coinse, falsework can be employed for these bridges only to a very limited extent, hence it was necessary to resort to the u.se of the cantilever. Thr^ e of the eight structures were designed as ordinary cantilevers, two as simple truss-bridges, and tlu-(.e as cantilevers during erec- tion and simple spans afterwards. The last style of bridge is very economical of both metal and money, and will bear fiuther investigation and extension, so as to be made appli- cable to crossings where the ordinary cantilever bridge would otherwise be adopted. Its modus operandi is us follows ; At each side of the river there is erected on false work a sim- ple span having its chords and certain of its web members (or for short spans all of them) stiffened for eirciiou stre.-ses. Then over each pier is built a toggle consisting of horizontal CANTILEVKIl HKIDCJIiS. 6;j P u y lO ti|»l)er-chord eyebais mid adjustiiLIe veitirals, by means of wliicli one lialf of the central spun is cunliUivered over tlie stri-am to ineol the otiior liidf, after winch the toggles are to he removed. This mctliod of ereclion can be understood by reference to the diagram in Fig. 1. \M^m^ i^'^ Fio. 1. One of the three oases nuintloned had rather peculiar con- ditions, which necessitated the adoption of another expedient. About nndstream there is a narrow rocky island that reaches to about the elevation of extreme high water. Near the edges of this island, as shown in Fig. 2, will be built two small piers, Fia. 2. each of which will support one end of a long 8p:in. Between the end shoes will run a temporary strut, and from each ped- estal will spring a temporary post to support the temporary top-chord eye bars that run from hi[) to hip. The rectangular panel is braced with Iciiipora'-y adjustable diagonals, and the top chord is hinged at the nuddle and connected to tlie pedes- tals by other temporary adjustable rods. The.»e two sets of adjustable rods permit of liie raising or lowering of one span at a time. By means of this device more tlian one half of each span can be cantilevered out to meet the remainder there- of, which will be erected on falsework. u DE PONTIBUS. It is intended to erect the cantilevercd portions of all three bridges with their ends higher thiin they will be in their final position, so tliat no raising, but only lowering, of tlie weight of the arms by tlie toggles will be necessary. The author is of the opinion that these toggles will work much more easily, and will prove in the end less costly, than the wedges used for adjustment in the erection of the Red Hock Cantilever I' !ge, a description of which was given by Samuel M. Rovve, 31. Am. Soc. C. E., in the Transactions o( that Society for 1891 In one of the throe true cantilever bridges for the proposed Japanese railroad an expedient lias l)cen adopted by the author which ma}' lie worthy of description. One approach to the structure, as shown in Fig. 3, is through a tiinuel end- \TxKK^/KKfWlX\fc ^s^ Fig. 3. ing in the face of a vertical wall of rock. It was at first intended to use this rock in lieu of one anchor-arm of an ordinary cantilever by letting the main i)osts lie close to its vertical face and tying the top chords well back into its ma.ss ; Imt a study of the contours of the rock showed that it dipped ofif to one side of llie line in such a way as to render such an anchorage of uncertain strength, so it was decided to increase the lengths of the suspended span and far cantilever-arm sufficiently to cut out the near cantilever-arm, and thus let the end of the suspended span roll on two small pedestals at the luouth of the tunnel. Five eighths of this span will be erected by toggles fastened into the rock, and tlie reuiaiidng three eighths will be cantilevcred out also by toggles from the end of the far cautilever-arui. This method requires more metal than does the one first contemplated ; nevrrlheless it is the cheapest, everything considered, that can be adopted. The rock-anchorage js amply .strong for the dead-load pulls CANTlLLVhIt i;Ull)>. 6J on il (liiriiijr erection, iilthoiiL'li. as before staled, it is uot silt- liciciitly ielial)le for resisliim tlit! e/rtcts of live lomls. Tlie liest nulliod of atlatiiiiig the suspended span of an oniinaiy cantilever bridge is by lianuers from inclined end posts on the cuntilever-arnis. For such suspenders narrow eye-bars sliould be used ; and it, is generally better to hinge Ihein at the middle. This is because they are subjected to transverse bending, due to longitudinal expansion and con- traction of suspended span from both changes of temperature and the appli(;alion and icnioval of the live load. Narrow bars can spring slightly without being overstrained, and a rotation of the eyes ou the pins will thus be prevented. Such a rotation would eventually enlarge the eyes and cut notches into the pins, necessitating for some futun; time e.vpenslvt repairs. A 8us]iended span thus hung is free to move longitudinally under thrust of train, but its ends are tightly held in a lateral ilirection, so that all wind loads are carried i)r()perly to the bottom (!hords of the cHUlilever arms ; aiul excessive longi- tudinal motion is prevented by the continuity of the track. In cantilever-arms it is better and moie economical to use inclined posts as wtll as vertical ones over the piers, so that the various loads will be canied more directly to the masonry. To insure the travel of the wind stresses down the transverse bracing between these inclined i)osts, instead of up to the apex of the top chord and down the bracing between the vertical posts, the author leaves out oik; pair of diagonals of the upper lateral system between the said apex and the tops of the inclined posts. The same expedient is used also for the anchor-arms and between the hips of the suspended span and the canlilever-arins. All bracing between oppo.site veilical posts and between opposite inclined po>ts should be made very rigid ; and in douliletrack .structurts all the sway-bracing slnnild be i>ro- portioned to carry as a Uw load, with the i)roper allowance for impact, the greatest shear which can come upon it from loading one side of the tloor oidy. (Jreal cure is necessary in designing the pedestals over tbo 66 DH I'ONTinrS. inuiii pit'is 80 lis to cjiiry tl.c loads lioin the llircc hciivy posts to tlie iiiusonry willioul ovcrslniiiiiiig any of lliu rnotiil in tlio pedestal, and so us lo distribute Uie total prtss\ire \iiiiforiiily over tiie masonry bearini;. L'ntil recently tlie autlu»r has adopted pediistals bidll of [)lales and sliapes, but has lately deei(Jed to try steel castings, us the i)ouud price for these has now come down to sonietliing like a leasonable figure. The (liHiculty iu finding room for the proper number of rivets for Hi ladling together llieir component parts renders built petles- tnls chimsy and expensive. The anchorage details reipiire special care, and no rules can be given U> govern their designing, for the reason that the conditions vary for all crossings. The following hints, tliougli, nuiy be of use to the designer : First. The anchor-bars should be made as long and as nar- row as practicable, and should be divided into sliort lengths by pins, for tlio sami; re ison as given in llie case of the sus- l)enders of tlie suspended span. Second. All anchorage details should be accessible to the paint-brush, excepting, of course, those portions of the bot- tom girders wliich are buried in the masonry. This result is accomplislicd by leaving wells in the ancliorages of suflicient size to permit tlie passage of a man to do the painting. If these wells are at any time i)artially filled willi water tempo- rarily by the rise of the stream, no harm will be done, pro- vided that the painting of the metal-work therein be always attended to pro|)erly. T/ii'rd. Concrete foi ancliorages is always i)referable to masonry, because it can readily be made to lake any reipiiied form. If necessary, its exterior can be protected against alnasion from ice or drift by facing with granite or other hard rock. Fourth. There sliould be an iudepcudeut anchorage against wind-pressure, obtained by sliding surfaces of steel, one of each pair of same forming purt of a heavy detail which is rigidly attached lo the bottom of the end floor-beam, and the other forming part of a heavy detail that is anchored lirml^ to the musoury. (.ANTir.KVKU ItUIIKlKS. 67 Fiflli. Th(! tops of the michor-picrs should be Tiuuic al)Ho- IiiUly \val»'i-li<;lil witlioiil iiilcifciiiif; with tlic loiiiiiUKlinnl < xpan.sioii of tlic anchor arm, .so as lo i)rcvt'nt nisliiig of Ihe iiit Iwicc as great as the inaxinuini uplift on the .said aiiciioi-picr, when tiie eU'nct of impact is duly included. A few observations concerning some of the largest caMti- lever bridge.s yet built may be of service to the reader : Till' largest structure of this type in the world is the bridge at (^ueensferry over the Firth of F()rlh, the main portion of wliich consists of two spans of 1710 ft. each, with central spans of yr)() ft. each, and two anchor-arms of 680 ft. each. The length of the tower-span over ilie centre pier is 200 ft., and that of each of the two other toweu-spans is 145 ft., making the total length of the main structure 5410 ft. Tho design for this bridge and a complete history of its <'oii- slruction are given in a .special work published by Engineer- in;/ (London). The exceptions whic;h the author would take to this design are as follows : Fir.st The suspended s|>an8 are just about one half as long as lli(!y ought to be foi- both appearance and economy. Second, The structure should have been made piii-con- necled for both case of erection anil certainty of stress dis- tribution. Tliird. A single .systc;*.". of cancellation for the webs of the girders would have been more scientiilc than the double sys- (iin adopted, and would not have been any more expensive. Fourth. The structure as a whole, fronj the point of view of American i-ngineers, was unnecessarily expensive On the other hand, though, the labor Involved in both the designing and buil.lin;^' of this bridge was immense ; and the successful compljtioM of the stru(;ture is a great credit to all concerned in its dosigiung and construction. The C'lutiiover bridge having the next longest span is the Lausdowne Bridge over the Indus Kiver at Sukkur, ludiu 68 I)K roNl'IHU.S. Tt consists of Ji siiiirlt' spun of 8','() f(. willioul iinclior-Jirnis, \}.iC hitter being replaced by guys, uiid witli ii suspended span of 200 ft. Tiie uppeanince of llie bridge is bizarre in the extreme, and the slnu;lure is economic in i;eillier weigiit of material nor cost of sliop work. Compared witii an American bridge of llu; .same span, capacity, and strength, the weiglils of ineial in llie 820-ft. span only would be about iu the ratio of unity to 0.7r>. The ci!. itilever having the ne.xt l';:-.j;est span, viz., 790 ft., is the railway britige at Mcm|>his over the Mississippi River. This structure is botii unsightly and uneconomical of uuiterial. Its layout of spans is \infortunate (but the War Department, jtnd not the designer, is lesponsible tor this), and the truss depths are far too small for both economy and a[)pearaijce. The re.vt longest cantilever span is that of the lied iiock Bridge over the Colorado River on the Atlantic and Pacific R-dhvay. Tliis structure consists of a main span of 660 ft. and two anchor-arms of Ki.'} ft. each, the length of the sus- pended span being 830 ft. Tlie widtli between central i)lants of trjisses is 15 ft. ; and the truss tlepth varies from 55 ft. for the su.spended span to 101 ft. for the vertical posts over the main j)iers. As tiie author is liie peison responsibli; for its layouc, his criticism thereof will not be of much valu' Tlie bridge was designed to meet certain conditions, economy in first cost being the prime reepiisite ; consctiuently the subjci I of a'sllietics did not receive great consideration. Engineers Hud architects ditfer fundamcnlally in llieir opinions concern- iug the architectural elTect in this structure. Some approve its appearancte; other.s characterize it as harsh in its outlines. Tlie relations between lengths of su,spended span, cantilever- arms, and anchor-aruis, and those of width and depth, although very hurriedly determined, hav(! since been found to be just about the best practic;ii)ie. This bridge, as before stated, is •■■::>cribed ver}' fully in the Tranmctionx of tiie American bociety of Civil Engineers Tor 1891. There are many other <'antile "<'r bridges having main spans of from 400 ft. to 500 ft, nr more, but sjiace will not permit thfiir enumerfiliou. SI n (JANTII.F-VKfl T5UID0KS. Many expedients have Ixvii used to connect the metal-work (if tlie juspeudcd spans of cantilever bridges, and considerable troubi'' f.as often been experienced in doing tbe work, owing to variations in bolh Icnglh and elcviition. Tbe author is of the opinion iliat but little ditliculty will be experienced if the following precautions be taken : First. See that the entire triangulation is so accurately done that there will be no jiossibility of an error exceeding one i|Marler of an inch in the ilistunc-c between centres of pins over main piers. A perusal of Cliapter XXIII. will show that this is perfectly feasible. Second. See that extra precautions are taken l)y the inspect- ors during tlie manufacture of tiie metal-work to insure that all lengths of main member, shall be absolutely correct. Third. See that the tapes used in shop .and tield arc of exactly the same length. Fourth. Use toggles like those described in this chapter for elfecting the adjustment. Fifth. Arrange to have the meeting ends of the chords a tritli' high, so that lowering and not raising will be necessary. Si.ith. Arrange matters so tiiat when the eiuls of the metal- work come together they will be a trifle apart ratiier than tend- ing to lap, for it is nuicii easier to heat the chords slightly by sus- pending beneath thein siicetsof metal containingslow tiresthan it would l»e to cool them by packing ice around them in cloths. Referring now to (lie before-mentioned special investiga- tions nuide b}' Mr. lledriok, inc (piestions set him for solution at the outset theri'()f were the following . Firs/. 'I'lie ratio of the economic length of suspended span to tJiat of the total o|)etMng, Seco7id. Tlie most economic length of anchor-arms when the total lenizth l)cfweeii centres of :ii\cborages is given, and when the main piers cat) he placed wherever desired. Third. The relations between the weights of metal in die suspended spun, cantilever-arms, anchor-arms, anchorages, main pedestals, juid nncliorspans. Fourth. 'IMie best proportionate length for anchor spans, and the comparative weights of nu'tal in those of dilferent lengths. to hK I'ONTtias. Fiflh. Tlie nitio of weigliis of inetiil in cantilever bridi^es of to tlu )le- l)ii(lge8 li tliu siiint viirious ly uiinilKT of spiius. Mr. Ilt'diick's luclliod of (U'tciniining tlie economic functions was t<» (like llic dutii on liaiid for tlie |)roposed Japanese caiili- ievcr biidgea, exact weiglits of inetul liuving been conii)Uted for structures of I{"20-ft., 400-ft., and 500-ft. openings, and, i»y varying tiie layouts so as to use longer and sliorlcr suspciidctl spans and longer and shorter ancliorarnis, obtain, by at;tual designs and estimates, the weights of nielal for a suUicient nundxr of layouts to indicate the desuc 1 mininui. In deterniiiung the economic length of susjKMidcd span for a certain opening, the Icnglh for the ancliorarnis was tiri-t assumed to be one fourth of said o|)ening, then the total weight of metal in the entire briroportion. On ihe otlier hand, tliougli, it has been found by trial tliat, with tht; thi(!e-eightlis ratio, there results a more sightly layout than can b(! obtained with the one-half ratio. Next Mr. lledrick tabulated the various component truss and lateral weigiils of several of the typical cantilever bridges designed in tin; author's ollice, the leading diiiiensions for which are given in the following table. P'rom these weights he constructed the curves shown on I*iate X, from wliicli can be found the total weight of metal in the trusses and lateral systems of any three-span cantilever bridge, wlieii the weight per lineal foot of tlw; trusses and laterals in the suspended span is known. Tiiis weight, by liic way", is, on tli 'A a U O (n H Q o s « - s i. ^ -7. Ml- i, z. Si — M a; (U ^5 5r> O 2 S u _ is '-> H ■QO o ^ '-/: _ -• ^ /J ~ b^ TO x = St ^ ic I.-? "*' ^J CI Tl IM "a CO ■/i Mv. V Ci ?» ^-4 *.^ 5 t s -r •^ / • ^ *-- a> t-« - »— < T-H »— 1 "" H 9 3 o s 03 5 •r» Ti — X -' i ^ re IT o ,S L. ♦* z -i ^ ^ ^ iX « •^ *- :s J^ rt '^ 1— 1 *-H »-H CO -•i ?f5> — 5 o t* IC o ^ ■^ iT «o « l- ^. - 'W ^-^ T— * 1— < co ^ V -■ o o c -:( => o CO ^ ?» ?» ys o o LO o CO i- o c o o o 0> O O to TO c~ -r i.o :s X' -^ ^f- -.'"^ c •" -•- * c s - := c ." : ." a to to to a o o If) - -;i - $ 5 j. ^ r* '-fi 2*-'— .'"-''-i.T'^?' i #i^:'; l»K PONtlJiLTS. beiuj^ refiuiied mainly for stiiTt'iiiiig coriiiiii truss members to resist erection stresses. Of course, if false woik be used for the suspended si)!in, tlie eight i)er cent excess will not l)e iuhled. The curves of percentage's are based on two assumptions, viz.: first, the panels througliout the entire .structure are of ecjual length, and, second, the lengths of the cautilever-arms and anchor-arms arcthe same. The first assumption is nearly always correct, for there is no advantage to be gained l)y var^-ing the panel lengths in tlie various portions of tlie bridge. If the leugtlis of cantilever and anchor arms are un- equal, the average weight of metal obtained for the latter by use of th(! curve will liave to bo corrected by the formula T where T' is the correct, final weight of truss and lateral melal in the anchor-arm, T is the weiglit of same found by the percentage curve, and r is tlie ratio of lengtii of cantilever- arm to that of anchor-arm. It should be oi)served that, in applying the percentage curves to structures having subdivided panels like lliose of tlic Petit truss, the niain or double panel is to be \isc(i as the basis of calculation. The me'hod of ai)plying the percentage ("'rvesisas follows: Let us tuUe any opening and assume tiiat tlie. , .ue six panels in eacli cantilever-arm, and that the weight per foot of truss and lateral metal in tlie suspended .'■pan i.« /f- the n-inel length being ;), aiid ;>//> = W. It is to be o' i,.' ti . this method is applicj'.ble for any proportitMiate li ,, u (»f suspemkd span. The weight of metal in the lloor system, being independent of the span length and simply a function of .le panel length and of the distance belv.een liUsses, is not oousiderec! in tiu; investigation, but is, of course, to be added when figuring the total weight of metal in the structure. The weight of truss and lateral metal in the cantilever-arm will be \.2W-\- \AW]-\MW I '2.0]F+3.4Tr-f 3.0Tr-11.65TK. CANTI! KVKIl HftlDOES. n The weii,dit of metal in the piiiiel over the pier is, iiccordiiig to the directions on the diagnun, 1.8 X •.i.OW= 5.4 U'. Let us assume that there are only five panels in the anchor arm, then the trial weight 7' will he Q.irtW't i.7.-)}r-f 2.nK+ 2.r)ir+ :].oif= lo.ioir. Suhstittiting in the formula gives ^„^1.L 10 1^3 G^^jj^j 2 i> \ W. It will he seen from tlu-se calculations that the full iiercent- ages given* for till' I'lui jjiiiici points of cantilever and anchor arms are to be used, al'hou!,di in reality there is hut a half l>anel lengtii for eacii point. This is ctiuscd by the heavy details reipnred at these points for aiijustment and anchorage. All erection metal at liie end of a suspended span is a-ssunud to belong to the cantilcvci- arm. Should in any case the panel lengths be unecjuul in dill'erent portions of the structure, it will be a simple matter to use the curves hy finding avcriiLre weights per foot for two assumed cases of equal panel leiiLrths, one makiiiij iIk- arm greater and tiie other makinsi: it less in length than it aclualiy is, and interpolating properly between the results tor tin; retpiired average weight ))i'r foot for the arm. The total weight of nietai in tin; two anchorages of any three-span cantilever bridge can be taken '\i five jK'r cent of the grand total weight of metal in the said three spans, and the weight of nu'tal in the pedestals on nniin piers at four per cent of same. Of course, conditions va.'-y fill have to be increased to 'iJnr, not withstanding the fact that the entire top chords nniy then l)e built of eye-bars; but siirh short spans would probably be barred out by consideration for navigation interests. The percentage curves of Plate X will not bear a rigid criticism, in that they nuike the weight of njctal dejieud ujton th(! number of panels. It is j>resu|>posed, however, tiiat tiie panel length adopted i.n the most appropriate one for the bridge; and the curves will be found ((uile accurate wliencver llie pniper jianel length is used. With long jianels tlie weight of nietal pt^r lineal foot found by tiie curves for cantilever and anchor arms is h"^s tlian that found therebj- for short panels. This is as it should be, but to a limited extent only ; for it can b(! found by trial that an abnormally siiorl or abin)rmally long panel length will give results too gnat or loo small wlien checked by C()m|tutatioiis of weights made from actual designs. 'i'iiese per(;enlage curves enabled Mr. lledrick to solve reatl ily the n< \t problem, viz., given the total distance iH-twe.ii centres of anchorages and cund tiiis length to be two tenths of th(! tt)tal di.stance between the am b uages. It must not be forgotten that for every dollar saved by re- ducing the toinl weight of metal through the shortening of tlie anchor-arms, it will be necessary to spend about twenty cent.s for extra concrett' in the anchorages. On t ids account, for tlie conditions assiinud, tlie truly economic length of < :i< ii CAN'TILKVKIJ JUUlXiKS. rs ;tli. ally foot not- II lie be (inehoruim of u tluct'-spiiri cHiililcvcr will gciiorally be a littlt; ^renter lluiii Iwent.y per cent «)f the total (lisljuicc between (Seniles of anelKM-iiircs. When, liowever, the problem is to determine the economic lenifth of anchor-arm for a lixeil distance between main piers, the result will be (piite (lifTereiil , ])C(rause, within reasonable limits, the shorter Hk; anclior-arm tiic smaller will be its total weight of metal, iiiid because tresth; approach is much less expensive than anchor-arm. It would not, for cvidciil reasons, l>e advisable to make the length of anchor- arm less than twenty per cent of that of the main ()i)eiung, or say lifteen jiir cent of the total distance between centres of ancliorages. With this length there would probably be no reversion of stress in the cliords of the anchor-ium, even when impact is considered. Generally, though, the ajtpear- aiKie of the structure will be improved by using longer ani;hor- arms than the infe'ior limit just suggested. In respect t > the best proportionate length of anchor- spans, the latter weigh so much per lineal foot for all cases liiat the shorter they are made the greater the economy , but, as before stated, it i.i improbable that navigation interests would ever jiennit of their being made shorter than one half of tlie main openings. In respect to his lifth and last prol)lem, Mr. Iledrick obtained the following restdls . The total weight of metal in a threo-span caittilever railroad bridge, tloor system incluiied, is to tiie total weight of metal in a simple-span Inidge of thnvocpial openings, for which false work is to be used throughout, ;is unity is to 0.0. The cor- responding ratio for the case of the i-eatre span, elected with- out false work, is unity to (U. For a very long brinding liridge of equal, sim|ilc spimsand the .sanu! mimber of piers, the spans being erected on false* work, is its imiiy to 0.75. For Ihe case of allern;ile sinip'e spans erected without fal.se 7« i\K I'ONTIfU'S. work, the ratio would bo as unity to 0.8. Tlii'so results were obtained by assuming average probable conditions ; but the longer the sinii)le spans and the greater the total lenglii of structure, the less will be the variation in weights of cantilever and simple-span l)ridges, althougii it would reipiire very long spans and a great total length of structure to change material- ly the rat. OS found. It is therefore evident that, when economy in first cost is considered, as it always ought to be, there will seldom, if ever, be any need for considering the adoption of cantilever bridges with anchor spans, because structures wiih simple spans are l»oth cheaper and better. It is also evident that in many eases it is advisable, from (lonsidcrations of botji rigidity and economy, to adopt a bridge consisting of three simple spans, with the middle one cantilevered from the others, rather than the ordinary three-span cantilever bridge. When each of the side si)ans is as short as one-half of the middle span, or even shorter, there will be no dillieulty experienced in the erection, and no great ])rovision will be res depliis, the autiior's practiee is to nialie the least disiancc iH'tW( en parallel trusses one tw<;nt}'-sevenlh of the main openini;: the least distance between axes of vertical posts over niairi piers, wlieii the trusses converge towards the suspended span, one Iwent^-iifth of the said opening; the truss dcptli for the suspended span, when the cliords are parallel, from one fifth to one sixth, or for very long spans even one seventh, of the s]mn; ai;d the height of tlie vertical posts over main i)i(!rs not to exceed foui', or preferably three and a half, times the lu-rpendicular dis- tance between their axes. For through cantilever bridges llie author generally makes the height of these posts about lifteen per cent of the length of the main opening. For tlie sake of appearance the centres of the toi)-chord i)ins in cantilever-arms are i)lac(d on arcs of parabolas, the vertices of which are located at the hips of the suspended span; and the anchor-arms are laid out to the same curve, beginning at the toi>s of the ])osts over the main piers. In concluding this chapter, u ciieck on the correctness of percentage curves for weights of cantilevers will be given by applying the (Mirves to the published estimated weights of metal in tlu; various members of the longest cantilever bridge that has ever yet been designed in detail, viz., the proposed 2'iOI)-ft. span (meas»u-ed between ceidres of main piers) for the North Jtivcr Bridge at New York City. This proposed .structure was tiesigned by the Union Hridgc; Company. Tiie total weight of metal in trusses and laterals of the 730- ft. susi)en(led span is 10,400,1)00 lbs. 'I'lic trusses, which are of the I'etit type, are divided into six nuiiu pancils of 1'.20 ft. eacli;conse(iuently the panel weight is 10,-100,000 : (i^l.TJW.OOO lbs. In the cantilever arm theie are six and five eighths nuiin panels; conseijuently the weight of trusses ami laterals there- for will be i.20W-]-\A0W~\ l.fi-jVr f 2.00 )Kj- 3.40 W^ t 3.00 IK i- i X 3.«0 W = i;}.!»0 W '- 34,01)0.000 Ib.s. 78 i)K PONTinrs. Tj'.u'h !iricli()r-arni is 840 ll. li>iiir, niid is divided into seven doiihle piiiicls, mid tiieic aie seven and (ive-ei,L;iitlis loads to be considfiod; eonse(iiicntly the wciglil of trusses und laterals llierefor will be 0.75 ]V \ 1.75 W \ 2.10 W-j- 2.50 JK-j- 3.00 TK-|- 8.75 \V-\ 1.75 W -\- I X 5.(15 ]V z^ 22.13 W - :j^;551,000 Ihs. 'I'liis weight niiist l)e rediicod, ouiii.iT to the fact, tliiit Ihe length of tlie cantilever arm is only si.\ sevenths of tliai i>\' Ihe anchoi-arni, making /• — 0.y57. 7" = -(1 + ;•) = '-''^'''•"*"\l.t?57) ^ 35,009,000 lbs. The total weight by the curves for the two cantilevc i and unciiur-arms is therefore 2(24,090.000 f 35,009,000) = 119,308,000 ll)s. The total weight of metal given in the i)ui)lished cslimatc for trusses and laterals for the two cantilever and anchor arms, after deducting 11,500,000 lbs. for weight of metal in liie anchorages and ignoring the allowance for sundries (which was prol)ably put in for prudential reasons), is 119,700,000 lbs., making the dillerence 802,000 ibs. , or about one (piarler of one per cent. This is an extremely accurate check, and proves that the curves are reliable ; nevertheless the author would not guar- antee them to give any such close coincidence for all cases. Since tiiese pages went to press Ihe author has been engaged on the making of a jireliminary design with a delailcd oil male of weight of metal for a proposed double-lrac;k railway and Inghway cantilever bridge, with a central opening of 1,(500 ft., to cross the St. Lnwremo Hiver near Queb(!c, Canatla. The result of the estimate as far as it has been car- ried gives another excellent check on tiie accuracy of oue of the curves ; as the error for Ihe (.antiiever arms is only one- eiglith of one per cent. The anchor arms have not yet been (Jt; tailed. CHAPTER VI. AKCHE8. TnK. tirch is II nitlier iituoinnioii type of strnrliire in Anier- icu, liccHiisc the conditions wliicli iiiukc il fconoiuirai aic unusual. For docp gorges with rocliy sides, or lor shallow slreiinis witli lock boUoin and natural abuliucnts, arclies are eminently jjiopcr and cconoiuical. But when a steel holloni chord is needed to take up the liirusl between springing points, all the economy of the arch vanishes. The advantages of tlie arch are a possible economy of metal and an ieslhetic api)earance, while its disadvantages are a lack of rigidity and, for most types, an uncertainly concerning the nuiximum stresses in the members. Arches are sometimes used for large train-sheds, in which their architectural ellect is certainly very flue, but tliey require about twice as much metal as docaiitilevered trusses supported on colinnns; conscniuently they can be adopted only when appearance is an extremely important factor in the design. When bridge foundations have to be built on piles or on any oilier material that is liable Ui slight .'settlement, or when the abutments roidd po.ssibly move laterally even a mere trille, it is nut [uoper to adopt an arch superslruelure; for any seltle- nuMit or any motion whatsoever in eitlier piers or abutments would up.«et the conditions assumed for the compulations, and thus cau.se to be increased to an uncertain amount some of the stresses for which the superstructure was proportioned. This criticism does not apply to the three-hinge 1 arch, but even this design reiju ires good, solid abutments ann firm foundations for piers. Arches cau be erected on false work, by cantilevering, or by 7D ^>. lAAAGE EVALUATION TEST TARGET (MT-3) // ■^>. f/. S %1 I u 11.25 U^|2^ 12.5 ■ so "^~ H^B 2.0 - 6" 1.8 JA lil.6 V] vQ v: # %!^ "•^yV"^ 7 Photographic Sciences Corporation 23 WEST MAIN STREET WEBSTER, N.Y. 14580 (716) 872-4503 \ S «^ 4^ 80 DK PONTIBLS. biiiUUiig vertically the two halves aud lowering them by cables till they meet at ti»e centre. Whichever of these methods is the easiest and cheapest is the one to adopt. A very easily erected arch is shown in Fig 4. The pieces marked AB are temporary, and are to be used only during erection. They can be made of timber, so as to be lemoved Fio. 4. readily after the arch is coupled at mid-span, or may be of steel, and be left in as idle members, solely for the sake of appearance. It will be seen from the diagram that the structure is a cantilever during erection, and afterwards consists of an arch span and two simple spans. This type of bridge probably requires a little moie metal than would an ordinary segmental arch with trestle-approaches, and possibly is not quite as rigid as the latter, but the saving of cost in erection will fully oflfset these disadvantages. With three hinged arches there is no ambiguity whatsoever in the determination of stresses, but in all other cases there is. There are four cases all told, viz.: 1. Arch without any hinges. 2. Arch with one hinge (at crown). 3. Arch with two hinges (at abutments). 4. Arch with three hinges (a crown and abutments). These four cases can be reduced to three, because there is no good reast the arch without hinges, so as to obtain tlie greatest po.ssible rigidity, even at the expense of certainty in computing stresses. For arched train -sheds, the two-hinged arch of crescent shape will generally be foiuul the most satisfactory. Winle the author was engaged on the i)reparation of this ciiapter he received a copy of Prof. Malverd A. Howe's new book, entitled " A Treatise on Arches." This work, which is entirely mathematical in character, is certainly the most com- plete book on arciies that has ever been written, and appears lo cover the entire subject of stres.ses in arches of all kinds in a most satisfactory manner, altiiough, of course, the author cannot vouch for the correctness of Prof. Howe's figures without checking the matliematics from start to linish, a task which lie feels is too great for both his spare time and his advancing years. It is probable, tliough, that the autlior will have the book checked some time by one of his assistant engi- neers, in ca.se tliat he has t > make another design for an arch. Meanwhile lie is satistied to assume that all of Mie mathemati- cal work is correct, because of Prof. Howe's established repu- tation as both a mathematician and an engineer. Prof. ^lowe has tabulated the res\ilts of his compulations in a very con- venient form, so that his forniulw can readily be applied in designing, especially for preliii'inary designs and estimates. In .spite of its discouragingly mathematical appearaiute. Prof. Howe's book promises to prove of great practical value to designers in utruclurul st(cl ; :in(i its author is certainly to be 82 1)E PONTIRUS. commended for the immense cllori lie liiis put fortli in accom- plisliiiig for the eugiueeriug profession such a luoorioua piece of work. Prof. Howe finds the reliiiive weiglits of metal in a 416' arch with a 67' rise, for Cases Nos. 1, 3, and 4, to be as follows ; Case No. 1, no hinges 1.00 Case No. 3, two hinges 1.21 Case No. 4, three hinges. 1.30 The author is of the opinion that, if he were to make three such designs for coinpaiison, there would not be such great diflfe'-ences in the weights, because constructive reasons will cause the designer to use only a few diireront sectional areaa in the chords of an arch, while Prof. Howe's students, who, as he stales, made the calculations from which the tabulated ratios were determined, probably proportioned the section of each panel length of each chord for the greatest stress to which it could 1)6 subjected. This would be eminently proper in making such a comparison; but the resul'sof the computatiouH would not agree with similar results obtained by a bridge specialist. It is ditiicult to nwike a proper comparison in re8j)ect to economy lie. ween arched and simple truss bridges, owing to the fact that the piers differ for the two cases; but a fair one can be obtained by assuming that steel braced piers are used to siipport the deck span. The author has had occasion lately to design in complete detail for a British Columbia railroad a 260- ft. arch bridge, shown in Fig. 5, h.'.viug a rise on the centre line of 59 ft., and to compute the exact weight of metal in same. For the sake of comparison, he has since designed according to the same specifications a 260 ft. deck-span, having a truss depth of 35 ft., resting on steel braced towers 36 ft. high. The total weight of ntetal tor the arch design is 2,111 pounds per lineal foot, and that for the truss design, including the towers, is 2,542 pounds per lineal foot, showing a saving of about twenty per cent in favor of the arch. ARCHES. 83 As for the relative rigidities of liiese two structures, tlicre is very little doubt that a con)|)urisou of the tiuislicd bridges under load would result ia favor of the simple spun. la making the prelimluary study for the arch bridge herein Fio. 5. referred to, there was prepared a cimiparative design for a three hinged arch, in which each half oi each arch consists of a lenticular truss as shown in Fig. 6. Contrary to the author's surmise, this design did not prove ..--A tAiR>ra9!*»m:Mi Fio. 6. to be any more economical than that with the circular arch, t,hfi toUl weights of metal in thu two structurea being ahinn 84 DE PONTIBUS. exactly the same. Tlie circular arch was, therefore, adopted on account of its superior appearance. Couceruiiig the rehuions between the principal dimensions for arch bridges of various types but little can 1x5 said, for tiie reason that but little is known, because of the sci'vrcity of such bridges in ihis country. In most cases the length of span a*:d the rise are determined by the existiig conditions at the cross- ing. For any given span, the greater the rise the less tlie ellect of uniform load stresses, but the greater the elfect of partial load stresses, and vke verna. Again, for any given span and rise, the -\rch depth does not affect the uniform load stresses materially, while it does so affect the partial load stresses; and as the latter are inferior in importr.nce to the former, it results that the depth of an arch for economy of material will be very much less than the best depth for an ordinary truss of the same span. The arch depth, too, will depend upon whether the arch has fixed ends and continuous crown, hinged ends and continuous crown, or hinged ends and hinged crown. For the Urst type, a varying depth increjising from the centre to the ends is econojnic; for the second, a varying depth increasing from ends to centre is l>est; while for the third, a constant depth from end to end seems prefer- able. Again, the arch depth will depend considerably upon the style of web, i.e., whether it be plate, open-riveted, or pin-connected. The best depth or depths to adopt for any case should be given a special study, in miikiug which Chapter VI of Prof. Howe's book will be found of great assistance. In respect to the style of curve to adopt, whether circular, parabolic, or elliptical, the author's preference would generally be for the circular on account of its simplicity, although the parabolic might theoretically give better results. In the author's opinion, a plate-girder arch should be made without hinj-'is, an open-webbed riveted arch either with or without hinges, and a pin-connected arch with hinges. In the latter case, it is only the web members that should be pin- connected, for the chord members should be riveted up and fully spliced from end to end. There slioitld be only a jingle system of cancellation used in webs of arches, so as to ARCHES. 86 Avoid as mucb as possible ninbiguity in tbe stress distribution. Riveled connections arc preferable to pin connections for the diiigonals on acconnt of rigidity, but are more expensive for erection. Hard and fast rules for the miuinmm spacing of outer arches of bridges for various spans and rises cannot well be given. The narrower tbe structure within reasonable limits the less the cost, but the less also the rigidity and the lateral resistance to overturning from wind-pressure. In the a60-ft. span herein rt rred to, the author made the distance l)etween c«mtral planes of a hes twenty-two feet, which was as small a distance as he dared to adopt, notwithstanding the fact that economy of first cost was an important factor in the design. An approximate rule to work by might be to make the perpendicular distance between outer arches not less than one third of the height from springing point to grade. In concluding this ciMipter, the author desires to call atten- tion to the fact that there is still a great deal to be learned about the designing of arches ; and to suggest that some pro- fessor of civil engineering, who is well posted on bridge de- signing and who has time to spare, could spend several mouths to the great advantage of the engineering profession in deter- mining the proper relations of span length, rise, arch depth, width between exterior arches, etc., for the variims styles of arch, and in ascertaining the relative econonnies of the latter. CHAPTEK Vn. TRESTLES AND VIADUCTS. But little need be said in this chapter concerning the design- iug of trestles and viaducts, as that subject is fully covered in Chapters XIV and XVI. However, as the latter chai)lers are BpecificHtioiis, and are written in very concise form, it seems advisable to give Ijcre certain cxphuiatious of the reasons for the rules and directions titoreiu formulated, even at tlie risk of repetition of a few matters. The best layout for a trestle or viaduct is the one which will make the cost of tho structure a minimiun, provided that the speciUcations used in designing will insure for any layout tlie requisite strength and rigidity. As stated in Chapter III, the greatest economy will exist when the cost of the bents and their pedesUvls is equal to the cost of the longitudinal girders and longitudinal bracing. On this account it is advisable to make the t( 'ver spans shorter than the intermediate spans, taking care, however, pot to have the former too short for either appearance or proper resistance to traction. In general, tower spans should vary in length from twenty to thirty feet, although for very low structures it may son)etinies Ije advisable to go a few feet below twenty. For the intermediate spans the length generally varies from thirty to sixty feet ; but for very low structures with heavy rolling loads the economic length may be found to be less than thirty feet, in which case it will be perfectly legitimate to re- duce the span length to suit the economic conditions. The reason for adopting sixty feet for the superior limit is because trestles and viaducts are nearly always erected without falsework by starting erection at one end of the structure and dropping the members down by means of an overhanging 86 TUKSTLES AND VIADUCTS. »t traveller running on top of tli(! crectdl portion of the work. With tower spiins of thirty feet and inlermediiite spjins of sixty feet, the tniveller will have to reach out ninety feet to erect n tower, which is about the extreme jiracticable limit. However, .should it be nece.s.sary to use more or less falsework, longer spans than 8i.\ty feel would pn)bai)ly be economic. The niost economic layout for a highway viu'luct with wooden joists is alternate towers and spans that consist entirely of joists, the limiting lengths of span being about twent}' feet for the towers and twenty-four feet for intermediates, which latter length is the greatest span u ed in general prac- tice for 4" X 16" wooden joists. It is not leu;itimale in such a design to rely on the wooden joisls of I he tower spans to act as a part of the longitudinal tower bracing. In railroad trestles the longitudinal girders should abut against and rivet into the webs of the columns, the latter being bent just below the longitudinal girders when the legs are battered. Tiie author has lately adopted this detail in some trestle designs for a British Columbia railroad, and has found it to be very satisfactory. For double track structures, the columns at tops are to be spaced a distance equal to the sum of the perpendicular distance between the longitudinal girders of one track and that between centres of tracks, and the legs may be made vertical up to a limit of !d)out twice the perpen- dicular distance between axes of opposite columns. For single-track structures, it is generally best to space the longitudinal girders and tops of columns ten feet centres, al- though an eight-foot spjicing is legitimate. The former spac- ing gives greater rigidity to the structure, but necessitates the use of deeper timber ties. By using very deep ties a greater girder spacing may be adoptcsd; but this is not necessary, un- less very long intermediate spans erected on falsework be em- ployed. It is not worth while to use a batter for columns less than one and a half inches to the foot, and it is never economical lo use one greater than three inches to the foot. Tlie smaller the batter the less the total weight of transver.^e bracing, but the greater the tension stres.ses on the columns. As a rule, it ss 1)R POM I HUB. is best to keep these teii>^ioii sirt'sscs low or even to make tliein noil existent; but in iiigli iresllcH it becomes necessHry to per- mit and provide for tliem. It is wbeu trestles are on sliarp curves tliat great butters must be used, in order to providf the bents, and liracing longitudinally between the vertical columns of alter iiate adjacent bents. In very high trestles, especially when located on siiarp curves, the combinations of coitunn stresses for live load, dead load, traction load, centrifugal load, and wind load run ex- tremely high, and demuiiti great column sections ; conse- quently in such cases it becomes necessary for the designer to use considerable good judgment so as to reduce the toUil stress to reasonable limits. For instance, the traction stresses can be cut down to leca than one half b}^ riveting tlie longituditud girders of an intermediate span to the towers at both ends. This reduces the thrust of train owing to the iiicreased length of structure used for determining tlie etiuivaleiit uniforni load, and fixes the tops of the towers so as to make a point of con- tratlexure at mid-height, tlius reducing the lever-arm and therefore the bending moment to one half. Again, unless the grade be heavy, it is often legitimate to assume that tiie velocity of train is materially lessened by the sliarp curve by the time tliat the train reaches the high portion of the trestle; and, as the centrifugal load varies as the square of tlie velocity, the stresses from this load will be greatly re- duced by the assumption. Again, the prevailing high winds and the centrifugal loads may act against each other instead of together, and the com- bination may be lowered in amount by recognizing this fact. In short, the designer in such a c.-ise can use his judgment to great advantage, and thus stive considerable metal that is not really needed, although it might be required if a strict ad- berence to the .specifications were enforced. TRESTLES ANH VIADUCTS. 80 Till! best style of bniciiigfor l)()th Vac longitudinal ami trans- verse faces of the towers consists of stiiTdiagoniils, each formed of four angles with u single line of lacing, all of stdd diagonals being riveted to the columns and to each other where they in- tersect by means of plates, and no horizontal struts being used except at top and bottom of towers, where they are necessary to make the bracing a complete system. The panel-points of the longitudinal bracing should coincide with those of tiie transverse bracing, although near the top of the tower the panels of the latter may be divided on account of the small distance between columns. In cheap structures, expense can be saved by making the diagonals of the sway-bracing of adjustable rods, and putting in horizontal struts at the panel points, which struts should always be riveted at their ends to the columns ; because pin- connected struts do not stiffen the columns sufficiently to war- rant the figuring of the latter as fixed at the panel points. When adjustable diagonals are adopted, the employment of horizontal struts at top and bottom of towers on all four faces is even more imperative than it is when stiff diagonals are used. The author has seen trestles without such bottom struts, in which the columns have been moved considerably out of place by the rods contracting in cold weather and drawing the column feet together. Six months afterwards the rods elongated and hung in festoons, so were prom|)tly tightene't the coliunn spring laterally under changes of temperature, '•iking care thtit it be proi)()rtioned properly to resist tl:e stresses due t(» such springing when (he s.ime are combined with the other stresses to wliich the column is subjected. Fixed ends for columns of solitary bents are much more con- ducive to rigidity of structure than are hinged ends. The question of sliding ends for longitudinal girders will be treated in the next chapter, which will deal with clev;iled railroads, the expausi(m i)ocket8 being the same for such structures lis for railroad trestles. The best sections for colunuis are either two channels laced or four Z bars with a web plate or lacing. If the columns have to carry transverse loads, they should have solid webs instead of lacing, so as to transmit the siiear elfectively from top to bottom. For light work, four angles in the form of an I with a single line of hieing will suffice. All columns when spliced should have their splices located about two feet above the panel points of the column bracing. Failure to so locate them will add materially to the cost of erection. All such splices should be made full, more espe- cially when the tension on the column runs high. In proportioning anchorages, the pedestal weight should i)o made not less than twice the greatest net uplift from tlie column, due account being taken of the buoyant effort of the water la case of a possible submergence of pedestal. to t'U- soil III. ital is of (li- tiiu CHAPTER VIII. ELEVATED RAILROADS. '^1 E author has liitely written for the AmeHcun Society of Civil Kugineers a Iciigtliy pnper on this 8ubj»^i:l. It has been very tlioroughly distMissed by the engineering profession, and tlic discussions have been answered in an cxliaustivu resume by tlie author and his assistant engineer, Ira G. Iledriclf, Assoc. M. Am. Sor. C. E. The original paper, the discus- sions, and the resume hav<; been published in the Transactions of the Society for 1897, Vol. XXXVII ; P'ld any one who de- sires to make a special study of the subject of elevated rail- roads will do well to read all that has been published thereon in tlie said Transtictions. There will, however, be given in tliis chapter a compen- dium of tlie contents of tlie paper for the use of those who have no time or inclination to wade through the two hundred pages that it occupies. Live Loads. The proper live load to assume in designing an elevated railroad is the greatest that can ever come upon it, and is de- termined by ascertaining the weights of engines and empty cars that are adopted at the outset, then computing how many passengers can be crowded into the latter and assuming that the average weight per passenger is one hundred and forty pounds. The live loads for elevated railroads, unlike those for surface railroads, do not increase from time to time, but remain constant. In fact the late tendency to operate the roads by electricity rather decretises tnem, for the weight on 91 92 I)E PONTIBUS. lUe axles of a motor car produces smaller bending moments than that ou the axles of a locomotive. After the distribiUioii of the live load on the various axles of the entire train has been determintd, it is well to prepare a diagram of equivalent uniform loads and one of total end shears similar to those for the Compromise Standard System of Live Loads for Railway Bridges given in Chapter XIX, in order to facilitate the computing of stresses and bending moments. Floor. The style of floor in general use on elevated railroads con siHts of timber ties with four lines of timber guard-rails, dosed floors of buckled plate carrying timber ties in baliiist being employed at crossings of important streets and boule- vards, so as to prevent dirt and moisture from falling upon l)eople passing beneath. Such a closed floor has been advo- cated for the entire line, and certainly it would be an improve- ment upon the oi)en floor ; l)ut the increased expense involved is likely to interfere seriously with its adoption for future elevated railroads. The ballast over the buckled i)late in tight floors is necessary to prevent noise from passing trains, whicrli, unless some effective sound-deadener be adopted, would be simply deafening. There is one important inci- dental advantage in employing a closed floor, viz. , that tie elevation of the grade is tliereby rel)ile-girder for open-wt-bbed construction. In many of tlie older elevated roads there is no connecting pliite between the diagonal and the ciiord, l)ut one leg of each of tiie angles in the diagonal is riveted directly to the vertical legs of the chord angles. This detail involves the use of either two or four rivels to the connection, which is evidently very bad designing, as there should be more rivets used, even if the diagonal stresses do not call for more on purely theo- retical consideration.s. Wlieie the theoretical number of rivels is very small, additional rivets should be used for two reasons, viz. : first, one or more of the rivets are liable to be logse, and, second, there is nearly always a torsional moment oa each group of rivets, owing to eccentric coniiectiou. n. FAILURE TO INTERSECT DIAGONALS AND CHORDS OP OPEN-WEBHED GIRDERS ON GRAVITY LINES. It is very seldom indeed that the designer even attempts to intersect at a single point all of the gravity lines of members assembling at an apex. Tlie failure to do so involves large secondary stresses, especially in the heavier members. By using connecting plates, it is always practicable t(» obtain a proper intersection ; and it is always better to do this than to try to compensate for the eccentricity by the use of extra metal for the main members. III. FAILURE TO CONNECT WEB ANGLES TO CHOI DS BY BOTH LEGS. Some standard bridge specitications stipulate tluit in case only one leg of an angle be connected, that leg only shall be counted as acting, although this stipulation is generally ignored by the designer working under such specifications. ELEVATKJJ UAILKOAUS. It is seldom, indeed, tliat bulli Icj^s are cotiiiecled. In uiclur to settle llie question of ihe necessity for this reqiiinjinent, the author has liad madi', in connection with his North weslcrn Elevated work, a series of tests to deslruclion of full-sized meuibers of opcn-webljed girders, attached in the testing nmchiue as nearly as pradicable in the same way as they would be attatliecl in the structure. It was intended to settle by these tests the following points ; rtrst, effect of connccliug by one leg only ; second, effect of eccentric connection ; and, third, the ultimate strength of star struts with tixed ends, each of these struts being formed of two angles. As these tests are not yet finished, their results cannot be given here. The principal deduction to be made from tlie tests thus far com- pleted is tiiat an equal-legged angle riveted by one leg only will develop about 75;* of the strength of the entire net section, while a G' X 3^" angle riveted through the longer leg wi]| develop about 90;*. It is therefore more (conomical lor short diagonals to use unequal-logged angles connected by tlie longer leg than to employ supplementary angles to try to develop liie full strength of the piece. In fact, the ex|)eri- ments made up to date indicate that these supplementary angles will not strengthen the diagonal essentially However, further experiments may show the contrary. IV. FAII-UKE TO I'llorOKTION TOP CHOllDS OF OPEN-WEUBED, LONGITUDINAL OIHDEUS TO KE8IST HFCNDING FUOM WUEEI. LOADS IN AUDITION TO THEIR DIRECT COMPRESSIVE 8TKE88E8. This neglect is common enough iu the older structures, and the fault is a serious one, although the stiffness of the track rails and that of the ties tend to distribute the load and thus reduce the bending. V. INSUFFICIENT BRACING ON CURVES, Too often in the older structures the curveil portions of the line are no better braced than are the straight jxjrtions. A substantial .system of lateral bracing on curves extending over the entire width of the structure and carrii'd well into the tops au iiuportaut one, for it is the lirst part of the metal-wurk l(; 98 DE PONT I BUS. resist the swiiy of trtiins. Not only should the top flanges of adjaceut girders be connected by rigid lateral bracing, bul the bottom flanges should be stayed by occasional cross-l)rHcing frames, one of the latter being invariably used at each ex- pansion end of each track. VII. PIN- CONNECTED, P()NY-TRIJ88 PPAN8 AND PLATE GIRDERS WITH UN9TIFFENKD TOP FLANGES. These defective constructions are noticeable in some of the older lines, but, fortunately, n(>t often in the newer. What the ultimate resislaiuc of the pony-truss structure is no man can tell without ti-sling it to destruction ; but, in the opinion of most engineers, it is nuicli less than it is assumed to be by those designing pony-truss bridges. VIII. KXCESS OF KXPANSION .TOINTH. Too many expansion joints in an elevated railroad are nearly as bad as too few. In the former cast; the melal is ovenstrained by the vibration induced by the lack of rigidil}', while in tiie latter case it is overstrained by extreme variations of tem- perature. There are elevated roads in existence with expan- sion joints at every other bent, and there is at least one with them at every bent. For long spans there slnndd be expan- sion provided at every third bent, and for short spans al every fourtli bent. IX. RESTING LONGITUDINAL OIUDKUS ON TOP OF CROSS- GIRDKRS WITHOUT RIVETING THKM EFFECTIVELY THERETO. This is l)y no means an uncommon detail, especially in the older structures. It is conducive to vibration, and its only advantages are ease of erection and a cheapening of the work by avoiding tield-riveting. X. CROSS-GIUDERS SUBJECTED TO HORIZONTAL BENDING THRUST OF TRAINS. BT The resistance that can be offered by a cross-girder to hori- zontal bending is very small ; nevertheless, cio-ss-girders are ELKVATED RAILROADS. 