The Project Gutenberg EBook of Transactions of the American Society of Civil Engineers, vol. LXVIII, Sept. 1910, by James H. Brace, Francis Mason and S. H. Woodard This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at www.gutenberg.org Title: Transactions of the American Society of Civil Engineers, vol. LXVIII, Sept. 1910 The New York Tunnel Extension of the Pennsylvania Railroad. The East River Tunnels. Paper No. 1159 Author: James H. Brace, Francis Mason and S. H. Woodard Release Date: July 1, 2006 [EBook #18722] Language: English Character set encoding: ISO-8859-1 *** START OF THIS PROJECT GUTENBERG EBOOK TRANSACTIONS OF THE AMERICAN *** Produced by Juliet Sutherland, Taavi Kalju and the Online Distributed Proofreading Team at http://www.pgdp.net AMERICAN SOCIETY OF CIVIL ENGINEERS INSTITUTED 1852 TRANSACTIONS Paper No. 1159 THE NEW YORK TUNNEL EXTENSION OF THE PENNSYLVANIA RAILROAD. THE EAST RIVER TUNNELS.[A] By James H. Brace, Francis Mason, and S. H. Woodard, Members, Am. Soc. C. E. This paper will be limited to a consideration of the construction of the tunnels, the broader questions of design, etc., having already been considered in papers by Brig.-Gen. Charles W. Raymond, M. Am. Soc. C. E., and Alfred Noble, Past-President, Am. Soc. C. E. The location of the section of the work to be considered here is shown on Plate XIII of Mr. Noble's paper. There are two permanent shafts on each side of the East River and four single cast-iron tube tunnels, each about 6,000 ft. long, and consisting of 3,900 ft. between shafts under the river, and 2,000 ft. in Long Island City, mostly under the depot and passenger yard of the Long Island Railroad. This tube-tunnel work was naturally a single job. The contract for its [Pg 419] construction was let to S. Pearson and Son, Incorporated, ground being broken on May 17th, 1904. Five years later, to a day, the work was finished and received its final inspection for acceptance by the Railroad Company. The contract was of the profit-sharing type, and required an audit, by the Railroad Company, of the contractor's books, and a careful system of cost-keeping by the Company's engineers, so that it is possible to include in the following some of the unit costs of the work. These are given in two parts: The first is called the unit labor cost, and is the cost of the labor in the tunnel directly chargeable to the thing considered. It does not include the labor of operating the plant, nor watchmen, yardmen, pipemen, and electricians. The second is called "top charges," a common term, but meaning different things to different contractors and engineers. Here, it is made to include the cost of the contractor's staff and roving laborers, such as pipemen, electricians, and yardmen, the cost of the plant and its operation, and all miscellaneous expenses, but does not include any contractor's profit, nor cost of materials entering permanent work. The contractor's plant is to be described in a paper by Henry Japp,[B] M. Am. Soc. C. E., and will not be dealt with here. The contractors carried on their work from three different sites. From permanent shafts, located near the river in Manhattan, four shields were driven eastward to about the middle of the river; and, from two similar shafts at the river front in Long Island City, four shields were driven westward to meet those from Manhattan. From a temporary shaft, near East Avenue, Long Island City, the land section of about 2,000 ft. was driven to the river shafts. TUNNELS FROM EAST AVENUE TO THE RIVER SHAFTS. The sinking of the temporary shaft at East Avenue was a fairly simple matter. Rough 6 by 12-in. sheet-piling, forming a rectangle, 127 by 34 ft., braced across by heavy timbering, was driven about 28 ft. to rock as the excavation progressed. Below this, the shaft was sunk into rock, about 27 ft., without timbering. As soon as the shaft was down, on September 30th, 1904, bottom headings were started westward in Tunnels A, B, and D. When these had been driven about half the distance to the river shafts, soft ground was encountered. (See Station 59, Plate XIII.) As the ground carried considerable water, it was decided to use compressed air. Bulkheads were built in the heading, and, with an air pressure of about 15 lb. per sq. in., the heading was driven through the soft ground and into rock by ordinary mining methods. The use of compressed air was then discontinued. West of this soft ground, a top heading, followed by a bench, was driven to the soft ground at about Station 66. Tunnel C, being higher, was more in soft ground, and at first it was the intention to delay its excavation until it had been well drained by the bottom headings in the tunnels on each side. A little later it was decided to use a shield without compressed air. This shield had been used in excavating the stations of the Great Northern and City Tunnel in London. It was rebuilt, its diameter being changed from 24 ft. 8-1/2 in. to 23 ft. 5-1/4 in. It proved too weak, and after it had flattened about 4 in. and had been jacked up three times, the scheme was abandoned, the shield was removed, and work was continued by the methods which were being used in the other tunnels. The shield was rather light, but probably it would have been strong enough had it been used with compressed air, or had the material passed through been all earth. Here, there was a narrow concrete cradle in the bottom, with rock up to about the middle of the tunnel, which was excavated to clear the shield, and gave no support on its sides. The shield was a cylinder crushed between forces applied along the top and bottom. With the exception of this trial of a shield in Tunnel C, and a novel method in Tunnel B, where compressed air, but no shield, was used, the description of the work in one tunnel will do for all. From the bottom headings break-ups were started at several places in each tunnel where there was ample cover of rock above. Where the roof was in soft ground, top headings were driven from the points of break-up and timbered. As soon as the full-sized excavation was completed, the iron lining was built, usually in short lengths. It will be noticed on Plate XIII