Concrete Construction: Methods and Costs

The subaqueous concrete bucket shown by Figs. 27 and 28 is made by the Cyclopean Iron Works Co., Jersey City, N. J. Fig. 27 shows the bucket suspended full ready for lowering; the cover is closed and latched and the bail is held vertical by the tag line catch A. Other points to be noted are the eccentric pivoting of the bail, the latch unlocking lever and roller B and C, and the stop D. In the position shown the bucket is lowered through the water and when at the proper depth just above bottom the tag line is given a sharp pull, uncatching the bail. The body of the bucket turns bottom side up, revolving on the bail pivots, and just as the revolution is completed the bail engages the roller C on the latch unlocking lever and swings the lever enough to unlatch the top and allow it to swing down as shown by Fig. 28 and release the concrete. The stop D keeps the body of the bucket from swinging beyond the vertical in dumping.

Figures 29 and 30 show the subaqueous concrete bucket made by the G. L. Stuebner Iron Works, Long Island City, N. Y., essentially the same bucket, omitting the cover and with a peaked bail, is used for work in air. For subaqueous work the safety hooks A are lifted from the angles B and wired to the bail in the position shown by the dotted lines, and a tag line is attached to the handle bar C. The bucket being filled and the cover placed is lowered through the water to the bottom and then discharged by a pull on the tag line.

DEPOSITING IN BAGS.—Two methods of depositing concrete in bags are available to the engineer; one method is to employ a bag of heavy tight woven material, from which the concrete is emptied at the bottom, the bag serving like the buckets previously described simply as means of conveyance, and the other method is to use bags of paper or loose woven gunnysack which are left in the work, the idea being that the paper will soften or the cement will ooze out through the openings in the cloth sufficiently to bond the separate bagfuls into a practically solid mass.

The bag shown by Fig. 31 was used to deposit concrete for leveling up a rough rock bottom and so provide a footing for a concrete block pier constructed in 1902 at Peterhead, N. B., by Mr. William Shield, M. Inst. C. E. Careful longitudinal profiles were taken of the rock bottom one at each edge of the footing. Side forms were then made in 20-ft. sections as shown by Fig. 32; the lagging boards being cut to fit the determined profile and the top of the longitudinal piece being flush with the top of the proposed footing. The concrete was filled in between the side forms and leveled off by the T-rail straight-edge. In placing the side forms the longitudinal pieces were placed by divers who were given the proper elevations by level rods having 10 to 15-ft. extension pieces to raise the targets above the water surface. When leveled the side pieces were anchor-bolted as shown to the rock, the anchor-bolts being wedged into the holes to permit future removal. The concrete was then lowered in the bag shown by Fig. 31, the divers assisting in guiding the bag to position. The mouth of the bag being tied by one turn of a line having loops through which a wooden key is slipped to hold the line tight, a sharp tug on the tripping rope loosens the key and empties the bag. The bags used on this work had a capacity of 2¼ cu. ft. To permit the removal of the side forms after the concrete had hardened, a strip of jute sacking was spread against the lagging boards with a flap extending 15 to 18 ins. under the concrete. The forms were removed by divers who loosened the anchor bolt wedges.

In placing small amounts of concrete for bridge foundations in Nova Scotia, bags, made of rough brown paper were used to hold the concrete. Each bag held about 1 cu. ft. The bags were made up quickly and dropped into the water one after the other so that the following one was deposited before the cement escaped from the former one. The paper was immediately destroyed by submersion and concrete remained. The bags cost $1.35 per hundred or 35 cts. per cu. yd. of concrete. Concrete was thus deposited in 18 ft. of water without a diver.

DEPOSITING THROUGH A TREMIE.—A tremie consists of a tube of wood or, better, of sheet metal, which reaches from above the surface to the bottom of the water; it is operated by filling the tube with concrete and keeping it full by successive additions while allowing the concrete to flow out gradually at the bottom by raising the tube slightly to provide the necessary opening. A good example of a sheet steel tremie is shown by Fig. 33. This tremie was used by Mr. Wm. H. Ward in constructing the Harvard Bridge foundations and numerous other subaqueous structures of concrete. In these works the tube was suspended from a derrick. Wheelbarrows filled the tube and hopper with concrete and kept them full; the derrick raised the tube a few inches and swung it gently so as to move it slowly over the area to be filled. Care being taken to keep the tube at one height, the concrete was readily deposited in even layers. Concrete thus deposited in 18 ft. of water was found to be level and solid on pumping the pit dry.