9D rarely protected from the bending effects of thrusts of trains. Wlial saves these cross-girders from faihire is the fact that coniiimity of tlie Iraclc tends to distribute the thrust over a II umber of bents. Nevertheless, it is not legitimate to depend on this fact, lor, especially on sharp curves, the tendency is to carry the thrust into the ground as dinctly as possil)le. in the author's opinion, tiie only propir way to provide for this thrust is to assume that 20% of the greatest live load between Iwoad- jacHiit expansion points will act as a horizontal thrust upon ihc columns betwein these Iwoexpansion points ; and all parts of tht! metal-work should be proportioned to resist this thrust properly. liy running a strut from the top of each post diagoiuilly to th(! longitudinal girder at a panel point of its sway-bracing, the horizontal thrust is carried directly to the post, and a lior- izont^l bending moment on the cross-girder is thus prevented. Such (construction should invariably be used where the condi- L'ons require it. XI. CUTTING OFF COLU.MNR HKT-OW THK BOTTOM OF CKOS8- (JIUDKUS AND KESTINd THK LATTRK THEHKON. This style of construction, which until lately was almost universal, is extremely faulty in that there is no rigidity iu the connection, and the culumu is thus made more or less free- ended al tlie top. It has been said that no harm is done to the colunui by making it free ended, as it can then spring belter when the thrust is applied. Unfortunately this reas')ning is fallacious, because the few unlucky rivets which connect the bottom of the crosH.girder to the lop of the column lend to produce a fixed end, and arc, in consequence, racked excessively by the thrust of the train. In all cases the column should extend to the top of the cross-girder, and should be riveted to it iu the most effective manner practicable. XII. PALTRY BRACKETS CONNECTINO CROSS-GIRDERS TO COLUMNS. Brackets are often seen compo.sed of a couple of little angles attached at their ends by two or three rivets. Such brackets are merely an aggravation, and are sure to work loose sooner or later. Althougli it is impracticable to compute the stresses in this detail, good judgment will dictate the use of solid- webbed brackets riveted ri^gidly to both cross jjirder and CQJ- 100 DE P0NTIBU8. umii so m to stiffen the latter and check the transverse vibra- tion from pnssiug trains. XIU. PROPOKTIONING COLUMNS FOU DIRECT LIVE AND DEAD LOADH AND IGNORING THE EFFECTS OF BENDING CAUSED BY THRUST OF TRAINS AND LATEUAL VIBRA- TION. The practical effects iirpose. The span, with all that it carries, weiglis about 290 tons, and the counterweights weigh, as nearly as may be, the same. As the cables ami their counterbalancing chains weigh fullj' 20 110 DE PONTIBUS. tons, Ihc total weight of the moving mass is almost exactly 600 tons. Should tiie span and the counterweights hecome out of bal- ance on account of a greater or less amount of moisture, snow, dirt, etc., in and on the jtavement andsitlewallvS, it can be ad- justed by letting water into and out of hnllast-tanks located lieneiith the floor, and, should tlds ndjustinent be in.'iurticient, provision is made i'or adding small weights to the counter- weiglits, or for placing such weigiits on the span. As the counterweights tints l)alancc the weight of the span, all the work whicli the machinery has to do is to overcome tlie friction, bend the wire ropes, and raise or lower any small un- balanced load that there may be. It lias been designed, liow- ever, to lift a considerable load of passengers in case of neces- sity, although tlie structure is not intendcnl for this purpose, and should never be so used to any great e.tlent. The span is steadied while in motion by rollers at tlie lops ami bottoms of the trus.sos. There are bolli Iransver-se and longitudinal rollers, the former not touching the columns, un- less there is sufficient wiiul-pressure to bring them to a be.nr- ing. The longitudinal rollers, though, areatla(;h(.'d to spri;igs, winch press them against the columns at all times, and take up the expansion and contraction of the trusses. With the rollers removed, the briilge swings free of the columns ; and, since the attachments are purposely made witak. the result of a vessel's striking the bridge with its hull will Ik; to tear tlieni away and swing the span to one side. Should the rigging of the vessel, however, strike the span, theelleet will be simply to break off the masts without injury to the bridge. This lat- ter accident Inis happened once Jilrendy, the result being ex- actly what the author had prcdi<;ted. There is a spe(;ial np- ])aratus, consisting of a heavy scpnire timber set on edge, trimmed on the rear to lit into a ,«(c(l < hannel wliich rivets to tlie cantilever-brackels of the sidewalk, and fa<;ed with a 6 X 0-in. heavy angle-iron, to act as a cutting edge. This detail is a very ellecliveone for destroying the masts and rig- ging of colliding vessels. Tlie bridge is designed to carry a double-track street rail way, vehicles, and foot-passengers. It has a clear roadway of 34 ft. between the counterweight guides in tin; ti)wers, the narrowest part of the structure, and two cantilevered side walks, each 7 ft. in tlie clear, the distance between central planes of trusses being 40 ft., and the e.xtreme width of sus- pended span 57 ft., except at the end panels, where it is increased gradually to (};{ ft. The roadway is covered with a wooden block pavement 34 ft. wide between gUiiidiails rest- ing on a 4-in. pine floor, that in turn is suppoiKMl iiy wooden «luiU8 which are bolted to 15-iu. I-beam stringers, spatted MOVAIM.Ii HUIDGKS IN GENERAL. HI about 3 ft. 3 in. from cciiire to centre. These stringers rivet up to the webs of tlic tioor-beams, und Ijcueuth tbein run {liiigonal angles, wliich rivet to the bottom tlange of each stringer, and thus form a very eflicient lower lateral system. 'I'lie sidewalks are covered willi 2-in. jnne planks, resting on 3 X l!i-in. pine joists spaced about 2 ft. from centre to centre. Tlie span is susjx inled at each of the four upper corners of the trusses by eight steel caldcs, which lake bold of a pin by means of cast -steel clamps. Tliis pin passes througii two liangerplates whicii jiroject abovr the truss, and are riveted veiy elloctively to tiie end post by means of the portal plale- ginler strut e purpose at its highest position. In addition to the butlers there will be automatic, electric cut-offs to remove the power before the deck reaches either end of its travel, besides powerful brakes to bring the moving mass to rest (juickly whenever the operator nuiy so desire, atid always automatically at the higheftt and lowest points of travel, in order to relieve the bulTers. The main sheaves are five feet in diameter and five inches wide, with eight radial arms. They are each cast in one piece and keyed to a seven-inch steel axle, that rests on two pillow-blocks each eight inches long, fitted with bronze bear- ings. The pillow-blocks rest on short posts riveted into long transverse girders that rest on the top chords and cantilever out beyond them about tivn feet at each end. These posts are to be well braced longiludinally. The supporting detail between the transverse girders and the top chords is such as to distribtite the load properly over the latter. There will be at each of the four corners of the moving deck two rollers for transverse motion and two for longitudinal motion, all acting on the faces of the columns that uphold the supporting span. The transverse rollers do not act unless there be sufficient wind-pressure ou the deck to move it laterally ; but the longitudinal rollers act whenever the deek MOVAMLK HKIDGES IN ORNEKAL. 117 is moved, iis they are backed by springs that jiress ihein at all times ngainist the columns. When the dock is at its lowest position it will be held firmly to the piers, with n proper provision for longitudinal expansion, in such a manner as to relieve entirely the guide- rollers from carrying the wind-pressure, so that they can act only when the deck is raised. The machinery-houso will l)e about twenty-two feet square and fourteen feet high under the eaves, capped by a dome, and finished in an ornamental style. The door is to be of I beams supporting a four-incli plank floor. All main sheaves are to be covered with ornamental hous- ings, and all gears are to be covered with small galvani/.ed- iron hinged housings. The velocity of the lifting deck will be limited to one foot per second by means of an autoniiilic governor attached to the electrical machinery. The lime required lo either raise or lower the deck the full height will therefore be about one minute. To provide for a possible breakdown of the electrical machinery, a man-power apparatus will be employed, con- sisting of two capstans connected lo the main shaft by means of gearing located in llie m.nciiineryhouse, and operated by levers working in horizontal pianos. The moving deck and counterweights will be balanced when the deck is at midlieight. On this account there will be a constant tendency to hold the deck from vertical motion at both ends of its travel, because of the unbalanced weight of the wire cables. In one sens(! this will be a decided advan- tage, but it will necessitate extra power to start the mass in motion. Again, the deck will be balanced for ordinary con- ditions of weatlier, but il is probable that the weight will be increased by moisture, accumulated dirt, etc. This, if it exist to a moderate extent, will be an advantage, in that it will tend to hold down the deck on the piers ; but, as before, it will require increased power to start motion and to operate. However, the amount of power available will be large enough to meet all conditions of loading and contingencies. Should 118 r>K ro NTT rims. tlie deck become lighter tlmn the counterwelphts by reason of the drying of tiie timber in tiie floor uad .screens, it will 1)0 necessary to ndd to its weight by louding it > but this condi- tion is not likely to exist, for what wciglit is lost by drying will be fully nmde tip by ftcciimnliited dirt in spite of all the l)r('C!»utions that may be taken to keep the Hoors clean. Wliether this jiroposod .structure will ever be built is pro >- lemalical, nltliough there is a fair chance of its being finished some day with modifications tending to ciieapen the work. It wojld be a great satisfaction to the author to complete this bridge because of the novel design for the lifting deck. Floating draws are a type of structure that cannot be recom- mended except as a temporary expedient. Tlie author liad occasion once to design one of them, but the necessity for its use did not develop, so it was not built. The ol)jection8 to lloating draws are as follows : 1. 'I'rouble from rise and full of water, necessiialinir ((hi- slaiit adjustments. 2. The depression of tlie dniw under the live load and llio consequent changing of the grmle. 8. Possible disaster from injury by ice or drift. 4. Trouble froiu leakage. 5. Clumsiness o* method of opening and closing the draw. As there an; nn advantages to oflf.set these disiulvantages, unle.ss it be possibly a small .saving in first cost of span, it is not likely thai there will be nuich call for tioating draws. In concluding this chapter, it may be well to summarize somewhat and indicate -Khnl kinds of draws should be used at various crossings. For streams bearing a moderate amoiml of tnirtic with cross- ings located in country districts or in unimportant cities, rotating draws are tlie cheapest and consecjuently the most appropriate ; but for great trutttc and for important (titles bascule and lift bridges are the best, the former for spans up to about one liundred and fifty feet, and the latter for longer spans. The choice between the bascule and the lift for all doubtful cases should be determined simply by the question of first cost. CHAPTER X. RKVOLVIN'O DUAWnUIDOEg. Revolvino (Iniw-spans nrc required when bridges across imvignble Ktreaiiis are not high enough above tlie water to provide tiie proper vertical clearance for passing vessels. Hefore taking up the discussion of draw-spans, it will be well to consider the relative advantages and disadvantages of high and low bridges for the crossing of such streams as the Mis- sissippi, the Missouri, and the Arkansas rivers. As r rule, there is very little difference in the first cost of a high and of a low bridge for such a crossing, what little there is being in favor of llie latter and seldom amounting to more than ten per cent. lOacli pier of a low bridge is (cheaper than the corresponding pii,;r of a high bridge ; but this saving is offset by the co«t of the pivot-pier, which is extra. The superstructure of a low biidge may be a trifle lighter than that of the corresponding higii bridge, but the more expensive metiil-work of the draw-span generally over- balances this. It is in the low, short trestle approaches that the lov, bridge costs less than the high one. As these approaches are generally built of timber, they have to be renewed about once in every eight years, and the cost of renewal is a regular fi.Ked charge, which lessens the annual net income from the bridge. Herein lies the superiority of the low bridge for such cross- ings. Nor is this its only advantage, for, by its adoption, there is generally avoided a considerable climb at each end of the structure. On the other hand, the low bridge involves some expense for operation, which is (piite an important matter when there is much river traffic, but which is of slight importance when 119 120 DE POi»fTrBUS. the draw lias to be opened only a few times per season, us is the case witli bridges over most Western navigable streams. Everytbiug considered, wbenever tbere is any cboice be- tween a bigh and a low bridge for tbe crossing of any impor- tant Western river, tbe aulbor favors tbe low bridge, not so irucb because of its lower first cost, but on account of tbe smaller expense for maintenance. Tbe different kinds of revolving draw-spans recommended are described in detail in Cbapters XV ami XVII. Tbey may be operated in various ways, for instance by manpower, steam, electricity, gas or gasoline engines, or water. Wber- ever an unfailing supply of electriciiy is available, tbat source of power is tbe best and cbea[)est. Steam is appropriate for large, beavy draws wbcre electricity is not available. Gas or gasoline engines are best suited for comparatively small spans in country districts ; and water-power can sometimes be employed to advantage where tbere is a fall of wat(!r near tbe biidge. It does not pay to use storage batteries for operating draw- bridges. Concerning tin's questi(m tbe author feels that he can speak as an aulliorily, for he once made tbe experiment, and it was a failure. For a wliile the machinery worked to perfection, but soon tbe batteries began to leak, and the leak- age gradually increased to such an extent that the batteries would not hold their charge for three consecutive days ; so the electrical power was given up, and the bridge has since been o^ erated by baud. Gasoline engines, everything considered, are probably the best source of power for operating the average draw-span. Tbe author has lately designed some small draws to be oper- ated therel)y; but the maciiinery has not yet been installed, so be cannot report concerning bow such engines act. In respect to the power required to operate draw-spans, the author uses an average of the Boiler formulie, viz., H. P. = 0.0125 TV t> 550 ' where W= total load on rollers in pounds, and v = velocity on pitch-circle of rack in feet per second. RKVOI-VINO nRAWBUIDGES. 121 The author obtained a tine cliecli on tlie correctness of lliis formula wheu testing tiic draw-span of liis Jefferson City highway bridge. This span of 440' weigiis 660,000 pounds, and was opened by four jnen in four minutes and tifty sec- onds, Tiie power applied by the men was measuied by dy- namometers, and from the length of their pith and from their pull the horse power was coni[)uted. It proved to be just a little less than unity, so near in fact that it was called unity. Tlie velocity v was, on the average, 0.066 feet per second. Substituting in the formula gives H. P. = 0.0125 X 660,000 X 0.066 ^ 550 = 0.99. It is possible that, if the experiments were to be made again, a greater divergence from the formula would be found, for the reason that the bridge is liable to work more easily after it has been operated a while. The computation of stresses in ordinary diaw-spans involves more or less ambiguity. The assumptions upon which the calculations arc biised are the following : !. The truss-rods in the tower are so light that they cannot transfer any shear pa^^t the pivot-pier; consecjuently, witii a live load on one arm oidy, the said Jirm acts entirely independ- ently of tlie other, thus making tlx- draw for tiiis loading con- sist of two simple spans. 2. For live load on both arms, the reactions are to be found on the a.ssumption that the draw is a continuous girder on four points of support, and by a formula based upon the 'I'lieorem of the Three Moments with a constant monient of inertia. Plate IX gives a diagram from which can be read at a glance the percentage values of the reactions for any balanced load plnany, and inspected by first-class inspectors during both manufacture and erection, they would be able to relieve themselves from the responsibility ; but if they were to do all this no bridge failure (;ould occur, or at least the chances for such occurrence woidd be extremely small. Some such method as Ibis for placing the responsibility for bridge dis- asters upon county commissioners will l)e established by law some day, perhaps in the not very distant future ; and the sooner the belter. As matters stand now, each new bridge horror stirs up the indignation of the populace, which vows thai this time the guilty parties shall be brought to punisliment ; but the inves- tigation generally drags, personal intluence is broii^^ t to bear, njoney is often used judiciously, and the result is that nobody is held responsible, and the disaster is soon forgollen. If each stale were to adopt standard speci(ic;itioiis for high- way bridges, and if there were a proper officer appointed to see thai the counties live up lo them, much good would be accom|)lished. The second edition of the author's pamplilet on highway briilgos is now exhausted, but no third edition will be issued, for the reason that le jen n'en nmt pan la chandelle. There will be given, however^ in Chapters XVI, XVII, and XVIIJ 13S I)E I'ONTIBUS, of this book complete 8|)ociti(:alioti8 for the (Icsigning of iiigb- •way bridges of nil isiuds ; consequeiitly this treatise may be said to replace the pamphlet that is now out of print. There is considerable difference, though, belweeu the old speciticiitions and the new, due to two reasons, viz., first, there have been great advances made in bridge-building in the last eight years ; and, second, the author .las concluded to abandon the attempt to conciliate ,hose who desire to build cheap structures, so has cut out Class I> from his spccitlca- tions, and iias strengthened up and imi)roved the other "classes" in several particulars, notably by raising the mini- mum thickness of metal from one (quarter to tive 'xteeulhs of an inch. The weights of bridges designed according new specifications will be somewhat greater than those designed according to the old ones, but the structures will be corre- spoiulingly better. Moreover, the new s|)ccilications will be found to be more rational, scientific, and generally satisfac- tory than the old ones, especially in the feature of impact allowance. It must be remembered that the author does not claim that the formula which he specifies for impact will pro- vide exactly for the greatest possible impact on all parts of all highway bridges; but he does think that it will always be great enough, and he knows that any structure in the design of which it is used, and which is proportioned by the 8|)ecifi- cations of this treatise, -will be well and properly designed in every part. CHAPTER xn. COMBINED BRIDGES, As a rulo, bridges for eiinying Imth rallwny and highway truffle are located in or near larg< ities, allhough an occa- sional structure of this kind is found in country districts. Tlie princii)al advantage of this type of bridge is the saving in flrst cost, and its principal disadvantage is a reluctance to cross over it on the part of timid drivers, whose horses may be frightened by the trains. The saving in first cost of a combined railway and hfgliway bridge as ccmipared ^\ith two separate bridges for railway and higiiway traffic is considerable , because the piers for the combined bridge are but little, if any, more expensive than those for the railway bridge, and because the extra metal for the superstructure of the former in comparison with that of the latter is v"ry much less in weight than the weight of metal required for a separate high i 'ay bridge. The prejudice against combined bridges on account of danger is almost wholly unfounded, for horses soon become accustomed to railway trains, and, when screens are em- ployed to hide the hitter, but little trouble is experienced on account of frightened horses. These screens may be made either slatted or close, the former offering less resistance to the wind, and the latter being the cheaper. The advent of the electric railway has somewhat compli- cated the question of designing combined bridges, for now it is often necessary to accommodate three or four kinds of traffic, viz., railway, electric, wagon, and pedestrian. When a highway structure has to carry a single-track elec- tric line in addition to the ordinary highway travel, the 188 154 DE fONTlBt'S. author classes it simply as u highway bridge, for it is seldom necessary to strengthen it inaterially, because of the electric- railway load, except in the floor and primary trass nienibcrs. But if a structure lias to cany either a single or double truck electric railroad only, the author treats it as a railroad bridge. Combined bridges may be divided into tlie following classes : 1. Structures having a single deck for all kinds of tratlic, the railway occupying the centre of the bridge, and the elec- tric railway lying close to one truss. 2. Structures having a single-track railway at the middle, a narrow footwalk on each side of same inside of the trusses, and cantilever brackets outside of the latter to carr}' wagon- ways and electric lines. Tiiis arrangement may be varied by running the electric cars over the main railway track, tlnis leaving the wings free for wagou traffic. 3. Structures liaving a double-track railway inside of the trusses, with long cantilever brackets outside carrying wjigons and electric lines next to the trusses, and pedestrians outside. This arrangement may be varied, as in Case 2, by carrying the electric trains on either one or both o' the main raihviiy tracks. 4. Structures having a double-track railway inside of the trus.scs, with short, cantilever brackets for wagon and electric- railway traffic outside, and either a single passagewjiy over- head at the middle for pedestrians, or two passageways for same on overhead brackets outside of the trusses. As before, this arrangement may be modified by running the electric trains over the main railway tracks. 5. Double-deck structures carrying railway trains on one deck and wagons, electric trains, and pedestrians on the otlier. The pedestrians may be accommodated either inside tiie trus-ses or, preferably, by exterior walks on cantilever brack- ets. The railway may be placed either above or below to suit the existing conditions, or the.ro mny be cither a single- track or a double-track railway both above and below, with wagon-ways and pcdestrian-wavs outside of the trusses cither above or below. Class No. 1 is the cheapest possible kind of combined bridge. COMBINED BRIDGES. 135 and at the same lime ihe most unsatisfactory, for wlieu h railroad train is about to pass over the bridge ail wagon and electric-railway tnivel must be kept off, and because pedes- trians must look oKt sharply for their safety when ou the structure with a railwa}' train crossing. Their danger is really greater, though, when jin (tlcclric train is passing a team or teams. The least allowable clear width of bridge for this class of structure is twenty feet, the electric cars running on a third rail and on one of the rails of the main railway. The author has built a large bridge of this class, and it has never given any trouble from the combined tralHc, which, however, is, up to the present, rather light. (.Mass No. 2 is a very satisfactory type of structure. The atithor has designed and built several bridges of this kiiul, the largest of which is 'he (combination Bridge ('ompany's bridge over the Mi«so\iri River at Sioux City, Iowa. It consists of two draw-spans of 470 feet each and two fixed spans of 500 feet each, the distance between central planes of trusses being twenty live feet. Class No. 3 is also a satisfactory type of structure. The author has built a large bridge of this class, viz., the one across the Missouri River at East Omalia, Nebraska. This class of stru(;ture involves very heavy metal-work ; but it is not une(!onomical. Class No. 4 is an unusiial type, and is not likely to be called for very often, although the author once had o(;casion to figure on a bridge of this kind. Class No. 5 is a very good comi»ination that can be modified to suit nearly any conditions of » ombincd trallic. A good example of this is the design described in Chapter IX for the Kansas City and Atlantic Railway C-'ompany's proposed bridge over th(; Missouri liiver at Kansas City, Mo. The East Omaha Bridge just referred to ailords an excellent exaniple of how to keep down tiie first cost of a structure and yet build it .so that it can be enlarged later after the bu-niness devel()i)s. The design for th(; linal structure involves a draw- spun of MO feet and a fixed spun of 5(50 feet, carrying ii double track railway between the trusses, a Mubined wagonwuyaud 136 DE PONTIBUS. electric railway outside of facU truss, aud a pedestrian-Way outside of eucli wagou-way, the bridge crossing the river at right angles ; while tlie present structure consists of the 520- foot draw-span, without the wings, aud three single-track combination spsins of 192 feet each, all the piers except the pivot-pier being built of piles and timber, and the centre line of structure making an angle of eleven degrees with the centre line of the final bridge. The deck carries a single railway track ill the middle and an electric line by means of a third rail to one side. All four classes of travel U3e this deck. The only portio;! of the existing structure that is really finished is tlie pivol-pic", which consists of a double steel cylinder forty feet in diameter sunk by open dredging to bed-rock, wliicli lies one hundred and twenty-two feet below extreme low water. The completion of the draw-span will be a very simple matter, consisting merely of adding the cantilever brackets with tlieir stringers and flooring and laying the electric-railway rails thereoti. The remaining jners for the final strucliue can all be put in, and the fixed span can then be placed on them without interrupting trattic, because of the deflection downstream of the present temporary structure. When the new biidge is completed, all that it will be necessary to do is to rotate the draw eleven degrees, so that the traflic iiuiy be Iransferrod thereto. Afterwards the old pile piers and tje combination spans can be removed at pleasure. In designing combined bridges of all clas.ses except No. 1, a considerable economy of metal may be elfectei' egitimately by keeping tiie total live load as low as is proper with reference to the theory of probabilities. For instance, in Class No. 2 the live load for trusses can be determined by adding to the equivalent uniform live load, given in the diagram ou Plate III or Plate IV, a much smaller highway floor load per lineal foot of span than that prescribed in the specificationg for highway bridges, because when the greatest train load is ou the bridge, the chances of having a heavy highway live load are very small. The longer the span the smaller may the live load per scpiare foot of floor be taken when finding the total live load for the trusses. t!i tl li bi til hi n( m fo li COMBINED l{ HI DOES. Vdl Again, in Cliisses No. 3 and No. 4 it would be legitimate to take liie live load per lineal foot for the raihva}' equal to twice the car load per liueal foot, and add thereto a small highway live load as in the last case. Finally, in Class No, 5 it would be proper for a four-track bridge to make the live load for the tru.s.ses equal to four times the car load per lineal foot, and ignore entirely the liighway live load ; for the greatest combined live load would never amount to four times the tar load. This was the metliod pursued by the author in determining tiie live load for the trusses of I lie proposed Kansas City and Atlantic Hallway Company's bridge referred to iu Chapter IX. CHAPTER XITI. DETAILING, It is ouly within h few years that much attention has been given to detailing by bridge engineers, tlie old custom having been for the engineer to figure the diagram of stresses, or, as it was llien called, tlie strain-sheet, and pass it over to a drafts- man (too often a cheap one) to make tlierefrom the working drawings of the bridge, using probably some old drawings of another l)ri(lge as a guide for the detailing. Concerning tlie evil eflfects of such a course of action the engineer who docs much inspeci ion of existing structurescan speak authoritatively. If questioned upon the subject, any such engineer will .say that nearly all l)ridges which fail or which are condemned and removed, are deficient in strength of details rather than in strength of main members. Some ytniva ai;'> the author had t)crasion to examine and report upon nearly all the bridges on two hundred miles of the main line of an iin|)ortant Western road, with the rcsull that he found it necessary to condemn almost all of them. A few have since been repaired, l)ut most of theni have been taken out and replaced. In mo.st of these condemned structures the detailing was so faulty that the bridges were gradually racking to pieces, and no amount of patchwork wotild have made them really serviceiblc. It is true tluit the main members were considerably overstrained by reason of the increase in rolling load.s, but had the details be<'n first-class the structures would have been standing today. Jiist here it may be w U to mention that the inspection of these biidges caused the author to establish for himself the 188 r)ETAILII^G. 139 following priuciple. which, ns it does not pertain to bridge- designing, is not given in Clmpter II : " lu nine cases out of ten tlie proper wuy to strengthen a weak bridge is to take it out and replace it with a good one, throwing the old metal into the scrap heap." In railway-l)ri(lge designing, for a number of years, the average ratio of weiglit of details to weight of main members has been gradually increasing; and the end is not yet, because the average l)ridgedeHigner has still a great deal to learn con- ceridng the importance of good and efflcient detailing. As long as contracts for bridges are awarded to bridge companies on competitive designs, anti the structures are paid for by tlie lump sum instead of by the pound, just so long will the science of detailing be ignored, and just so long will bridges be built which will eventually wear out, simply for want of a little more metal distributed just where it is needed, viz., in the details. The author feels that lie cannot speak too forcibly concern- ing the importance of thoroughly scientitic detailing for all kinds of metal-work; for what avails it that a st'-ucture have an e.\(!ess of section in every main member, if a single important detail be lacking in strength? If the author were in a position where he had to cut down the weight of a structure even as much as thirty y.'r cent, he would unhesitatingly take the metal almost entirely out of tlie sections of the main members and leave the detailing practically unchanged. A structure thus designed would long outlast one of the same type in which the weiglit of the details and that of the main members were reduced in the same proportion. A few years ago the standard text books on bridges ignored entirely the subject of detailing. Later they have taken cog- nizance of it by illustrating certain details in common use, both good and bad (generally the latter), but have failed to state the fundamental principles that sho(dd govern the designing of all details. These general underlying principles and complete instructions as to how to detail scientifically the author has endeavored to give in Chapters II, XIV, XV, XVI, and XVII of this wo- k. The bridge designer, by studying these chapters 140 DE PONTIBUS. carefully, mastering all of their contents, and, while making his drawings, applying the principles therein given, will be able to evolve structures that, to say the least, will be a great improvement on the average structure in common use. CHAPTER XIV. GENERAL SPECIFICATIONS GOVERNING THE DESIGNING OF STEEL RAILROAD BRIIKIFS AND VIADUCTS AND THE SUPERSTRUCTURE OF ELEVATED RAILROADS. GEXEBAL DESCBIPTIOi^ MATEniALS. All parts of the structure, except ties, foot-plauks, and guard-timbers, sliall, for all spans of ordinary lengtlis, be of mediuin steel, excepting only that rivets and bolts are to be of soft steel, and adjustable members of either soft steel or wrought iron. For very long spans high steel may be used for top chords, inclined end posts, pins, eye-baia in bottom chords, and those in mai!> dingonals of panels where there is no reversion of stress when impa::t is included. It may be used also for the web-members of cantilever and anchor arms in cantilever bridges where the variation of stress is com- paratively small and where the impact cannot be great. E.\- cepting for purely ornamental work, cast iron will not be allowed to be used in the siiperstructure of any bridge, trestle, or elevated railroad, cast steel being employed wherever important castings are necessary. CU08H-TIE8, FOOT-PLANKS, AND GUARD-TIMBERS. Cross-ties, foot-planks, and guard-timbers shall be of long- leaf. Southern yellow pine or other timber whicli, in the opinion of the Engineer, is equally good and serviceable. The wooden tloor shall be so designed as to ensure safety from passing trains for the ndlroad employees. The spaces between ties sliall, in general, not be less than five (^)) inches nor more than six (6) inchps wide. The sizes pf ties shall be such 09 lit 142 DE PON TI BUS. to give the requisite resistftnce to bending, under tlie assump- tion thai the load on one pair of wheels is distributed equally over three ties, the cfTect of impact being considered. No tie shall be less than seven (7) or preferably eight (8) inches wide, nor less th!U\ six (6) inrlies deep, nor less than ten (10) feet long, e.\ee])t in the ea^eof elevated railroads, wlu're llie Icnglli may be rediieed to eignt (8) feet for a spacing of live (5) feel between central planes of longiludinal gliders. Ties shall bedapjied to a full and even Ixaring not less than one-half (A) inch onto Ihe stringers ; and each aherna'e tic; siiali be secured thereto at each end by a three-quailer (Jj inch hook bolt. All timber bolts shall be of soft steel, with cold-pressed threads. Outer guard-timbers shall be 6 ' X H" laid on tiat, dapped one (1) inch onto the ties, and placed so thtit their inner faces shall be just twelve (13) inches from the gauge planes of rails. Wiiere inner guard-timbers are em])loyed, they shall be 6" X 8" on flat, dapped one (1) inch onto the ties, and place.l so that their outi ^aces shall be just five (5) inches from the gauge-planes of rails. Each guard-rail must be bolted to each alternate tie by a three-quarter (J) inch .screw-bolt, the head of whieii shall be countersunk into the wood by means of a eup-siiaped washer. Each guard- limber nui.st be spliced over a tie with a half-.-ind- lialf joint of at least si.\ (0) inches lap, througii which must pass a three-quarter (|) inch screw-bolt. Guard-timbers shall e.\tend over all piers and abutments. Steel rails or heavy steel angles well fastened to the ties may be substituted for the inner wooden guard-rails, or the inner guards may be omitted allogelher if the Engineer so direct. IlEUAIMNQ ArrAHATUS. At each end of every bridge or trestle, there is to be placed a rerailing apparatus that will, in the n)osl etlective manner practicable, retiirn to the track any derailed car or locomotive that is not more than half the wjdth of track gauge out of line. GENERAL DESCRIPTION. 143 BUCKLED-PLATE FLOOHS. If the Engineer so desire, a bucklcd-plute floor with ties in biillast may be used iusteiid of the wooileu floor, in whieh case llie size of the ties may be reduced to 6" X 8" X 8'. All buckled plate floors must be thoroughly diaiued so as not to retain water, and the upjier surface of the buckled plate nuist be protected from rusling by u liberal use of the best obtainable preservative coating. SUVEUELEVATION ON CUKVES. On curves the outer rail will be elevated the proper amount for the degree of curvature and for the assumeil medium velocity of tnuns ; and this elevation must be framed into ties, as no shims will be allowable anywhere under ties or rails, ex- cepting ill the case of very sharp curves requiring a superele- vation exceeding three (S) inches in five (5) feet, on wliich long shimming timbers are to be bolted to the top flanges of the outer longitudinal girders, or short, substantial ones to tops of ties, so as to give the reciuired superelevation. The formula to be used for total superelevation on standard- gauge roads is 0.3277 7'^ E= U where E is the total snpereh,'vation in feet of the exterior rail above the interior rail, V is tiie assumed velocity of train in miles per hour, and Ji is the radius of the curve in feet. The total superelevation is to be obtained by depressing the inner rail and elevating the outer one equal amounts, thus preserving the grade of the centre line. SPACING OF STIUNGEU8, GIRUKRS, AND TllACKS. In general, stringers for through bridges shall be spaced eight (8) feet centres for single track bridges and six (6) feet six (6) inches for double-track bridges and half-through plate- girder bridges. In elevated railroads the spacing of the longi- tudinal girders may be made as small as five (5) feet centres. Peck plate-girders may be spaced from six (6) feet to ten (10) 144 DE PONTIKUS. feet centres, the usual distiince being the nearest even foot to one tenth (jio) of the span; but in high trestles the spacing shall, preferably, be ten (10) feet, and never less than eight (8) feet. The standard distance between centres of Imcks on tangent for surface railroads shall be thirteen (13) feet, while for ele- vnled railroads it shall generally be twelve (12) feet. gPACINO OF TRUSSES. From centre to centre of through trusses \hv. perpendicular distance shall not bo less than seventeen (17) feet, or one twentielli (gV) of the span length. From centre to centre of deck, i>in-connected, or riveled tnisses the perpendicular distance shall not be less than leu (10) feet or one thir- teenth ( i\) of the span |<^ i ^ length, except in the case of elevated rail- roads, where open- webbed, riveled gir- ders are adopted. These may be spaced according to the direc- tions given for plate girders. CLEARANCES. The clear opening on tangent shall not be less than that shown in Fig. 7. On curved track, the horizontal distance from the centre of track to clearance line shall be increased at all points two (2) inches for each degree of curvature. BASE or RAIL Fio. 7. GENERAL DESCKU'TIOX. 145 EFFKCTIVB LENGTHS. Effective lengths shall be as follows : For pin-coiinected spans, the effective length shall be the distance between centres of end-pins of trusses. For riveted girders, it shall be the distance between centres of bearing-plates. For stringers, it shall be the distance between centres of cross-girder webs. For cross-girders, it shall be the perpendicular distance be- tween central planes of trusses. For columns and posts, it shall be tlic greatest length be- tween points of axis that are rigidly held in the direction in wliich the strength is being considered. These elfeclive lengths are to be used in calculating mo- meuts, stresses, and working strengths. EFFECTIVE DEPTHS. >%.' Effective deptlis shall be as follows : For pin-connected trusses, the perpendicular distance be- tween gravity lines of chords, which lines must pass through centres of pins. For plate-girders and open-webbed riveted girders, the per- pendicular distance between centrelines of gravity of upper and lower flanges; but never to exceed the depth from out to out of flange angles. STYLES OF BUIDGES FOU VAIUOUS SPAN LENGTHS. For spans under fifteen (15) feet, rolled I beams. Fc" spans between fifteen (15) feet and eighty-five (85) feet, plate girders. For sjmns between eighty-five (85) feet and one hundred and twenty-five (125) feet, "A" truss, pin-connected spans, or riv- eted, open-webbed girders of single cancellation. For spans between one hundred and twenty-five (125) feet and one hundred and seventy-five (175) feet, riveted, open- webbed girders of single cancellation, or pin-connected trusses 14« ])K i'()NTini;s. tles'gnc'd willi speciul reforcuce to cxlieino rij^idity in till parts. For spans cxcei'diiig one huiulicd und .sevenly-tive (175) feet, piii-conne(;tt'd spans. The use of pony truss bridges of any iiind is proliibited, excepting only halfllirougli, plute-giider spans, in whicii ilus top llunges me held rigidly in phue by bmcliets riveted to cross-gilders tliat are spaei d generally not to exceed liflcen (15) feet apart. In general, double-trick I)ridges shall have only two lrus.se.s, iu order to avoid spreading the tracks. KOIl.MS OK TKi;SHKrt. The forms of trusses to be used are as follows : For pin-connected spans up to one hundred and twenty - five (12.1) feel, the "A" truss. For opeu-vvel)bed, riveted girders, tlie Warren or tri.'ingulnr girder with verticals dividing the panels of the top chords; also the Pratt truss. For deck-spans having top chords supporting wooden ties, the Warren or triangular girder with verticals dividing the panels of the top chords. For spans between oni; hundred and twenty five (l2o) fet!t Hud about two luindred and fifty (250) feet, Pratt trusses with top chords either straight or polygonal, For spans exceeding two hundred and fifty (250) feet, Petit trusses. It is understood that these linnting lengtiis are not fixed ab- solutely, as the best limits will vary somewhat with the num- ber of tracks and weight of trains. MAIN MEMBEllS OK TKU8S-BUID0K8. All spans of every kind >hall have end floor-beams, riveted rigidly to the trusses or girders, for supporting the stringers. Stringers are to be riveted to the webs of the cross-girders. Iu general, all trusses shall have main end posts inclined. All trusses shall be so designed as to admit of Jicr-urate GKNKUA'. DESCllirTION. 147 ciilciilatioii uf nil ulrcsscs, excopting only sucli iiniiiipurlant cases of ambiguity us that iiivo'vud by using two Htiff diag- onals in a middle panel. All liiteral bracing and other sway-bracing siiall be rigid both above and below, i.e., tlie sections nuist be capable of resisting compression, adjualable rods for such bracing being allowed only in towers of druw-spans and in lower lateral sys- tems of deck-bridges. T)je stilt diagonals of lower lateral systems, which shall be of double cancellation, shall be riveted rigidly to the string- ers where they cross them, so as to transfer in an effective nuinncv the thrust of braked trains to the truss-posts without causing a horizontal bending on tlie cross-giniers. All through-spans shall have stiff portal I)racing at each end, connected rigidlv to the inclined end iK)Sts. The said portal bracing shall be made as deep as the specified dcr head- room will allow. When the height of the tru.ssos is great enough to permit it, thet^ shall Imj used at each panel point a rigid bracing frame riveted to the top lateral strut and to the posts, and carried down to the clearance line. When the truss depth is not great enough for this detail, corner brackets of proper size, strength, and rigidity are to be riveted between the posts and the upper lateral struts. Deck-bridges shall liave stiff, diagonal braces between oppo- site vertical posts, which bracing, as a matter of precaution, shall have suflicient strength to carry one half of a panel-truss live load with its impact allowance ; and the transverse bracing between the vertical or inclined posts at each end shall be sutticicntly strong to transmit properly to the masonry one half of the wind-pressure and centrifugal load (if there be any) which is carried by the entire upper lateral system of the span. The lower lateral systems of deck-bridges shall be made of adjustable rods in alternate panels, thus leaving every other panel unbraced, and forcing the wind-pressure from below up the vertical bracing and to the ends of the span by the upper lateral system. 148 DK PONTIBL'S. Suspenders or hip verticals and two or more piinel lengths of bottom chord nl each end of each span sliall, preferably, be made rigid members, excepting that in " A '' trusses the bottom chords and i ni'ic; verticals are to be of eye-bars. A.l floor-b( a ns are to be riveted to the truss-posts in truss- spans, except. ng in the case of Petit trusses when the suspenc'- ers are of eye bars. In these, tioor beam hangers may be i:sed, provided they be made of plates or shapes, and that they be stayed at their upper ends against all possibility of rotation. ! CONTINUOUS SPANS. Except in tlie case of swing-bridges or cantilevers, constu-u- tive spans are not to be made continuous over the points ol support. TKESTLK TOWEU8. As a general rule, each trestle- bent shall be composed of two columns battered from one and a half (Ij) to two and a half (21) inches to the foot, the bents being united in pairs to form towers. Ench tower thus formed shall be thoroughly brac-'d with rigid bracing on all four faces, and shall hnve four horizontal struts at the base. In each intermediate liori- zoutal plane of division, formed by the panels of the tower bracing, there is to be a pair of diagonal adjustable rods to bring the colunms into proper position and to retain them there. The feet of the columns must be attached to anchorages capable of resisting twice the greatest possible uplifting ; and the details of the melalwork connecting the anchor-rods to th<; columns must be such as to make the metal woik and pedestals act as a single piece, so that, if tested to destruction by overturning, the bent would uol fail between the super- structure and the substructure. While it is desirable to have suflicient l)ase to prevent any tension from coming on the anchor-bolts, it is not advisable on this account to nuike the batter of the columns too great, especially in very high trestles. GKNERAL DESCllIPTION. 149 When trestlo-bents become unduly wide, a vertical column is to be jiliiced midway belweeu the legs so as lo divide up the truiisverse and horizontul sway-biacing. Care must be taken to jv.ovide properly for expansion and contraction at column feet both transversely and longitudinally. In elevated railroads, the towers can be placed at about every fourth span or, say, every one hundred and fifty feet, or can be dispensed with altogether, when the conditions so re(iuiie, by strengthening the columns properly to resist traction, thrust of braked trains, and the longitudinal compo- nent of diagonal wind-pressure. ADJUSTABLE MEMBERS. It is preferable to avoid altoget.icr the use of adjustable members in trusses, as well as in sway-bracing. If the struc- ture must hv made as cheap as possible, adjustable counters may be employed ; but it ix advisjvble to confine their use, as before stated, to diagonals in towers of swing-spans and in lower lateral systems of deck-bridges. CAMKEU. All trusses must be provided witii such a camber that, with the heaviest live load on the span, the total camber shall never be quit(! taken out by deflection. With parallel chords, sutti- cient (imb(!r will be obtained by making the top-chord sections longer than the corresponding bottom-ehord sections by one eighth (i) of an inch for each ten (10) feet of length. One half of the camber after a span is swung is to be taken out of th(! trac!' by dapping the lies, unless this would cut too deeply into the timber. Plate girders and shallow, open-webbed, riveted girders should not be given any camber. EXPANSION. Ever}' span nmst be provided with some means of longitud- inal expansion and contraction due to changes of temperature 150 DE PONTIBUS. over a range of cue lumdrcd mid fifty (150) degrees Fabreii heit. Spans up to eighty-five (85) feet in length, or in certuiii cases up to even one hundred (100) feet, may slide on pinned surfaces; but those of greater lengtli must move on nests of turned rollers. Occusionally a rocker end is permissible; but this method of expansion ia always to be avoided if practi- cable. ANCHOUAGE. Every span must be anchored at each end to the pier or abutment in such a manner as to prevent the slightest lateral motion. Vut so as not to interfere with the longitudinal motion of the trusses or girders du(! to changes of tcmperaiure or loading. NA.HE IT.ATES. Tlie niimes of the designer, manufacturer, and builder of every bridge or trestle, also the dute of erection, must l)c :it tached thereto in a prominent position and iu a durable man- ner. LOADS. The loads to be considered in designing bridges, trestles, and elevated railroads are the following; and all parts of same are to be proportioned to sustain properly the greatest stresses produced thereby for all possible combinations of the various loads. A. Live Load. B. Lnpact Allowance Load. C. Dead Load. I). Direct Wind Load. E. Indirect Wind Load, or Transferred Load. F. Traction Load. G. Centrifugal Load. IL Elfeets of Changes of Temperatu/e. In calculating the stresses caused Ijy a uniform moving load, the load shall be assumed to cover the panel in advance of the panel point considered; but the half-panel load going LOADS. 