that there is a depression in the rock between Station 65 and the river shafts, leaving all the tunnels in soft ground. As this was directly under the Long Island Railroad passenger station, it was thought best to use a shield and compressed air. This was done in Tunnels A, C, and D, one shield being used successively for all three. It was first erected in Tunnel D at Station 64 + 47. From there it was driven westward to the river shaft. It was then taken apart and re-erected in Tunnel C at Station 63 + 63 and driven westward to the shaft. It was then found that there would not be time for one shield to do all four lines. The experience in Tunnels C and D had proven the ground to be much better than had been expected. There was considerable clay in the sand, and, with the water blown out by compressed air, it was very stable. A special timbering method was devised, and Tunnel B was driven from Station 66 + 10 to the shaft with compressed air, but without a shield. In the meantime the shield was re-erected in Tunnel A and was shoved through the soft ground from Station 65 + 48 nearly to the river shaft, where it was dismantled. There was nothing unusual about the shield work; it was about the same as that under the river, which is fully described elsewhere. In spite of great care in excavating in front of the shield, and prompt grouting behind it, there was a small settlement of the building above, amounting to about 1-1/2 in. in the walls and about 5 in. in the ground floors which were of concrete laid like a sidewalk directly upon the ground. Whether this settlement was due to ground lost in the shield work or to a compacting of the ground on account of its being dried out by compressed air, it is impossible to say. The interesting features of this work from East Avenue to the river shafts are the mining methods and the building of the iron tube without a shield. [Pg 420] [Pg 421] [Pg 422] EXCAVATION IN ALL ROCK. Where the tunnel was all in good rock two distinct methods were used. The first was the bottom-heading-and-break- up, and the second, the top-heading-and-bench method. The first is illustrated by Figs. 1 and 2, Plate LXIII. The bottom heading, 13 ft. wide and 9 ft. high, having first been driven, a break-up was started by blasting down the rock, forming a chamber the full height of the tunnel. The timber platform, shown in the drawing, was erected in the bottom heading, and extended through the break-up chamber. The plan was then to drill the entire face above the bottom heading and blast it down upon the timber staging, thus maintaining a passage below for the traffic from the heading and break-ups farther down the line. Starting with the condition indicated by Plate XIII, the face was drilled, the columns were then taken down and the muck pile was shoveled through holes in the staging into muck cars below. The face was then blasted down upon the staging, the drill columns were set up on the muck pile, and the operation was repeated. This method has the advantage that the bottom heading can be pushed through rapidly, and from it the tunnel may be attacked at a number of points at one time. It was found to be more expensive than the top-heading-and-bench method, and as soon as the depression in the rock at about Station 59 was passed, a top heading about 7 ft. high, and roughly the segment of a 23-ft. circle, was driven to the next soft ground in each of the four tunnels. The remainder of the section was taken out in two benches, the first, about 4 ft. high, was kept about 15 ft. ahead of the lower bench, which was about the remaining 11 ft. high. EXCAVATION IN EARTH AND ROCK. About 2,500 ft. of tunnel, the roof of which was in soft ground, was excavated in normal air by the mining-and- timbering method. In the greater part of this the rock surface was well above the middle of the tunnel. The method of timbering and mining, while well enough known, has not been generally used in the United States. PLATE LXIII Starting from the break-up in all rock, as described above, and illustrated on Plate XIII, when soft ground was approached, a top heading was driven from the rock into and through the earth. This heading was about 7 ft. high and about 6 ft. wide. This was done by the usual post, cap, and poling-board method. The ground was a running sand with little or no clay, and, at first, considerable water, in places. All headings required side polings. The roof poling boards were about 2-1/2 or 3 ft. above the outside limit of the tunnel lining, as illustrated by Figs. 3, 4, and 5, Plate LXIII. The next step was to place two crown-bars, AA, usually about 20 ft. long, under the caps. Posts were then placed under the bars, and poling boards at right angles to the axis of the tunnel were then driven out over the bars. As these polings were being driven, the side polings of the original heading were removed, and the earth was mined out to the end of these new transverse polings. Breast boards were set on end under the ends of the transverse polings when they had been driven out to their limit. Side bars, BB, were then placed as far out as possible and supported on raking posts. These posts were carried down to rock, if it was near, if not, a sill was placed. A new set of transverse polings was driven over these side bars and the process was repeated until the sides had been carried down to rock or down to the elevation of the sills supporting the posts, which were usually about 4 ft. above the axis of the tunnel. The plan then was to excavate the remainder of the section and build the iron lining in short lengths, gradually transferring the weight of the roof bars of the iron lining as the posts were taken out. This meant that not more than four rings, and often only one ring, could be built before excavation and a short length of cradle became necessary. Before the posts under the roof bars could be built and the weight transferred to the iron lining, a grout dam was placed at the leading end of the iron lining, and grout was brought up to at least 45° from the top. Such workings were in progress at as many as eight places in one tunnel at the same time. Where there was only the ordinary