Another method of handling a tremie was employed in constructing the foundations for the Charlestown Bridge at Boston, Mass. Foundation piles were driven and sawed off under water. A frame was built above water and supported by a curbing attached to certain piles in the outer rows of the foundation reserved for this purpose. In this frame the vertical members were Wakefield sheet-piling plank, spaced 6 to 10 ft. apart, and connected by three lines of double waling bolted to the verticals at three different heights. This frame was lowered to the bottom so as to enclose the bearing piles. The posts or verticals were then driven, one by one, into the bottom, the frame being flexible enough to permit this. The spaces between the posts or verticals were then filled by sheet-piling and the frame was bolted to the curbing piles. This curbing afterward supported the traveler used in laying the concrete. Thus a coffer dam was formed to receive the concrete as shown in Fig. 34. The 1-2-5 concrete was deposited up to within 5½ ft. of the mean low water level, the last foot being laid after water was pumped out. The tremie used to deposit the concrete was a tube 14 ins. in diameter at the bottom and 11 ins. at the neck, with a hopper at the top. It was made in removable sections, with outside flanges, and was suspended by a differential hoist from a truck moving laterally on a traveler, Fig. 34. The foot of the chute rested on the bottom until filled with concrete; then the chute was slowly raised and the concrete allowed to run but into a conical heap, more concrete being dumped into the hopper. As the truck moved across the traveler a ridge of concrete was made; then the traveler was moved forward and another parallel ridge was made. The best results were obtained when the layers were 2½ ft. thick, but layers up to 6 ft. thick were laid. If the layer was too thick, or uneven, or if the chute was moved or raised too quickly, the charge in the tube was "lost." This was objectionable because the charging of the chute anew resulted in "washing" the cement more or less out of the concrete until the chute was again filled. To reduce this objection the contractor was directed to dump some neat cement into the tube before filling with concrete. A canvass piston was devised which could be pushed ahead of the concrete when filling the chute. It consisted of two truncated cones of canvass, one flaring downward to force the water ahead, and the other flaring upward to hold the concrete. The canvass was stiffened and held against the sides of the chute by longitudinal ribs of spring steel wire; the waist was filled by a thick block of wood to which all the springs were attached; and to this block were connected additional steel guides to prevent overturning and a rope to regulate the descent. Very little water forced its way past this piston and it was a success, but as the cost was considerable and a piston was lost each time, its use was abandoned as the evil to be avoided did not justify the outlay.

The chute worked best when the concrete was mixed not quite wet enough to be plastic. If mixed too wet the charge was liable to be "lost," and if dry it would choke the chute. An excess of gravel permitted water to ascend in the tube; and an excess of sand tended to check the flow of concrete.

In constructing the piers for a masonry arch bridge in France in 1888 much the same method was followed, except that a wooden tremie 16 ins. square made in detachable sections was used. This tremie had a hopper top and was also provided with a removable cap or cover for the bottom end, the latter device being intended to keep the water out of the tube and prevent "washing" the first charge of concrete. The piers were constructed by first driving piles and sawing them off several feet above the bottom but below water level, and then filling them nearly to their tops with broken stone. An open box caisson was then sunk onto the stone and embracing the pile tops and then filled around the outside with more broken stone. The caisson was then filled with concrete through the tremie which was handled by a traveling crane. The crane was mounted and traveled transversely of the pier on a platform which in turn moved along tracks laid lengthwise of the caisson. The tube was gradually filled with concrete and lowered, the detachable bottom of the tube was then removed, allowing the concrete to run out. The tube was first moved across the caisson and then downstream and back across the caisson, and this operation repeated until a 16-in. layer was completed. The tube was then raised 16 ins. and the operations repeated to form another layer. There was almost no laitance. From 90 to 100 cu. yds. were deposited daily.

