IRLF 3Dfi GIFT OF The Engineer in Field and Office New Ideas for Securing Uncom- monly Quick, Accurate and Economical Results. Reprinted from the Engineering News-Record. ENGINEERING NEWS-RECORD Tenth Avenue at Thirty -Sixth Street New York 1918 Copyright 1918 McGraw-Hill Publishing Co., Inc. PREFACE Civil Engineering is an art of which it can never be said that it "stands still." It keeps pace with human needs and human necessities are ever expanding. Every year every month almost every week witnesses fresh demands for more conveniences, quicker transportation, more effec- tive safeguards for public health and comfort. In meeting these demands, civil engineers are, happily, most efficient. The profession is continually making new discoveries putting forth new inventions suggesting new applications of former theories devising more effective designs improving upon the old methods. Recognizing its obligations to the field it serves, the Enginering News-Record not only opens its columns freely to the authors and originators of these new ideas so that their experiences may be passed along for the benefit of others, but is continually seeking out the new and helpful. These advances thus become incorporated into the common prac- tice long before they get into the books or become part of the teachings of the schools. The best of the new ideas which have recently appeared in the News-Record have been collected for this volume. The editors feel that many of them are too valuable to remain hidden in the files of a periodical, even of a journal whose back numbers are so frequently consulted as are those of the Engineering News-Record. They are confident that in this accessible and permanent form this new material will be helpful to every engineer who wishes his work to be abreast of the times. It is Civil Engineering brought up to date. For the most part, in the case of the contributed articles we have let the engineers tell their own story. But little editorial responsibility has been assumed, beyond the elimination of occa- sional personal and local references in the articles as they originally appeared in the columns of the Engineering News-Record. The mere fact that they were admitted to these columns is sufficient guarantee of their good faith and of the editors' judgment of their worth as contributions to current practice. 380101 CONTENTS I Of Value to the Highway Engineer II Interesting Lights on Building Construction-Theory, Design and Methods III Some Capital Things in Foundation Work IV New Ideas in Designing and Building Bridges and Dams V Helps for Municipal and County Engineers VI Present-Day Water-Works Construction, Maintenance, Op- eration and Repairs VII Interesting Solutions of Problems in Sewer Construction VIII What Engineers are Doing in the Fields of Flood Control, Irrigation and Hydraulics IX Practical Pointers for Railway Civil Engineers X Helpful Suggestions from Recent Concrete Construction XI Results of Recent Tests in the Laboratory XII Practical Hints for the Surveyor XIII Draftsmen's Kinks XIV Simple and Effective Methods for Filing Engineering Data The Engineer in Field and Office Of Value to the Highway Engineer Traction Is a Straight-Line Function of Tire Width It has been recognized for many years that the narrow steel tire is enormously destructive to earth and gravel roads. The formation of a rut begins practically with the passage of the first vehicle improperly equipped. The U. S. Department of Agriculture has shown that the effort necessary to draw a certain load is an inverse ^ ** *** ^ ** ^ ^ ^ ^ ^ ^ x <^ ' 200 300 400 SOO 600 700 800 SOL RDunds \\feight (Gross Load) Per lineal inch width oFftre. Graph Shows Uniformity of Test Conditions function of the tire width, and in this manner forwards another argument against excessive load concentrations on the road surface. The accompanying curve is plotted from data obtained under similar conditions of moisture and atmosphere. The earth road was plowed, graded and rolled between each series of tests, and its condition throughout the investigation was probably as uniform as the nature of the test warranted. A gross wagon load of 5000 Ib. was maintained, the tire width being varied from U in. to 6 in. The Engineer in Field and Office The unit draft decreases directly as the weight per inch width of tire decreases, until a weight of 250 Ib. per inch of tire is reached, as is indicated by the curve. The draft for a 6-in. tire is larger than that for a 5-in. tire. This was uniformly true, and indicates that there is not only nothing gained by increasing the width beyond a certain point, but there may be a positive disad- vantage in so doing. Handling Hot Oil on Maintenance BY H. M. LUKENS The Los Angeles County Highway Department has a unique system of heating and distributing oil in the maintenance of its paved ways. The oil is too cold for use when it reaches the job in trucks from the main depot. It is poured into buckets that have previously been arranged in two rows with another row on top. A portable distillate and water burner has proved very efficient. As many as 80 buckets of oil can be heated to 500 F. in about 15 min. About 15 or 20 gal. of oil at this temperature is placed in the hand oiler for direct application to the patch or street. The hand oiler is made of sheet iron, with a boiler about 2$ ft. in diameter and about 3i ft. high. Inside this boiler hangs a cast-iron pot of about 20 gal. capacity, the intervening space between the pot and the boiler being used as a firebox. This apparatus is so suspended on two iron wheels that the center of gravity is low enough to keep the boiler upright. On one side is riveted an iron handle with which to move it, and opposite and on top of the boiler is the hand crank for the oil pump, which is fastened to the bottom of the oil pot. The pump, of the rotary type, is driven by sprocket and chain from the hand crank and is capable of developing 6 to 10 Ib. pressure per square inch. To the pump is attached 20 ft. of f-in. metallic hose with a piece of f-in. pipe 3i ft. long at the discharge end. This end is plugged so that under pressure a spray is formed which may be regulated by a throttle. The maximum capacity of the machine is about 125 gal. per hour. It has been found from experience that best results are had from heating the oil to the desired temperature in the buckets and using it in the hand oiler without any additional fire in the firebox. Fire in the oiler causes the bottom of the oil pot to cake up, which in time breaks loose and clogs the pump. Highway Engineering Concealed Wood Strips for Transverse Joints Many of the troubles attending the use of transverse joints in concrete pavements can be avoided by installing at the bottom of the slab a I -in. wooden strip of about half the depth of the concrete. In this manner planes of weakness where transverse cracking may concentrate, are definitely located. Four joints of this type were installed in 1915, and withstood the following winter without cracking, although fine hair cracks appeared in June, 1916. These cracks were very narrow and followed a zigzag line around the individual stones of the concrete directly over the wooden strip. The zigzag cracks nowhere departed more than 2 in. from a straight line. These cracks were exposed to traffic all summer, and no spalling of any kind has been noticed. The result was so satisfactory that in 1916 10 consecutive joints of the same type were installed on each of two roads. These have gone through one winter and are now in excellent shape. Fine hair cracks have appeared over most of the joints, but the fact that the pavement has not yet cracked over a few indicates that the slabs might have been made somewhat longer than 30 ft. in this particular locality. In no case has the pavement cracked, except over the wooden strips. The advantages obtained by this method over other types of joints are: (1) No spalling along the joints; (2) no interference with scr ceding or floating, and a much smoother pavement; (3) the joints are cheaper and easier to install, the wood may be of cheap material and need not be creosoted; (4) better appearance of the road. The advantages of this method over nol joints at all are that by choosing the proper length of slab to correspond to the width of pavement and to climatic conditions the cracks will be minimized and will run square across the pavement at regular intervals instead of occurring haphazard and running in every direction. It is absolutely necessary that the top of the joint be kept below the surface of the slab a sufficient distance to give body enough to the concrete over it to prevent spalling. This minimum depth has not yet been definitely determined, but from observations already made, a distance of 2i in. near the edges and 3 in. near the center of the pavement seems to work out very well. Should this distance be much greater, it probably would necessitate decreas- ing the length of the slab in order to insure the cracking at the joints only. The Engineer in Field and Office A simple way to prevent the wooden joints floating up through the concrete is to drive short pegs at intervals of about 7 ft. across the road and nail through the strip into the pegs. Bent wires hooked over the joint would answer the same purpose. The strips should be installed by means of a taper installing board in the usual manner. Exact Wear of Concrete Pavements Measured BY A. N. JOHNSON Consulting Highway Engineer, Portland Cement Association, Chicago The usual methods that have been devised for measuring the wear of macadam roads are open to considerable error, aside from usually being rather clumsy, but the method here proposed over- comes many difficulties and is also extremely simple and convenient. A drill hole i in. in diameter is made in the pavement. In the case of a concrete pavement it may be 2 or 2 in. in depth. In the <- .......... -r ..... & MachmeFvced Steel -Flat Sea te 6"- ........... * sFine Thread Surface fofPare/Tjent efe- 5 "%"Diarr>. Copper P/ug ^CementGrout With This Device Wear of Concrete Can Be Easily and Accurately Measured bottom of the drill hole is set a copper plug about I in. long, with its top made semi-spherical. The plug is embedded in the hole with cement paste. A short steel scale divided into inches and hundreds, small enough to be placed in the drill hole, is provided. At its end is a small button with conical undersurface, which is to rest on the spherical end of the copper plug. The hole is filled with a wood plug, left flush with the surface, which is to be removed when readings are to be made. Highway Engineering A convenient method of measuring is to have a U-shaped handle, which will span about 6 in., having flat ends on the legs. Drawn between the supports is a tightly stretched thin thread or wire 1 in. from the surface of the pavement. This is to be held so that the thread will be over the center of the drill hole. The flat ends of the legs rest on the pavement surface and the measure- ments are made to the thread. The thread may first be placed longitudinally with respect to the pavement, then transversely, and in as many other positions as desired. Grasses To Protect Road Embankments BY W. C. BUETOW Division Engineer, Wisconsin Highway Commission, La Force, Wis. Much damage to embankments and slopes that cannot be prevented even by the provision of proper drainage and keeping cattle from trampling the slopes can be taken care of by the planting of well-chosen grass, which should be properly cared for until a mat or blanket has formed. Grasses do not form a mat or blanket very quickly, due to the variance of the climatic conditions. The action of frost is the most disturbing factor, and on account of the depth of its penetration the upper stratum of the bank is kept in a loosened state that is very susceptible to the spring rains. When selecting grasses to be planted on the slopes, it should be kept in mind, therefore, that a grass with a good system of roots is to be used, as well as one able to withstand the hot rays of the sun. Grasses that will grow on a sunny bank probably will not take root and form a mat on the shaded slopes, in which case a different kind of grass is necessary. The protection to the bank lies almost wholly in the ability of the roots to interweave so minutely as to form a system to reinforce the surface stratum. Hungarian brome grass, Canadian blue grass, fescue grasses and Western wheat grass are good to plant. In many cases sod is handy to get and, when properly cut and placed, offers very good protection. The sods must be held in place by wooden pegs until the grass has taken root. Seeding and planting are equally successful when conducted properly. The surface in each case must be prepared in order to make it at all conducive to a good growth. The seed cannot be merely distributed on the slope and left with the hope that it will catch. If necessary, the ground should be manured before any seeding is done and if a sandy slope is treated, the surface must be The Engineer in Field and Office protected from the wind with straw or brush until the grasses are able to withstand the elements. Different kinds of soil require a certain kind of grass or a combination of several kinds. For shaded places a combination of Kentucky blue grass, wood meadow grass, crested dog's tail and various leaved fescues are well adapted. Clay soils can be treated with a mixture of Kentucky blue grass, English rye and fancy red top. The rye grass gives an early quick result, the red top makes a bottom grass, and the blue grass is the permanent feature. Sandy soils require a quickly growing binding grass that will withstand the drought. Creeping bent, Rhode Island bent and fine-leaved fescue are grasses that answer this purpose. Western wheat grass has unusual ability to grow on the sunny side of an embankment. Creeping bent, Canada and Kentucky blue grass and crested dog's tail are quick growing, deep-rooting grasses that will bind the soil until such time as the more permanent grasses are in possession. Sweet clover and creeping honeysuckle are recommended for planting in cuts. Dump Road Drags Before Pulling Across Railroad Crossings Men operating road drags should dump their drags or unhook one end of the drag chain and pull the drag endwise whenever a railroad crossing must be crossed, according to the "Bulletin" of the Iowa State Highway Commission. Railroad men are complaining that this precaution is not being taken. They have appealed to the commission to do all in its power to get dragmen to watch this point. Pulling the drag across the tracks in the usual manner leaves the space between the rails and the planking on each side filled with dirt and small stones. Each passing train packs this material down, and the next dragging fills the space again. This continues with every dragging ; and if a good-sized hard hand-stone becomes packed in this space, there is the best possible situation developed for a derailment. Derailments of hand cars and motor cars are of almost daily occurrence because they are not heavy enough to crush the dirt and stone sufficiently to allow the wheels to retain a safe hold on the rails. There have been many derailments of trains traceable to this cause, but so far, no serious wrecks. Highway Engineering A New Subgrader Design Instead of the usual scraper blade at right angles to the center line of the road, a new subgrader used in California has four pairs of plow-shaped blades arranged to balance the thrust. These are set at an angle with the center line, and as a result the tendency to pile up earth ahead of the grader is avoided. This feature makes possible more work of a satisfactory character with fewer laborers than is obtainable from the ordinary type of subgrader. A feature of the new device is the quadrant and lever regulation of the blade, whereby the height of the blade can be adjusted without the use of special tools. This arrangement also permits raising the blades until the scraper rests on a central circular plate, or turntable, in which position it can be easily swung parallel to the center line of the road to allow space for passing vehicles, or so that it may be trundled off the graded roadbed. Angle irons are provided as guides to keep the wheels on the header. Since the blades shear through the material rather than pile it up ahead, depressions in the subgrade are avoided to a great extent. The grade may be controlled to a nicety on account of the clean action of the plow blade, and thus a considerable saving in concrete is effected. It is notable that the setting of the blade at acute angles has reduced the tractive effort below that required by the ordinary grader. The most effective work has been done when the grader is hauled by cables running from its extreme ends to the tractor, which is kept about 30 ft. in advance. Earth-Road Maintenance by s Contract BY C. S. BENNETT State Engineer and Inspector, Greenup, Ky. After considering various methods of maintenance for the new roads of Greenup County, Ky., it was decided to adopt the contract method. The specifications as finally drawn up embody provisions for dragging, cleaning ditches and culverts, cutting weeds on the county's right-of-way, maintaining a crown, etc. Bids were then asked for and work on each section was let to the lowest responsible bidder. Mileage was considered the basis for payment, while slides and washouts were paid for by the cubic yard. The work was usually let in five-mile sections. A road drag, two drag scrapers, one wheel scraper, shovels, and mattocks, were furnished to each contractor, being charged against him with the provision that at the expiration of his contract he account for all tools supplied to him. In addition, two large blade 8 The Engineer in Field and Office road graders were purchased by the county, being lent to the various contractors as they were needed in shaping up the roads in the spring season. A contractor was required to own at least one team. The work was under the supervision of the county road engineer, who notified the contractors when they were to drag the roads and when to cut the weeds on the county's right-of-way. He also issued to them instructions as to other details of the work. Payments were made quarterly, the payments being so distributed as to make a maximum payment fall at a time when the greatest amount of maintenance was necessary. Thus, for instance, a contract dated Jan. 1, 1916, stated that payments would be made to the contractors as follows: 20% Apr. 1; 40% July 1; 15% Sept. 1 and 25% Dec. 31. The maximum payment, 40%, was made at the end of the spring season, during which time the greatest amount of work is necessary to keep the roads in condition. In addition, each contractor was required to give bond for the faithful perform- ance of the work called for in his contract. Strict adherence to the terms of the contract, especially in regard to dragging the roads, was insisted upon. The cost of maintenance on seven sections of road for the year 1916 is as follows: Length, Contract No. Miles 45 Dl 4 45 C 1 5 45 Gl 5 45 G 2 5 45 E 1 3 45 A 1 5 45 G 3 3 From a comparison made with similar work done in other counties under various other methods, the writer is led to believe that the contract method, properly looked after, is the cheapest method of continuous maintenance for earth roads. Peaked Subgrade Provides Better Underdrainage A concrete or brick slab laid with a flat bottom upon a flat subgrade affords no possibility for underdrainage. As a result, seepage through the pavement at various points such as expansion joints, along car rails, and through semiporous spots keeps the subgrade moist and affords a good chance for frost upheavals with the resultant cracking of the rigid surface. While there is but slight importance attached to such seepage if the construction is on a sandy or porous soil, a clay or heavy soil brings about conditions which cannot be disregarded. Bid per Bid per Bid per Total Cost Milerper Maintenance Cubic Yard for Earth Cubic Yard for Rock Mife $45.00 0.35 0.40 $48.50 56.00 0.37 0.37 61.05 40.00 0,24 0.35 40.35 60.00 0.30 0.30 76.22 53.00 0.35 0.40 53.00 100.00 0.20 0.50 138.00 75.00 0.35 0.60 75.00 Highway Engineering 9 The ordinary precaution of laying a slab with a crowned bottom upon a crowned subgrade does not entirely alleviate these conditions, as the slab is laid directly upon the heavy soil, filling every crevice and blocking a free movement of the water from under the crust. ^ * on each Side offibacf c*ndfill.edwt+h crushed Stone or Grave/ Shpe of -Side Drains %'per-fl; Peaked Subgrade Brought Level with Sand or Gravel Affords Best Possible Drainage The newer specifications, which require 2 or 3 in. of sand to be laid upon the flat subgrade, and upon this sand bed to lay a flat-bottomed slab, are undoubtedly an improvement, but it is questionable whether this method will prove to be entirely sufficient. The subgrade should not be crowned, but peaked, as indicated in the illustration. On this a sand covering about 2 or 3 in. in thickness at the center can be well rolled to a flat surface. This will allow a free movement of the water between the bottom of the slab and the top of the impervious subgrade and should afford adequate protection against the troubles outlined in the foregoing. Oiled Shoulders Resist Erosion A mountain road in San Bernardino County, Calif., with a maximum grade of 5% presented a serious problem in the protection (TANQGNTS) (ctmvtof Oiled Road Shoulders, San Bernardino County, California of high embankments from the wash of storm water running off the paved surface. The problem would have been simple if funds 10 The Engineer in Field and Office had been available for concrete curbs and gutters. Instead, the following plan, illustrated in the accompanying cross-sections, was adopted: The shoulders for 2 ft. in width were raised from 6 to 8 in. by filling with the adjacent gravelly soil, and were rounded off to easy curves. The shoulders were then given two coats of 75% asphaltic oil, each coat being properly sanded. Cutouts for the water were provided at safely placed points and at culverts. This method of construction formed an earth curb, and the pavement acted as the gutter. The oiled shoulders have a sufficiently hard surface to resist erosion and also serve to protect the edge of the macadam from breaking down under wheel traffic. There was a considerable saving over standard practice, the cost being only about 2c. per lin.ft. of shoulder. J. S. Bright, Jr., Engineer, San Bernardino County Highway Commission, San Bernardino, Calif. Curb with Integral Expansion Joints BY E. E. KIRKPATRICK City Engineer, Bartlesville, Okla. Better quality of work, better and more positive expansion joints, less risk in construction, better protection against drying, elimination of the danger of mortar scaling, elimination of the dam- age caused by removing templets and forms while the work is green, and a better top facing owing to the fact that it is placed immediately after the body concrete has been deposited, are the advantages derived from the new specification for concrete curbs recently adopted by Bartlesville, Oklahoma. The curb, of the di- mensions shown in the illustration, is blocked off into sections not exceeding 7 ft. in length by transverse expansion strips not less than -fa in. nor more than T % in. in thickness. These strips, placed before the concrete is deposited, are composed of asphalt-roofing sheets, asphaltic felt or other elastic asphaltic material cut to the form of the cross section of the curb. Highway Engineering 11 The concrete is spaded thoroughly, the expansion strips being held in place by thin boards or metal strips that are raised as the concreting progresses. The mortar top dressing is placed as soon as the form is filled. Dressed lumber is used for forms, and the concrete is mixed to such consistency that with proper spading a perfect face is obtained. Save Trouble by Using Well-Built Forms As the engineer watches the ordinary contractor set his side forms for concrete roadwork, he easily realizes that not enough consideration is given to their construction and grade layout facts _ *"->.gxf' I &= <~6>\ wood Form notched crs FI0.4 Stands Much Rough Handling which later on are sure to cause much extra trouble and loss of time. The forms should not of course be built to a degree of over- nicety, but should be so constructed as to prevent their becoming 12 The Engineer in Field and Office easily bent, warped or worn out of shape. Another feature often- times overlooked is the desirability of staggering the joints, thus preventing an undulation forming across the road in case the joints are out of grade. To obtain a method of easily supporting side forms for grade, 2 x 2-in. wooden stakes are driven to the proper elevation of the bottom of the form; and they, in conjunc- tion with alignment stakes which are not shown, and together with the method of staggering the joints, make a satisfactory arrange- ment. Figs. 2 and 3 show details of wooden side-form construction, Fig. 2 showing the angle iron extending 6 in. beyond the end to insure that adjacent forms will be maintained at the same elevation. Fig. 3 shows a modified design that prevents the angle iron from becoming easily bent when the forms are roughly handled, while Fig. 4 shows a side form made entirely of wood. Drainage Where Highway Parallels Railroad BY C. S. BENNETT Highway Engineer, Greenup County, Kentucky In constructing a section of a state-aid road in Greenup County, Kentucky, the highway paralleled the railroad for a distance of one-half mile. The railroad fill is 8 ft. higher than the road surface gpS^Wpsp WTwvmgi '^TWvv.