IMAGE EVALUATION TEST TARGET (MT-3) P 1.0 I.I ■ 40 mUkwn " ■■" |2.0 ^;. /S« '•y '/ Hiotographic Sciences Corporation 33 WEST MAIN STMET WnSTU.N.Y. I4S«C (716) •73-4S03 ^" CIHM/ICMH Microfiche Series. CIHM/ICMH Collection de microfiches. Canadian Institute for Historical Microreproductions / Institut Canadian da microraproductions historiques <^ TAchnicai and Bibliographic Notes/Notes tecliniques at bibliographiquas Tha inatituta has attamptad to obtain tha bast original copy availabia for filming. Faaturas of this copy which may be bibliographically uniqua, which may alter any of tha imagas in tha raproduction, or which may significantly changa tha usual mathod of filming, ara chackad balow. n Colourad covars/ Couvartura da couleur Covars damagad/ Couvartura endommag6a Covars rascorad and<'or laminatad/ Couvartura rastaurte at/ou nalllcul6a I I Cover title missing/ La titra da couverture manque I I Coloured maps/ Cartas g^ographiquas en couleur □ Coloured ink (i.e. other than blue or black)/ Encra da couleur (i.e. autre que bleue ou noire) I I Coloured plates and/or illustrations/ E D D n Planches et/ou illustrations en couleur Bound with other material/ RelW avac d'autres documents Tight binding may causa shadows or distortion along interior margin/ La re liure serrie peut causer de I'ombre ou de la distortion la long de la marge int^rieure Blank leaves added during restoration may appear within the text. Whenever possible, these have been emitted from filming/ 11 se peut que cartainas pages blanches aJoutAes lors d'une restauratlon apparaissent dans la texta, mais, iorsque cela Atalt possible, ces pages n'ont pas AtA fiimAes. Additional comments:/ Commentairas supplAmentaires: L'Institut a microfilm^ la mellleur exemplaire qu'll lui a 4t6 possible de se procurer. Les details do cat exemplaire qui sont peut-Atre uniques du point de vue bibliographiqua, qui peuvent modifier une image reproduite, ou qui peuvent exiger une modification dana la mithoda normala da filmaga sont indiqute ci-iS^ssous. r~l Coloured pages/ D This item Is filmed at tha reduction ratio chackad balow/ Ca document est filmi au taux da reduction indiqui ci-daasous. Pagea de couleur Pages damaged/ Pages endommagtes Pages restored and/oi Pages restaur^as at/ou pelliculies I — I Pages damaged/ I — I Pages restored and/or laminated/ Pages discoloured, stained or foxed/ Piges dteoiorAas, tachattes ou piqutes □ Pages detached/ Pages ditachtes HShowthrough/ Transparence □ Quality of print varies/ Quality InAgale da I'lmpreasion □ includes supplementary material/ Comprend du material suppl6mentaire □ Only edition available/ Saule Mition disponible Tl to Tl o\ fil O b< th sii ot fil sii or Pages wholly or partially obscured by errata slips, tissues, etc., nave been refilmed to ensure the beat possible image/ Les pages totalemant ou partieilemant obscurcias par un fauiilat d'arrata, una pelure, etc., ont 4ti filmtes k nouvaau da fa^on h obtanir la maiiiaure image possibla. Tl sh Tl w M dii er be "2 re( m( 10X 14X 18X 22X 2SX aox ^ 12X lex aox a4x 2tX 32X 1 tail* idu odifier une mage The copy filmed here has been reproduced thanks to the generosity of : University of Saskatchewan Saskatoon The images appearing here are the best tjuality possible constidering the condition and legibility of the original copy and in keeping with the filming contract specifications. Original copies in printed papnr covers are filmed beginning v/ith the front covef and ending on the last paf;e 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 printea or illustrated impression. The last recorded frame on each microfiche shall contain the symbol ^^> (meaning "CON- TINUED"), or the symbol ^ (meaning "END"), whichever applies. IVIaps. 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 upper left hand corner, left to right and top to bottom, as many frames as required. The following diagrams illustrate the method: L'exemplaire f ilm* fut reprodult grice A la gin^rosit^ de: University of Saskatchewan Saskatoon Les images suivantes ont tti reproduites avec ie plus grand soin, compte tenu de la condition et de la netteti de rexemplaire filmA, et en conformity avec les conditions du contrat de filmage. Les exemplaires originaux dont la couverture en papier est imprimte sont film6s en commenpant par ie premier plat et en terminant soit par la dernlAre pege qui comporte une empreinte d'impression ou d'illustratlon, soit par Ie second plat, salon Ie cas. Tous les autres exempilaires originaux sont filmfo en ccmmenpant par la premiere page qui comporte une empreinte d'impression ou d'illustratlon et en terminant par la dernlAre page qui comporte *in9 telle empreinte. Un des symboies suivants apparattra sur la dsrnlAre image de cheque microfiche, selon ie cas: Ie symbols -^ signifie "A SUIVRE", Ie symbole ▼ signifie "FIN". Les cartes, planches, tableaux, etc., peuvent Atre filmte A des taux de rMuction diffirents. Lorsquj Ie document est trop grand pour Atre reproduit en un seul clichA, il est film* A partir de Tangle supArieur gauche, de gauche i droite, et de haut en has, en prenant Ie nombre dMmages nAcessaire. Les diagrammes suivants illustrent la m*thode. rrata :o pelure, 1 A □ 32X 1 2 3 : ,t .-- ■ t : 4 5 6 Wl SS^^^i^^^ """U^ ja. J. . "1 M . J- t H '^ t ' 1 -> - tMJ^^n^ili>^imiSp>l^i2:OT 1! ELHVATION OF BR ScaZe 80 ft. (^, ol/v inch/ t I Sr'f/.e Fi//y /,;^,J, '/•. ^'^Z^? '^^? anikh- EI!Cn .-t^. 1 i-T-i-fi ! ' . i rl i M I ! I ; n M 1 I I l-l 1 ! HI 'J I ■- i I -' ■ f t -H \- , • ^M i ''ii I -i ! )- 1-! I H It ' H k..^ .,■14, •yv "rv^'-^^'f-v^^Krij — \TION OF BRIDGE. /I I ! 1 Fift.7. TT" L _ji — -III — h — ^1- I'Vj.." C!)! /•;■ Scalg'iOft to o-rL inch, Fio. P). i'JV/' .(J ■Ij Mil ^|i — 1 II' ii 1 """" " ' 1 Jll ill jII III ill- -III 1 /' ./x /> f.V ^B =^ . t1iil 4# SiUlion o/'i'S' "s/xiii I V.I. ir^H it"'*' •^r,«'' J.. *^ ■>-'.. J^f!> c-IH< •'i'!*' '"t^f-'i it' «■», r> * »\ "C'V^V; " ;: «. •■,*•'■.: .