Still another example of tremie work is furnished by the task of depositing a large mass of concrete under water in the construction of the Nussdorf Lock at Vienna. This lock has a total width of 92 ft. over all, and is 49.2 ft. clear inside. The excavation, which was carried to a depth of 26.24 ft. below water level, was made full width, between sheet piling, and the bottom was filled in with rammed sand and gravel, forming a kind of invert with its upper surface horizontal in the middle and sloping upwards a trifle at both sides. A mass of concrete having a total thickness of 13.12 ft. was built on this foundation in the center where the upper surfaces were 13.12 ft. below the water level. Concrete walls were carried up at the sides of the lock to a height of 3.28 ft.; these walls were 8.2 ft. thick. The methods used in placing the concrete were as follows: Three longitudinal rows of piles were driven on each side of the axis of the lock, these piles supporting a 6-rail track about 7 ft. above the water level. Three carriages spanning the full width of the lock transversely moved on this track. Each carriage had three trolleys, one in each of the main panels of the transverse pile bends. These trolleys each carried a vertical telescopic tube, by means of which the concrete was deposited at the bottom of the lock. These tubes or chutes were of different lengths in the three carriages; the first ones deposited the concrete up to a level of 23 ft. below the surface; the next set deposited the concrete between that level and 19.7 ft., and the last set completed the subaqueous work up to the final height of 16.4 ft. below the surface. The tops of the tubes were level with a transverse track extending the full length of the carriage. The ends of these tracks just cleared the outside rows of piles, which, on one side of the lock, supported a distribution track parallel to the axis of the lock. Dump cars running on this distribution track delivered the concrete to smaller dump cars on the carriage tracks, and in turn these smaller cars dumped into either of these chutes on each carriage. The carriages were moved from end to end of the lock, the whole area of the lock coming under the nine chutes, inasmuch as each chute moved one-third the length of the carriage. The concrete was deposited in three horizontal layers 3.28 ft. thick, the layers being built in comparatively narrow banks, so that the different layers would key together and form a corrugated mass. The chutes were shortened as the concrete was deposited, three layers being placed successively. The main body of the bottom and the side walls were built by this method, and then the water was pumped out and a 2.3 ft. layer of concrete rammed over the bottom and completed with a finished surface 9 ft. thick.

GROUTING SUBMERGED STONE.—Masses of gravel, broken or rubble stone deposited under water may be cemented into virtually a solid concrete by charging the interstices with grout forced through pipes from the surface. Mr. H. F. White gives the following records of grouting submerged gravel:

In experiment No. 1 a reservoir 10 ft. square was filled to a depth of 18 ins. with clean gravel ballast (1½ to 2-in. size) submerged in water. A 2-in. gas pipe rested on the gravel and was surmounted with a funnel. A 1:1 Portland grout was poured in. After 21 days set the water was drawn off, and it was found that the grout had permeated the ballast for a space of 8 ft. square at the bottom and 6 ft. square at the top, leaving a small pile of pure cement mortar 6 ins. high about the base of the pipe; 16 cu. ft. of cement and 16 cu. ft. of sand concreted 100 cu. yds. of ballast. In experiment No. 2, under the same conditions, a grout made of 1 part lime, 1 part surki (puzzulana or trass) and 1 part sand, was found to have spread over the entire bottom, 10 ft. square, rising 5 ins. on the sides, and making the concreted mass about 3½ ft. square at the top; 25 cu. ft. of the dry materials concreted 100 cu. ft. of ballast. In experiment No. 3 the ballast was 2½ ft. deep. A grout (using 8 cu. ft. of each ingredient) made as in experiment No. 2 covered the bottom, rose 14 ins. on the sides and made a top surface 4½ ft. square; 32 cu. ft. of the dry materials grouted 100 cu. ft. of ballast. In experiment No. 4 the ballast was of bats and pieces 3 or 4 ins. in size laid 7 ft. deep. A grout made as in experiment No. 2 (using 88 cu. ft. of each ingredient) concreted the whole mass to a depth of 6 ft. up the sides, and 2½ ft. square at the pipe on the surface of the ballast. Mr. White says that a grout containing more than 1 part of sand to 1 of Portland cement will not run freely through a 2-in. pipe, as the sand settles out and chokes the pipe. Even with 1:1 grout it must be constantly stirred and a steady flow into the pipe maintained. The lime-trass grout does not give the same trouble.