-t <-.* VillllllllllllllHIIIIIIIIIIIIIIHIMIIMMllUMIIIIllllllllilinHllliliiiiiiiinnilMllliilHHIII'HIIIinilllfJIIIiilljilllllllllll/llllMII \ Culver i Pipe \ ^llillllilllilliiilllilllllliiniiiiNhiiNMnl'nilllliiiiniiiiliHIiluviiiii.iiiiiiiiiMiiiiiiiiiiiiiiijiiiiiiiiuaiiiiiniiiiiiiiiiiiMMiiiiniii) Drop Inlet Takes Care of Drainage Where Railroad Parallels Highway for about half of this distance. As all culverts extended through both fills, it was decided to adopt the inlet method used on paved roads and streets. Excavations were made in the ditch line over the existing pipe culverts, all of which were 48 in. in diameter and of corrugated iron, and a section was sawed out of the top of Highway Engineering 13 the pipe. Short sections of 12-in. corrugated culvert pipe were then placed on end, and fastened to the large pipe by means of an easily made metal collar through which several bolts were run. The top of the riser pipe was provided with a concrete apron 6 in. thick, the width of the bottom of the ditch, and extending about 3 ft. along the ditch. A few small reinforcing bars were inserted in the green concrete, and formed an effective grating which excluded foreign materials. Unusual Road Drainage Lincoln Highway, east from Wooster, Ohio, at this point dips from a comparatively level stretch to a grade of some 6 to 8%. Across the ridge where the road begins to dip is an intersecting road at right angles. At this point there was a culvert pipe on both Unusual Drainage, Lincoln Highway, Ohio sides, while open ditches were provided for drainage of the side hill and to carry considerable accumulated water from the more level place above. 14 The Engineer in Field and Office To avoid scouring on the deep side ditches on the improved road, about 300 ft. of 30-in. corrugated-iron culvert pipe extending from the crest of the hill to a creek at the foot of the hill has been installed. Apparently it is not intended to be covered. Drainage water from the level stretch on the opposite side of the road passes through a culvert pipe at the crest and then is conducted diagonally across the roadbed into the large culvert pipe by a segmental cast- iron culvert pipe. Stone Header Prevents Failures in Paving Baltimore, Md., has adopted a detail of pavement construction to prevent failures at such points where a hard paving material like granite block, brick or wood block adjoins a softer material like asphalt. This detail is as follows : A well-jointed stone header, 12 in. or more in depth and 4 or 5 in. in width, is constructed across ^.-Wearing Surface ... vitrified Block Sunken Stone Header Prevents Failures in Paving the street parallel to and 14 in. below the finished surface. The concrete base is placed on one side of the header for the block pavement, and on the other side for the asphalt pavement. The blocks are laid tight against the header, which in case of a 3i-in. vitrified block gives a 2-in. bearing. When the asphalt is laid, the binder course is brought flush with the header; and the topping or wearing surface is laid over the header to the finished contour of the street, thus leaving the header 14 in. below and entirely hidden from view. Road Surface Affects Tractive Effort A series of tests has just been completed in California which was planned to determine the tractive effort required to move vehicles over various types of roads. Tests of a similar character Highway Engineering 15 which were previously made, notably by the U. S. Office of Public Roads and Rural Engineering, afforded no data suited to California conditions or which could be used as the starting point for the further series of tests in tire wear and gasoline consumption which is contemplated. 100- o- TEST NO. 50, CONCRETE 83lb. TESTN0.46. OIL MACADAM REVERSE (DOWN GRADE) Dynamometer Curves Vary Much In planning the tests, it was decided to concentrate effort on securing constant conditions for a short length of time during which measurements could be made, rather than to make long runs and average the results. This policy was considered the best where experiments were to be made on the open highways and on several types of roads. One vehicle was therefore equipped and moved at the same speed over the various road surfaces to be compared. For this service a standard farm wagon was selected, equipped with steel axles of equal length and 38- and 46-in. wheels in front and rear, respectively, all wheels having 4-in. tires. The gross load was 3 tons, consisting of rice in sacks. The speed was kept very close to 2.4 miles per hour, and tests were run on level grades or, 16 The Engineer in Field and Office where slight grades were unavoidable, test runs were made in both directions and results averaged. At the time of the tests the usual sunny summer weather obtained, with maximum temperature of 105. The tractive effort was measured by a small dynamometer attached to the tongue of the wagon, which recorded the momentary pull at all times during the test runs and the total pull for the test period. The measurement of distance is accomplished by unwinding a cord of definite length from a drum within the device. In a 50-ft. run an error of 6 in. in cord length, due to stretching or any other cause, would introduce an error of only 1% in the result. The record of tractive effort is made by the pencil actuated by a lever attached to the strong spring compressed by the pull. The integrating device operates on the planimeter principle, but as the measuring wheel makes one revolution for each 4 in. and can be read to one-thousandth of a revolution, the instrument is 10 to 20 times as sensitive as the ordinary planimeter. RESULTS FAVOR UNSURFACED CONCRETE As shown by the accompanying table and curves, the resistance encountered on oiled surfaces was considerably more than on TYPICAL, RESULTS FOB VARIOUS SURFACES Tract. Resist. Test No. Kind of Road Condition of Road Location Total Per Ton 29-30-31 Concrete (unsurfaced) . . Smooth, excellent Near Davis 83.0 27.6 * 11-1 2 Concrete (unsurfaced) . . Smooth, excellent Near Davis 90.0 30.0 26-27-28 Concrete (%-in. surface as phaltic oil and screenings) ... Smooth, excellent Near Davis 147.6 49.2 13-14 Concrete ( % -in. surface as- phaltic oil and screenings) ... Smooth, excellent Near Davis 155.0 51.6 9-10 Macadam (water-bound). Smooth, excellent Near Davis 193.0 64.3 22-23 Topeka on con- crete) Smooth, excellent Near Davis 205.5 68.5 Gravel Compact, good condition.. Near Davis 225.0 75.0 t45-48 Oil macadam ... Good, new Near Sacramento 234.5 78.2 J46 47 Oil macadam .. . Good, new Near Sacramento 244,0 81.3 33 Gravel Packed, in good condition. . Near Davis 247.0 82.3 18-19-20 Topeka on plank Good condition, soft, wagon left marks On Causeway near Davis . . 265.0 88.3 34 Earth road .... Firm, 1%-in. fine loose dust Near Davis 276.0 92.0 24-25 Topeka on plank Good condition, but soft. . . Near Davis 278.0 92.6 1-2-5 Earth road Dust % to 2 in Near Davis 298.0 99.3 3-4 Earth Mud, stiff, firm underneath. Near Davis 654.0 218.0 6-7 Gravel Loose, not packed Near Davis 789.0 263.0 * Graphic record indicates that the load was being accelerated when test was started, t Drawn with motor truck at 2% miles per hour. J Drawn with motor truck at 5 miles per hour. concrete. It was pointed out that this difference would be less at Highway Engineering 17 lower temperature, but the tests were made under normal conditions in the central California valleys. Curves for Tests Nos. 2, 3 and 7 The supervision of the tests has been in the hands of Prof. J. B. Davidson, of the University of California, who is the inventor of the dynamometer. Side-Hill Fill Held by Trees In the construction of one of the roads in the northern part of New York State the side-hill fill has to be made over a wooded area. The usual procedure in such cases is to cut off the timber, leaving the ground clear before beginning to make the fill. In this case it has been decided to leave the trees standing on the supposition that by so doing the fill will be less apt to slip. The only clearing of trees which will be allowed will be in the space included between shoulder lines. 18 The Engineer in Field and Office Pavement Along Rails Should Be Lowered BY E. W. WENDELL Assistant Engineer, New York State Highway Department, Albany All pavements, within the line of possible contact, should be laid below the plane of the tread of a car wheel. Consider the case of a T- rail such as is shown in the illustration though it should be borne in mind that the above rule holds for any other type of rail. The rail head is 2i in. wide, while the wheel tread is 3 in., meas- ured from the gage line. Hence, even without any wear in the tread, if the pavement is flush with the rail top the tread will ride in contact with the paving surface for the width of the overhang. But the rail head wears, and not in- frequently the tread grooves, so that the over- hang is below the head on tangent track. While perhaps the railway company should be pro- hibited from using such equipment, it must be recognized that such cars are owned and oper- ated; and as long as an adverse condition exists, engineers should meet it with proper design. The speed of cars being generally regulated by law to a certain safe maximum, there is no necessity for elevating the outer rail on curves to an amount that will distort the cross-section of the street and spoil the section. The illus- tration shows a 2-in. elevation, but this is more than will be found in practice except in very unusual cases. This section shows clearly the details and dimensions used to overcome the destructive effect of the overhanging tread, but even on tangents at least a Hn. difference in grade should be maintained between the rail head and the pavement surface directly adjacent to it. On curves the amount of superelevation must first be decided upon, as from this is determined the difference in elevation between the rail head and the pavement surface on the inside of the curve. Building Construction 19 The following are good rules to follow in determining these points: (1) Place the outside rail (D in the section) at the eleva- tion it would have were the track on a tangent; (2) in the case of double tracks, give B the same elevation as D, if there is not much longitudinal grade at the curve. This plan leaves the half of the street on the outside of the curve with its normal crown, and gives the other half a proper slope toward the gutter. If there is much grade, an inlet may be located before the curve is reached, to catch the gutter drainage, so that D can be placed even with, if not below, the gutter grade; and B will be at the elevation of C, with A below B the amount of the outer eleva- tion. This method results in track drainage for the side of the street at the outside of the curve and thereby eliminates the gutter as a factor in the drainage. Interesting Lights on Building Construction- Theory, Design and Methods A Standard Notation for Beam Flexure The engineer should be as familiar with the application of the formula for beam flexure as with the use of the multiplication table. Because this formula is one of the most important in the whole subject of stresses, a standard system of notation is desirable. Sixty years ago Rankine gave in his "Applied Mechanics" the fundamental equation, M = pl/y. Later writers have departed from the notation of Rankine. A few present-day notations will be quoted. TEXTBOOKS STEEL-MILL, HANDBOOKS Church M=^ Bethlehem... . m = - = fS n Heller M = - Cambria M = gj- = pS X-i Kent M = - Carnegie . . M = - = fg n Ketchum M= ^ Jones & Laughlin Mr = - = fS Lanza M = ^ Lackawanna. . . . M = J 1 = PS Ci Merriman M = - f c Trautwine M = It is true that variance in notation causes little trouble to engi- neers making constant use of the equations, but to those who use 20 The Engineer in Field and Office them only occasionally it is a cause of annoyance. The writer suggests the following as a standard notation: A STANDARD NOTATION FOR BEAM FLEXURE b = Width of beam, in inches d = Depth of beam, in inches Section = Section cut by plane perpendicular to longitudinal axis of beam A = Area of section, in square inches c = Distance from the neutral axis to the extreme fiber of section, in inches I = Moment of inertia of section about neutral axis, in inches to the fourth power Iv = Moment of inertia of section about axis parallel to neutral axis, in inches to the fourth power v Distance between these axes, in inches r = Radius of gyration about neutral axis, in inches L = Length of span, in feet = Length of span, in inches W = Total load uniform'y distributed, in pounds w = Uniform load per foot of length, in pounds = Load concentrated at any point, in pounds M = Total bending moment at any section, in inch-pounds = Stress, in extreme fiber of section in pounds per square inch = Section modulus, in inches to the third power E = Modulus of elasticity, in pounds per square inch (steel = 29,000,000) D = Maximum deflection, in inches ^I X | = Axes of coordinates x, y = Distances of any point from axes of coordinates, in inches General formulas : Iv = I + Av 2 ; I = Ar 2 ; S = - ; M fS = - The letter n is universally used for the elasticity ratio between steel and concrete, or E 8 -r- E c For this reason the letter c is used in the above notation for the distance from the neutral axis to the extreme fiber of section. Stitch-Riveting of Struts In a built-up member, fastenings between the parts are required only at points at which these parts tend to move relatively to each other. The usual practice in designing compression members, how- ever, is to fasten the elements together in such manner as to guard against any increase in the slenderness ratio in any element of the column. Is this safe enough? Suppose we have a strut made of two 6 x 3i x i-in. angles spaced i in. apart, of length 96 in. Assuming the two angles to act to- gether, r = 1.45 in., and l/r = 66.2. The minimum r for this angle (Fig. 1) is 0.67 in.; hence, if we put the pin rivets at the center of length, we have a slenderness ratio of 48 -=- 0.76, or 63.1, which is slightly less than that of the main column. In Fig. 2 is given an elevation of the column looking at the face of the outstanding legs. Here is shown by means of dotted lines the deflection that the column would probably take if given a load aproaching its maximum capacity. There would be no tendency at the middle section for one angle to move axially rela- Building Construction 21 lively to the other. The column would act, as far as the deflection illustrated in Fig. 2 is concerned, as a column of length equal to twice the distance between pin rivets. In this case r = 0.97 (see Fig. 1). Multiplying 0.97 by 66.2, we have 64.2 as twice the allow- able spacing of pin rivets. This calls for connections at each third-point of the length. \! At 4 M / ,5 FIG. 1 " PIG. 2 With a middle rivet only, the slenderness ratio for parallel axes is 96 -^ 0.97, or 99. By the A.R.E.A. formula the allowable stress for this ratio is 9070 Ib. per sq.in., while the allowable stress for the entire column section, slenderness ratio 66.2, is 11,370 Ib. per sq.in. Under the latter load the angles considered as independent elements would be loaded 25.3% over their allowable stress. Method of Erecting a Large Steel Dome The 92-ft. dome of the Wealthy St. Baptist Church, Grand Rapids, Mich., has a steel frame formed by eight main arch mem- bers 35 ft. in span, with 19-ft. rise, framing into an octagonal 22 The Engineer in Field and Office crown diaphragm, 22 ft. wide across the points. The arches are tied together at the heel by four trusses and four sets of angle ties. The arches are 2 ft. deep at the top and 5 ft. at the outer ex- tremity. Three lines of beams parallel to the base ties carry the wooden ceiling and roof joists. The lateral bracing consists of a system of rods together with a line of struts in the center of each bay at right angles to the roof beams. A steel monitor frame 8 ft. high surmounts the structure. The erection procedure was as follows : A derrick of the required height was raised, and the eight sides of the diaphragm were riveted up around its base. With two sets of blocks, the ring was raised to the final elevation, 45 ft. above the floor, and light timber falsework placed underneath. The arches were raised with a gin pole, bolted in place, and the base ties erected. The roof beams, struts and rods were then placed. Graphic Solution for Fink Roof Truss A simple general solution for the graphical analysis of the Fink type of roof truss can be made without resorting to the usual expedient of substituting auxiliary members or using partly ana- stress Diagram Modified Fink Truss Analyzed Graphically lytical methods. For example, consider the modified Fink truss shown in the diagram. The only load affecting the truss members LM and LK is CD, and hence these stresses can be readily found graphically. Building Construction 23 Select any convenient point M' on the line drawn through d in the force diagram parallel to the rafter of the truss. Draw M'U parallel to ML, giving L' on the line through c. Draw UK' parallel to LK, and WK' parallel to MN, as shown. The triangle L'WK' thus formed gives the true stresses in ML and LK, and also the true relative position of the point K' with respect to the parallel lines through c and d. The true position of K', however, must be on a line through h parallel to HK, and therefore lies at k in the stress diagram found by K'k drawn parallel to the rafter. With k located, the remaining stresses are found as usual. The method can be applied to any loading and to unsymmetrical Fink trusses. Design of Concrete Retaining Walls Diagrams for the designing of retaining walls are simpler and more convenient than formulas. For the purpose of illustration "TYPICAL SECTION HOR.PRESDIAGRAM I 08/0 Values of Various -Functions. Fig. 1. Dimensions and Quantities Walls Without Surcharge two cases will be considered (1) that of a gravity wall without surcharge, and (2) that of a gravity wall with railroad surcharge 24 The Engineer in Field and Office of Cooper's E-55 loading, with center line of track 8 ft. 6 in. from the face of the coping. A gravity retaining wall of the type shown in Fig. 1, which is safe against overturning, will also be safe against sliding and crushing the masonry, and with a suitable toe may be built upon any ordinary foundation. We may then assume a minimum factor of safety against overturning and find the widths of the base for various heights. Assuming that the slope of repose of the filling TYPICAL SECTION HOR.PRESJ)IASRAMS Values of 'Various Functions Fig. 2. Dimensions and Quantities Walls with Surcharge is li to 1, the horizontal pressure on a vertical plane is, according to Rankine, 0.286 of the vertical pressure. Earth has been taken at 100 Ib. per cu.ft., and concrete at 150 Ib. per cubic foot. The diagram, Fig. 1, gives the width of the base of the neatwork for a factor of safety against overturning of exactly 2. Using a minimum footing projection of 6 in. at the toe and 1 ft. at the heel, the entire wall always has a factor greater than 2. The maximum compressive stress in the concrete is 90 Ib. per sq.in., and the wall will be on the verge of sliding when the coefficient of friction is about 0.3. The dead-load for the case of a surcharge wall is considered the same as in case 1. The live-load has been taken as 11,000 Ib. per lin.ft. of track, distributed from the base of rail and end of tie Building Construction 25 as shown in Fig. 2, the limits of the distribution being the face of the wall and a line 6 ft. 6 in. from the center line of the track midway between tracks, assuming a second track 13 ft. from the first. The ratio of horizontal to vertical live-load pressure has been taken the same as for the dead-load. The portion of the live-load pressure uniformly distributed over the width 6 has been assumed as adding to the weight and stability of the neatwork, while for the entire wall the width & + 1 nas been used. No dis- tribution has been considered beyond the plane of the face of the wall, no matter what the length of the toe might be. The horizontal-pressure diagrams are shown in Fig. 2, and the widths of the base