■* : ..v#^j. ^i' - *;•;'• ■"-J, 'A: TO THE PUBLIC. •*<'^ii I' f'.V*' ^ '(•'.••■■jI An c'xporionce of nearly a (luartor of a century in construct in jj; and reconstructing railway bridges under the most diifi''ult circumstances, and the observations of a professional career extending over a much longer period, having convinced the subscriber, that there are more errors in bridge architecture irom the employ- ment of elements which are superfluous and deranging, tliun from the lack of those which are sustaining and ivciprocating. a sericH of experiments was instituted, extending througli several years, under circumstances t'livoralile to obtaining reliable data for the purpose of estaldishing principles which were not generally ad- mitted; and which have resulted in the " Istmietrical Bridge Truss," herewith suljmitted, for which a patent liiis been obtained. The review, by Mr. Charles Mac- donald. civil engineer, of the ])rinciples, both mechan- ical and economical, which are involved, will be found not only interesting to the })rofessional j'eader, but also useful to bridge builders at large. A Company, composed chietly of the younger mem- hersof the profession, has been formed for the purpose of introducing the new system into use ; and as it ':>^'' m ■t' if >ti i* •%i I .»<"•«>''::•' "5 '•"■■■I'* \ •* '.'' ■■ ' ■ til-- .'«••„ \i} *' ■ - .:^i ^ ■ «•■, 'f. ■ \ - . •-, . • "• ' •'Nv TO THK P1JBMC. rests its claims for popular favor on the solid basis of practical experience, f!;overneil and harnioniy,ed by ac- curate science, jrreat confidence is felt in tbe advan- tages which will accrue to the railway interests at lar 4.;'^*!i^ V - % INGOIU'OllATIOK. VS i Tfie Isometrieal Truss Bri(ij!;o Company has .been incorporated under the act of Assembly of the State ol' Pennsylvania for such purposes made and provided, and the business objects of said company are fully set fortii in the followinir extract from the corporate agreement " In order that we may be and remain a body cor- porate, under the name and style of The Isometrieal Truss Bridge Company, for the purchase and sale of United States h'tters-patent, granted to J. Button Steele, on the 9th day of April, A.l). 18G7, for the Iso- metrieal Bridge Truss, and also for the purchase and sale of other United States letters-patent for improve- ments in truss bridges, the sale of rights and granting of licenses under said patents, or any of them ; also tor the construction of bridges, and for enjoying all the rights and j>rivileges conlerred by said act of As- sembly," The Board of Managers being desirous of offering every facility to parties wishing U) test the merits of the Isometrieal Truss, will furnish plans and estimates and, if desirable, secure the erection of bridges upon this system in any part of the United States. The office of the company is situated in Pottstown, [if] !t;>^:: J -i "J -^'l'* ••• • 1' INCORPOnATION. Montgomorv County. Pennsylvnnia, and all applioa,- tiv/ns for liconisi^s shouUl l»e a;(' ('nm|i:my. Pottstown, Pii. >.'■' ■1 :'• • V; ii^- ■•V ■ • - .\--^ Tc^pies oi' the aooonijianyinii, pamphlet can he oh- tained hv addressinii the Seeietar\- as al)()ve. \\h(» will • >- • also he in a jtosition lo lui'iiish any other information which mav Ic' desired. ■ V 1 ^-V ■.--. - ' 4..,.. ... 4. THE ISOMETRICAL TRUSS. Wk propose in the following pagos to give a 1)i'iof de- scription of the peculiaritieH of a ('(^nibination in ln'idge con- stnu'tion, to whieh the name of The rnonietiieal Trus?* ha> IxM'ii applied ; and to the proper nnder8tainling of the same, it will l»e nqcessarv to look tsoniewliat in detail into the general theory of strains in girders having horizontal flanges. The writer approaehes a suhjrct, whieh in its entirety involves so nuieh to interest the engineer, with eaution. ami. it is to he hoped, a dne proportion of modesty. The ohjeit ti) he attained is sim[)ly truth ; and if in this attenipt, a jii>t and seareliing eritieism shall have demonstratt'(l the fallaty of the views expressed, sojiie good will at leasr ha\(' heen done in e.xposing the errors of a few for the henetit of all. The general laws regnlating the aetion of strains in a lieain supported at hoth ends and loatled throughout its It'iigth, have heen thoroughly inv«'stigated hy some of the most emitient physieists of the day. And it may not l)e |ti'einatnre to assunie that the conclusions arrived at are "ri )ricnv as lollows 11< All heams uniforndy loaded and supported at hoth ends. iii'c Nuhjeeted to liorizontal aiid verti«al strains. The liori- zonfal strain is a ma.ximum at tlie central cross section, iind decreases toward eacli [)oint of support in p»»p(M'tion jr 'J A; :« V Mi f t, ti 8 THE ISO METRICAL TRUSS. . • it ''■'•. >*■ ■■•I ■.•♦•*■ ■ '* V.***- ■ •>; . fi, ■ ■■ ■•• J ■' t.- »!%■■<*?♦■' - • ...... Vjr . .■■- .-.C:*-- .» ^.-^^ "tsJi .".-•*"■•.', ■•* *'" f •• ^■^.♦A■ ■ H to the ordinates of a parabola. The vertical strains are zero at tiie centre, and increase directly with the distance toward each point of support, where they become equal to oue-half the uniform load. In ca5tion. Fig. 2. Fig. 2 represents a case in whieli tlie-load \\V is com- municated to the top chord hy means of vertical ties: should the load he applied directly to the top chord, stmts would be substituted for tics. The force 1\ evidently tends to pro\f ] .'^»««>., ♦■■■., (^''>l»".^.|rf. ■> k y- ••■ . "' .'!^v, ■ ■,>,■ •-■.*.. I " fe'V .; • ,.' • • • ^'^^ ''• »:''■ VT»'- . '■ V /♦•■^ (•« . «.i ;i •.•».' ■ . • ' ♦ h '" * .