ground-water to contend with, the driving of the top heading drained the ground very thoroughly, and the enlarging was done easily and without a serious loss of ground. Under these conditions the surface settlement was from 6 in. to 2 ft. Under Borden Avenue, there was more water, which probably came from a leaky sewer; it was not enough to form a stream, but just kept the ground thoroughly saturated. There was a continued though hardly perceptible flow of earth through every crevice in the timbering during the six or eight weeks between the driving of the top heading and the placing of the iron lining; and here there was a settlement of from 4 to 8 ft. at the surface. TUNNELING IN COMPRESSED AIR WITHOUT A SHIELD. When it became evident that there would not be time for one shield to do the soft ground portions of all four tunnels under the Long Island Railroad station, a plan was adopted and used in Tunnel B which, while not as rapid, turned out to be as cheap as the work done by the shields. Figs. 6 and 7, Plate LXIII, and Fig. 1, Plate LXIV, illustrate this work fairly well. The operation of this scheme was about as follows: Having the iron built up to the face of the full-sized excavation, a hole or top heading, about 3 ft. wide and 4 or 5 ft. high, was excavated to about 10 ft. in advance. This was done in a few hours without timbering of any kind; but, as soon as the hole or heading was 10 ft. out, 6 by 12-in. laggings or polings were put up in the roof, with the rear ends resting on the iron lining and the leading ends resting on vertical breast boards. The heading was then widened out rapidly and the lagging was placed, down to about 45° from [Pg 423] [Pg 424] the crown. The forward ends of the laggings were then supported by a timber rib and sill. Protected by this roof, the full section was excavated, and three rings of the iron lining were built and grouted, and then the whole process was repeated. PLATE LXIV, FIG. 1.—TUNNELING IN COMPRESSED AIR WITHOUT SHIELD. PLATE LXIV, FIG. 2.—T-HEAD AIR-LOCK. PLATE LXIV, FIG. 3.—CUTTING EDGE OF CAISSON ASSEMBLED. PLATE LXIV, FIG. 4.—CAISSON SUPPORTED ON JACKS AND BLOCKS. CONCRETE CRADLES, HAND-PACKED STONE AND GROUTING. Had the East Avenue Tunnel been built by shields, as was contemplated at the time of its design, the space between the limits of excavation and the iron lining would have been somewhat less than by the method actually used, especially in the earth portions. This space would have been filled with grout ejected through the iron lining. The change in the method of doing the work permitted the use of cheaper material, in place of part of the grout, and, at the same time, facilitated the work. The tube of cast-iron rings is adapted to be built in the tail of the shield. Where no shield was used, after the excavation was completed and all loose rock was removed, timbers were fixed across the tunnel from which semicircular ribs were hung, below which lagging was placed. The space between this and the rough rock surface was filled with concrete. This formed a cradle in which the iron tube could be erected, and, at the same time, occupied space which would have been filled by grout, at greater cost, had a shield been used. As soon as each ring of iron was erected, the space between it and the roof of the excavation was filled with hand- packed stone. At about every sixth ring a wall of stone laid in mortar was built between the lining and the rock to serve as a dam to retain grout. The interstices between the hand-packed stones were then filled with 1 to 1 grout of cement and sand, ejected through the iron lining. The concrete cradles averaged 1.05 cu. yd. per ft. of tunnel, and cost, exclusive of materials, $6.70 per cu. yd., of which $2.25 was for labor and $4.45 was for top charges. The hand- packed stone averaged 1-1/2 cu. yd. per ft. of tunnel, and cost $2.42 per cu. yd., of which $0.98 was for labor and $1.44 was for top charges. ERECTION OF IRON LINING. The contractors planned to erect the iron lining with erectors of the same pattern as that used on the shield under the river, mounted on a traveling stage. These will be described in detail in Mr. Japp's paper. Two of these stages and erectors worked in each tunnel at different points. The tunnel was attacked from so many points that these erectors could not be moved from working to working. The result was that about 58% of the lining was built by hand. At first thought, this seems to be a crude and extravagant method, as the plates weighed about 1 ton each and about 20,000 were erected by hand. As it turned out, the cost was not greater than for those erected by machinery, taking into account the cost of erectors and power. This, however, was largely because the hand erection reduced the amount of work to be done by the machines so much that the machines had an undue plant charge. The hand erection was very simple. A portable hand-winch, with a 3/8-in. wire rope, was set in any convenient place. The wire rope was carried to a snatch-block fastened to the top of the iron previously built; or, where the roof was in soft ground, the timbering furnished points of attachment. The end of the wire rope was then hooked to a bolt hole in a new plate, two men at the winch lifted the plate, and three or four others swung it into approximate place, and, with the aid of bars and drift-pins, coaxed it into position and bolted it. Where there was no timbering above the iron, sometimes the key and adjoining plates were set on blocking on a timber staging and then jacked up to place. LONG ISLAND SHAFTS. The river shafts were designed to serve both as working shafts and as permanent openings to the tunnels, and were larger and more substantial than would have been required for construction purposes. Plate X of Mr. Noble's paper shows their design. They consist of two steel caissons, each 40 by 74 ft. in plan, with walls 5 ft. thick filled with concrete. A wall 6 ft. thick separated each shaft into two wells 29 by 30 ft., each directly over a tunnel. Circular openings for the tunnel, 25 ft. in diameter, were provided in the sides of the caissons. During the sinking these were closed by