of the neatwork for dead-load and live-load are plotted on the diagram, as are also the values for dead-load only, The factor of safety against overturning for the neatwork is ex- actly 2. The maximum compression in the concrete is 103 Ib. per sq.in., and the wall will be on the verge of sliding when the co- efficient of friction is about 0.3, as before. The limiting values for the height of wall have been selected so as to make the diagram general for all heights likely to be encountered. It is not intended, however, to indicate that gravity walls of the extreme heights shown could be economically built. For estimating purposes, the quantities for the higher walls would be on the safe side. Forms Designed To Leave Storage Space As the street had to be kept open for traffic, storage for 1250 yd. was provided in a building recently erected in Toronto, Ont., E^lS* Floor STORAGE Wide Space Between Bents for Storage by employing specially designed form supports. The use of bents with a wide space between them was made possible by unusual story- 26 The Engineer in Field and Office height of the first floor 18 ft. 9 in. and the fact that there was no basement. The bents were 11 ft. apart at the top and 17 ft. apart at the bottom, this construction being carried over an area six bays wide and three bays deep. The fourth bay, in front of the aisles thus formed, served as a connecting passage, and bents were built lengthwise of it for the full width of the six bays. Teams would drive into any of the first five bays, dump and return to the street by way of the connecting passage and the sixth bay, which was used for return only. Each storage space thus provided had a length of about 75 ft. and a width of 17 ft., giving a storage capacity of 250 yd. Two bays were reserved for sand, two for stone, and one for cement. Gypsum Roof Slabs of 10-Ft. Span A roof covering composed of gypsum T-beam slabs 10 ft. long, having a clear span of 9 ft. 8 in., is a feature of a steel-frame building 190 x 200 ft. recently erected at Racine, Wis. These long slabs, designed as the outcome of numerous tests made at the University of Illinois, are of T-section, 15 in. wide and 8 in. deep, with a thickness of H in. for the top, or flange, and 2i in. for the rib. The ends are closed by diaphragms 15 in. wide and 2 in. thick. Each beam has two -in. rods in the bottom of the rib, one of these being bent up at the ends as a shear rod and looped so as to increase the value of the bond stress. In the top flange is embedded a steel- wire mat of No. 14 gage and 4-in. mesh. The weight is 16 to 17 Ib. per sq.ft. The joints are plastered on the outer surface. These beams were made at the site, in wood forms, at the rate of about 300 sq.ft. per hr. They only required 15 min. to set, from the time the mold was poured until the forms were knocked down, and were erected in place and walked upon within three hours after being poured. They were designed to carry a uniformly distributed live-load of 50 Ib. per sq.ft. with a safety factor of 4. Tests of five sample beams 24 hr. old were made with loads of 2200 and 2400 Ib., or 200 to 218 Ib. per sq.ft., with respective deflections of 0.034 and 0.070 in. The total loads were then increased to 3500 Ib., causing slight horizontal shear cracks to appear, but without other signs of failure. In the roof construction these slabs, or beams, are carried by light steel trusses for the saw-tooth portions and by steel purlins, etc., for the flat portions, all of the supporting steel being spaced 10 ft. c. to c. In general the flat roof portion has a pitch of 1J in 12, while the saw-tooth construction pitches about 7 in 12. When Building Construction 21 a flat ceiling effect is desired, instead of the ribbed, or beam, ceiling formed by these T-beams, the slabs may be of H-section, with the top and bottom flanges of equal width. These weigh from 22 to 24 Ib. per sq.ft. for the live-load noted above. In such a roof now being erected the beams weigh 22 Ib. per sq.ft. Crane Runway Columns Carry Roof A yard crane with supports designed to carry a roof over the yard is an interesting feature of the new foundry plant for the James A. Brady Co., at Chicago. The material yard is between two buildings and is served by a 5-ton electric traveling crane. tf ~lLaw.i rtsnj .........^ **4RnkiM''ie ?'-(,* ^"Cross Bracing i.-*6v.-y*-v.- vi T*t5CW^V4^c & r^i ^% fiSf^r^ T/iis Runway for a Yard Crane Is Designed To Carry a Roof Elevation of Cross Bracing The crane runway consists of 20-in. I-beams carried by brackets upon columns about 42 ft. high. The runway is 180 ft. long, with columns and trusses spaced 20 ft. c. to c. The columns are of H- section, with a 10-in. web plate and four angles 5 x 3 in. Longi- tudinal bracing is provided by a double line of 6-in. channels, form- 28 The Engineer in Field and Office ing an inverted T-section, between the columns, at a height of 15 ft. above the ground. Horizontal and vertical diagonal bracing is provided in three of the bays. Temporary Stringers Carry Elevated Tracks An opening 4 ft. wide in the structure of the Metropolitan Elevated Ry., Chicago, was necessitated by changes in the bents and spans where the line crosses the new Union Station. One new bent was 4 ft. from the old one, and the adjacent girder span had /Pt:to"x\ i NEYfBEKT ^6- >|<~/0^ NEWPAHELSPUCED TO GIRDER* Stringers Carrying Track Rails Across Gap in Elevated Railway During Alterations to be lengthened by splicing on an additional panel. Traffic could not be interrupted; and while the work was being done, each rail was carried across the gap in a stringer of trough section, the ends of the stringer resting on the track ties at each side of the opening, as shown in the accompanying drawing. Tar Paper Deadens Noise of Trucks To deaden the noise from passing trucks rumbling over a con- crete floor, a factory in Norwich, Conn., has successfully used a heavy tar paper pasted to the floor by paint. The method of appli- cation is as follows: The floor is first given one coating of a gray cement paint. On the following day, when the paint is thoroughly dry, a second coat is applied. At the same time one side of a five- Building Construction 29 ply tar paper is painted; and when both paper and floor are still wet, the paper is carefully laid wet side down on the floor and rolled with either a roller or wide-tired truck until all signs of air pockets beneath the paper disappear. The surface seems to improve with age and very effectively reduces noise at a low cost. Excess Volume of Flared Column Heads In computing the volume of the capital or flare in a column for girderless concrete floor construction, a simple method is to deter- mine the volume of the flare outside of the column shaft, which is itself figured as running from the top of one floor slab to the bottom of the next slab above. Formulas for this additional volume have been deduced by Edgar H. Mosher, of Washington. Where V = volume of concrete to be added to the volume of a straight column figured to the bottom of the slab; D, diameter of flare or top of capital, and d, diameter of column, the formulas are as follows: For square columns, T/ V^ **/ v^ 6 For round columns, _ (D- dY(D 7.63 For octagonal columns, 7.23 These formulas are deduced from the truncated cone formula, assuming 45 to be the slope of the flare. Special Forms For Light Floor System The designers claim a considerable saving in weight and hence in cost over any other fireproof floor of equal strength for a floor consisting of rows of small concrete beams or joists of T-shape poured monolithically with their supporting beams or girders between forms made up of spacing members or boxes of wood and cores of bent sheet metal. The accompanying perspective shows the general layout of the system, which may be used either on steel- frame or on reinforced-concrete buildings. In building the floor, a line of struts is placed in each lower T-beam and capped with plank that serve as beam bottoms. On 30 The Engineer in Field and Office these plank as joists is then laid expanded or ribbed metal that forms a ceiling base. Then metal forms bent to the shape of one side of the T-beam are placed over each plank and held in position by properly cut brace boards between beams and by stiffener planks along the top. In the T-form thus made, the proper reinforcing is placed and the concrete poured to make the T-beam. Perspective of Layout of System When these beams have set, the bracing boards are knocked out and the beam metal forms released to be taken upward and used on the next floor above. The beams are then overlaid with another sheet of expanded metal, and the usual cinder concrete floor is placed. For the upper (floor) layer a fireproof board could be used in place of expanded metal. Building Construction 31 Concrete Cantilever Beams Cantilever beams of reinforced concrete, designed to support an outside entranceway leading from a bridge over the Fox River, and flat-slab floors without column capitals, to facilitate the placing of partitions, are two principal features in a recently completed eight-story hotel at Aurora, 111. Above the cantilever entranceway for three panels of its length the mezzanine floor was also extended over the river on columns supported by the cantilever beams, the heaviest of which is here illustrated in detail. The building is 70 x 100 ft. in plan, with its main entrance on the south side and with a second entrance at the northwest corner, where ! the hotel adjoins the new bridge. This out- > side entrance, which is indicated by the photo- . graphs, cantilevers 10 ft. over the river, and > results in saving space within the building 1 lines to such an extent that one more store > is now available which would have been elim- ? inated if this scheme had not been adopted. In order to facilitate the locating and rear- > ranging of partitions, flat-slab floors supported ^ on square columns designed to eliminate the I column capitals usually used were adopted. 1 The first-floor slab is 7 in. thick; the slabs 5 used on the upper floors are 6 in. thick. No beams were required except the cantilevers already described and beams for window lin- tels and to frame around elevator and stair openings. The design of the slabs proved to be economical, four-way reinforcement of i-in. rods being used. The forms for the cantilever beams were also constructed as a cantilever timber frame- work, owing to the fact that the river bottom beneath the outer ends of these cantilevers consists of boulders, dirt and rubbish, making it impossible to drive piling on the shore. 32 The Engineer in Field and Office Window-Frame Clamp Saves Time and Room When setting window frames in stone or terra cotta faced build- ings, considerable time can be saved, as well as avoiding the obstruction of floor space with long braces, by using the clamp shown in the photograph. The long piece is a 2 x 4 to each end of which is nailed at right angles a piece of board about 6 in. long. Five nails are used in each end. The distance between the inside edges of these boards is 1 in. greater than the distance between the outer face of the jamb and the face of the brick backing. The window frame is set up when the first jamb stones have been placed, after which one of the braces is placed at each side. Clamp Can Be Quickly Adjusted The braces may be tightened with wedges, or by driving down one side until the brace sets tight. They will hold the window frame securely in place until the brick work is finished. Anchor High Form to Face of Stone Retaining Wall Anchor bolts were used as shown in the sketch to secure the form for facing with concrete an old retaining wall of dry masonry which supports a section of a railroad line that runs along the face of a cliff about 100 ft. high. The wall extends down to a ledge about building Construction 33 50 ft. below the track. This ledge was about 10 ft. wide outside the old stone wall, and the working space was still further reduced by the thickness of the new wall, which was 4 ft. at the bottom and 18 in. at the top. The anchor bolts were of i-in. round rods, one High Form on Cliff Face Anchored to Old Wall end of each of which was split for about 2 in. to receive a wedge. Holes 5 in. deep were drilled into the stone of the old wall, into which the split ends of the rods, with small wedges inserted, were driven. As it was impossible to remove the bolts after the forms were stripped, they were nicked with a hacksaw about 1 in. inside the face of the form. When the forms were taken off, a few sharp blows with a hammer knocked off the ends of the bolts inside the surface, and the holes were plastered up. Some Capital Things in Foundation Work Use Pier Form as Inside Cofferdam The pier-shaft form was set under water and did double duty as inside cofferdam and concrete form in constructing the shaft of the pier for the Division St. bridge in Spokane. The cofferdam had been sealed under water, and two layers of bracing, which had been required during the driving of the sheeting and the placing of the seal, were then removed. This work was done by a diver, taking out the crossbraces but leaving the tie-rods in place. The form for the pier shaft, which was built of 2 x 8-in. 34 The Engineer in Field and Office tongue and groove sheeting, placed vertically, was stood up in the cofferdam, brought to line and well braced inside with the help of the diver. The cost of this form exceeded by a surprisingly small amount that of similar work done in the open. The extra width of the cofferdam and the batter of the pier left a space 28 in. wide at the top and 16 in. wide at the bottom between the outside of the form and the piles. Around this space, about a foot of concrete was deposited under water, sealing any leaks between the concrete and the bottom of the form. This space was then carefully puddled with a mixture of gravel, clay and manure, gravel being used to save the clay, which was expensive. This formed a puddle-wall cofferdam for practically the cost of a single line of sheeting, plus the cost of placing the puddle material. Having tongue and groove walls on both sides, the dam proved very tight. The method also saved considerable time, the initial time for constructing the cofferdam being much less than if it had been made with two rows of sheeting, and the form and puddle work being carried out while the concrete seal was setting, which time would have been lost. A pump was put inside the form, the space unwatered and the seal concrete carefully cleaned. The bracing inside the form was taken out as the pier concrete came up. Cofferdam on Uneven Rock Bottom A cofferdam of Wakefield sheet piles, which was framed and put in place as a skeleton before setting most of the piles, was used successfully on the practically bare and very irregular rock bottom at the site of one of the piers for the new Division Street bridge in Spokane. This bridge was a structure 70 ft. wide and about 600 ft. long, consisting of three concrete arch spans with concrete viaduct approaches. The pier between the second and third arch spans was located in water which was 23 ft. deep at the lowest point, at low-river stage. Being only a short distance above the falls, there was a strong current at the bridge site. Soundings showed that the site of the pier was crossed diagonally by a vertical ledge, the bare rock at the downstream end of the pier being about 15 ft. below the water, while the rock at the upstream end, where the depth was greatest, was covered with 3 or 4 ft. of sand and fine gravel. It was planned to construct the pier on a concrete seal placed on the rock under water, and it was necessary to use a cofferdam because the irregular bottom made a crib out of the question. Foundations 35 The cofferdam, the skeleton of which was built in the water, towed to place and set before the insertion and driving of most of the piles, was made 2 ft. wider and 3 ft. longer than the base of the pier. Three layers of outside and inside waling, between which the sheeting fitted, were framed and spaced at the proper Frame for Sheet-Pile Cofferdam Supported by Driving Down Slotted Piles distance apart by bolting them to sheet piles placed at each cross- brace, 16 ft. apart. In these piles, through which passed the long bolts shown in the drawing, which held each layer of bracing together, were cut H-in. by 4-ft. slots to permit driving them after the dam was placed. So that they would not bind after the frame had been drawn tight with the long bolts, 1-in. pipe sleeves, \ 36 The Engineer in Field and Office in. longer than the thickness of the pile, were placed around each end of each long bolt in these slots. These sleeves, bearing against plate washers inside the rangers, prevented their being clamped tightly to the pile. The rectangular part of the frame was 12 x 73 ft. inside the sheeting, but the total inside length of the cofferdam was 89 ft. on account of the two pointed nose sections. In addition to the posts shown in the drawing between the three tiers of bracing, the frame was X-braced for its entire length in a vertical plane, and each layer was also X-braced with 2 x 6-in. planks. Pile bents, braced above and below water, had been driven out to the site of the pier for construction purposes and to serve later as a support for the arch centering. Extending the entire length of the pier and about 3 ft. from it, this falsework afforded a good anchorage along one side of the dam. On this falsework was a large stiff -leg derrick. ,The frame was moored and held plumb by the derrick while being weighted with bags of sand until it grounded on the downstream end. The upstream end was weighted to float level with the lower end, after ' which the slotted piles were driven, where the 4-ft. slots were long enough to permit it, to a bearing. These piles were then bolted securely to the frame, serving as legs to hold it in place while the remaining Wakefield piles were set around it. Special piles were made where necessary to secure tight closures. When all the bays were filled, the piles were driven to rock with a drop hammer, handled in swinging leads by the derrick, and bolted to the frame, more weight being added to overcome the increasing buoyancy. The piles were built up of 3 x 12-in. plank, surfaced one side and one edge, and were dry when placed, so as to swell tight. The only openings were now the slots, the irregulari- ties on the bottom, and in some cases a hole as much as 3 or 4 ft. high where the slots did not permit driving a pile to bottom. The large holes and slots were boarded up by a diver. Concrete Bases for Old Bridge Piles Replacing piling under wood bridges still in fairly good repair as to caps, joists and flooring has been one of the big items of expense in Sedgwick Co., Kan., for years past. The county is now doing a great deal of permanent bridge and culvert construction, but there are still many pile bridges over the larger streams which will have to be maintained for a number of years to come. To reduce this expense concrete bases have been placed under the piles Foundations 37 that have rotted dangerously at the ground line. Replacing a pile always necessitates tearing up the deck of the bridge, removing at least two or three joists and frequently the cap, and sometimes requires falsework for the piledriver. Where it has been possible to work on the ground and there was soil that would stand, a repair gang has put in a concrete base, retaining the upper part of the old pile. The chief difficulty has been to get down past the root of the pile. This has been surmounted in most cases either by pulling or digging out the lower part of the pile, or by moving the top of the pile slightly. -.own/e ^Lagscrews Base for Old Pile A 7-in. post augur drills the holes to a depth of 5 to 8 ft. Sometimes the holes are reamed out by a "jabbing digger" or a pile spade, which operation allows a mushroom base to be formed. Where shale or a very hard clay subsoil is encountered even though only a few feet under the surface no attempt has been made to penetrate it; nor have these holes been put down or bases built where there is surface water. In a number of instances where sand was struck below the surface, this sand was thoroughly mixed with a sackful or less of cement, thus forming a base that holds satisfactorily. This construction, however, is never attempted except where the surface soil is firm and hard. After filling the hole with concrete, a round sheet-metal form, a little larger in diameter than the hole, is placed at the surface. In the form three or four iron straps are placed, so that they will 38 The Engineer in Field and Office project 8 to 12 in. above the concrete. Each strap is drilled for two lagscrews, and the straps are placed so that the pile slips between them. The top of the base is pressed down or cupped, using either a wood form or a trowel. This device helps to keep the pile firmly in place. The concrete sets for 48 hours before the pile, which is secured with the lagscrews, is placed. Under favorable conditions this method of repair is much cheaper than replacing old piles with new, and it seems to be more lasting. Cushion Blocks Reinforced by Vertical Rods With a specially designed cast-steel hood fitted with a reinforced wood plug rapid driving of steel sheet piles has been accomplished at Dam No. 39, Ohio River. As many as fifty-four 25-ft. Lackawanna piles have been driven with one driver in one 8-hour shift. The Woodjkjshion Secrion A- ''-> Round Pins Transmit Blows and Save Cushion Block hood is so constructed that either straight web, gusset, corner or T piles can be fitted into the grooves in its under side. It is held to the hammer by wire lines or by rods passing through If in. holes. A circular hole in the head permits an oak block to be inserted as a cushion to receive the direct blow of the hammer. This block, 11 in. in diameter, is bound with a 1 x 2-in. iron collar. This prevents it from mashing up at the top. Holes are drilled in the block to receive l-in. round iron pins which transmit the blow from hammer to pile follower. Foundations Spreading Piles Pulled Back By Turnbuckles The column footings for the three-track elevated railway along Stillwell Ave., Brooklyn, N. Y., were constructed across a swamp in which the mud varied in depth from 15 to 30 ft. The piers were supported on spruce piles 30 in. on centers, which were driven to refusal with a 2000-lb. drop hammer into the sand and gravel under- lying the mud. Fill made p "^ * 6 '' torefain fl ft 2'/&Wire Footings M$\ ; Cables _ _CinderFHI_ Spreading Footings Restored by Turnbuckles Over the greater part of this swamp there was an ash fill 6 or 8 ft. in depth, which had been in place some 6 years. This fill served to give lateral stability to the footings. At one point, however, near a creek, there was no fill on the original salt marsh. Here it was brought home forcibly to the engineers that piles driven in soft material, although able