Mr. W. R. Knipple describes the work of grouting rubble stone and gravel for the base of the Hermitage Breakwater. This breakwater is 525 ft. long, 50 ft. wide at base and 42 ft. wide at top, and 68 ft. high, was built on the island of Jersey. Where earth (from 0 to 8½ ft. deep) overlaid the granite rock, it was dredged and the trench filled in with rubble stones and gravel until a level foundation was secured. Cement grout was then forced into this filling through pipe placed 8 to 10 ft. apart. The grouting was done in sections 12½ ft. long, from 7 to 10 days being taken to complete each. Upon this foundation concrete blocks, 4×4×9 to 12 ft., were laid in courses inclined at an angle of 68°. The first four courses were laid by divers, the blocks being stacked dry two courses high at a time. The joints below water were calked by divers and above water by masons, and a section was then grouted. When two courses had been laid and grouted, two more courses were laid and grouted in turn, and so on. In places, grouting was done in 50 ft. of water. The grout should be a thick paste; a 30-ft. column of grout will balance a 60-ft. column of water.

CHAPTER VI.

METHODS AND COST OF MAKING AND USING RUBBLE AND ASPHALTIC CONCRETE.

Two kinds of concrete which vary in composition and character from the common standard mixtures of cement, sand and broken aggregate are extensively employed in engineering construction. These are rubble concrete and asphaltic concrete.

RUBBLE CONCRETE.—In constructing massive walls and slabs a reduction in cost may often (not always) be obtained by introducing large stones into the concrete. Concrete of this character is called rubble concrete, and the percentage of rubble stone contained varies from a few per cent. to, in some cases, over half of the volume. The saving effected comes partly from the reduction in the cement required per cubic yard of concrete and partly from the saving in crushing.

The saving in cement may be readily figured if the composition of the concrete and the volume of the added rubble stones be known. A 1-2½-5 concrete requires according to Table X in Chapter II 1.13 bbls. of cement per cubic yard. Assuming a barrel of cement to make 3.65 cu. ft. of paste, we have 3.65 × 1.13 = 4.12 cu. ft. of cement paste per cubic yard of 1-2½-5 concrete. This means that about 15 per cent. of the volume of the concrete structure is cement. If rubble stone be introduced to 50 per cent. of the volume, then the structure has about 7½ per cent. of its volume of cement. It is of interest to note in this connection that rubble masonry composed of 65 per cent. stone and 35 per cent. of 1-2½ mortar would have some 11½ per cent. of its volume made up of cement.

The saving in crushing is not so simple a determination. Generally speaking, the fact that a considerable volume of the concrete is composed of what, we will call uncrushed stone, means a saving in the stone constituent of one structure amounting to what it would have cost to break up and screen this volume of uncrushed stone, but there are exceptions. For example, the anchorages of the Manhattan Bridge over the East River at New York city were specified to be of rubble concrete, doubtless because the designer believed rubble concrete to be cheaper than plain concrete. In this case an economic mistake was made, for all the rubble stone used had to be quarried up the Hudson River, loaded onto and shipped by barges to the site and then unloaded and handled to the work using derricks. Now this repeated handling of large, irregular rubble stones is expensive. Crushed stone as we have shown in Chapter IV can be unloaded from boats at a very low cost by means of clam shells. It can be transported on a belt conveyor, elevated by bucket conveyer, mixed with sand and cement and delivered to the work all with very little manual labor when the installation of a very efficient plant is justified by the magnitude of the job. Large rubble stones cannot be handled so cheaply or with so great rapidity as crushed stone; the work may be so expensive, due to repeated handlings, as to offset the cost of crushing as well as the extra cost of cement in plain concrete. On the other hand, the cost of quarrying rock suitable for rubble concrete is no greater than the cost of quarrying it for crushing—it is generally less because the stone does not have to be broken so small—so that when the cost of getting the quarried rock to the crusher and the crushed stone into the concrete comes about the same as getting the quarried stone into the structure it is absurd practice to require crushing. To go back then to our first thought, the question whether or not saving results from the use of rubble concrete, is a separate problem in engineering economics for each structure.