bulkheads of steel plates backed by horizontal steel girders. The shafts were sunk as pneumatic caissons to a depth of 78 ft. below mean high water. There have been a few caissons which were larger and were sunk deeper than these, but most large caissons have been for foundations, such as bridge piers, and have been stopped at or a little below the surface of the rock. The unusual feature of the caissons for the Long Island shaft is that they were sunk 54 ft. through rock. [Pg 425] [Pg 426] It had been hoped that the rock would prove sound enough to permit stopping the caissons at or a little below the surface and continuing the excavation without sinking them further; for this reason only the steel for the lower 40 ft. of the caissons was ordered at first. The roof of the working chamber was placed 7 ft. above the cutting edge. It was a steel floor, designed by the contractors, and consisted of five steel girders, 6 ft. deep, 29 ft. long, and spaced at 5-ft. centers. Between were plates curved upward to a radius of 4 ft. Each working chamber had two shafts, 3 ft. by 5 ft. in cross-section, with a diaphragm dividing it into two passages, the smaller for men and the larger for muck buckets. On top of these shafts were Moran locks. Mounted on top of the caisson was a 5-ton Wilson crane, which would reach each shaft and also the muck cars standing on tracks on the ground level beside the caissons. Circular steel buckets, 2 ft. 6 in. in diameter and 3 ft. high, were used for handling all muck. These were taken from the bottom of the working chamber, dumped in cars, and returned to the bottom without unhooking. Work was carried on by three 8-hour shifts per day. The earth excavation was done at the rate of about 67 cu. yd. per day from one caisson. The rock excavation, amounting to about 6,200 cu. yd. in each caisson, was done at the rate of about 44.5 cu. yd. per day. The average rate of lowering, when the cutting edge of the south caisson was passing through earth, was 0.7 ft. per day. In rock, the rate was 0.48 ft. per day in the south caisson, and 0.39 ft. per day in the north caisson. At the beginning all lowering was done with sixteen hydraulic jacks. Temporary brackets were fastened to the outside of the caisson. A 100-ton hydraulic jack was placed under each alternate bracket and under each of the others there was blocking. The jacks were connected to a high-pressure pump in the power-house. As the jacks lifted the caisson, the blocking was set for a lower position, to which the caisson settled as the jacks were exhausted. After the caisson had penetrated the earth about 10 ft., the outside brackets were removed and the lowering was regulated by blocking placed under brackets in the working chamber. The caisson usually rested on three sets of blockings on each side and two on each end. The blocking was about 4 ft. inside the cutting edge. In the rock, as the cutting edge was cleared for a lowering of about 2 ft., 6 by 8-in. oak posts were placed under the cutting-edge angle. When a sufficient number of posts had been placed, the blocking on which the caisson had rested was knocked or blasted out, and the rock underneath was excavated. The blocking was then re-set at a lower elevation. The posts under the cutting edge were then chopped part way through and the air pressure was lowered about 10 lb., which increased the net weight to more than 4,000,000 lb. The posts then gradually crushed and the caissons settled to the new blocking. The tilt or level of the caisson was controlled by chopping the posts more on the side which was desired to move first. The caisson nearly always carried a very large net weight, usually about 870 tons. The concrete in the walls, which was added as the caisson was being sunk, was kept at about the elevation of the ground. There was generally a depth of from 5 to 20 ft. of water ballast on top of the roof of the working chamber. The air pressure in the working chamber was usually much less than the hydrostatic head outside the caisson. For example, the average air pressure in the south caisson during January, 1906, was 16-1/2 lb., while the average head was 62.5 ft., equivalent to 27 lb. per sq. in. Under these conditions, there was a continued but small leakage into the caisson of from 15,000 to 20,000 gal. per day. In the rock the excavation was always carried from 2 to 5 in. outside the cutting edge. As soon as the cutting edge was cleared, bags of clay were placed under it in a well-tiered, solid pile, so that when the caisson was lowered the bags were cut through and most of the clay, bags and all, was squeezed back of the cutting edge between the rock and the caisson. Table 1 shows the relation of the final position of the caissons to that designed. The cost of rock excavation in the caisson was $4.48 per cu. yd. for labor and $10.54 for top charges. The bottom of the shaft is an inverted concrete arch, 4 ft. thick, water-proofed with 6-ply felt and pitch. As soon as the caisson was down to its final position and the excavation was completed, concrete was deposited on the uneven rock surfaces, brought up to the line of the water-proofing, and given a smooth 1-in. mortar coat. The felt was stuck together in 3-ply mats on the surface with hot coal-tar pitch. These were rolled and sent down into the working chamber, where they were put down with cold pitch liquid at 60° Fahr. Each sheet of felt overlapped the one below 6 in. The water- proofing was covered by a 1-in. mortar plaster coat, after which the concrete of the 4-ft. inverted arch was placed. While the water-proofing and concreting were being done, the air pressure was kept at from 30 to 33 lb. per sq. in., the full hydrostatic head at the cutting edge. After standing for ten days, the air pressure was taken off, and the removal of the roof of the working chamber was begun. The water-proofing was done by the Union Construction and Waterproofing Company. TABLE 1.