151 to the forward pjiiicl point will be ignored; or, in other words, tlie uiiiforni load will lie treated as if eoiiceutrated at the various panel points. In deck-spans on .sharj) curves, after the centre curve for each rail and Hit; centre lines of the longitudinal girders are laid out, th(! approximate extra live load on the outer girder due to tin; projection of the curve of the rail beyond its centre line near nnd->|)an is to be computed and added to the regular live load; but the corresponding excess of dead load from the flooring, being .small, is to be ignored. As the superelevation provides for an ecjual distribution of the live load on the rails for tlic assumed medium velocity of trains, there will be au excess of live load on the outer girder due to the velocity be- ing sometimes arealer tlum this; but the said excess is so small that it is to be ignored. The excess of live load on the inner girder, due to the ve- locity of train being sometimes less than that assumed for de- termining the superelevation, is offset by the reduced load due to the projection of the centre line of the rail near mid- span beyond the centre Hue of the girder; so it also is to be ignored. LIVE LOADS. The live load to be used in designing any railroad structure shall be taken from the "Compromise Standard System of Live Loads for Railway Hridges and the Equivalents for Same," which is given in Chapter XIX and in Plates I, II, III. and IV. lu single-track bridges but one of the seven classes of load- ing given can be used for any span; but in bridges liaving more than one track two or even three cla-sses of loading can be used in tin; same span, if so desired bj' the Engineer: for insumce, (.'lass W could be adopted for stringers, Class X for cro.ss-girders, and Class V for trusses, thus utilizing the theory of probabilities. The etiinvidcnt live loads given on tlie diagrams are to be used instead of the actual wheel concentrations. For elevated railroads the live loads are generally to be very 152 I)K PONTIBUS. much lighter than that of Class Z of the Compromise Stniid- iiril System; but ihe said loads will Imve to be determined for each iudividiial system of elevated railroad, so iis to provide for the greatest train load thai can ever come upon the struc- ture, but for no more. IMPACT ALLOWANCE LOAD. The impact allowance load is to be a percentage of the equivalent uniform live load, found by the formula P = 40000 L + 500' where Pis the porccntago jind /, the length in feet of span or portion of span tluit is covcicd b_v the live load, when the member considered is subjected to its maximum stress. DKAO LOAD. The dead load is to include ilio we'ght of all the metal and wood in the structuie, t'.xccpting lliat of those portions resting directly on the abulmonts, whose wciuflits do not aflfect the stresses in the trusses; alsci uiiy other pernianenl load that may be carried l)y tin; strueturc. The following uint weights are to be assimied in estimating the dead load : Creosoted lumber four and one-half (4i^) pounds per foot board measure. Oak and other hard woods four and a quarter (4^) pounds per foot board measure. Yellow pine three and three-quarters (3}) pounds per foot board measure. Wliite pine and other soft woods two and three-quarters (SJ) pounds per foot board measure. Kails and their fastenings, sixty (60) pounds per lineal foot per track. Two thirds (|) of the dead load shall be assumed to be con- centniled at the panel points of the lower chords in through- LOADS. 153 bridges and at tliose of the upper chords in deck-bridges; and one third (l) of the dead load at the panel points of the upper cliords in through-bridges and at those of the lower chords in deck-bridges. If in any bridge; design the dead load assumed should differ from that computed from the diagram of sections and the detail drawings by an amount exceeding one (1) per cent of the sum of the c(iuivalent live load and actual dead load, the calculations of stresses, etc., are to be made over with a new assumed dead load. WIND LOADS. For railroad bridges the wind loads per lineal foot of span for botli the loaded and the unloaded chords are to be taken from the curves given in Plate VII. The wind loads for the loaded chords include a pressure of three hundred (300) pounds per lineal foot on the train, the centre of wiiich pressure is applied at a height of eight (8) feet al)ove the i)ase of rail. For determining the requisite anchorage for a loaded struc- ture, the train of empty cars shall be assumed lo weigh one thousand (1000; pounds per lineal fool. In trosth; towers (lie colunuis iind transverse bracing shall be proportioned to resi,st the i'nilowing wind-pressures in addition to all other loads. 1st. When the structure is loaded, four hundred and fifty (450) pounds per lineal foot on stringers and cars, and two hundred ami fifty (250) pounds for each vertical foot of each entire tower. 2d. When the structure is empty, three hundred and fifty (IMO) pounds per lineal foot on stringers, assumed to be con- centrated one foot above the centre of stringer, and three hundred and tlfly (350) pounds for each vertical foot of each entire tower. The wind loads for longitudinal bracing are to be taken aa seven tenths (0.7) of those for the transverse bracing. In flgiuing greatest tension on colunuis and anchor-bolts, computations are to he made for bolh tiie loaded ami the un- 154 DE PONTIHUS. loaded slructure, in douhh-tiack trestles placing IJjo tialu of empty ears on tlie leeward Iraek. All wind l<«uls are to be treated as moving loads INDIRECT WIND I-OAD OU THAN8FKKRED liOAD. For both throui^h and deek spans, even with polygonal top chords, the transferred load is to be assumed to produce a tension in the leeward bottom elioid that is constant from end to end of span, and a similar release of tension on the wind- ward bottom chord. For trusses with parallel chords this assunii)tioii is correct, provideil that all tlie windjiressure travels directly to ends of span by the horizontal bracing; wiiile for trusses with polygontd top chords the asstnnption is a com- promise, the travel of wind-pressure being ambiguous. The transferred load at one pedestal is to be found by multiplying one half of the total wind load on the top chord by the average truss depth and dividing the product by the perpendicular distance between central planes of trusses. TRACTION LOAD. The total traction load on anj^ portion of a structure is to be taken as twenty (20) per cent of the greatest live load that can be placed on that portion of .said structure. In proportioning the towers and colunuis of trestles and elevated railroads, llie towc-rsand columns between consecutive expansion points are to be assumed to receive no aid from neighboring towers and columns, but must be figured for the greatest possible traction load between said consecutive expan- sion points. No percentage of impact is to be added to traction roads. CENTIUFUOAL LOAD. The centrifugal load is to be computed for the greatest probable velocity of trains by the formula C wv' INTENSITIES OF WORKING-STRESSES. 155 wlicrc iJ is the ceutrifiigul load ptr liueiil foot, w is the cquiv- iilcMit live loud per lineal fool, v is the velocity of train in feet jwr second, and li is the radius of the curve in feet. All portions of the structure affected by the centrifugal lo!i(l are to be figured to carry properly the stresses induced by the said load in addition to all other stresses to which they may be subjected. No percentage of impact is to be added to centrifugal loads. EFFECTS OF CHANGER OF TEMPERATUUE. In ordinary structures changes of temperature will not affect the stresses in the menjbers, provided, of course, that proper l)recuulion be taken to permit unrestricted expansion and con- traction. But in all arches, excepting only those hinged at both ends and at the crown, the stresses caused by the assumed extreme changes of temperature must be computed and duly considered. INTENSITIES OP "WOBKING-STRESSES. The following intensities of working-stresses (i.e., pounds per Sipiare inch of cross-section) are to be xised for all cases, except where wind loads are ctmibined with other loads, under which conditions the said intensities are to be increased thirty (;{()) per cent. But when high steel is employed the metal is to be strained tifteen (15) i)er cent higher for all ca.ses than herein specified, even after the said thirty (30) per cent has been added to allow for wind stresses. Tension on eye-bars in bottom chords and main diagonals, and on lateral rods 18,(X)0 pounds. Tension on shapes in bottom chords, main diagonals and laterals, on eye-bars in sus- ^)ende^s and hip verticals, and ou soft-steel adjustable truss members 16,000 '* Tension on net section of jdate-girder fianges (assuming one eighth of the area of the web to act as a part of each fiange), extreme fibres 156 1)E PONTIBUS. of rolled I benms, ami on shni)es in body of suspenders, hip veilicals and hanger-plates (there being 50 per cent increase of net area for section througii eyes) » 14,000 pounds Tension on adjustable truss members of wrought iron • • 13,000 IJending on pins 37,000 Bearing on pins and rivets (measured upon the projection of the sennintrados upon a di ametral plane) 22,000 Shear on pins and rivets 12,000 Shear on webs of plate girders 10,000 ' ' For fleld-rivels the intensities for bearing and shear are to be reduced twenty-flve (35) per cent. Compression on top chords 18,000 — 70 - ; I Compression on inclined end posts. . , 18,000 — 80 -; Compression on all other struts with I fixed ends 16.000 - 60-; r Compression on all other struts with I one or two hinged ends 16,000 - 80 -; T where I is the unsui>ported length of the strut in inches and r is its least radius of gyration in inches. Compression on end stiffeners of plate girders. 14,000 pounds. Tension on extreme fibres of long leaf, Southern, yellow-pine timber in bending, the effect of impact being considered 2,000 " BEARINGS DPON MASONRY. All bed-plates must be of such dimensions that the greatest pressures on the masonry, including impact, shall not exceed those iriveii in the following table: INTEN81T1KS OF W()UKlN(J-STIlK8Si:S. 157 Matorial I'eiiiiissible Pressure *'**®"*'- per Square Inch. Am. Nat. Cement Concrete 130 pounds. Brickwork laid in Cement 170 " Portland Cement Concrete 200 Ordinarily Good Sandstone 200 " Extra Good Sandstone 250 " Yellow Pine or Oak on Flat 300 " Ordinarily Good Limestone 300 " Extra Good Limestone 400 " Granitoid 450 " Granite 550 " KEVEK8IN«-8TUE8SE8. In case stresses reverse, the areas required for both tension and compression, including impact in each case, are to be fl;jrured separately, and three fourths (J) of Ihe smaller urea is to be added to the larger are.i in order lo obtain the total sectional area of the piece. The rivets, however, are lo be figured for the sum of the two stresses, both impacts included. The effect of reversion of stresses in case of wind loads is lo be ignored when computing sectional areas of members and the number of rivets required ; but, of course, wherever rever- sion of stress occurs, the piece must be stiffened so ns to resist compression. NET SECTION. The net section of any tension Hange or member shall be determined by a plane culling the member square across at any point. The greatest luunber of rivet-holes whicli can be cut by any such plane, or whose centres come nearer than two and a half (2i) inches to saiil phuie, are to be deducted from the gross section when computing the net area. BENDING MOMENTS ON PINS. In figuring the bending moments on pins, the stresses shall be assumed as concentrated at centres of bearings. 168 DE PON Tin US. COMBINATIONS OP STRESSES. Tn tlie girders of pl.-ilc tinier spans mid of (1»'(U, o|)t'ii-\\('l)l)('(l, rivi'tt'd-irirdor sp;ins, tin; only stresMes Mint nct-d to ho cimi- sidorcd are tliose ransed by the live, iiiipaci, dead, mid cen- trifugal loads. The trusses of tlirough-hridges will be nlfected by the live, impart, dead, direct wind, and indirect wind loads ; and in exceptional cases ;i' > liy tlie centrifugal load. The trusses of deck bridges will l)u all'ected by all of Ihewj loads. In uo ca.se will tlie traction load allecl tiie trusses of bridges to such an e.vtent as to reciuire consideration ; conseipicntly the only l^rovisiou for traction load required in through and deck bridges is adequate rigid bracing to carry it from tlie track to the trusses without subjecting any portion of the structure to an improper loading, as, for instance, the flanges of cross- girders to horizontal bending. In hri(lf/iH of all kinds the various loads herein specified sliall be combined without any reduction ; but in trestles, more especially very liigb ones, it will be legitimate, wlien combining the stres.ses from the various loadings, to reduce some of them or even to ignore some entirely, in order to avoid proportioning for any highly improbable or impossible combination of loads. For instance, when a trestle is situatid near the middle of a sharp curve or near the apex of two lieavy rising grades, it would be incorrect to assume a higii velocity of train. In such cases as these the element of in- dividual judgment in combining the stresses from the various loads and in assuming the sizes of the latter cannot well be eliminated. IJENDING ON TOP cnOKDS. For cotnbined direct stresses and bending (m chords, the moment is to be computed by the compromise formula CO.MIMNATIONS OF STUKHSKS. 159 wlu'it' IFistlic totiil tnmsvcrso load in pounds on llie pitice, including impact,, and I is tin; longtii r)f tlic jiiccc in iiiciics. Tlio extreme (Ibrc-stross for the conibiuiitioii kIihII n(»l ex- ceed Hixtcen tlioiisand (16,000) pounds ; and tlie nionunt at mid iKincl is Id be ussunit d the same in amount as that at the |)anel points. Top chords subjected to transverse loading should be made as deep as economy of metal will permit. HKNDING ON INCI-INKD END POKTP. In i»roporlionin!r inclined ciul posts of trusses of throuf^h- bridges for a combiuntion of all tin- loiid-. liciein specified, fo- gctlier with the l)cnding caused by llie wind-pressure \vlii
  • er cent of the specified intensity of working-stress, the effect of such bending may be ignored ; but, if it does so exceed, its effect niu.st l)e combined witli tliose of the other stresses, vising, however, foi- determining the sectional area, an inten- sity of worliing-stress ten /lO) per cent greater than that specified. 1()0 l)E I'ONTIIJUS. GENERAL LIMITS IN DESIGN INO. The following ^eiuMiil limits sliall be lulhered to in design- ing bridges, I resiles, viaducts, uud the line-work of elevated railroads : No nieial less than three-eighths (f) of an inch in thickness shall be used except for lilling-plates. Tlie least allowable thicknesses of webs of rolled 1 beams shall be us follows : 24" I beams f" webs. ao' •' vv " 18" " i" " 15" " iV' " 13" " t" " No channel less than Icn (10) inclus in deplli shall be used, except for lateral stmts, in which eight (8) inch channels may be employed. No angles less than '■]" X ^s X §" s lall lie used, except fi)r /acing. No eye- bars less than four (4) inches deep or three (juarters (J) of an inch tlii( k sliall be employed ; and the; depths of eye bars for chords and main diagonals shall not be less than one fifty-fifth {j'b)"^ t'"^ length of the horizontal projection of same. No adjustable rod sliall have less than one square inch of cross-section. The shortest span kiigth for trusses with polygonal lop chords shall be one huiulred and seventy-five (175) feet. The limit of span length in which the stringers can be riv- eted continuously from end to end of span sliall be two hun- dred (200) feet. Beyond this limit sliding bearings nnist lie used at one or more intermediate panel points ; and in no span shall there be a length of continuously riveted stringers exceeding two hundred (200) feet. For all couipression-members of trusses and for eohnnns of viaducts an;, elevated railroads the greatest ratio of unsup- ported length to least radius of gyration shall be one hundred (100), excepting those menibers whose main function is to OKNKRAL I'UINCll'LKS IN UESKJNIN'f;. Ifil resist tension. In those the limit may be miscd to one liuu' drod iind twenty (liJO). The concspondinf; limit for till struts belonging to sway- bracing shall be one hundred and forty (140). OENERAIi PRINCIPLES IN DESIONINQ ALL STBUCTUBE8. In designing all structural metal-work the following prin- ciples are invariably to be obsrrved : 1. All members must be straight between patiel-poinis, as curved struts or ties will under no circumstances be allowed. 2. The axes of all members of irii.ssiisor gii(l(rs and those of lateral systems coming together at any ape.\ of a tru.ss or girder must intersect at u point, wiitnever such an arrange- ment is practicable ; otherwi.so the greatest care UMist be em- ployed to ensure that all the induced strcs'lanes at right angles to each other is to be attained wherever practic able; but in designing top chords and inclined end posts this rule cannot be fol- lowed. 5. In both tension and compression members, the centre line of applied stress must iiivarialiiy coincide with the axial right line passing through the ceu'res of gravity of all cross- sections of the member taken at right angles thereto. C. The principle of symmetry in designing must be ivinied even into the riveting; and groups of rivets must be made to balance about centre lines and ceiilra' planes to as great an extent as is practicable. 7. In all structural metal-v/ork, excepting only the ma- ph'mer^ for operating jyovaliie bridges, uo torsiojj on any 1G2 DR PONTIBIS. member shiill be jjermittcd, it it can possibly be avoided ; otherwise, the gR'»tt\st care must be taken to provide ample streuglh and riL'^'ily for every portion of the structure af- fected by Hiicb torsion. 8. lu designing all pin-connected work ample clearance for packing must be provided, and ample room must be left for assembling members in confined spaces. 9. In bri(l>;i's, trestles, and elevated railroads the thrust from braked trains and the traction must be carried from the stringers or longitudinal gii. 17. Slur SI ruts formed of two angles with occasiomd nLojJ. OKNEIIAL IMIINCU'LKS IN DIvSKiN I N'(i. I(i3 pieces of ans^le or plafe for .stayinj^ same a'*e not to be used, for Ix'ttcr results iire oliluiued by phiciiig tiie augles in the form of a T. IH. In all mail) members having an excess of section above tlial called for liy the ureale.s(. combination of stresses, the entire detailing is to be proportioned to correNpond with llie utmost working capacity of liie member, ami not merely for the greatest total stress to which it may be subjected. In this connection, though, the reduced capacity of single angles (•oune(;ted by one leg only must not be forgotten. 10. Di'signs mu t invariably be made j-<>tlial all n)etal-work after erection shall be accessible lo the paint-brusli, (ixcepting, of course, those surfaces which are in contact svith eacli other or with th(! masonry. This reiiinrement 'ules out all clo.sed coiunuis of every type and description. 20. In general, details must always be proportioned to resist every direct and indirect stress that may ever come upon them under any probable circumstances, without siibjecting any portion of their material to a stress greater than the legitimate (^orresjionding woi king-stress. ~M. In all designs simjjlicity in l)otli main members and de- tail-; is to be considered of the greatest imitortaiu'c. '2:1. In all structures rigidity is to lie considered quite as im- j)ortant an clement as nnre strength. 2;V Structures on skews are to be avoided whenever it is practicable to do so. 24 The use of more than a single system of cancellation in bridges shall be eontined entirely to lateral systems and sway- bracing, except that at mid-panels of trusses two rigid diag- ileh of rivets in all classes of work in the direction of the stress shall never exceed six (G) inches, or sixt< eii (IG) times the thickness of the thinnest outside plate, nor be less than three (8l dianu'ters of the rivet. At the ciidsof compics sioii inenilicrs ii shall not exerted four (1) times the dianu'ter of the riveLs, for a length ecjual to twice the width of tin meml)er. When two cjr more tldcknesses or plate luc riveted togellu f in compiessioii-memlxMs, the outer row of livels shall not be more than four (4) diameteis from the side edge of the plate. Jfo rivet-hole centre shall be lejw ihan ore and a Imlf (l^J •II RIVET I NO. 165 diameters from tlie edge of u plate, and, wlicnevcr practieable, this distance is to be increased to two (2) diameters. The rivets wlien driven must completely till the holes. The rivet-lieads must in generfd be round ; ancl lliey must be of uniform size for the same-sized livets lliroriuhoul the work. They must b" neatly made and (toncentric with the livct-holes, and nuist thoroughly pincli the connected p'eces together. Itivets with flat heads shall be preferred to countersunk rivets; the iicight or tiiickness of the Hat head shall be three eighths (|) of an inch Rivets shall not 1 -ouutersunk iu plates less tlan seven sixteenths (j'^) of an inch in thickness. Flanges of sfingers and girders carrying the vertical load from the tiesshall liave their rivets spaced uniform!}' from end to end. and at the miinnuim distance emiilo3'ed. Whenever possible, all rivets shall l)e machine-driven, and the machines must ])e capable of retaining tlie applied pressure until after the upsetting is completed. Field-riveting must be done with a button sett : the heads of the rivets must be liemispherical, and no rough edges must be left. All rivets in splic-e or tension joints are to be arranged synunelrically so that each half of any tension-member or splice-plate shall have the same uncut area on each side of its centre line. No rivet, excepting those in shoe-plates and roller or bed l)lates, is to have a less diameter than the thickness of the thickest plate through which it passes. The effective dianuiter of any rivet shall be assumed the same as its diiuneter before driving; but, in making deduc- tions for rivet-holes in tension-members, the diameter of the boles shall be assumed on(> eighth (J) of an iiudi larger than that of the rivet. In the elfective are.-i of riveted members, jtin, boll, and rivet boles shall be count'.'d out for tension, und boh and j)iu lioles shall be counted out for compression. 166 DE PONTIHUS. DETAILS OF DESIGN POK ROLLED I BEAM SPANS. Rolled I beiims used as lougitudiriul girders shall have preferably a depth not less than one twelfth (,'3) of the span. They shall be proportioned by their moments of inertia. I beam spans may have eitiier one or t^vo beams per rail. In the former case the spacing should be si.x (G) feet six (0) inches, and in the latter case two (2) feet six (6) iiicl»es between contiguous girders. Willi two lines of stringers per track, there will be recpiired a bracing-frame at each end of span and di.'igonal bracing between the top llanges, unless the span be less than ten (10) feet in length, in which case the diagonals may be omitted. With four lines of stringers per track, no diagonal bracing will be required, but thna' (:>) bracing-frames at I'iwh end will be used, with three (3) more at mid-span when the span length exceeds ten (10) feet. Each I beam is to have at each end a pair of stiffening anules, c^ne of which will form a portion of the end bracing- frame. These are to fit tightly at both top and bottom against the llanges. Under each end of each I beam there is to be riveted a bear- ing plate of proper area and thickness to jHstribute the load uniforndy over the masonry, said plate being bolted effectively to the latter with due provision for expansion and contraction. DETAILS OP DESIGN FOR PLATE GIRDER SPANS, Plate girders shall have preferably a depth not less than one tenth ( j*jj) of the span. All plale girder-;, whenever it is practicable, shal' <)udt witlKjut si)lices in the web , and, whcni such l)eco -i: neces- sary, the smallest possible lunuber of .siuie shall be t.dopted. Tlie splice-pl.ites and rivets for the splices shall be such r to develop in every respect the full strength of the net section of the web, the main splice-piates extending from Mange to flange and having at le.asl two (2) rows of rivets on each side of the joint. In addition to these, each tlange shall be spliced by two cover-plates on lop of the vertical legs of DKTAILS Ob' DESIGN FOR I'LATlXil KUEK SPANS. i(n tlio lliUige aiii^lt'S. Tlicse iiiiist be long (Miougli to develop by tbo coiinecling rivets nl least Iweuty-tive {■i')) per cent more than Ibe fill, slreni;tli . Every non-continuous Hange-piece shall be fully spliced so thai tlie splicing plates and rivets shall have a calculated streng ii at least twenty live ('J:")) pei' cent greater than that of tlie section spliced. Field-splicing of plate girders will never h^' allowed for fixed spans, except in structures lor foieign countries. At least one half of every llange section must consist of angles, or else the heaviest sections of the latter iuu.st be u.^ed ; and the innnbcr of cover-plates must be made as sin:ill as practical)le, in no c;ise exceeding thiee (3) per flange. The lengths of ilii'sc covci plates must be such as to make them l)i'ojecl at each end not less than nine (11) inches lieyond the point dctcrmiiu'd by the calculations for the reiiuisite resistance to bemling. Where two or three cover-plates per flange are used, they shall be of < ({u.-d thickness, or shall decrease in thickness out- ward from the angles. The cover plates shall not extend more than four (4) nchcs or e'ght (8) times the thickne-s of the outer pl;Uc beyond the ou'er line of rivets. Willi cover- I>lales more than fourteen (II) inches wide, four 4) lines of rive!s shall be used. 'i'lie compression. tlaiiges of plate girders shall be made of the ^anie gross section as the tension- flanges; and they shall be .so siilfened laterally that the unsupported length shall never exceed twelve d'J) limes llie width of flange. In deck spans there arc to l)e bracing fraie-e^s at t!ie ends nm] at int(M-medi;ite points not more than fifteen ( l.^» feel apart; .and iliere is |o ho an eireeiive system of diagonal ht-cing of 1G8 DE PONTIJU.'S. iitiglt'S between the toi) fliiiigos of the contiguous girders for euch truck. Ill half-through npiins the girders are to be divided up into l)auels geneially not exceeding fifteen (lij) feet in leiigtii. If u steel floor system be \ised, tliere lire to be brucliets of web- pliites !iud angles at the ends of the cross-girders extending to the top liaiiges of the longitudinal girders, so as to stay tlie laiter eft'eclively ; while, if a wooden floor system of ties rest iug on shelves or on the bottom flanges L used, there are to be steel cross frames with solid webs, of the greatest de|)lli obtainable, with similar brackets at their ends for the same pui'i'ose. Ilalf-lhroiigh plate-girder spans are to have a rigid, double intersection, lower lateral system of angles riveted to- gether by ])lates and angles at their intersections and to the botlom flanges of ilic steel stringers, if the hitter be employed. Webslillciicrs shall be placed at the ends of plate-girder spans, also iil all points of concentrated loading and at inter- mediate i>oints at di.^tances not exceeding either the depth of the girder or Ave (5) fiet, except in the case of shallow girders where the shear, including impact, does not exceed Ave thou- sand (5000) pounds per square inch of web section. Under such circumstances the spacing of inteiinediate sliireners may be made as great as three (3) feet six ((5) inches. All sliireners must bear lightly at top and bottom against the flange angles. Under end stilTcners tlitre must be flUfra flush with the flange angles, but intermediate slill'eiiers shall, preferidily, be crimped. End stilTeiiing angles shall in no case be less than 3i" X 3J ' X f , net, and must have siiflicitnt area to curry the entire enii shear, including impact, willi ihe specified intensity of woik- iiig-slrcss, no relianci' being placed on the fllhT'^. The sections of intermediate si liTening angles shall not be less than those given in the following tabC. I^eiipth of Wirder. Dimensions of Angles. Up to 50' n\ X :H" X §" From 50' to 70' 4 X H X | From 7t» to 90' X3i XI avu as cas vati wit nici OPKN-Wi;i?BUI), KIVKTHl) (il i{im:k-sfaxs. IGD In proportioiiiugllie flanges of plate girders, one eiglitb (J)of tlie gross urea of the web is to be jisstiiiied !is coneeiitnited at till! centre of gravity of eacii flange; or, in otlicr words, after iiaving found tlie net sectional area reiiuired for tbe tension- flange by ignoring the resistance of tlic web to Ijcndiug, tiiere is to be subtracted llierefrom one eigbth (J of the gross area of the web- plate. At the ends of all plate girders tliere must l)e suflicient rivets in eadi flange to transfer properly thereto from tlie wei) the total end-shear in a distance eijual to the ellective deptli of the girder. At the euds (,f cover-plales the spacing of tiie rivets which attach the covers, for a length eijual to at least twice the width thereof, shall be male the minimum used in the tianges. Under each end of each plate girder there is to be riveted a bearing plate of proper area and tiiickiiess and thf)rongiily stiffened so as to distribute; the load luiiformly over the ma- sonry, said plate being bolted effectively to the latter with due provisiou for expansion and contraction. DETAILS OF DESIGN FOB OPEN- WEBBED, KIVETED GIRDER-SPANS. All open-webbed, riveted girders for both deck and half through bridges shall be riveted up completely in tin.' siiop, as flelil-rivetiug will be allowed only for the lateral bracing, except in structures for foreign countries. In open-webl)ed, through, liveted ginUrs, however, lluMon- nection of main memheis will have to be by ticld-rivets. In such cases all of the trus.s-meMil>ers will have to be assembled in tiie shop, after which the rivet-holes for the cwuieclions shall be reamed so as to ensure perfect filtintf in Iho tiehl. The u:?e of shallow opcii-weltbed, liveted girders shall be avoided whenever possible, for the reason that they aie tpute as expensive and never as satisfactory as plate girders. In case, though, of their being reipiired, as for instance in ele- vated railroads i>ocnpyiiig city streets, they are to be jyrovideil with short, substantial web-plilesat the ends and al all inter- mediate points where connections are made to other girders. i:o I)K PONTIHUS. lu i>iop()rtioiiiiig the wcb-mcnibers of such girders, the Hpccitied intensilii's of working stresses arc to be reduced fioiii ten (10) per cent for (»" X 3,j' angles to twenty-tive (25) per cent for eciuiil-legged angles, with proportionnle amounts for angles of intermediate inequality of legs, so as to compensate for the secondary stresses due to the eccentric grip of tlie rivets. In no case will it be jierinissible to use lljits instead of angles for web-menibers, but tci'S may be employed, |)rovided their heads be wide enough to permit of satisfactory riveted con- nections. A.t all intersections of web-members with chords, connect- ing phites are to be used; for it is not perniiss'lile to attach web angles directly to chord angles willujut using an inter- mediary ])late. The exact intersection at a point of all gravity lines of girder- members assembling at any apex must l)e adhered to in tiie designing of open-webbed, riveted girders. In designing all riveted connections, the greatest care is to be taken to make connecting plates and groups of rivets l)alance about centrelines of stress, especially where pa> said angles be spaced very far apart, jis in later il struts connecliiig deep top chords, they are tf) be placed on the corners of a rectangle, wi.h their legs turned inward, and laced on all four faces of the box .strut thus formed. Eye-bars are to be used for !ill bottom chords and main diagonals that do not require to be stiffened. Counters, when employed, can be of either roiuids, squares, or flats. These and all other jidjustable members are to h;ive their ends etdarged for the screw-threads (unless soft-sleei, cold-presse(i threads be used) so that the diameter at the bottom of the thread shall be one eighth (^) of an inch greater than that of the body of a round rod of area equal to that of the atljustable piece. In short spans, two angles riveted back to back, or even a single large angle, may be used for lower Lateral diagonals ; but for long sp;ins the tliagou'ils nre to be niadc; of four angles U' I ir. UK F'ONTIIirS. iu the form of an I with ii single line of liicing. When two angles are U'^ed, u single i)Iate must not lie depended on to form the splice at the intvisedion of the diagonals, but two angles, each not less tlnn two (2j feet long, are to be placed benciith or on top of the spliced angles, so as to form ii full splice* ill respect to rigidit}' as w(;ll as strength. Diagonals for upper lateral systems and vertical sway- bniciiig shall, preferably, be built of four angles in the form of an 1 with a single line of lacing; hut, for structures where this section would involve an extravagant use of metal, two of the angles, one at top and one at bottom, may be omitted, thus nuiking each strut consist of two angles laced, provided, of course, thiit where the struts cross they shall be rigidly connected by two plates of ample size. This luibahinced section for such diagonals is to be avoided Avhenever it can be done without undue use of metal. In no case, though, will it be permissible to use angles in tension that are not capable of resisting properly the possible compressive stresses, witii d\ie regard for the specified limit of ratio of tuisupporled length to least radius of gyration. In designing transverse lateral and overheaii points. If the panel length exeeiul thirty (30) feel, there shall be a bra(;ing-fiame at mid-length l)etweeii the contigJious stringers of each Iraek; but for all shorter jmnels the rigid lower lateral diagonals which are riveted to the bottom flanges will stiffen the latter sullicienlly. In respect to stiffening anglts for stringers, the rules governing those for plate-girder spans are to be followed; but the end sliffeners are to be faced or otherwise treated so as to make the stiingers of e.xael leigth throughout, and so as to effect a uniform bearing of the end stiffeners against the webs of the cross-girders. In respect to pro|)ortioning of tlanges and numlier of rivets required, the r\iles given for plate-girder spans are to apply also to stringers. The said rules are to apply also to (mdss- girders, as shall also those relating to stiffeners, splices, cover- plates, and size of compression- flanges, that are given for plate-girder spans. Wherever it is necessary to notch out the corners of the cross-girders to clear the chords, the greatest care must be taken to provide an adeipiate means for transfei- ring the shear to the jxtsls without im[)airing either the strength or the rigidity. If necessary, in through-bridges the web of the cross-girder can be divided into three parts so as to let the end portions project above the top llange and form brackets that will afford opi)ortunity for using an ample ninuber of rivets to conne(;t to the posts, and will .strengthen P'operly the otherwise weakeneil cro-ss-girder. In order to carry the. thrust of trains from the stringers to the posts through the lower lateral diagonals, the latter and the stringers aru to be madt; to form complete hoiizontal trusses by running angles between stringers at the leve. of the bottom tlanges. In single-track bridges two piec(;s of angles per iiaiu'l running transversely between siringersat the iuter- iiecition of \hr latter niib the diagonals will sufflce; but in i IMAGE EVALUATION TEST TARGET (MT-3) A {./ ^ .^ti. < ^ 174 DE PONTIBUS. doiible-tmck bildgt'S tlierc will be n-qtiired two such nngles per pttiiel belweeu inner stringers, and four diiigonal angles pe. panel to run from where the Intend diagonalH intersect the outer stringers to whore the inner stringers nieot the crttes- girders. AH plates, angles, and channels used in built members of trusses must, if practicable, be ordered the full length of the member; otherwise the splices must develop the full strength of the member, without any reliance being placed on the abutting ends for carrying corapre&sion. But. in total splices at the ends of sections perfect abutting of the dix'ssed ends is to be relied upon. However, the splice- platis even there must be of ample size and strength for both rigidity and continuity. Tlie unsupported width of plates strained in compression, measuring between centre lines of rivets, shall not exceed thirty-two (33) times their thickness, except in the CMse of cover-plates for top chords and inclined end posts, where the limit may be increased to forty (40) times the ihickiHjss. Where webs arc built of two or more thicknesses of plate, the rivets that are used solely for making the several thicknesses act as one plate shall in no ca8en sides of all compression-members composed of two rolled or built channels, with or without a cover-plate, shall be stayed by tie plates at ends and by diagonal lacing- bars or lacing-angles at intermediate points. Lacing-lmrs may be connected to the flanges by either one or two rivets at each end; but lacing-angles, which are used for members of heuvy section only, must be connected by two rivets at each end. The tie-plates shall be placed as close as practicable to tiie ends of the compression-members. Their thickness shall net be less than one- fiftieth (7('|;)of the distance between t!ie centre lines of the rivets by which they are connected to the flanges, unless said tie-plates ])e well stiffened by angles, in widdi case they uiay be made as thiu h9 three eighths (|) of an inch. LACINO AND PIN-PLATES 175 The Icugth of a tie-plate shiill never be less Ihiiu its width, or one and one-half (IJ) times the least dimension of strut (unless it be close to a web diapiiragm of tlie member, in which case it may be as short as twelve (12) inches), antl seldom greater than one and one-half (U) times its width. The thicknesses of lacing-bars shall nevei be less than one liflielh (j'o) of the Knglli between centres of the end rivets, nuasuring between inniost rivets in case that there be more than one rivet at each end. The smallest section for a lacing- bar shall be one and three quarter (1 J) inches by three eighths (j) of an inch, which size shall b;; used for channels under nine (9) inches deep; and the largest section shall be two and a half (2A) inches by one-half ( j) inch, which size shall be used for channels lifteen (15) inches deep. For intermediate sizes of channels, the sizes of laeing-btirs shall be inlerpolaied. For all built channels of greater depth than fifteen (15) inches, and for nil cases where a lacing-bar would require a greater thick- ness than one-half (j) inch, angle lacing is to be used, the smallest section for same being 3" X 2i ' X I", and the largest 2i" X 3*" X f '. For two (3) inch lacing-bars and two and a half (2i) inch lacing angles, three-quarter (J) inch rivets are to l)e used; and for two nad a half (2A) inch lacing-bars and three (3) inch lacing-angles seven-eighths (|) inch rivets are to be adopted. In general, the inclination of lacing-bars to axis of member shall be about sixty (60) degrees ; but in meml)ers of minor importance and in tension- members the said inclination may be made slightly Hatter. Pin-plates shall be used at all pinholes in built members for the double purpose of reinforcing for the metal cut away and reducing the intensity of pressure on pin ami bearing to or below the specified limit. They shall be of such size as to distribute properly, through the rivets, the pressure carried l)y such plates to both tlanges and web of each segment of the member ; and tliey shall extend at l"*"^* six (6) inches within the tie-plates of said member, so as to provide for not less than two (2) transver.se rows of rivets there. When the pin ends of compression -members are cut awaj 176 DE P0NT1BU8. into jaw-plates or forked ends, for the purpose of packing closely the various nieuibers connected by the pin, tLw'se juw plates or post extensions slmll be considered as columns, the thickness of each of which shall be determined by the follow- ing formula : p= 10.000-300;; where p is the greatest nllowable intensity of working-stress (impact being consideied) ; I is the unsupported Iciiglli in inclies, measuring from llie centre of the pinhole to the centre of the hist transverse line of rivets beyond the point at which the full section of the member begins ; and t is the tottd thickness in inches of one jaw. The length I is alwiiys to be made as small as pnicticnbU'; and, iu cases of unavoidably long extensions, the plates are to be stiffened by an interior diaphragm composed of a web wiili four, or sometimes only two, angles. It is always better, whenever practicable, to avoid cutting away the ends of channels ; but, if Ihey must be trimmed, the ends must be reinforced so that the strength of the member shall not be reduced by the trimming. In riveted tension-members, the net section through any pinhole shall have an area fifty (50) per cent in excess of the net sectional area of the body of the member. Tlie net sec- tion outside of the pinhole along the centre line of stress shall be at least sixty-five (65) per cent of the net section through the pinhole. Pins are to be proportioned to resist the greatest shearing and bending produced in them by the bars or struts which they connect. No pin is to have a diameter less than eight tenths ( f%) of the depth of the deepest eye-bar coupled thereon. No truss-pin is to have u smaller diameter than three and a half (8^) inches, and no lateral pin, if any such be used, a diameter less than two and a half (2i) inches. Lower chords are to be packed as closely as possible, and in such a manner as to produce the least bending moments on the pins; but adjucfut e3'e-bAis iu the same panel must never have to pi-( CHORD-PACKING, ROLLERS AND I'KDK.srALS. 177 less than a ODe half (i) inch space between them, iu ordei- to facilitate 'painting. The various menihers attached to any piu must be packed as closely as piiiclicable, and all interior vacant spaces ni'ist be filled with steel tillers, wliere their omission would permit of motion of any member ou the pin. All bars are to lie in plnnes as nearly as pfwsilJle parallel to the central truss-plane, no divergence exceeding one eighth (J) of an inch to the fool being permitted. In detailing I struts composid of four angles with a single lino of lacing, the clear distance between backs of angles shall never be made less than three-quarters (J) of an inch, in order to permit the insertion of a small paint-brush. Th*; greatest allowable pressure upon expansion-rollers of lixed spans, when impact is considered, shall be determined by the equation p = 600rf, ■where p is the permissible pressure in pounds per lineal inch of roller, and d is the diameter of the latter in inches. The least allowable diameter for expansion-rollers is three (3) inches. Rollers shall be enclosed in boxes made practically dust- tight, but which will not retain water, and which are so •leslgned that the sides can be readily removed for the purpose of cleaning. These boxes must be so designed as to permit of a free movement of the rollers in the longitudinal direction of span sufficient to take up the extreme variations in length due to temperature changes and deflection, and at the same time prevent any transverse motion of the end of the span. All sboe-plates, bed-plates, and roller-plates are to be so stiffened that the extren.e fibre-stress imder bending, when im- pact is included, shall not exceed sixteen thousand (16,000) pounds. Pedestals shall be either of cast steel or built up of ulates and shapes. In built pedestals, all bearing surfaces of the base-plates and vertical bearing-plates must be planed. The vertical plates must be secured to the base by angles having at least two rows of rivets in the vertical legs ; and the said 178 DK I'ONJIHUS. vertical plates must bear properly from cud to end upon the base. No buse-plule, vertical plate, or coiiiiuctiiig angle shall be less in thickness than three quarters (J) of an inch. The vertical plates shall be of sufHcient height and must con- tain enough metal and rivets to distribute properly the loads over the bearings or rollers. The bases of all casl-steel i)ede8- tals shall be planed so as to bear properly on the masonry or rollers. All rollers antl the faces of base-plates in contact therewith are to be planed smooth, so as to furnish perfect contact between rollers and plates throughout their entire length. All pedestals whether built or cast must have one or more diaphragms between webs, carried up as high as the general detailing will i)ermil, .so as to liansmit transverse horizontal thrust to the base without overstraining the webs by bending in their weakest direction. Heads of ( ye-bars are to bo made of such dimensions that when the bars are tested to destruction they shall bieak in the; body and not in the eyes; and in case of lo()|)-eyes, so that they shall not fail in the welds. Rods with bent eyrs shall not be used. In loop-eyes, the distance from the innc point of the loop to the centre of the pinhole nuist not be less thnn two and one h;ilf (3i) times the diameter of the pin. and the loop must lit closely to the pin throughout its entire semi- circumference. DETAILS OF DESIGN FOR TRESTLES AND ELB- VATED RAILROADS. The sections of main members of trestles shall, generally, be as follows; Columns, two channels laced with Uanges turned either out or in, two channels with I-beam web between, four Z bars with web-plate, four Z bars with a single line of lacing inside and occasional stay-plates outside, or four angles with a single line of lacing inside , diagonals in transverse and longitudinal bracing, and nil bottom hori/.ontal bracing struts, four angles with a single line of lacing; horizontal transverse bracing struts al top of lowers, bracing DKTAILS OF TRESTLES. 179 frames of tiuglus ; luiigitudiual slnits ut top of towers, plate girders ; aud longitudiiml girders, plate-girder spans, or occasionnlly, for very long spans, open-webbed, riveted girders or pin-conuccled trusses. The detailing for longitiidiuiil girders of trestles and elevated railroads and the bracing between sanic shall comply with the specifications governing the designing of plutc- girder spans and the floor systems of pin-connected spans. In '/uueral, the transverse tind longitudinal bracing of trestle >wers shall consist of a (loiible-cancellation system of stiff dia. >nals without auy horizontal struts, except at the bottom bc-lween pedestals. The latter struts must be strong enough to move the column feet upon their sliding-bearings when said struts are expanded or contracted by (changes of temiierature. Provision must l)e made for holding some feet rigidly, and for sliding some in one horizontal direction only, and others in an}' horizortal direction, at the same time holding them all down so that they shall not be lifted jK'rceptibly by the wind-pressure. Sliding-plates are always j)referal)le to rollers for pedestals of trestles. They shall be planed extremely smooth, and so as to bear properly at aii parts. Occasionally, in solitary bents, it is permissinie to use hinged ends for columns at pedestals ; but it is generally better to nuike them fixed, and to figure the columns for the greatest bending produced in them by transverse loads and extreme changes of temperature. The tops of trestle columns are to be made vertical by bending them just beneath the longitudinal girders where the latter are riveted to them ; and the upi)er transverse struts must be made as deep as the longitudinal girders, and must be riveted effectively to the columns. Corner biackets of double webs are to be used for connecting the columns to the horizontiil struts and bracing-diagonals, and at the same time to strengthen the column at the bend. Additional strength- ening is to be gi>v'n by using a solid web or diaphragm in the column extending from (he top thereof to a point about two (3 feel btlow the bend. ISO I)K i'ONTlRl'S. All Splices lu cultiinns twc to bo full, butt splices, located preferably about two ('J)feft above the poiuJs where the sway- diagonals connect, shiuglcsplicing being avoided because of the trouble it gives during erection. The best span lengths for trestles are generally those which make the total cost of structure a minimum, the tower length varying from twenty (20) feet for low trestles to thirty (80) feet for very high ones, and the intermediate spans varying from thirty (30) to sixty (60) feet for the same limiting heights. Any length of girder exceeding sixty (60) feet would probably necessitate the employment of a too long, heavy, nnd expensive traveller, or else the use of bents of falsework between the towers. For elevated railroads the sections of main members shall he as follows : Longitudinal girders, preferably plute girders, or, if necessary, open-webbed, riveted girders; cross- girders, plate girders ; columns for structures without longitudinal or tower bracing, two channels with an I beam riveted between ; and columns for structures with longitudinal or tower bracing, four Z bars with ii web-plate. All columns for elevated railroads are to have both ends fixed, being held rigidly at tl:? top by either the longitudinal girders or by deep struts that carry the thrust of braked trains from the track to the columns, and their sectional aieas arc to be figured accordingly for both direct innd and bending. Longitudinal girders in elevated railroads shall, generally, be riveted into the cross-girders and not rest thereon, except under certain conditions for the sake of clearance beneath, in which case the top flanges of the half-through girders must be stayed at the ends and at intermediate points, as specified for plate-girder spans On all curves in elevated railroads, special lateral bracing of angles, riveted at intersections to the longitudinal girders and carried over and riveted to the columns, must be employed. Where brackets for columns can be used advantageously in elevated-railroad work, they must be put in, and must be built of solid web plates and angles. In general, the limiting length of structure between expao-r DKTAIL8 OP ELEVATED RAILUOADS. 181 stoD points shall be about one hinulrcd and fifty (150) feet. If Ibis length be exceeded materially, the culunins may have to be strengthened to resist the bending caused by changes in temperature. All expansion-pockets are to be so detailed as to throw the loitd from the longitudinal girder as close as possible to the web of the cross-girder; and sufficient rivets are to be used in connecting the pocket to the cross-girder to provide for l)oth the direct shear and the bending moment from the eccentric load. All anchor-bolts at column feet are to extend well up above the base-plate, passing inside of a curved plate that is riveted to the column, and which siipports a heavy washer-plate to leceive the anchor-bolt nut. The space between the curved plate and the anchor-l)olt aftei erection is to be filled with Portland-cement grouting. All column feet are to be raised so far above the ground that no dirt, snow, or moisture can collect around them and remain there. The boxed spaces at column feet are to be filled wiih Portlund-ceuient coiirrele made with small broken stone. The bases of pedestals are always to be made large enough to prevent all possibility of settlement of foundations. In figuring Die pressure on tlic base of I he pedestals it is not suf- ticient to recognize only the direct live and dead loads, but it is necessary also to compute the additional unequal intensi- ties of loading caused by both longitudinal and transverse thrusts. CHAPTER XV. SPECIFICATIONS FOR KAILROAI) DRAW-SPANS. Thk speciticRtions given in the preceding chapter for fixed spans apply also tu draw-spuus, except where otherwise stated in the following pages. GENERAL DESCRIFTIOX. MATERIALS. The speciflciitious previously given apply also lo draw- sp.ins, except tliiit cast iron nmy be used for the centre cast- ings on top of pivot-piers, for anchor- pieces in the masonry, for shafting boxes, and for ritiUchairs, and some other castings of minor importance. Tlie use of Idgh steel for drawbridges will not be permitted. 8TT;,ES ok BRID(}KS KOH various span LEN0TH8. For spans up lo one hundred and sixty (160) feet in length, plate-girder spans should be u>-epins; althougli in certain cases it may, for the sake of appeamnce, be made a little greater. The truss depth at, the inner lups should be from one ninth (^) to one tenth (j'^) of the total length of span. The truss depth at outer hips for spans up to foijr hundred (400) feet will gener- ally be determined by the clearance required. For longer spans it should be between one fourteenth {^) and one lifteenth ( ,'j) of the total span length. The length of the centre panel will, in most ctises, bo made equal to the perpendicular distance between ad 8l'KCIKICATr()N'S I'OU UAILUOAD DHAW-Sl'ANS. 185 carried by one arm of the aaUl truss or girder wbeu tliu span is being swung, which proporiiun is to be tul^en from tlie fol- lowing table: SpatiH. Up to 150., 150' to 250'. 250' to 350'. 