In planning rubble concrete work the form of the rubble stones as they come from the quarry deserves consideration. Stones that have flat beds like many sandstones and limestones can be laid upon layers of dry concrete and have the vertical interstices filled with dry concrete by tamping. It requires a sloppy concrete to thoroughly embed stones which break out irregularly. In the following examples of rubble concrete work the reader will find structures varying widely enough in character and in the percentages of rubble used to cover most ordinary conditions of such work.

Where the rubble stones are very large it is now customary to use the term "cyclopean masonry" instead of rubble concrete. Many engineers who have not studied the economics of the subject believe that the use of massive blocks of stone bedded in concrete necessarily gives the cheapest form of masonry. We have already indicated conditions where ordinary concrete is cheaper than rubble concrete. We may add that if the quarry yields a rock that breaks up naturally into small sized blocks, it is the height of economic folly to specify large sized cyclopean blocks. Nevertheless this blunder has been frequently made in the recent past.

Chattahoochee River Dam.—The roll-way portion, 680 ft. long, of the dam for the Atlanta Water & Electric Power Co., shown in section by Fig. 35, was built of a hearting of rubble concrete with a fine concrete facing and a rubble rear wall. The facing, 12 ins. thick of 1-2-4 concrete, gave a smooth surface for the top and face of the dam, while the rubble rear wall enabled back forms to be dispensed with and, it was considered, made a more impervious masonry. The concrete matrix for the core was a 1-2-5 stone mixture made very wet. The rubble stones, some as large as 4 cu. yds., were bedded in the concrete by dropping them a few yards from a derrick and "working" them with bars; a well formed stone was readily settled 6 ins. into a 10-in. bed of concrete. The volume of rubble was from 33 to 45 per cent. of the total volume of the masonry. The 1-2-4 concrete facing was brought up together with the rubble core, using face forms and templates to get the proper profile. The work was done by contract and the average was 5,500 cu. yds. of concrete placed per month.

Barossa Dam, South Australia.—The Barossa Dam for the water-works for Gawler, South Australia, is an arch with a radius of 200 ft., and an arc length on top of 422 ft.; its height above the bed of the stream is 95 ft. Figure 36 is a cross-section of the dam at the center. The dam contains 17,975 cu. yds. of rubble concrete in the proportions of 2,215 cu. yds. of rubble stone to 15,760 cu. yds. of concrete; thus about 12.3 per cent. of the dam was of rubble. The concrete was mixed by weight of 1 part cement, 1½ parts sand, and a varying proportion of aggregate composed of 4½ parts 1¼ to 2-in. stone, 2 parts ½ to 1¼-in. stone and 1 part ⅛ to ½-in. stone or screenings. The sand was one-half river sand and one-half crusher sand. The following shows the amounts by weight of the several materials for each of the several classes of concrete per cubic yard:

The average composition of the concrete was 1-1½-3½. Its cost per cubic yard in place including rubble was 38s 9d per cu. yd. or about $9.30. In proportioning the mixture on the work use was made of the device shown by Fig. 37 to weigh the aggregate. The measuring car is pushed back under the stone hopper chute until the wheels drop into shallow notches in the balanced track rails; stone is then admitted until the lead weight begins to rise, when the car is pushed forward and dumps automatically as indicated.

Other Rubble Concrete Dams.—Rubble concrete containing from 55 to 60 per cent. rubble was used in constructing the Boonton Dam at Boonton, N. J. The stones used measured from 1 to 2½ cu. yds. each; the concrete was made so wet that when the stones were dropped into it, it flowed into every crevice. The materials were all delivered on cars, from which they were delivered to the dam by derricks provided with bull-wheels. On the dam there were 4 laborers and 1 mason to each derrick, and this gang dumped the concrete and joggled the rubble stones into it. Records of 125 cu. yds. per 10 hours, with one derrick, were made. With 35 derricks, 20 of which were laying masonry and 15 either passing materials or being moved, as much as 21,000 cu. yds. of masonry were laid in one month. The amount of cement per cubic yard of masonry is variously stated to have been 0.6 to 0.75 bbl. The stone was granite.