•♦. , 12 THE I80METRICAL TRUSS. practically, tliat the most economical angle with the verti- cal in a triangular system 18 45", hence, if all other things were equally appropriate, the proper form for a parallel girder would involve a right-angled triangular web. In practice, the nearest approach to this is the Warren girder, in which the triangles of the web are equilateral ; and inasmuch as the relative economy of this system is only 15 per cent, greater than the theoretical case, its many practical advantages have earned for it the first position. The question is now reduced to the Warren girder as the most economical form of bridge truss, having parallel flanges, and the conclusion is general in its application whether the structure be of iron or wood. At a subsequent period of the investigation we shall see that economy is not the onlv advantiiajeous circumstance connected with a triangular system. Fig. 6 will serve to represent what is known in this country as the triangular truss. It differs from the original Warren in the introduction of vertical ties and posts, which arc rendered necessary in order to afford intermediate points of support between the apexes of the main triangles. It is evident that so long as the spans to which the simple AVarren combination is ai>plied, remain within such limits, that the unsupported distance in each panel shall not require additional strength in order to enable it to sustain the distributed load, the vertical members in Fig. 6 may be dispensed with ; but unfortunately for the success of the system this condition limits the extreme i^pans to a very small margin. Thus in the case of a rail- road bridge required to sustain a rolling load of one and a half tons per lineal foot, and at the same time to present sufficient resistance in any one panel to the weight of the heaviest locomotive — say forty-five tons — as applied at the points of contact of the driving wheels, we should be ob- liged almost from the start to introduce vertical supports, or otherwise increase the diinensions of the chords beyond what would be required to resist a horizontal strain due to the whole load, since the panel load per lineal foot may greatly exceed, possibly even double that quantity. Assuming, then, that the vertical members are a practical \m THE I80METRICAL TRUSS. 13 Pig. 6. Fig. 7. 'iC' »^ ''v I Br" I ►/*i i<.' lA ^^■ V ! .1, ■ ■■■.■' i. ;!»:■•■■■ V. f ,■•;-. C ■H ♦■■ ^•« •• .'"'■■. ■ '■ sS^'- ■ ■•I'T <\ .1 TV ^'■' • ■:'■''■ l:\ «••*.,•< 14 THE ISOMETRICAL TRUSS. necessity for all eases witliin the reach of our present in- vestigation, and tliat their introduction was not rendered theoreticallv necessary in order to sustain tlie uniform h)ad; in other words, that they are totally incapahle of opposing any resistance to change of form in the panels; therefore the material tlnis applied is a dead load upon the useful members, jind consequently necessitates an additional amount of material in them to transfer this load to the abutments. ISo much for the question of economy; but there is another feature in the use of a single system of triangles, which begins to assume important proportions as soon as the spans are increased beyond tlie ordinary limits now considered safe; which is due to the increasing dimen- sions of the diagonals nearest the points of support. It is evident, from a consideration of Fig. 6, that these mem- bers must be capable of sustaining a strain equal to one- lialf the entire load upon the truss multiplied by the ratio between the diagonal and perpendicular. Now as the span increases, all other things being }>roportional, so must this strain increase, and as a natural conseipience the sectional area of resisting material ; and when this area reaches a certain limit it becomes necessary to introduce expensive appliances to fulHl the necessary conditions. In the case of a Howe truss, this limit is reached at u span of one hundred and tifty feet, as will be seen in the discussion of that particular system in a 8ubse(pjent portion of this ])aper. Bearing these facts in mind, it is presumed that we are now in a position to appreciate the introduction of two independent systems of triangles in the arrangement of the web. Referring to Fig. 7, it will be observed that the diagonals represented by full lines, form, with the top and bottom chords, a continuous Warren girder, precisely similar to Fig. 6, with the exception of the vertical mem- bers. These members, we have seen, are rendered prac- tically necessary in the iiansmission of intermediate load to the diagonals. They add to the dead weight to be car- ried, and do not in any way increase the general strength of the structure. Supjtose, as in Fig. 7, that these verticals be dispensed with, and a second Warren system introduced '^ • m TUE ISOMETIUCAIi TRIJ88. 15 as represented l>v the dotted (liaj^onals; it is perfectly clear that the first system will he relieved of one-luilf the load which it would otherwise he called upon to sustain, and that this amount will he taken ui» and curried to the abutments hy the second system ; at the same time all the practical considerations involved in the vertical members have been done away with, and no more sectional area, and con- sequent material, is recpiired in the diagonals than would he called \'or in the single full system (Fig. 6), supposing it practically capable of being adapted to a moving load without the use of interm diate supports. Hence, since ail niemberB which do not assist in preserving the general etiuilibrium liave been removed, and the remaining ones subjected to an equal share of duty imposed in very nearly the proper theoretical direction, it may with safety be as- sumed that a web system involving a double set of triangles, substantially as indicated in Fig. 7, possesses both theo- • retical and practical advantages over any other now in use, coniinending it to general introduction. Such are the main features of what has been called the Isonietrical Truss. The principles upon which it is con- structed being general, it is ecpially applicable either in wood or iron; but as the requirements of railroad enter- prise in this country demand more than any other the introduction of n cheap and substantial wooden railroad bridge, we shall devote the greater portion of our remain- ing space to the application of the isometripal principle to wooden bridge construction. Within the limits of a mere pamphlet, it would be quite impossible to attempt anything like an extended examina- tion of the different kinds of bridges in use throughout the country. Suffice it to say, that although there are a great variety, and many of them fulfilling all the require- ments of the original design, yet practically there are but a very few systems acknowledged to combine sufficient advantages t<) insure their continued adoption; and it is ' with these we propose to occupy our readers' attention, by way of instituting a comparison between their internal economv and that of the Isometrical Truss. I i. •Vj* m •Mi m •r ^li (.