850 to 450'. Over 450'.. Ration of U|)lift to bfitd Load. These uplifts are to be adopted both for finuinp; the uplift stresses in trusses and for proportioning the ctid-liftiug mu- chinery; provided, however, that for the latter purpose no assumed uplift 1)e less than twenty thous!) of an inch in diameter. In no case shall the upper truck be less than two and one- (piarter (2J) inches, or the lower truck less than two and one- half (2^) inches thick, measuring on the central cylindrical surface of the drum. Tl»e lower track shall be anchored to the top of the pivot- pier with bolts uot less than one (1) inch in diameter, nor less than lifteeu (15) inches long, set in place with Portland cement grouting. These bolts arc to be made of soft sicel, with cold- pressed threads and hexagonal nuts at top, and with spMt e'.ids and wedges at the bottom. They are to be placed in pairs opposite on the inside and outside of the track, and are to be spaced not to exceed eighteen (18) inches between cenlres. Tlie lop of the pier is to be levelled off with neat, Portland cement mortar, and the lower track is to be set in sjime. It shall be made one and one-half (IJ) or two (3) inches higher in the centre than at the edge, so that the water will drain toward the latter. A snuUl gutter or depression in the top of the pier is to be made just inside of the lower track, and at the bottom of this depression drain-holes are to bo put in, leading the water from the gutter down ou the outside of the pier. These drain-holes are to be at least two (3) inches in diameter; and the tops are to be protected with screens, so as to prevent choking. They are to be spaced uot to exceed ten (10) feet between centres. The rollers shall be of cast steel, and are to be ::late that rests on the box girder. The si/.e of the anchor-rod is to be dotermineil by assuming an unbal- anced upward wind load of live (5) pounds per square foot on the total area of the horizontal projection of one arm of the span. The cap-plate for holding down the top connection-plate for the radial struts is to be attached to the toj) of the centre cast- ing by means of a bolt tapped into same. Tiii-t bolt is to be at least one and one-quarter (1^) inches in diameter. The rack for turning the span is to be made in short sec- tions, not over four feet long, so that in case of breakage oidy a small portion of the rack need be replaced. These rack segments are to be bolted to the lower track with tap-bolts not less than fifteen sixteenths (||) of an inch in diameter, and spaced not to exceed fifteen (15) inches between centres. Tliere must be enough of them in any case in any one sec- ment of the track to resist, with a good margin for con'in- gencies, the entire shear (including that due to the rotating moment) caused by ih' effort of the pinion or pinions that engage with said segment. The letist allowable thickness of metal in the rack shall bo one and one-eighth (1^) inches. The ends of the rack segments are to be planed so as to secure close contact, and the abutting ends are to be bolted together with turned bolts at least seven eighths (i) of an inch in diameter. The bottom of the rack and that portion of lower track SPKrlPlCATlONS POU HAILROAI) DRAW-SPANS. 100 upou wbicli I lie mck boiirs are to be planed smouth. The width of the hiise of the rack sliuU beat least two thirds (}) of itH height; and ribs bracing tlie vertical portion to the base Khali be provided at distances not exceeding eighteen (18) inches. Dndnage-holes not less than ilirec fourths ()) of an inch in diameter, spaced not njore tlian two (2) feet between centres, shall be l)ored in tlie lower-track segments, starling just back of the ra(;k and leading to the outside of the track. The girders (tver the drum siiall be so arranged as to dis- tribulti the load over it properly. Tlie number of bearing points required will depend upon the length of span, the distance from centre to centre of trusses, the total load to be carried, and the economical sizi! of pivot-pier. The arrange- ment of the supporting girders in turn depends upon the number of bearing points to be used For ordinary single- track bridges up to three hundred (iJOO) feel in length a very good arrangement of girders over drum is secured by nniking the diameter of the drum and the length of centre panel equal to the distance from (-entre to centre of trusses; then the mid- dle points of both tlie longiliullnal and the transverse girders will be directly over the web of the drum, thus furnishing four points of bearing. Four more points of bearing are secured by putting in short diagonal girders, which connect to both transverse and longitudinal girders and bear on the drum at their centres. This arrangement gives in all eight (8) points • of support. The longitudinal, transverse, and diagonal girders over the drum shall be .so designed that 'heir rigidities will be such that when deflected under llie load the extreme fibre-stress will be about the same in all the said girders. The bottom-chord .stresses iu the centre panel can either be carried by the longitudinal girders, or the bottoni' chord sec- tions can be continued through the centre panel, the longitu- dinal girders being placed above them, and steel chairs being inserteil beneath their centres to furni.sh bearings on the drum. In case that the bottom-chord stresses are carried by the longitudinal girders, ample provision must be made for 200 DB PONTIBUS. them, as well as for the bending stresses, in designing the sections for these girders. Wliere the clearance over tlie waterway will permit, metal can be saved by lett'ng tbe top flange of the lougitudiual girder form the bottom chord of the truss. In any arrangement of girders over the drum, bearing-plates at least one (1) inch thick must be used between the top flange of the drum and the bottom flanges of the girders, in order to make the points of concentration well defined, and so as to transmit the load properly from girders to drum. All girders bearing on the drum are to have stiffeners on both sides of their webs at all points of concentration; and in no case are the stiffeners to be crimped, but are to have fillers beneath. They must have close bearings at top and bottom flanges, and are to be proportioned in the same manner as previously specifitd for those on the drum. The rollers, tracks, drum, and girdeis over drum shall be completely assembled in the shop before shipment, all holes being reamed to fit and the sections being match-marked. Every roller must have a true bearing on both the upper and the lower tracks during a complete revoluiiou of the draw. Before the assembling of the rollers is done there must be nnirked on both the upper and the lower track segments a circle of the sjime diameter, which circles will come a trifle inside of the exterior ends of all rollers; then, after the turn- table is perfectly adjusted, each roller is to be marked where these circles touch it. After the turntable is disconnected each roller is to be set up properly in a lathe, and the exterior periphery is to be chamfered off exactly to the points marked, so that •.i'hen the turntable is set up in the field, if the exte- rior of each roller is brought exactly to the circles on the two tracks, the rollers will all be in their proper positions. These lines on the tracks will serve also afterwards to line up the rollers whenever the turntable is to be adjusted. SPECIFICATIONS FOR hAlLHOAD DRAW-SPAKS. 20l MACHINERY FOR TURNING THE SPAN AND LIPTING THE ENDS OF S/.Mdi POWER. When a draw-span is to be opened frequently, some kind of mecimniciii power must be used. The kind of power Ijol adupled to any particular span depends upon a number of conditions, more especially the location of the bridge. A gasoline-engine is an economic and convenient form of l)ower for small spans which do not require more than twelve (12) or fifteen (1;"5) horse-power to operate. Duplicate electric motors, where direct connections can be mode with electric-liglit or street-railway power-plants, are very efficient, convenient, and reliable; but in no case is it safe to depend upon sturagebalteries for power. Tlie use of electric motive power is therefore confined to bridges located in or near towns or cities. Where over twelve (12) or fifteen (15) horsepower is re- (juired for operating the spans, and where electrical conuic- tions cannot be made, the steam-engine is the best form of power to 11:36, except i)ossibly in some special cases where water-power c;in be had conveniently. Except in tjje case of short, light drawbridges, whenever met hanical power is employed it is necessary to apply the same to the rack by two pinions located diametrically ojiposite each other. If with this arrangement tl»e tooth pressure be still too high, it will be necessary to repbiee each jiinion by a l)air of pinions located as close together as practicable. With pinions located far apart some kind of an equalizer must be enjployed to divide the work equally between them, on ac- count of the unavoidable, slight irregularities in tlie tooth- spacing of the entire rack. When electrical power is adopted, the equalizing may be done by means of electrical connections between the duplicate motors; but with any other power a mechanical equalizer between the two ladial shafts must be employed. There will be no equalizing needed between tho P8%OVlNCi^i_ HDRARY VICTORIA, B. C. 202 DE POKTIBUS two piuioiis of each pair, on account of their being piuctd so close logelher. With the eqiiftlizing arrangement just specified, it is legiti- mule lo iissunie an equal division of work among all the pillions I hat engage the rack. No matter what mechanical power be used, all spans must he provided also with hand-operating machinery. le' on •METHOD OP DETERMININQ POWER REQUIRED FOR OPERATING THE SPAN AND LIFTING THE ENDS. The power required for turning any span is to be deternuued by the following formula. (1) H.P. = .01 25 IFp 550 where W ~ total load on rollers in pounds, and « = velocity oil pitch-circle of rack in feet per second. The value of v is to be determined by the formula V ■— itD 4^' where D = diameter of pitcli-cirele of rack, and t = assumed time in seconds for luriiliig tlie draw tiirough one fourth (j)of a revolution. This method gives the power required under ordinary conditions; but it is always necessary lo figure also the power required to open the span against an assumed un- balanced wiiul-piessiire. This is to be delermined as follows: The unbalanced wind-pressure on one arm is to be lak(!n at five (5) pounds per square foot of the exposed surface of the floor and both trusses. Let P = total unbalanced wind load on one arm in pounds, p.nd V = velocity of travel of its centre of pressure iu feet per second; then (8) H.P. = 650' SPECIFICATIOKS Poll UAlLROAl) DIIAW-SPANS. 203 The value of v is to be deterniiued by assuiniug iv certain time t, in seconds, for turning the draw one foiirtli (}) of a levolulion. Let I = distance in feet of tlie centre of pressure on one arm from the centre of tbe drum; then (8) Ttl 2t' For Mechanical-power Turning macJiinery tbe greatest H.P. required is to be determined us follows: Case 1. — {(i) By Formula (1) determine tbe il.P. required for turning tbe span in the least time in which it is probable thai the said span will evjr need to be opened. Case II.— {a) By Formula (1) determine the H.P. required for turning tbe span in twice the time assumed in Case I. (b) By Formula (3) determine the H.P. required for ojieratiiig the draw against tbe unbalanced wind load in twice tbe time assumed in Case I, and add together the two amounts of H.P. determined by (a) and (6). Tbe sum will be the greatest H.P. required for Case H. The greatest pressure on the teetli and toision on shafts found for these two cases is to be used, the metal being strained on tbe extreme libre as hereinafter specified; but the said teetli and shafting must also be figured on the iissuinpiions thai the entire availaljle capacity of the machinery is recpdred ni<:rely lo hold the draw from turning under an exx'ssive unbalanced wind-pressure, and that under these conditions tbe metal is strained twice as high as hereinafter specified. For Hand 'luniliin inachinery the H.P. required to turn the span in tbe leant time in which it is probable that It will ever need to be opened by man-power is to be found b}' the formula previously specified; theu tlie number of men re- (pnred to perform this work is to be deterniiiud by assuming that six (0) men are equivalent to one H.P. In proportioning all parts of the Jiand-operating macihinery there slndl be assumed on the levers as many men as are required l)y the above method, each man exerting a horizontal thrust of one hundred and twenty (130) pounds. Under such conditions the S04 I)H I'OXTIBUS. metal is to be strained the same us lioreiiiaftei' specified for macbinery operaled by mechanical power uuder ordinary con- ditions. DETAILS OF MACHINERy. OPERATING MACHtNERY. All gear- wheels are to be of cast steel with cut gears. To (letenuine the size of any gear-wheel, the tooth pressure on the pitch-circle is first to be found as follows : For gears moved by mechanical power only, where II. P. ■= horse-power to be transmitted by gear, v = ve- locity in feet per second at its pitch-circle, and P=: tooth- pressure. For gears moved by hand-iwwer, P = 120NM, where N= number of men, M = multiple of lever over gear under consideration, and P — tooth-pressure. Having thus determined the tooth- pressure, the pilch can be foimd by the following formida ; p = .035 \'\P, for gears in which the face is equal to 2| times ti e pitch, where p = pitch, and P— total tooth-pressure. Til is allows an extreme fibre-stress on the teeth of eight thousand (8,000) pounds per square inch, which is to he the standard intensity for all teeth under ordinary conditions of operation. Beve'-gears are to be consiilercd as only three fourths (5) as strong as spur-gears of the same pitch and face. The use of bevel-gears with very thin teeth will not be allowed, even though they be of standard pattern ; but special bevel- gears with thicker teeth than usual will have to be manufac- tured. SPECIFICATIONS FOK RAII.KOAI) DRAW-SPANS. J^'Oo All gears are to be key-seated ami fliiished in accordance with the practice of llie best machine-sliops. All pinions gearing into the rack and into the large spiir-wiieels are to be shrouded on top, and tlie extra sirengtii obtained by this shrouding is not to be counted upon in pro[)ortioning the size of the teeth of the pinion. All shafting is to be of cold-rolled steel, and is to be provided with couplings, collars, and keys for gears. All couplings must be strong enougli to develop the full strength of the shafting, and must be keyed to the sumo, tl;inge-couplings being preferred. All couplings are to be placed as near the bearings as praciicable. Suitable collars are to be u.sed wherever they are necessary to hold the shafting from moving longitudinally. Tlie greatest allowable length .of any shaft between centres of bearings is to be determiuel by the formula z = 75 v'f7^ where L — the iinstipported length in inches, and d = diameter of shaft in inches. The diameter reqtiired for any siiaft is to be determined by the following formula: = W IT' where rf = diameter required, H.P = the horse-power to be transmitted, and JV=the number of revolutions per minute. Tins will allow for all bending lliat will come on any well- designed and properly supported .'^hafl under ordiiuuy con- ditions, and provides for an extreme fibre-stress of about twelve thousand (13,000) pounds per squaie inch, under the assumption that the twisting moment and the bending moment are about equal. Every shaft, however, after being designed by the i)receding formula must be checked as follows, and if found weak must be pro|ierly strengthened either by increasiii:;- tiio difinu'tcr or by reducing the lever arm or arms of the bending mommU 200 I)E PONTIIJUS. First, laud the twisting inonifiit iiiul ihe Ijoiiding moment (including lliat ciuisi'd by the weiglil of llie sliiil't ilticll) by cnmpuling tlie toolb-piessure, wbidi is tlie foKx- producing directly these moments, culling the twis;ing iuomh nt T iind tlie bending moment M. The eqiiivrdent twisting monunt for !i combination of tlies-e two moments is given by llic ec^uation T == if + \/M-' + T\ where 7" is the equivalent twisting moment. The corresponding extreme tibre-stress is to be found by the equation 2" /=5.1 (/» wliere d is the diameter of. the shaft, and /is the extreme libie-slr(!'ss. This shoidd never exceed twelve thousand (12,000) pounds per scpiaie inch for all ordinary conditions of operation, or twenty-four thousand (24,000) pounds per square inch for the unusual conditions of the machinery stalled by the unbalanced wind-pressure when working at its utmost capacity. In no case is any shaft of less than two and one-quarter (2|) inches in diameter to be used for any part of the machinery of draw-spans. Siutable cast-iron boxes are to be provided for all bearings. All boxes, bearings, couplings, collars, etc., are to be nuide in accordance with the best machine-shop practice. Tiie boxes for the liiie of shafting running to ends of span are to have wooden shims beneath them so that the shaft can be aligned perfectly after the span is swung. The hand-power turidng-machinery is to bo so arranged that the levers cnn be conveniently applied to slnifls near the centre of s]ian for bolli the turning and the end-lifting machinery. Shafts must also be provided for apjilying the hand-power levers to the end-lifting nuichincry at eadj end of the span. Suitable hand-levers arc to be i)roviik'd for as many men as are required for operat ng the draw. These levers are to be constructed entirely of steel, ii.\cej)ting only SPECIFICATIONS FOR RAILROAD DRAW-SPANS. 207 the small wooden quarter-rounds at tlie ends by which the men tuUc hold. All machinery shall be so arranged that the span can bc turned completely around in cither direction, and so that it is reversible in every partictdar. KND- 1.1 KTINO AI'l'AUATUS. The ends aie to be lifted and locket' by means of a toggle mechanism to be operated by screws at each end of tlie span. The entire machinery is to be made strong enough, with the previously "^[lecilied intensities of working-stresses, to exert an upward force on each end of each truss equal to the assumed uplift in case of mechanical power ; or to transmit to the end rollers the greatest force that the men can exert on the hand levers, assuming that iis many men will be applied then to as are nqiiired for the turning-machinery, and that each man exerts a horizontal thrust of one hundred and twenty (120) pounds, straining the metal the same as in the case where tiie power is mechatdcal. In case of mechanical power, all the teeth and shafting must also be ligured on the assumption that the en lire available capacity of the machinery is required merely to start motion, and that under this condition the metal is strained twice as high as herein specified. The size of screw required is to be determined by the following formula : where d = diameter of screw at base of threads, and P = axial pressure on screw. The axial pressure is to be determined for the two following cases, the greater pressure thus found being adopted; Case I.— _ 2iih^ h ' where Ii= total assumed upward reaction at one end of span, // = K'eatest rise of ends when end lifts are applied, and h = travel 0/ iiut on screw juecessarj to produce the rise A'. 208 Di; lOXTIBUS. The factor two (2) is used to allow one hundred (100) per cent for fiiclion. Case 11. — P = 80^i^, where M = the number of pounds pressure the screw will exert for one pound applied on the lever, and N = number of m. II on said lover. By using eighty (80) instead «)l' one hundred and twenty (120) in the above formula, tliere is made an allowance of tliirty-liiree and a tinrd ('i\i^) per cent for friction, which is certainly lower than it will ever be under ordinary wurLiug conditions. Assundng the coelhcienl of friction low in this case nnd = 10,000 - ?^~, z where I = greatest ui\supporte(l distance between tillers, ex- cept in links in which only one filler is used between two fiats, when il is to be taken as the entire distance from centre to centre of end- pins, i = thickness of each link, and p = the in- tensity of working compressive stress. In no case is the diameter of any pin used in a toggle to be less than two and a half (2^) inches. Rail lifts arc to be provided in connection with the euil- lif;ing toggle, and the mechanism therefor is to be so designed tliat the rails will not begin to rise until the end rolh'is have been drawn from their bearings on the end shoes The rails shall be lifted so as to clear by one (I) inch all parts over which they mu.st jiass in ttirning, under the assnmpiion that the temperature of the top chords is higher by tiurty (30) de- jprOes Fahrenheit than that of the bottom chords. SPECIFICATIONS FOR UAILUOAU DRAW-SI'AXS. 209 aU- Suitable guide-chairs for tKc rails near the cuds of tlio span are to be provided beneath the same ou at least flfteeu (15) ties from each end of the span. These chairs must be either spiked or boUed to the lies, and must hold tiie rails tirmly in place. Guide-rods such as are employed in ordinary switcli- work are to be used every six feet between the portions of the rails resting in the guide chairs. Tlie strength of all parts of the rail-lifting machinery is to be determined by computing the force necessary to deOect the two rails the required amount in a distance of twenty {20) feet, and adding tifty (HO) per cent thereto for friction. If considered necessary for any particular span, latclies are to be provided for holding the ends in place ; but under ordi- nary conditions the track-rails and the end rollers are all that will bti recpiired. In double-track drawbridges special attention must be piiid to the designing of not ordy the lifting-gear, but also the trusses themselves, in order to ensure that, under the most unfavor- able circumstances possible, there shall be no lifting of the ends of trusses off their supports. If such a lifting were pos- sible, the result would certainly be the derailment of an enter- ing train, and consequently disaster to the span. To prevent such uplifting the trusses must be deep and very rigid, and the lift of the ends roust be from one (1) to two (3) inches, ac- cording to the length of the span. 8HOUB AND END-BEAKINQ UOLLEKS. Rollers are to be provided beneath the end-pins of trusses and attached to the span by means of links which form a part of the toggle. The rollers must be bored so aS to fit ever the pins at the bottom of the links. Both the pins and the inside of the rollers mu'^t be finished very smooth; and provision must be made for oiling I he bearings between them. Tim allowable intensity for bearing between rollers and pins shall be ten thousand (10,000) pounds per square inch of horizontal projection of pin inside of the roller. No roller shall be less than six (6) inches in diameter, and the pins in:9{de of st\me shall not be less than three and seven- 210 DE rONTIBUS. sixteenths (J},''^) iiicUes net in diunieler. Tliu pluy between rollers and their pins shall not be over one tliirty-seeond (j,'g) of an inch. The links forming tlic support for the ends of trusses are to be proportioned by the formula p = 10,000 - 300 where p = intensity of working compressive stress, I = greatest unsupported length of one link, and t = thickness of same. In all drawbridges where, on account of infrequent opera- tion combined with great changes in temperature and great length of arms, there is a tendency to drag llie rollers longi- tudinally on llieir bearings, the detailing of the link siippoils must be sucii as to provide sulHcient rigidily to overcome the fricti(m of tiie rollers on their bearings, and thus permit the lifting apparatus to accommodate itself to extreme changes of temperature without overstraining any of its parts. Tbe bearings for rollers on the shoes shall be cup|)ed one- eighth (|) inch or more in depth so as to provide ample bearing area, using an intensity of ten tliou.^and (10,000) poimds, impact being included in the caiculaled load. The shoes to receive the end rollers may be made of either cast or structural steel, and are to be anchored firmly to the masoiny The two shoes at one end of span are to be connected (o each other by means of adjustable rods not less than one and one- half (li) inches in diameter, and strong enough to take up the entire thrust from the toggle. Shimmiug-plates varying in thickness from one fourth (i) to one half {\) of an inch and of a total depth of not less than three (3) inches are to be used bencnth the shoes so as to pro- vide adjustment for the ends of the span. Shoiilders must be provided on the slioes to furnish a bear- ing for the rollers when they are lowured l)y tiie toggle. Each shoulder must be turned so as to fit the roller exactly, when the axis of the pin through the said roller is in the vertical plane of the truss. The height of these siioulders above the bottom of the rollers shall be about one ihird of the diameter SPECIFICATIONS FOB RAILROAD DRAW SPANS. 211 of mU\' rollers, but never enough to involve Iho jxtssibility of collision with the draw-spun during ils revolutir)n and when tlie lop chords thereof are thirty {',]()) degrees Fahrenheit wanner limn the bottom chords. All purls of tiie end-lifting niuchiuery must be finished in ucconluiice with the best niachinc-shop practice, and all sliding surfaces shall be provided with oil-holes that are easily accessible. In all cases end tioor-beanis with double webs shall be us-ed, in order to provide proper support for the end-lifting machinery. Whenever spans are to be floored for highway traflic, all koyliolcs for applying hand-levers are to be provided with suitable cast-iron caps. Whenever practicable, the end-lifting toggle machinery is to be assembled in the shops to make sure that it will work satisfactorily. HOUSES AND SUPPORTS. Wherever mechanical power of any kind is to be used for operating any draw-span, a suitable house is to be provided for same. The size of the house required will depend upon the kind of power to be used, and the amount thereof. All l)!irts of the house shall be durable and strong, and shall be finished in a first-class and workmanlike manner. A sufHcieut number of windows is to be put in to light properly all parts of the building. The house shall be placed high enough in tlie tower to give the required clearance beneath its supports, and, where shallow trusses are used, it shall be placed entirely al)ov(! the span. The supports for the house shall be designed to curry the weiglil of the latter and that of all machinery to b(! placed therein, together with a proper allowance for live load. In general, steel beams shall be used for llie joists sup- l)()rtiug the floor, and all parts of the latter shall be made strong enough to carry three hundred and fifty (850) pounds per square foo*. The weight of the house and its machinery must always be 212 I)l<; I'ONTIHUS, considered in pioporlioning all parts of the 8tru(!lme which will he iiffuctt'd by Iheso loads, whether the span is to he pro- vided with niecluuiieal power at first or not, as it may become necessary later on to put it in. The wind load on the house imist also he considered in proportioning the tower posts and all bracing between them. CAMBER AND DEPLECTION. The lengths of all truss members shall be such that when the assumed uplift is applied at the ends of the span, and when the greatest live load is on the structure, the centre lines of the bottom chords from end to ind of span will lie in a horizontal plane. Tiie vertical- movement of the ends, from the condition of no stress in the chords, vvlien the weight of the finished span is supported on the falsework, to the condi tion of the span swung, must be very carefully figured, as upon this will depend the camber increments or decrements in lengths of members, the cleaiauces, adjusHnenls, etc. CHAPTER XVI. GENERAL SPECIFICATIONS GOVERNING THE DESIGNING OF STEEL HIGH WAY BRIDGES AND VIADUCTS. GENERAL DESCRIPTION. CLASSIFICATION. Hroirw^AY bridges sball be divided into three classes, viz., Class A, wLicb includes those that are subject to the continued application of lieavy loads ; Class B, which includes those tliat lire subject to the occasional application of heavy loads ; and Class i), which includes those for ordinary, light traffic. In general it may be stated that bridges of Class A are for (li-nsely populated cities ; those of Class B for smaller cities and manufacturing districts ; and those of Class C for country roads. MATERIALS. All parts of the structure, excepting the flooring or paving. ahnll, for all spans of ordinary lengths, be of medium steel, excepting only that rivets and bolts are to be of soft steel, and adjustable members of either soft steel or wrought iron. For very long spans high steel nuiy be used for top chords, in- dined end posts, pins, and eye-bars in botloni ciiords and in main diagonals of panels where there is no reversion of stress when impact is included. Cast iron will not be allowed to be used in the superstructure of any highway bridge or trestle, except for purely ornamental work, cast steel being employed wherever important castings are necessary. 313 214 DE PONTIBUS. JOTST8, PLANKS, GUAHD-TIMBEUS, AND WOODEN IIANDKAILS. Joists, plunks, gutird-mi's, haud-rails, and all otbir tiiiiber portious of the structure shall be of long-leaf, Southern, yellow pine, or otlier timber which, iu the opinion of the Engineer, is eqimiiy good and serviceable. The sizes of the limber joists shall be sucli as to give the requisite resistance to bending, the elTocl of impact being con- sidered ; but no joibt shall be less than three (3) inches wide or twelve (12) inches deep. As a rule the depth of a joist shall not evceed four (4) limes its width. Otherwise, the joists shall be properly bridged at distances not exceeding eiglit (8) feet. They shall be proportioned by the formula o wiiere M \h the greatest bending moment in inch-pounds upon a joist, H is the intensity of working-stress in pounds, // the width of the joist in inches, and d tlie depth of same in inches. Joists shall be dapped at least one-half (^ inch upon their bearings, and shall have their tops brought to exact level before the planks are laid thereon. They shall be spaced not to exceed two (2) feet between centres; shall, preferably, lap by each other so as to extend over tile full width of the tloor-beam ; and shall be separated half an inch, so as to permit the circulation of air. The out- side joists, however, shall abut so as to provide; flush surfaces from end to end of .span. Floorplaidis for the main roadway shall be at lensl three (:!) inches thick and from eight (H) to ten (10) inches wide, and shall be laid with one-quarter (|) inch openings. Each plank shall be spiked to er.ch joist on which it rests by two (2) seven (7) inch cut spikes, the holes for which shall be bored in order to avoid splitting the timber, or else by two (2) seven (7) inch wire uails. SFKCIFICATIONS FOli STEEL HIGHWAY BRIDGES. 215 Whenever n wearing-floor is used, the lower plunks in\ist be planed on the upper side and sized to a uniform thickness, and the wearing-floor nuist be planed on the lower side so us to ensure a perfect bearing between upper and lower floors. Floor-|)laidvs for fooiwnlks shall be at least two (2) inches thick and not nuich more or less than six (6) inches wide, and shall be laid wilh one-half (I) inch o|)enings. Each of said plauks shall be si)iked to each joist upon which it rests b}' two (3) six (()) inch cut spikes, the hoUs for same being bored. All planks shall be laid with the heart side down. Tliere shall be a wheel-guard of a scantling not less than four (4) inches by six (6) inciies on each side of the roadway to prevent wheel hubs from striking the trusses. It is to lie laid on its flat, and blocked up from tlie floor by shims at least one (1) foot long, six (0) inches .vide, and two (2) inches thick spaced not more than seven (7) feel between centres, each sliim lieing sj)iked to the floor by four (4) four and-a-half (4j) inch cut spikes. The guard rails are to be bolted lo the floor through the centre of each shini by a tlnci'Mpmrter (J) inch bolt, which nuist also i)ass through the joist beneath. When the guard-rails are bolted to the wooden hand-r.iil posts, the bolt-heads are to I)e countersunk into the guard-rail, so as to make a flush surface on the inner face of same. The joints in the guardrail are to be lap-joints, at least six (0) in(rhes long, each located symmetrically over the niiddli! of a shim. When a bridge is on a heavy grade, the inner, upper corners of the guard-rnils are lo l)e covered with steel angles fastened to the timber by counlersiudv screws, spaced about eighteen (18) inches apart, so as to protect the guard-rails from the injurious eflfecls of using them instead of wheel-brakes foi- heavily-loaded wagons. Wln.n wooden hand-r.uls are employed, they are to be nuule of piiu!, the posts being 4" X 6" X 4' 6'' to 5', with two (2) runs of 2" X 0' timbers— one on its Hat and the other below on edge to support the flrst for a hauil-rail — and one (1) run of 2" X 12" hub-plaidi. The posts are to be spaced not to exceed ten (10) or, prefer- ably, eight (K) feet apart. The hand-railing is to be flrmly 21(5 DE PONTIBUS. attached to the bridge, and rigidly binced. When tlie rigidity of a band-railiug is dependent upon that of the outer joists, the latter must be properly bridged aud stilt'ened. Any other wooden haud-railiug of equal strength and rigidity, and which is satisfactory to the Engineer, will, however, be accepted. When iron hand-railing is employed, it is to be of a linn, substantial pattern, pleasing to the eye, and rigidly attached to the trusses or floor-beams. Both through and deck bridg jf "o to be provided with a hand-rail on each side, not less Uum three and a half (3^) feet high above the floor. lu case there be any liability of a horse jumping over this railing, its height nuist be increased to four and a half (4^) or Ave (5) feet. There must be a hand-rail on the outside of eao'.i sidewalk, nut less than three and a half (3}) feet in height above the floor. Fr.OOKINO ON APPUOACHES. All floor-timbers, guards, and railings sliiiU extend over all piers and abutments, and make suitable connection with the embankments at the ends of the 8tru(;ture. Aprons or rover- joints of steel plate shall be provided at the ends of spans, if required. The floors of the sidewalks shall e.xtend to and con- nect with the floor of the main roadway, so as to leave no open space between them. BTUKKT-nAHiUOAD TKACKS. Should 'theie be one or more street-railroad tracks crossing the bridge, ihere must be directly under each mil a joist or stringer, properly proportioned to resist the effect of the total maximum load on the rail ; and the bending eilect of the concentrated loads upon the tloo; beams must be duly con- sidered. The rails shall be So laid as to ofl"er as little obstruction as possible to the wheels of vehi(;le.s. ' PAVKD FLOORS. Where paved floors are adopted, tht pavement shall be of the best of Its kind, and shall be built according ti. 'he latest SPECIPICATIOKS FOR STKKL llfGHWAY HUIDaES. 21^ and most approved specifications. Piived Hoors (ire always lo be supported by steel slringc^rs, preferably of rolled I beams, spaced generally not to exceed three (3) feet six (6) inches be- tween centres. For asphalt or stone-blook pavements, a buckleil-plate tloor, with coucrele thereon, f>h:iil be used. Tlie surface of the pavement must be thoroughly drained so as not to retain water, and the upper surface of the buckled plate, before it is covered with the concrete, must be protected from rusting by a liberal use of the best obluiiiable preservative coating. When wooden-block paving is adopted, it may rest on a timber floor from four (4) to five (5) inches thick, which in turn rests on and is spiked to timb'-r shims that are bolted ef- fectively to the stejl stringers. All paved floors must be pitched so as to drain transversely to the structure; but plank floors need not be pitched, as the water will druiu through the quaiter-iuch openings. CLKAKANCKS. Tiie smiillest allowable clrar roadway slmll be twenty (20) feet, measured between inclined end posts, excepting for cheap (!ountry bridges, where it may be reduced to eighteen (18) feet, or even to fourteen (II) IVct, in case that the bridge be so short that no piovision need be made for teams passing thereon. The smallest allowable clear headway shall be fourteen (14) feet, except for bridges in (;ities wliere the ordinances require a greater height. The corner brnckels may, however, en- croach on the specified clear headway, provided they do not extend either laterally or downward more than five (5) feet. KFFRrTIVK I,KN«TI!S AND OBl'THS. See Specifications for Railroad Slnictures. BTVLES OF BHIDOKS FOU VAltlOUS SPAN LEN0TU8. In general, spans of and below twiiily (20) feet are to con- sist of rolled beams or simply wooden "jists; spans from 218 DK PONTinUS. twenty (20) to sixty (60) feet, of plate gink-rs, spans from sixty (60) to ninety (90) feet, of open- webbed, riveted girders of single cancellation, or pin-connected "A" trusses; and spans exctediug ninety (90) feet, of pin-conuected trusses. ' I' :.«' of pony-truss bridges of any kind is prohibiiod, ex- cept! ly half-through, plate girder spaos, in which the lop flaL^ i are held rigidly in place by brackets riveted to cross-girders that are spaced generally not to exceed fifteen (15) feet apart. FORMS OF TKU88E8. The forms of trusses to be used are as follows : For pin-counecled spans up to ninety (90) feet, the "A" truss. For open webbed, riveted girders, the Warren or Triangu- lar g'nler, with verticals dividing the piinels ; also tiie Pratt truss. For deck-spans carrying joists on tlie lop chords, liic Warren or Triangular girder with verticals dividing tiie panels of the top chords. For spans between ninety i9()) feel and about two hundred and fifty (250) feet, Pratt trusses willi top cliords either straight or polygonal. For spans exceeding two hundred and fifty (250) feel, Petit trus.ses. It is undeistood that these limiting lengths are not fixed absolutely, as the best limits will vary somewhat with tiie width of bridge and the live load to be carried. MAIN MKMHKRB OF TKUH8 HUIDOEB. All spans of every kind shall hav(! end tloor-beams, riveted rigidly to the trusses or girders, for supporting the joists or stringers. Steel stringers are, preferably, to be riveted to the webs of the cross-girtlers, but wooden joists are generally to rest on top of the latter. lu geuer.il, all lru8.ses shall have main cud p(^sts inclined. SPPXIFICATIOJ^S FOR STEEL HIGHWAY imiDGES. 5il9 All trusses shall be so designed ns to lulinit of uccurate calculiitious of all stresses, excepting only such uuiniportunt cases of uiublguity us occur whea two stiff diagonals are used iu a middle panel. lu important bridges with steel stringers, all lateral bracing and other sway-bracing shall be rigid above and below ; i.e. the sections must be capable of resisting compression, adjust- able rods for such bracing being allowed only iu towers of draw-spans and in the lower lateral systems of deck bridges ; but, in cheap country bridges, the lateral and other sway diagonals may be adjustable rods. Tiie stiff tliagonals of lower lateral systems, wliich shall be of double cancellation, shall be riveted rigidly to all the steel stringers where they cross them. In the trusses of important bridges counterbracing the web shall be effected by using stiff diagonals, hut in cheap bridges it may be done by using counters of adjustable rods. All through-spans shall have portal bracing at each end, carried as low as the specified clear headroom will allow. The portal struts shall l)e riveted rigidly to tlie web or both flanges of tiie incline 1 cml yjosts. liiveling portals to one flange only will not lie allowed. Wlien the height of the trusses is great enough to permit, transverse, vertical sway-bracing shall be employed ; other- wise, corner brackets of proper size, strength, and rigidity are to be riveted between the posts and the upper lateral struts. Deck-bridges shall, as a matter of precixulion, have sway- diagonals between opposite vertical posts of sutticient strength to carry one half of a panel-truss live load with its impact allowance ; and the transverse bracing between the vertical or inclined posts at each end of span shall be sufliciently strong to transmit properly to tiie masonry one half of the total wiud-prtssure carried by tlie upper lateral .system of the span. The lower lateral systems of deck-bridges may be made of adjustable rods in alternate panels, thus leaving every other panel unbraced, and forcing the wind-pressure from below up 320 t)Vl POXTIBU.'^. the verticnl braciug ;iiid to the ends of the span by the upper latenil system. In important bridges, suspenders or hip verticals and two or more panel leugtiis of bottom chord at each end of span shall, preferably, be made rigid members, except that eye-bars are to be used i )r bottom chords of " A " truss bridges. Ail Hoor-bean s are to be riveted to the truss-posts in truss- spans, excepting in the case that eye-ltars be used for suspend- ers or hip verticals. In such cases 11 'or-beam hangers may be used, provided Ihey be made of plates or shapes, and that they be stayed at their upper ends against all possibility of rotation. CONTINUOUS SPANS. See Siieciflcations for Railroad Structures. TUESTLE- TOWERS. In general, the descriptive specifications for railroad trestles are to ])e followed in designing highway trestles or viaducts, except that in cheap s'ructures all sway-diagonals of towers may be made of adjustable rods, with hori/ontiil struts at the panel points, provided that the struts be rigidly riveted to the columns. CAMBER. All trusses must be provided with such a camber that, with the heaviest live load on the span, the total camber shall never be quite taken out by detlection. With parallel chords, sufficient camber will be obtained by making the topclioid sections longer than the corresponding bottom-chord sections liy five thirty-seconds (/j) of an inch for each ten (10) feet of length. Phite girders and shallow, open-webbed, riveted girdeis should m)t be given any camber. EXPANSION, ANCHORAGE, AND NAME PLATES. See Specifications for Railroav system and primary-truss members are to be figured for these electric train loads when passing either the road-roller or the heavy wagon-load; and the trusses as a whole are to be figured for a uniform load found by combin- ing the equivalent electric load, considering it to occu|>y ten (10) feet of roadway, together with its impact allowance, with the regular uiuform live load per square foot of floor on the remaining width of ciear roadway, together with its proper 1 r.pacl allowance, provided that the equivalent live load per lineal foot for the cars, plus the proper impact allowance, ex- ceetl %he regular live load for a ten (10) foot width of roadway, 324 |)K I'ONTIP.US. phis its proper impact iilluwancc. If it do not so exceed, the rcgidiir uniform live loud is to be employed. IMl'ACTAIiLOWANCE LOAD The impact-Hllowaiice load is to be a percentage of the uni- form live load, found by the formtda P = 10000 /. + 150' where P is the percentage and L the length in feet of spau or portion of span that is covered by the live loud, vvlien the member cjnbidered is subjected to its maximum stress. DEAD LOAD. See Specifications for Railroad Structure. WIND LOADS. For highway structures the wind loads per lineal foot of span for both the lojided and the unloaded clio.ds are to be taken from the curves shown on Plate VIII. This diagram was figured for a clear roadway of twenty (30) feet. For wider structures, the wind l')ads are to be iucrea.sed two (2) i^er cent for each foot of width in excess of twenty (20). The wind loads given on the diagram have been computed from detailed designs for simple spans up to seven liundred and fifty (750) feel in length, but beyond this limit they have been assumed; consequently, in designing spans of greater length than this, it will be necessary to check tlie assumed wind-pressure after the sections are proportioned, using an in- tensity of twenty-five (3.")) pounds per square foot. The intensities employed in preparing the curves varied from forty (40) pounds for very short spans to twenty -five (25) pounds for very long ones. For viaducts, the wind pressure on the empty strmture is to be assumed as three hundred (300) pounds per linea' foot on the spans at the level of the floor, and two hundred and fifty (250) pounds for each vertical foot of each entire tower, SPECIFICA1I0.S> FOlt HTKKI- lIKillWAY HKIDfJKS. 32.5 The wind loads for longitudinal brnciiii!: me to bo taken us seven tenths (0.7) of those for the transverse bracing. For vialucts carrying electric trains, tiie wind loads are to be taken from the Specifications for Railroad Sinictures. All wind loads are to be treated as inomng 'ouds. INDIIIECT WIND LOAD OR TltANHFKIlRED I-OAD. See Specifications for Railroad Structures. TKACTION liOAD. See Specifications for Railroad Structures. CKNTIJIFUGAL liOAD. See Specifications for Railroad Structures. EFFECTS OF CHAN(}EB OF TEMPEKATURE. See Specifications for Railroad Structures. INTENSITIES OF W^OBKING-STRESSES. See Specifications for Railroad Structures. HEARINGS UPON MASONRY. See Specifications for Railroad Structures. REVER8ING-8TUES8ES. See Specifications for Railroad Structures. NET SECTION. See Specifications for Railroad Structures. BENDING MOMENTS ON PINS. See Specifications for Railroad Structures. COMBINATIONS OF STRESSES. The Specifications for Railroad Structures under this head ing are to be followed, with this exception : in bridges and viaducts that do not carry trains, the live load and the wincj load are assumed not to act simultaneously. 226 |)K I'ONTIIHS, BKNDINO »)N Tor CHOUDH. See Specirtcatioiis for Railrojul Striu;tiire.s. UGNDINO ON INCLINKD ENO I'OHTH. The Speciti(;HliouH for Kitilroiul Blnictiires under tliift head- ing lire to be follo'wed, witli tliis exception: in bridges that do not carry trains, the live iimil and the wind load are assumed nut to act simultaneously. BENDING DUE TO WEIOHT OK MEMUEK. yee iSpeeificalions for Railroad Structures. OENEUAIi LIMITS IN DKHIONINO. The folliiwing geiural limits shall be adhered to in design- ing highway bridges and viaducts : The perpendicular distance between central planes of trusses shall never be less than one twentieth (j'^) of the span. The length of any bracket cantilevered beyond a truss or girder shall never exceed one Iialf of the perpendicular distance between the central planes of adjacent tiusses or girders, unless there be more than two trusses to the span. No metal less than five sixteenths (/g) of an inch in thick- ness shall be used, except for filling-plates; and in important bridges this limit shall be increased to three eighths (g) of itn inch. The least allowable thicknesses of webs of rolled I beams shall be as follows: 24" I beams V*«" webs. 20 " " J 18 " " {g 15 " " f 12 " " }g No channel less than six (G) inches in depth shall be used exc pt for lateral struts, in which five (5) inch channels may be employed. No angles less than 2|" X 2^" X rs" shall be used except for lacing, SI'ECIKICATION.S FUlt STKKL IIKHIWAY HRIDCSES. 127 No cye-bui8 icHH ihuii thrvc (3) iuclieudcep or fiveuiglilhs (I) of Ml iiH'lt thick slitill be uinplo^ed; uiid the depths of eye-bars for chords and niaiu diugoimis shall uot be less than one sixtieth (^'g) of the lioii/.oittal length o( same. No Hiijustablv rod shall have loss than three quarters (J) of a stiuaru iu(;ii ul cruss-section. The shoilest span Icn^ih for trusses with polygonal lop chords shiill be one hundred iind sixty (IGO) feet. The limit of span length in wiiich stct'l stringers can be riveted continuously from end to end o' span shall be two hundred (200) feet. Beyond thin Hunt slidiugbearings must be used at one or more intermediate panel points; and in no span shall there be a length of coniinuously riveted stringers exceeding two hundred (200) feet. For all compression- members of trusses and for columns of viaducts the greatest ratio of unsupported length to least radius of gyration shall be one hundred and twenty (120). excepting those members whos»^ main function is to resist tension. In these the limit may be raised to one hundred and fifty (150). The corresponding limit for all struts belonging to sway- bracing shall ale') be one hundred and fifty (150). OENERAIi PBINCIl'LES IN DESIQlJING ALL HIGH- WAY STBUCTUREt>. See Specification for Uinlroad Structures. RIVETING. In general, the specifications for rivt ting given for railroad structures shall apply also to highway structures, except that in the latter tlte diameters for rivets may be reduced to three quarters (J) of an inch for ordinary work. PETAILS OP DESIGN FOR ROLLED I-BEAM SPANS. Rolled I beams used as longitudinal girdi-ns shall have, preferably, a depth not less than one fifteeulh (^b) of the span. Tlicy shall be proportioned by their moments of inertia. The spacing shall generally not excee(| three (3) feet six (6) inches. S'SH DK PONTIHUS. Provided Unit woe den shims be boiled to the top tluiigcs for spiking the plunks iliereto, no swuj'-bracing will be retjuircd ; but otherwise it must be ustd. Eucli I beam is to )>iive at eacii end n pair of stiffening angles, titling tigiitly at both top and bottom to the tlanges, to carry the load lo the nnisomy and to form part of the end braciug-friiines. Each i)air of girders is lo have a bracing-frame at each end; and under each end of each 1 beam there is to be liveti d a l)earing-plate of proper area and thickness (never Kss than live eighths [^] in(;li) t.» distribute the load unifonnly over the masonry, 8aie stiffened. CounkTs, wium employed, can be of either rounds, squares, or tliils. Tliest! mid all other adjustable nu-nibcrs are lo lia\e their ends enlarged for tlie screw-threads (unless soft steel, cold-pressed threads be nsed), so that the diameter at the bot- tom of the thread shall bo one eighth {\] >f an inch gr.aler than tiiat of the body of a rour.d rod of area eqnal to tliat of the adjustable piece. Diagoinils for upper lateral systems and vertical sway- bracing siiall, preferably, be l)uilt of foiw angles in the form of an I, with a single line of lacing; but for structures where this section would involve an extiavagani use of metal, two of tlie angles, one at tlie top and one at the bottom, may be onutted, thus making each strut consist of two angles laced, provided, of conrse, th;it where the struts cross they shall lie rigidly connected by two plates of ample size. This unbal- anced section for such diagonals is to be avo'ded whenever it can be done without undue use of nu^al. Ii. ,o case, though, 230 DE PONTIBtJS. will it l)e permissible to use angles in tension that ure not capable of resisting properly Uie possible compressive stresses, with due regard for the specitied limit of ratio of uiisup- polled length to least rn.lius of gyration. In cheap highway bridges the lateral diagonals may be made of adjustable reds with right and left clevises at their ends, by wiiich they are to be connected through pins to corner- plates that are riveted to both the lateral ijtrut and the truss member. The ordinary detail consisting of two or three short pieces of angio riveted on top of the cover-plate, and between two of which the rod lies, will not be permitted. Where adjustable rods are employed, the struts to the ends of which they attach must be figure 1 for a total compressive stress e(|ual to the suiu of the components (in the direction of said strut) of the greatest allowable working-stresses on all of the adjust- able rods meeliug at one eiul of said strut. While this method gives an excessive stress for llie strut, I lie etfut will be a desirable error on the side of safety and riiiidity. In designing transverse lateral and overhead struts and their connections, it must be remembered that their main function is to hold rigidly the chords or posts to jdace and line, and not merely to resist as columns the greatest calculated direct stresses to which they nniy be subjected. For this reason such st'^uts should have ample section for righlity, and the conrcctiug plates at their ends should grip both connected m« mbers effectively. Where built stringeis are used for the floor sysleui, they shall be made without cover-plates, and generally of the economic depth in respect to total weight of metal, hut never less in depth than one flfleeuth {j\) of the span. No splices w!ll be allowed in their flanges nor any in their webs, pro- viiled that sutticiently long web-plates are procurable. The compressioiitlanges shall be made of the same gross section as the tension-tlanges, and they shall be so stiffened that the tinsupported length shall never exceed sixteen (10) times the width of tiauge. Rigid diagonal bracing of angles is to be used between the top tlanges oC such stringer-, luiless they be held rigidly in place by the dooring; and rigid bracing-frames V r n ti sPK(;iprcArroNrs forstkml Hr(ji£\VAY hhidoes. 2;]l TllL- are to be oniplryed between tbe ends of adjiKent stringers at all expansion points. Wliere sucli stringers are used, tbe lower lateral system must invariably consist of rigid sectiotis, eaeh piece being riveted to eiicli stringer wbere it crosses tbe same. Ill respect t'> stiffening angles for siringers, tbe rules govern- ing Ibose for pl.ile-girdcr spans are to be followed ; but tbe end slill'eners are to be faced or otherwise treated so as to make the siringers of exact length liiroughout, and so as to effect a uniform bearing of tbe end slilTeners against tbe webs of the cross-girders. In respect to tlie proportioning of tlangcs and number of rivets required, llie rules given for plate-girder spans are to apply also to stringcns. The said rules are to apply to cross- girders, as shall also those relating to siitfeners, splices, cover- plate^i, and size of eoiupressiou-rtauges, that are given for plate-girder spans. Wiierever it is i' saiy to notch out the corners of the cross-girders to clea Iiords, the greatest cure must be taken to provide an adequaie means for trans ferring tbe shear to the posts without impair! t; ojther tbe strength or the rigidity. If necessary, in through-bridges, the web of the cross-girder may be divided into three parts so as to let the end portions project above tbe top fl mge and form brackets that will afford opportunity for using an ample iiun;- ber of rivets to connect to the posts, and will strengthen pr^ *- erly tbe otherwise weakened cross-girder. All plates, angles, and cliannels used in built members of trusses must, if practicable, be ordered the full length of tlie member ; otherwise the splices must develop the full strengii of the niemb(;r without any reliance being placed on tbe abut- ting ends for carrying compression. But in total splices at the ends of sections perfect abutting of the dressed ends is to be relied upon. However, the splice- plates even there must be of ample size and strength for both rigidity and continuity. The unsupported width of plates strained in compression, measuring between centre lines of rivets, shall not exceed tliirty-two (32) times their thickness, except in the case of 232 rK PONTIBUS. covei-plules for top chords imd inclined end posts, where the limii may be increased to forty (40) times the thiciiness. Where webs are built of two or more thicknesses of plate, the rivets that are used solely for making the several thick* nesses act as one plate shall in no case be spaced more than (12) inches from each other, or from other rivets connectiog said component thicknesses together. Tiie least allowable thickness for such compound web-plates shu'l be one (1) inch. The open sitlis of all compression-members composed of two rolled or built channels, with or without a cover-plate, shall be stayed by llf-plates at inds and by diagonal lacing bars or lacing angles at intei niL'diate points. Lacing-bars may be couneclod to the tianges by eilh.'r one or two rivets at each end ; but lacing angles, which are used for members of lieavy section only, must be (;onnecte(l by two rivets at eacli end. The tie-plates shall be placed as close as practicable to the ends of the coniprcssiou-niembcrs. Their thickness shall not be less than one tifiieth (Ko)of the distance between the centre lines of the rivets by which tliey are connected to the flanges, unless said tie platen be well stiffened by angles, in which case they may be made as thin as three eighths (!