to support the load brought upon them, have little lateral resistance, and that a structure sup- ported on piles under such conditions is liable to injury if a horizontal force is brought to act on them. The piers at this point are 7 ft. 2 in. by 9 ft. 8 in. in plan, sup- ported by 12 piles each ; they carry a load of 145 tons. The contractor laid a standard-gage track on a cinder fill about 2 ft. deep between the two lines of piers, and used a 50-ton locomotive and flat-cars 40 The Engineer in Field and Office to haul structural steel. The load of 35 tons per car axle which was thus brought to bear on the underlying mud forced it out laterally against the piers. It sprung the piles and spread the opposite piers apart 2 in. in the bent in question. In order to repair the damage, two I-in. wire cables with turn- buckles were passed around the piers. By drawing these up, it was possible to pull the piers back to correct position. An earth fill was made around the footings to retain them in their proper position, the cables being left in place as an additional precaution. Freezing Ground Acts Like Hydraulic Jack Some 16-ton concrete piers in a Middle-West city were heaved this past winter by as much as 3 in. and subsequently settled back to their original elevation. This most unusual and extreme condition cannot be explained by ordinary frost action, but can be accounted for by the piers becoming the pistons of hydraulic jacks in which frost produced the moving pressure. This occurred where a bridge was being built in the course of track-elevation work. Some heaving also occurred at an adjacent street. The abutments and piers at both streets were poured during the summer and autumn of 1916. The bridge steel could not be placed until the following season, hence the piers and abutments carried no load other than their own dead weight. Where the most heaving occurred, only the center and south rows of piers were appreciably affected. The north row was fairly well drained and much drier than those to the south where all conditions were favorable to waterlogging. The bottom of the north side piers was about 4i ft. below the ground surface, of the center piers 4 ft., and of the south piers 3i ft. The discovery was made in March, 1917, that the south row of piers had heaved by amounts varying from 0.07 to 0.22 ft. The center row showed heaving ranging from 0.01 to 0.09 ft. The north row showed nothing in excess of i to i in., most of which could have been in the original setting. After this was discovered, levels were taken at intervals of 3 to 5 days. As the ground thawed the piers settled back into position until the highest corner of any grillage was but 0.06 ft., or |-in. above what it should be. A settlement of as much as 0.18 ft. or 2i in. is shown. The usual explanation would be that the frost penetrated to a level below the bottom of the piers and heaved them by direct action, but this does not seem adequate for the following reasons: Foundations 41 (1) It is very doubtful if the frost penetrated as low as the bottom of the piers. (2) If the frost did penetrate to below the bottom of the piers there could not have been more than from 6 in. to 1 ft. of frost at the most. This thickness of frost could not possibly heave a pier nearly 3 in. Water expands about ^ of its volume upon freezing. The expansion of water-soaked soil would not be greatly in excess of this if as much. That would mean but 0.1 ft. upheaval for a whole foot of frost under the pier. If the pier is considered analogous to the piston of a hydraulic jack and the water or semi-fluid mud is forced under the pier by the pressure from the freezing expanding strata nearer the surface, the action can be understood. Piers Were Plungers in a Big Natural Jack Let AB represent the surface of the ground. Suppose the ground to freeze to a depth d, such that the frozen layer becomes rigid and unyielding. Let the frost then penetrate to an additional depth e. The layer e expands in freezing. It exerts a pressure both upward and downward. The frozen mass d is unyielding and if the weight of the pier is less than the force required to break the rigid layer, the water and semi-fluid clay will be forced upward like a piston. Such an action would account for any amount of heaving. At the other location, where there is practically no depression, only 3 or 4 piers out of a total of 42 were found to have been raised. The heaving varied from i in. to li in., only one pier having been heaved the latter amount. But at this street no pier bottom was less than 5i ft. below the ground surface. It is certain that frost did not penetrate to that depth. It is unfortunate that test borings for depth of frost penetration were not made at both streets in order that more exact information might be had on this point. 42 The Engineer in Field and Office Locomotive Tows Caissons to Place The caissons launched for the bridge at Moncton, N. B., were quickly and easily towed to place by a locomotive at slack tide and moored before the treacherous currents of the Petticodiac River had an opportunity to carry them away. When the first caisson was completed, it was lowered by jacks to sliding ways of 12 x 12-in. timber 24 ft. long. Each of these ways rested on one of the four main ways, which were spaced on 14-ft. centers under the caisson, giving an overhang of 13 ft. at each end. The main ways, which were braced and blocked continuously, had a slope of about 1 to 9 and were run into the ground at the lower end to prevent their sliding. The struts between the main ways were weighted with rails to keep the ways from floating at high water. The forward ends of the sliding ways were also weighted with rails to sink them as the caisson floated, when they were pulled clear by lead ropes attached to their rear ends. The caisson was launched one hour before high tide, while the current was still running in slowly. As expected, it reached the end of the ways before floating, giving time to get a li-in. steel towing line aboard before the tide rose sufficiently to carry it clear of the ways. This line ran through snatch blocks on scows moored at pier 3 and at the site of the caisson, and was led from there to a third sheave in line with a siding on the opposite shore, where the line was attached to a switch engine. As the caisson floated, the loco- motive started slowly, pulling it over to the first scow at pier 3. The shackle pins on the mooring lines of this scow were knocked out and scow and caisson together hauled over to the second scow. The latter, moored near the location of the pier, had on board the anchor lines up and down stream. These were quickly transferred to the caisson, which was then dropped downstream to approximate posi- tion. The entire operation was completed before the outgoing tide interfered. Bearing Test on Confined Wet Sand A concrete slab-and-girder mat, with the wet treacherous sand underneath confined by a ring of interlocking steel sheet piling, supports the new boiler house and coal-storage plant of the New York Steam Co., at Burling Slip and Water St., in downtown New York City. The load of the boiler room averages 2.6 tons per sq.ft. of the entire foundation ; for the coal-plant mat the load is 5.4 tons. Before the Building Department would permit such a foundation to Foundations 43 be laid, it had to be convinced by tests of the safety of the method. The sketch shows how the loading test was made and gives the settle- ment by curves. The test arrangement is really a model of the foundation pro- posed. A steel sheet-pile box was driven to a depth of 26 ft. below curb and the material inside excavated to a depth of 15 ft. A concrete slab 2 ft. thick was placed on the sand bottom below Lead in Tons per Sq.ft. Loading Test for Unusual Foundation groundwater. The slab was loaded with pig iron to give a maximum load of 6 tons per sq.ft., and readings were taken on the four corners. The greatest settlement after 237 hours was 0.061 ft., the average settlement for this period being 0.047 ft. The time curve shows that there was no settlement between load applications. 44 The Engineer in Field and Office New Ideas in Designing and Building Bridges and Dams Clear Form of Notes for Bridge Stake-Out The following form of note keeping for stake-outs of bridge bents has been found simple, quickly made and of ready reference "HIGH BRIDGE" 5take-ou+ Mark. on rock near left stake = elev B.M.*2 1095.49 1 127.82 *bot. cap 6.5. __416./!0954bFtqs. H. I. * 1099.65 3/3241 '32'-05*Heioh< r d - 4.44 \ 3 60 lt rELFootinqs l095.4|-qdd7.'QO 10.60-10-07^ 4- Offset stakes I4'-07^" Bent @Sta280t9fl5 Mark, top of left stake=elev^ J.PT.U/ Instrument "Transit AJ Y ) Z7 "'7 B.M*4 1113.55 1127.64- bah cap. B.S = 0.33 r 1 109.03 = F ^rllU9.UJ = rtqs. j 9JI6.6I 'iS'-OTl^igM r add 7.00 =9.07= 9'-Ol" "I ' 1113.86 levFtqs->H09.03 y add 7.00 -9.07- 9'-QJ add Offset stakes Record Bridge Stake-Out Operations in This Manner on the job. If the bridge is on a grade, it is necessary to have available, preferably in the same notebook with the stake-out records, the information indicated in the accompanying table. DATA ON "HIGH BRIDGE" Station 280 + 80.5 280 + 98.5 281 + 16.5 Bent No. 12 13 14 Grade (1%) Floor Surface Bottom of Cap 1129.82 1129.64 1129.46 1127.82 1127.64 1127.46 With the above data and benchmark levels readily accessible, the staking out of the bridge bents is undertaken, and the operations so performed are recorded in the manner suggested by the illustra- tion of a typical field notebook opening. Bridges and Dams 45 Temporary Hinges for Concrete Arches The common assumption that, with an arch curve laid out to conform to the dead-load equilibrium curve the dead-load produces no bending moments in the arch, is materially in error. This error arises from the arch shortening produced by the dead-load compres- sive stresses and is similar to a fall in temperature; it results in a reduction of the horizontal thrust, with a consequent divergence of the true pressure line from the assumed arch axis. The pressure line then passes above the axis at the crown and below the axis at the springing, thereby increasing the compression in the outer fibers at the crown and in the inner fibers at the springing. The magnitude of the foregoing error depends upon the propor- tions of the arch. It may be neglected in arches having a rise greater than one-fourth of the span ; but in flatter arches it becomes increasingly serious. From some designs that have been worked out, in which the depth of rib was uniformly 3% of the span length, it appears that the addition to the dead-load stresses on account of rib-shortening may attain considerable magnitude, especially in the flatter arches. A method of avoiding these additional stresses is to provide hinges at the crown and springing during the erection of the arch. Temporary hinges were employed by the late George S. Morison in the construction of masonry arches; they have also been used in a number of European bridges. These hinges should be closed by pouring concrete into the joints only after the full dead-load is on the structure and the shortening and shrinkage changes have taken place. By selecting the temperature for closing the hinges, the range and effect of the subsequent variation may be minimized. The above advantages of the three-hinged construction apply mainly to the conditions during erection. For the finished structure, a hingeless type is to be preferred on account of the greater rigidity and the greater security against crown settlement. Temporary hinges eliminate the shrinkage stresses without involving the difficulty in the construction of bulkheads caused by interference with the steel in the case of reinforced construction. Circular Curve for Arch Design In laying out an arch curve for the first trial design, a simple circular curve is ordinarily satisfactory. With this curve drawn, the weights of the arch segments and superimposed filling are figured and the resulting equilibrium curve constructed. With this pressure line as a new arch curve, the deadloads may be revised and a second 46 The Engineer in Field and Office equilibrium curve drawn ; as a rule, however, the first curve may be retained without the revision of a second trial. For ordinary concrete arches with earth-filling up to a level line, the ideal arch curve will be found slightly higher at the haunches than a simple circular curve. The following table gives the amount of this deviation, as found from actual designs, for different rise- ratios : Deviation from Circular Curve Ratio of Rise at Haunches, in Per Cent. to Span of Rise 0.25 4.3 0.20 4.2 0.15 3.7 0.10 35 0.07 3.1 In the flatter arches, the deviation from a circular curve is barely noticeable. Theoretically, the ideal arch curve is the equilibrium curve for the dead-load plus one-half of the live-load covering the full span. This curve would be the exact mean of the two extreme curves obtainable by placing the live-load alternately on the two halves of the span. In practice, however, on account of the usual small ratio of live- to dead-load, there is no material difference in using the equilibrium curve for dead-load alone. Approximate Methods for Arch Design The elastic theory when applied to arches with fixed ends is not only time-consuming, but is essentially a method of analysis and not directly of design. Hence there is a real need for satisfactory formulas or methods for determining approximately the crown thickness and the proper form of the arch ring. It is necessary in order to determine the stresses to assume the dimensions of the arch rib. Experience must be the main guide in this primary assumption, and the designer has two classes of aids: 1. Emperical formulas, which are crystallized expressions of past experience ; 2. Approximate methods, by which crown thrusts and lines of pressure may be determined. Formulas of the first class are the modern representatives of the experimental proportions which probably guided the ancient arch builders. Methods of the second class are an outgrowth of the line of pressure theories developed for voussoir arches, and are an improvement over class 1, principally in that a greater range of conditions can be considered, and that more particulars of the form of the arch can be determined by their use. Bridges and Dams 47 Joseph P. Schwada has developed formulas for the thrust T and the depth d at the crown in terms of a coefficient K representing the ratio between the average crown stress and the maximum stress in the arch. The following equation has been derived for the thrust, based upon the assumptions used by Mr. Schwada, but using the general formulas derived by C. Tourtay in 1902. In Mr. Schwada's notation : This equation gives slightly lower values than Mr. Schwada's equation, all terms being identical except those in the parenthesis containing R and d. The depth at crown is T d = 144/Jf Mr. Schwada presents a valuable series of diagrams and tables for the value of the coefficient K, and for the solution of his equations. The process of the design of an arch ring is then as follows : 1. Tabulate the general conditions which can be at once deter- mined or assumed, and the known factors such as span, rise, loads, etc. 2. Make a rough assumption of the crown and springing line thickness (possibly using a formula of class 1). 3. Compute the value of thrust T, and if necessary correct the assumed crown thickness. 4. Construct a line of pressure with horizontal thrust (pole distance in the force polygon) equal to T. 5. Choose a value for K which will suit the type, the rise, and the span of the arch, and compute the value of the crown thickness d. 6. Choose a curve for the arch axis which will fit the line of pressure constructed as in 4. 7. Vary the thickness of the arch and place the reinforcement in accordance with the thrusts shown in the force diagram, but making proper allowance for the effects of moving loads and temperature. 8. Analyze the arch thus determined by the elastic method. It will be found that a very close approximation to the best design is determined by the first seven steps, and only minor changes may be expected from the full elastic analysis, while the labor is far less than a preliminary analysis by the elastic method. 48 The Engineer in Field and Office Culvert Pipe Used as Basis for Floor System to Bridges A reversion to an old system of highway bridge floor, with some modern variations, is seen in the use of corrugated iron culvert Underside of Colorado Bridge Floor Where Culvert Half Rounds Are Used as Forms pipe forming the basis for concrete arch floor spans. In the old designs light gage corrugated steel in the new type half sections Concrete Floor of Ohio Bridge To Be Placed on Culvert Sections as Forms of heavy-gage culvert iron is used, with additional reinforcement of the concrete floor against expansion. Bridges and Dams 49 One view shows such a floor as installed on a through-truss bridge in Hamilton County, Ohio. For this bridge the culvert forms span between the bottom flanges of I-beam stringers which are framed into floor-beams. The other view shows the under side of the floor of a deck-girder bridge at Boulder, Colo. Here the main girders, 41-ft. span, are of reinforced concrete and the floor, a concrete slab reinforced against expansion with l-in. rounds on 10-in. centers in both directions, is poured on half-culvert sections spanning steel 24-in. I-beams which are spaced 4 ft. apart and placed parallel to the main girders and span between the abutments. In this bridge the I-beam stringers are stiffened by cross-braces of steel bars, as shown. Bridge Erected With A-Frame An example of the solution of a difficult erection problem through taking advantage of the exceptional natural conditions is seen in the methods adopted in the case of the Pool Point Bridge on the Elkhorn extension for the Carolina, Clinchfield & Ohio Railway. This bridge crosses a deep basin in which there is always from 50 to Center Support Was Provided for Cantilever Erection 90 ft. of water. The stream is of torrential character, and above the site of the bridge is a splash dam which at times discharges large numbers of logs which would carry out any ordinary falsework. These special conditions made it impracticable to use falsework of the usual type. The character of the rocky side walls suggested the use of an A-frame or arch type of center support and cantilever erection from the north shore for the 270-ft. main truss span. This design was therefore made of the riveted type, with provision for compression in the lower chords and proper sections throughout for cantilever erection beyond the center-panel point. The A-frame legs were made of derrick booms. 50 The Engineer in Field and Office The main truss span of nine 30-ft. panels, 270 ft. long, flanked on the north by a 50-ft. girder span and on the south by a 70-ft. girder span is composed of two trusses spaced 19 ft. apart on centers. The main span weighs 613 tons, and the order of procedure in erection was as follows: The end post and 50-ft. plate girder span having been erected by the derrick car, the car could be advanced and the first panel placed with a temporary wooden bent at panel point Lj. The car was then advanced and the A-frame legs were placed on the north side and held in position by guys. Following this, lines having been passed to the other bank, the south legs of the A-frame were swung. The shore ends of the A-frame had cast-steel bolsters supported on concrete skewbacks and the tops were provided with cast-steel shoes for supporting the truss. The A-frame being in place, the remaining panels up to the center were erected, the splices being located as shown on the diagram, and the center panel point blocked up on top of the A-frame. From this point cantilever erection proceeded regularly panel by panel until the shoe at the south end had been placed. The deflection of the cantilever end was about 8 in., and care had been taken to block the center high enough to bring the south end about a foot high, allowing for this deflection. Jacks were then applied and this south end lifted until the span was free at the center. The A-frame was removed and the span lowered to the masonry. Radial Bracing for Large Cofferdam In constructing a steel sheet-pile cofferdam 46 ft. in diameter for the Pennsylvania R.R. it seemed desirable to use a quantity of 8 x 16-in. 18-ft. timber, which was on hand, for bracing. As the distance across the cofferdam was more than twice the length of the timber, a wood pile was driven in the center of the cofferdam before excavating; and a "hub" supported by this pile was constructed. Excavation has been carried 18 ft. below water level, and two sets of waling and struts have been placed in this manner. The method has been very successful. Forms Have Sliding Cantilevered Studding Cantilevered vertical studding which does not have to be taken down to set the next lift of forms is used to support the forms for La Loutre dam, being built on the upper St. Maurice River. By the use of such supports the interior spaces where concrete is being Bridges and Dams 51 deposited are kept clear of all tie rods. Two pieces of timber on edge are held parallel and 1 in. apart, being double bolted through a space block at the upper end. This arrangement gives a slot the whole length of the support, so that it can be slid vertically upward for another lift of concrete. In this respect the scheme differs from previous forms of same type. The holding bolts are tightened when the form and support are in place for proper alignment. If there is any overhanging, the heel or lower end of the support is blocked outward. Form framing here is horizontal and lagging is vertical. Holding bolts 1 in. in diameter are spaced at 2-ft. intervals. The bolts are wrapped in a sleeve made of paper which has been dipped in tar and dried. At the inner end is a 3 x 3 x T 5 F -in. plate. Very little difficulty has been found in screwing the bolts out. Large square washers at the end span both timbers of the support. Automatic Flood Gates for 17-ft. Dam The 17-ft. concrete dam for a hydro-electric plant on the Cedar River at Nashua, Iowa, is equipped with a number of automatic flood gates each 46 ft. long and holding back a maximum of 7 ft. of water. The installation of the flood gate is worthy of comment, because this particular type is new to this country. It was imported from Switzerland and service has already proved its value. It is simply a walking-beam with a gate hung at one end and a concrete counterweight hung at the other. The gate is hinged at the bottom. When the water rises above the required elevation, the gate turns down, which in turn raises the counterweight. The higher the water the farther it opens the gate. When the gate opens, the leverage between the counterweight and fulcrum increases and that between the gate hinges and fulcrum decreases, thereby overcoming the increased weight of water