^* ,'*W>.. ■♦'.'# !*•*•' 'Ui 16 TOR laOMKTRICAIi TRUSS. The Howe bridge is probably tlie most extensively adopted wooden truss in the United States. Why it is so is probably owing to a variety of circumstanees, some of which may possibly not have tlieir origin in the scientific merits of the combination; but the fact of its continued success must l>e taken as evidence in its favor, and it may therefore be assumed to be one of the best wooden bridges in general use. The original Howe truss is too familiar to those interested in bridge construction to need any minute description. Fig. 9 will serve to illustrate its gen- eral design. Previous to its invention, what was generally known as the Burr bridge, commanded the greatest slnire of attention ; but upon the introduction of railroads, it became necessary to devise a structure possessing the ele- ments of stability under a moving load to a greater extent than could be obtained in the Burr. The Howe patent eertaiidy effected a great improvement in dispensing with the arch in combination with a parallel truss, and in the practical application of the counter brace, so as to admit of a certain amount of rigidity under a moving load ; but it soon became evident that even this arrangement was not adapted to spans of more than one hundred and fifty feet, if confined to the simple truss, by reason of the dif- ficulty in providing for the increased vertical strain near the points of support. This will ap[)ear more clearly when a particular case is subjected to the test of calculation. Suppose a single-track railroad bridge to be required of one hundred and fifty feet span, the weight of a maximum applied load being taken at one and a half tons per lineal foot, and the weight y>t' structure at a half ton, making a total load of two tons per lineal foot of bridge. The total weight to be sustained by the bridge will be three hundred tons; and since very nearly one-half of this strain is transmitted to the abutments through the web on each side the centre, it will be necessary to proportion the rods and braces in the last panels to resist such strain. The exact figures are one hundred and forty tons for each end of the bridge; taking a divisor for safety of five tons per square inch of section in iron, we require twenty-eight THE 180METRICAL TRUH8. 17 8qnui'C iiK'lics in all for (lie end rodn. In general i>ractice there are six rods in the last panels, or three I'or eaeh truss. Dividing twenty-eight by six we have 4,^^ sqvuire inches tor each, or 2". 44 diameter for round iron. The strain upon the braces will of course be increased by the ratio of the diagonal to the per[>endicular ; but these members l)eing in wood, liance8 for the purpt>se of increasing the number rather than the diameter of end members. It may be insisted that the assumed rolling load of one and a half tons per lineal foot is excessive, and in justiKca- tion of the objection reference might be made to excep- tional cases wherein the llowe truss liad been used for railr(»ad purposes, on somewhat greater spans and much less dimension of rods; l)ut we are compelled to adopt this seemingly high limit, from a ^u-actical knowledge of the extreme load which is sometimes brouii:ht to bear on the bridges of tlie Philadelphia and Heading Railroad, and from a desire to maintain the practice of bridge building fully up to the requirements of the age. In order to over- couie the ditHculties above mentioned, what was called an improvement in the llowe truss was effected by falling back upon the discarded arch of the I'urr in combination with the llowe; a!id this new arrangement has been at- tended with a measure of success; but that there are serious objections to the combination of the arch and truss for rail- road })urposes no practical bridge builder will deny; and it is onlv because no better svstem presents itself that the Improved Howe retains the Held. This is written with a somewhat intimate knowledge of the peculiarities of other wooden bridges, which, in theory, may possibly lay cUiim to merit. For example, the McCallum Intlexible Arch Truss, as explained and illustrated by the inventor, is one of the most beautiful of wooden-bridge combinations, and iX. >; :\7 :^:3 s »•,■ hi. i-tJ m t 111 '^'-^ MB'*'. .•■.'•; 18 THE ISOMKTRICAIi TIIU88. ;•■■" .VJ .... '^ .. •■» _ i^*"-^". ■■ "•• • F "'^>^ .V (it- • i yet wlion siibjoctod to the test of time it is utterly iniprac- ticablo: witii('j*« the McCallinii bridge earrying tlie North- ern Central JIaiUvaj over the Susquehanna River above Ilarrirtburg, which has just been replaecd by an Improved Howe. Next to tbe Howe, and in many particulars superior to it, is the VV^hippie Truss, a inoditication of the Pratt (Fig. 10), whicli we have seen is but a substitution of the diago- nal rod and vertical jtost, for the vertical rod and diagonal brace of the Howe. The practical advantages possessed by this truss over the Howe, consist for the most part iu a saving of material over the piers in case of a through bridge, and in the first two panels beyond the pier in the case of a deck bridge; there are also some minor a.2 feet. Centre to centre of chords, 20.438 feet. Lcngtii of diat^onals, 23.0 feet. Lenjjjth of each panel, 11.8 feet. Weight of bridge, half ton per lineal foot. Rolling load, one and a half torn? per lineal foot. Number of panels*, 14. The numbers attaclied to tlif ditterent members repre- Hont the maximum strain, in tons of two thousand pounds, to which tiiey can possibly be subjected Oy the action of the moving load and the weight of the bridge itself; and *:.vi 4. t, •'I'll' iM-' 'C if "TT^^^l^^^^aarsBT- »«:.' '*r i ■ t ^ ■ ■ ■"' I' (I -4 1 ^W ... * ■ ■ 1 ■J- '=!