—RELATION OF THE FINAL POSITION OF THE CAISSONS TO THAT DESIGNED. Location. Long Island City. Shaft. North. South. Corner. High. East. North. High. East. North. Northeast 0.21 ft. 0.08 ft. 0.05 ft. 0.32 ft. 0.15 ft. 0.28 ft. [Pg 427] [Pg 428] [Pg 429] Northwest 0.22 " 0.08 " 0.02 " 0.00 " 0.15 " 0.12 " Southwest 0.27 " 0.14 " 0.02 " 0.18 " 0.45 " 0.12 " Southeast 0.23 " 0.14 " 0.05 " 0.39 " 0.45 " 0.28 " Location. Manhattan. Shaft. North. South. Corner. High. East. South. High. East or West. North or South. Northeast 0.23 ft. 0.74 ft. 0.38 ft. 0.00 ft. 0.06 ft. east. 0.04 ft. south. Northwest 0.00 " 0.74 " 0.22 " 0.08 " 0.06 " " 0.13 " north. Southwest 0.11 " 0.31 " 0.22 " 0.21 " 0.45 " west. 0.13 " " Southeast 0.46 " 0.31 " 0.38 " 0.04 " 0.45 " " 0.04 " south. The cost of labor in compressed air chargeable to concreting was $3.40 per cu. yd. After the roof of each working chamber had been removed, the shield was erected on a timber cradle in the bottom of the shaft, in position to be shoved out of the opening in the west side of the caisson. Temporary rings of iron lining were erected across the shaft in order to furnish something for the shield jacks to shove against. The roof of the working chamber was then re-erected about 35 ft. above its original position and about 8 ft. above the tunnel openings. This time, instead of the two small shafts which were in use during the sinking of the caisson, a large steel shaft with a T-head lock was built. This is illustrated in Fig. 2, Plate LXIV. The shaft was 8 ft. in diameter. Inside there was a ladder and an elevator cage for lowering and hoisting men and the standard 1-yd. tunnel cars. At the top, forming the head of the T, there were two standard tunnel locks. MANHATTAN SHAFTS. A permanent shaft, similar to the river shafts in Long Island City, was constructed at Manhattan over each pair of tunnels. Each shaft was located across two lines, with its longer axis transverse to the tunnels. Plate XIII shows their relative positions. They were divided equally by a reinforced concrete partition wall transverse to the line of the tunnels. On completion, the western portions were turned over to the contractor for the cross-town tunnels for his exclusive use. South Shaft.—Work on the south shaft was started on June 9th, 1904, with the sinking of a 16 by 16-ft. test pit in the center of the south half of the south shaft, which reached disintegrated rock at a depth of about 20 ft. Starting in August, the full shaft area, 74 by 40 ft., was taken out in an open untimbered cut to the rock, and a 20 by 50-ft. shaft was sunk through the rock to tunnel grade, leaving a 10 or 12-ft. berm around it. (Fig. 1, Plate LXX.) The erection of the caisson was started, about the middle of January, on the rock berm surrounding the 20 by 50-ft. shaft and about 15 ft. below the surface. Fig. 3, Plate LXIV, shows the cutting edge of the caisson assembled. The excavation of the small shaft had shown that hard rock and only a very small quantity of water would be encountered, and that the caisson need be sunk only a short distance below the rock surface. Therefore, no working-chamber roof was provided, the caisson was built to a height of only 40 ft., and the circular openings were permanently closed. The assembling of the caisson took 2-1/2 months, and on April 2d lowering was started. Inverted brackets were bolted temporarily to the cutting-edge stiffening brackets, and the sinking was carried on by methods similar to those used at Long Island. The jacks and blocking supporting the caisson are shown in Fig. 4, Plate LXIV. As soon as the cutting edge entered the rock, which was drilled about 6 in. outside of the neat lines, the space surrounding the caisson was back-filled with clay and muck to steady it and provide skin friction. As the friction increased, the walls were filled with concrete, and as the caisson slowly settled, it was checked and guided by blocking. The cutting edge finally came to rest 31 ft. below mean high water, the sinking having been accomplished in about seven weeks, at an average rate of 0.50 ft. per day. The final position of the cutting edge in relation to its designed position is shown in Table 1. A berm about 4 ft. wide was left at the foot of the caisson below which the rock was somewhat fissured and required timbering. The cutting edge of the caisson was sealed to the rock with grout on the outside and a concrete base to the caisson walls on the inside, the latter resting on the 4-ft. berm. Following the completion of the shaft, the permanent sump was excavated to grade for use during construction. North Shaft.—The north shaft had to be sunk in a very restricted area. The east side of the caisson cleared an adjoining building at one point by only 1 ft., while the northwest corner was within the same distance of the east line of First Avenue. As in the case of the Long Island shafts, the steelwork for only the lower 40 ft. was ordered at the start. This height was completely assembled before sinking was begun. The caisson was lowered in about the same manner as [Pg 430] [Pg 431] those previously described. The bearing brackets for the hydraulic jacks were attached, as at the south shaft, to the inside of the cutting-edge brackets. The east side of the caisson was in contact with the foundations of the neighboring building, while the west side was in much softer material. As a consequence, the west side tended to settle more rapidly and thus throw the caisson out of level and position. To counteract that tendency, it was necessary to load the east wall heavily with cast-iron tunnel sections, in addition to the concrete filling in the walls. Soon after sinking was begun, a small test shaft was sunk to a point below the elevation of the top of the tunnels. The rock was found to be sound, hard, and nearly dry. It was then decided to stop the caisson as soon as a foundation could be secured on sound rock. The latter was found at a depth of 38 ft. below mean high water. With the cutting edge seated at that depth, the top of the caisson was only 2 ft. above mean high water, and as this was insufficient protection against high tides, a 10-ft. extension was ordered for the top. Work, however, went on without delay on the remainder of the excavation. The junction between the cutting edge and the rock was sealed with concrete and grout. The caisson was lowered at an average rate of 