at every stage of gate opening. Gravity and constantly shifting leverage are its features. The gate is of steel and was entirely fabricated and set before grouting in. Creosoted plank was used for decking, well bolted on. Leaking around the ends was prevented by a leather bearing on the concrete. A leather bearing on a curved plate prevented leakage over the hinge where the gate fastened to the rollway. The counterweight was formed and poured in place, being shored up from the crest of the dam. Concrete Closing Slabs Slide to Place On the construction of the dam at the new hydro-electric plant at Hiram, Me., on the Saco River, water was allowed to flow through The Engineer in Field and Office the dam during the main building period, three openings being left in the lower part of the deck. Closure was then accomplished by three reinforced concrete gates which were cast upon the deck of the dam as shown and each held in position by a wire rope attached /7& close Ga+e saw Loo here ^issyS?:\ #'& 6ATE''* : \ $>> ; 6ATE ' p^^iM^I^;- &.4^fiM^*^*4*i'*A S e c -t- i o n Sec-Hon through Ga-f-es Ups-t-ream Eleva-fion Slabs Slid Down Deck of Dam on Greased Tarpaper To Make Closure to log anchored under the opposite wall of the dam by a second rope and hook. Each closure gate was reinforced both ways with l-in. square rods spaced 4 in. on centers and was provided with a U-bolt by which the wire attachment was made. To prevent adhesion between the closure slabs and the deck of the dam, two layers of single-ply roofing paper, greased between layers, formed the base upon which the concrete was poured. The closure gates were about 16 ft. high, 18 in. thick and 15 ft. wide, sliding downward between guides composed of 4-in. tees flanked by 4-in. by i-in. plate on each side. By sawing off the log above each gate on the top of the dam the slab was released and slid into place seating against a concrete recess at the bottom of the deck, as shown in the cross- section. Bridges and Dams Slipping Bridge Abutment Saved A casual inspection of the 100-ft. steel highway bridge which was built some time ago without engineering supervision showed that the anchor bolts were bent and the shoes were pressed tightly against them. There was no information available as to the exact location of the abutment or as to the batter of the faces, but it was very easily seen that one of the abutments had moved. An investigation indicated that while one abutment rested on hardpan or rock, the other which is about 18 ft. high above water level and of unknown depth below probably rests on a fine sandy loam. The stream channel shifting had eroded this deeply. To prevent further erosion, two lines of piles were driven near the toe of the slope and the whole slope covered with a heavy riprap of stone. Holes were drilled through the abutment near the juncture of the wings and face, and 14-in. rods, 34 ft. long, were Abutment Retained by Concrete-Incased Rods Passing Through Concrete Deadmen placed in trenches running back along the roadway. These rods were incased in concrete for their entire length and anchored to concrete deadmen at the end. To further strengthen the abutment, a narrow trench was dug along its back and filled with concrete the extra width providing 54 The Engineer in Field and Office sufficient breadth to receive the bearing when it was replaced in its proper position. When the concrete at this point was hardened sufficiently, the nuts were tightened on the tie-rods, a new anchor bolt hole drilled and the bolt grouted in. To lift the bridge while widening the seat beams, struts made of pine logs flattened on two sides were attached at the first panel points, as is indicated in the line drawings. A beam rested on top of these struts and passed under the top chord. The bottom chords, stiffened by means of 4x6 timbers inserted between the floor-beams, completed the provisions against reversed stresses. Arches Destroyed by War Replaced Quickly The exigencies of battle in the north of France have required the rapid and stable reconstruction of a number of masonry arch bridges that had been more or less completely destroyed by the German or by the Allied forces. These bridges are generally in an area where timber and cut stone are scarce, and their reconstruction must be done rapidly without the aid of the needed quota of skilled artisans. To meet these conditions, concrete arches placed without the use of falsework have been successfully employed in a number of cases. Method of toying Up Arch Fig. 1. Restoring Arches in Destroyed Stone Bridges in Battle Area of France Cement can more readily be brought forward than any other structural material, and sand and gravel are local products, so that concrete, which can be made by unskilled labor, is doubly effective for such work. A novel feature of the reconstruction is the use of old iron and a minimum of timber for arch centers, which can be readily erected, thus saving time and labor. Bridges and Dams 55 The first operation in the reconstruction of one bridge was to build the light timber framework carrying the footway and erect thereon the towers for a construction cableway. From this cableway a series of centering ribs made up of old steel rails was placed. These rails, which were found in the neighborhood, weighed 60 Ib. to the yard. They were cold bent to the proper curve, in two sections and spaced 20 in. c. to c. clear across the arch. At the abutment they were bolted to a bedplate that was held by a hook bolt driven into the masonry. These curved rails were used as the basis of a thin concrete arch that in itself served as the center for the main arch. This procedure was adopted rather than placing the main arch immediately upon falsework hung from the steel ribs themselves, because the rails were not sufficiently strong to act as centers. Cross-Section of Falsework Center 60- Ib. Rail Bearing Of Kail . a+ Abutment Elevation of Falsework Center Fig. 2. Details of the Steel-Rail Arch Centering Used on Meurthe River Bridge The centering consisted merely of timber joists and a floor. This concreting was done in two parts, a 1:2:4 concrete was placed for its uniform thickness of 10 in. from abutment to abutment, and for the full width. On this a concrete arch rib and two abutment sections were first placed and the intermediate sections last. The concreting of this shallow section could be done in one morning. Ten days was allowed for this concrete to set. Meanwhile the top of the rib was laid off in 19 voussoirs, and a vertical dividing wall was erected across the arch at each voussoir division line. This dividing wall was made of a wire mesh, large 56 The Engineer in Field and Office enough to hold the aggregate, fastened to f-in. and f-in. vertical rods tied in at the bottom to hook bolts that had been left emerging from the centering concrete. These frameworks having been placed during the 10 days allowed for the setting of the centers, concreting was carried on across the arch rib in the voussoirs so laid out, placing them across the bridge so as to impose the least eccentric loading on the centering arch. The progress of voussoir deposition is shown. All this concreting for one 6-ft. arch could be done in two 10-hour days. After the main arch rib has achieved a sufficient set, the centering arch can be removed, although this is not necessary. Meanwhile, the superstructure of the arch can be erected in a continuous process following the construction of the main arch rib, and the roadway put into service in a minimum of time. Top Forms in Arch Concreting Top forms were considered necessary on the new Chesapeake & Ohio Northwestern Ry. work at Sciotoville, Ohio. It has long been usual in concrete bridgework to omit top forms near the crown of the arch where the slope is very small, but of late it has become rather common practice to omit these forms far down the arch rib slope and to attempt to screed the top surface to line. Concrete so placed is apt to pile up, causing an excess of arch-ring depth at the bottom of the section poured and a thinning of the ring near the crown. Therefore a top form for the haunch section of the arch ring, which extends from the springing line to a crown section placed in advance about 3 ft. on either side of the crown, was used on this bridge. A Small Bobtail Draw In building a swingbridge across a neck of Lake Lucerne at Stansstad, Switzerland, for 22 m. clear opening, the engineers for the Canton Nidwalden adopted the bobtail swing type but detailed it in such manner as to make substantially a single-leaf swing. The short arm ends just beyond the turntable, so far as the bridge floor is concerned. It extends out under the approach floor, however, as counterweight for the long arm. The approach deck over this counterweight has supports at one side only, the other side being left clear to allow the counterweight to swing out. The result of this arrangement is that rocking action due to live-load on the short arm is nearly eliminated, making it possible to dispense with wedging Or tight latching at the outer end of the long arm, as well Municipal Engineering 57 as with bearings under the short arm. The operating machinery is thereby simplified, and the required power reduced, which (the bridge being hand-operated) means that the time required for opening and closing is reduced. The end of the long arm has one lower and two upper track-wheel supports, and these enter between tracks inclined slightly upward, as the bridge closes. There are also two flat bearings, one under each girder, but they are adjusted to be barely in contact when no live-load or wind is acting. The turntable is center-bearing and has two side wheels and two at the quarter-points nearest the channel. Field Determination of Bridge Skew In reconnoissance work for ordinary highway bridges no great accuracy is required in determining the skew; even a variation of 5 being seldom important. A method of determining this skew in the field is to take an ordinary 6-ft. rule, hinged about its 3-ft. joint, place one arm in the line of the road and the other arm in the line of the stream; the distance between the two ends is the chord of the angle of skew of the bridge. The following is a table by which the angle can be determined : Radius equals 36 in. Chord 18 in. 20 in. 22 in. 24 in. 26 in. 28 in. 30 in. 32 in. 34 in. Angle 29 00 32 20 35 39 42' 45< 49 C 52 C 56' 40' 00' 20' 40' 20' 40' 20' Chord 36 in. 38 in. 40 in. 42 in. 44 in. 46 in. 48 in. 50 in. Angle 60 00' 63 40' 20' 20' 20' 20' 40' 67 71 75 79 83 88 00' Intermediate angles can easily be interpolated. Helps for Municipal and County Engineers Tile Drains Under Curbs in Syracuse The accompanying sketches show how the City of Syracuse drains all pavement foundations. The curbing is set in a block of concrete 18 in. wide and 12 in. deep. Under this block of concrete is 18 in. of cobblestone filling, along the edge of which is laid a 3-in. tile drain which empties into the sewer catch-basins. Wherever there are street railway tracks a 5 x 16-in. Medina curb or header is placed against the ends of the ties and a similar construction, as shown in the right-hand sketch, is used. It is said that in the 58 The Engineer in Field and Office spring, when the frost is coming out of the ground, these drains empty a continuous flow into the sewer catch-basins, and there is MZDINA CURB Fbrvemen-f- 3"Tile- Standard Drainage Details for Concrete Curb (Left) and Special Curb at Railroad Crossing (Right) no question but that they save many pavement troubles. They are used in all kinds of subsoil and beyond the price of the tile they add practically nothing to the cost of the pavement. Types of Drinking Fountains Tested Tests of 77 drinking fountains of 15 different types showed that due to improper design all were possible sources of infection of the users. No less than 80% of the fountains and the water from 11% of them contained streptococci, although none were These Fifteen Types Showed 80% o/ the Fountains and 11% oj the Water Samples Infected found in the water supplied to the 18 buildings in which the fountains were located. The infection was due to contact between the lips of users and the structure of the fountains, or to water falling back from lips Municipal Engineering 59 to fountains, owing to vertical discharge of the jet. To keep lips away from the fountain structure and water from falling back on it, and to prevent water from being retained in fountains with Three Tests of This Protected Drinking Fountain Showed No Streptococci Infection cup- or ring-shaped depressions, the fountain shown herewith was designed. Three bacterial tests showed no streptococci on either the fountain or the water discharged from it. When To Haul, When To Waste and Borrow A simple formula by which to calculate the economical length of haul beyond which it is preferable to waste and borrow may be developed as follows: Take two adjacent sections, one in cut, the other in fill, and each containing the same volume of material V, measured in excavation in both cases. Under this condition the material taken from the cut will just make the fill, and therefore, provided the haul from cut to fill is of a certain length, the total cost of grading the two sections, with all the material from cut used in fill, will be the same as by the system of borrow and waste. These two conditions may be expressed by the equation V(a + cte/100) = V(a + 6) + Vc in which a equals cost of excavating and loading, in cut, per cubic yard; 6, cost of hauling and dumping wasted material, per cubic yard; c, cost of borrow and fill, not rolled, per cubic yard; d, cost of hauling and dumping material taken from cut to fill, per cubic yard hauled, and x length of haul, center of gravity of cut to center of gravity of fill, in feet. 60 The Engineer in Field and Office Eliminating V and a from the equation and reducing, = c whence x = 100(& + c)/d Very likely d will be found practically constant for the entire job, but b and c will need to be estimated separately for each cut and fill, and will no doubt show considerable variation. Irregular Street Intersection Area Calculations It is almost always necessary to compute the areas of some very irregular street intersections at the time of their improvement. Unless some planning is done beforehand, the proper measurements, which will result in a determination of the area with any degree of precision, are not made in the field. The accompanying diagram of a hypothetical intersection serves to illustrate the use of a special Computations of Areas of This Nature Simplified curve by which the calculations are simplified, and its employment will make the fieldwork as simple as possible. Only a very brief explanation will make clear the proper procedure to follow in the field. Let us assume that the areas of H, I and / are desired. While, of course, it is not exactly true that the curves at the intersection are always arcs of circles, they may be considered so for practical Municipal Engineering 61 purposes, and very close results will be reached if proper averages are taken. To compute these areas, the average radius and the internal angle must be determined. Since the ratio of T to E is the same \ 20 60 90 30 40 50 60 70 Values of ^ in Degrees Enter This Curve with Average of Linear Measurement To Get Average Angle as that of C to 2M, as may be seen from the small diagram given with the curve, the one curve is all that is necessary. To determine, for instance, the area marked J, the procedure would be as follows : l = ff r2 - 91 C 23.8 With this ratio, using the curve, -^ = 43 . * . A = 86 5.07 Average 2.53 equals "Ratio." A Then or 16.0 = 16.0 tan 43 0.933 11.9 11.9 sin 43 0.680 or 17.2 or 17.5 34/7 Average 17.35 62 The Engineer in Field and Office Area inclosed by the tangents and the radii, 17.35 X 16.0 equals 277.5 Area of segment of circle, 86/360 X X 17.35 2 equals . 226.0 J 51.5 sq.ft. 5.7 sq.yd. If the curve is considered as a parabola, the area might have been calculated in the following manner, but it does not seem at all certain that the result is as near the actual value as is obtained by the method outlined above. Area- 2 X Area = ^= 43.7 sq.ft. J = 4.85 sq.yd. It may be seen that the first method is more likely to give a better result than the one based upon the curve being a parabola, because more measurements are utilized. However, either assumption will simplify the computation over the usual method. Sidewalks Flushed Over Tops of Parked Automobiles The downtown sidewalks as well as the streets in Chicago are flushed every night. To expedite the sidewalk work, there has Extension Pipe When Not in Use Folds Over Top of Tank recently been added to one of two tank trucks a pipe extension to clear parked automobiles. It is connected to the discharge line Municipal Engineering 63 from the pump with a stuffing-box joint and has a knee brace of the same 2i-in. pipe as the remainder of the line. A single outfit can flush the whole 72,000 sq.yd. of area in the territory covered in an 8-hour night at a cost of 20c. per 1000 sq.yd. It is usual, however, to let the two trucks work at the job, although from the Parked Automobiles Are Not Disturbed by Sidewalk Flusher second one the hose must be dragged behind vehicles near the curb. The remainder of the night's work is to flush the 120,000 sq.yd. of pavement. The fan nozzle is 6 in. wide and delivers a stream i in. thick. It is played toward the gutter and carries the dirt and surplus water ahead of it. No complaint of damage to walks or calking has been made. When the second tank is equipped, the connection will be made near the front of the tank, so that there will be no danger of bending the pipe, should the truck run it into something when half-way extended. A Modern Intersection for Paved Roads It is appreciated by automobile drivers that the loss of momentum and time due to slowing down to turn a 90 corner within the limits of a 50-ft. highway is about equivalent to that required to travel some 1500 ft. on a tangent of similar grade. Another difficulty with the usually designed intersection is the lack of sufficient clear-sight to enable the driver to realize the approach of a motor on an intersecting road. To overcome these difficulties, the design shown in the accompanying illustration has been put forward; and as may 64 The Engineer in Field and Office be seen, within the limits of the proposed highway the clear-sight is 132 ft., though in reality it would be much more than this if the regulations concerning vegetation were enforced. While the total area of pavements in the proposed design is 1703.6 sq.yd., only 739.6 of this is necessary for the curve connection shown, the remaining 954 sq.yd. constituting 482 lin.ft. of main-line pavement, 18 ft. wide. Pavements 18 ft. wide being ample for two A Construction*! Proposed Intersection Gives 136-Ft. Clear-Sight lines of motor traffic at high speed, they need not be made still wider on curves where speed is necessarily reduced. Likewise, there would seem to be no good reason for widening 10-ft. pave- ments on curves; 15- or 16-ft. pavements, however, should be widened to 18 ft. Brick and concrete pavements are generally crowned 2 in. The preservation of this crown at the intersection would prove ob- jectionable, so it should be reduced at the intersection to about i in. and tapered out to the full crown in a distance of about 10 ft. in each direction. Sections like G, H, J, K, L, P, Q, R should be built monolithic with the usual convexity of surface at /, K, although J is depressed. Municipal Engineering 65 Areas like H, J, K will come out warped surfaces easily built by an experienced contractor. The 10 construction joints shown should be nothing more than planes of cleavage. Sections like K, E, F, J, which are built last, have their corner elevations fixed by the main pavement. The usual convexity of surface is preserved, therefore, and the inner edge F, J is depressed to meet the required elevation. In areas like K, L, E the surface of the ground should be kept about 1 in. below that of the surface of the pavement adjacent. The catchbasins and drains will keep the ground dry. Setting Street-Corner Radius Stake In setting radius stakes for street corners where the angle of the intersecting streets is not a right angle, the solutions shown in the following sketch save considerable time in finding the correct location for the radius stake. Two methods may be used: (1) with transit at E, the inter- section of the curb line and the center line of the intersecting Radius State u t. ,i Diagram for Setting Street-Corner Radius Stake street; (2) with transit at F, the intersection of the center line of the two intersecting streets. Considerable time and instrument work are saved by using the method with F as instrument point. For method 1, transit at E, the formulas are: x = tan A = sin B R y = R cot i B z = R sin A 66 The Engineer in Field and Office For method 2, transit at F, the formulas are: = x a tan o = sinB -^- tanB R + M y = R cot\ B L = x+ y + R + M sine Street Assessments in a Hillside Town It is believed that the results obtained by this method for making street improvement assessments in a hilly town with crooked streets and irregular lots are more nearly equitable than those obtained by the ordinary front-foot method. Fitting Assessments to a Steep, Crooked Street Arrows indicate upgrades. Heavy line shows assessment district boundary. Frontages given in feet and tenths The following factors are taken into consideration: (1) Front- age; (2) area; (3) assessed valuation; (4) the benefit or damage to each lot by virtue of its new position with relation to the graded street; (5) correction for over-assessment, where the rates would be excessive. The method of making the assessment is as follows : The entire cost to be assessed is distributed in amounts directly proportional to (1) the frontage, (2) the area and (3) the value as shown in Columns 5, 6 and 7 of the table. The mean of these values is then Municipal Engineering 67 computed (Column 8). To this mean a benefit or damage factor is applied varying with the relative position of the lot with reference to the street surface before and after grading. The figure obtained in Column 8 in the case of lots in Class 1 (considerably damaged by grading) are multiplied by ; lots of Class 2 (slightly damaged), by ; lots of Class 3 (neither benefited nor damaged), by 1; lots of Class 4 (slightly benefited), by $; and lots of Class 5 (consider- ably benefited) , by i. The results after this operation are shown in Column 10. A wider and more refined variation could be applied in cases where the conditions warrant it. The sum of the quantities in Colmun 10 is seldom equal to the total cost to be assessed, so that the difference between the sum of Column 10 and the desired total is distributed in proportion to the values found in Column 10 (added in the case of the example). The corrected quantities are