-*v iw* ..■■'■■V Wf| 20 THE ISOMEiUICVL TRUSS Fix. 8. Fig. 0. "^ ■ /j THE ISOMETUICAL TRUSS, 21 Fig. 10. Fi^. 11. 1 m I •»'•• •*"' t^A, ^ J ' V, ..^.iiWiiU. 1/ '!■ ill ''V <*■ ^' "■'** Vrf i; <.^- *' • ■ ^^■- ■ lii^i.. .''♦:■„ 22 THE ISOMETRICAL TRUSS. .*^' .U'V-, •• ' •^vrV- r.'-. • V Iv'--. ! I* *. ■■■«»- ■■• '^liJ^'^ ■■,'■ ''■"■ ^.'^ • , ,■"• •*..•■*.• • if' ■-'•••. "rfH t'. w^ rs: the signs* + or — represent compression and tension re- spectively. It will be observed that the central members of the Isometrieal Truss are subjected to both + and — strains. This, of course, is to be understood as due to the counteraction of a moving load, and is provided for in the other systems by members introduced expressly for the purpose. In the Howe truss we have indicated the coun- ter braces, which are not theoretically required, by dotted lines; they are always introduced in practice, but are not required uidess it be to hold the angle blocks against which the main braces abut. The method by which the strains were calculated in Figs. 9 and 10 is too well known to need any explanation. In Fig. 8 each system of tri- angles was treated independently, and the load upon each panel, including the weight of panel itself, was supposed to be applied at the point of each panel division. This is* not strictly true in practice, since that portion of the load which is fi^ed will be partly sustained at the top of each division by the brace, but the error in effect is insigniti- cant. It may be interesting to notice that the effect of a change of load from the bottom to the top chord would be to trans- fer the amount of the minus strains with the signs changed to the compression members in each panel, and vice versa. The first thing that attracts attention in the examination of these figures is the great disparity between the strains upon corresponding mend)ers of the Isometrieal Truss and the other two; but in order to reduce these figures to a tangible shape it will be necessary to determine the actual amount of material required to resist the strains, in each system, and then determine the relative percentage. This may be effected with sufficient accuracy for our present purpose by adding the total strains having the same signs in each, and multiplying the result by the theoretic length of the members corresponding. Thus the total minus strain in the Isometrieal including counter svstem is 444 tons; assuming a divisor, for safety, of 5 tons per scjuare incli for wrought iron, this will re<|uire HS^^f^ square inches, which, multiplied by 23.6 feet, the lengtli of a diagonal, ■*^9*mi THE ISOMETRICAL TRUSS. 23 (^ '1M jiiid rlivided by 144, gives 14 ,«^ cubic feet of wrongbt iron for one-half the bridge In tlie same way the amount of iron in the Howe trusts is found to be 17/jj cubic feet, and in the Whipple 20,»5 cubic feet. For the plus strains wood has been assumed as the resisting medium, and a corre- sponding divisor for safety of one-half ton per square inch. The results for compression are — Isometrical, 138 cubic feet. Howe, 258^5 Whipple, 166^^5 " By comparing these quantities with each other and ar- ranging them according to their relative position we have, calling the Isometrical 100 — Table I. &r Name. Rplatifefcon- omy of Iron. R»'lativ« econ- omy of wood. AvfrauB pcon- oroy. — Remarks. Isometrical Tru.ss, . . 100 100 100 Web sys- tem alone Whipple " . . U3 120 13U in each case. Howe «» . . 122^ 187 154/, Thus it appears that the Whipple involves an expendi- ture of 43 per cent, more iron than the Isonietrical in the web system, and 20 per cent, more wood, and the Howe truss 22| per cent, more iron and 87 per cent, more wood. It is possible that some slight variation of these figures might be obtained by a different arrangement of the num- ber of panels, but nothing in general practice will warrant a more favorable expectation for either the Whipple or Howe. Again alluding to the subject of ordinary limiting span, we find the end diagonals (Fig 8), subjected to a strain of 9') tons. Taking as before, a divisor of 5, this will require IH s(piare inches of cross-section, and as there are three rods in each panel, or six in all, we have for each 3.16 square inches, or 2" round iron. Figures 9 and 10 give for cor- responding ntembers 153 and 152 tons, which for the same divisor and number of rods, call for 2",Yo ''<>"'>d iron, #3 .•I. V (^f■ t^H '^y \^:AtiJir^.jL.'i. -Zi .S- JjTTJlS-y-lS m ''37 ■•■■■ r' Jl 24 THE ISOMETRICAL TRUSS. rii'i..,"".,-^ ._♦ >,#.^. -' 41 > -ft* «J ■•"■•/?■;" ■J' ■ ' '. "<>i 4i . '«** ! s I'l i 1 .4 which, as we have soon, is bevond the ordinary limit of size; while on the other hand, the Ison.etrical mityht be increased very considerably in span betore the limiting size of 2"h would be renuired. The above results are but the reduction to figures of the general conclusions arrived at in the earlier part of the discussion, and altboiigh they are confined to but three systems of wooden luidi^os, it is believed that sufficient evidence has been deduced to establish the claim of the Isometrical Tru-'ss to the first position among structures of this class. Although our remarks have been confined to two promi- nent types of wooden bridges, by way of comparison, the same conclusions will apply to structures in iron possessing the same general features. Thus, for example, in the dou- ble system of wrotiiiht iron columns and diaironals made use of bv Mr. Linville in his maarisons,namel3', the increased facilities for long spans, and in this respect the Isometrical has only a very slight advantage over