|)of an inch. The leiigtii of a tie-plate shall never be less than its wi»lth, or one and one-half (1|) times the least dimension of strut (un- less it be (;lose io a well diaphragm of the member, in which case it. may be lUiwie n- short as twelve (13) inches), and seldom grenicr than one ami one half (1^) times its width. Tlie thicknesses of lacing-bars shall never be less than one tiflielh (B'o)of the length between centres of the end rivets, measuring between inmost rivets in case that there be more than one rivet at each end. The smallest section for a lacing-bar shall be one and three- quarter (If) inches by five sixteentlis (/j) of an inch, which size may be used for channels und* r eight (8) inciies deep; and tiie largest section shall be tW" and a half (2^) inches by seven-sixteenths (/ff) inch, whieli size shall be used for chan- nels fifteen (15) inches deep. FK I'ONTIBUS. DETAILS OF DESIGN FOR VIADUCTS. Tlie specifications for ilie " Dctiiils of Design for Trestles and Elevated Riiilroads" are in general to be followed as far as lliey will apply in llie designing of highway viadiicis, l lie ]>rincipal variation being that, for clieap structures, adjustable rods with clevises nviy b • substituted for tiie stiff diagonals in tile four faces of the braced towers, by adding, of course, hor- izontal struts at the panel points of the transverse and longi tudinal bracing. The.se struts must be riveted to the columns by means of wide plates to which the clevises at'ach, and must never be pin- connected. Corner horizoital plates are to be employed for attaching the horizontal adjustable rods by means of clevises, each of said plates being riveted to both a transverse and a longitudinal bracing strut. The detailing for tlie loiigitudiiuil girders of viaducts and the bracing between same shall comply with the specifications for detailing highway plate or open -webbed, riveted girder spans; and the specifieat >hs for wooden floor system, paving, hand rails, etc.. shall be the same for highway viaducts as for highway bridges. CIIAPTEIi XVII. SPECIFICATIONS FOR HIGHWAY DRAW-SPANS. Thkse s|)eciflcation3 will be given i)riiicipall3' l)y reference lo the previous speciflciitions for Railnmd Structures, High way Bridges, and Railroad Draw-Si)ans. OENEBAL DESCRIPTION. CLASSIFICATION. See Speciticalioud t\>L- Highway Bridges. MATERIALS. See Speciflcatious for Railroad Dra\v-.>5pans. JOISTS, PLANKS, GCAKD-TIMIiERS, AND WOODEN HAND-RAILS. See Speciflcatious for Highway Bridges. FLOORING ON APrUOAClIES. See Specifications for Highway Bridges. HTEEL RAILROAD I'HACKS. See Specifications for Highway Bridges. PAVED FLOORS. See Specifications for Highway Bridges. CLEARANCES. See Specifications for Highway Bridges 237 238 PE TON TI BUS. EFFECTIVE LENGTHS AND DEPTHS. See Specifications for Hiiilruiul Structures. 8TYI.ES OF H1UI)(!KS FOU VAHIOIIS SPAN liKNGTHS. For spans up to one luimlred iind forty (140) feet in length, plate-ginicr spans are to be used. Thehc plate-girder spans may be made to act as continuous girders over tlie pivot-pier, or may have pin-conntclions over the drum, so that when the live load is applied they will act as two separate spans. The former style is generally i)referable as a matter of economy in time of operation, there being no important reason for raising the ends to any great extent, as there is in the case of railroad diaw-spans. For spans hetween one hundred and forty (140) and two hundred and twenty-five (225) feet, pin-connected Pratt trusses with parallel chords are to be used. For spans between two hundred and twenty-five (225) feet and three hundred (300) feet, pin-connected Pratt trusses with broken top chords are to be employed. For spans of over three hundred (300) feet, pin-connected trusses witli subdivided panels are to be adopted. It is understood that these limiting lengths are not fixed absolutely, ns the best limits will vary somewhat with the width of bridge and the live load to be carried. The proper truss depths for all ca.ses cannot well be specified, as tliey will depend upon various considerations, such as appearance, economy, width of structure, etc. In all cases the top chords are to be of rigid members, and inclined posts are to be used at ends and over drum, as specified for railroad draw-spans. MAIN MEMBERS OF TRUSS DRAW-SPANS. See Specifications for Highway Bridges. LOADS. See Specifications for Railroftd Draw-Spans. SPKCIFICATIONS FOR HKillWAV DUA WSl'ANS )iii\) LIVK LOAUB. See Specifications for Iligluvay Uridines ; imd, for llio manner «)f applying live loads to (Iraw-spans, see Specificatioua for Hailroad Draw-Spans. IMPACT- A LLOWANCK LOAD. See Specifications for Highway Bridges. DKAI) LOAD. See Specifications for Railroad Structures. ASSIIMKD UPLIKT HTHESSKH. See Specifications for Itiiiiroad Draw-Spans . Tlie inferior limit of uplift for designing tlie niacliinery of ligl.t iiigliway drawbridges is to be taken at ten thousand (10,000) pounds at each of the four corners o ' the span. "WIND LOADS. See Specifications for Highway Bridges. For method of s using tlic wind loads, see Specifications for Railroad Draw- j Spaus • INDIRECT WIND LOAD On TTtANHFRHKED LOAD. * See Specifications for Railroad Structures. For method of dealing with this load, see Specifications for liiiilroad Draw- Spans INTENSITIES OP WORKING-STRESSES. See Specific aious for Railroad Structures. Bh AMINOS UPON MASONUY. See Specifications for Railroad Structures. REVEHSINO STRESSES. See Speciticalious for Itailroad Structures. NET SECTION. See Specifications for Railroad Structures. _< 240 DE I'ONTIHUS. BENDING MOMKNT8 ON PINS. See Speclficulious for liiiilrotul Structures. COMIU NATIONS OK STUE88E8. See Specitications for Riiilroad Draw -Spans. It is to be observeii, liowever, that, for spans which do not carry trains, tlie live load and the wind load are ftssumed not to act simultaneously. KENUING ON TOP CHORUS. See Specifications for Railroad Structures. BENDINO ON INCLINED KN'D P08T8. The Specifications for Railroad Structures under this head- ing are to he followed with this exception : in bridges that ilo not carry trains, the live load and the wind loail are assumed not to act simultaneously. BENDING DUE TO WEIGHT OF .MEMBEH. See Specifications for Railroad Structures. OENERAIi LIMITS IN DESIONINO. See Specifications for Highway Bridges. GENERAL PRINCIPLES IN DESIONINQ. See Specifications for Railroad Structures. RIVETING. See Specifications for Highway Bridges. DETAILS OF DESIGN FOR PLATE-GIRDER DRAW- SPANS. The specifications for the corresponding item in tije Specifi- cations for Railroad Draw-Spans are to be followed, with the following exceptions : Ist. The perpendicular distances between central planes of girders will be made to suit the ^ leral requirements; and. IIK Si SPECIFICATION'S FOIt HI(ii..VAV l>l{A\V-SI'A NS. '341 2(1. Al least eiglit (8) poiiils of support on the drum will be needed. DETAILS OF DESIGN FOR PIN-CONNECTED DBAW-SPANS. The specifications for Ihe correspond inj? item in tlic Spcci- licalions for Hifj;iiw:iy Bridij:c.s are to be followi d, and in addi- lion tliurcto tiioso given umior the Inwling " Details of Des'gn forTrnssesof Draw-Spans" in llic Specilieiitions for Kailroad Draw-Spans are to bo employed, except that tiie use of adjust- able members for lateral diagonals will be permitted in the case of cheaj) highway draw-spans. DETAIL.S OF DRUMS AND TURNTABLES. Ill general the Speeilicalions for the corresponding item in the Specifications for Railroad Draw-Spans sliall be followed, ex(!ept that, for liglit, highway draws, tlie limiting thicknesses, etc., may be reduced to the following : Top flanges and webs of drums— three eighths (|) of an iiicii. Bottoffk flanges of drums— five eighths (f) of an inch. Upper track segments — one and three-cpiarter (IJ) inches. Lower tr.ick segments — two (2) inches. B:'aring-plates over drum — three (piarters (}) of an inch. Centre casting on pivot-pier — one (1) in(!h. Anchor-bolts for same — one and one-eighth (IJ) inches in diameter and two and a half (2i) feet long. Rollers — ten (10) inches in diameter and six (6) inches face. MACHINERY FOR TURNING THE SPAN AND LIFTING THE ENDS OF SAME. See Specifications for Railroad Draw-Spans. METHOD OF DETERMINING POWER REQUIRED FOR OPERATING THE SPAN AND LIFTING THE ENDS. See Specifications for Railroad DiawSpans, 242 DE PONTIBUS. DETAIIiS OF MACniNEBY. Ol'ti'* i'lNO MACIIINEIIY. See Specifications for .arale oi)eralion, by rolling an ingot and cutting off the scrap. The originid ingot shall have at least twice the cross-sectional area of the slab, and the latter shall be at least six times as thick as the plate. All finished material coming from the mills must be free from seams. Haws, or cracks, and must have a clean, smooth tinish. COMPOSITION OK ROIXRD BTEEL. The greatest allowalile percentages of certain principal in- gredients of the various kinds of rolled steel shall be as given in the following table: Iiigredi^iit.^. Percentages. Soft Steel. 0.5 ();i 0.04 0.04 0.00 Medium Steel. Hl«li Steel, T'hoHphoniR (acid steel). . I'liosplioiiis (Imslc steel). Sulphur Silicon JiutiKanese O.OC 0.04 0.05 0.05 0.70 0.07 005 0.06 0.80 34G Dli PONTlBfS. These percentages apply to drillings taken from the edges of plates and tlie exterior of shapes, bars, or Hals. If, how- ever, the drillings be taken from tlie middle of plates or the heart of other sections, the percentages given in the table are to be increased twenty-flve (25) per cent. IDENTIFICATION. Each ingot shall be stamped or marked plainly with its proper melt-number; and this melt-number must be stamped or pamted plainly on all blooms, billets, or slabs made from such ingots in order to identify the material throughout its various processes of manufacture; and the melt-numbiT must be stamped plainly on each piece of linishcd material. Rivet and lacing steel, and small pieces for pin-plates and stiffeners, may be shipped in bundles, securely' wired together, wiili the i)low or ■melt number on a metal tag attached. W' OKNKKAI. PKOVISIONS ON MKTUODS oK TKsriNO. Rivet-rods and other rounds are to l)e tested in the form in which they leave the rolls, without machining. Test-pieces from angles, plates, shapes, etc., shall be rect- angular in shape, with a cross-sectional area of preferably about one half (i) of a .square inch, but not less, and shall be taken so that only two sides are machine fiaislieil, the other two having the surface which was left by the rolls. Should fracture occur outside of the middle third of the gauge length, the test is to be discarded as worthless if it falls below the .standard., If any test-piece Itave a manifest flaw, its test shall not be considered. In case that one test-piece falls .slightly below the require- ments in any particular, the Inspector nuvy allow the re-testing of the lot or heat by taking four (4) addilioind tests from the .said lot or heat; and, if the average of the five (5) shall show that the steel is within the requirements, the metal nniy l)e accepted; otherwise it shall be rejected. te in ■' SPRCrFIOATIONS FOK STEEL IN BRIDGES, ETC. 247 Drillings for ciiemical analysis may be taken eitlier from tlie preliminary test-piece or from the finished malerial. The speed of the machine for breaking test-pieces shall not be less than one-qiiarler (J) inch per minute, nor more than three (8) inches per mlniile. Material wliich is to be used witliout annealing or furliier trL'atnu'nt is to b(^ tested in the condition in which it comes from the rolls. When thcmnterial is to i)<; annenled or ()liier- wise treated before use, the specimens representing such ma- terial maybe similarly treated before testing; but they shall also give standard elongation, reduction, and fracture before annealing. TENSILE STRENGTH. The ultimate tensile strenglii per square inch on test-pieces for all tliree kindsof rolled sled used in structiual metal-work shall be as follows : Soft steel r)0,()00 lbs. to 00.000 lbs. Medium steel . . . 60,000 lbs. to 70,000 lbs. High steel 70,000 lbs. to 80,000 lbs. not be ELASTIC MMITS. The least allowable elastic limits obtained from test-pieces and determined in the usual manner by the dro[) of the beam shall be as follows : Soft steel 30,000 lbs per .square inch. y Mum steel... 3"), 000 lbs. per square inch. High steel 40,000 lbs. per square inch. ELONGATION. The percentages of elongation sliall be obtained from the test-pieces after l)rcakiug t)ii an original length of eight (8) inches, in which length must occur the curve of reduction from stretch on both sides of the point of fracture. The least allownlile elongations for the various kinds of rolled structuial steel shall be as follows. 348 IJE I'ONTIHUS. Percentage of Elongation. Shape. Soft Steel. IMediiini Steel. High Steel. Roiiuds (exceDtin^ Dins) 29 29 27 20 26 24 23 21 20 Pins 18 Angles antl bars 22 Plates under 40" wide Plates 40" to 70" wide, and webs of beams and channels 20 19 Plates over 70" wide Flanges of beams and ciiannels 18 UliUUCTIO.N OK .\KE.\. Tiie reduction of area, inuii.surc'd on test-pieces, for the va- rious kiuds of rolled slruc'.ural steel shall be as follows : Percentage of Reduction of Area. Shape. Soft Steel. Medium Steel. High Steel. Rounds (exceutlne nins) 50 48 44 40 40 40 38 37 86 88 Pins Angles and bars 34 34 Plates under 40" wide Plates 40" to 70" wide, and wel)s U 34 Plates over 70" wide Flanges of beams and channels 30 HEND1NG TE8T8. Spcciineiis of soft steel shall be capable of bending to one hundrfd and ciglily (IHO) dciirees and closing; down Mat u[)on tlicinsulvcs, without cracking, when either liot, cold, or qiteuchcd. Specimens of mediiitn steel, when heated to a dark orange nud cooled in water at .seventy (70) degrees Fahrenheit, or when cold or hoi, shall be capable of bending one hundred and eighty (180) degrees around a circle whose diameter is equal to llic tliickncss of the test-piece, without showing signs of cracking on the convex side of tlie bend. SPKCJIFICATIUNS h'OU STEKF. IN BliIJKiE:S, ETC. ;.'49 Specimens of liigh sleel when qiieiielied in a siniiliir rniuincr shiill be ciipuble of l)i'n(ling ninety (90) degrees around a circle whose diameter is equal to twiee the thickness of the test-piece, and one hundred and eiglity (180) degrees, either hot or cold, without showing signs of cracking on the convex side of the bend. UKIFTINO TESTS. Punched rivet-lioles in medium steel, pitched two (2) diam- eters from a slieared edge, nuist stand drifting until their diameters are fifty (50) per cent greater than those of the original holes, and must show no signs of cracking the metal. Higii steel must stand the same test, except that the increase in diameter is to be twenty-five (25) per cent instead of fifty (50) per cent. FUACTURE. All broken test-pieces for all three classes of steel must show a silkv fracture of uniform color. NUMBKK OF TEST- PIECES. At least three (3) tensile tests and three bending tests shall be made on specimens tvom dilTc rent ingots of each melt. The bending tests mny, if desired, be in;ide on the broken test-pieces of the tension tests. If material of various shapes is to be made from the same melt, the specimens for testing are to be so selecteil as to represent the dilfereut shapes rolled from such melt. All tests are to l)e made by the Contractor for the In.spector without charge. The Insi)ector will Ix; permitted considerable latitude in respect to the number of tests re(iuired, reducing same when the metal runs uniforndy and increasing .same when it does not. Lots for testing .shall not exceed twenty (30) tons in weight; and plates rolled in universal mill or in grooves, or sheared plates,, shall each constitute a separate lot, as shall also the angles, channels, or beams. 250 T)K I'ONTIHUS. TESTS OF FUtiL-SIZED ETE-BARS. Full-sized eye-bars may be tested to destruction , provided notice be given in advance of the number and size required for this purpose, so tliat the material can be rolled at the same time as that required for the structure. The number of tests of full-sized eye-bars will depend upon the size of the order and upon the reguliirity of tlie results of the tests. In general, for small orders, the number of tests shall be about tliree (3) per cent of the number of eye-bars in the order, but never less than two bars for an order for a single span. For large orders the number of tests shall be about two (2) per cent of the number of eye-bars in the order. Should the Inspector find the bars to be very uniform in strength, elasticity, and duc- tility, and fully up to the specifications, he will be at liberty to reduce the number of tests of full-sized l)ars. In the case of testing long bars, it will be allowable to choose a bar at ran- dom from a nuiul)er of finished bars, cut it in two, and upset the cut end of eacli piece, tlius making two test-bars. Full sized bars of medium slcjd must show an ultimate, tensile strength of at least sixty thousand (60,000) potmds per square inch for bars of one (1) inch thickness and under, and not less than fifty six thousand 156, 000) pounds per stiuare incii for bars of two (3) inches thickness and over. Bars witli thicknesses betweei these limits must show proportionate strength. The elongation shall lie not less than fourteen (14) per cent in a gauged length of ten (10) feet; and the elastic limit shall not be less than fifty-five (55) per cent of tiie ultimate strength of tlie bar for bars not over one inch thick, or less than fifty (50) per cent of same for bars of two (2) inches thickness and over, with proportionate percentages for bars of intermediiite thicknesses. For high steel the limits just specified shall be changed as follows : Ultimate strength, 70,000 to 65,000 pounds. Elongation, twelve (1'^) per cent. Elastic linut, fifty-two (53) per cent to forty-seven (47) per cent. SPRorPICATroNS Poll STRRL iK MUIDnES, ETC. 351 Any lot of steel burs which meets the requirements of tlie preceding paragraph shall be accepted, if none of the bars which break in the eye show an ultimate strength, elastic limit, or elongation less than that specified for the body of the bar, unless one fourth (J) of the full-sized samples so tested break in the eye. In case of failure to meet any of these requirements, the lot from which the sample bars were taken will be rejected. All full-sized sample bars which break at less than t' e ultimate strength .specified, or do not otherwise fill the .speci- fications, sliall be at liie expense of the Contractor ; luiless, in case of tiiose that break in tlie eye, he siiall have made objec- tion in writing to the form or diinen.sio:is of tiie heads before making the eye-bars. All others .shall be paid for by the pur- chaser at the contract price of finisiied metal-work on cars at shops, less tlie scrap value of the broken bars. PIN METAh. Pins up to si.\ (6) indies in diameter may be rolled, but al)ove that diameter they shall be forged. The rounds from which the pins are to be turned must be true, stiaight, and free from all injurious Haws or cracks. All forged pins shall be reduced from a single bloom or ingot until perfect homo- geneity is .secured throughout tlie wliole mass. Tiie blooms shall have at lea,st three (3) times tlie sectional area of the finished pins. No forging shall be done below a red heat. VAUIATION IN WEIGHT. Except in the case of sheared plates orilered to gauge, a variation in cross-section or weight of rolled material of more than two (2) per cent from that specified may be cause for rejection. For the .said sheared plates the permissible excess variation shall run from four (4) per cent for plates five eighths (I) of an inch or more in thickness to eight (8) per cent for plates three eighths (f) of an inch or less in tliick- ne.s.s, the variations for intermediate thicknesses being directly interpolated. bV. PONTIBUS. Should Ihf shipping weiglit of any eiitiru oidcr exceed hy more timii one (1) per cent the weiglit couipuled from liie approved siiop drawings, tlie amount in excess of tlic said one (1) per cent will not lie paid for, unless in the entire order tlie weight of plates exceeding thirty six (36) inclieH wide bo greater than thirty (30) per cent of the whole, in wliicli case the allowable variation shall be increased to two (!i) per cent. wiuniGiii inoN. All wrought ircui, if any be used, must be of tbc best quidily obtainable, tough, ductile, librous, and of a ludforni quality; also straight, smooth, and free from cinder-pockets or injurious Haws, buckles, blisters, or cracks. No steel scrap shall be used in its miinufacture. The tensile strength, determined fron\ test-pieces in the same nuinner as spe(;ilied for steel, shall not fall below fifty thousand (50,00(1) pounds per stjuare inch; and the elastic limit shall not be less than twenty-six thousand (20,000) pounds j)er square inch. The alongalion, determined in tlif same nuinner as specified for steel, shall not be less tlian twenty (20) per cent. All wrought iron must bend cold oi;e hundred and eiglily (180) degrees, without sign of fracture, to a curve the inner radius of whicli equals the tiiickness of the piece tested. Soft steel is to be used instead of wrought iron wherever piac- ticable. CAHT IHON. Except where chilled iron is specified, all ciastings shall be of tough, gray iron, free from injurious cold-shuts or blow- holes, true to pattern, and of a workmanlike finish. Sample- pieces one (1) inch square, cast from tlie same heat of metal in sand-moulds, sliall be capable of nuiintaining on a clear span of four (4) feet six (6) inches a central load of five hundred (500) pouiuls when tested in the rough bar. All castings shall be straight and out of wind, with proper and approved uniform thickness of melal, ami shall liave perfect, sharp, SPKfJIFlCATIONS FOK HTKIWi IN HIllIMJKS, K'IC. -ly.] and vAeau lines, angles, ami muuldings, till re-entrant angles being properly filleted. CAST HTKKL All steel fastings slinll he made of acid open hcarlli steel containing from twenty tive inindicdtlis (O.'j,")) to four tenths (0.4) percent of carbon, and not more than Die following per- centages of other ingredients : Phosphorus, five hnndredihs (0.05). Sul[)hiir, live hundredths (0 05), Manganese, eight tenths (0.8). The ultimate tensile strength .shall run from sixty five thou- sand (65,000) to .seventy-live thousand (75,000) pounds per square inch; tin; elastic limit shall not be less than one half (i) of the ultimate stiength; and the elongation of tc^t specimens in two (3) inches shall not be less than tifteeii (15) per cent for rtxtd castings or seventeen (17) per cent for movable castings. All steel castings shall be carefully and uniformly annealed, and shall be true to drawings, smooth, clean, and free from blowholes, spongine.ss, and all other defects. All corners therein shall be properly filleted. TKSTH OF KOLLKR8 FOK DHAW-8PAN8, The Contractor shall make, at his own expense, under the direction of the Engineer or his duly authorized representa- tive, for each draw-span, tests, not exceeding thieu (:!) in number, of full-sized cast rollers; also any tests of s]>ociinens of the metal for the same that may be considered necessary by the Engineer to determine its quality. OTHER TESTS OF FULL-SIZE MEMBERS OR DETAILS. The Contractor shall make, at his own expen.se, tmder the direction of the Engineer or his Insjieclor, such oilier tests of full-size members or details as the Engineer may prescribe, 254 1)K I'ONTJHCS. provided tliat llic siiid inumbeis or details are sliniliir to those used oil the work, mid provided lliul the toliil cosi lo tlie('oii- tnictor of such extra tests docs not exceed one (pmrter of one per cent (O.'i.W) of the total conlract price of the work. WOllKMANWlllI'. All metal shall be carefully straightened before being turned over lo the shops. All workn)aushi|) shall be tirst-class in every particular, and all portions of metal-work exposed to view shall be iieally Ihiished. All idle corners of plates and angles, such for instance as the cuds of the unconnected legs of angle lacing, shall be utally cliaiiifercd olf at an angle of about forly-tive (45) degrees, so as U) give a siglitly linisii to the work and to avoid beniling of s lid corners during shipment and erection. As far as practicable, all parts siiall be so constructed as to be accessible for inspection and paiuling. All punched work shall be so accurately done th'U, after the various component pieces are assembled and before the ream- ing is commenced, forty (40) per cent <>f the lioles can be entereil easily by a rod of a diameter one sixteenth (j'j.) of an inch less than that of the punched holes; eighty (80) per cent by a roti of a diameter one eighth (J) of an inch less thsin same; and one liuii;lietl (100) per cent by a rod of a diameter one quarter (|) of an inch less than same. Any shopwork not coining up to this requireinent will be subject to rejection by the inspector. SHEAUED EDGES. All sheared and hot-cut edges shall have not less than one (piaiter (J) inch of metJil removed by pinning to a smooih, linished surface. Lacing-l) irs, fillers, stay-plates, and stringer- bracing connecting plates only will be exempt from this re- quirement. RE-ENTUANT COllNERS. No sharp or xinfi!let{!d reentrant corners will be allowed anywhere in the work. SPECIFICATIONS FOR STEEL IN UKIDOES, ETC. 255 ANNRAMNO. Ill all cases where a steel |)iece in which the full strength is rwjuired has been partiully heated or bent, the whole piece must be subsequeiitl}' annealed. In pieces of .secondary im- portance, where the bending is slight, said bending is to be made cold, and no annealing in such cases will be required. Crimped web stiff eners will not reciuire annealing. All eye-bars shall be carefully and uniformly annealed at a dark orange heat. UIVKTS. Rivets when driven must completely t^.ll the holes, Lave full heads concentric with the rivet-holes, and be machine-driv jn whenever practicable. The machine must be capable of re- taining the applied i)re8sure after the upsetting is completed. The rivet-heads must be full and neatly rinished, of approved hemispherical shape, in full contact with the surface, or be countersunk when so reipiired, and of a uniform size for the same-sized rivets throughout the work; and they must pinch the connected pieces thoroughly together. Flattened heads may be used in certain places, if necessary for clearance. Except where shown otherwise on the drawings, all rivet diameters are to be seven eighths (J) of an inch. No loose or imperfect rivets will be allowed to remain in any part of the metal-work. RIVET-HOLES. Rivet-holes must be accurately spaced; the use of drift-pins will be allowed only for bringing together the several parts forming a member, and they must not be driven with such force as to distort the metal about the holes. The distance between the edge of any piece and the centre of a rivet-hole must never bo less tiian one and a half (]«) inches, excepting for lattice bars, small angles, and wlu.re especially shown otherwise on the Engineer's drawings; and, wherever practica- ble, this distance shall be at least two (3) diameters of the rivet. 256 UK I'ONTIBUS, rUNCniNO AND TIEAMINO, All rivet-holes in steel-wo) ,., if puncbcd, sliall be made with a punch oue-eighth e fai c I true and sciwaro or to exact bevel ; and the end-stifT<.'ncrs must l)e placed perfectly tlusii with these planed ends, so ns to tif'ord a i)roper bearing. Filling-[)lates beneatli end-stilTi'ning angles must be practically flush with said angles, and must m no case project outside of same at tiie bearings. If a good and satisfactory jol> of woik cannot be obtained l)y tins nieihod, the end-stitl'ening angles shall be made one eighth (J) of an inch thicker, and the entire ends sliall be planed after the slilTening angles are riveted on. No web-plate will be allowed to pioject beycjnd the flange angles, or to recede more than one eighth (^) of an inch from faces of same. All filling and splice i)lates in riveted work must fit at their ends to the flanges sunicienlly close to be .sealed, when painted, against the adinissim of water; bit they need not be tool-finished, unless so specially indicated cither i>n the drawings or in the si)ecitications. EVE-BAUS. Except in the case of loop-eyes, no weld will be allowed in ihc bf>dy of the eye-bars. The heads of the eye-l)ars sindl be made by upsetting, rolling, or forging into shape. A variation from the speci/ied dimensions of ihe beads will be allowed, in thickness of one thirty-second (.,'.,) of an incii below and oiw sixteenth {f\) of an inch above that specified, and in diameter of one fourth (\) of an inch in cither direction. Eye-bars must be perfectly straight before Ixu-ing. Loop-eyes shall invariably bv' made of wrouglit iron, as steel cannot be relied upon to afford a proper weld. hv riNIIOLKS. All pinholes must be bored tiidy parallel and at right angles to the a.xe« of tlie nvembers, urdcss oliierwis<' >iiown on the drawings; and, in piece-; not adjustable for length, no variation of more than one tiiirly-second (;jV'<'f "" 'iK'h will be allowei] in the length between centres of pinholes. rioH 1)H I'ONTinUS. Pinholes in e^e-bars imist be in the centre of the heads, and on the centre line of tlio bars. Bars, whicli are to be i)lace(i side by side in the structure, shall be bored at the same temperature, and shall be of such e(iual length thai, when placed in a [tile, the i)in at each end will pass throngii the iioles at the same time without forcing, PINS. AH plus shall be turned accurately to a gauge, and shall l>e linished perfectly round, smootli, and straight. All pins up to three and one-half ['.]h) Inches in diameter shall fit tiie pin- holes within one liflieth (./„) of an inch, and all i)ins over three and one-half (;{A) inches in diametcH" shall lit their holes within one thirty-second (.j'j) of an inch. The (.'ontractor must provide steel pilot imls for all pins to preserve the tineads while said pins are being driven. TUUNED BOLTS. When members are connected by bolts which transmit shearing stresses, the holes must be reamed i)urallel, and the bolts mubl be turned to a driving tit TUUNBlICKLliS, Nl TH, TlIItKADS, ANU WASHKUS. All sleeve-nuts, turnbuckles, and clevises must lie made so strong and still that they will be at>le to resist withou;, rupture the ultimate pull of the bars Avhich they connect, and without distortion the greatest twisting force to which they could ever t)e subjected. They must be made so that the tiircailcd lengths of IJie rods (Migagcd can be verified. The dimensions of all .square and hexagonal nuts, except those on the ends of pins, shall be such as to develop the full strength of the })oily of the adjustable member. No round- headed bolts will \h', allowed. Washers and nuls luust iiave uniform bearing. Cast or wrought iron wasliers must i)e used under the heads of all timber bolts when the bearing is on tiie wood. si'KcincA TioNs I'oit ,sri;i:L in i;i;ii)(ii:s, ktc. ^i.M) All ihrciids, (!.\(:(;[)t, Iliose 0:1 llic oinis of pins, iiiiist be of the Uuitt'tl Sliiles staiidiinl. Eacli udjusling nut must be provided with lui effect ive nut-lock or check-waslier, HOI.l.KUS. Hollers '-hall ]h; turned iiccurately to a gauire, and n)ust be linislied in'ifecljy round, and to tlie correct diameter or diameters, fidrn end to end. Tiie tongues and grooves in I)lates and rollers must til snugly, so as to prevent lateral motion. Roller beds must be i)bined. ANCtlOK-BOI/rS. All bed-plates and bearings must be fox-bolted to the masonry or attached to concrete by anchor-plates. The Con- tractor must furnish all bolts, drill all holes, and se! the bolts to i)lace with Portland-cement grouting. All anchor-bolts are to be of soft steel witli cold-pressed threads; and the threaded portion of all such l)olts tested to destruction shall develop a greater s.iength than tliat of the unthreadc.'d poi'tion of same. Tiie lengths of t'le nuts for all ao[), the other two in the field, before erection. Pins, bored pinholes, turned friction-rollers, and all other polislied surfaces shall be coated with white lead and tallow before shipment from the shop. After the structure is ere<;ted the metal-work shall be thorougbly (;Uansed from mud, grease, or any other objection- able material that may be found thereon, then thoroughly and evenly painted v.itli two (2) coats of paint of any kind that the Engineer may adopt. All tiircvf coats of paint given \o the metal-work are to be of distinctly dilTen nt shades or colors ; and the second coat must be allowed lo dry thorougiily befo.e the third coat Is ajjplied. No thinning of paint with turpentine, benzine, or other thinner will be allowed without special written iiermission from the Engineer No painting is to be done in wet or freezing weather. All painting is to be done in iv thorougli and workmaidik maimer, to the satisfaction of the Engineer, and no paint whatever is to be used on the .structure without lirst being approved by the Engineer. All the mafrrials fur painting shall be subject at all times (o the closest inspection and chemical analysis; and the detection of any inferior (juality of such material, in either siiop or field, shall involve the rejection of ail such suspected material ;it hand and the scraping and repainting of those portions of SPKClFlCATlONS I'Oll SIKKL IN HKIDGES, ETC. 201 tlio work whicli, in the opinion of tlio Eiigiiiecr, wen; de- fectively painted on uc oiint of such inferior niateriid. All recesses which would retain water or through which water could enter must be lilled with thick jiainl or some water-proof cement before receiving final painting. All sur- faces so close together as to |Mcveiit tiie iiiseiliou of paint- brushes must be painted thoroughly by using a piece of cloth instead of the brush. SHIPPING. All parts shall be loaded carefully so as to avoid injury in transportaiion, and shall be at the Coiitru(;l(jr's risk until erected and accepted. In shipping hjiig plate girders great care is to be talien to distribute the weight properly over i1k; two cars that support them, and to provide means for peiinitliiig tlie cars to pass around curves without disturl)iiig the loading. In both the handling and shipHient of metal-work every care is to be taken to avoid bending or straining tlie pieces or damaging the paint. All pieces bent or otherwise injured will be re- jected. TIMBEK. All timber must 1)e of the best (pinlity, sawed true and out of wind, full size, and free from wind-shakes, large or loose knots, decayed wood, sap, worm-holes, or any other defect that would impair its strength or durability. ERECTION. The Contractor shall furnish all staging and falsework, and shall erect, adjust, and p.aint all of the metal-work ready for the timber-floor. He shall also furnish and lay the latter and put on tlie track-rails, unless there be a written agreement to the contrary. The Contractor shall employ suitable mechanics for every kind of mechanical work, and shall, at the recjuest of the eu- 2G3 I)K I'OXTIHUS. giiiet;r, dischiirge any workman whom the said Engineer shall deem incoinpelent. negligent, or untrustworthy. All nmterial of whatever kind siiall he suhjeet to inspection and approval at any time during the progress and until the final completion of the work ; and tlie entire work shall l)e conslnieted in a siihstanlial and workmanlike manner, anil to the satisfaction and acceptance of the Engineer. DKKKCTIVE WOHK. The Contractor, upon being so directed by the Engineer, shall remove, rebuild, or make good, witiiout charge, any work wliich the said Engineer may consider to be defectively executed. The fact tiial any defective material in the struc- ture had been previously accepted l)y the oversight of the Company's engineers or inspectors shall not be considered a valid reason for the (Contractor's refusing to remove it or make it good. And uiu.il such defective work is removed and made good, tiie Engineer shall deduct fif»m the jiartial pay- ments or th(! final payment, as the case may Ix-, whatever sum for siKiii defective work as may, in liis opinion, appear just and equitable. niKKCTIONM TO CO.NTK.VCTOH. Tn case that the (-"ontractor shall not la; present upon the work at any lime when it may la; ni'ccssar}' foi the; Engineer to give instructions, the foreman in < liari^c for the lime being sinvll receive and obey any orders liiat tlie Engineer nniy give. The Contractor slnill commence work at .sucli points as tin,' Engineer may direct, and shall conform (o his ilir'ctioni a^io tlie Older and lime in widch Mie diirerent piirls of tiie work shall br done, as well n.s to the force requireil to coritplete the work Ht llie date spccifled. CI USING THOUOfi.M FAKES. The Contractor and his employees shall so conduct their operations as not to clo.so any thoroughfare by land oi water SPECIFICATIONS lOll liltKCrnON OV UUIDGES, KTC. 203 wilhoul the wiiituii (roiiseiit of the proper aulliorities of such thorough fare. KKSI'ONSIBIMTY FOR ACCIDENTS. The Contractor sliall assimu' and be responsible for all acci- dents to men, animals, and materials before the acceptance of the structure ; and must remove at his own ex])ense all false, work, rubbish, or other useless material caused hy liis oper- ations; and such work sliull be included ms a part of the work to be performed. The Cctnlractor shall place sutlicienl and proper guards for the prevention of accidents, and shall put up and maintain at night suitable and sullicient lights. DAMAOES. The Contractor shall indemnify and save harmless tlie Com- pany against all claims and demands of al! parlies whatsoever for damages or compensation for injuries arising from any obstructions (Mealed by the Contractor or his er.iployees, or from any neglect or omission to provide proper lights and signals during the construction of the work. ALTERATION OF l'I..\NS. The Engineer shall have the power to vary, extend, in- crease, or dimiinsh the ([uantity of the work, or to dispense with a portion thereof during its pn)gr(!ss without impairing the (KMliact; and no allownnce will be made the (.\>n tract or e.Kctpi for the work aclutiUy done. In case any change in- volve the exe( ution of work of a class not herein provided for. the (Contractor shall jierform the sanu; and be paid the aoti^l cost tiieniof phis the percentage for prolil agreed upon in the contract. In this case the Contractor must furnisii the Engi- neer with satisfactory vouchers for all labor and material expemled on the work. STKICTNK88 OF IKSPECTION. All materials and workmanshij) will be thoroughly and carefully inspected, and the Contractor will be lield at all 264 m: poxN'TiBUS. times tt) tho spirit of tlie sivcificatious; but nothing vvill be done by tlio Conipun^-'s engineers or inspectors to give tin: Contnictor needless \v(»ny or iuino^iinee, the intent of botli spec ilicitt ions and inspection being simply to obtain for the (Jonipjiny work that will be liist-class in every particular and a credit to every one connected with its designing and con- strucliou. SPIRIT OF THE SPECIFIC ATTONB. The nature and s|)irit of Ihe-e specilicalions are to pro- vide for the work herein enumerated to be fully completed in every detail for the purpose designed; and it is hereby under- stood thai tiie Contractor, in acc(!pling the contract, agrees to fmiiish any and every thing ne(!css;iry for such construction, notwitlistanding any omission in the drawings or speciticik- tious. ENOINKEK. Whenever ill these specirtcalions the term "Engineer" is employed, it is understood tlial it is to mean the Engineer of the Company or the duly author- ized representative of same. TENDERR. Tenders for all work, whenever it is practicable, shall be made on schedule prices, lump-sum bids being accepted only for such parts as .steam or electric machinery, which could nut well be paid for by the pound. All Lenders are to be made in strict accordance with ihe plan;; and specifications submitted to liidders by lh(i Engineer; and no bids based upon suggested changes in same will be considered. In awarding contracts, pieference will be given to those bidders in whose shops the piece-work system is least em- ployed. CHAPTER XIX. TFIE COMPROMISE STANDARF> SYSTEM OK LIVE LOADS KOR RAILWAY BRIDGES AND THE EyUIVALENTS FOR SAME. In 1893 the author published a Hiiiall paiiiphh't, now out of print, whicli bore the above title. Its contents are reproduced iiere instead of in a second edition. The various h\v.\)h talien in its preparation were as follows : In 1891 tiie author presented to the Aniericnn Societ}' of ("ivil P^ngineers a paper entitled "Some Disputed Points in Railway-Bridge Desigi.ing," in which he advocated the adop- tion of a few standard train-loads for railroad bridges, instead of the ahnost innumerable ones then in u^e, ollVrcd a set of loads for discussicn, and urged that the " lupuvalent rniform- Load Method ' of coniputitig stresses be adopted instead of the burdensome method of wheel concentrations that liad been in vogue for the preceding ten years. This p.-iper received a very thorough discussion, from which it was evident that bridge engineers and railroad engineers, as a whole, would be glad to settle upon a few standard loading-*, and to adopt some simple etpdvalent method of (computing stresses. Most of those who desired the ab;indonmeut of the " Coiu'entrated Wheel-load Method," advocated tlie adoption of the " Ecpiiv- alent Uniform-Load Method," but a few favored either the "Single" or the "Double Concentration Method," with a constant car-load. Tliis paper, with the discu.ssious, was published in the Feb- ruary and March 1892 number of the Transnctionn of the American So(!iety of Civil Engineers, and was reviewed very generally by the tcchiucal ))r(>ss, attention being paid princi- pally to the subject of equivalent loads. These reviews started 365 ',*<;(; l)K I'ON'rilJl'S. ii s(!iies of It'ltcrs by tliciuilliordiid others, that wcic printed iit first ill tlic lidUroad (hizette, and L.ti-r iilsu in tin; Kngineeriny Record, in wliicli li'ltcrs the suliject of eciiiivuleiils was thor- oiigiily and cxhiiustively treated. These jiroved tliat the " E(iiiivaieiit L'uiforiiiLoad Method " gives results wliicdi are ueeiirate eii()ut,di for idl practical jjiirposes, and tiiat iieiliier the "Single Concent liiled-Load Metliod " nor \\\v "Double Concent rated-IiOad Method " gives results coinciding at all closely with those found by the theoretically exact method of " Wheel Cniicentiatioiis." Ill November 1892 the author sent a circular letter to all the chief engineers of railroads in the United States and Can- ada who were members (in any grade) of the American Socii'iy of Civil Engineers, and to every other memlHjr of thai society connected wilh or specially interested in tht; designing, build- ing, or operating of railroad bridges. 'I'lds letter soli(;ited a ballot on certain " Disputed Points in Hallway IJiidge Design- ing." foremost among which were tiioscjof standard live loads and a siin|)le ecpiivalent method for (•omputalion. Tin; num- ber of responses received was as great as could liave been ( \- pected ; and thu result was that about eighty-two per cent of those who voted favored and eighteen per c(;iit opposed the adoption of " a iStandard System of Live Loads for Railway Bridges " similar to that proposed by the author. Eighty-two per cent also of those who voted were in favor of abandoning the " Concentrated Wheel-Load Method," and eighteiii per cent were in favor (»f retaining it. Of lh?> ^>. IMAGE. EVALUATION TEST TARGET (MT-3) V // .'^Z i 1.0 ;£|21 12.5 1^ 1^ lu I.I J.-^ L25 iU 11.6 2.2 20 1.8 V] /I >> •^^ '/// 7 Photographic Sdences Corporation 23 WEST MAIN STREET WEBSTER, N.Y. 14580 (716) 872-4503 \ s ^ '2(iS DK POVTIBUS. taken from the "Compromise Standard System," and that tbe "Equivalent Loads" thereof are to be employed. In this "System" will lie found from " Class Z" to "Class T," inclusive, a close approximation to any live load that an engineer is likely to want to use ; and if, for a certain c:ir- load, some engineer should prefer a heavier or ligi^ler engine- loading, I'O can obtain practically what he wislies by speci- fying that one class is to be used for floor systems and primary- truss members, and another class for main-truss members. The author does not advise this, however, except in the case of double-track bridges, where it would be advantageous to use a certain class for floor systems and primary-truss mem- bers, and a lighter class for the trus.ses, because the chanct-s of there being two, full, maximum train-loads on the span at the siiMie time are generally very small. It might be well to carry this idea even further by specifying, for instance, "Class V " for stringers, " Class W " for floor-beams and pri- mary-truss members, and "Class X" for main-truss members of doubk'-tiack bridges. Such a method would be in accord- ance with the theory of probabilities ; but it would not apply to single-track bridges, for which the locomotive and car loads of the " Compromi.se Standard System " have been properly adjusted. The "Equivalent Uniform-Load Method" reduces to a minimum (he labor of making computations of stresses in bridges. The correctness of this statement will be rendered evident by the ensuing explanations of the use of the method. As for its exactness, if any-one has any doubt whatsoever about its closeness of approximation to the theoretically cor- rect method of wheel-concentrations, let him read the author's letter in the Railroad Oazette of July 28, 1893. An inspection of Table I of that communication shows that no reasonable man can object to the "Equivalent Uniform-Load Method" because of its want of exaciness. In designing a bridge, one conimences naturally with the stringers, then passes to the floor-beams; and afterwards to the trusses ; so let us follow this order. COMPROMISE STANDARD SYSTKM OK LIVE LOADS. 2i)9 STRINOEKS. From Plate III find tlie equiviilont live loiul per lineal foot for a spun eqiiul to the panel length, add to same the assumed weight per foot of two stringers iind I lie Hoor they supporl, and divide the sum hy two, calling the result ic ; then find the total bending moment at mid-spun hy substituting in the well-known formula, where I is the panel length in feet, and M is the recjuired mo- ment in footpounds. Should the total end shear be required, it can be found for each stringer by adding together the end shear given on 1*1. II and the total weight of one stringer \n\)x the floor tliiit il carries, and dividing the sum by two. FLOOH-HEAMS. In proportioning a floor-beam, tlio important thing to ascertain is the total concentration at the point where two stringers meet. The. live-load concentration is to be found by multiplying together the panel length and the equivalent uni- form load per lineal foot given on PI. Ill /or a spin etjiinl to twice the panel length, and dividing the product by two. It is unnecessary to describe here how the dead-load concentration at each stringer support is to be found. Nor is it necessary to do more than merely mention that the live-load concentra- tion obtained for the floor-beam is the same as that required in finding stressesln primary-truss members. TRU88E8. These can be divided into two kinds, viz., those with equal panels and parallel chords, and those in which the panel lengths are unequal, or the chords are not parallel, or both. In the first case, the stresses can be determined most expedi- tiously by substitution in tabiduted formuUp, and io tbc secom) case by the grapliical method. 270 I)E I'ONTIIU'.S. Case I. From Plate IV find the equiviileiit uniform live load per lineal foot for the given span iengtli and multiply same by the panel length, calling the product L. For siiigh; trad; bridg s this must be divided by two. All the live load stresses in midu-truss members of siugle-iulersection bridges can be found by substituting this value of L in Table XVII. Just here it is proper to remark that the " EiiuivaUnI Uni- form-Load Method " is not applicable to trusses of multiple intersection ; but the most approved modern practice in bridge. engiueeriug does not eouutenaiice the building' of trusses or girders having n\ore tlnin a single system of canceli.;ti<)n Tlie "Eq\iivalent Uniform Load Method" does howevt r, apply to trusses wiih divided panels, such as the Petit truss; Itui as this style of truss nowadays involves almost inv.'.rialily a polygonal lop chord, its trea'raent herein will come under Com II. Where trusses have unequal panels or chords not parallel, the first step to take is the finding of all the dead -load stresses by tlie graphical method, starling from one end of the span and working towards the middle, where the last stress is checked by the method of moments, and the correctness of the entire graphical work is thereby proved. The next step is to find from PI. IV., as in Case 1, the equivalent live-load per lineal foot for the span, and there- from the value of the panel-truss live-h)ad L. Next set a slide- rule for the ratio of dead load per lineal foot and the equivii- lent live load pei- lineal foot for the span, and, by referring to the dead-load stresses already found, read oil from the rule all of the live-load stresses in chords and inclined end posts. Next assume that there is an upward reaction at one end of the span equal lo 1,000 pounds, 10.000 pounds, or 100,000 pounds (according to the size of the bridge), caused by a load placed at the first panel point from the other end of the span, then find graphically the stress in each web- mem her from end to end of span, caused by this assumed upward reaetiop. (JOMl'UOMISK STAN'DAUI) SYSTKM OF LIVE LOADS. ~ Tlieii calculutc the value of the live-load reaction for the iiiiiximum stress in each web-member by meaus of the slide- rule and the following formula and table, in which n is the number of panels in the span, 7i' is the number of the panel point at the hend of the train, coiniting from the loaded end of the span, and C is the coefficient of — . n Live-load reaction for the head of train at n' = X L n of n' C H' C n' 18 C n' C 1 1 1 28 91 19 190 2 3 8 36 14 105 21) 210 3 6 9 45 15 120 1 «' 231 4 10 10 55 16 136 ! 22 253 5 15 11 66 17 153 33 270 6 21 12 78 18 171 24 i 300 Then, still using tiie slide-rule, find the greatest live-load stress in each web-meinber by the following equation : Stress required = Actual Reaction Stress from Assumed lleaction X Assumed Reaction' Where the panels are divided as in the Petit truss, and where inclined subposts are emi)lo3'ed, the tensile stress in the upper half of each main diagonal thus found will have to be corrected by subtracting therefrom a stress equal to -- sec. .1, where A is the inclination of the diagoniil to the vertical. But when inclined subties are used instead < "■ iu(tlined sub- posts, the correction just referred to will api»ly only to the compressive stresses in the lotoer halves of the main diagonals. The reason for making this correction, as will be at once evi- dent to any one who is accustomed to finding stresses in Petit trusses, is that the metiiod above outlined ignores the sub- division of the panels ninn ascertaining by graphics the stresses caused by lheas.«uiued u])ward reaction. 272 DB P0NTIBU8. Tu comparing the equivalent loads for spans of one Imn- dred feet, ijiven on PI. Ill, with those given on PI. IV, au apparent discrepancy will be noticed. This is dne to the fact that PI. Ill is for plate-i-irder spans, for which the equivalent loads were ohtained fronj the bending moment at mid span; while PI. IV is for iru.ss bivas, for which the e(ivuvaleut loadttare the average of those at all of the panel points. hi CHAPTER XX. TIMBER TRESTLES. TiMBKR trestles can be divided into two geueml classes, via : Fii'Ht, Pile-trestles, or those in which each bent is formed of several piles, a cup, and transvt-rse sway-bracing ; and, Second, Framed Trestles, or tiiose in wliicii each bent is composed of squared timbers framed togetiier and braced. Owing to the excessive length of piles r(qiiire>! for greater height.'