given in Column 11. An inspection of Column 11 shows that Lots 66, 68 and 69 are assessed an amount in excess of 50% of the tax assessor's valuation a percentage which has often been held to be a confiscatory rate. These amounts are then reduced to a sum slightly below 50% of the value, and the excess is then distributed at a uniform rate over the remaining lots, giving the final values shown in Column 12. The example given is for the improvement of Central Ave. shown on the diagram (Fig. 1). METHOD OF COMPUTING STREET IMPROVEMENTS ACCORDING TO FRONTAGE, AREA AND VALUE, WITH BENEFIT FACTOR ADJUSTMENT Lot Frontage, Area, Ft. Sq.Ft. Value , Cost Distributed According to > Frontage Area Value (1) (2) (3) (4) (5) (6) (7) 16 20 59.5 84.0 6,900 12,000 $1000 800 $152.26 214.96 $177.94 309.45 $211.05 168.84 21 0.0 23,500 1000 0.00 606.01 211.05 66 197.6 7,100 400 505.66 183.08 84.42 68 80.0 4,000 300 204.72 103.15 63.32 69 179.8 3,300 300 460.10 85.10 63.32 Mean of Class of Col. 8 with Col. 10 Final Cost Lot Cols. 5, 6, 7 Benefit Benefit Factor Adjusted Distribution (1) (8) (9) (10) (11) (12) 16 $180.42 1 $120.28 $126.37 $136.08 20 231.08 3 231.08 242.33 260.95 21 272.35 4 317.73 332.97 358.20 66 257.72 3 257.72 270.17 199.00 68 123.73 5 164.96 172.94 149.00 69 202.84 2 169.05 177.44 149.00 Home-Made Portable-Pump for Manholes A portable centrifugal-pump outfit to unwater manholes has recently been put together in the shops of the Lincoln Park Com- mission, Chicago, for use in draining the electrical underground 68 The Engineer in Field and Office distribution system. The equipment is mounted on low wheels and can be hauled by an automobile or a truck. The 12-hp. 1500-r.p.m. marine-type gasoline engine is direct-connected to a centrifugal pump rated at 500 gal. per min. at 2000 r.p.m. Accessories to the pump and engine are a 23-gal. gasoline tank, 5-gal. automobile radiator, a 20-gal. circulating-system tank and a 66-gal. primary tank. With the available speed, 4-in. intake and 3-in. discharge hose, 300 gal. per min. can be pumped against a 15-ft. head, including suction and discharge, but water may be lifted 30 feet. Control of the entire outfit, including the spark and gasoline levers, the outlet and air-vent valves and the valves on the circulating system, are all grouped at the front end of the machine within easy reach of one man. Means have been provided for obtaining easy access to any part by making the inclosing walls either hinged or entirely removable. The pump has proved its value in pumping dry a line of ten 3-in. ducts that lie along the shore of Lake Michigan and are arranged to drain from one manhole to another. A detailed statement of the cost of constructing this outfit in the shops of the commission is given in the accompanying table. COST OF BUILDING EMERGENCY WATER PUMP MATERIAL 1 marine engine and muffler $90.00 1 radiator 25.00 To build truck for engine pump and wheels 225.80 To install engine on pump truck 69.07 Painting pump truck 21.54 To supplying suction hose, valve and fittings 125.89 Three IJ-ft valves 2.82 1 cap .10 12 bolts, various sizes .12 5 pipe plugs, various sizes .36 5 bushings, various sizes .45 12-ft. water pipe, various sizes 2.34 28 nipples, various sizes 2.24 4 tees, various sizes .30 6 unions, various sizes .99 56 screws and nuts, various sizes 1.00 19 ells, various sizes 1.61 6 padlocks, various sizes 660 Miscellaneous material 24.13 Total $599.86 LABOR Mechanic, 31 hours at 50c " i a A ft -^ ^i^n "^ ; ;; The Engineer in Field and Office to the first time interval of the flood where O is zero. The operations may be carried out in tabular form if desired, without drawing the curves. When the accumulated-outflow curve has been completed in the manner described, the desired outflow hydrograph of the stream may be readily obtained from it. Small Wells Help Drain Irrigated Land Rising wells or boreholes sunk below the level of tile drains may materially assist the drainage of irrigated lands that are underlain by shale. This applies particularly to such lands in the Rocky Mountain States, where the water in the shale is under pressure and where methods employed in other sections of the country have not proved successful. Deep tile drains are required, not less than 6 ft., and as much as 8 ft. deep in many cases. The depth to the water-carrying strata, however, is much too great for ditching; and as pressure conditions exist in these strata, the sinking of wells permits the water to rise into the drains. Owing to the character of flow in shale ground, the area of influence of a relief well is not large, and from two to six wells per 100 ft. of trench may be necessary. Their maximum depth is usually 20 ft. below the drains, and in general the flow is encountered at a depth of about 15 ft. In most cases a 2-in. hole is sufficient to care for the water. The well should be located near the end of a tile, and one end of the tile then chipped so as to leave a 2-in. hole over the well. If the banks will not stand and the wells must be driven as the tile-laying progresses, the well should be a few inches to one side of the drain and connected to the hole in the latter by placing a half -tile over the well. If the well is directly under the tile, it is likely to be closed by sediment washing into it, especially if the flow is weak. Modified Pitot Gives High Accuracy The pitot tube has long been used as a device for measuring the flow of liquids and gases, but only when used with the utmost care have the results proved uniform. Many experimenters have worked with modified forms in the endeavor to reduce the variation in results, but it is evident not only that the data obtained are variable in the hands of different men, but that the same tube may have different coefficients. Irrigation and Hydraulics 103 In order to correct this latter defect the "Hydraulic Shunt-Flow Tube," was devised. This device is a tube so arranged that it may be introduced into the stream with the tip directed against the flow and yet maintain at the tip the same pressure that existed before the introduction of the tube. The water flows into this tube and may be shunted into a small container and weighed, leaving the velocity undisturbed from the normal. Valve Common' Pitot Tube FIG. I Formula: V' H* Pitot Head. C Coefficeni- a Acceleration of Gravity Hydraulic Shun* (Modification of tfie Pilot Tube) Formula: FIG. 2 Q* flow from Tube A * Area pf lip Opening ( Tip Coefficient Almost as Simple as a Pilot Tube The velocity of flow at the tip of the tube will be equal to the quantity of water collected in the measuring tank, in a measured time, divided by the area of the tip all quantities being measured in the usual units. It is possible to demonstrate mathematically that turbulent flow should not affect the coefficient of the tip. Theoretically, of course, the tip coefficient should be unity under all conditions, but a series of experiments undertaken with this in view show that it varies less than 1% the pitot coefficient under like condition varying more than 4%. Dams of Boulder-Filled Wire Baskets Two hydraulic dams recently built in California consist of units, or baskets, of poultry netting filled with coarse gravel and rock These units are built in place and are laid like shingles on a roof. Into the netting are passed very coarse gravel and rock up to the size of a man's head. When a sufficient quantity has been deposited, the top is leveled off with a straightedge, and the selvage edges of the 104 ."."'; The Engineer in Field and Office netting are drawn together by means of a piece of strap iron with a hooked end. The selvage edges are then fastened together with wire, and the ends are folded in and similarly fastened. Discharge Plotted with Novel Curve A quick and easy method has been developed for showing the variations in quantity of water passing a given point in a stream. The common method of obtaining a continuous chronological record of the fluctuations of the height of water in rivers is by means of a gage with a clock and float attachment. This curve is used in connection with what is known as a "station-rating curve" to determine the amounts of discharge. The usual method of plot- ting a discharge curve is by reading from the station-rating curve the quantities corresponding to critical points on the curve of gage heights and then plotting these results against time on a separate sheet of cross-section paper. The method herein described permits plotting the discharge curve on the same paper with the curve of gage heights. It also eliminates the time consumed in reading the gage heights, then referring to the station-rating curve to obtain the discharge, and finally plotting that value. This method presupposes the use of cross-section paper on the recording gage, as is usual with those gages which have a rectilinear rather than a circular movement of the recording pencil. It is necessary, first, to plot the station-rating curve with the same scale of heights as is found on the gage-height record. Then draw a straight line from the origin at an angle of 45 with the axes. Next, cut out with a sharp knife the sections lying between the station-rating curve and this 45 line. Finally, paste a piece of cardboard at the bottom of this curve, keeping the lower edge straight and parallel to the horizontal axes. This station-rating curve is then superimposed on the gage- height record so that the horizontal lines representing the same gage height coincide. A scale of quantities precisely like that on the station-rating curve is then laid off on the vertical axis of the gage-height curve, the zero of this scale starting at the origin of the station-rating curve. To plot the discharge curve, all that is now necessary is to move the station-rating curve horizontally along a straight-edge until it intersects the curve of gage heights at any point A. Then follow the vertical from this point until it in- Irrigation and Hydraulics 105 tersects the 45 line. This intersection A' is a point on the discharge curve. Repeat this operation for all critical points A and then connect the points A' with a smooth, free- hand curve. This gives the discharge curve as in Fig. 1 plotted on the same paper with the gage-height rec- ord and corresponding to the time scale thereon. To determine the total dis- charge, planimeter the area under the discharge curve thus plotted. That this method is cor- rect is due to the selection and arrangement of the scales of height and quan- tity and to the fact that the ordinate of any point on a straight line passing through the origin and making an angle of 45 with the axes equals its abscissa. The operation of this method of plotting will be facilitated if the sta- tion-rating curve is made on tracing cloth and with no vertical or horizontal lines except the axes. The little extra time required in the preparation of the sta- tion-rating curve is many times recovered when plot- ting the discharge curve, especially if, as is often the case, the gage-height curve extends over a long period of time. 106 The Engineer in Field and Office Movable Flume on Hydraulic Fill Dam During the construction of a large hydraulic fill dam recently completed in the West, considerable time was saved by the use of an inclined runway for the flume from which material was deposited on the dam. Although the flume was moved by hand, it was only necessary to interrupt the flow for short periods while the delivery flume was skidded along the incline to the desired new position. Three borrow pits were used, from which, by means of hydraulic giants, material was sluiced into three main flume lines. From these mains the material was conveyed through flumes along the upstream and downstream sides of the fill. By the use of gates the streams were discharged from the flumes at the desired intervals toward the axis of the dam. The flume box, 12 x 24 in. in section, was built up of li-in. boards and was paved with 6 x 12 x 6-in. hemlock blocks set on end. Im- mediately above the blocks 1 x 6-in. projecting strips were nailed on either side. This box was supported on 2 x 6-in. stringers, which in turn were carried by 4 x 6-in. caps on the low sliding bents spaced 15 ft. apart and inclined on a slope of 5 to 1. The 6 x 6-in. inclined caps were supported on pieces of the same size resting on the ma- terial previously deposited and tied together by cross-bracing. The flume proper was moved up the incline by means of a lever and chain device at each bent. The lever consisted of an iron bar near the lower end of which was fastened a chain connecting with the framing of the flume box. At the extreme lower end of the bar a second chain was attached which passed over an iron claw fastened to the upper end of the inclined 6 x 6-in. cap. This lever was operated by one man at each bent. With the lever in an upright position, pulling it down through the quadrant, which it was possible to describe, would move the flume up the incline a few inches. This gain was then caught by taking up slack in the chain at the claw, and the operation repeated. Short lengths of lateral flumes were attached to the openings in the upstream side of the flume box to facilitate control of the flow. Doors for closing the openings were provided so that material could be discharged at any desired point along the crest of the fill, a duplicate line of flume being used along up and down stream faces. Of course each time the flume was moved it was necessary to establish a new connection at the point where it was fed from the supply flume. Sluicing was carried on practically continuously night and day in order to save time, and the movable feature of the flume was considered to be a factor in the progress made on the job. Railway Civil Engineering 107 Hydraulic Elements of Semicircular Flume The accompanying curves give the hydraulic elements of semicircular flumes, partly full, in simpler form than curves before published. They can be platted in a few minutes from the elements UVl' 01016 0.015 0.014. 0.013 0.012 0.011 6 o.oio Z 0.009 E0.008 C 0.007 \ \ \ V \ \ \ \ t* r = AREA OF WETTED SE6M SQ HYDRAULIC RADIUS, FT. FLUME NUMBER ENT FT \ \ \ \ "" \ \ , 1 \ Vs z. & " S \ 0.005 0.004 C.003 0.002 1 c \ \ \ \ \ \ \ \ \ \ \ Z 03 0.4 05 OJ6 0.7 05 0.9 LO U tfc 13 U 15 16 Curves for Semicircular-Flume Computations of circular segments of radius unity, tabulated in almost any book dealing with the flow in circular conduits, by using the commercial number of the flume, which is the length of the curved flume sheet in inches, instead of the diameter. The wetted perimeter, which is seldom used, can be easily found from the area, and the hydraulic radius or the curve can be drawn if desired. Practical Pointers for Railway Civil Engineers Accurate Stadia Profile at Low Cost A stadia profile was run last summer over the Illinois Central R.R. from Champaign to Centralia, 111., a distance of 125 miles, for the railway-engineering department of the University of Illinois. Top-of-rail elevations were desired at intervals of 300 ft. for use in connection with grade-resistance investigations. The requirements that the stations should be tied-in to the mile posts and that the 108 The Engineer in Field and Office linear error in any mile should be not more than 10 ft., which is as close as the mile posts can be indicated on the dynamometer-car charts, were easily met. Using 300-ft. sights, there would be 18 readings to each mile; hence, assuming an accuracy in the stadia readings of 1 in 300, the probable error in the length of a mile, according to the method of least squares, would be 1/18, or 4.35 ft., which is well within the allowable limit of error. This made it possible to use the stadia instead of the usual method of chaining. In starting the work the rodman held the rod on the top of the rail opposite a mile post, and this point was recorded as Sta. 0. The levelman then paced a distance of 300 ft. and took a stadia reading K) 20 JO 4O 50 60 70 SO 9O 100 HO 120 150 MILES Compensating Tendency in Linear Errors of Stadia with the instrument only approximately level. If the paced distance was within 5 ft. of that indicated by the stadia sight, the instrument was finally leveled, and accurate level and stadia readings were taken. Any error in excess of 5 ft. was corrected by making a new set-up. The rodman then advanced to Sta. 3, approximately opposite the instrument, and a level reading was taken. The rodman then paced 300 ft. in advance and gave the levelman a trial reading for distance. The levelman signaled the correction necessary to locate Sta. 6, which point was marked on the rail with keel and used as a turning point. The process was repeated at distances of 300 ft. until Sta. 48 was reached. The instrument was set up as nearly midway between Sta. 48 and the mile post as could be determined by eye. Readings to Sta. 48 and to the mile post were made and added to the distance previously covered. In this way the distances between mile posts for 122 of the 125 miles were obtained. Railway Civil Engineering 109 The distances between mile posts as given on the Illinois Central's official profile were assumed to be correct, and the differ- ences between these distances and those obtained by stadia meas- urements were considered the errors in the stadia measurements. In very few instances were the mile posts 5280 ft. apart, as indicated by the railroad chaining. The distances between posts varied from 5265 to 5285 ft.; hence, not knowing what reading to expect, the observer was entirely unbiased in taking the last reading in each mile. In every case the stadia measurement was made and recorded before the chained value was taken from the profile. The maximum error in any one mile was 14 ft. In 7 miles there was no error; in 67% of the miles the error was 5 ft. or less; in 96% it was 10 ft. or less, and the average error per mile was only 4.5 ft. conforming closely to the probable error of 4.35 ft. expected accord- ing to the method of least squares. As some of the errors were plus and some were minus, the accumulated error at no point was very large. The total error at the end of each mile has been plotted in the accompanying curve, and it is seen (1) that the largest error was 60 ft. at the end of the fiftieth mile, (2) that the error at the end of the one hundred and twenty-second mile was 12 ft., and (3) that at eight points along the line it was zero. Laying Out Crossovers Between Curved Tracks Crossovers between non-parallel curved tracks can hardly be located by any exact mathematical formula. Cut-and-try methods on paper on a large scale are the main recourse; but it is very difficult to lay out the flat curves without resort to geometrical con- struction, the lines of which would cover the paper if many trials Offset- Scats \ Frocr Fbrrrf^ ' Tangent to Curve? Templet Aids in Paper Locations of Turnouts were made. A transparent templet like that shown can be used to advantage. The base line can be set tangent to the curve by the offsets, the circle being placed at the desired frog point, and the line of the frog can be laid or pricked off. Thus a number of trials can be made without confusing the drawing by any consider- able number of lines. 110 The Engineer in Field and Office Climate Should Govern Hump Profiles Three hump profiles for gravity railroad yards, designed respec- tively for cold, moderate and warm climates, were adopted on the recommendation of the Committee on Yards and Terminals at the recent convention of the American Railway Engineering Association and will be substituted for the profile now in the "Manual." As the drawings indicate, 4% grades, reduced to 1% over the track scales, are recommended for cold climates while for more favorable climates the grades should be lightened or shortened, or both. NQ.J, FOR COLD CLIMATE NO. 2, FOR MODERATE CUMATE WftJ. ft?/? VWffM CLIMATE Colder Climates Require Steeper Grades The committee points out that the problem cannot be solved precisely, since the speed developed by cars depends not only on the type of car, but, with the same type, on the length, whether loaded or empty, the lubrication, efficiency of maintenance, tempera- ture, time standing before being pushed over the hump, head winds, care with which the tracks are brought to and maintained at the profile grade and, lastly, the timidity or assurance of the car rider. Railway Civil Engineering 111 In the estimation of the committee there are these seven salient features in the design of a hump: 1. A short grade steeper than the approach grade to bunch cars. 2. A level grade over the summit, these grades to be of such length as to form the tangent to a reverse curve connecting the approach grade with the first descending grade. 3. A short grade from the summit, to separate the cars quickly and give them the desired speed for weighing. 4. A light grade over the track scale. 5. A moderately steep grade, the rate and length depending on the traffic, to the end of the ladders. 6. A grade through the ladders sufficient to maintain the speed throughout the turnout. 7. A light grade that will just overcome in the length of the body tracks the speed already acquired. Rails Located Before Concrete Is Cast The sketch shows a method of setting U-bolts to hold down track rail by which much time was saved over spacing the rail with templets in constructing track for ore bridges at a Buffalo dock. Filler LEO WALL Appro*. 