0.53 ft. per day. The size of the shaft below the cutting edge was 62 ft. 7 in. by 32 ft. The average rate of excavation during the sinking in soft material was 84 cu. yd. per day. The average rate of rock excavation below the final position of the cutting edge was 125 cu. yd. per day. There were night and day shifts, each working 10 hours. Excavation in earth cost $3.96 per cu. yd., of which $1.45 was for labor and $2.51 for top charges, etc. The excavation of rock cost $8.93 per cu. yd., $2.83 being for labor and $6.10 for top charges. The final elevations of the four corners of the cutting edge, together with their displacement from the desired positions, are shown in Table 1. RIVER TUNNELS. The four river tunnels, between the Manhattan and Long Island City shafts, a distance of about 3,900 ft., were constructed by the shield method. Eight shields were erected, one on each line in each shaft, the four from Manhattan working eastward to a junction near the middle of the river with the four working westward from Long Island City. Toward the end of the work it was evident that the shields in Tunnels B, C, and D would meet in the soft material a short distance east of the Blackwell's Island Reef if work were continued in all headings. In order that the junction might be made in firm material, work from Manhattan in those three tunnels was suspended when the shields reached the edge of the ledge. The shields in Tunnel A met at a corresponding point without the suspension of work in either. An average of 1,760 ft. of tunnel was driven from Manhattan and 2,142 ft. from Long Island City. PLATE LXV, FIG. 1.—SHIELD FITTED WITH SECTIONAL SLIDING HOODS AND SLIDING EXTENSIONS TO THE FLOORS. PLATE LXV, FIG. 2.—SHIELD FITTED WITH FIXED HOODS AND FIXED EXTENSIONS TO THE FLOORS. TUNNELS DRIVEN EASTWARD FROM MANHATTAN. Materials and Inception of Work.—The materials encountered are shown in the profile on Plate XIII, and were similar in all the tunnels. In general, they were found to be about as indicated in the preliminary borings. The materials met in Tunnel A may be taken as typical of all. From the Manhattan shaft eastward, in succession, there were 123 ft. of all-rock section, 87 ft. of part earth and part rock, 723 ft. of all earth, 515 ft. of part rock and part earth, 291 ft. of all rock, and 56 ft. of part rock and part earth. The rock on the Manhattan side was Hudson schist, while that in the reef was Fordham gneiss. Here, as elsewhere, they resembled each other closely; the gneiss was slightly the harder, but both were badly seamed and fissured. Wherever it was encountered in this work, the rock surface was covered by a deposit of boulders, gravel, and sand, varying in thickness from 4 to 10 ft. and averaging about 6 ft. The slope of the surface of the ledge on the Manhattan side averaged about 1 vertical to 4 horizontal. The rock near the surface was full of disintegrated seams, and was badly broken up. It was irregularly stratified, and dipped toward the west at an angle of about 60 degrees. Large pieces frequently broke from the face and slid into the shield, often exposing the sand. The rock surface was very irregular, and was covered with boulders and detached masses of rock embedded in coarse sand and gravel. The sand and gravel allowed the air to escape freely. By the time the shields had entirely cleared the rock, the material in the face had changed to a fine sand, stratified every few inches by very thin layers of chocolate-colored clayey material. This is the material elsewhere referred to as quicksand. As the shield advanced eastward, the number and thickness of the layers of clay increased until the clay formed at least 20% of the entire mass, and many of the layers were 2 in. thick. At a distance of about 440 ft. beyond the Manhattan ledge, the material at the bottom of the face changed suddenly to one in which the layers of clay composed probably 98% of the whole. The sand layers were not more than 1/16 in. thick and averaged about 2 in. apart. The surface of the clay rose gradually for a distance of 40 ft. in Tunnels A and B, and 100 ft. in Tunnels C and D, when gravel and boulders appeared at the bottom of the shield. At that time the clay composed about one-half of the face. The surfaces of both the clay and gravel were irregular, but they rose gradually. After rock was encountered, the [Pg 432] [Pg 433] [Pg 434] formations of gravel and clay were roughly parallel to the rock surface. As the surface of the rock rose they disappeared in order and were again encountered when the shields broke out of rock on the east side of the Blackwell's Island Reef. East of the reef a large quantity of coarse open sand was present in the gravel formations before the clay appeared below the top of the cutting edge. In Tunnels C and D this was especially difficult to handle. It appears to be a reasonable assumption that the layer of clay was continuous across the reef. Wherever the clay extended above the top of the shield it reduced the escape of air materially. It is doubtless largely due to this circumstance that the part-rock sections in the reef were not the most difficult portions of the work. While sinking the lower portions of the shafts the tunnels were excavated eastward in the solid rock for a distance of about 60 ft., where the rock at the top was found to be somewhat disintegrated. This was as far as it was considered prudent to go with the full-sized section without air pressure. At about the same time top headings were excavated westward from the shafts for a distance of 100 ft., and the headings were enlarged to full size for 50 ft. The object was to avoid damage to the shaft and interference with the river tunnel when work was started by the contractor for the cross-town tunnel. PLATE LXVI, FIG. 1.—REAR OF SHIELD SHOWING COMPLETE FITTINGS. PLATE LXVI, FIG. 2.