^, pile-treslles should riin-ly, if ever, exceed thirty feel in height, while framed trestles, if i>r.n>erly di-s'gneii for rigidity as well as for strength, maybe carrie 1 up to much greater heights, the economic limit being probably about one hundred feet. l'IliE-TUK«TIiK8. ITie Itents of a pile-trestle should contain at least four pilcjs each. Where tlie trestle does not exceed ten feet in height, the piles may be driven vertically, and no sway-bracing need be used, i)rovi(led that the piles have a good penetration in reliable material. For greater heights of trestle than ten feet, the two outer piles of each bent shcmld be given a batter of from two to three inches (o the vertical foot. Each bent should also be bmced with one or two sets of sway-bracing, each composed of two 3" X 10" yellow-pine diagonals, thor- oughly bolted to the piles, wherever they cross them, by J" bolts. Wherever the piles are of irregular .sizes, they should be trimmed off so as to make the diagonal bracing tit prop- erly. The piles for such bents should be so spaced laterally as to 373 274 i)E rovTinus. 1 I: givo great I nu is verse rigidity to the structure, and at the same lime afford .•unj>le support for tlie cnps. A good spacing is us follows : Distance centre to centre of outer piles, 11' 0" ; dis- tance centre to centre of two inner i>il\is, 4' 6 ". The caps should b^j at least VI" X 14" X 14', placed on edge and attached to the piles by means of I" drift-bolts. For ordinary pile-trestles in fairly firm soil no longitudinal sway-bracing will I)e required for heights below ten feet; but for heights between ten and twenty-two feet, single-deck, longitudinal sway-bracing should be used in every fifth panel, so as to prevent the structure from moving longitudinally as a whole because of thrusl of trains. For heights greater than twenty-two feet, each allernate panel should be braced longi- tudinally by double-deck bra<'ing, so as to hold the piles at mid-height, and thus strengthen them as cohnnns; the trans- verse sway-bracing for tliese cases should also be double-deck for the same reason. For ordinary pilc-tresthvs up to twenty-two feet in lieight tlie panels should be a trifie less than fourteen feet in length, while for greater heights either the same length may be used or alternate panels may be made from twenty-four to twenty- eight feet long by trussing tlie stringers, according to which of the two methods is the more economical. The stringers under each rail should be built of three runs of timber generally sixteen inches deep, the sizes being deter- mined from the loading and by using an intensity of two thousand pounds for the extreme fibre, when impact is in- cluded. The stringer timbers are to be separated from each other at tiie panel points by means of timber packing-blocks, which are to serve also as splice-timbers. These limber blocks should be at least three iiiclies thick and six feet ir. length, and should have at least four bolts through them. They are to be separated from the stringers by small cast-iron fillers three quarters of an inch thick, so as to prevent the timbers from coming in direct contact with each oilier. The splice- timbers must be made wide enough to project an inch or two below the bottoms of stringers, and must be nolchid over the caps sous to hold the .stringers firmly in j>lace. The diatauCf TIMHKU TUKSTLKS. 215 from centra to n!Ulre i)f middle stringers sliould l)c live ffci. Iiileriiicdiale (.'.ist-iruu sepunilors witli l)ult.s sliuiild be ii>ed between ndjiiceut stringer-timbers, ut distuuces not tu exceed Jive feet centres. Tlie lenntli of (he striager-timljers for ordinary trestles sbould be iwentyeigirt feet, .-o as to extend over two panels, iind thus .stiffen i!ii: floor system materially. The ties should bo 8" X 8" X 10". They should be dapped over the stringers at least one-lialf inch und S|)aced thirteen Inches fronj centre to centre. Insirovided by ilriving piles and cutting them off above the ground, by using timber sills, or by building .small masonry piers. In all such trestles it will be necessary to brace the struc- ture thoroughly, both transversely and longitudinally. All framing of bents should be done in such a manner as to tie all parts firmly together. For very high trestles it will l)e economical to increase the lengths of alternate panels to twenty-five or cvpu thirty feet, And truss the stringers, 276 DE POXTIBra. The lougitiidiniil brnciiig sliould consist of diiiguiinlH of lim- ber of suitable (iiincnsiuns, ill alternule puiiels, witii liuri/oii- tnl struts iimdc continuous iit bracing piincd points liirougliout all panels. In addition to the transverse and longitudinal bracing pre- viously described, all trestles on sharp curves should be pro- vided with a lateral system composed of timber diugoiials spiked to caps and to bottoms of stringers. What has been said in regard to lioorlng for pile- trestles applies also to framed trestles. The compression-members, when impact is included in th : stresses, are to be proportioned by tlie formula; j) = 1700 - 0.4r J for long-leaf yellow-pine and hard woods, and p ~ 1000 - 0.2 for white pine, short-leaf yellow pine, and soft woods, in which forniulte ^ and /> are respectively, in the same unit, tlic unsupported length and the leiist transverse dimension of the strut. / SPECIFICATIONS FOR TIMBER TRESTLES. CLKAKINU AWAY UUHBIBU AND PUI<:l'AKI^(^ OIIOUNO FOB STARTING WORK. Before beginning work on any trestle, all rubbish 'ogs. trees, and brush must be cleared awaj', and all combustible material must be burned or removed for the entire width of the right-of-way. DIMENSIONS All posts, braces, stringers, ties, guard-rails, sills, and all timber generally, shall be of the exact dimensions given and Ji^ured in the diawings. No variilions fro:a these wjll bp TIMBER TRESTLES. 2:7 iiilowed. except upou tUe written consent of the Engiueer or his duly authorized lepreaeutative. / DRAWINGS. The drawings will be made to the scales indicated, but in all cases the tigures are to lie followed in preference lo the scale, where there is any discrepancy between llie two. The draw- ings are lo l)e followed exactly, excepting in cases of errors or omissions, which must be referred to the Engineer for correc- tion or for additional information. TIMBER. All timber shall be of good quality, and of such kinds as the Engineer may direct. It must be free from wind-shakes, wanes, black, loose, or unsound ktiots, sap, worm-holes, and all descriptions of decay, or any other defect wiiich would impair its strenglli or durability. It must l)c sawed true and out of wind, and to exact dimensions. Under no circum- stances will any timt)er cut from dead logs be allowed to be placed in any portion of the structure; but all timber must be cut from living trees. PILEB. The piles are to be cut from good, live trees of such varieties of timber as may be selected by the Engiueer. They must be straight, sound, and perfectly free from wind-shake.i, wanes, large, loose, black, or decayed knots, cracks, worm -holes, and all descriptions of decay; and they must be stripped of all bark. If square piles are to be used, they must be hewed sijuare and not sawed, but must be as free as practicable from axe- marks. Square piles must be at least twelve (12) inclies across the face, and must show not more than two (2) inches of sap across the corners. The sizes of round piles will depend upon their length, but in no case shall they be less than nine (9) inches in diameter at tips. They shall be so nearly straiglit that a right line, taken in any ravbieh is liable lo Hplit or injure tliem, liie end» are to be pro- tected by cast or wrought iron shoos. Whenever iu driving it becomes apparent that the hammer is splitting or injuring the head of a pile lo any inaleriai ex- tent, the top Is lo bo banded by u heavy wrouglit-irou ring while the pile is being driven. All piles must be cut otT at tops to an evact line so that the caps will bear evenly on all the piles of the group. All piles injured in driving, or that are driven out of place, shall either be cut oil or withdrawn, as the Engineer may elect, and others shall be driven in their stead. W^henever the heads of the piles are of greater diameter than the width of the caps, they are to be adzed olT at tlie tops at an angle of about forty-live (45; degrees, so as to be Hush with tiie sides of the caps. All piles must bj accurately spaced according to plans, and those beneath the track-stringers must be driven vertieaUy. All battered piles must he diiven to the angle shown on ihe drawings. Where piles of dilt'erent diameters are used in the same bent, the large piles must be ad/ed otT wliere tiie diago- nal braces cross them, sotliul the diagonals will not be bent out of Hue. FKAMING. All framing must be done to a close fit, aiul in a thorough and workmanlike manner. No bloeking or shimming of any kind will be allowed in mal'iug joints, nor will any open joints be accepted anywhere on the work. All joints, ends of posLs, ends of piles, etc.. and all surfaces of timber which are to be phiced in direct contact with otUtiV timber or with masonry, must be thoroughly painted with hot, creosote oil, and then covered witli a good coat of hot asphal- TIMRKIl TllKSTI-KS. 2:9 tiitn, or Hiiclt olhur iimteiiiil ur iiiiilfrial.s as iiiiiy be .selectfd Uy the Kngiiieer. All lu»leH of Hiiy kind, wiiieli me bored in any of ihetlniberH, lire to be tboroiiglily Naiumted with hot n-sitliuiliim, and all bolts and faiM s of waHhei'H, wliieh are to be placed in direct contact V I the timber, are to be warmed and dipiKid in a vnt of the stuuv. lhs, All bra(Mng timbers are to be bolted to piles, caps, or other timbers wherever they cross ilr.-m The ends of all stringers shall be tirmly attached to caps by means of drift-bolts, timber cleats, or some other melliod which, in the opinion of the En- gineer, is ci-iuaily gooil. For slru(!turey on curves, th ; superelevation of outer rails is to be provided for by bevelling tlii' ties, not to exceed three (8) inches in tive feet, or by dapping inner stringers on caps not to exceed two (2) inches, or, where the recpiired .superele- vation is too great to be provided for by either of tlu' two methods named or by a combined use of tliem, by cutting oflf the tops of the piles on an inclined plane. The last methou is not to be resorted to unless it be absolutely necessary, and then extreme care must be taken to cut the piles off so that their tops will lie in a true plane. In no instance is tliis to be done when framed bentsare used, as the inclination can then be given in cutting the tops of the bent posts to receive the caps. METAL-WOIIK All bolts, nuts, and dowel- pins shall be made of soft steel or wrought iron of the same quality as that specified for adjust- able members of bridges iu Chapter XVIII. Preference will be given to screw-bolts of soft steel with coldpressed threads. All bolts must be practically perfect in every rcspec t, and, 280 DE PONTJBUS. wherever necessary, they must be provided with uuts and threads of the standard size required for tlieir diameter. The thicliuess of a nut shall not be less thau the diameter of the rod for which it is intended, and the side of a square nut must not be less lh:in twice the diameter of the bolt. The heads of all bolts shall be of the same size as the nuta required for tlie screw ends. All screw-bolts, drift-bolts, and dowel-pins shiill be made truly straight before being driven, and all nuts must be screwed up light against the washers. All nuts and heads of bolts must have heavy O.G. washers between them and the timber. All washers are to be made of cast iron of good qual- ity, and must be sufficiently large and thick to provid. properly for distributing the pressure due to the greatest allowable ten- sion in the bolts over the area of the washers. They must bi; finished in a neat .'uul workmanlike munner, and must be free from air-holes, cracks, cinders, and other defects. All spacers are to be made of cast iron, unless otherwise speciticd, and must be of the same quality and finish as specified for the washers. ERECTION, DEFECTIVE WOKK, DIRECTIONS TO CONTRACTOR, CLOSING THOROUGHFARES, RESPONSIBILITY FOR ACCIDENTS, DA.VIAOER, ALTERATION OF PLANS. STRICTNESS OF INSPEC- TION, SPIRIT OF THE SPECIFICATIONS, ENOINKER, AND TEN- DERS. See Chapter XVIII. CHAPTER XXI. INSPECTION OF MATERIALS AND WORKMANSHIP. Unless all tlie iimterials used in ti stnictnic and all work, matisbi]) during tlie Viirioiis slages uf manufacture ul liie shops and of constrUcliou '".i th« field be subjected to conipetenl and lionest inspection, much of the benefit obtained by scientific design and thorough specifications will be lost. For many years most of the inspection of structural metal- worli was a sad f.'irce ; and, in consequence, the general public placed but little confidence in inspecti(m, witli t.ie result that a large portion of the bridge-work of the country was left entirely to the tender mercies of the manufacturers, who nat- urally worked for their own interest and not for that of the purchasers. Latterly, however, owing to the efforts of a few flnst-class inspecting bureaus, the status of inspection has been somewhat improved, although it is far from bi ing to-day what it ought to be. In making this Itist statement the author speak, idvisedly, in that he has suffered considerably, even of late years, from bad inspection in such matters as the insertion of a rust-joint in a turntable between the bottom of drum and top of upper-track segments, where no such filling was allowed in cither plans or specifications; badly matching holes in field conneciions; pinholes too small for pins; important members omitted in shipping; eye-bars made longer than called for by the drawings ; great recesses In webs xnd fillers at ends of girders; and shop-paint applied over half an inch of frozen mud. Such things, to say the least, are extremely annoying, and often cause great expense during erection. Primarily, the blamo for such errors must fall on the in- 261 Jjs2 DK PONTIBTS. spcctors, for such eirregious bluuders slioiiM never esc;a|)C' lluir observttliou. But tliey are by no menus entirely to blame for the fact that tlie inspection of sU'uclunil steel in gcMenil is not wliat it ought to be ; because biidi of them are the rtiilroad managers and promoters of birge nnlerprises, who do not rec- ognize ihe necessity for tirst-class inspection, and wlio are often not willing to pay one half of what such inspection is worth. Here again, though, the inspectors are to blame, for the rtascm that in the keenness of their competition for work they have cut prices to such an extent as to make it impossible to do proper inspection without losing money. When pinned down to facts they have to confess this. Tiie coolnes.-- of some of the "small fry" inspectors is often amusing The author was once baided over the coals by one of this class who had put in a low bid for some inspection, and whose tender had been re- jected because of the low figure, the work having been awarded to one of the regular inspecting bureaus at about fifty per cent more than the unsuccessful bidder asked. After expressing his mind pretty freely, be fired this parting shot : " Well, I never intended to do thorough inspection for you, anyhow." The inspection business has been utterly demoralized in times past by just such action as that contemplated by this inspector ; for it was the general custon\, and is yet to a cer- tain extent with some inspectors, to take contracts for inspec- tion at whatever figures the purciiasers are willing to pay, then handle the work so as not to lose money on the contract, regardless, of course, of the interests of their em|)loyer8. Strange tales concerning inspection come to the ears of engineers— such, for instance, as passing car-load after car-load of metal-work that was not seen by .the inspector until after loading for shipment ; but such tales need verification, which, of course, it is nobody's business to give them. Tiiere is no doubt, though, of some of them being authentic. In one case )D the author's experience the inspector left his work for ten days in charge, of one of the bndge-coinpant/'s ahipping-clerka, without notifying either the author or his direct employers, the inspection bureau, of his contemplated absence. 8uch iNSl'ECTloy OF MATERfALS ANM) WOllKMA XSHIP. 283 nctions us this inukc one entertuiu doubts soinetiincs as to whether inspection reiilly pays. It is possible that the general deniorulizatioii of nietiil iu- speclion by iiisufflcient prices aud Iseeu conipelilion has low- ered the quality tliereof to such an extent that even tlie highest possible prices would not. make it, for some time to come, what it ought to be ; because not only are the assistant in- spectors lucking in i)roper tniiniiig and thoroughness, but the manufuclurers have become accusloined to u ceriaiii class of inspection, and would deem it a hardship to be subjected to much more rigid requirements. Eventually, however, the re suiting improvement iu mauufivcture of metal-work would be an advantage to the mauufrcturers as well as to the pur- chasers. A decided improvement iu inspection can be brought about only by concerted action on the part of the principal inspecting bureaus and inspectors of the country, backed, of course, by tlic aid of all engineers who are directly interested in the de- signing and building of structural metal work. If these in- specting bureaus auR PONTIBCS. of Ibis propusiliou to form nn nssociation of inspectors, term- ing it a " trust." Strictly speaking, it certainly would par- take of the nature of n trust, but it would bo a good and worthy one, whose main ohj ct would be to effect a much- needed reform. On the s:une basis the American Institute of Architects is a trust, for the reason that it establishes a miui- nnnii fee of five per cent for the making of plans and specifi- cations and for the services of an inspector on all building work; and surely such an organization should not be con- demned on this account. On the contrary, the architects have set the engineers a good e.xuniplo in forming this association ; and, until engineers follow their lead in this particular and es- tablish minimum fees for professional work, the engineering profession will fail to attain its higliest degree of ctticiency, and will therefore not be properly recognized as a profession by the general pubMc. Returning to the (piestion of the inspection of stnxctural steel, the author herewith presents, as his idea of what good inspecticm sliould consist, his standard instructions to the iu- spccling bureau which he employs and to its inspectors at mills and shops. First. Study carefully the engineer's drawings as soon as they are finished, and make out a list of special points and features that will require e.xtra ciive in the shops to secure good workmanship and proper fitting, then niake out a type- written report of these and submit it without delay to th", Engineer. Second. Study carefully, as soon as they are finished and approved, all shop drawings, so as to beco.ine thoroughly familiar with the entire construction. Third. Make sure that metal of uinform character and of the strength, elasticity, and ductility specified is furnished by the rolling-mills, folio vin^- the metal from one process to an- other from start to finish, and making sure that the test-pieces broken represent coirectly the metal they are supposed to represent. Fourth. Check tiie chemical analyses of the metal occasion- ally, so as to gee that they are properly made, taking care th>it INSPECTIOiN' OF MATRRFALS AND WORKMAXSKIP. 285 the Contractor is Inforinetl as to what piece the samples arc takeu from, so that lie can make a check lest, if he so desire. Fifth, See lliut all the various tests iiulicjilid in the specifi- calioDS are made faithfully, the iiunl)er of same depending- upon the relative uniformity of tiie moljil furnished. Sixth. Make sure that all the puncliing is done with sucii care that the assembled parts will come together so as to mnku the rivet-holes match so accurately that when the reaming is tinisheil there shall be no imgidiir holes. Seventh. Make sure that all pieces are cut to exact length and proper bevel, that all web-stiffening angles bear perfectly at top and bottom against tlange angles, and that there are no loose rivets. Eighth. Wherever rivels with flattened heads or counter- sunk rivets are calleise in the structure, no matter how many limes it may have already been in.spected and passed. Twelfth. See that all metal-work is proi^erly cleaned by the most approved methods and apparatus bjjfore the tirst coal 0/ 2H6 DE PONTIHUS paiiil is applied, and tliiit tlio latter is allowed to dry thor- oughly before the metul-work is loaded ou the cars for ship- ment. Thirteenth. See that all shop painting is thoroughly done, and that proper paint, mixed so us to comply with the specirt- cations, is invariably used ; and make an occasional chemical analysis of the paint, taking care thai the Contractor is noticed of the contemplated test aftiir tiie samples ary taken, in order that he may make a check analysis, if lie so desire. Take special care to prevent any pieces of metal from being riveted together, unless the contiguous faces be first thoroughly painted. Fourteenth. Insist upon the discharge of any employee of the Manufacturing Company who wilfully violates or continues to violate the specifications and the instructions given by the Engineer or his inspectors. Fifteenth. While endeavoring in every possible way to ob- tain good work, avoid as nuich as possible doing anytiiing to annoy or harass the Contractor; but, on the contrary, take special pains to aid him in every legitimate manner to finish his work quicUly an. I inexpensively. Sixteenth. Formulate and prepare for each large piece of work the best practicable method of recording progress and reporting thereon, and divide up tiie total work into groups or sections, so tiiut the notes may be easy for reference. Tliis should be done by the inspecting bureau, and should not be left to the shop inspector. Seventeenth. Send into the office of the Engineer regular weekly reports concerning the progress of the work, any special reports that from time to time appear to be recpiired, the tabulated results .of all tests of materials, and copies of all .shipping bills. Eighteenth. Make sure that all shipping weights are correct by seeing the metal weighed, and keep account of the weight ©fall metal sent out on the work, as the Contractor will be paid by the pound. It will be necessary for the inspecting bureau to check all of these weights against the shop drawings to show how they agree or disagree. A detailed statement o.f TNSI'KCTIOX OF MATKKIALS AND WOWKSfANSlIIP. 287 both sots of weights must be sent to the Engineer upon the coinpic'tiou of tlio contract, or, at his request, upon the com- pletion of an}' (Jelinito portion tliereof. Nineteenth. Tiie inspecting bureau shall, under no ciroum- stances whatsoever, intrust respousil)le work of any kind to iiisufticiently trained assistants. Wiien new inspectors are to be broken in, they must receive their trainii g in such a way as not to jeopardize in tiie slightest degree the quality of the material or workmanship. Twentieth. Finally, and in short, do all you can to make the structure in every sense of the word a credit to uU concerned in its designing and construction. The author has had m-ide for him lately by Mr. 11. T. Lewis, one of his inspectors, a rather interesting series of tests to de- termine the average accuracy of punched rivet-holes. These tests were made after the metal w.is assembled for reaming by inserting rods of various diameters in the assembled holes. From the results of these tests the author has prepared the following clause for the speciticalions given in Chapter XVIII. "All pupphed work shall he so accurately done that, after the various component pieces are assembled and l»jfore the reaming is commenced, forty (40) per cent of the holes can be entered easily by a rod of one si.xteenth (^jf) of an inch less diameter than that of tiie punched holes; eighty (80, per cent by a rod of a diameter one eighth (i) of an Inch less than same; and one hundred (100) pir cent by a rod of a diameter one (pnirler {\) of an inch less than same. Any shop-work not coming up to this requirement will be subject to rejection by the inspector." It will be noticed that this specilicaiion does not reject absolutely all work that does not come up to its exact require- ments, the inspector being allowed some latitude in distin- guishing between simple and complicated sliop-work, imi)or- tant and unimportant connections, and the assembling of few and of luinu'roiis component pieces. If the Association of Inspectors herein suggested were estab- lished, it could do good work for the engineering professiou hy laying out a. series of tests of full-sized jDciubersautJ details 288 DE PONTIBUS. of bridges and other stnictunil metal-work, to be made from time lo time as a portion of tlie inspection for large contracts. Tliis would need tlie assistance of tlie consulting engineers, who, In preparing their specifications, sliould include, as a part of the work of the manufacturers, the making, under the supervision of the inspectors, of certain tests of full size parts, it being understood at tlie outset that the results of such t«'8tF shall be of direct value to the accomplishment of the work covert il by the specifications The author has for the past five years been endravoiing in this way to obtain some much needed information concerning tlie strength of both main members and details of bridges and elevated railroads; but his attempts to have the tests mtide have not always proved suc- cessful As foi- the p»*oper price lo pay for firsl-class inspection, the author would slate that some tliree years ago he s«d)mitted to several of the principal inspecting liureaus a draft of instruc- tions to inspectors S AND WUKKMAN.SHIl'. ;280 Secoud. See thai the metal-work goes together properly and expeditiously, and report to the Engineer all necessity for chipping or filing on account of bad shop-work. Third. Watch carefully the riveting to see that no burnt rivets are used, that all field-rivets are driven in accordance with the specifications, and that uo loos*' rivets are left in the work. Fourth. See that all vacant spaces in the nietal-woik are completely filled with paint-skins or other water-proof male- rial before the painting is begun Fifth. In elevated-railroad work see that during the erec- tion of the metal- work the lengtlis of the girders are sutliciently correct to prevent all po.ssihility of using up the spaces pro- vided for expansion, assuming tlie greatest temperature of the metal to be one liundred and twenty-five degrees. See also that the expansion and contraction of the structure cannot in- jtire the stairways. Sixth. In drawbridges, see that the masonry of the pivot- pier is levelled off with the greatest accuracy, and that the lower track-segments are set to exact position and level, thus making a perfectly conical surface for the r;)ller8. See also that the latter are adjusted so as to bear evenly at top and bot- tom against both upper and lower track-sei^mcnts. Seventh. See that the ends of draw-^nans are properly adjusted by means of the shimmiiig-philes on the rest piers and those in the bottom chords near the pivot-pier. Make sure that in every particular the draw is veversible end for end; and see that all shafting is properly aligneu .othat there will be uo binding in any of the bearings. Eighth. See that, before the operating machinery is tested, all slidingor rolling surfaces are thoroughly lubricated, and that the turntable is cleared of all obstructions, such as nails, etc., on tl" lower track-segments. Then make a ti.'st of the machinery and compute therefrom the horse-power required to operate the draw. Ninth. See that all anchor-bolts are set in exact position and to correct level, and that they are projjerly grouted in. Tenth. In placing the bearings for arches, take tlie grc/i] 290 DE PONTIBUS. est cnre that the ceiilres are set to exact position and level, and that the bearings for the metal-work on the masonry are perfect. Eleventh. Whenever there are any atlju.stal)le rods used In a structure, see llmt they are properly tightened before the work is left, taking care thai they are not screwed up more tightly than is really necessary. (b) kailb. First. Examine all rails us soon as received, so as to see that there are no poor ones which have escaped the rail-inspector's eye, or which have been loaded for shipment after being re- jected. Inspect also all other track-metal, such as angle-bars, bolts, and biiices, so as to see that they are of the correct type and are delivered in good shape. Second. See that all rails are l;iid to exact line and levi-l, and that they bear properly everywhere. Third. In draw-spans, make sure that the track-rails at the ends will not interfere with the operation of the draw. (C) PAINTING. First. See that, after proper cleansing and retouching with paint, the metril-work receives its first field-coat of paint as soon as practicable after erection, and that the next coat is applied as soon as practicable after the first field-coat is thoroughly dried, but in no case before. Second. Make sure that all paints used are of the proper color, qualit}', and consistency, and that no adulterants or thinners are used ; also, that all paint is properly applied. Third. Look carefully to the painting of all close spaces between metal, and see that it is done efTeclively with a piece of cloth, according to the specifications. Fourth. See that all jtortions of the metal-work, which are to rest on the masoiny or which are to bo embedded in the concrete, receive their two field-coats of paint in due time, so as to dry thoroughly before the said metal-work is erected. INSPKCTION OF MATEIUALS AND WOKKMANMim'. 291 {O) EXCAVATION. First. Watch carefully all excavation so as to make sure that it is (lone in strict acconinncu with the spcciflcntioiis and witii the City Ordiiiaucus, if there be any. See tliat, in doing the exciivalioM and in building the structure, the Contra(;to.r does not obstruct public tralHc. Second. In foundation-work in cities, see tliat all pipes and sewers are nu)ved properly and coupled or spliced effectively after being uncoupled or cut, '{''liinl. Whenever there is any doubt about the proper re- sistance of any foundation, test it by loading it by means of a pn»porly designed and built apparatus. Always ram thor- «)ughly any foundation wliere liie resistance to loud would be increased by such rauuning. See tliat the materiul from the sides of (he pits is prevented from falling in. Fourth. See that all surjilus material is removed expedi- tiously from City streets, and that, whenever any piece of con- struction is completed, all falsework, rubbish, etc., are re- moved from the site and are deposited in an unobjectionable place. i(!h are in the me, so Led. (e) foundations. Firet. See that the bed-rock is always properly prepared to receive the caisson or masonry, as the case may be, letting the caisson into the rock so as to provide an even bearing around the cutting edge, and levelling or stepping off or filling up with concrete to receive the latter. Seconil. In elevated-railroad work, see that wherever colmnns are located in the street their feet are properly en- cased in concrete, and that cast-iron fenders are correctly set around the columns and filled with concrete and grouting, (hen sealed effectively against the ingress of water. See also that, after the columns are tip ami encased, the pavement is relaid in a substantial manner, to the satisfaction of the City au- thorities. Third. When large steel cylinders are used, see that the^ 292 DE poNTiiura. n re kept well braced with limbers on tlio inside durinj,' sin K- ing, HO lis to avoid all possibility of collapse. Fourtli. Soe tliiit proper gindes arc provided for all cais- sons and cylinders, so that Ihey can l)e kept in exact Iiorixontal position during the entire linking. Fifth. Sec that the tops of all pit;rs arc properly tliMslied off to receive the snperstructnrc, taking care lliul all bearings are made perfectly smooth and to exact level. (f) caissons. First. In l)uilding timber caissons, see that the plans are followed exactly, and that the full ({uantum of timber bolls is used ; also, thai short timbers are not put in where long ones are called for. See that all timbers are properly framed. Second. In sinking caissons, see that tlicy are never allowed to deviate matoriidly from correci, position, and that all errors of position are corrected as soon as oossible after they are dis- covered. Tliird. In (iliing working-chambers of caissons, see that the concrete is packed tightly against the roof, and that no voids whatsoever are left therein. (a) CEMENT AND CONCHiSTE. First. Test all the cement, according to the special instruc- tions therefor, so long before it is needed for use that the Contractor shall not be delayed by such testing. Second. See that all cement is housed .so as to be protected effectively from the weather, and that no dampened or other- wise injured cement is allowed to be used on the work. Third. Inspect as soon as delivered, and if possible before it is dumped on tlio ground, all sand and broken stone, so as to make sure that they comply in every particular with the specifications ; and insist always upon all of these materials that are rejected being removed immediately from the vicinity of the bridge site. Fourth, See that strong and proper forms for concrete are wspd in the cpusiruclirn of all j[)cdestals, and that all visjbJii INSPECTION' OP MATERIALS AND WOUKM ASSlIII'. 293 portions of tUe hitter uic tinislicd oil smooth, tiiu top Hiiiface hi'iiig hroiijjht to exact elevation and made perfectly level. Fifth. See tliat all concrete is mixed accordiiiij to the speci- lications, that it is put in place inunediately after mixing, aud that it is thoroughly rammed. Sixth. See that no injury is done to the concrete in remov- ing the timber forms, or, if any be done, that it be properly repaired ; also, that the timber be left in whenever its removal would tend to injure the work. Seventh. When concrete is placed under water, .sec that either a tremie or proper collapsing-bucket be used, and that the water be not permitted to injure the concrete. See al.so tliat all such concrete is mixed extra rich. (ll) IMLINQ AND TUKSTLEWOUK. First. See that all piles conform in size, quality, and straightness with the requirements of the specifications, even if they have been already passed by the timber inspector be- fore shipment, and reject any that are unfit for use. Second. See that all piles ai'c driven straight and in proper position, and that the tops are not unduly injured in driving having the said tops banded, whenever necessary to prevent splitting. Third. See that all piles are cut off at the exact elevation required, and that the caps are properly drift-bolted thereto. On curves see that the superelevation is obtained properly, and not by shimming up on the caps. Fourth. See that all sway-bracing is bolted effectively to the piles and caps. (I) TIMBER, FLOORING, AND HAND-RAILS. First. Inspect all timber as soon as delivered, marking plaiidy all rejected pieces ; and see that all such pieces are removed from the vicinity without delay, in order to prevent their being put into the structure without the knowledge of the resident engineer. It is, of course, permissible to use the ^94 DE PONTIBUS. good portions of rejected timbers ; but iu doing so great rare should be exercised to prevent tlie workmen from putting uiiy poor materiiil into tiie work. The fact tluit all the timber re- ceived had been previously uccupted by tlie timber inspector is no reason for using unsatisfactory material ; moreover, sometimes it happens that ti nbers which the inspector iins never even seen are marked with his stamp and siiipped. Second. See that the floor system is properly laid and attached to the metal-work, that each rail bt-ars effectively upon every tie which it crosses, and that the rails are laid stiaight, evenly, and to exact grade. Third. See that the hand-railing is brought to proper align- ment, and is held there in a permanent manner. Fourth. See that all joists in highway bridges are properly dapped on floor-beams so as to bring all of their uppur sur- faces to exact elevation or elevations ; also, lliat all inter- mediate joists lap past each other far enough to reach entirely across the top flanges of the floor-beams. See that the outer lines of joii^ts abut and run continuously, and that they are effectively spliced on the inside. Fifth. See that all joists in which the depth exceeds fo;ir times the thickness are bridged at distances not to exceed eight feet, and that when the hand-railing depends for its rigidity, upon that of llie outer joists the latter are well bridged and otherwise stiffened where the posts are attnched. Sixth. See that alternate bolts attaching guard-rtiils to floor pass through both the flooring ami the outer joists, and that all holes through the latter are bored in the central plane of tlie joist. (j) MASONRY. First Inspect all stone as soon as received, so as to see that It has not been injured in transit, and that it is satisfactory in every particular, even if it lias already been passed by the stone inspector. Second. See that all stones are thoroughly cleaned and wet before being laid. Third. See that all mortar is mixed in the proper propor- IJJSPRCTIO^f OF MATKRIALS AND WORKMANSHIP. S<)5 lion, and that it is used ou the work belore any set has occurred. Fourtli. See that all joints are thoroughly flushed with niortiU', and that no voids are left anywhere in tlie masonry. Fifth. See tliat all coping-sloues are set so that the lop of the pier will lie in a truly horizontal plane, and that they are kept in place by proper clamps and dowels as per plans. Si.xth. See tiiat the exposed joints are all cleansed and jiointed in a thorough and workmanlike manner, and in accorduuce with tlie specilications. (k) geneual instructions. First. See that all proper precautions against accidents to the public and to the workmen be taken during erection, and that no glaringly careless man be allowed on the work. Second. If there be more than one Contractor on the work, see that no friction arises between contractors, and that their combined work is flnish6(l in good shape. Third. WiuMe doing everything in your power to obtain gooil work, avoid as much as possible worrying or hat assing the Contractor, and use every legitlr'iate endeavor to aid him to complete his work expeditiously and inexpensively. Fourth. Finally, and in short, study the specifications care- fully, and do all that you can to insure the structure's being in every respect a credit to all concerned in its desigidng and construction. In respect to the testing of cement on construction work, the following iiiNtruclions, which the author has pre[);ired for his resident engineers, will give the reader all necessary infor- mation, it being understood that no brands of cement are ever used except those which either the author or his a.s.sistaut3 have previously tested thoroughly by long-time tests, and which have proved to be perfectly satisfactory : First. In testing cenient in the tield, remember that it is not a series of laboratory tests which you are to make, but that your object is simply to see that you are receiving and using cement of an average quality of the standard brand or brands 296 DE PONTIBUS. adopted, and that it comes up to the geueral requirements of the specitications. Second. Look out for irregularities in the quality of the cement, so- as lo avoid using any that is either loo old or too fresh, or wliich has been iMJuietl by dampness. Third. Test tirst for fineness, second for soundness, third for activity, and fourth for rise iu temperature, rejecting all cement which is imtit for use liecause of non-compliance with the specifications in these particulars. Fourth. It will seldom, if ever, be necessary to resort to the boiling test, which is essentially a laboratory test ; although it may prove useful iu an emergency to determine conclusively whether certain cement is fit for use or not. Fiflii. Test all cements for the ten.sile strength of neat briquettes, making one day and seven-day tests. Never pass cement on shorter time-tests than seven days, as the one-day test is by no means conclusive. Sixth. Make, more for your own satisfaction than for any other reason, a few sandbiiquette tests for seven and twenty- eight day.s, so as to know the value of the mortar which you are using. It would not do to rely on sand-briquette tests for the acceptance or rejection of cement, as this would delay the work too much. Seventh. You will often have to use your judgment about passing or rejecting cement that is needed for immediate use and which falls in some compaiativel}' unijuportant point to quite fill the reciuirements of tlie specitications. Rather than delay the Contractor materially, pass such cement, provided that in your opinion its use will iu no way injure the quality of the work ; but, on the oilier hand, if the rejection of the saitl cement will not delay the Contractor seriously, in.sist on its complying with the specifications in every particular. Be careful not to let the Contractor run iu any poor cenienl or force it upon you because of any assumetl or real necessity for haste in completing the construction. In respect to inspection of stone for masonry, the author offers, as his idea of what stone-inspection should be, the fol- lowing instructions to stone-inspectors, it being understood INSPhCTION OF MATKUIALS AND WORKMANSHIP. '^D"? that they apply only to stone from quarries that have been previously investigated and found satisfactory: First. Ileject all stone conlaining iinj'^ dry seams. These seams are often very hurd to delect ; but u>ually l)y a cnreful inspection of the surface of the stone they may he found. Sometimes a mere line is all the evidence of the existence of such seams, while in olhe'' cases they show nion; jilainly. Second. Reject all stone containing seams called "crow- foot," which are either open, or which are liable to dissolve out after exposure to the weather. Third. See tiiat no stone is quarried at a time when it is liable to freeze before the quarry-sap is out of it. Stone sliould be quanied at least a montli before it is allowed to freeze. Fourth. See that no powder or other explosive is used in quarrying the stone, excepting to remove ledges of useless stone, and even then make sure that no stone to be used is injured by the explosives. Fiftli. If the stone be of such a character that tlie quarry, bed cannot be told at a glance, the Inspector must nuirk eacli stone in such a manner that it will be sure to be laid in the wall on the .said quany-bed. Sixth. Reject all stone whicli is taken from any portion of the quarry that is affect el injuriously at any lime by frost. Seventh. See that all stone is handled carefully after being taken from the quarry, so that no cracks are developed or other injury done thereto by rough usage. Eighth. See that all stones arecut to the exact dimensions called for by the plans, and that they comply in every partic- ular with the specilications. In respect to inspection of limber, both in the woods and at the sawmills, the author's instructions to his timber-inspectors, as follows, will be found useful : First. Study well and compare with the mill pet)ple all order-bills, looking carefully to the various lengths, widths, thicknesses, bevels, numbers of pieces, etc., so as to make sure that your order-bills check properly against those furnished to the mill people and against the partial order-bills furnished 298 DE PONTIBirS. by the latter to their various employees, so as to avoid all pos- sibility of errors. If any be fomid, correct them yourself, if possible; but, if imt, refer them to the Eiigiueer for correc- tion. Second. Each timber-inspector is to be provided with a special stamping jpimmer of his own, that lias a characleristie mark which will identify all timber passed by him. He is to keep this hammer at uU times in his own possession, so th&t it can be put to no illegitimate use by interested parties ; and under no circumstances is he to lend it to another inspector. Third. Each timber-iMsi)ec'tor must study carefully the 8|MJcifications furnished him, and must be governed thereby; nevertheless, there will be occasions when he must trust to his own judgment as to what timber is tit and what is until for the recpiired purpose, fy probing with a wire jill hollow or bird-eye knots, and, should the hollow be over one inch deep, reject the limber. Eightli. Check lengths of cutting gauges every day, as they are liable occasionally to be knocked out of position. Ciieck widths and thicknes.ses at each change of the machine. Ninth. In inspecting i)iles, look carefully to their straight- ness, and see that they comply in this and in every other par- ticular with tiie si>eciflcations. Tenth. See that due care is used in handling and loading timber so as not to bruise it ; and under no consideration allow it to be floated in the water after it is cut and dressed. Eleventli. Keep a daily record of all timl)er accepted, .so that the Engineer may be informed on short notice as to how much of any bill has been cut. Twelfth. Notify the Engineer or other proper party of all shipments, and keep an accurate account of everything shipped, so that upon shojt notice a statement in respect to any uncompleted order can be made, giving the amount that has been shipped aud the amount that reuidins to be for- warded. Thirteenth. Tlie Cliief Inspector must make regular monthly reports to the Engineer or other proper party or parties con- cerning the progress of the work, (pnilily of timber furnished, etc.; and must send in montiily statements of all moneys 300 t)E Po>rntn:s. received and expended by him in couneclion with his work of inspection. Fourteentli. Use every endeavor not to cause by your in- spection iuiy more biiudling of material than is necessary for doing your work thoroughly ; and do nothing to give the mill people needless wo* ry or expense. In concluding th s chapter, the author desires to emphasize his previous stateiaent that, in order to obtain a truly tirst- class structure, it is necessary not only to design it properly and prepare thorough specifications for its building, but also to provide a corps of competent and honest inspectors, who will from start to finish examine carefully and test all mate- rials that are to be used, and who will see that the entire manufacture and erection are done in strict compliance with the specifications. CHAPTER XXII. DESIGNING OF PIERS. The object of tlii.s chapter is not to provide the bridge- builder with either a complete specifiailioii for building piers of all kinds or full directions as to sinking them under nil possible circumstances, but to indicate to the designer, tirst, how to determine the best kind of piers to use at any pro- posed crossing, and. second, how to proportion them. Text- books on substructure do not generally cover tiiis ground, but deal mainly with masonry specifications and methods of sinking piers. The reader who desires to learn anything con- cerning piers which is not given in this chapter is referred to Baker's " Treatise on M:isonry Construction " and Pattou's " Practical Treatise on Foundations." In determining the layout of spans for any important cross- ing, the first (juestion to settle is what method of pier-sinking to adopt, for upon this will depend to a certain extent the span lengths. The three principal methods in common use are as follows : 1. The Coffer-dam system. 2. The Pneumatic j^roce-ss. 3. The Open-dredging process. The use of coffer-dams is, or should be, limited to crossings where the bed rock is not more than fifteen feet below the ordinary stage of water, and where there is no great, sudden rise anticipated. This method always figures low in the pre- liminary estiiiiate, but is generally found to run much higher when the total cost of the finished structure is co:npute(l. The author nearly always discourages his contractors from attempting to use this jnethod ; and thus far l>is experience FROVhNC'^L LIORARY VICTORIA, B. C. 305; DE rONTIHUS. proves that, when they fiiil to adopt iiis advice about it, I hey are generally sorry therefor by the time tiie work is finished. Coffer-dams are liable to give trouble in several ways: first b}' leakai^c, second by flooding, and tiiird by collapsing. If a Contractor gets tliroiigh a large piece of (!ofTer-dam foundation work witiioul accident or trouble of SiUie kind he is in great luck. For bare bed-rock, movable coffer-dams may be emjiloyed ; but they are troublesome to con.slruct, and are sometimes very (lifHcult to remove because oi a deposit of sand taking place while the piers are being built. Tiie pneumatic process for sinking piers is in most cases the best one to employ, the only objection to it being the exces- sive cost of installing the plant, even if one has a complete outfit at his disposal. Its great advantages are that it enables the contractor to overcome, in thecheipest and most expeditious manner possible, all obstacles that may be encountered in sinking; and that it ensures the obtaining of a satisfactory foundation for the piers. It can be used for depths as great as one hundred feet or even more, although there is consider- able danger to the workmen when the depth exceeds eighty or ninety feet. Most of the bridge-piers which the author has put in have been sunk by the pneumatic process ; and he has no hesitatio!i in recommending it as the most satisfactory, all-around method in probably nine cases out of ten which occur in a consuliing engineer's practice. The open-dredging process is suitable for very deep founda- tions, or for putting down caissons that are to rest on the sand, or for bed-rock foundations where there is no liability of great scour. For large piers this process is much cheaper tlian the pneumatic on account of both the smaller cost of i)lant and the more rapid progress in sinking. In case, however, that ob- stacles be encountered, such as trees or large boulders, the expense for sinking is liable to run high, as these obstacles may have to be removed by a diver or divers, which always involves great expen.se. The author has piit down three large piers by the open-dredging jn'occss, two to u depth of uinetv nnsioNiNo OF piers. 303 / f(H't iind one to a dcptli of one hundred and twenty- two feet below extreme low wnter, and lian encountered no trouble worth mentioning during the sinking In the case of the greater depth, a mass of bouklers was found overlying the bed-rock. This was penetrated as far as practicable by ex- cavating the boulders and laying bare the bed-rock near the centre of tiie pier, then tiring charges of dynamite iil the bot- tom till the cylinder refused to sink any farther, after whidi it was tilled up with concrete. It is probable tliat, if one were to try to sink small pi«.rs to any great depth by the open-dredging process, difflcully would be experienced because of the lack of sufficient weight of pier as compared with the large amount of skin friction. The latter in sand is generally a little less than six hundred pounds per square foot of vertical surface. On the East Onudia bridge the author arranged to reduce this friction by means of small wjiter-jets placed around the circumference of the cylinder about every six feet in height ; but these were found to be unnecessary, so were not used. In case of striking clny or any other sticky substance, such an attachment might prove of great service. The open-dredging process is liable to abuse by the builders of cheap highway bridges, who, in order to save a little in first cost, use it to sink cylinder piers of small diameter moder- ate distances to bed-rock, which may in these i)l:i(es be laid baie or nearly so by excessive scour. With this [n-ocess it is generally not practicable to anchor the cylinders tirndy to the bed rock, but with the pneumatic process it is. There is still anoliier style of foundation besides the three described, viz., that which involves the use of piles. These piles may either support a timber grillage, upon whieli to rest the pier proper, or may run up into the concrete body of the pier. This class of foundation is of a cheap or I»K I'ONTIHUH. 4. Coinbiiied lliiusi from lightest possllilc limkcd train, mid ft wiiidpressure ou tiuiii uud structure fiitiul to one half of llmt 8pecilied. Next delenniue by judgment the propi;r batter, and hiy olT the pier to scale; then divide it by horizontal planes from four to six feet apart, and comii/.e tlie weights of iil! the |)ortion8 of the masonry between hese planes, making a proper reduc- tiou for weight of water for those parts which would be sub- merged by an average stage of river. Next compute the wind-pnssure on each vertical division of the pier, down to the assumed stage of water, in a direction parallel to the spans, using the same intensity and direction for the wind-pressure us were adopted in finding the longi- tudinal thrust from wind-iiressure on the spans. Next lind graphicidly for all four cases the curves of pres- sure from the vertical and horizontal loads at top of pier, com- bined with the weights of the various divisions of the latter and the wind-pressures thereon, and see that none of the said curves at auy plane of division pass outside of the middle third of the section at said i)lane. If any of tliem do, the batter will have to be increased, or, if all the curves fall mucii inside of the middle third points, it will have to be decreased; and in eitiier case the graphical computations will have to be made again, and so on until a satisfactory batter is found. The author is aware of tlie fact that this method of design- ing piers is not in general use, and it is (juite possible that he is the sole engineer who adopts it ; nevertheless he maintains that it is the only proper way to design masonry piers. The single concession wliich he would be willing to make on the .score of economy would be to assume that a certain small portion of the thrust ou a span is taken up at the roller end. But if the rollers are in good working order the amount of thrust that they will resist is very small indeed — so small, in fact, that the author prefers to neglect it entirely. The ordinary method of i)r()portioning piers is to make them as small as possible under coping and baiter them all around, or at least on the sides, one-half inch to the foot. In some cases this will suffice, but in others it will not. One of the largest insuillci untraini these pi omy in properlj An in general) superstri properly lever bri out four inch to author t pieis to nppearan him, and inches to tory appe In neaj stream, ( I lie proj strength sure. A of piers i Construe enough resist pro combinat Neverth* high piei ■wliere th( as a mat overlurni length of neatest v cocked-hi Where DKSKJNINO OF IMKKS. 