10'- >^ of Dock Rail Holds Anchor Bolts While Concreting Altogether, 2400 ft. of 125-lb. rail was laid in this way. It is anchored by 1-in. U-bolts 24 in. apart, the ends of each bolt passing through a bedplate on which the rails rest, and the nuts on each side holding 112 The Engineer in Field and Office down cast-iron clips that retain the rail. To space these anchor bolts, templets of 2-in. pine were first used, but the method was abandoned after the completion of the first 120 ft., because it delayed concreting and there was difficulty in maintaining the bed- plates at the proper elevation. The method shown in the sketch, which involved hanging the rail in place before concreting, was substituted. A pipe spreader is used at each anchor, as shown in the sketch, to hold the bolt, bedplate and clips in position. The rail is leveled by the wing-nuts at the top of the hanger rods and lined by slipping the block under each of these nuts one way or the other, before concreting. While the con- crete is still green, the top of the rail is checked for grade. Short 1 x 4-in. spreaders from the form studs to the web of the rail hold the latter to line. Making Borings for Embankment Subsidence Borings were made to test the subsidence of embankments in connection with the Government valuation of the Chicago, St. Paul, Minneapolis & Omaha Ry. The tools used on the work were a carpenter's 2-in. auger, a 2-in. pod auger, a short drill, two bars, 5 and 8 ft. long respectively, both of 1-in. drill steel, a supply of 2-in. single-strength pipe and couplings for casing the hole in soil that u-"' Extension Bar for Boring Auger caved, a pipe lifter and pipe holder, wooden mauls for driving the casing, a shovel and post-hole digger for use in going through the ballast, a short piece of heavy log chain with hook and eye, two pipe wrenches and a supply of extension rods. The augers and drill each had a shank 4 ft. long with an eye at the upper end so that an extension rod could be hooked on as the hole was lowered. These extension rods were of i-in. round steel, 8 ft. long, with an eye at Railway Civil Engineering 113 one end and a hook at the other. As some of the holes were more than 50 ft. deep, to unscrew joints every time the auger was pulled up for cleaning would have been slow work. With the hook connec- tions the rods were disconnected as fast as they were pulled up, and there was no chance for sections to separate while in the hole. There was one rod 4 ft. long to use in connection with the longer ones, so that there was never more than 4 ft. out of the ground at a time. The method followed in making these tests was to put a hole through the ballast with the shovel and post-hole digger, then set in a length of the 2-in. casing and use the auger the rest of the way, lowering the casing as the hole progressed. If the fill was of clay or any material that would stand without caving, it was often possible to complete the hole with only the one piece of pipe, as that would keep the ballast out of the hole. Where the fill was dry sand, or if there was water on the sides of the fill, it was necessary to keep the casing close to the bottom of the hole; and in some holes better progress was made with the casing driven lower than the hole, the material being bored out inside of the casing. In a number of holes small gravel stones were found which caused a great deal of trouble, until the writer had a 3-ft. piece of H-in. pipe (the largest that would go inside the 2-in. casing) fitted with an eye at one end so that the extension rods could be hooked in. This pipe was churned up and down in the hole until the gravel had become wedged in the pipe. In this way any stone that would go in that pipe could be removed. At times the stones were too large to be removed in that way, and it was necessary to drive the casing down until the stone was wedged in it. The casing was then pulled and cleaned. In some holes it was possible to replace the casing without losing any of the hole, but at other times it was found that from 5 to 50% of the hole had filled and would have to be bored out again. How Softened Concrete Lining To Tunnel Will Be Repaired Plans have been made to repair the concrete lining to the Cascade tunnel of the Great Northern Ry., which has softened due to the formation of sulphur compounds from sulphur in the locomotive gases and the cement. As the lining was put in principally to protect the rock face against disintegration and not for strength, and as the steam locomotives are now superseded by electric trac- tion, it is thought that repairs made now will be permanent. 114 The Engineer in Field and Office The proposed method of repair is as follows: (1) Clean the entire area of sidewalls and arch, removing all disintegrated por- tions. After cleaning, the surface of the concrete is to be sprayed with an alkaline solution for neutralizing the acid in the concrete. (2) Drill 2-in. holes where necessary to provide additional drainage. (3) Drill 4-in. holes in the arch where necessary and fill cavities existing back of the concrete lining with sand filling or grouting. (4) Replace all disintegrated portions of the lining with a coating of concrete by a cement gun. The specifications provide that where the coating is over 3 in. deep it shall be supported by wire mesh cut to fit and fastened to the old concrete by spikes driven into holes drilled in the concrete 24 in. apart. The mixture for use in the cement gun will be one part portland cement and 3i parts sand, mixed dry. Turntable for Handling Relay Rails The turntable shown in the accompanying sketch is proving a big labor and time saver in handling rails for the new car-repair yard of the Pennsylvania R.R. now under construction at Greenville, New Jersey. The 85-lb. rails used are second-hand, and the ball of each is badly worn on one side. Since it is therefore necessary to place the un- worn side on the inside of the track being laid, it happens that many r Device Saves Much Time in Handling Old Rails of the rails have to be turned end for end before placing them. Previous to building the turntable it required considerable maneuv- ering by a gang of at least six men to turn one rail. With the turntable, however, which is set up about 18 ft. from the track being laid, two men can turn a rail with ease. The device was made complete for $8. Railway Civil Engineering 115 Open-Tank Creosote Treatment of Timber Open-tank treatment of timber is desirable for interurban and the smaller steam railroads that have a number of timber bridges and other timber structures to maintain. Such a plant is convenient for treating fence posts, paving blocks and the like on very short notice. The Virginia Railway and Power Co. has operated an open-tank treating plant at Norfolk, Va., since May 1, 1914, using dead oil of coal tar from its own gas-works as a preservative. Water-gas tar was tried as an experiment for a few months and finally abandoned because of the small saving and its doubtful value. Yellow pine, mostly of merchantable grade, has been the only species of timber treated in the open tank, and has varied in size from 2 x 4-in. to 14 x 14-in. timber of all lengths. A number of pine poles have also been satisfactorily treated. The penetration obtained has been from 12 to 20 Ib. per cu.ft. of timber. Well-seasoned timber is desirable for open-tank treatment; in the case of green timber it is necessary to keep it in the tanks until it becomes well seasoned from the heated oil. The method of treatment is, first, to place the timber in the tank and weight it to prevent floating, and then cover it with oil. The steam is turned on for about eight hours, at approximately 100 Ib. pressure, the oil being kept at about 200 F. The steam is then cut off and the oil and timber are allowed to cool over night. The next day the timber is removed from the tank and placed on the storage piles by the derrick boom. The following figures give the actual cost of treating at this plant for one month. One foreman (who also operates the electric derrick) at $3, one fireman at $1.50 and four laborers at $1.50 per day are required, working under the bridge supervisor. A total of 39,098 ft. b.m. was treated. The costs were as follows: Cost per Item Total M Ft. B.M. Dead oil of coal tar. 7375 gal. at 6Jc $479.38 $12.23 Coal, 6800 Ib. at $3 per ton 9.10 .23 Labor, including foreman 83.50 2.14 Maintenance of plant 20.00 .51 Interest on $3000 investment 15.00 .38 Total expense for one month $605.98 $15.49 Average penetration, 19.6 Ib. per cu.ft. of timber. The drawing shows the layout of the treating plant. The smaller tank is used for treating only in emergencies. The dead 116 The Engineer in Field and Office oil of coal tar is brought from the gas-works in a 2200-gal. tank-car that is fitted with a section of pipe to allow filling the treating tanks directly. .--Tomb Car in Position -for Filling SPUR TRACK 1 \OPEN TREATING TANK 10'*4O \OHhecrtedbyrSfeam Coil covering Bottom of Tern If STORAGE TANK +-/80 ! 7 Jeo 50 |*> ^30 Ijjzo 10 I \ ... PffOf Q >ERCV NCRE V5/572 TEROA NCYFC DJM2G >R ff 7 ^ s / JH \ 1 1 \ 1 | \ v V/TH ONE-) ^ r nisa MLF7 MIS VEST "ENCY REN51 ABOU1 ~HISL 1ST | s I U) s | WITH THE USED IN 1 ""*" \ -^, SLOPPY CONCRETE SOMETIMES 'WAD WOP.K4NDINBUH.D1N6 CTION, TWO-THIRDS TV THREE- OF THE FVSSIBLE STRENGTH ' \OFTHECONCRETEISLOST > *. "***^ i FOL IRTH5 / g \ 70 80 90 100 110 120 130 MO 150 160 170 180 190 ZOO Per Cent, of Water giving Maximum Strength How Excess Water Affects Concrete Strength This curve is reproduced herewith. In it the ordinates represent the relative strength of concrete expressed as a per cent, of the maximum that can be secured from a given amount of cement and the same aggregate. The abscissas indicate the relative quantity of water used in the mix, considering the amount that gives the maximum strength as 100%. It will be noticed that there are no definite figures given, because the proper amount of water varies with the method of handling and placing the concrete, the condition of the aggregate, the temperature of the outside air and a number of other factors. As a rule, the amount of water that gives the maximum strength produces a mix that is too stiff for most purposes, though in cast-concrete products a mix even drier than that which gives the maximum strength is sometimes desirable, because the molds can Concrete Construction 129 thus be removed within a short time. The general conditions, however, are shown in the curve. It will be observed that concrete strength increases rapidly with the quantity of water up to the optimum condition. With any further increase in the amount of water there is a rapid falling off of the strength, so that with an amount of water about double that required for the higher strength the concrete has only about 20% of the maximum strength. In building concrete roads the consistency which is recommended corresponds to about 105 to 115% of that giving the strongest con- crete. The economies resulting from handling the concrete are more important than securing the maximum possible strength for a given amount of cement. It has been observed that many road contractors insist on using water varying between 130 and 200% of that corre- sponding to the highest strength, with the effect on the strength noted in the curve. Had this curve extended beyond the 200% water, it would have been practically flat, thus indicating that the maximum of damage is caused at that point. The proper quantity of water will vary with the quantity of cement and the size and grading of the aggregate, and to a less degree with the nature of the aggregate. The water required for a sand and crushed-stone aggregate is not appreciably different from that required of a sand and pebble mixture, provided the grading of the aggregates is similar. But there is no direct criterion for determining in advance the best quantity of water for WATER REQUIRED FOR ROAD CONCRETE , Mix N Approximate Mix as Usually Water Required Volume of Expressed (Gallons per Sack of Aggregate /^-Aggregate N Cement) Cement After Mixing Cement Fine Coarse Minimum Maximum 1 5 1 2 4 6 6J 1 4i 1 2 3 5f 6| 14 1 li 3 5| 6 13 1 11 2i 5 5i concrete being placed on a road. The concrete should be mixed so that only a small quantity of free water will appear on the surface after leveling and striking off, which gives concrete of a jelly-like consistency. The principal difficulty in the way of attempting to determine in advance the proper quantity of water is due to the fact that aggregates are generally damp, and the degree of dampness is not uniform. In the case of concrete made of sand and pebbles, or sand and crushed stone, well graded in sizes up to 1 in., the accompanying table indicates about the quantity of water that should be used for the mixes commonly employed in concrete roads. It is assumed that the aggregates are in a room-dry condition, which does not always prevail in work. 130 The Engineer in Field and Office Effect of Hydrated Lime on Concrete The Bureau of Standards some time ago started an exhaustive series of tests looking into the effect of hydrated lime on concrete, as regards strength, consistency, watertightness and workability. A preliminary announcement gives some conclusions, based on six months' tests, as to the effect on strength. The results should be used with caution and hold only for hand-mixed, tamped concrete, of rather dry consistency and stored in a damp closet. There is no evidence at present to show the effect of the hydrate on concrete which is either machine mixed, poured or aged in air, and interpretation of the present conclusion to obtain such information is not justifiable. The following conclusions have been reached: (1) The substitution of hydrated lime for cement causes, in general, a diminution of the compressive strength of the concrete. This is most pronounced with 1:2:4 concrete. (2) Hydrated lime causes a less diminution of strength in a 1 : 14 : 3 concrete than in a 1 : 2 : 4. Results indicate that the sizing of the ingredients may be one of the important factors in determining the value of hydrated lime in concrete. (3) At least up to six months, there is no appreciable difference in behavior between high calcium and high magnesian hydrate. (4) The diminution of strength caused by hydrate when 1: 14: 3 concrete is stored in air is not nearly so great as when the concrete is stored in the damp closet. Link Expansion Joints on Viaduct Measurements taken between extremes of temperature for a year on the linked expansion joints of the Colfax-Larimer viaduct in Denver indicate that all joints work as nearly equally as could be hoped for. The structure consists of alternate towers and suspended spans, the latter having lengths one-half the distance between bents and being supported by links upon the overhanging girders of the towers. With favorable results on the structure proper, on which to base the design of the joints in the pavement, the problem was to make provision for movement approximately every 80 ft., in such a way that there would be as little annoyance as possible to traffic. Plates or angles projecting in roadways are somewhat of a nuisance, and the thickness of the asphaltic pavement did not render a built-up joint feasible. The- joints adopted have undergone the test of an exceptionally severe winter in Denver, some being pulled apart at least I in. Concrete Construction 131 without as yet showing any need of repair; and as for ease of riding, a crack in an asphaltic pavement gives less shock to a vehicle than almost any other form of joint. One other point in the design of the viaduct having to do with expansion, which has caused some anxiety, is the laying of sidewalk slabs of 4-in. thickness directly upon the old concrete of the super- structure itself. Although these slabs are completely severed by transverse joints at intervals of about 5 ft., they are monolithic with the curb ; and there was some fear that, being exposed to the atmos- phere, their rates of contraction would be different from that of the M 18, Gal Steel 21" Wide before Elastic Joint Sidewalk Exp. Joint Exp j oint in Expansion Joints Are Different at Track, Roadway and Sidewalk concrete underneath, which difference might result in surface cracking and possible shattering after long exposure. There was some thought of painting the joint with tar pitch or asphalt paint; but it was finally concluded that this would be ineffective, since the concrete of the underlying superstructure, being more or less lumpy, the painting would not overcome the resistance of the mechanical bond. This possible trouble could have been avoided by making the sidewalk monolithic with the superstructure, but previous experience along this -line showed that it would be impossible to obtain either a surface to grade or a good alignment of the curbing. Double Shell Concrete Chimney A special arrangement of the interior shell and the use of adjustable steel forms to give an exterior taper are the features of the reinforced-concrete chimney at the plant of the Rockford Paper Box Board Co., Rockford, Illinois. It is 187 ft. high with an interior diameter of 7 ft. throughout. The outside diameter is 10 ft. 4 in. at the footing and 9 ft. at the top, with a taper of 0.056 in. per foot, commencing at the top of the breeching. The inner shell has been carried nearly to the top in order to protect the outer wall and to improve the draft efficiency 132 The Engineer in Field and Office If the gases are only slightly cooled and come in contact with a relatively thick portion of the outer wall, they heat its inner surface while the exterior may be exposed to very cold weather. As concrete is a poor conductor of heat, this condition results in severe internal stresses, and observations of many chimneys built in this way indicate that the most serious cracks occur about the level of the top of the short interior wall. I : : i ! Elevation at Breeching f .... 7 > ....>,- Plan at Breeching 22' 1 ^ Plan at Footing Concrete Chimney with Inner Shell Extending Full Height of Shaft To create a draft in the air space, in order to protect the outer wall from high temperatures and to prevent the collection of soot, openings in the outer wall are made at the base of the air space. The inner wall is made as thin as practicable in order to minimize unequal expansion and is 4 in. thick for its entire height. The thickness of the outer wall of the chimney illustrated reduces from 12 in. at the bottom of the air space to 4 in. at Concrete Construction 133 the top of this space (beneath the cap). In nontapering chimneys the variation in thickness is made by offsets on the inner side, but in this case the inner side is vertical and the reduction is made by the tapering of the outer side. This taper was given by the use of adjustable steel forms. These were in lifts 2 ft. high, and the diameter was reduced at each lift. The forms were found very flexible, and the same ones used for the stack were made to serve also for the enlarged head, or cap. The concrete is a 1:2:4 mix, made with gravel i to 1 in. in size. It was hoisted in buckets in an inside tower, which carried the working platform. The reinforcement consists of steel rods and hoops in each wall, the vertical rods being hooked to the rods in the base of the footing. This footing is built directly upon a stiff clay formation, no piles or other special treatment being required. Results of Recent Tests in the Laboratory Results of Wood-Decay Investigations Field and laboratory studies of the U. S. Forest Service indicate that much more care should be exercised in the selection of timber and in the construction of buildings to avoid conditions favorable to decay. Any one of the following causes may result in rapid de- terioration of the building: (1) Use of green timber, (2) allowing the timber to get wet during construction, (3) allowing the timber to absorb moisture after the building is finished because of leaks or lack of ventilation, (4) use of timber containing too much sap- wood, (5) use of timber which have already started to decay. The avoidance of these conditions will, as a rule, prevent decay. In special cases preservative treatment is necessary, zinc chloride and sodium fluoride being better than creosote for buildings. Studies to determine the extent to which lumber is attacked by fungi while seasoning in lumber yards are being continued. Specific cases were studied, showing how sound lumber is infected by partly decayed lumber before shipment is made. Simple rules were formu- lated for restricting the spread of fungus in lumber. Tests to determine the effect of various amounts of resin in the southern pines upon their durability indicate that it does not depend directly upon their resin content. About 1500 pieces of wood, representing 50 different species, are under test to determine their relative durability. At the end of 134 The Engineer in Field and Office three years, all of the conifers with the exception of cypress, red- wood, yew, and the cedars, have decayed, as have also most of the hardwoods. Kiln-Dried Douglas Fir Timbers Tested Bending tests were made upon four kiln-dried Douglas fir beams and minor specimens cut from them by the Forest Products Lab- oratory. Beams 1 and 2 were dried under high velocity, low-super- heated steam, at a temperature of 225 F. for a day, 230 F. for the succeeding seven days, and 240 F. for the remaining four days. Beams 3 and 4 were kept at a temperature of 220 F. throughout the run, the steam being superheated for the first five days, when the humidity was reduced to 70%. During the next four days the humidity dropped gradually to 25%, where it was maintained during the remaining seven days. Beams 1 and 2 had a modulus of rupture of 6620 and 4910 Ib. per sq.in. respectively, and beams 3 and 4 a modulus of rupture of 4740 and 6160 Ib. per sq.in. respectively. All beams failed by horizontal shear due to the severe checking. The results are very encouraging for the development of the process for kiln drying Douglas fir beams in structural sizes. Zinc Borate Retains Fire-Resisting Power The Forest Products Laboratory has just fire-tested a small shingle roof section, painted with a zinc borate paint, which has been exposed to the weather for nearly three years other shingles specially treated with the same paint being used as a control. The results show that the paint had resisted the action of the weather without losing its fire-retarding properties to any marked extent. White Paint Saves Gasoline In considering the effect of the different types of rays of which light is composed, it is found that the calorific, or heat-producing, rays are conducted by painted or finished objects in widely varying degree. This fact should be studied. When black or dark-colored paints have been used in painting large tanks containing light dis- tillates, rapid absorption of heat takes place, and considerable losses by evaporation are apt to occur. White or light-colored paint should Laboratory Work 135 therefore be used for the finishing coats on oil-storage tanks. Paints presenting a high gloss are, moreover, less absorptive of thermal rays than those presenting a matte surface. RISE IN TEMPERATURE OF BENZINE CONTAINED IN SMALL, TANKS PAINTED IN VARIOUS COLORS (GLOSS FINISH), WHEN SUBJECTED TO HAYS OF CARBON ARC FOR PERIOD OF 15 MINUTES COLOR Rise in Deg. Tin plate 19.8 Aluminum paint 20.5 White paint 22.5 Light cream paint. . 