—SHIELD WITH LOWER PORTION OF BULKHEAD REMOVED. The shields were erected on timber cradles in the shaft, and were shoved forward to the face of the excavation. Concrete bulkheads, with the necessary air-locks, were then built across the tunnels behind the shields. The shields were erected before the dividing walls between the two contracts were placed. Rings of iron tunnel lining, backed by timbers spanning the openings on the west side, were erected temporarily across the shafts in order to afford a bearing for the shield jacks while shoving into the portals. The movement of the shield eastward was continued in each tunnel for a distance of about 60 ft., and the permanent cast-iron tunnel lining was erected as the shield advanced. Before breaking out of rock, it was necessary to have air pressure in the tunnels. This required the building of bulkheads with air-locks inside the cast-iron linings just east of the portals. Before erecting the bulkheads it was necessary to close the annular space between the iron tunnel lining and the rock. The space at the portal was filled with a concrete wall. After about twenty permanent rings had been erected in each tunnel, two rings were pulled apart at the tail of the shield and a second masonry wall or dam was built. The space between the two dams was then filled with grout. To avoid the possibility of pushing the iron backward after the air pressure was on, rings of segmental plates, 5/8 in. thick and 13-7/8 in. wide, were inserted in eighteen circumferential joints in each tunnel between the rings as they were erected. The plates contained slotted holes to match those in the segments. After the rings left the shield, the plates were driven outward, and projected about 5 in. When the tunnel was grouted, the plates were embedded. The bulkheads were completed, and the tunnels were put under air pressure on the following dates: Line D, on October 5th, 1905; Line C, on November 6th, 1905; Line B, on November 25th, 1905; Line A, on December 1st, 1905. This marked the end of the preparatory period. In the deepest part of the river, near the pier-head line on the Manhattan side, there was only 8 ft. of natural cover over the tops of the tunnels. This cover consisted of the fine sand previously described, and it was certain that the air would escape freely from the tunnels through it. To give a greater depth of cover and to check the loss of air, the contractor prepared to cover the lines of the tunnels with blankets of clay, which, however, had been provided for in the specifications. Permits, as described later, were obtained at different times from the Secretary of War, for dumping clay in varying thicknesses over the line of work. The dumping for the blanket allowed under the first permit was completed in February, 1906. The thickness of this blanket varied considerably, but averaged 10 or 12 ft. on the Manhattan side. The original blanket was of material advantage, but the depth of clay was insufficient to stop the loss of air. The essential parts of the shields in the four tunnels were exactly alike. Those in Tunnels B and D, however, were originally fitted with sectional sliding hoods and sliding extensions to the floors of the working chambers, as shown by Fig. 1, Plate LXV. The shields in Tunnels A and C were originally fitted with fixed hoods and fixed extensions to the floors, as shown in Fig. 2, Plate LXV. A full description of the shields will be found in Mr. Japp's paper. The shields in each pair of tunnels were advanced through the solid rock section about abreast of each other, until test holes from the faces indicated soft ground within a few feet. As the distance between the sides of the two tunnels was only 14 ft., it was thought best to let Tunnels B and D gain a lead of about 100 ft. before Tunnels A and C opened out into soft ground, in order that a blow from one tunnel might not extend to the other. Work in Tunnel C was shut down on December 23d, 1905, after exposing sand to a depth of 3 ft. at the top, and it remained closed for seven weeks. Work in Tunnel A was suspended on September 29th, 1905. By the time Tunnel B had made the required advance, it, together with Tunnels C and D, was overtaxing the capacities of the compressor plant. Only a little work was done in Tunnel C until July, 1906, and work in Tunnel A was not resumed until October 22d, 1906. [Pg 435] [Pg 436] TUNNELS DRIVEN WESTWARD FROM LONG ISLAND CITY. Materials and Inception of the Work.—The materials met in Tunnel A are typical of all four tunnels. From the Long Island shafts westward, in succession, there were 124 ft. of all-rock section, 125 ft. of part rock and part earth section, 22 ft. of all-rock section, 56 ft. of part rock and part earth section, 387 ft. of all-rock section, 70 ft. of part earth and part rock section, and 1,333 ft. of all-earth section. PLATE LXVII The materials passed through are indicated on Plate XIII. The rock was similar to that of the Blackwell's Island Reef, and was likewise covered by a layer of sand and boulders. The remainder of the soft ground was divided into three classes. The first was a very fine red sand, which occurred in a layer varying in thickness from 6 ft. to at least 15 ft. It may have been much deeper above the tunnel. It is the quicksand usually encountered in all deep foundations in New York City. The following is the result of the sifting test of this sand: Held on No. 30 sieve 0.6% Passed No. 30, " " No. 40 " 0.4% " No. 40, " " No. 50 " 0.7% " No. 50, " " No. 60 " 2.4% " No. 60, " " No. 80 " 14.9% " No. 80, " " No. 100 " 54.0% " No. 100, " " No. 200 " 8.0% " No. 200 " 19.0% 100.0% This means that grains of all but 4% of it were less than 0.0071 in. in diameter. The 19% which passed the No. 200 sieve, the grains of which were 0.0026 in. or less in diameter, when observed with a microscope appeared to be perfectly clean grains of quartz; to the eye it looked like ordinary building sand, sharp, and well graded from large to small grains. This sand, with a surplus of water, was quick. With the water blown out of it by air pressure, it is stable, stands up well, and is very easy to work. It appears to be the same as the reddish quicksand found in most deep excavations around New York City. The second material was pronounced "bull's liver" by the miners as soon as it was uncovered. "Bull's liver" seems to be a common term among English-speaking miners the world over. It is doubtful, however, if it is always applied to the same