3or \: litrgcst bridges in tbc Uiiilcd Sliitcs lius piers l)uilt witli sucli insullicieiit biitlcr tlial il is evident ut n gluQce, to eveu i;ii uiitruined eye, timt something is wrong. By tlie way, one of these piers is cniclied from lop t«^ l)ott()ni. owing to false econ- omy iu the design, but not lieeause of its failure to /Igure properly for the curve of pressure. An inherent sense of fitness in the mind of the designer will generally lell him, when Ijo looks at a scale-drawing of the superstructure and piers of a bridge, whether the latter are properly proportioned. In the case of the : ,cd liock canli lever bridge over the Colorado Kiver the piers were first laid out fourteen feet wide under coping, witli a batter of half an inch to the foot, and the drawings were submitted to the author for his criticism. He immediately pronounceil the piers to be proportioned incorrectly, simply because of their appearance. Their proportioning was then turned over to him, and lie found by trial that a batter of one and a quarter inches to the foot was necessary. This batter gave a satisfac- tory appearance to the entire layout. In nearly every case the leugtli of the piers up and down stream, determined by the minimum size under coping and the proper side-batter for thrust, will provide sufficient strength and stability to resist both current and wind pres sure. A thorough investigation of resistance to overturning of piers down-stream is given iu Baker's " Treatise on Masonry Construction." In il he proves that any pier which is large enough under coping, and which has ordinary batter, will resist properly the overturning tendency of the worst possible combination of loads from wind, current, and floating ice. Nevertheless, in long-span, single-track bridges with very high piers, crossing swift streams that carry thick ice, and ■where the structure is exposed to high winds, it is advisalile, as a matter of precaution, to test the piers for down-ptream overturning according io Prof, leaker's method. Should the length of pier parallel to the stream be found insufflcient, the neatest way to obtain the requisite stability is to put in a cocked-hat just above the elevation of extreme high water. Wljere a masonry pier rests on bed-ro nose, and will thus cause a cracking of the masonry and a uplitling-off of the front of Ihe pier. Such a disastrous result of the violation of llie principle of symmetry in designing is by no means unknown, even in important railroad bridges. In respect to timber-and-concrete caissons for masonry piers, the following general remarks will prove useful to the designer : There should be an offset of not less than two feet all around the base of the masonry, and preferably a little more at the ends, so that in case the caisson be located a little out of place the masonry can be shifted thereon so as to bring the pier into proper position. The number of courses of 13" X 12" timber in the roof of the caisson should never be less than four and seldom more than eiglit. Any less number than four would be liable not to give the roof the proper stiffness during the sinking, and any more than eight would tend to cause an undue settlement of the l)ier on account of the compression of the timber, which always takes place. The designins; of the roof and sides of I lie working-chamber should bo lone with the greatest care, so as to preveut all possibility of collapse, and tlie cutting edge should be shod with steel plates to protect the timber when the caisson is piussing through bouldirs or logs. Tiie roof-timbers, if possible, should always be of full length, and the spacing of the bolls therein should not e.\ceed four feet. The vertical timbers on the outside of the working-cliamber should be carried well up into the roof, shoulderi d, and firmly bolted thereto. The crib above ti.'! working-chamber should DESIGNING OF PIERS. 309 be sbeiitlied so as to reduce the friction duriug siuking. The drift-bolts should be seveii-eigliths-mch rouuds driveu into three-quarter-iiicli bored holes. The tilling of the working- chamber with concrete should always be done with the great- est care, using extra ricli concrete, so that there shall be no voids between the concrete and the roof. Portland-cement of the best quality should be used for filling the working- chamber and shafts ; but it is legitimate to employ an extra- good q\uiliiy of Americau natural cement for filling the crib in case that it be necessary to keep down the expense. How- eve:', Portland cement is always prefera'>Ie. In respect to caissons built of steel and concrete but little need be said, except that great care should be taken to design the working-chamber strong enough to resist properly the weight of the concrete above and the unequal pressures from boulders below. The metal below the roof of the working- chamber should not be less than one-half inch in thickness, and all parts near the cutting edge should have thicknesses varying from three (pinrters of an inch to an inch. All joints in the cutting edge siiould be full spliced, as should also those in the roof of the working-chamber. Timi»er grilliig<'s resting on piles should have, preferably, not less than four courses of timber, although often but three and more rarely two are emp'oyed. As the grillage is gener- ally wider than the masonry, it takes about four courses to distribute the weight uniforndy (or nearly uniformly) over the piles. In case of an unusually wide grillage, more than four courses would be necessary, or else the masonry should be widened by means of footing-courses. Such grillages should be built with care, so as to have a level bottom; and all piles should be cut oft to exact level, otherwise there will be un- equal bearing between piles and grillage that might cause serious damage to the masonry. Brick piers are not common in America, probably because, until lately, it has been difil ilt to obtain jiroper brick. In the author's opinion, piers built exclusively of hard-burned clinker brick and mortar of the very best quiilily of Portland cement, nnxetl in the proportion of one i)art cenumt to 310 DK PON Tift US. two parts sand, and having tliiu joints perfectly filled, are better than the average masonry pier, for the reason that tlie bricks will never disiulegrale, while the average stone vised for bridge-piers will. The author lias not yet had occasion to build any brick piers; but he intends to give tiiem a trial on the first opportunity. Unprotected '-one -h piers are satisfactory for Southern rivers, where tic ''< • ' frost is not severe, and where there is no ice of any juiioiuit carried by the stream. The author used this style of pier for the Arkansas Kiver bridge of the Kansas City, Pittsburg, and Gulf Railroad near Redland, Ind, Ter. SevLM-al othtr Southern bridf^es iiave piers of this type and Ihps lar they have proved satisfactory, Tlieir chief reccni- mendation is their cheapness. In order to ensure their being properly built, nothing but the best qualities of cement and sand should be employed, and the mortar for the concrete should be mixed rich, cspeciMlly uv.m the exterior of the piers. Some engineers give the work a skin-coating of ri'jh mortiir; l)Ul the author prefers to use fuuly l)r()ken stone and extra- rich mortar for six or eight iiiches all around, and to not at- tempt to smooth down the •;>( Mor. Of course it is practi- cable to put on a .skin-cou's > f'.o U will stay, and M) that it will not have a streaky !i ;■;■, r;,ic ■ but to do this requires more fare tiian liie average wo; .r .jm is inclined to take. Steel shells filled witli concr<'le ;.!.i.ue very satisfactor}' piers, provided they be not used in salt or brackish water, wliich would rust them out in a short, time. These piers are appli- cable where good stone for masonry is < ^pensive, and where the piers must be protected from the abrasion of ice or from the excessive »old, which would tend to disintegrate even fairly good concrete. Such piers can l)e built in the usual form of masonry piers ^ ■'; rounded ends all the way up, or. when they pass much u-^^' high water, they may run olT into two cylinders with bracin.,; urt-veen. Butt-sp! ces are prefer- able, and the; , .lice-plates below the mud linoshould be placed on the in^.de soss to oiler as little resistance as possil)le to sink- ing. This style of pier is a favorite one of the author's, for the DESIGNING OP PIERS. 31i reason tliat it is both sightly and inexpensive. When taken to task for using it, as often hapi)ens, lie replies, " Good con- crete protected with steel is better than poor masonry." In respect to the thickness of stiiel to use, the author's prac- tice is to adopt lialf an inch below tiio ordinary stage of water and tliree eighths of an inch above, although for cheap bridges he occasionally shades these thicknesses one sixteenth of an inch. For the coping of such piers stone may be employed ; but it is preferable to put on a moulding of siieet metal, as this is more in keeping with the vest of the pier. This .style of cop- ing has been criticised on the plea that it is false, and that it has no direct function ; nevertheless, the author considers it eminently proper to use it, and that us function is .simply to beautify the construction by relieving the harsh outlines. Where stone coping is not used, the top of any kind of con- crete pier may be finished off with either ricli concrete of small broken stone or with granitoid, mixed in tlie proportion of one part of Portland cement, two parts tine granite screen- ings, :ind three parts of small cr>isiied granite. Cylinder piers filled with concrete are the most common kind of pier in America, and they are certainly the worst; nevertheless they have their i)lace in good construction* when they are jiroperly designed and binlt. Their abuse is due mainly to tiie builders of cheap highway bridges, who think that if the top of the cylinder is simply large enough to hold the pedestals, tiiatisall which is necessary, no matter how high the piers may be, how great may be the scour, or what kind of foundati(m there is. If piles are employed as a foun- dation, they put in all tliat their small cylinders will hold, and never dream of its being necessary to figure how many tons each pile will have to sustain. Cylinder piers are legitimate construction in places where, under the worst possible conditions in respect to scour, they will have a firm grip in solid material, say not less in depth than twenty per cent of the height of the entire pier. Cylinder piers will iK.t often stand tiie test of the curve of p jsures herein (h-scribt'd for masonry piers ; but this is not 312 DE PONTinUS. accessary, because they can resist tension on one side in both the metal ami the couciele, if the latter be of the correct qual- ity ; i.e., the cylinders can act as beams to resist the horizontal thrust of wind and trains in the same way as do the columns of elevateel rod with an adjustment. A number of these chains were used for each pier, the rods passing through heavy timbers ut the t a I fo wi so ini de W( DESIGNING OF PIEliS. 315 rear of the pier near the top. By screwing up on these ad- justments the tops of the two piers were moved baclc a little. Provision is nmde for future scour by leaving some spare chain beyond the point at whicii the rod takes hold, so thai one chain at a time Ciiu be loosened, lowered, and retightencd. These East Omaha bridge-piers will probably la.st a long time yet, although when they were put in no one anticipated that they would be needed for more than eiifht years. In sinking piers the greatest care should always be taken to start them in e.\acl position and to keep them there. The in- stant it becomes evident that a pier or cylinder is get' i\s out of correct position, it should be moved back, even if it i^e neces- sary to slop the sinking entirely until the true position be re- covered. Generally it is feasible to build a frame of piles and heavy timbers around each pier or cylinder, so as to guide it to exact position at all times, barritig a slight springing of the piles, which, however, can generally be guarded against. Some sixteen years ago the author had occasion to sink four eight-foot cylinders in the Des Moines River by open dredging to bed- rock, so as to form a single pier, tlie axes of the cylin- ders being located on the corners of a twenly-four-foot square, irnfortunatcly the author in making the design, owing to in- experience, had provided no allowance for variation of location. The foreman of construction informed him that it was abso- lutely impossible to sink those four cylinders so correctly that the struts would lit between their tops ; consequently the author was compelled to undertake the superintendence of the work him.self. He built a strong frame of piles and heavy timbers, all thoroughly braced, around the space to be occu- pied by each cylinder, cutting out the horizontal timbers to tit the curve of an eight-foot circle, and even gouging out places for the rivet-heads to pass. One of these horizontal guides was located close to the surface of the water, and the other some nine or ten feet higher. The cylinders were dropped into these guides and sunk to bed-rock. After all four cylin- ders were in place, and partly iillt'd with concrete, the struts were inserted between their tops, and were found to furnisli a driving fit. This resull was, perhaps, due as much to good 316 DE fOKTiBirS. luck as to good management ; but the experience taught tho author a lesson which he has never forgotten, and which he desires to impress upon all young designers, v./ that in pre- paring any substructure design it is essentuil to provide liberally for all possible variations from correct position m all parts of the work. ,, In respect to the designing of pedestals for elevated la 1- louds and the delerminati(,n of the bearing capacities of soils, the reader is referred to the author's before-mentioned paper on Elevated Railroads published iu the 1897 Transactions of the American Society of Civil Engineers. CHAPTER XXIII. TRIANOULATION. The necessity for extreme accuracy in the tHangiilntion for piers of long briiiges is not generally recognized ; hence result errors in pier location that sometimes riquire the lengthening or shortening of the superstructure, or which involve tlie adoption of an unanticipated skew. There is no excuse wlial- 8oever for any such errors in location, because the method of triangulution adopted should provide a check against not only blimders, but also even trifling variations from correctness of posiiion. antl because the Contractor should inviui!il)ly. at the outset of his work, talie sncb precautions as will prevent the occurrence of any variation in sinking in excess of that pro- vided for in the Engineer's plans. In the triangulations for bridges over large rivers, such as the Missouri, the author makes a practice of measuring each base-line five times and each angle thirty times ; and no point Is ever located without using a check from another base-line, thus providing an intersection of three lines, which theoreti- cally should be a mathematical point, but which actually varies therefrom, generally about a quarter of an inch, and sometimes even as much as one half of an inch, in sights of about one thousand feet length. The author has tried lioth iron rods and steel tapes for measuring base-lines, and has adopted the latter as the more accurate. The objection to using rods is that it is almost impossible to run a line a thousand feet long with three rods that must always be made to actually touch each other with- out sometimes disturbing slightly the position of two of the rods, when either liftinj^' or putting down the third rod. With 317 318 DB PONTIBUS. iv reliable steel tape properly handled, the extreme error in a iniinbcr of measurements of the siinio line should he less than one quarter of an inch in one thousand feet. This would make the probable error of the avenige length considerably less 'han that amount. If any measurement sliow a greater variation from the average than one quarter of an inch to the thousand feet, it siiould l)e rejected, and another measurement should be maih; to replace it. This presupposes comparatively level ground for tiie base-line ; hence, if iho ground bo very irregular, a greater variation may be allowed. It should, however, in no case exceed one-half inch per thousand feet. Tiie tape line used should he a new one for each structure, and it should be tested at the bridge shops in comparison with their standard. As a matter of precaution, it is well to test it in the field with another tape that is to be set aside Jis a reserve and not used unless an accident happen to the primary tape. For very long and important bridges, especially cantilevers with ^ong spans, it would be well to have the tape tested by the Bureau of Weights and Measures at Washington, D. C, or by some other testing bureau of recognized standing — such, for instance, as that of the Washington University at St. Louis, Mo, The charge for such testing is usually merely nominal. As the coefficient of expansion is not the same for all tapes, it might be advisable for extremely accurate work to have the coefficient determined for the tape to be used; but in most cases of long-span bridges this would be an unnecessary re- finement. A fifty-foot tape is long enough, and is in many respects preferable to those of greater length. The author has no use for extremely long tapes to measure distances directly between pier centres either during sinking or after the piers are finished, be- cause this method of measurement is by no means as accurate as that of intersecting three lines on each pier and using two independently measured base-lines. The only direct measure- ment that is of any real value, and which can be obtained before the falsework is up, is one made on the ice. In such a measurement care must be taken not to let the tape touch the ice, but to rest it on plugs driven on perfect line into holes TUlAN(a l.ATiON. 319 therein mid cut off to exiict level. Tlicre is no inoie diJlicuU nioasureiiient to miike correctly than oue with m long steel tape between two (listunt jMjints without intermediate supports; he- cause, in tho tirst place, the doultlo nieusurcnuiiit on shore and in correct position is a slow and tedious one to niuke, involv- ib^ us it does the use of the level to obtain the sig, which must be exactly alike in l)oth cases, and, in the second place, the conditions of wind and temperature are likely to ynry to such an extent as to cause errors that are very difficult to correct. All base-line measurements should be made in cloudy weather, or just after sunset, or even at uight; and the tem- perature should be noted for each lifty fiel measured, as all lengths must be reduced to those for an assumed standard shop temperature of seventy degrees Fahrenheit. Even slight variations of temperature will cause errors of importance in the length of an ordinary base-line, the change in length per degree of temperature and per unit of length being ahout 0.0000066. For a base-line of one thousand feet and a varia- tion of one degree tiie change in length would be eight oue- bundredths of an incii. This, it is true, is no great amount, but there is always a liability of tliere being a difference of as much as ten degrees between the average temperatures for measurements made on two different days, and as much as two or three degrees In a single measurement of a base-line. In using a steel tape it is better to start from the one-foot mark rather lli from the end, imless the ring be placed back of the zero-point. The author's method of measuring a base-line on compara- tively level ground is to run in a line of stakes of at least three inches by one inch section and from two feet upward in length, spaced at intervals of alxjut ten feet and put in to exact line and level, with a large flat headed tack driven to line on each stake, and the- true base-line scratched with a knife along tlie toj) of each tack. The line is measured by stretching the tape with a uniform pull of six pounds over the line of stakes, keeping the one- foot mark or the zero- mark, as the case may be, over the centre that is cut on the hub at the 330 DE PONTIBUS. end of till! biuse-lliic, iinil scratching,' with ii knife on the tack where the fifty- foot mark on the; tape comes, IIum* starting from this point to measure anollier forty-nine or fifty feet, ami so on until the centre of tiic hulj at the far end of the Ixise-llne is reached. The next time that the line is measured tlje lirst length should be thirty-nine or forty feet, so »» to avoid using the same tacks; and each succeeding first length should be ten feet shorter. This not only involves the use of fresh tacks for each measurement, but also prevents any manipulation of the tape so as to make the partial measure- ments agree with ' e made previously. In case that a Uly level line cannot be obtaineo put in and use in- termediate hubs on the base-lines Base-lines, whenever it is practicable, should be run approx- imately at right angles to the longitudinal axis of the bridge; but this is by no means essential, and it is folly to try to iMake the intersection exactly at right angles, except in the following case, which represents an ideal system of triaugu- lation that can rarely be utilized, on account of the existing conditions of shores, and obstructions both natural and arti- ficial. The said ideal system consists in running four base lines, as shown in Pig. 8, all exactly al right angles to the centre line of the bridge, and laying off thereon distances equal to those {ipux the buse-Une to pier 'jenlres, so that all lines pf DE PONTIBUS. sight will intersect Ihc ceutre line at angles of exactly forty- five degrees. The a'lvantagc of this sj'stem lies in the fact that all the piers are located by direct sight without having to measure the angle, the only angles requiring measurement being the fotir right angles between the base-lines and the centre line of bridge, and the four other iingles required for determining and checking the distance between base-lines along the bridge tangent. The lengths of base-lines for ordinary systems of tiiaugu- lalion will generally be regulated by local conditions. They 4 3 \ ^y«^ ' 1 jo 90 \ \ \ \ *(1 N X Pier / _X^ \ y y ^v Pidr / / \ / 2 \^ ' / \ \ \ \ Fig. 8. slioidd usually be about as long as the total length of bridge, or, when there is a base-line on each side of the river, us long as the perpendicular distance between opposite base- linos; but, if necessary, they may bo made as short as seven tenths of same. Too short base-lines will give too sharp in- tersections, and therefore sometimes loo great variations from correctness ; nevertheless, sharp inlersections can be employed at times by taking e.xiru puins with the work and by employing an extra intersection us u cheek, in case that uuy discrepancy occurs. )i\ TRIANOULATION. 323 Aflor llie base-lines are measured and the hubs are put in, tlie next step to talce is to measure the six principal angles of thu triangiilation. Tliese should be measured with I he great- est accuracy continuously around the limb of the transit, making fioin ten to thirty readings of each angle, according to the degree of refinement required. The instrument should be graduated for accurate work as fine as twenty seconds, or preferably ten seconds. A heavy transit with a good, solid tripod will usually give better results than those obtained by using a ligliter instrument. Tlie sun should never be per- nutled to shine on !he instrument when the angles are being observed, as it is impossible to make accurate measurements under such a condition. In keeping notes of triangulation-work a record should be made of the date, the temperature, the condition of the weather, the direction and approximate velocity of the wind, and the names of the trausitmaa and picketman. If long sights are to be taken, the picketman .should be pro- vided with a pair of field-glasses to enable him to see the transitman's signals ; otherwise much time and labor may 1)0 spent to no purpose. Long sigiits .diould never be taken lowan^s the sua when it can be avoided. Tiie error of all three angles in each cf the two main tri- angles should not exceed two seconds in important work. Of course it is not necessary t;> go to any such refinement in .shoi t-span bridges ; but in very long ones the error miglit wcll In reduced as low as one second. If the error in a triangle be I'oiuid too large, it maybe possible to avoid measuring all three angles again by looking over the notes and ascertaining from tlie weather conditions which angle is most likely to be at fault, then mea.suring this angle anew. If the second average angle reduces the total error in the triangle to within a proper limit, all right ; but if not, the other t>vo angles will also have to be measured a second time. On the same principle, if, in a group of measurements of one angle, one or two readings be fouud to differ greatly from the others, they nuiy be throwu out when obtaining the 324 DE PONTIBUS. It sometimes happens tliat both intersections of the bridge tungent with the base-lines cauuot be seen from one end of one of the hitter. In this case it will be necessary to put in a hub on the bridge tangent far enough ahead of the hidden point to clear the obstruction, triangulate to it, and measure the exact distance from it to the hub on the base-line. This expedient was necessary in the triangulation for the author's Jefferson City highway bridge. A check on the accuracy of the triangulation work is ob- tained by comparing the two computed lengths of the bridge tangent between the intersections thereof with the base Hues, or between one such intersection aud a fixed point on tlie tangent on the other side of the river. The disagreement in these two measurements should be withiu the lin.il of one half of au inch to the one thousand feet. To show how accurately such work can be done, the author would state that for the Jefferson City bridge he gave his resident engineer instruc- tions to allow no variation from correctness exceeding three eighths of au inch iu either the mdin triangulation itself or in the intersections for pier centres. His instructions were fol- lowed so faithfully that no error exceetling three sixteenths of an inch was allowed to pass in any part of the work. The •whole field-force once lost au entire haif-day iu rectifying an error of one half of an inch iu the intersections for a i)ier centre. This is an excellent record ft)r accuracy, considering that the dista. ce between base lines on the bridge tangent was a little over fi..een hundred feet. The author is generally not so rigid iu his retiuirements for exactness as he was in this case, the reason for such strict instructions being the fact that this was the resident engineer's first experience in important triangulation. The triangulation for the author's Sioux City bridge, made by LeeTreadwell, Mem. Am. Soc. C. E., with a bridge tangent about twenty-two hundred feet long between base-lines, was probably just as accurate as that for the Jefferson City bridge, because the errors in distances between pier-centres measured pn top of the falsework n- re actually inappreciable. TRIANGULATION, 326 After the main triungulation for a bridge is finished, the next step is to compute tiie angles to the various points on the piers that will be needed during llie sinking. For a single cylinder pier it will suffice to triangulate to the centre only, and for a pier composed of two cylinders a triangulation to the centre of each cylinder will be enough ; but for a rectan- gular pier it will be necessary to locate not only the centre, but also anotlier point near the periphery, in order to prevent tlie pier from being rotated ab(;ut its vertical axis in going down. After the calculations are completed a triangulation- sheet should be prepared, on which should be shown all of the triangulation with the various distances on all lines and the exact angles for all deflections. Foresiglits should next be located for the bridge tangeni and for all pier points, so that the Iraiisitman fliall never be under the necessity of turning oil an angle when locating a pier. The position for any foresight is generally determined by convenience, but it should be chosen so as to avoid any IMobabilify of disturbance. I-'ach foresight, which consists 'f a substantial wooden targt L, 'is located by turning off the propei angle from the l)ase-line, and is then fixed immovably in position, iifier whic i. a series of from ten to tiiirty readings of the angle is made, thi (jornspouding et'iitie lines being marked on the target. Tlie average of all of these centre lines is then detenu ined, and is assinued to be the true centre, whicli is marked ( onspicuously on the target. Each target is to be marked also with its characteristic letter or n\imber, so that its individuality ni?iy be recognized by the transilman from the most dist I' ■ oint of observation. The angles for determining the conoei centre of any target sliould be laid off continuously on the limb of the transit. All foresights should be inspected occasionally so as to see that they have not beer disturbed, although any disturbance will be discovered, the firat time that the foresight is used, by the three lines failing to intersect in a point. When piers are to be built in open coffer-dams, 'he work of locating them is comparatively simple , for when 'hey are 32G 1)E PONTIBUS. once located little or no movement takes place afterwards. But when piers are to be sunk by the pneumatic process or by open dredging, great care must be taken at every step, because the pier is always eitiier moving or liable to move at any mo- ment. In sinking piers by either of the two last-mentioned processes, the resident engineer should keep such notes that from them he can report daily as to the exact horizontal posi- tion of the cutting edge of the caisson, the position of the top of the pier, the elevation of the cutting edge, the iucliuatiou of the axis of the pier to the vertical, and the amount, if any, that tlie pier has been revolved around its vertical axis. Tlie Contractor can contluct his operations with much more cer- tainty of landing the pier in its true position, if he be kept informed as to its relative position every day. If temporary staging be used around the jner, from wliich to conduct the operations of construction, keeping track of the various motions of the pier will be a comparatively easy task, for the approximate alignment can be obtained from tempo- rary points located on the staging, which points, however, need occasional checking to see that the staging has not shifted sliglitly. If there be no st'iging, all locations will have to be made by triiingulation, and, as before staled, two points on each pier will be needed in order to detect rotation. When the caisson has reached a considerable depth, however, the liability to rotate is greatly lessened. After all that may be said, the work of keeping the pier in correct position will be dependent on local conditions and many varying recpiirements. In respect to the levels, care should always be taken to pre- serve su('h measurements as will enal>le the leveller to keep a record of the vertical distance fvom the cutting edge to the top of the crib at each of the four corners. This will be necessary in order to determine how much the said cutting edge is out of level. In giving the final elevations for the copings of the piers, it will sometimes be found necessary to lake very lonij fore- TRlANGlTLATION. 33 Oi^ sights, owiug to the impracticability of setting up the level near the piers. In such cases a backsight should be taken to a bench-mark about the same distance from the iustvumeut as the pier is therefrom, and in the opposite direction, so as to off- set a possil)le slight lack of adjustment in the level, and to compensate for the curvature of the earth. CHAPTER XXIV. OFFICE -PRACTICE. As there has been almost nothing yet written concerniii!? the way in which work is handled in a Consulting Engineer s of- fice, the author has concluded to close this little treatise with a chapter on " Oftice Practice " ; and as no two engineers pur- sue exactly the same methods, and as the author is naturally more familiar with his own than with those of others, he will deal herein solely with the established practice of his own of- fice, which practice is the outcome of over ten years of special effort to secure the best possible results both expeditiously and economically. LAYING OUT WORK. This chapter being confined entirely to oftlce-work, it will be assumed at the outset that all such field data as profiles, maps, plats of borings, etc., have been secured. In bridge-work it is necessary to determine the following First. The Purpose for which the Structure is to be used.— This being settled, there ensues the fixing of the live load, the clearance between trusses, and the clear heiglit above base of rail or surface of roadway. Second. The Clear Height between Standard High Water and the Lowest Part of Structure. - If the stream be a navigable one, the minimum clearance will be regulated by the requirements of the War Department. In other cases the clear h(!ight will depend on the required elevation of grade of railroad or road- way, provided that the lowest part of the superstructure will never offer any ol)struclion to tioatiiig drift or ice during the highest floods. The minimum clearance should preferably be ten feet, and never less than five. 328 OPPICE-PRACTICE. 329 Where a low bridge is required over a navigable stream, some oue of the various kinds of laov.ible bridges described iu Chapters IX aud X must be used ; but for all ordinary cases the rotating draw is the most suitable type. Tliird. Best Span Lengths to adopt. — In many cases there will be no choice as to span lengths, whicli are liable to be deter- mined by such conditions as the reqidremcnts of the War De- partment, obstraction of stream by piers, danger fronj wasli- out during erection, etc.; but, wiicre the designer has any choice in the mailer, he siiould be governed by the principles of economy laid down in Chapter Ilf, taking care, however, lliiit he does not violate any of the principles of esthetics given in Chapter IV, unless he be forced to do so by circumstances that are absolutely beyond liis control. As stated in Chapter III, the greatest possible economy will exist when the cost of each pier is equal to one half of the cost of tlie trusses and lateral systems of the two spans which it hel[)3 to support. The determination of these economic conditious is, of course, a matter of cut and try; but after a few trials tlie economic span length can be approximated very closely. In making such calculations the trial weights of trusses and laterals can be found with sufficient accuracy by taking a span of known weight and computing therefrom tlie weights for the spans of the trial lengths by the following methods : A. The weight per foot of the lateral system is directly pro- portional to the span length, provided that the superstructure is not changed in width, which is generally the case. Should the width be changed, the new weight will have to be modi- fied accordingly, under the assumption that the weight varies about half as rapidly as does the width. B. To find the truss weight W i>er lineal foot of span of length I' from the corresponding known weight W of span I, the following approximate but quite accurate empirical for- mula may be used : This will give approximately the weight per foot of trusses 330 l)K PONTIIJUS. for any spun length, provided the live load per lineal foot re- main uncluinged, C. Tofliid for liny span length the truss weight 7" per lineal fool for a total load p' per lineal foot from the corresponding known weight 7' for a load /?, the following approximate em- pirical formula may be used . •=.-('+^^)- This is quite accurate for all ordinary spans, but for very long ones it gives too great a variation between 7" and T. After finding the value of 7", the value used forp' should be checked; and if there be any serious disagreement between the value assumed and that found, the substitution in the formula should be made anew, and so on until a satisfactory agreement between the said values of jt' be obtained. Fourth. General Layout of Structure. —The general layout should consist of a profile, a plan, and enough cross-sections to illustrate properly the entire substructure, superstructure, and approaches, all being made to exact scale. For long crossings, a scale of one fortieth of an inch to the foot is the most satisfactory, but for short crossings the scale should be made larger. The proportioning of the skeletons of the trusses should be done in accordance with the suggestions given in several of the preceding chapters, and the dimension 3 of the piers should be determined by the principles established in Chapter XXII. Each general layout should give the following information : Elevations of bed-rock, low water, standard high water, extreme high water, lowest part of structure, grade-lines and tops of piers, lengths of all spans between centres of end-pins or centres of bearings, distances between centres of piers, all leading dimensions of piers, heights of trusses, and lengths and kinds of approaches. As soon as the general layout is completed and finally adopted, the computations of stresses and sizes of members of spans may be begun. OPFICE-l'RACTICK. 331 For elevated railroads it is necessary to delcrmiae the following ; First. The number of tracks on the various portions of the line, and tlie clearances over streets and alleys. Second. The live load per track to be carried by the struc- ture. 2%ird. The location of the line, whether in the streets or on private property. Fourth. The style or styles of girder construction. In some locations the City Ordinances may require open-webbed girders, :is these shut out less light than do solid plate girders, while in otlier locations the plate girders would be per- missible. Fifth. The location of columns, whether in the street or on the curbs, also, for location on private property, the number of columns per bent. Sixth. The economic span length. As indicated in Chapter III, the greatest economy will exist when the cost of the longitudinal girders is equal to the cost of the cross-ginlers, columns, and pedestals. Where the columns are located in the street or on the curbs, due consideration must be given to the probable cost of removing underground obstructions, such as water-pipes, gas-mains, etc. With these points all settled, the calculations for propor- tioning all parts of the structure may be proceeded with. Where the structure is on a curve, it is best to make the bents radial whenever practicable. The exact location of each coluuni should be figured from certain known lines, and all ordinates for same should bo indicated on the layout. Much careful study should be given to the work of establish- ing each feature of the layout ; for, if mistakes be made therein, they are liable to cause great delay and expense later on. Roof-trusses and steel buildings will not be treated in this book, as it deals mainly witJi bridges, viaducts, and elevated railroads. The office work connected with the de- signing of roofs and steel buildings will, however, not differ 332 »E PONTIBUS. csscutially from tbat pertuiuing to the desiguiog of the other structures. CALCULATIONS. After the leading features of any proposed structure have been determined, and after tlje general layout thereof is com- pleted, the next step to itikc is the making of the calculations ncoessiiry to dcternuuc the stresses in all the parts and the proper sizes for same. For convenience in making to correct scale pen-sketches of the various portiims of the design, the author uses a cross- secliou paper divided into one-qiiaitor-inch squares, the slieets l)eii)g ten and a half inches wide by sixteen inches long, which size experience has shown to be the most satisfactory. At tlie head of each page are written the date, title of structure, and name of computer. At the beginning of each set of calculations the following general data for spans are given : First. Lengtli of span. Second. Number of panels. Third. The various truss depths. Fourth. Perpendicular distance between central planes of trusses. Fifth. Live load or loads to be used. Sixth. Wind loads for both upper and lower lateral systems. Seventh. Spacing of stringers. The dead loud from the track and ties in railroad bridges or from the timber floor or pavement in highway bridges is first determined, using the unit weights of materials given iu Chapter XIV; then the stringers or longitudinal t'rders are figured and proportioned, after which their weights and that of their bracing are computed. Next the floor-beams or cross-girders are proportioned, and their weights are figured. From all these weights the weight per lineal foot of the metal in the floor system is next found. As the lateral system can nearly always be designed before the trusses, it is generally best to compute the weight per Jiueal foot of the entire lateral system before the trusses are OPFICK-PRAOTICE. 3;j:j touched, because the dciid load fur llie hitler will be affected by the weight of the former. Next it is necessary to assume the weij^ht of metal per lineal foot for tlie trusses, using, if necessary, the formula' given previously in this chapter. This completes the data for the preliminary dead load, which will consist of llie following items : First. Flooring (timber, traclt, pavement, etc.). Second. Floor system (stringers, stringer-bracing, and Hoor- beams). Third. Lateral system (upper and lower lateral systems, vertical sway-bracing, and portal-bracing). Fourth. Trusses. In making up the dead load, the end tloor-l)cnms and pedes- tals must not be included, as their weight produces no bending moment on the span. The dead-load stres.ses in trusses are always found analyti- cally for spans with parallel chords and equal panel lengtiis; but for all other cases they are determined graphicjilly, and are checked by a single numerical calculation at the member where the graphics stop. Whenever it is practicable, in making arithmetical com- putations, the slide-rule is employed. For ordinary work, in which tlie totid stresses can be written with six figures, a twelve-inch slide-rule will give the stresses accurately in thousands of poumls ; but where the stresses are greater, Thacher's cylindrical slide-rule is employed. The live-loud stresses are found by the method explained in Chapter XIX. The computation of all stresses found analytically is facili- tated by determining the trigonometrical functions involved in the calculations, and multiplying the panel loads by them. By setting these products on the slide-rule and using the proper tabulated coefiicients, it is often practicable to read off a large series of stresses without resetting the slide. The dead-load stresses and the live-load stresses are written on separate diagrams on the calculation-sheets. Xhp impact stresses arp f opnd from the Uve-load stresses l)y 3;u J)K PONT I BUB. slide-rule from the fornmlie given in eitlicr Chapter XIV or Clmpter XVI. us the (luse nmy be, or from llie corresponding tublcH at the end of the book, and are written on a sepanile diagram. Next are computed all tic wind-stresses which could possi- bly allcct I he sizes of llu; sections of inain-tniss members, and these are recorded eitiicr on a separate diagram or on one of those already preiwired, in the latter case care being talien to indicate that eacli such stress is marked as a wind-load stress. Next the various combinations of all stresses are made and recorded on a new diagram, after which the reqinred sections of nil main meml)ers are figured according to the specitica- tioua, and are recorded on the .same diagram; then the actual sections are proportioned and recorded there also, The exact lengths of all members, including camber a"ow. anoes, are next figured and recorded on the lasl-menlioned diagram. Next the weight of metal in the trusses is estimated. For preliminary estimal'is 'he weights of details are percenlaged from recorded results of previous similar estimates ; but if the structure be of an unusual type or size, the details are sketched and tlieir weiglits are computed. Next the total weight of metal in the structure is figured, and tlie dead load is checked. If it does not agree with tliat a.ssumed witlii'j the limit of error set in the si)ecifications, a new dead load is assumed, and the entire computations of total stresses, sections, and truss weights are made anew. It is very seldom, liowever, that it is necessary to make thes(! calculations more than once, owing to the great mass of accumulated data concerning weights of metal in all kinds of bridges. lu making any set of calculations the computer .should check back on his work at short intervals, so as to see that no crior lias been made, because the effects of such errors often extend over all succeeding computations. In determining .stresses graphically, the frame-diagram shoidd be laid out on as large a scale as is convenient, and the load- diagram should be made a^ aroall as practicable ; for the lar^e OFFICK-PRACTICE. 335 (id, I at a of It est; of of frame pjlves great acciinicy in Inclinutions of members, which in llie all-iinportant point in graphicul computations, and the small load-diagrimi conilnes the graphics ton reasonable space. If tlie inclinations are correct, nccurate results will be obtained with a very small loud-diagram. Tlic autlior's limits of error for graphical work arc one quarter of one per cent at mid- span and one per cent at the far end of span. Siiould the error exceeci these limits, the graphical work has to be done anew. Smooth paper, sharp pencils, true triangles, and per- fect straightedges are necessary to secure good results, to which list should be added painstaking accuracy in every manipulation of the appliances. All calculations on the standard sheets are made in black copying-ink ; and when they are checked l)y another computer, as is the invariable oustoivi in the author's oOice, all check- marks and corrections r.re made in red ink, and each page checked is so marked and initialed by the checking ccmpulei', who not only verifies all the numerical calculations, but also follows carefully each step in the design so as to guard against all possible errors. Tlie work of checking is greatly facilitated , if all the steps taken are indicated plainly', so that they can be easily followed by the checker. Each result checked is ticked off with red ink. MAKINO DRAWINGS, Owing to the necessity for iiaving several copies made of each drawing, the latter is lirst laid out in pencil on brown paper, and is copied in ink on tracing-cloth. In some simple designs, however, the pencilling is done directly on the trac- ing-cloth ; but this is the exception rather than the rule. For convenience in handling and filing, it is very desirable to have all drawings made of a uniform size. After several years of experience, a size of twenty-nine inches in width and thirty- eight inches in length has been adopted as best suited for bridge pi ins. This .size may be used for all detail drawings and stress- diagrams, but it is often necessary to increase the length for profiles and general drawings. The drawing is always made on the rough side of the tracing-cloth, as it is 336 DE PONTIBUS. often convenient to do a couaideiabk amount of drawing and writing in pencil on tlie sheet. AnotUer reason for using liie rough side is that any erasure shows iess thereon than it would on the smooth side, and it is often necessary to do considernble erasing on tracings. As before stated, the first drawings to be made are the gen- eral profile and plan with cross-sections, to establish all the main dimensions of the structure. These drawings can be prepared before the computations are finished. Next come the streSvS-diagrams, which should contain the cambered lengliis of all members, the dead load, liv lineal foot of span. Twelfth. Clearance required above base of rail or floor. TJiirteenth. Kinds of materials to be employed in all parts of structure. Fourteejith. Diameters of rivets to be used. The stress- diagram proper may be simply a line-drawing, each main member being repreeentcd by u single right line, or all the main menibers may be drawn to scale by means of their periphery-lines. The latter method ia generally adopted be- cause of the improved appearance of the sheet which il affords. The scale for any stress-diagram should be large enough to OFFICE-PRACTICE. 337 give plenty of room between puuel points to contain all the necessary writing. After the stress-diagrams are completed, the detail drawings arc begun. There is considerable differouce in the methods employed by consulting engineers to convey to manufacturers an understanding of the design whicli they desire to have executed in the shops. Some insist that the only proper method for the engineer to pursue, if he desires his details to be followed, is to make complete working or shop drawings, ready to be turned over to the template makers, while others prefer to make what are termed general detail drawings, whicli show to exact .s(;ale all the details, and give all in>|)<)itant dimensions and the number of rivets in each connection, l)Ut wliich do not locate each rivet by figures, leaving the working druwings to i)e made l)y the manufacturer. When the latter method is adopted the working drawings must be sent in du- plicate to the engineer for his approval before any of the work is sent into the shops, the said drawings being checked by the engineer's assistants, not only to see that they agree in every important particular with the original drawings, l)ut also to make sure that they contain no errors of any kind. The latter metliod is the one which the author invariably employs, and for adopting it he gives the following reasons : First. Each bridge-shop has certain methods of doing work, which demand thai the working drawings be made in accord- ance therewith ; otherwise the cost of the manufacture is materially increased. These methods cannot be considered by the engineer, who has neither the time nor the inclination to go to the (rouble of acquainting himself with the various methods of all the leading bridge-shops of the country. t>econd. The miture of the work of a consulting engineer is not such as to ju.stify him in keeping together enough trained draftsmen to execute with sutlicient rapidity the large amount of drawing necessary, if the first-named method be followed. TJiird. The capacity for accomplishing work in a consulting- engineer's office when the second method is employed is prob- ably three times as great as it would be were the first method adopted, 338 PE PONTIBUS. Fourth. With llie careful aud thorough system of checking shop-drawings in vogue iu the author's office, all the advan- tages to be gained by making complete working drawings are obtained by the much simpler method of making complete detail drawings. Fifth. The manufacturer always appears lobe better pleased and satistied if the making of the sliopdrawings be left to him ; and the work of manufacturing the metal proceeds more smoothly in consequence. In starting a detail drawing, the first thing to be done is to lay out a sheet of standard size. If the subject be a framed structure, such as a bridge or roof trns.s, it will greatly econo- mize space on the drawing if the skeleton frame be laid out on u small scale, say thrce-ciglilhs or one-half incli to the foot, thus giving the proper inclinations of all membei"8, and if the details at all the panel points and connections be made to a larger scale, say three (juarters of an inch or an inch to the foot. The centre ofgnivily lines of all nudn members should coincide with the lines of the skelet8 JO 0.6'jr)l) 59 11 0.6811 60 18 0.6173 61 i;< 0.6134 i;8 14 (1.6098 63 \h (I 6061 64 16 0.6084 r>r< 17 0.59f8 1 60 W 0..')»58 1 67 10 0.5917 1 68 30 0..5888 69 SH 0.5848 70 0..'^i8!4 71 iS 5780 78 «4 0..5747 18 w 0..')714 74 10 0.5688 75 «? 0.5650 76 m 0..')618 77 •» ()..->.5H6 7H m O..Vm6 79 81 0.5585 80 an ()..5495 1 HI 8B 0.5405 88 S4 .'■485 83 86 (l.540."> 84 m 0.r.3r6 85 «7 (1 .')3»H 80 8B ().'.5:!19 87 8» 0.5891 88 40 0.53C3 89 41 0.5286 00 4« 580M 01 48 0.6I8I 08 4i 0.51.5.') 03 45 0..518H 04 46 0.5108 95 47 0.6076 90 48 0.5051 97 49 r^i& . 08 0.. '■)()( )() 0.1'.(75 I!I51 0.4986 4908 0.4K78 0.4K51 0.4S31 0.4 808 0. 1785 0.(768 O.K3;l 1717 ().4t;;),") 0. ir7;i 0. ICi.')! O.KWiO 0.4608 0.4.^87 0.4.')66 0.4516 0.4.'A'5 4.505 4181 0.1161 4444 0.4485 44<>:i O.IW. 0. 1367 0. 1348 0.4389 0. 1310 0.4898 0.4874 4855 0.4-.'87 1819 180.' 0.41S4 0.4167 0.4149 0. 1131 0.4115 0.4098 0.4088 0.406r> 0.4049 0.4033 90 0.4016 100 (l.4(l(X) 105 0.:!!t88 no :i846 115 0.3774 K'O 0.370.4 185 (1.36.36 130 .■i.-.7l i:!5 0.3.509 110 :;M8 145 (1 3389 1.50 ;i.i.33 1.55 (' 3879 Mil) (>..3.'-.'6 l6-> ii.-:i';5 170 (I..il85 17.-. 0.. 3(177 180 (1.30,30 185 (1 •.'985 190 0.8911 195 0.8899 800 8H.-)7 810 0.8778 3-.>0 0.8703 830 0.86.38 840 0.8.564 850 8.50<» 860 0.8 139 870 0.-.'.l8I 880 8.i86 -.90 (i.8-.';3 3(10 ■:-:-:'i 385 0.8105 :i-.o 8(K)0 375 0.1',»(I6 100 0.1818 4.50 0.1667 .500 0.15,39 5,50 0.1489 600 o.vm 650 18.50 700 (t.ll76 7,50 0.1111 800 0.10,53 8.50 0.1(100 900 (I.(l',i58 9.50 0.11909 1000 0.0870 fSftu^ TAr.i.i; III. 353 H o 5'. I V I *^ *" ^'^ — ^ — ^P ^ .^ W W " » •"• ^^ "-• •—• — • — »— — • »— « T— t "^ v; s^ ^ 2 i; i-Z ';£ '£ "^ ^ m '' *' " '^ ~- X'U^~'£'r^ — 3 2; I- <- 1- {- 1 - < - 1- qE « « Xj i cc « QC X 5: 3: cv c: c'. .3! c» S 93 MM ■< M ' ' ^ II U< ►s •/) u MM H OJ z; Ul H i'. O ^ Ct X ' - ^ -r IC T W» 7» ~^ — ■^ "^ T? cv ?o ri? cc ^ cc '.o ^ 7? 1"? S I- * ^ 4 - -.T iC — -» '^^ Tl W ?l 7> 7' 7' 71 7> 71 7» 7> -7) -r »S iO i.*^ iO ^. i.7 o *r: »rj »o 5 so * '-r ^5 :c -^ ic o ^r- 1^ I- i - $ 7. * i 5 i 7> f -^ X I fi ^ -J r i *^ =r -.c 7 5 5} ^ •& Q Ol * t- -^ --^ art -T 0? 7 . 7> --. c 5 r. y i - -i 1*5 --f -^ irt 5 ii ^O »f7 i(7 ti7 it7 »^ *C if7 LT t.'7 »i7 to tC T -■ ^ -• -^ .^r -* 'I' -* -• "i* 5CJ-«) I ~.l V. -• T» w-r 1--: V • - X - o — "- — — •-•-'- — — ?)«*;»■?»?» 354 DE PONTIBUS. OOOOOOOOQOOOOQOOOOOOOOO ^^ ^T I* O CO !i^ WJ '.' «. u- ^-^ --r \— ',_/ '.'.- ^fcJ WJ . r tij Sj ^- T % - ..^ C( — see. Xi-t-«oi.'^in'» — coc:xcci-ot5 ^T-—. — ooooo — oooooococsc. (55 c-. CJOi n a I o CO • ® p W QJ 8 "J c - M K 2 o "^ H M g H I- 1- 1- i- 1- 1- 1- X' 00 00 00 X Oi a. QC * X * o; oi OS oi * * •r is m o o «f; irt •." 1.0 »ra ••■: t— --o !£> to to ta S «c o (c t- 1- 1- i.v X V-. T 1-. c C'. 5 *. ?» is 35 — ' -r 1- o * '5 o -M ic x •-■ ■^ i-i o c ^ X J- V3 "5 m f T w T» fj ^ c c c X i- t- to "C t;? X tc 1.'; •c *^ *0 ti^ »S »rt I*; »o *c o *o »c 4rt "^ — -T -r T T .-.11^ o Ji-xo»s rr 2i" "^ 'f^ ■•^'-5& o>o '-■Mto • " c o ^- X oc ( - '^ '3 t': -^ w « {-(-i-t-^-i-f-l-t* *-*-(-*- I- -.£:-Oc£r':DO 312333 T>«rr»c;Dt-oco:o>-i^>w-"MMi' I-.r-I-l-..■-lr-.^-.r-.r-•.-.?l^J^^(^^^• TAHLK V. 2 y. o o < l-H a 03 cc Q p- I a w c * H g » ^ <6 < r^ " H E « H HI o tn i-xm^;:;2;OT3'i^tgi-ac^o>-Hir>TO»rir!-oi-cros