The Spier Falls Dam on the upper Hudson River was built of rubble concrete containing about 33 per cent. rubble stone. The concrete was a 1-2½-5 mixture, and the engineer states that about 1 bbl. of cement was used per cubic yard of rubble concrete. This high percentage of cement may be accounted for by the fact that there was a considerable amount of rubble masonry in cement mortar included in the total. The stones and concrete were delivered along the dam by cableways and stiff-leg derricks set on the downstream sloping face of the dam delivered them from the cableways into place. There were two laborers to each mason employed in placing the materials, wages being 15 and 35 cts. per hour, respectively. The labor cost of placing the materials was 60 cts. per cubic yard of masonry. The stone was granite.

Granite rubble laid in layers on beds of concrete and filled between with concrete was used in constructing the Hemet Dam in California. The concrete was a 1-3-6 mixture, and was thoroughly tamped under and between the stones. For face work the stones were roughly scabbled to shape and laid in mortar. The stone was taken from the quarry 400 ft. away and delivered directly on the dam by cableways; here two derricks handled the stone into place, the dam being only 246 ft. arc length on top, though it was 122½ ft. high. The cableways would take a 10-ton load; stones could be taken from the quarry, hoisted 150 ft. and delivered to the work in 40 to 60 seconds. Common labor at $1.75 per day was used for all masonry except facing, where masons at $3.50 were employed. Cement cost delivered $5 per barrel, of which from $1 to $1.50 per barrel was the cost of hauling 23 miles by team over roads having 18 per cent. grades in places. Sand was taken from the stream bed and delivered to the work by bucket conveyor. "Under favorable conditions some of the masonry was put in for as low as $4 per cu. yd." There were 31,100 cu. yds. of masonry in the dam, which required 20,000 bbls. of cement, or 0.64 bbl. per cubic yard.

The following novel method of making rubble concrete was employed in enlarging two old dams and in constructing two new dams for a small water-works. The available time was short, the amount of work was too small and too scattered to justify the installation of a stone crusher, and suitable gravel was not at hand. Sufficient small boulders in old walls, and borrow pits and on surface of fields were available, and were used with thin Portland cement mortar. One part of Alpha or Lehigh cement and three parts sand were mixed dry at first and then wet with just enough water to make the resulting mortar flow by gravity. This mortar was shoveled into the forms continuously by one set of men while other men were throwing into the mortar in the forms the boulders which were cleaned and broken so as not to be more than 7 ins. long. In general the performance was continuous. Three mortar beds were placed parallel with, and against, one side of the forms, with spaces of about 4 ft. between the ends of the beds. The boulders were dumped on the opposite side of the forms. Two men shoveled in all the mortar and did nothing else. While they were emptying one bed the mortar was being mixed in the preceding bed by two other men and the materials placed in the third bed by still others. Another gang was continually throwing in the boulders and small stones and still another was breaking stone. One man should keep the mortar well stirred while the bed is being emptied. About 20 men were necessary to do all parts of the work. The forms were of 2-in. planed plank tongued and grooved. Especial pains were taken to make the forms tight, and all leaks that appeared were quickly stopped with dry cement. Some pains were taken to prevent a flat side of large stones from coming in direct contact with the forms, but round boulders and small stones needed no care to prevent their showing in the finished work.

In conclusion it is interesting to note, perhaps, the earliest use of rubble concrete for dam construction in this country in constructing the Boyd's Corner Dam on the Croton River near New York. This dam was begun in 1867 and for a time rubble concrete was used, but was finally discontinued, due to the impression that it might not be watertight. The specifications called for dry concrete to be thoroughly rammed in between the rubble stones, and to give room for this ramming the contractor was not permitted to lay any two stones closer together than 12 ins. As a result not more than 33 per cent. of the concrete was rubble.