tlie Linville t: tiss. For the purpose of instituting an economic comparisoti between these two bridges in iron, a Linville truss has been arranged and calculated, jis in Fig. 11, having the same general data as before mentioned; and one word in this regard, befor<' passing to a comparison of figures. In order to prevent a multiplicity of on this princifJle, have fully realized the most sanguine expectations of their pro- jectors, and the public generally, is no argument in favor of the theoretical accuracy of their design nor of the prac- tical economy of their Tonstruciion, but rather that they are the best applications of the principles upon whicli they are constructed to practice, and most certainly entitle their originator to a very high rank in the special branch of the profession to which he has devoted his attention. It is not by any means claimed for the Isometrieal Truss tliat it is a perfect cj .^ •^ .■ ».»•,!(« A , 1 I. 26 THE ISOMETRIOAL TRUSS. to any and every other now in use, as the writer is well aware, that in tlie almost infinite variety of eircumstances which the engineer is called upon to provide tor, the par- allel girder must frequently he laid aside for other and more appropriate appliances. But of this we feel compelled to insist that it is t!ie most practical application of the true theory of the action of forces in a parallel girder now in existence. Hitherto our comparative investigations liave been con- fined to the two questions, of economy of construction and adaptability to long spans, and in those we have shown the Isometrical Truss to be founded on sound principles. But there are other considerations involve<\ in the discus- sion, which are of great importance in arriving at a proper conclusion, and they may be generally expressed by the terms compensation and adjustability. . The iirst is understood to be the abilitvof a structure to adapt itself to the varying circumstances of a changeal)le temperature and load; and tlie second, relates to the facil- ities presented for restoring it to a normal condition, ren- dered necessary by the deteriorating influences of time. In regard to compensation, the very nature of the construction of the Isometrical Truss, involving as it does, equal equi- lateral triangles in all directions, involves perfect uniform- ity of action under either circumstance of change of tem- peratures! or load, hence there can be no distortion of figure within the limits of a maximum load, and the truss is therefore practically rigid. In orAr to illustrate the necessity of proper adjusting facilities in a truss, the following diagram shows the change which takes place in any one panel of a parallel girder, under the action of a load or the deterioratinir influence of time. A deflection below the normal position of the truss manifestly causes a compressive strain on the diagonals a, a, a, and a tensile strain upon h, b, b. So long as this strain does not exceed the elastic litnit of the material, these members will of themselves force a return to the normal immediately upon the removal of a loud, and sii»je "I*', m THE ISOMETRICAL TRUSS. 27 Fig. 12. they are placed in tlie direction in which the panel changes form, ihey will roqnire to exert less eft'ort in accomplishing the ohject than in any other position. For the same reason, if the truss, after a lapse of time, shall have hecome permanently deflected, we have but to screw up the diagonals h, 6, b, to secure entire adjustment. The manifest superiority of this system over others in- volving an adjustment through vertical members, is now apparent, since the best possible way to a-)[>ly force to b, h, 6, in • order to shorten them, is in their own line of action, and if it becomes necessary, as in the Howe truss, for example, to shorten the vertical rods in order to effect the same result, much greater expenditure of force is required, and this accounts tor the fact admitted l)y all practical biidge build- ers, that when a Ilowe truss becomes j)er- nuiJiently deflected below the horizontal, expensive false works are necessary to efl'ect a restoration. We shall have occa- sion to notice the truth of this observation in the discussion of some experiments up(in models. During the earlier consideration of some of the questions treated of in this paper, the writer was fortunate enough to secure the perusal of an ably-written analysis of four well-known iron bridges by Mr. 0. iShaler Smith; and feeling that the conclu- sions therein arrived at would warrant a more expended range of favorable comparison for the Iso- nietrical Truss, t Synop (^ III) has been transferred direct. And it is perhaps needless to add that .\[r. Smith's well-established reputation is a sutK- cient guarantee for the correctness of results thereiii con- tained. f? r •ft-{ 1*1' M iii 4^1 .. . ■'. >jdA^ ..jv ' ' ^'AsSsilL^i&L Aud^ 28 THE ISOMKTKKAr- TRUSS. o j= ■— I .' >' Vv "•* • ;.• .1 ■1, > f ■ >■*• ■>: V I I' :,!, ,^fc'..•, I ■ *f ■ , hi* h*^ :« "x c '- • E 5 ~ pai. to •• >■< :/. N3 :3 :S ^ c > w -3 0; J; i c V' — • s ■5 « -*^ ^^^ ^ -r Ih ■r •._ H c .2 r = E-t: i ." = ■« .^ -1 w .^ ^ H -3 :5 =3 tpTf 5—5 «er D 5 7. .t; it a >.^ y: s MM W 1 w ', 1 ,m^ ^> cc ^^ . — z ^ *s* cs "^ o 02 c l-H Z Z '^*-^ C) ^ ■*-» ** St fc- t- "" •.^ C — ^s H a H ■K ?- C 1 xt a; :: - O '/J ^ :; = .»,._ - w *- 5 5S rs .i '-f "^ W) i: ^ JB.»q-pB01 <>l ]l|i4MM Jo n»i)B|.u itii; t>ti'U<);-) i 'pttooijjud •ojd WB tiniijj JO a'iio -Vduj :l)HI noiB.) am III p.tinnN -h« piioi JO jqJtitw t|jgil>»j)s Uj l!.>U.i(| qiStii.ijjri gn|dins 'OOI) -B|iio|Bo joj palling -HH fHiij) JO mXpiw •SHII.ll \(\ .iii.KMi |H!j.))">n JO )i(]l!.i.\\ |«n).>v •mmi aqt .(q '•OJo*! )oii '-jv 'Ji>«oj 'a.»!djojq«!«.\\ 'Wrtii.i) q.>iiA ni ? 2 » B r = e ^ 1) — I. ->- -a j= X rt — ?» 'f — MM'* M .t »r — — N M • I — — — o — « -t «-- "^ C^ :C O , M t- -f O ,«C «C -t .rt I -*• T -»• M X XXX M 3» X S M = - X ?t r. X c> ?■) "M e-» I- 1— 1 4 ?l M M M M ^ X ifl 31 TT -^ *■« ■^ -f ^ ri X ^■l rt ^- ■ -T -r ;^ -r n M ?l * ri *c «c ■M M ■M r» ?1 fl e-i M O 00 -o y. "T •B HM ?> 5C tc X |BI JrtlBUl Jo JI|J(1,1 »« C -t -rt M ' " 1 X -r X •£ e* s-» M M as B C = —; rt u _ — «c H o = o o^c = s c« ^i' e-f IM* t— «-. »>• 1- IM M I-l M ~. •^ iC rt = M 5> i-1 lO Oi X >n CC 1- •o >« I'. M T» ■M r: . c M M «c '>\ M O e-i -^ "^ Cl e to •w » ^^ o •* Cm ^ R _ w •-» 1 rr e! a B M* "^ B m >. V JS ^ u 3 •jj S 6t S >% B « a 3_ H u i> c: B 9i •*J X 1) K «>» e> B s. a Wj X V 1. in •- B S c s k o H 3i .2 s « a a -.b E J: T _ B ~ St s 5.= 2 " »j — aT " — B ""I aj — S • — *^ w « " V •^ *■ M ^■^ So .5 s-5 i * ca 3 — J3 u '• - « J. t. 3> -*■ «C -r S M 31 THE I80METRICAL TRUSS. 29 The triangular truss mentioned in tins connection, is [(recisely the same as has been noticed in Fig. 6, and the marked su[»eriority of the Isometrical system over this ar- rangement has, we trust, been fully established. From our previous discussion of the provisions for coi^pensation and adjustment in this system, we feel compelled to assume the responsibility of placing it absolutely Hrst in every respect except, perhaps, its adaptability to all spans and positions; and in this particular an equality is claimed with the Fink. The original application of the Isometrical principle to iron bridge construction, is believed to have been made on the riiiladelplna and Reading Railroad, in the year 1863. A series of experimental trusses of GO feet clear span were constructed by the patentee in the sliops of the company at Pottstown, and afterwards put up at different points on the line, where they were subjected to the most severe trials ca- pable of being apjdicd in practice, and the results were in all cases highly satisfactory. These same bridges, it may I; .' remarked, are now carrying the heavy trade of this well- known railroad without the sliglitest indication of weak- ness or the expenditure of a dollar for adjustment. It was not, however, until the summer of 1865 that any attempt was nuide to a[»i)ly these principles to wooden bridge con- struction, but the results then obtained, although throuirh the medium of experiments upon models, were full and conclusive. A brief description of these experiments may not l»e uninteresting to our readers. The lirst model was constructed in the IMiiladelphia and Reading Railroad Company's shops, at I'ottstown, on the general design represei»ted in Fig. 7- It consisted of a single truss of :25 feet span, having its parts propoi'tioned upon a scale of J the size of a bridge 200 feet span, but no p^rovision was made lor a system of counterbracing. After placing the niodwd tirndy on its supports, and contining the upper chord so as to prevent buckle, a uniform load of castings* was aj»plied to the lower chord, of 3500 pounds, i)roducing a deflection of 1^ inches ; but in order to test the eliect of a want of counterbracing, the t..«,tings from one-half the mi m, m mf mi f >.» i 'hi- tj/a 4 !(•*■*• IM;. ■ - "^?. ^ ■ . » • y ' .■ ' ^" .>•■.;.. •..,■" •la-' ' ^•' ■'-!•■ "> ■-•••''4; .•■■V' ♦ ■ t'l • ■*• k^ ■, '■ * "i* . V f * • ,- . '',■■ '•■^■■■ I;* * ■ .•.» ■ , . I* 30 THE ISOMETRICAL TRUSS. truss were rernoved'jnul it was noticed that tlic braces near- est tlie middle hecame lo(3se. In order to remedy tliis evil, the entire load was removed and iron tie-rods intro- duced alongside of the braces, and in onler to make room for them,, their entire section had to be cut out of the chords, top and bottom. After this had been accom- plished, anf each panel, while the Isometrical retained three, thus giving it a superiority of one-third the amount of brace tind>er; in all other re- spects both trusses were alike, so far as the nature of their designs would permit. Although this single exception is to be regretted, it is not believed that the vast difference between the results obtained can be attributed, even in part, to it, since the tensile system was the same in each, and the Howe truss had not liad ita chord section de- creased one-ninth by the ad ■ ■ • * ••' represent; and for that reason it is not jn'ojiosed to draw any inferences in favor of the ori«j^inal (U^sign for two hun- dred feet span of wliicli this truss is a model ; but viewing these results in tlie liglit of a eonipnrison of systems, tlie evidence is undoubtedly in iiivor of the triangular web. This after all is but the veritication by experiment of the general conclusions arrived at in tlie earlier discussion of the subject, and the later numerical results obtained by a comparison of strains. "We therefore feel qualitied in passing to general conclu- sions eminently favorable to the Isometrical Truss. And in submitting these views to members o/ the profession who may }ionor them with a careful inspection, we desire again to repeat that if they cs'Ji be demonstrated to l>e clearly j?t variance with correct reasiuiing, we shall be only too glad to protit by the criticism. GENERAL CONCLUSIONS. The Isometrical Truss, founded as it is upon the most practically accurate web system, is necessarily the most economical form of parallel girder now in use, either of wood or iron. And from a coiiii>arison of four well-known iron bridges as instituted by V. Shaler Smith, and sub- mitted in Table III, this remark nniv be extended to cover the best examples of suspension trusses. Next to economy of construction comes the question of general availability for all spans and positions. We have not thought it nee .-sary to discuss at any great length the comparative me'it:' of the different trusses in this respect, since it involves a-.i amount of detail which would be any- thing but interesting, and at best quite unnecessa'y to the formation of a generally accurate oi»inion; and after all the J great desideratum of a perfect web carries with it advan- tages which are etjually applicable to short or long spans through or deck bridires. COMPENSATION AND AD.JUSTMENT. As to the elements of compensation and adjustment in the Isometrical Truss, we have seen that owing to the ecjui- Bj.;' tj THE I8OMETRI0AL TRUS8. 83 lateral arrangements of its internal members, no possible distortion of figure can take place under a change of tem- perature; hence the relative strength of the structure always remains the same. In the iron truss the only effect of a high temperature is to increase the total length of the chords; but in the application of the principle to wood, since one of the diagonals in eacli panel must be iron and the other wood, the effect of an expansion of the iron is simply to lighten the original camber of the truss to a sufficient extent to bring the brace again into play, and the rigidity of the structure remains unchanged. In fact this is the only case in which compensation is rendered necessary, since in the iron truss, supposing wrought iron to be used throughout, all parts expand equally, and consequently no . c'aml)or is lost. Adjustments are rendered necessary when, by reason of the shrinkage of a portion of the material, as in a wooden bridge, or from defects in workmanship, the structure is deflected below the horizontal ; and to effect this the diag- onals pointing downward towards the centre must be short- ened, since we have seen that a lengthening of these mem- bers by a rise in temperature will cause depression in a woodon structure, hence reversing the operation will cer- tainly restore the proper elevation. It would scarcely seem necessary, after the general discussion of the action of strains in the web, to devote more time to the question of adjustment, but to add the evidence of experiment we will mention that the model Isometrical Truss was subjected to a variety of trials, by the writer, with a view to determine the facility with which it might be restored to its proper camber, after a supposed shrinkage. All the diagonals were brought to a proper bearing, and the camber of the bottom chord noted carefully, afterwards the entire truss >|j was caused to assume a heavier camber by simply screw- ing up the diagomil rods, heretofore mentioned, with the thumb and finger, and it was easily seen that any desired elevation within practical limits could be secured by this means. On the other hand, the application of a wrench, with great force, to the rods of the model Howe, alongside. 'mi W' 1« :;*! -> • L «i ■■*■; • . i~.r. ^ ,'.«, 34 THE ISOMETRICAL TRUSS. . «?' had no other eftect than to sink the bolt heads and their washers into the chords, and no change of camber was pro- duced Avhatever. The large plate accompanying this paper represents one span of a single-track through bridge, now being built by the writer, for the Perkiomen Railroad Company, over the Schuylkill River near PhaMiixville. All the parts are pro- portioned to resist the strains, as represented in Fig. 8, and particularly in the braces, is a large margin added above the requirements for safety. It will also be noticed that the connections over the pier and the anchorage involve a considerable expenditure of materia] beyond what is gen- erally called for in the best railroad bridges. This has been added as a measure of safety, and while in all ordinary cases this item of expense might be dispensed with, the increased feeling of safety in the reduction of the effective span twenty feet, and the perfect immunity from the action of wind, will fully compensate for the increased expendi- ture of material. The several isometrical sketches will serve to illustrate the arrangement of braces and ties, counter braces and counter ties. The angle block is believed to combine ad- vantages over either the original oak block or the later cast-iron Howe, in insuring a uniform bearing over the entire chord without the danger of fracture, as in cast-iron, or of rapid decay, as in the simple oak bearing. We have proposed locust as the best material for the filling of this block, but tliere are many other species of wood which would answer almost as well ; circumstances of conve- nience had much to do with its selection in this case. The connection of the diagonal rods with the bottom chord by means of a 2" pin directly through the centre of the chord, admits of a considerable saving of material hi the bearings, and the application is only rendered possible by the reduced strain involved in this system. It would, of course, be out of the question to admit of attachments of this kind in either the Howe or Whipple, unless the chords were considerably increased beyond the requirements for horizontal strain. A similar remark migh*: be extended ^■HHIIHH THE I80METR1CAL TRUH8. 35 to the bearing block connections on the top and bottom chords. In the original Howe bearing block, it was found that the material of the chord was eventually crushed and destroyed at the plane of contact, owing, of course, to the enormous strain on the braces and rods near the points of support ; but as we have seen that the Isometrical princi- ple permits a reduction of 50 per cent, of this strain, it will readily be admitted that this element of safety fully com- pensates for all such disturbing causes. In packing the chords with oak blocks and keys, the writer is well a vare that there is great liability to rot at the planes of junction, by reason of the deleterious acids contained in the natural oak, but being in a position to test the ettcct of the Bur- uottizing process, In neutralizing this destructive tendency, the experiment is made, with an eye to its future applica- tioii to such parts of the truss as may not be subjected to transverse strain. But inasmuch as the annexed plan of bridge was intended merely as an example of tlie applica- tion of the principles involved in the Isometrical system, it will be unnecessary to examine further into practical details, which, after all, must be left to the test of extended experience, and which cannot affect the general conclu- pions already siibmitted. mi m i- J: !.•♦■.*; *^M '..♦1 I! Ui