thing. In this case it consisted of layers of blue clay and very fine red sand. The clay seemed to be perfectly pure and entirely free from sand. It would break easily with a clean, almost crystalline, fracture, and yet it was soft and would work up easily. The layers of clay varied in thickness from 1/16 in. to 1 in., while the thickness of the sand layer varied from 1/4 in. to several inches. The sand was the same as the quicksand already described. The "bull's liver" was ideal material in which to work a shield. It stood up as well and held the air about as well as clay, and was much easier to handle. The third material was a layer of fine gray sand which was encountered in the top of all the tunnels for about 400 ft. just east of Blackwell's Island Reef. It was very open, and had grains of rather uniform size. During the starting out of the tunnels from the shafts, and for more than a year afterward, the roof of the working chamber in the caissons and the locks previously described under the Long Island shafts took the place of the bulkhead across the tunnels for confining the air pressure. The first work in air pressure was to remove the shield plug closing the opening in the side of the shaft. This being done, the shield was shoved through the opening, and excavation begun. At the start the shields were fitted with movable platforms, but no hoods of any kind were placed until after the rock excavation was completed. METHODS OF EXCAVATION. The distribution of materials to be excavated, as previously outlined, divided the excavation into three distinct classes, for which different methods had to be developed. These three classes were: First.—All-rock section. Second.—Rock in the bottom, earth in the top. Third.—All-earth section. [Pg 437] [Pg 438] The extent of the second and third classes was much greater than that of the first, and they, of course, determined the use of the shield. Shields had not previously been used extensively in rock work, either where the face was wholly or partly in rock, and it was necessary to develop the methods by experience. The specifications required that where rock was present in the bottom, a bed of concrete should be laid in the form of a cradle on which to advance the shield. All Rock.—At different times, three general methods were used for excavating in all-rock sections. They may be called: The bottom-heading method; the full-face method; and the center-heading method. The bottom-heading method was first tried. A heading, about 8 ft. high and 12 ft. wide, was driven on the center line, with its bottom as nearly as possible on the grade line of the bottom of the tunnel. It was drilled in the ordinary manner by four drills mounted on two columns. The face of the headings varied from 10 to 30 ft. in advance of the cutting edge. After driving the heading for about 10 ft., the bottom was cleared out and a concrete cradle was set. The width of the cradles varied, but was generally from 8 to 10 ft. The excavation was enlarged to full size as the shield advanced, the drills being mounted in the forward compartments of the shield, as shown by Fig. 1, Plate LXVII, which represents the conditions after the opening had been cut in the bulkhead, but before the new methods, mentioned later, had been developed. PLATE LXVIII The sides and top were shot downward into the heading. The area of the face remaining behind the heading was large, and a great number of holes and several rounds were required to fire the face to advantage. As soon as firing was started at the face, the heading was completely blocked, and operations there had to be suspended until the mucking was nearly completed. The bottom-heading method was probably as good as any that could be devised for use with the shields as originally installed. All the muck had to be taken from the face by hand and handled through the chutes or doors. By drilling from the shield, some muck was blasted on to the extensions of the floors and could be handled from the upper compartments. At best, however, the shield with the closed transverse bulkhead was a serious obstacle to rapid work in rock sections. The full-face method was only used where the rock was not considered safe for a heading. A cut was fired at the bottom, together with side holes, in a manner quite similar to that adopted in the first set of holes for a bottom heading. The cradle was then placed, in lengths of either 2.5 or 5 ft., after which the remainder of the face was fired in the same manner as for the bottom-heading method. The closed transverse bulkhead with air-locks, as shown in Fig. 1, Plate LXVI, was placed in the shield in the hope that it would only be necessary to maintain the full air pressure in the working compartments in front of the bulkhead. It was also thought that some form of bulkhead which could be closed quickly and tightly would be necessary to prevent flooding the tunnel in case of blows. While no attempt was ever made to reduce the pressure behind the shield bulkhead, it was obvious from the experience with Tunnels B and D, while working in the sand between Manhattan and the reef, that the plan was not practicable, and that the closed bulkhead in the bottom was a hindrance instead of a safeguard. As soon as rock was encountered in those tunnels at the west edge of the reef, the contractor cut through the bulkheads and altered them, as shown in Fig. 2, Plate LXVI. Taking advantage of the experience gained, openings were cut through the bulkheads in Shields A and C, while they were shut down near the edge of the Manhattan ledge. In erecting the shields at Long Island City in May and June, 1906, openings were also provided. These shields had to pass through about 700 ft. of rock at the start, the greater portion of which was all-rock section. It was at that point that openings were first used extensively and methods were developed, which would not have been possible except where ears could be passed through the shield. The bottom- heading method was first tried, but the working space in front of the shield was cramped, and but few men could be employed in loading the cars. To give more room, the heading was gradually wi...