Abutment for Railway Bridge.—Figure 38 shows a bridge abutment built of rubble concrete at a cost of about $4.50 per cu. yd. The concrete was a 1-2½-4½ mixture laid in 4-in. layers. On each layer were laid large rubble stones bedded flat and spaced to give 6-in. vertical joints; the vertical joints were filled with concrete by ramming and then another layer of concrete placed and so on. A force of 28 men and a foreman averaged 40 cu. yds. of rubble concrete per day. The following is the itemized cost per cubic yard, not including forms, for 278 cu. yds:

Some English Data on Rubble Concrete.—Railway work, under Mr. John Strain, in Scotland and Spain, involved the building of abutments, piers and arches of rubble concrete. The concrete was made of 1 part cement to 5 parts of ballast, the ballast consisting of broken stone or slag and sand mixed in proportions determined by experiment. The materials were mixed by turning with shovels 4 times dry, then 4 times more during the addition of water through a rose nozzle. A bed of concrete 6 ins. thick was first laid, and on this a layer of rubble stones, no two stones being nearer together than 3 ins., nor nearer the forms than 3 ins. The stones were rammed and probed around with a trowel to leave no spaces. Over each layer of rubble, concrete was spread to a depth of 6 ins. The forms or molds for piers for a viaduct were simply large open boxes, the four sides of which could be taken apart. The depth of the boxes was uniform, and they were numbered from the top down, so, that, knowing the height of a given pier, the proper box for the base could be selected. As each box was filled, the next one smaller in size was swung into place with a derrick. The following bridge piers for the Tharsis & Calanas Railway were built:

It is stated that the construction of some of these piers in ordinary masonry would have taken four times as long. The rock available for rubble did not yield large blocks, consequently the percentage of pure concrete in the piers was large, averaging 70 per cent. In one case, where the stones were smaller than usual, the percentage of concrete was 76½ per cent. In other work the percentage has been as low as 55 per cent., and in still other work where a rubble face work was used the percentage of concrete has been 40 per cent.

In these piers the average quantities of materials per cubic yard of rubble concrete were:

448 lbs. (0.178 cu. yd.) cement.
0.36 cu. yd. sand.
0.68 cu. yd. broken stone (measured loose in piles).
0.30 cu. yd. rubble (measured solid).

Several railway bridge piers and abutments in Scotland are cited. In one of these, large rubble stones of irregular size and weighing 2 tons each were set inside the forms, 3 ins. away from the plank and 3 ins. from one another. The gang to each derrick was: 1 derrick man and 1 boy, 1 mason and 10 laborers, and about one-quarter of the time of 1 carpenter and his helper raising the forms. For bridges of 400 cu. yds., the progress was 12 to 15 cu. yds. a day. The forms were left in place 10 days.

To chip off a few inches from the face of a concrete abutment that was too far out, required the work of 1 quarryman 5 days per cu. yd. of solid concrete chipped off.

Concrete was used for a skew arch over the River Dochart, on the Killin Railway, Scotland. There were 5 arches, each of 30 ft. span on the square or 42 ft. on the skew, the skew being 45°. The piers were of rubble concrete. The concrete in the arch was wheeled 300 ft. on a trestle, and dumped onto the centers. It was rammed in 6-in. layers, which were laid corresponding to the courses of arch stones. As the layers approached the crown of the arch, some difficulty was experienced in keeping the surfaces perpendicular. Each arch was completed in a day.

In a paper by John W. Steven, in Proc. Inst. C. E., the following is given:

Mr. Martin Murphy describes some bridge foundations in Nova Scotia. Rubble concrete was used in some of the piers. The rubble concrete consisted of 1 part cement, 2 parts sand, 1 part clean gravel, and 5 parts of large stones weighing 20 lbs. each and upwards. The sand, cement and gravel were turned three times dry and three times wet, and put into the forms. The rubble stones were bedded in the concrete by hand, being set on end, 2 or 3 ins. apart. No rubble stones were placed within 6 ins. of the forms, thus leaving a face of plain concrete; and the rubble stones were not carried higher than 18 ins. below the top of the pier. One cubic yard of this rubble concrete required 0.8 to 0.9 bbl. of cement.

ASPHALT CONCRETE.—Asphalt or tar concrete in which steam cinders or broken stone or gravel and sand are mixed with asphaltum or tar instead of cement paste are used to some extent in lining reservoirs, constructing mill floors, etc. Such mixtures differ in degree only from the mixtures used for asphalt street paving, for discussion of which the various books on paving and asphalts should be consulted. The two examples of asphalt concrete work given here are fairly representative of the mixtures and methods employed for concrete work as distinguished from asphalt work.