    p C5 X <- o S 1.'^ -T ;? 'M ?5 — o S X QC (^ tj ij rp ;:^ 2? " ?>'T-^Xi 7>-fi3 C-. o, f- 3 » o -r oj 7) C> •- C S X X ;- "^ ift -r -p TT T> ^ ^r ro -' -<■ ^ 5C X tc cc « TC cr cc T' c» .1 ^.t ?> ?> ?> ■?> *» ti t^ — I i» ^!S^$$S«^SS^S8^$S9;i:3l!;^?!9$i$$ *< f lO *n lO *ft ift »o *c »^ »o *c.- i't •■ . ^ -^ "■ -t -^ •r -* •* — -T ^ * -.'NTO-*«C'C-xo»o — •^^»7f•lr:^c^■.(JP9»o»-JJJ^2 ;J5G Die I'ONTIlUa, -< en •J e-1 H w Op002-:;OOQ:5CiQC;OOOQOOOOQOOO .. -i . — Q&csi* 1-35013^ Oi QD fij » - « - ^ ir^ o -r -T T^ ■?> -: ? O) C2 03 OV C> 0> Cd O C) C7S Oi C> C« 7. OOOCOOOOO'- — — — — — '-— 1- — CI T»*IT>r»^ —.— ..-^ — .-100000000000000000 Ti ■?» w -T in » I- ao o: o — T» or -T 4^: re I- X m o — T» w "T m m le: in ta ta in ta i.^ lO !C to t> ^ — O ^. cr. 7D ■»; I - -2 «o ic iTj 1/7 »o i^ i.'i i/; »0 . ". »C if^ »f^ I* »(t u^ i.'^ t-"^ »0 T -r -r -f -r •r •t *• -^ •wl k »-'?I?!'»'W»9l-X3»0 — •Mt?»»tO«CI-'nO»0 — ->«''>i»i TABLE VtU 35 Oi [ ^ ?5 Table VII. CENTRIFUGAL FORCE IN PERCENTAGES OF LIVE LOAD. V X 100 C.F.= 32. a X ft" V = Velocity in feet per second. R = Radius in feet. De- Velocity in Miles per Hour. Kfee. 10 10 20 !?o 30 :t.) 40 60 60 1 0.12 0.23 0.20 0.53 0.46 0.'.3 0.73 1.40 1.05 2.10 1.13 2.80 1.87 3 73 2.91 5.8-.' 4.20 o i<.U) 3 0.35 0.47 0..58 0.7:1 1.05 1.31 1.40 1.86 2.3 J 2,19 2.92 3.05 3.15 4.20 5.25 4.28 571 5.00 1 S .4 11.65 14. 57 12 .59 4 7.40 9.33 10.78 .>> 7.14 20.99 6 0.70 0.8-,' i.,5; 1 HI 2.79 3 20 4.. 37 5.10 0.30 8. 50 9 99 11.19 13.05 17.48 20 39 25.17 r 7 31 ','9.30 8 m 2.10 3. 7! 5.H2 8.39 11.42 14.91 23.30 »i.Rr> 9 1.10 2.36 2.02 4.19 4.00 0..55 9.43 10.48 12.84 14.'J6 16.77 18.03 20.80 29.11 37.74 10 7.2^ 41.92 11 1 .28 2. HO 5.12 8.00 11.52 15.68 20.49 32.01 IJ 1.41) 3.11 5 58 8.73 12.59 17.11 22.35 31.89 13 1..51 3.40 0.0') 9 45 13.61 18 53 24.20 11 1.63 1.74 3.00 3.92 6.M 10.17 10 90 14.65 15.70 19.94 2l.;i0 :.'6.05 27.90 1.5 0.9; 16 1.86 4.18 7.43 11.02 16.73 22.77 17 1.98 4.44 7.90 12.31 17.77 84.19 18 2.09 4 70 8.30 13.00 IS. 81 25.60 19 a. 21 4.96 8.8-.' 13, lO 19.85 27.00 ao 2.32 5.22 9.28 14.51 20.88 28.42 21 2.44 5.48 9.74 15.22 21 91 29 82 aa 2.55 5.74 10.20 15.94 82.95 81.23 23 2.06 5.99 10.05 10 65 23.97 32.63 24 2.78 6.25 11.11 17.37 25 00 34.02 2.5 a.H9 0.51 11.57 18 08 20.02 :i5.42 ac . 3.01 ;M2 6 70 12.02 12 17 18 79 nt..5<» 27.05 28 07 .30.81 38 20 27 7.02 28 •■i'£i 7.2: 12.93 20.20 29.09 39.59 29 8 34 7.N3 1 3.-37 20.91 30.11 40 97 80 3.40 7.78 13.8.1 21.0J 3 I.I-.' 42 3.5 31 .s y, 8.(« 1t.2S 22.32 12 13 13 73 82 S «8 8.29 14.73 23 ai .33.15 45 11 83 3.M> 8 54 15.18 23.^2 31,10 40.49 34 3.91 s.;9 15.62 21.42 35 111 17 8,') 35 4.02 9.04 10.07 25.12 30.15 49 20 36 4.13 (1,2^ 10.51 25.81 37.10 .5 .M 10.95 20 .50 3><.15 51.9-.' 38 4 » S.79 IT. 39 •-'7.19 39.14 .53 27 39 4 40 10. (M 17 84 27 S8 40 U 51.02 40 4.57 10.28 18.27 28.. 57 41 12 55.97 4! 4.68 10 53 18.71 29.24 42.10 57. .30 48 4 70 10 77 19 15 29.93 43.09 58 (iO 43 4.!K) 11.02 19.,5S 30.01 41.08 59 99 44 5. 01 11.20 20.01 31.29 45 04 01.29 46 5.11 11.. 50 20. 14 31.95 40 (X) 02 01 46 5.2-.' 11.71 20.87 .32.03 40 97 47 5.:« 11 90 21.30 :i3..30 47 95 48 5.41 12.23 21.73 33.98 48 92 49 5.54 12 10 22.15 34,03 49.85 SO 5.05 12.';i 22.58 .35.30 .'W.82 NoTK, — The stepped line shows the Uniitiiig percentajjes for a super* elevation of 4" for outer r:sil. 358 DK rONTlliL'S. w -^ a; 73 < W ' «< Oh t^ -«! H (A pet o CD » ;«; <: w a CO ■t-O- + X33 ^ 1.3 0^ 4> N .3 O Oj 3? JJT S?s ^'§S §S!2 X lO ** T t- iT. ^ tr.rji ic -r « J- ■* as 1 1- irt — » w :c l> — •:)«- A -O l- — CM- C-. t- «■:>(- ot »:> ir 05 JO ?» -- rj ic OS « c» >-• ".TTJ 05 .^ or in OS 3"- OS Ti tM.'; IS OS o: T» ■>» in IS C5 c> ?) If; in X ?» xxxxxxxxxxxxxxxxxxx Mi^ aaha MIJ9 («M atf r KiJ! mIjB i^x fOX aya Hf Mto Mts **r Mtfi MIlT Mto Mv MllA xxxxxxxxxxxxxxxxxxx ■n I • 00 c. c >- iSi- vc cj o — I- ct 0-. o — 7» 00 iSOi.-CCOSOlO'.Ot-OOClOSOl-QCOSO'-t- TAbLE VIII. 350 '« o < o « o n 25 -r-, O- + ~» HI --±_v — y t— ( ^^ H (^ ^ H Oh < > O H O w Q 2« ^t o I -/; ! -C i ~ = tl jT 1 X 1. " C3 c. K « (w _ 1. — «. ^ hi •/. ■^ M •S-.- 1 7* r C 1 1 m • 1 !? ^£1 :r p 2! " a' tr 3 w ••" '-o f '-^ o I"; 15 « / «tt «li «,„,«,. ,„„,„„,„,„ „^ ,„ ^^ _^ _^^ xxxxxxxxxxxxxxxxxxy ^ ^ • ' 2 "T" O — ?» JO -T *r. :Ci I- C -" •Tl ?? T ».- 'S N ^ ^ „ „ ^* — ^ ;^ ■"■ '■■ * *-- ~ — — 7' re o * " 2 n Zi ^ ^" £ ^ ""^ "^ '■" ^ ^ '=' ~ '?' f'' C 43 ~ a ad 3G0 nK PONrliJL'd. O) H ■Ji O (1h >* « li y. »-4 a « o .--©--. V(l^ K + O -S'J» t'-' n h ^« 1 'i5 ^ ^ •-I 72 1 Jl > w 1 !! H ' W -!l .- rA ij ' *- "^r K a, ' /''^^^ ^1 n ( — j^ a- - - (. X K^; ^/ Eb ■ O (fi y 1 E- Si ' K ■'^ : C5 Cx] --Xr-* Pt ' < QQ O w 7.» « •— O > - O.S £-3 ^oiS M i- 5S 1 - i* O -T ^> 1-Cs Ji •5f. ssr^n ^j 1- -»!- 5^ ■♦ T C5 7i ?5 ?i ir; t- 1-^ XO rfeo •«• -T i ^isl 1-- f5 1< ft 1> xxxxxxxxxxxxxxxxxx xxxx»xxx«xxxxx^. XXX ?{!;2 aI n ;: 'ji ■£ l: ^ 2 Ji ?i cl '•:^ =» ^ !? m :-».»5: ; 5 « - ww^wOBei F^jBrtoc "OBB ""^ "^M cwn CUM as < - -/. o — o .-• J" ;_ ^ „ „ ^ ,, r^ ^ ?» •?» IJ W SI « -*» ~*-* •** -4* *• 3d X C5 in ic -o I ■ I • :/v a- 'T. o — '?» 5C '■■^ * ' - V-, ,^ r-> ,-1 ^^ r-i t-f »-• »-i •-» ^^ .T» CJ SJ O W CO — 9 2.5 3-~ Table ix. 3f)l Tabi.k IX. BENDING MOMENTS ON PINS. Moments In In.- MdMientM In In • 1 t Moments ill In.- a H>8. for Fibre c lbs fnrFitnv Il)s. for Fibre e •a Stress of £ Sti-ess (if 8tre.>i8 of E ar.ooo ;i5,100 27.(XX) 35, UX) 27,0(¥) 35,1(X) i lbs. per sq. n. ll)s. per Rq. in. .2 ll)M. per sq. in. lbt<. per S(|. in. c3 2 lt)H. per S({. in. li)s. per sq. In. "Uf ai'.HX) 27570 801300 1119700 ii»6" 4104r>rA)() 4f.-.'00 ''4 KIIOKK) 1313100 12 4.580.500 59.547(K) •4 4H(H) 5:^8(H) •% 1IH;3.)(K)' 1382;i(X) 12^ 4VJ50tK) 614254 5il932(X) 7791200 •'% I'.'ftWO 1642(X) 1627S00 2116100 13^4 6I66(I(X) 8015800 1.39800 181700 ^7H 17(X)70() 22109(.0 mi fi342300 824600(1 15ia00 200,t00 S^/^ 1775700 23aS400 131^ 6.521700 847820(1 i 1C96(X) 220.500 18.')3(XX) 2408900 13»Ji 6704600 8716000 '>Ht 186100 24 mx) 9 1932300 2518100 13.% 13% 6890900 8968200 1^ 203500 264500 9)4 2013!XX) 2618100 7080500 9204700 •v'2-,'000 2H80(X) !44 2097900 2727.300 14 72738(X) 9455900 ■i?! 2415(K) 314000 9% 2184:^(X) 2839600 '•l'/^ 74700(X) 9711000 a62200 310900 9UJ 9W 2272!XX> 29.54800 ml ';6707(X) 9971900 284100 369300 2363(i(X) 3072700 7874(KX) 10236200 307100 3992(K) 9% 2157000 3194100 141^ 8081 1(X) 10.505400 5 331 3(X) 430700 2.V.2.5(X1 3318300 |H% 82920* X) 10779(XX) •"iJ^l 85«S(H) 463fKX) 10 2(i.')0!KX1 ;W4(i2(H) jl4?i ml 85(XM0O II058;«X) ^ 3S3r.O0 498700 loki 275 1«X) 3.576700 87245(X) 11841900 411000 535100 1014 2854400 3710700 15 89462(X) 1I(»0I00 •'>^ 441000 573:«X) IO?8 317!)."i00 4l3-i400 16^ 11907300 1.5479.500 587.^)00 698700 loa. 3292900 4280S00 17 13022!X)0 169298(X) U 572500 744300 10% 3409000 4432500 I'l''^ 142063(X) 18468:.'00 «^1 (109200 792000 11 3.'^.28100 4586500 18 154.591(X) 20096800 tiM 647100 841200 ll'-H 3649!XX) 4714900 IS^ 16783500 21818600 «% 68C700 892700 HM 3774300 4906600 19 1 SI 81 300 23635700 0^ 728000 940400 ii«/h ;W01500 5072000 ^'■H I!M;.546(X) 25.55 10(X) c?*i 7707(X) 10019(X) 111^ 4031400 5240800 90 21205800 27567500 6^4 815300 1059800 Note.— 270(K) lbs. is tlie allowalile stress, exclnilinp wind. 35100 lbs. is the allowable stress, including wind. ^-^-v ^J^^ •v^, •> IMAGE EVALUATION TEST TARGET (MT-3) 4 (./ .% V4 f/x 1.0 IS 1 1.1 11.25 2^ 1 2.5 '^ 1^ 12.2 2.0 S Bfi ' Ufi — 6" 1.8 U IIIIII.6 V] vl 7: 7 >^ Photographic Sciences Corporation « ^\ « •'' V sv k ^. ^^ Ci^ 33 WEST MAIN STREET WEBSTER, N.Y. MSBO (716)872-4503 If 4 J /J bG2 1)E I'ONtlDL'S. Table X. BEARING ON PINS. BeaiiDK. Diam. of Pill. 82000 Lbs. per Sq. In. 44000 4(I80U 49.500 .5-.i300 •VjOOO .•irsoo OOfiOO 6.S:^00 6(!000 6>*8()0 Tl.'iOO 74^00 :700t) 7DH00 Hg.'sOO avioo 88000 90H00 9.3.500 06.300 90000 101800 i04r)00 107300 110000 118800 11.5.500 118.M0 181000 18.«0() 129.300 138000 134800 137500 140300 14.3000 146800 14^500 161300 88600 Lbs. per Sq. In. 57800 00800 04400 67000 71.500 75100 78700 88800 85800 80400 93000 m»oo 100.00 103700 lO'.doo 110800 114400 118000 181600 125100 188700 1.38:300 185000 130400 14:iOOO 140600 1.50800 i,5;jroo 1.57300 160000 161500 168000 171U00 17.5800 178a00 188300 186000 189500 103100 106600 Dial. I. of Pin. Bearing;. 28000 Lbs. per Sq, In 154000 156800 159.500 108.300 165000 167800 17a>00 173800 176000 178800 1MI500 184:300 187000 180800 108500 lOoSOO 108000 800800 203500 806300 209000 811800 214.500 817:300 220000 888800 88.5.500 888-300 8:31000 8:33800 8.36.500 ••:30.3OO 248000 841800 847,'>00 8.50:300 25:3000 855S00 8.58500 861300 88600 LbR, per Kq. In. 800800 80:3SOO 807100 810000 814500 818100 88K00 825810 288«(K) 232400 886000 239500 84-3100 816700 850:i00 85:3800 8.57 1(K( 261(K)0 864IX)0 268100 871700 27.5.300 278900 888400 886000 889600 893200 8116700 300300 303000 ;tO7.')00 311000 314000 .318.'00 :381S00 3.'.5:300 388U00 3;38r>oo .336100 ;330600 Note.— 82000 lbs. per sq. in. is the allowable stress excluding wind. 28600 " ' " " including " Table xi. 363 Table XI. INTENSITIES FOR FORKED ENDS AND EXTENSION-PLATES OF COMPRESSION MEMBERS. Formula: P = 10000 - Wr. I 1 P 1 9700 •1 WOO 3 9100 4 H800 5 8500 6 8200 7 7900 8 7600 9 7300 10 7000 P I t P 6700 31 3700 MOO 22 8400 6100 23 3100 6800 24 2800 5500 25 2500 5S0O 26 2200 4900 27 1900 4600 28 1600 4300 29 Mm 4000 30 1000 304 Di: Pont I BUS. i o » 3J •< U » a o z, S "< ES 03 m a u i 3 Is > tit _c 'u n «i» «-i» 8 S 2 US g I s s "9 ;:;;!? ^^ "58 OXi ^« ?*0 TO- OT CC gfi ** 1- o o» t- Ho» !"2 9a j-ta '♦'5 9 •O ecu A I- OOD — .> « .I O — I- c» w Sac IN I- •» 1- S 2 « 8 fe ?i Si: eotxi in 'T t-o onio -i-i- Q — i— — « It*- (-— 2; »l» « s I. c J3 CO Hf "w 'St'' e»s rrB iftw 9o *(*» wcr f« ^rt in-* i?}'* •PH!J JO doiis 09b, ojb, fiotc, cob, cqBc, cnb. •jnpi^ Suuvag « ja u J5I*. ^2- T» CI I- ¥» w -J m ^ i- m Xi «s o i- OjCt, COfe, rCt. C<3l«, OSfc, C0&, •13AIH JO ja)<>iii«|(i •^M lorxi l-|«0 r+- It o •5 <£ ^ L Em 1 ii ») V > 8 L. <- s t- V > ■o S i. s ja OQ «5 OS! s O n H Hi n H CO cc I >-. o CQ U H A 55 O 00 (Q C5 O a, o u a H O n o 2 = OS o X o a E OS ca I s o £^ H a fa- O O H TABLE XIII. 36d M 253 S«e(..t.oo iifN*'^ V» !9$S :f:3* sf:j?s ifSS CO •a c U e c s 3 »;:f2?' sr:«^ sf^psr stJ^sr Sf:^2r ^;|!:^ ijr:«^'» ^:i?:s isr:?:^? xRs xj;;;^ 2r:«is 2^:?;^ ::s:j^ :? S i( 6t 9( of K $ c> sraf^f jj!^-^^ sr:^^ gr^sr sr:i?3r jsaKS « 3f*is :?:«»' 3r:f;? sf^f sr ::sa?» a- ¥ .2 P^a ^owi.Qoo.o-2;5Si£SU22?5?J?{|l55^S 366 DE PONTIBUS. cn en u o s 03 O «) GO > -' ^E H ^ d II « U CO O 5 OJ H< 2© ^5 QQ HU «J MO O* 8° ea Wo U |S Ho 03 Jo *J< 5o wB iJW OOJ $!;« fag ==2 a i!0 c:o nOSp OtJcc w S^S SCO -"in t-o»o ■nSi-So Tt •^Si-S^^n OD S'5!58K5:S8§?S 45»ir3Si-5sS?»*S»^ eJSe<5^if;Si-SoS!oirtr^S t25;S§S!?SgSS8?J§?Sr:82 *™ •-«•-»*-«*-* 1-1 C^ SS5???S'5SSSS88SSr:82g 1^ ^ ,^ ^^ v^ ^^ (71 ^) O Cfi Q ^z; (too O4 —•■-"»->— ^ i-i CI C» T» ■?»« « * o 3 a s o ■5? • e» %< el to M c.S '2!25iS89i8SS;-:S8lf???r:g2gSIS ^| 1, T" 4 — c " •'I O b Is n So is " o I* g1 b, GO 73 Q « o c a: y. o r H , 5c n u> a ft - < u ft '/; <«; 2^ w . • s ►J i'. o B5 O ^ r^ CO 0- ^ft„ o 5 5 ss w •J o P S y. ti cc o <. , 05 3 i', ^«§ 'fi <'. < '^ 'A 'A 3 tf u H c« f- <; Q ►- J K H w2^ K > w t^ W 57 n* ft ca Si? S 8 ?5 (» H <; a; S^ a H W ftt-g !^ « Ch >^ )r w a! O •Q& 93 a :r :« :« ^ ig •O I « c E o :« :« :f! :? :f! :s! :s! <£ m 00 ^ ^ I- o wx> » 7» 4C X 1-1 'J" I- o ^ »- c» ©» *i r.- « TO w T 'T ^ irt ic m © •e o « a Pi ^o-• '- "N o» »0tSiSiSiei^ 368 DE I'ONTIBUS. Table XVI. INTENSITIES OF WORKINd-STRESSES FOR VARIOUS MATEKIALS. TENSION-STRESSES. Eye-bars IHOOO lbs. per sq. in. Shapes 18000 " ' " " FlaHK»H of door-beams and striiif^ers (countini; in % of web) 14000 • Hip vert lcal« (eye-li(irs) IBOOO ' " " (Hhaitet*) and liiiiiffer plates * HIKX) ' Adjustaltle menilwi's, soft steel . . ...... ItKKX) ' •' •• wmiiKht iron i:«00 • Lateral rods 18000 siiapes ItJOOO • CUMPKEMSION-8TRK8HK8. Top-ohords 18000 lbs. - 70- per sq. in. Inclined end posts Intermediate posts and KiibdliiK<>i>als I^Ateral struts (no impact for wind loads) Columns of viaducts (fixed ends) 18000 lbs. - 80^ r 16000 lbs. - 80- r 16000 lbs. 16000 ;bs. J »( II ti II II li *■ II il 60- r 00- " " •• r {I = unsupported length; r - radius of gyration, both in same unit.) Forked ends and exleusion^plates. 10000 ibs. -300^ i« li fti (/ = length in inches from centre of pinhole to first, rivet beyond point wlierc full section of member beting; ( = thickness of plate.) Rollers, allowing for impact, static load (XKkt per lin. in. *' " •' moving load SOOd (d = diameter of rollers in inches.) .i .% It SHKAKINO-STRKSSKS. Webs of plate-girders, medium steel, net section . 10000 lbs. per sq, in. Pinsandrivets 12000 '* " '' " BBNDtNG-STHBSSBS. Extreme fibre of rolled sections of medium steel, impact included 16000 Ibs. per sq. in. Extreme fibre of timber beams, impact included . . 2000 " " " " ♦ Increase net section through eye 50 per cejit ovpr tliat of bo<1y ojf TABLE XVI. 369 Table X\ I— (Continued.) INTENSITIES OF WOKKIN(J-H PRESSES FOR VARIOUS MATERIALS Impact, railway bri(JKeK, / = Iiiipaci, high auy bridKusii, I =: 400 L + 500" 100 L + 150" (L =: Length in feet uf span.) For reversiog-BtreHKfH tlgiire tli« arean ret|iiire4l forbotli tension and uoinpreHMion and add % of the Icsxer area to tlie ^Mfaler. For combined dead, live, and wind load stresses strani 30 per cent hiKher ibaii tor dea < ^ c = ^'^ \Ufl IS i=--!= S ;:->3.2-««» 3 >-J; : : - : H •f-+ ++++ 11 a 4 tn ,r. tn >r. m tn « ■?» — oo •-• 06 fe:, = = =, . t kJ: : : : - : H H-+-f 4--+ ++ ^ »r « Cf r- O -- *» i » . ^ . - a ^ • - - • - • a> 1 1 n -c :c >- in o •■O i W K 91 e» — — o s k); : - : : : H +++++++ V in irt o m o If: "5 g. •»r cc ■?» I- o o « o ■p* • - ; ; ; ; ^^ 1 1 * »— M«»-e-. w g •-J; : : : : : -. H + ++ + ++4 4- S iC •»• M 5J '-'©•-' '» g. b: i - : J : • *-< 1 1 ^^ S u >-;; ; : : i i ' H ++++++++ 1^ in lO u: m irt in 'f: "5 4> in-fMif»^»o — b.. = .-. :: ^^ 1 1 1 ^^i(^^E^-55 V S TABLK XVII. .'171 •2 s I (A ri CO o 1-1 u _] Q is --I a < u p o: u o >', OS C/3 Ed H £ 0} ^ ^ oS^o AJ3 eta a 4-1 a B (A 0. a c ^ -•a -St IS H ■51 a looino , mtet-od »-»•. i ^.*i r^ w* ^ « « « 4+++ i : •: : ^44+ H •«4: 44 4+ •rtOtOO >o 10 in 10 c «o »^^^ ^ « wt-'oo » k^ « « ■• at ^ © 0' s. 1 * ^ 00 1 P3 3j s u "53 B 3 * J i 8 1 100 ++++ 4444 444 44 44 "••l-O*© C0©|.-0 4> loiraio 100 inio a b: : r ^ J * • s. "do "©» d© d) 1 *=', 3 * ' '1 •C « lO «oao«--ifto 09 09 ^ X C> '?# ^ 00 ©» ■?* ^ r-I 1 1 s ©■SHOff* 91 to 2 1 kJ ' -■ ■• ' '^ * a « « ++++ + •n in in 1 44444 to tft m m 10 1 ^ "5 a 9»^©d 4444 9(C0 44 »^d • — 44 ^ ad ^>n 3 •^o *"" KJOiO ©« rs p« w^ ^i i i '. tf « < « « n w H "a! a ++++4 44444 1 4444 444 444 an 9( -« in ■V W TJ i-i •n in >n o 10 ©in ininio a. ■ii^oai oj'VTr r^do SJ J S 3 kM - e 1 S 3 3 i i : 1 1.0 tf3 in © «n©93eoi-co ia> 1 »ii ir^^ «»*Mf-«** 2 K^n^n. tf>>^ i-t »-. t-N »-( .-* ecin«o •^j : : ; -- ^: i : i : : : < * * 444444 -+•44444 44444 444 444 »o ifi in viicioinioio 4} a eoei — orH wmos 0)r<© 1 «©« tei-ao rr e<3 IN -^ ©' d b:..: .« .• d : J J C4 1 I- 1 ^ j« E J3 £ ^^sa& tH) u S OCifclfc.Qj 1 S§^i^<§ S 5 -3 > Cd p U > « ^ II II II m o •< H CO S5 o kH H < > J • 1 y q; B- P 1 to fc 4 '0AiSi*a iMMM^aO C I CO s .<|i.«i l» W«^ «» f t - OS »» :3 C* «T «0 ■// C. W s •-• S<5 O <0 X O 1-1 91 e 91 •S ^ 7* T •'". «C X Oj O " •-•5*M-riOWt-XO>0— *» W " ^ TO ^* iC O ;C 4- at. 05 Os^* W — 1-« n •.-# >: TO « •«''»r r o I -4 O "Tl o z n m z z m > z D -I m z D 'v H O o r > CO 03 ijc- _21,000'< 3--^ gf.OOO* :;■■ ^21,000* r ■ '^ 21 ,00^ - °riN 18.000i* £.yp > CO CO T©- l»,OCK0 48,O0O X PLATE I ^ zT 18^." -I O ---, 31,000>" f p34 .bon* Lit ,000*' j 'L ai.ooc* ,-»■ n. ^ - 7^|S 18.000i< l"_a^.OO^ "liMjLiioon* ^i^Hl3*.oop ««34,00li* T'^ 21.000'^ ^ ~ 2) ,0 00 #ar^^ > CO CO iTiOgofe 20X00* 80,000" -l*- !^ 80.000 " J p — >^ ?tN n.oofl * H O o I O -n O z m m ik.iLogo'^; Ti-^ 19 ,000** f^b^ 19,00(^ fH 10,000^ 4'^c:: 19,00OX o IS.OOQ^ H > 85.006 » 'Z 25.000* < '2a!oooX.g aSjOogx ^ ,t-K IMOO^o gy^b; 18.0 00* "^ sEli5iQOO* o T^PS 18. 0001 m iF ' „ m O 1.5,000''' ? r D > SSjtMOO* z ^Fl8,000>< S 18,000* ^ O j> a CO ■n O 3> 3> -< CO o o m CO -4 I m O O > X r m -D 33 O O :i z ■T, -^ o m o m ^ z J> H z 31 o > 3) H a to ■< z CO —4 0) ^ p fe ^ ^ TOTAL END SHEAR it 3 « jB p j5 § fe £ S a ^ 13 ,fe ^ 3 ,a S 8 .•^ * *^ *^ •^ « •» ••^^ * TOTAL END SHEAR l^ rOTAL END SHEAR IN POUNDb. P L AT LCr. OTAL END SHEAR IN POUNDS. >•> .- EQUIVALENT UNIFORM LOA jn p o. s> p -» j-i I i § s § g z PB o -n ? " ? -t, rt- (T o » ST II g " ^,-» CL ^" *N -■ 3- ?-° • C < o ^^ i. ?■ 5 5- EQUIVALENT UNIFORM LOA LENT UNIFORM LOAD e. Si IN POUNDS. PLATE III. LENT UNIFORM LOAD IN POUNDS. 100 120 14,ti00 5,500 5,400 5,;)00 S.liOO 0) 5,100 O z 5,000 3 O n um z 4,^00 Q < •1.700 O -J 4,000 «c 4,500 o 4,I(X) u. *s 4,300 D H i,:U) Z UJ 1.100 _l < l.lXjO > _) •■,!KX) O UJ 3.S00 3.700 I 3.000- 3,500 3.400 3,300 3,100 3,000 •^60 2«0 300 320 :M0 300 :{«) lOO i^D HO 460 4!«J r)00 SPAN IN FEET. iw:T":rT::":::::::::"::::::"":::"::i:::::-:::" X lJ^^.x -14-, -r-- I ._ — -.-..- _-. _4 ""X 4 LIVE : - - .x.-xx: xxx-X : x x ; :x_.^" : xx xx-i x ±x r u.i»c. wu xL-j: X ' " -^- - .... '_ ... '_, _x . .X '. - ,..4..'...;. "L. _il- ^ ' J T x! . _J-t-^ -| ----- - . - 4,. -L 1 I , TVi , , *. ,., » ,...,.., -,.,.. ...-.....--. a.. .--..-... . V t 1 xj ■^ __ j * "''^1. f^ i^ ■" ' I* ll ^i ! 1 ^s 11^ " -J ; 1 ■ ' I'n ' |- ' I ■ ' _(_ li- 1 — j i •*, J_|- 1 — 1-J J--_LJ-j ._'._! _ 1. . 1 , ^j 1 — . ^ 1 t* s t ] S r" lUU V ^ J') "'.''. " i ^ ^ " " -^ UL -M-^v■t s; -- - ■-- s^- ' - -- - -4- - ' -^ -T - i I ** "N ^ s 1 UJ . 1 1 ^ . I ^ ^ 11 1 ■*< *^ i"'li ;■"■ '■ fj*^ — •■ — ' '" ■ ' ''■■■' ^^. "-■■ --!■.- -.,^-L-fc 4^*1:--- f^x:^!-^./: ■ Tx ( :,:;:xE-i:"^%:f:-xI:$xxx?x:pi-x u) ^s ' ^ St_ 1 1 "^ 1 "" "* £ 1±^l::;?ix:-.xxx:' +. xl^f-'-^.x: v xxxx :-:-| ±±'r£-ixx: + x :: ■ ,. ^ fl%#[flwiii ffl , „ ' ' ^ s Jz "^ *" ^ * X ^^ ^ "r!'x iitlrli'S-i" --- i-i--- -it"""" ""^4. — jj-j-- - S^, 70 "tt: r 1 ' ■ x: — ^~-" ----Xl^-.^ -. -^4-- 1 -- -- --..- .. - .L. - ..^q^Lg!- ^ "»■ ^ **^*il. il ^ S 1 ' ^*^ **ifc W: • 1 '"■■■_■'' I" _■ ^"<*^ t u"J_"" "^ - Sk, ^ ^ ^ fc 4 L ' 1 -... ^LiA o-L. 1 {"'^''^i^r - n^ ' ^z"~ ,::' "":■..:,:■/_.■■:._ ..'-.,: ■:""-:' :j_.'., ■...':: 1 n-">^ -X X-- j-X . I .1 i ..--±..-. i-l -- . ... r ■* T AQ : ... ■"""..." " """ ....-..--..- - .- - - -(-^ ^.., . ^j. . "0 50 J 100 150 200 250 300 350 -1(K SPAN IN FE! PLATE Y. 3()0 350 KX) SPAN IN FEET, 100 1^ 14D 160 LENGTH OF SPAN IN PLATE VI. )AD FOR ELECTRIC CARS FOR HIGHWAY BRIDGES. J.A.L. Waddell, Consulting Engineer, Kansas City, Mo. n 4ari rtrr A m 1,800 z 1,600 < O J 1,400 1,200 1,000 800 600 400^ mt!fs^mii.y li» 140 160 180 LENGTH OF SPAN IN FEET. m 300 WIND LOAD IN POUNDS. 0) Tl > m m H WIND LOAD IN POUNDS, PLATE VII. >000 WIND LOADSFOR HIGHWA'^ J.A.L. Waddell, Consulting Engineer, Kamsas Citv, Mo. SPAN iN FFE7 PLATE VIII. .DSFOR HIGHWAY BRIDGES. J.A.L. Waddell, Consulting Engineer, Kansas Citv, Mo. 1000 SPAN IN FFET. im 30% m 'M • ONE ARM OF DRAW PLATE IX. DIAGRAM OF REACTIONS FOR BALANCED LOADS ON DRAW SPANS. COMPUTED BY J.A.L. Waddell, Consulting Engineeii, Kansas City, Mo. -5 short mid. 40ft \ ^ ■ . . I-: ?TrM:;;-f li Hi :!;m 450 400 u. •m 350o 11 Tr;-, 300 350 ttC! \t<.'.:: 200 : m ^-m ;:;t. 1^ mi': 150 il ii. Hi 100 For weight of tower panel, multiply the percentage given for panel ' anchor arm next to tower by 1.8. For Span3 with Subdivided Panels, a main or double panel is to } used as the basis of Calculation. Metal in A no borages = 5 )< of total weight of entire superstructure. Metal on Main Pier8=|- of weight of metal in Anchorages. The weights of Lateral Sy'> ems are determined in the same manner 3 the truss weights. 50^ 10 10 98765432 NO. OF PANEL FROM END OF CANTILEVER ARM. INDEX. PAGE " A " truss-bridges, advantages of '■ A " truss-bridges, history and description of 3 Abandoned drawings 342 Abuse of man-power nuulunery. , 128 Abutting ends 174 Abutting ends, workmanship or 257 Accessibility to paint-brusli 25 Accidents, responsibility for 203 Accuracy in base-line measurements 317, 318 Accuracy in punching 254, 287 Adherence to principles in designing 4 Adjustable members 149, 171 Adjustable rods for draw-spans 102 Adjustment of draw-spans 128 Adjustment of rollers for draw-spans 19() Adoption of subpunchiiig and reaming 9 Advantages and disHdvaiiliiges of high and low bridges.. 119 Advantages of combined bridges 133 Advantages of lift-bridges 113 Advantages of medium steel 8 Advantages of soft steel A Advantages of the arch < .» /Esthetic reform in bridge-building 45 ^Ksthetic reform in bridge-building, opposition to 45 /Esthetics in cantilever bridges 61, (52 .Esthetics in design 39 .Esthetics in design, fundamental princijile of 47 iEsthctics in East Omaha draw-span 48 (Esthetics in general engineering constructions 42 Esthetics in painting 52, 53 /Esthetics in j)ier designing 53, 54 Alteration of drawings 243, 263, 342 Ambiguity in stresses in ordinary draw-spans 121 Ambiguity in trusses 219 373 zu INDEX. PAnE American Institute of Arcliitoets 284 Analytical nictliud of liiuiin^ stresses 3.'{;; Ancliora;j;(' a Anchor-arms, economic lengtlis of 7P Anchor-arms of cantilever iu'idges, stresses in 5^ Anchor-bars for cantilever bridges 6(i Anchor-bolt connections to columns 181 Anchor-bolts, specifications for 259 Anclioring spans to bearings 24 Anchor-piers of cantilever bridges. w«'iglit of masonry in. (i7 Anchor-spans of cantilever bridges, proper lengths for.. . 75 Anchor-spans of cantilever bridges, weight of metal in, 73,74 Angle measurements 323 Angles connected by one leg only, proportioning of 170 Angles connected by one leg only, tests of 27 Annealing 255 AnneaJing test-specimens 247 Appearance of curvature 20 Appearance of structure to be governed by location. ... 46 Approaches, flooring on 210 Approaches, ornamentation of 53 Appropriate ornamentation for bridges 47 Arches 79 Arches, cantilevered 80 Arches, curvature of 84 Arch depi hs 84 Arches for higlnvay bridges 81 Arches, future investigations concerning 85 Arches, hinges for 84 Arches, spacing of . . . . ." 85 Architects' charge^i against American bridge-designers for want of taste 40 Architectural effect, provision for 17 Arkansas River Bridge 310 Arrangement of members in pairs 21 INDEX. 375 PAOB Asspnihlinp of turntahlos in shops z^Hi AHS(» Best span-lengths for bridges ;}2!l Best span-lengths for trestles 8G, 180 Bevel gears . 2i)\ Binding calculations 345 Binding drawings 344 Blunders in shopwork 282 Bolt-holes in timber trestles 27!) Bolts in timber trestles 27!' Bottom-chord packing 17 Brackets for pinions 193 Brackets in elevated railroads 180 Brick piers 30!i Bridge approaches, ornamentation of 5."> Bridge-designing stiil in process of development 43. 44 Bridge disasters 131 Bridges psigned by nianufacturers 3 Brown paper for drawings 34(5 Muckle-piat«' lloors 143 I'uilt members, workmanship on 25(! Builr stringers for highwav bridges 230 Burnt rivets .' 289 Cables for Halsted Street Lift- Bridge Ill Caissons 292 (^lissons of steel and concrete 309 Caissons of tind)er and concrete 308 Calculations 332 Calculations, binding of 345 Calculations, filing of 344, 345 Calciilations, graphical 3.34 Camber for draw-s])ans 212 Camber for highway bridges 220 Ciunbcr for railroad bridcres 149 Canal Street Fold Cantilever-arms, Bridge, Chicago 107 stresses in 68 m INDEX. sn PAOE Cantilever bridfjea 55 Cantilever bridges, a-stlu'tics in (>1, «i2 Cantilever bridf^es, andiorage against wind-pressure. . . . (Hi Cantilever bridires. anchorage details for (Ui Cantilever biidgcs. auclior-biu-s for 00 Cantilever bridgfs, combinations of stresses in GO Cantilever bridge designing, system i/ation of 57 Cantilever bridges, economic relations of 57 Cantilever bridges, erection stresses in 61 Cantilever bridges, t»xpansion and contraction in 07 Cantilever bridges, linding stresses in (il Cantilever bridges, imi)act for 01 Cantilever bridges, live loads for (iO, 01 Cantilever bridges, pedestals for (Ui Cantilever bridges, stresses in 58 ("antilever bridges, widening of, over main jtiers 02 Cantilevered arclies : 80 Cantilevering simple spans during ere(!tion 70 Cantilevers for roof-trusses 57 Care of machinery for drawbridges 12!) Cast iron 245 Cast iron, specifications for 252 Cast-iron trimmings for decoration 52 Cs'.st steel, specifications for 25)^ Cenient an Cheeking preliminary drawings 24.'J 378 INDEX. PAGE Chocking shipping weights 286 ( lu'cking sh()i)-st sections for 00 Column' for elevated railroa Combinations of stresses in highway briilges 225 Combinati(ms of stresses in railroail bridges 158 Combinations of stresses in railroad trestles 158 Cond)inations of stresses in trestle cohunns 88 Combinations of stress«'s, recording 334 Combined bridges 133 Comparative economy «)f arches and simple truss spans.. 82 Comparative imp(» Compromise Standard System of Live Loads, etc 2()i'i Con<'eptrated loads for highway bridges 222, 22:1 ( oil' iusion 340 Concrete for anchorages of cantilever bridges (Mi Concrete, injury tt) 293 Concrete mixing 2!)3 ( oncrete piers 310 Concrete piers, forms for 292 ('(mcrete, testing materials for 292 Connecting-plates for riveted giiders. 170 Connecting web-angles to chords by o. " mily 96 Connection for shoes at ends of draw-spa - :*I0 Connection of columns to masonry . . 181 Connection of suspended spans 69 Connections of tloor-beams to posts 148 Consideration of iiuality. freijuency, and probability of stresses 22 Consultation concerning architectural features oO Contem|tt of engineers for architectural features in de- signing 44 Contents of drawing-sheets ,338 Contents of stress-diagrams 336 (^tntiiuiity of stringers in railroad bridges 160 C« ntinuous spans 148 Co.itraetion, provision for 23 Convergence of engineering ))ractice and architectural ideals ,47.48 Correction for temj)erature in base-line measurements.. 319 Correctness of impact formula' 7. 8 Cost of traveller alFected by truss depth 32 Cdunterbricing in highway bridg«'s 219 Counters 171 Counterweighted basciile bridges 106 Counterweights of Ilalsted Street Lift-Bridge 108 Couplings for shaftings 20o Cover-plates for plate girders 167 380 INDEX. Crimping of web stiffening angles 94 Cross-section paper for bridge calculations 332 Cross-ties 141 Crowfoot seams 297 Cupped bearings for end rollers of draw-spans 210 Curvature in chords, ai)])earance of 20 Curvature of arches 84 Curvature of top cliord 51 Curved members in English bri Dej)ths of arches H t Deptlm of drums 123 Depths of longitudinal girders for elevated railroads. . . . 102 Depths of trusses for draw-spans IS:? Designing, errors in fi Designing of piers 30 1 Des Moines River liridge piers 315 Detail drawings 243 Detail drawings, method of making 337. 338 Detailing 138 Detailing of members having excessive strength 23 1 INDEX. 381 PAQK 94 332 141 21)7 210 20 84 51 . 20 10 ), 300 ?2, 74 . 254 . 00 . 311 . 200 . 2t)3 . 142 12, 331 . 330 , . 332 .. IH . . ir)2 . . 333 122 184 33.; 122 . 40 . 202 200 . 212 2. 283 . 30 . 14") St . 123 . 102 . ! K3 r> . 301 .. 315 243 37.338 . 138 . . 23 PAOK Details of dpsi^n for hi»Th\vay viaducts 230 Details of dcsififn for open- webbed, riveted girders 100 Details of design for pin connected spans 170, 228 Details of design for j)late-girder drasv-spans 188, 100 Details of design for plate-girder spans lle concentrated load method 200 Double rotating cantilever draws 103 Drainage of ])ivot piers 128. 105 Drawings, tiling of .344 Drawings for trestles 277 Draws, double rotating cantilever 103 Drift-bolting 270 Driving (ield-rivets 24 Driving piles 278 Driving piles into cylinders 313 Drawbridges for various span lengths, styles of 182 Drawers for tiling drawings 344 Drawings 243 Drawings, dimensions of 344 982 INDKX. PAnE Drawin^'-sliet'ts, contoiits of ;W8 iJrawing.s, making of '.VM} Oia\v-a|»ans, adjuHlnient of 128 Draws, jniU-hack 104 Drifting tests IV) J)rillings for choinical analysis 24<), 247 Drums, details of 1!)2 Drums for Ilalstcd Street Lift-Bridge 112 Drum webs 1!);{ Duties of bridge specialist .'{ Dynamite in pier-sinking, use of ;{(>;{ Ease in designing 1;; Kast Omalia Bridge i;55, K{(i Kast Omaha Bridge piers 3(13, .'{J 1 Eiist Omaha draw-span, equalizers for 12r> East Omaha draw-span, rising of ends 12(i, 127 Eeeentric loading from sidewalks 222 Eeonomieal conditions for cantilever bridges r)(i Economic depths for triisses with polygonal top chords.. 32 Economic depths of plate girders 33 i.'conomic depths of trusses 30 Ecv>nomic functions of cantilever bridges 70 Economic layouts for viaducts 87 Ecoi.omic length of anchor-arm for fixed distances be- tween main piers 7") Economic length of main s])an in cantilever bridges for fixed distances between anchorages 74 Economic length of suspended span 70 Econonnc panel lengths 3.3 Economic princi|)lc for any layout of spans 30 Economic rehitions of cantilever bridges 57 Economics for crossings involving danger of washouts. . 38 Economic span lengths 33, 34 Economic span lengths for elevated railroads 02, 0.3 Economic span lengths for trestles 80, 180 Economic span lengths, mathematical demonstration of. 34 Economy, comparative, for arches and simple tru.ss spans 82 Economy in design 30 Economy in roof-trusses r>7 Economy in trestles and elevated railroads 30, 84 Economy in trusses with parallel chords .31 Economy, necessity for IT) Effective depths Hf) EflFeetive diameter '.f rivet lOf) Effective lengt s . . 145 Er Er Er ces I'AOE . . . :t:is ... ;wr) . . . 128 . . 104 . . 24') 24(». 247 . . . . l!>2 . ... 112 . . . . VM :{ ;$():{ . . . . i:! i;i5. 131) 303, :{ii .... I2r> 12(1, 127 .... 222 5(1 32 3:5 30 70 87 7"» hords.. be- lies for . 74 70 33 3(1 r)7 liouts. . 38 33,34 1)2, 03 . . . . 8(1, 180 tion of. 34 e truss 82 30 57 30, 84 .... 31 .... 15 145 .... 105 .. .. 145 INDEX. 883 PAGE KfToc'tive .strength of plate-girder webs in bending 109 Ktteet of impact, provision for 10, 17 Ktl'eet of weight of traveller on erection stresses 00 Kffects of changes of temperature 155 Elastic limits 247 Electric lights for ornamentation 52 Electric motor.", for draw-spans 201 Electric railway loads 223 Electric railways on highway bridges 133 Elementary ])rinciplos in outlining spans 50, 51 Elevated railroads 91 Elevated railroa7 2:i «.K), 149 . UU . 95 . 181 . 177 . 14 6 5 . 222 . 17« . 151 . 15t'. , 2(»«) . 178 . 257 7 .. 14 .. I'M) . . 2:1 .. 95 i.. 97 .. 138 le- .. 96 .. 1«9 .. 24 .. 24 .. 15«) ... 189 }44, 345 ... 345 ... 292 ... 27 ... 261 ... 61 ... 12 ... 12 ... 67 ... 180 ... 167 ... 170 ... 118 ... 299 ... 216 INDEX. 385 PAQB Floor-planks for highway bridges 214 Floors for elevated railroads 92 Floors for wooden trestles. 275 Folding bridges 107 Foresights 325 Forked ends 176, 233 Forms for concrete piers 292 Forms of trusses for highway bridges 218 Forms of trusses for railroad bridges 145, 140 Formulae for truss weights 32!), 330 Formulw for wooden compression members 27(i Formula for impact for railroad structures 7 Foundations, testing of 291 Fracture 249 Framed trestles 273, 275 Framing timber trestles 278 Frequency of stress application 22 Full lengths of sections 174 Full-sized eve-bars, tests of 250 Full-sized members, testing of 288 Full-sized members, tests of 253 Fundamental axiom in architecture 42 Fundamental printiiple of aesthetics in design 47 Future investigations concerning arches- 85 CJas-engines for operating draw-spans 120 Uasliglits for ornamentation 52 (lasoline engines for operating draw-spans 120, 201 (Jear wheels 204 General economic principle for all structures, involving capitalized cost *of deterioration and repairs 38 General instructions to fleld-inspectors 295 General layout of structure 330 (ieneral limits in designing highway bridges 22() (leneral limits in railway-bridge designing 160 General notes on detail drawings 342 General principles for jiroportioning details 28 (Jeneral principles in designing all structures 161 General provisions on methoils of testing 246 (ileneral specifications for liighway bridges and viaducts. 213 General Specifications for Highway Bridges of Iron and Steel .' 130 Girders in trestles, spacing for 87 Girders over drums 199 Graphical calculations 334 Gravity axes, intersection of 20 grillages 303, 309 386 iNDrx. PAQK Guard-timbers 142 Guide-chairs for rail-lifts 209 Guidinj; caissons and cylinders in sinking 21)2 Ilulf-tlirough platc-fjinlcr spans KiS llalsted Street Lift-Iiridgo at Chicago 108 Hand-operating niacliiiicry 200 Hand-operating machinery, necessity for 202 Hand-rails for highway bridges 215, 21(» Hand-rails, ornamental 52 Hand turning-niacliinery, fornmhp for 20;} Harlem River bascule bridge at New York 107 Harshness of outlines of triiss-bridges 42 Heads of eve-bars ITS High steel,* use of 8, 141, 182, 244 Highway-bridge failures 130, 131 Highway bridges 130 Highway-bridge lettings 2 Highway bridges, classification of 213 Hinged ends for columns 00, 179 Hinges for arches 84 Hip verticals for highway bridges 220 Hip verticals for railroad bridges 148 Holes in timber for trestles 271) Horizontal bendinp on cross-girders of elevated rail- roads 98 Horizontal sway-bracing in trestle-towers 89 Houses for machinery of draw-spans 211 Howe's Treatise on Ardies 81, 82 Hubs, Irianguhition 320 Hydraulic buflers for Hui.ted Street Lift- Bridge 109 Ice-breaks for ])iers 308 Identification of metal 240 Ignoring bending in columns of elevated railroads 100 Imaginary superiority of cantilever bridges 55 Impact Impact allowance for highway loads 224 Impact allowance for railroad loads 152, 223 Lupact for cantilever bridges Gl Impact for highway bridges 8, 132 Impact method, importance of 7 Impact stresses, method of finding 333 Importance of chapter on First Principles 12 Importance of imjjact method 7 Importance of rigidity 15, IG Importance of scientific detailing 139 Improvement in highway-bridge building 131 LNDKX. 38; Tniprovonicnt in insj)Pftion Jnipioveim-iitH in desij^n for llalstcd Street Lift-Hridge. Inni|)iuity <>f lacing to nirry tninsversc load Inclined end ponts. bending on Inclined end posts of lii;,'h\vay bridf^es. bending on Inclined end posts, sectit)n8 of Inclining tniss-planes of cantilever bridges Incorrect assumptions in economic investigations Indexing Ind»'x of tulies India inks I ndices I iidin'ct wind load Indirect \vin i:i4 IH;-) 100 245 14 •>m 290 244 294 281 290 2G0 29»i 297 203 28() 340 28S 284 323 !)(> 101 97 97 155 20 15 5 9 317 177 107 i'33 388 INDEX. PAGR JofTpiaoii City Bridge, anchorage for draw-spi'Ti 124 •leffer«oii City Ikidge piers 313 Jefferson City Uridge triangulation 324 .JoiatH for highway bridges 214 Judgment in designing, necessity for 15, 23 Kansas City & Athinti(? Railway l>ridge 114. 135 K. C., 1*. & (J. Railway bridge over tlie Arkansas River.. 310 Keyiioies for hand-levers of draw-spans 211 Knots in timber 2U9 Labor-saving devices, use of 13 Lacing 174, 175, 232 Lacing, incapacity to carry transverse load 20 Lack of 0,'sthetic treatment in American bridge de- . signs, rea.sons for 39 Ijack of economy in cantilever bridges 55 l^ack of rigidity in cantilever bridges 55, 57 Lack of rigidity in suspension bridges 57 Ijansdowne cantihfver bridge 07, 08 Latches for draw-spans 209 Lateral bracing for highway bridges 219 Lateral bracing on curves in elevated railroads 180 Lateral struts, sections «)f 17i Lateral systems ior draw-s|)ans 192 Latenil systems, weight per foot of 329 Lattice girders, objection to 19 Laying out work 328 I^ayout of structure, general 330 Layouts for elevated railroads 331 Layouts for viaducts, economic 87 Least thicknesses of drums 19-1 Least thickness of metal in highway bridges 22(5 Legitimate and illegitimate decoration 4(5 Length of centre panel for draw-spana 183 Length of pin-plates 14 Lengths, effective 145 Lengths of base-lines 322 Lenticular arches 83 Lettering 339 Letting bridges by the pound 139 Levelling off tops of pivot piers 195 Levels in pier-sinking 320 Liberal allowance in estimating weights for details 28 Lift-bridges 108 Lifting deck 114 Tamitation of scope of all specifications 1,') ;.i;niting lengths of caj^tilever b"M'ketf». ....,,..,...,.. %W IND£X. 389 , 31) . 55 55, 57 . 57 07, «B . 209 . 219 . 180 . 171 , 192 . 329 19 328 330 331 87 191 22(5 4() 183 14 145 , 322 . 83 . 339 . 139 . 195 . 320 . 28 . 108 . 114 . 15 . m i PAOI Limiting lengths of continuous steel stringers 227 Limit of \vorking-Htre88 8 Jjimits in bridge-designing 160, 220, 227 Limits of crroiH for grapliics 335 l^inkw for toggles of draw nuicliinery 208 Jjist of data tor spans 332 jjive and dead loads, eoinparative importance of 7 Liv<^ loads for eantih^ver bridges 00, 01 \A\i' loads for eonihiiied hri Locating piers during sinking 32(> Location of base-lims 321, 322 Location of structiu'e as atl"«>cting its decoration 40 Long panels 51 Jjong sights in triaiigulation 323 Long steel taju", measurements with 319 Loop-eyes 178 ] ioose rivets 255 Lower-chord packing 17(5 Lower lateral systems of higliway deck-bridges 219 Lower lateral systems of railroad deck-bridges 219 Lower tracks of turntables 194 Lubricating sliding and rolling surfaces 28!> JjUmp-sum bids 2 Machinery for drawbridges, care of 129 Machinery for draw-spans 201 Macliinery for Halsted Street Lift-Bridge Ill Machinery-houses for draw-spans 211 Main central spans of cantilever bridges, stresses in. . 59,(50 Main nusmbers of highway truss-bridges 218 Main members of railroad truss-bridges 140 Making detail drawings, method of 337 Making drawings 335 Man-power macliinery 128 Man-power operating apparatus for Halsted Street Lift-Bridge 112 Man-power operating machinery, necessity for 202 Margin lines for drawings 339 Mask of ornamental cast iron 42 BdO li Measure of strength of structure 1(> Mechanical-power turning-machinery, formulaj for 203 Medium steel, reason for using 8 Memphis Hridge 08 Menominee Canal bascule bridge at Milwaukee 107 Metal 244 Metal- work for timber trestles 270 Method of checking finished design 21) Method of determining i)0\ver required for operatiiig and lifting draws 202 Method of making detail drawings 337 Method of measuring with steel tape 31!) Method of utilizing ecjuivalent loads 207 Methods of operating revolving draw-spans 120 Methods of pier-sinking 301 Methods of study of testhetics in any bridge design.. . . 4!> Methods of testing 240 Metro; olitan Klevate . 12 . 2ya . 41 . ir)7 . 103 . 18 25, 2() , . !;')() . . 2;")!) . . 31(> *> I nianufailurcrs to iinprovenicnt 10 Opjiositioii to lestlietic reform in bridge-building 45 Opposition to ]Udposed metliods ol design 4 Ornamculal i-asting, mask of 42 Ornamcnial hand-rails 52 Oniaiiiciitalioii for bridges. anj)ro])riate 47 Ornamentation of bridge approaches 53 Oi iiamciitat ion of elevated railroads 52 Ornament at ion of ])ortals 52 Ornamentation of viaducts 52 Outlines of truss bridges, harshness in 42 392 INDKX. PAGE Outlining spans, study of testlietifs in 50 Overturning of piers, resistance to 307 Paint-brush, accessibility to 25 Painting, aesthetics in 52, 53 I'ainting, inspection of 290 Painting, investigations for !) Painting, specifications for 25iJ Pairs, arrangement of members in 21 I'altry brackets in elevated railroads t)9 Panel lengths, economic 33 l*a])cr on Klevated liailroads 91 Part henon 43 I'aved doors 21(5 l'edestal-('a|)s for jlevated railroads 93 Pedestals 177 I'ech'stiils for cantilever bridges OtJ Pedestals for elevated railroads, sizes of 181 IVdestals of t«)\vers in llalsteil Street Lift-IJridge 109 Percentage-curves for weights of cantilev«'r bridges, 72,74,77,78 Petit truss, superiority of 19 Piece-work 2(54 Pier centres, triangnlati(m to 325 J'iers, designing of 301 Pier-sinking 315 J'ier-sinkiiig, methods of 301 Pile foundations 303 Pile piers 314 Piles in concrete 303 Piles, specifications for 277 Pile-trestles 273 Piling 293 Pin-connected trusses for elevated railroads 98 Pines 298 Pinholes : 1 7(5 Pinholes, specifications for 257 Pinion brackets 1 93 Pinions for draw-spans 125, 201, 205 ]'in-metal 251 Pin-plates 175, 233 Pin-proportioning 17(J Pins, bending moments on 157 Pins, specifications for 258 Pipes, removal of 291 J'ivot-piers. drainage of 128 Placing concrete under water 283 IKDEX. 393 pagk 50 307 25 )2, 53 290 9 259 21 99 33 91 43 21(> . 93 . 177 . ()G . 181 . 109 77, 7H . 19 . 2H4 . 325 . 301 . . 315 . . 301 . . 303 . . 314 . . 303 . . 277 . . 273 . . 293 . . 98 . . 298 . . 170 .. 257 . . 193 01,205 .. 251 75, 233 .. 170 .. 157 .. 258 .. 291 .. 128 .. 293 I i Planinor-arums 19,S Plate-jjfirder driiw-spans, field-splicing for 189 Platf {girders, economic depths of '.m Pneumatic process 302 Pouy-trusH bridges 2 IS Pony-trusses for elevated railroads !I8 Portal bracing for railroad bridges 147 Portal bracing in liighway bridges 21!) Portals, ornamentation of 52 Posts, sections of 171 Powdered dialk or talcum 34(» Power for draw-spans 12(», 201 Power for op«'rating .lelVerson City draw span 121 Power re(|uired foi Pieliminary data for elevated railroads 102 Preliminary dead load 333 Pre|)aratioii of Ijcd-rock to leceive masonry 291 Preservation of metal- work 10 Pressure on screw-threads 208 Prevention of rising ends 209 Price for lirst-class inspection 288 Principles of designing 12 Principles in designing, adherence to 4 i'robability of stress application 22 Projecting web-plates 257 Projection of i)iles into cylinders 313 Proper colois for painting bridges 53 Proper conditions for arch bridges "9 Proper distance between expansion j)oints in elevated railroads 95 Proper kinds of draws for various crossings 118 Proper lengths for anchor-spans 75 Proper live loatls for cantilever bridges 00, 01 Proper loads for piles in cylinders 313 Proper locations for cantilev(>r bridges 55 Proper locations for suspension bridges 57 Proper method of letting bridges 3 Proper pricj for lirst-class inspeition 28S Proper relations between outline diniensions of spans. . . 51 Proportioning details, general principle for 28 Projtortioning of built members 174 Proportioning of jdns 17(5 Proportioning turntabb's of draw-spans 128 Proportioning web-anglcs coniu'cted by one leg only. ... 170 proposed association of inspectors 283 394 IKDEX. . PAGE Protecting column feet 291 Protecting hubs 320, 321 Provision for clearance in packing 24 Provisi(»n for i>Hect of impact ](}, 17 l^iovision for expansion and contraction 23 Piovision for factor of ignorance 7 Provision for thrust transference 20 PwU-back draws 104 Punching, accuracy in 254, 287 Punching and reaming 250 Quality of stress, consideration of 22 tjuarry-l)ed . . . 297 (Quarry-sap 297 Quenching 248, 249 Packs f«)r turning draw-spans 198 Padial struts for drums 193 Kail inspection 290 IJail lifts for draw-spans 208 Kailroad structures, specifications for 141 Kails for elevated railroads 93 Kaising column feet 181 Raising colunui feet of elevated railroads 101 Katio ot" Icngtli to least radius of gyration 100 Katio of working intensity to elastic linnl 8 Katios of length to least radius of gyration in liighway Itridgcs 227 Katios of truss-depth to span-length 32, 33 Katios of weiglits of cantileviMs and lived spans 75 Keaiinng 250 Iveaming, necessity for 9 Keasoiis for lack of a'stiietic treatment in American bridge designs 39 Keasons for unsatisfactory conditions in existing ele- vated railroads 24 Keasons for using medium steel 8 Kecordiug of cctmbiiuitions of stresses 334 Jiecording of wind stresses 334 Kecording progress of shop- work 28(i Ked i{ock Bridge piers 307 Ked Ivock cantilever bridge 04, 08 Keduct ion of area 248 Reduction of lield-riveting to a nunimum 24 Kedundant members, objections to 1!> Ke-entrant corners 254 Referencing hubs 321 Reform in bridge-building, testhetic 45 f1 INDEX. 395 PAOK 21)1 ,:j-21 '24 (», 17 23 7 '20 104 '287 2m 22 297 297 249 198 193 290 208 141 93 181 101 100 8 y . 227 32,33 . 75 . 250 9 n . 39 . 24 8 . 334 . 334 . 28() . 307 04,08 , . '248 , . 24 , . 19 . . '254 , . 321 . . 45 PAOK Reiniorcing-Dlates at pinlioloa 14 Kojectioii t)t' innteii il 202 lU'latioiis iK'twccii priiu-ipal ilinienHions of arches 84 Relations iK'twi-eii various leading dimensions in canti- levoi" bridges 77 Removal of [)i|)es and sewers 291 Removal of surplus material "291 Repairing bridges 139 Reports of tindx-r inspectors 299 Re-railing api)aratus 142 Resistance to overturning of [)iers 307 Responsibility for accidents 203 Resting longitudinal girders on cross-girders 98, 180 Reversibility of draw-spans 207 Reversing-stresses 1 ;7 Reversion of sti'csses in liottom chords 32 Revolving drawbridges 119 Rigid bottom chords in highway bridges 220 Rigidity, importan»-e of 15, 10 Rigidity of architects' ideals 42 Rigid lateral bracing for railroad truss-bridges 147 Rigid lateral syst«'ms '24 Rigid sway-bracing for cantilever bridges (i5 Rigid to]) ( hords for draw-spans 183, 191 Rim-bearing rnxiis c«'ntre-bt'aring turntables 123 Rising ends of draw-spans 120, 209 Riv«'t -heads 255 Hixet-holes 255 Hiveting 104 Riveting in highway l)ri(lge8 227 Riveting, necessity for symnu'try in 22 Rivets 255 Rivets in direct tension 25 Road roller 222. 223 Rolled steel 245 Roller boxes 177 Roller plates 177 Rollers I77 Kolleifi for draw-spans 15)5 Rollers for draw-spans, tests of 253 Hollers of Habted Street Lift-Hridge 110 Hollers, specifications for 259 Koof-trusses 33I Roof- trusses, cantilevers for 57 Roof-trusses, economy in 57 Rust-cement 194 'i^^6 INDEX. PAQB Sand-briquette testa 296 Scales for drawings 338, 342 Scientitic detailing, importance of 131> Screws for draw-span machinery 207, 20S Sectional areas of web members for open-webbed, riveted girders 27 Sections and section lines (drawings) 341 Sections for columns of elevated railroads })4 Sections of compression llang(!S of plate girders lt»7 Sections of lateral struts 171 Sections of members for highway bridges 228, 220 Sections of meml)ers of elevated railroads 180 Secti Superelevation for wooden trestles 27!) Superelevation on curves of elevated railroads $)5 Superelevation on curves of railroad bridges 143 Superlluous metal, use of 17 Superiority of cantilever bridges, imaginary 55 Superiority of Petit truss 19 INDEX. 399 PAOR Supplomonlary aiifjlos for opon- webbed, riveted giniers. . 27 Surplus material, removal of 2{)l Suspended span, eeonomie leiif^tn of 70 Suspended spans, eonneetion of (59 Suspended spans of cantilever bridj^es, stresses in oS Suspenders for cantilever hridj^cs (»5 Suspcndeis for iiij,'li\vay bridjjes •22(l Suspeiulers for railroad bridfjes 14H Suspension bridges, lack of rigidity in 57 Suspension bridges, proper locations for 'u S\vay-l)racing for deck-bridges , 21'.) Sway-bracing for draw-spans 191 Sway-bracing for highway bridges 219 Symmetry in layout about a middle ])lane 48, 49 Symmetry in pier designing ;i()S Symmetry in riveting, necessity for 22 Symmetry of section, necessity for 21 Systemization of cantilever-bridge designing o7 Systemization of knowledge 34(5 System ization of work \'.i Table of data for various cantilever bridges 71 Tape-lines ;}1K Tapered reamers 25(> Targets ,'J25 Temperature, ell'ects of chajiges in 155 Temporary bracing for steel cylinders ;{12 Tenders ' 2(i4 Tensile strcnigth 247 Tension on anchor-bolts in trestles 148 Testing cement 295 Testing foundations 29 1 Testing full-size mendiers 288 Testing-niachiiu\ speed of '. 247 Testing, methods of 24(1 Testing openiting machinery 289 Testing [)aints 10 Testing taj)e-linea ;} 1 8 Test-j)ieces, annealing of 247 Test-pieces, number of 249 Tests of angles connected by one leg only 27 Tests of full-size eyebars 25(1 Tests of full-size meinbers and details 253 Tests of rollers for draw-sj)ans 25;j Tests of star-struts 28 Text-books on substructure .301 Thickness of track-segments 195 400 INDKX. PAGE Thinning paints 2G() Tlioroughtures, closing 202 Thorouglnifrts in inspertion 281 I'hiouda, speeificiitions for 259 Thrust transference, provision for 20 Tie-plates 174, 232 Tie-spacing for elevated railroads 102 Tiiid)er 201, 293 Tind)er-b<)lts 142 Timber caissons 308 Timber conijjrossion-nieiinK'rs 270 Timber, defects in 299 Tind)er, e.vtreiiie fibre-stress for 150 Timber for trestles 277 Timber, griUage 303, 309 Timber inspection 297 'i'indier inspectors, reports of 299 Timber piers 314 Timber portions of highway l)ridges 214 'I'imber j)ortions of railroad slnictiucs 141 Timber stringers 274 Timber trestles 273 IMtles for drawings 339, 342 Toggle-links for draw machinery 208 Toggles for semi-cantilevers 62, 03 Top chords, sections of 170 Tops of columns in trestles and elevated railroan 270 'ricstlc (•olumiis. Itultcr for 87 Tn'sllc coliiiiiiis. conibiniitioii of Htivssps in 88 'I'lcstli' coliiiimH, (If'tailH for 87 Tn-sl Ics HO Tn'stlcH, details for 87 Tn'stlcs, t'cniioiiiy in 30, 84 'i'r«'stl('-to\vcr III acing 179 Trestle-towers 148 Trestle-towers, hraeing tor 8!) Trestle-towers for liighwuy structures 220 Trestle-work, inspection of 293 Triangulation 317 Triaiigulation-liubs 320 Triaiigulation-shoets 325 Triaiigulation to pier-points 325 Triiiuiiiiif,' cliaiiiiel-ends 170 Triuiniing desij,Mis 2 Triuiming idle cornel h 254 True economy in design 30 True economy, necessity for 15 Truss depths, economic 30 Truss deptlis for draw-spans 183 Trusses, weights [ter foot of 32!>. 330 TnisHCH with parallel chords, economy in 31 'I'russes with ])olygonal toj) ( hords, ((conoinic depth of. . 32 Tubes, filing 345 Tunil)U<'kles, siiecilieations for 25S Turned bolts for turntables 1!)3 Turned bolts, specilicationa for 258 Turntables, details of 192 Turntables of draw-spans, proportioning of 128 Twist-drills 250 Two-angle struts 171 Two-rivet connections, objections to 25 Types of arches 80 I'ltiinate tensile strength 247 Unbalanced wind-pressure on draws 202 T'niforni sizes of drawings .335 Ifiiit weights of materials 152 Unit working i)ressure on screw-threads 208 Unnecessary cantilever bridges 55, 56 402 INDEX. PAOK Unprotefied ooiu rote j)i('is HlO UnsjitiHiiu'ldi V coiuliti*)!!-* of existing clevati'd ruilruadH, rcasoiiH for '21 I'iisii|)|Mir(c(l widths of jtlatcs in foiiiidcssioii 174 Uplift for liij^'liwiiy dra\\-«i»iiiis, as-mnuHl 'I'Mi r|»lift loads for diMw-spans IHI Upward dead-load reactions in draw-spans 12*2 list' of east-iron trininiinj,'s for deeoraiion ;V2 l's«' of dynamite in pier sinking ;{(),'{ Use of high steel S Use of jndgment in designing 15 Use of labor-saving devices i;{ I'se of powder in quanying 21)7 Use of slide-rule 3'.i',i Use of superlluous metal 17 Use of supplementary angles for open-webbod, riveted girders 27 Use of water-jet in pile-driving 314 Vain etForta of engineers to compromise with grace by ornamentation 41 Value of faculty of judgment 14 Van Buren Street IJascule J$ridge, Chicago 10(i Variation in cost of piers with s])an length 3.> Variation in weight 25 1 Variation of truss- weight with s|)an-length 3.'> Variation of weight of lateral system with span-length.. .'<;") Variation of weight of metal with length of opening in cantilever bridges 7<> Various tyj)es of arches HO Vertical posts, sections of 171 Vertical sway-bracing 21i> Viaducts ..." ^^i Viaducts, economic layouts of H7 Viaducts, ornamentation of A2 Violation of principles in designing 4. o Washers, s|)ecifications for 2r>s Wasbing drawings with benzine .'MfJ Waste of money in designing ."» Water-jet, nae of, in pile-driving. ..." 314 Water-power for operating draw-spans 120 Wearing-floors 2 1 5 Web members of open-webbed, riveted girders, sectional areas for 27 Web-plates at ends of open-webbed, riveted girders 100 Weba of drums 103 Webs of girders for elevated railroads 94 INDKX. 403 PAOB \V«'l)-si)li('OH in (Iniina I!»;{ \V('l)-s|)li<»'s ill piiih' jjinlfTs l(i(i Well .sliUViKTs for plain j^irdt'iH KJS W'cli-sliUViiiii}; struts, soctioiis ot 171 W'cij^'liiiij,' rciict ions for dniw simiis 122 Wcijrht of IJalstfMl Street, Lift-Hrid^e 1()!» \\«'ijf|it of iiiiisomy ill iiiielioi-|iiers of eiiiiliiever luidp's. (i7 Wei^^'lit of metal ill aiuiiora^ies of c.iiit ilever hiid<,'es. ... I'.i \\'<'i<,'lil of iiietiil ill aiiclior-spaiis of cantilever hridj,'es. 7.S, 74 W'eij^lit of tiiivelier, elleet of. on erection stresses (iO \\«'i<,dits of l)ri(l«,'es let by liiiiip siini and hy pound price. 2 W'eifjhts of trestle-aiicliora;^es W'ei^lits per foot of trusses ',i2\) Wells ill anehoin'res of eantilev-r l»rid};es {\(\ ^^'heel-J(ual•ds for iiif^lnvay I>rid{,'es 215 Wiiipple truss, ohjeelions to 1!) ^\ ideiiinj; cantilever hridj^'es over main piers (V\ \\'iilo\v mattresspH around piers 314 Wind loads for draw-spans 185 Wind loads for iiififliuay hridjjes 224 Wind loads for liiphway viaducts 224 Wind loads for railroad hridj^es 153 Wind stres.ses. recordinjf of 334 \N'inner Brid<,'e 114 Wooden eompressioii-inemlxTs 276 Wooden liand-rails for liij^hway l)rid«ifes 215 Wooden st rin<.'( IS 274 Wooden trest le liraciii