23.0 Light pink paint 23.7 Light blue paint 24.3 COLOR Light gray paint Light green paint Red iron oxide paint Dark Prussian blue paint.. Dark chrome green paint. . Black paint Rise in Deg. 26.3 26.6 29.7 36.7 39.9 54.0 Since white paints faintly tinted have given substantially the same heat-reflecting properties as white paints, the former should be given the preference, as they are more restful to the eye and more durable on long-time exposure. The Lead Content of a Leaded-Zinc Paint In states requiring the formula to be shown on paint packages, the lead content of all oxides excepting the lowest (under 5%) must 80 90 100 PbS04\! ZnO Amounts of Leaded Zinc and Sublimed Lead to Use for Desired Lead Content of Mix be stated. In all cases where the lead sulphate is actually basic (Pb 3 S 2 8 ), this is properly given as "basic lead sulphate." On the other hand, if leaded oxide be used in which the lead content is 136 The Engineer in Field and Office practically all present as normal lead sulphate (PbSOJ, the pro- priety of the designation may be subject to question. In practice it is often necessary or desirable to increase the basic lead sulphate content of the formula over that present in the leaded zinc available for use. Since basic lead sulphate (or sublimed lead) as marketed contains about 5% of zinc oxide, calculation of the required proportions of the two pigments required to yield the relative percentages is a rather complicated process. The accom- panying chart will furnish the required information at a glance. The figures to the left indicate the percentages of leaded zinc and those to the right of sublimed lead required to produce the required percentages of ZnO and PbS0 4 indicated at the bottom of the chart. The four diagonal lines represent the several grades of oxide. To find any desired percentages shown in the bottom rows of figures, using any of the oxides shown, follow the perpen- dicular line to its intersection with the proper diagonal; the required percentages of that oxide and sublimed lead will then be found on the horizontal line to the left and right respectively. For example: Desired percentages 50 each, using a 35% leaded oxide. The 50-50 perpendicular intersects the 35% diagonal at the horizontal line, which indicates at the left 76% leaded zinc, and at the right 24% sublimed white lead. The desired percentages will therefore be obtained by using 76 Ib. of the former to 24 Ib. of the latter. Tinned Copper Is No Better Than Tin Plate as a Roofing Material The unusual and interesting corrosion of the roofing material of the Library of Congress was recently investigated. This building has been covered since about 1893 by tinned sheet copper that has become covered within the last 10 or 15 years with small pits; in many cases, these pits have extended completely through the sheet. Such a condition is interesting, particularly in view of the fact that Washington is uncommonly free from smoke, which is ordinarily understood to be a strong accelerating factor in corrosion. The investigation has shown that the corrosion was due to no accidental inferiority of the material, but that it is to be considered as characteristic of all material of this type. It appears, therefore, that tinned copper is not superior in any way to tin plate for roofing material and that in view of its greater cost can no longer compete with it. Tinned sheet copper is used also for containing vessels such as milk cans and for fittings such as troughs, etc., for soda fountains Laboratory Work 137 and breweries. It is probable that such articles would also be sub- ject to pitting corrosion of the same type if they were not worn out by actual abrasion before the corrosion had proceeded far. To Prevent Volume-Change of Wood Pieces of air-dried wood were treated in a variety of ways to determine the best methods of retarding their absorption of moisture from the air and loss of moisture when placed in dry air. Specimens were treated with paraffin, impregnated with sugar, heated at high temperature, painted, varnished, coated with bakelite, impregnated with creosote, etc. Curves have been drawn showing the relative shrinkage and swelling which took place in these boards when exposed over a period of a year to atmosphere containing various humidities. The most effective results were secured by coating the wood with paraffin. Excellent results were also obtained from high temperature treatments and by impregnating the wood with sugar. Recent Pressure Tests of Welded Joints Tests recently made in the machine-construction laboratory of the University of Kansas show that the joints made by the oxyacetylene method develop average strengths against rupture from internal pressure as great as, if not greater than, those which may TESTS SHOW SUPERIORITY OF WELDED CONNECTIONS Size Pressure, Lb. per Sq.In. Pipe Type Joint At Failure Maximum Nature of Failure 2 in. Welded T 4400 4400 Tube seam split Welded T 2200 2200 Leak in tube seam Welded T 4750 4750 Tube seam split Screwed T 2350 2750 Sand holes in fitting Screwed T 500 2000 Sand holes in fitting 3 in. Butt weld 5300 Butt weld 4950 4950 Tube seam split Butt weld 4250 Coupling 3950 3950 Coupling split Coupling 3400 4400 Leak In coupling Welded T 3500 Welded T 4250 Welded T 3505 Screwed T 350 2700 Sand holes in fitting Screwed T 300 3100 Sand holes in fitting 4 in. Butt weld 5100 Pipe bulged Butt weld 3250 Coupling 300 3000 Leak at threads Coupling 750 2600 Leak at threads Welded T 3850 5100 Leak in weld Screwed T 1000 1950 Sand hole in fitting be ordinarily expected from screwed pipe connections. The data given in the accompanying table are supplementary, but confirm- atory of similar figures determined about a year ago. 138 The Engineer in Field and Office The results indicate that the strength of a welded pipe connection is practically the same as that of unwelded pipe; and although care- less welding might result in a leaky connection, if the line be tested when installed it should be immune from trouble in service. Special Pump Carries Pressure Over 5000 Pounds A special high-pressure pump of simple detail was designed and built for this work. It was connected to the specimens under test by means of a i-in. copper tube. A pressure gage, with a check valve opening toward the gage, was located on the tube the valve being necessary to steady the pressure in order that satisfactory readings could be obtained, because some of the samples carry pres- sures greater than 5000 Ib. per sq.in. before failure. Dense Wood Gives Nails High Values Tests of the efficiency of various types of wooden joints were made, and about 3000 nail-pulling tests were completed on nails driven into twenty-five different species of American timber. While the data have not been fully analyzed, it appears that the holding power of nails has a definite relationship to the density of the wood, and that there is practically no difference in strength between a solid beam and a wooden beam of the same dimensions made of two planks nailed together. Surveying 139 Practical Hints for the Surveyor Small Motor Cars on Precise Level Work Motor driven cars with flanged wheels for use on railroad tracks carry the outfit used by the United States Coast and Geodetic Survey in that part of their work which is upon railroad rights of way. An interesting novelty is the adding machine attached to one of the cars for recording the reports of the parties engaged in precise leveling. A rigid frame, which holds the calculating device, is bolted to the body of the car. The operator is provided with a comfortable cushioned seat over the wheel, where he can do his work conveniently. The added reliability and greater speed of the machine for noting the readings of the level rods makes it a valuable addition to the equipment of parties in the field. Another car carries the level, the tripod being securely mounted upon the frame of the motor vehicle, and the operator standing Coast and Geodetic Survey Party Carried on Light Motor Cars in the space between the tracks. Flat trays, or hand barrows, on each car are available for transporting the other instruments, tools and personal belongings of the party. The cars are light in weight, with frames formed of steel tub- ing, and are capable of traveling at a good speed with their load of instruments and passengers. A small gas engine under the frame supplies the motive power, and enables the party to reach its field of operations promptly and without fatigue. 140 The Engineer in Field and Office Scale on Transit Leg Fixes H. I. An easy method of establishing the height of instrument of a transit where this is needed, as in stadia surveys, is to fix a scale on one of the tripod legs by which the length of plumb-bob string and hence the H. I. may be gaged. The procedure of placing and using the scale is as follows: Set up the transit over the hub and measure the H. I. carefully and accurately. Have the point of the plumb-bob just touch the tack or wherever the H. I. is measured from. Swing the point of the bob to one of the legs of the transit and mark this point on the transit leg. By similar markings scratch a scale on the transit leg, using feet and tenths. Then at any time, with the point of the bob touching the tack at any set-up, the H. I. may be read directly on the leg of the transit by swinging the bob in an arc and reading the point of the bob. This will give results to the nearest hundredth, which is generally close enough. Set Slope Stakes a Foot Outside A deviation from the usual manner of setting slope stakes for railroad grading has been used successfully by the writer. Instead of the stake being driven slanting at the toe of the embankment or top of the cut, it is moved out one foot farther and driven down straight. Thus, each center-line stake is practically referenced by two hubs, which are much less likely to be displaced than if set at the edge of the slope. If the contractor is advised of the method of setting the stakes, it has proved easy and convenient for him to make his measurements accordingly. Survey Data Made Available for Field Use The engineer engaged in land surveying is often handicapped in the field by finding that he has not secured enough notes to cover the work to be done or by finding corners described by the notes at hand destroyed, together with the witness trees, and that a starting point some distance away must be sought. Again, the surveyor will often find other bearing trees than those described in the Government records; and unless he has the notes for these trees, they are worthless. First copies of complete section notes, giving all topography, are secured for each township in which work is likely to be done. By having the topography, a tedious search for a corner is often done away with, as a reference measurement can be made from Surveying 141 some stream or ridge. These section notes are written on specially printed blanks, each 8 x 11 in. and ruled as shown. After being copied, all sheets are carefully checked. To describe corners reestablished or reset, typewritten blanks are prepared for section, quarter-section and donation-land claim corners. These are filled out whenever the surveyor, or his assist- ants, resets or establishes a corner. South Bndr 5HEET | # tbt,,* .C, 35 Z Tu>p. 26 S R 5 w. rt-69' 47' W Between Section 35 "<* Srrtion 2. 2.OO 7 80 . Va r 1-4' E" . ZO 1 Ravine-. C. S. 31.00 Top of nd<5e. N Sc S 40. OO Sri Quarter Post from wbicb a Ye 1. Oak. 24" on. Bean. M.48*W. 132 IKs. W. Ook 18" Di,.BeBr..5.8r^. 487 ttfs 57-00 Top o?- n'd(5e. Be'^n 4^ cfescend. 63.00 L?ove Oak opemo. 75.00 Ivfove hillj ^nfer voll^u. 76.2.0 Trail M. ^ S 80.00 Sel Post Corner to Sections 2-3-34-35 from which a W.CbK. /<>" Dia. Bear,. S2S# . ?ff //ft . -L_^_ ___ Binding-'''^" "T" Highest serial number of pamphlets filed in box - Serial number of box --^ -^ M No. 19 Wire nails, j long, flat head-- .&.-:?< Filing Box Is Easily Made The filing of pamphlets by arbitrary serial numbers is far more simple, elastic and generally satisfactory than any attempt to classify or group these publications according to their subject matter. And any system of indexing, to be satisfactory, demands the personal attention of the person who is to make use of it, as the average office help cannot be expected to do indexing of this sort with any satisfactory degree of intelligence. Filing Data 157 Index of Details for a Structural Office Well executed and consistent plans are the prime indication of a good design in all classes of structures ; especially in building work, which involves a great many different types for the great variety of uses. This makes standardization of details and design somewhat more difficult than for the superstructures of highway and railroad bridges, but unless standardization is kept in mind constantly, the cost of making the design is quite likely to be so high as to show little profit; and at the same time the resulting structure may be more costly to the owner than if the reverse were true. A large structural-engineering office one employing a large force of draftsmen and designers, and carrying on work simultane- ously on several large buildings must of necessity have a good set of office standards for the guidance of the men and as an aid to the chief draftsman or engineer in obtaining uniform practice on all jobs of the same type, without necessitating his spending the greater part of his time giving instructions and personal supervision to each and every man or squad. The office standards for concrete building design should include tables and diagrams for the design of reinforced-concrete slabs and M \ beams (such as values of K = TJZ ) f r various combinations of unit stresses; areas and weights of steel bars; data on web rein- forcement of beams; reinforcement of spandrel girders; diagrams and tables for the design and detail of reinforced-concrete columns and footings and for steel columns incased in concrete; tables and details showing amount and method of reinforcing concrete stairs, and standard sheets showing the arrangement of lettering, dimen- sions, details, etc., on the floor plans. For steel building design the standards may include properties of various types of columns ; con- nection details; bracket details; cast-iron bases; typical details of spandrel beams, crane girders, steel stairs, steel stacks, tank floors, towers, etc. In order that original and special details developed which would not ordinarily be indexed in the general job file or included in the standards may be readily found, a card index of various details of completed designs in the office files, which are likely to be used on other work or referred to at various times, should be maintained. With such a file a new man in the organization can at once acquaint himself with methods of detailing certain portions of a standard or 158 The Engineer in Field and Office special building without a lengthy explanation on the part of the chief draftsman, thus saving much valuable time and increasing materially the speed with which the plans can be completed. Filing Catalogs and Pocket Maps Some of the material to be filed in an engineering library, particularly trade catalogs and folded pocket maps, is so diversified in size and shape that it is hard to handle. Thus, while it has come to be generally conceded that trade catalogs ought to be unified as to dimensions, and while this would be of advantage not only to the possible purchaser but also to the manufacturer, the catalogs continue to vary from the flimsy leaflet to the substantially bound volume. Yet, as the engineer uses them in making estimates of cost, in selecting equipment, and in other ways, he must have them readily available and well indexed. The general indexing schedule in use in one library is a form of the decimal system, by which the books and data are assembled under various classifications according to subject matter. This modified Dewey system was found to be too detailed for application to the trade catalogs, and it was necessary to simplify it; that is, K 6, Water Works Distribution System, including in its subdivisions the subjects of pipes, valves, distributing reservoirs, conduits, meters, etc., became in the simplified schedule, K 6, Water Works Pipes and Valves; while K 6.5, Meters, became more comprehensive and included Measuring Devices for Water. Within each classification it was decided to maintain an alpha- betical sequence of the catalogs according to the names of the firms issuing them. To do this an exceedingly flexible filing device was necessary, since it was planned to place the material without regard to size and shape and without segregating the bound volumes. Pam- phlet boxes would have meant considerable waste space, and vertical files either the segregation of the bound volumes or an unwarranted expenditure for filing units. Heavy manila envelopes were also considered for the filing of the leaflets and pamphlets, but were found undesirable both because they were incommodious, and be- cause when the envelopes were stacked the exposed ends offered on surface for indexing. It was decided to use a pasteboard receptacle open at the top, containing cardboard folders with tabs, such as are used for filing correspondence. The containers come in two widths, 1 in. and li in., and are 12 in. long just the depth of the open shelves upon which the catalogs were to be stacked. The exposed end of the Filing Data 159 unifile is curved and is covered with heavy brown paper. Upon this a sticker, bearing the classification number, was placed. Except in those instances where a whole container was devoted to the catalogs of a single firm, the indicating of the contents upon the end was avoided, so that as new material was added the cardboard folders could be shifted. On the tab at the top of each folder, however, the name of the manufacturer and the classification number were placed, while on the bound volumes and the thicker of the pamphlets strips were pasted with the firm name. In connection with this file a card index of the names of the manufacturers was maintained, with the classification number, or numbers, under which the catalogs of each firm were to be found. It has proved extremely useful to include also in this index the trade names of materials, since these so often differ from the name of the manufacturer or of the firm issuing the catalog. As the catalogs are assembled according to subject matter, a subject index has not been found necessary, but reference to an advertising index is made when doubt arises as to the manufacturer of some special material. This method has been tried out for five years, in the filing of 16 shelves of catalogs, and found satisfactory. From the very fact that they are designed to be tucked away, the pocket maps are likely to be tossed into a desk drawer or placed in a pigeonhole and forgotten. It is wise, therefore, to have a file of them in some conspicuous place where they will be readily accessible and will not be overlooked. Like the trade catalogs the pocket maps vary in size and shape, for while some seem designed for the coat or outer pockets, very many are long and flexible and of a form suit able for the vest or trousers pockets. A simple device for main- taining a neat and uniform file of such maps is the heavy manila envelope of standard size (4i in. by 10i in.) with its flap removed and a cloth tape tab H in. in length pasted on the closed long edge. On this tab the brief title of the map is written. The envelopes are filed on end in alphabetical sequence of the titles of the maps. The tabs are bent to one side and as they have been pasted at varying intervals on the long side of the envelopes it seldom happens that they overlap. If it should chance that a map is too long for an envelope, the latter may be slit at the top end. This file is kept in a glass-front case and stacked by means of tin rack ends. Rack ends placed behind the envelopes will keep them from slipping back on the shelf. In the general card catalog of material in the library, the maps are listed under the subject "Maps" and also under the in- dividual titles. 160 The Engineer in Field and Office Value of Carefully Kept Stadia Notes It is especially important that some definite form of note-keeping be followed in making stadia surveys, even if they are to be platted by the one who makes the survey. There are so many classes of information to be recorded, that if the work is not done systemati- cally, the value of the data recorded is lost through the impossibility of transferring the information contained in the notes to a scale IP. & 77? .as Ufff H.L F.S. Eleva. f.1 3.25- 6-7 /<*/ SJ 33.3 37.' 37-W 36.Z J7./ tSW; oMDlsf. rrTr J 6+33 ffem Poad turns L Field onK Hub in Rocrd Lowsp&ZOOb. Hub In Pocrd arks M.VM Field 20 KcfZ Ske Systematic Note-Keeping Helped by This Form map. This form must, of necessity, be adapted to the work to be done, and should, so far as practicable, conform to established forms of note-keeping. A form used in making drainage recon- naissance surveys is illustrated by the accompanying sample of notes which were recently taken in the field. Turning points and instrument points are numbered successively arabic figures being used for turning points, as they are usually placed on the stake by an untrained assistant, and Roman numerals for the instrument points. The turning points are available for later cross-line surveys. All instrument heights and heights of turn- ing points are calculated in the field as well as the stadia distance. A mental calculation is made by adding one-half the observed distance to the lower stadia reading as a check on the level reading on the central cross-head. This also serves as a check on the observed distance. "- Filing Data 161 In line surveys the distance from starting point to each instru- ment point and turning point is computed and recorded in the dis- tance column, this computation being done at night, or after return to the office. After being checked by adding the observed distance from starting point, the total distances are set down in ink and are of assistance in platting the survey. Location of Water Mains, Connections and Valves Kept on Card File Any card system of keeping records which has proven its value by twenty years and more of continual use, without radical change, is worthy of notice. In this class falls the card system for re- cording the location of water mains, valves, connections, fire hy- drants and other special fixtures, which is in use by the Water Department of the City of St. Louis. This system comprises the use of specially printed 4x6 cards kept in drawers and referred to ordinarily only by members of the office force of the distribution system. The accompanying illustration gives a very clear indication of the nature of the card and the information kept on it. Each size of pipe and each particular fitting has a special symbol, and the situations of all valves are marked, showing their position with reference to the curb or building line. The street running from bottom to top is an east and west street, and from left to right, a north and south street. The east and west streets follow each other alphabetically in the drawers; that is to say, Atlantic Ave. would come before West Ave., and all intersec- tions with north and south streets are represented by succeeding cards which start from the most easterly intersection of the east and west street with a north and south street. Therefore, to locate any intersection it is necessary only to know which of the streets is an east and west street. Each card is inclosed in an envelope of transparent paper, which keeps it from being soiled by the rather excessive handling to which cards of this sort are subjected. It is only possible to take the cards from the drawer by unlocking a special keeper-rod, the kkey to which is guarded carefully and is never allowed to leave the possession of the chief draftsman of the water department. 162 The Engineer in Field and Office t 1 t- | o t 1 5 _H 1 1 s> ^ O"> <0 Q_ c