Slope Paving for Earth Dam.—Mr. Robert B. Stanton describes a small log dam faced upstream with earth, upon which was laid an asphalt concrete lining to make it water tight. The stone was broken to 2-in. pieces, all the fines being left in and sufficient fine material added to fill the voids. The stone was heated and mixed in pans or kettles from a street paving outfit; and the asphaltum paste, composed of 4 parts California refined asphaltum and 1 part crude petroleum, was boiled in another kettle. The boiling hot paste was poured with ladles over the hot stone, and the whole mixed over the fire with shovels and hoes. The asphalt concrete was taken away in hot iron wheelbarrows, placed in a 4-in. layer rammed and ironed with hot irons. The concrete was laid in strips 4 to 6 ft. wide, the edges being coated with hot paste. After the whole reservoir was lined, it was painted with the asphalt paste, boiled much longer, until when cold it was hard and stone was broken to 2-in. pieces, all the fines being left in and sufficient fine material added to fill the voids. The stone was heated and mixed in pans or kettles from a street paving outfit; and the asphaltum paste, composed of 4 parts California refined asphaltum and 1 part crude petroleum, was boiled in another kettle. The boiling hot paste was poured with ladles over the hot stone, and the whole mixed over the fire with shovels and hoes. The asphalt concrete was taken away in hot iron wheelbarrows, placed in a 4-in. layer rammed and ironed with hot irons. The concrete was laid in strips 4 to 6 ft. wide, the edges being coated with hot paste. After the whole reservoir was lined, it was painted with the asphalt paste, boiled much longer, until when cold it was hard and brittle, breaking like glass under the hammer. This paste was put on very hot and ironed down. It should not be more than ⅛-in. thick or it will "creep" on slopes of 1½ to 1. After two hot summers and one cold winter there was not a single crack anywhere in the lining. A mixture of sand and asphalt will creep on slopes of 1½ to 1, but asphalt concrete will not. With asphalt at $20 a ton, and labor at $2 a day, the cost was 15 cts. a sq. ft. for 4-in. asphalt concrete. On a high slope Mr. Stanton recommends making slight berms every 6 ft. to support the concrete and prevent creeping. Asphalt concrete resists the wear of wind and water that cuts away granite and iron.

Base for Mill Floor.—In constructing 17,784 sq. ft. of tar concrete base for a mill floor, Mr. C. H. Chadsey used a sand, broken stone and tar mixture mixed in a mechanical mixer. The apparatus used and the mode of procedure followed were as follows:

Two parallel 8-in. brick walls 26 ft. long were built 4 ft. apart and 2½ ft. high to form a furnace. On these walls at one end was set a 4×6×2 ft. steel plate tar heating tank. Next to this tank for a space of 4×8 ft. the walls were spanned between with steel plates. This area was used for heating sand. Another space of 4×8 ft. was covered with 1½ in. steel rods arranged to form a grid; this space was used for heating the broken stones. The grid proved especially efficient, as it permitted the hot air to pass up through the stones, while a small cleaning door at the ground allowed the screenings which dropped through the grid to be raked out and added to the mixture. A fire from barrel staves and refuse wood built under the tank end was sufficient to heat the tar, sand and stone.

For mixing the materials a Ransome mixer was selected for the reason that heat could be supplied to the exterior of the drum by building a wood fire underneath. This fire was maintained to prevent the mixture from adhering to the mixing blades, and it proved quite effective, though occasionally they would have to be cleaned with a chisel bar, particularly when the aggregate was not sufficiently heated before being admitted to the mixer. A little "dead oil" applied to the discharge chute and to the shovels, wheelbarrows and other tools effectually prevented the concrete from adhering to them.

The method of depositing the concrete was practically the same as that used in laying cement sidewalks. Wood strips attached to stakes driven into the ground provided templates for gaging the thickness of the base and for leveling off the surface. The wood covering consisted of a layer of 2-in. planks, covered by matched hardwood flooring. In placing the planking, the base was covered with a ¼-in. layer of hot pitch, into which the planks were pressed immediately, the last plank laid being toe-nailed to the preceding plank just enough to keep the joint tight. After a few minutes the planks adhered so firmly to the base that they could be removed only with difficulty. The hardwood surface was put on in the usual manner. The prices of materials and wages for the work were as follows: