Vol

Great Salt Lake Crossing

New sampler speeds design of

H. V. ANDERSON, J. M. ASCE, International Engineering Company, Inc.~

P1anning the new Great Salt Lake trestle and the 11 miles of rock fill, Severe storms, whieh occur frequently crossing of the Southern Pacific Rail­ comprising the main part of the on Great Salt Lake, have subjected road required an extensive study of project, were completed in less than the fills to destructive wave action. the existing facilities and their pos­ two years. Even making no allowance To prevent water from washing them sible future, and a study of the pos­ for the type of equipment available away, concrete paving was placed over sibility of supporting a fill on the in 1904, this was a construction mar­ the rock. Finally, huge limestone riprap lake-bottom clays. 1 he Swedi;;h foil vel. was built up on the sides. Even though sampler proved very adaptable in pro­ The wood trestle across Great Salt the fill material has been eroded and viding undisturbed samples of the clay Lake bas had a long life. The salt there are flat slopes below the water from which soil-strength characteristics brine of the lake is a natural preserva­ surface, which assist in maintaining could be determined. Through a series tive. No living organism except brine stability of the riprap, the fills still of investigations, it was found fea.sible shrimp exists in it. No insect will require maintenance. to replace the existing trestle with a harm wood that has been immersed rock fill. in the lake. After continual soaking, Repair or replace is question For many years the Southern Pa­ wood fibers become hard and brittle Mounting maintenance costs, the cific has foreseen the necessity oi re­ and do not decay. Freezing and thaw­ necessity of redecking the existing building its Great Salt Lake cutoff. In ing action is limited. Concentrated salt trestle in the immediate future and August 1953, it contracted with the water has a less corrosive _effect on about every 25 years thereafter, plm International Engineering Co., Inc. iron than sea water, so the bolts used need for the replacement of the timber (IECO) for a study of the engineering in the trestle are in much better con­ piling in the next 25 years, dictated and economic factors involved in the dition than would normally be ex­ a thorough study of the problems and reconstruction or replacement of the pected after fifty years of service. possibilities to find some other means crossing. This work included taking un­ Despite the preservative effect of of crossing Great Salt Lake. The disturbed samples and making founda­ the salt, many timber members have following plans were studied: tion borings, laboratory tests, analyses, had to be replaced. A new deck, helper 1. Redeck existing trestle and replace with a engineering design, cost estimates, and stringers, and horizontal bracing strut5 new trestle after 25 years economic studies. were added between 1920 and 1927. 2. Reconstruct a new trestle and redeck it The early railroad builders aYoided To keep up with the increasing Eize after 30 years the problem of the soft clays in the of trains and freight loads, it has been 3. Construct a new parallel trestle and redeck bottom of Great Salt Lake by building necessary to strengthen the bents and existing trestle the original route around the north increase their size. Bracing piles were 4. Redeck existing trestle and construct a rock end of the lake instead of across it. added from 1943 to 1946 to provide fill after 25 years Although the advantage of a route additional longitudinal rigidity. 5. Construct a rock fill with 3,000 ft of trest le across the lake was always realized, A fire destroyed 645 ft of double­ 6. Construct a rock fill with 24,000 ft of ne" trestle it was not until 1902 that the con­ track trestle in 1956, after construction struction of the Lucin Cutoff was had started on the new fill. The pile;:. 7. Construct a concrete trestle finallv undertaken. Rock was hauled thought to be fire-resistant as a result 8. Build a rock fill 13 rn ::es long from-quarries at each end until founda­ of continual soaking with salt water The key factor in determining th< tion failures of the soft clays halted · spray, were destroyed as the fire con­ most economic means of crossing th( the operation. Only then did con­ sumed their centers. The bents were lake is the clays which make up tht struction of a wooden trestle begin. rebuilt by placing prefabricated pony­ lake bottom and which must provid( Although the lake was no more than bents on the original piles after the the foundation for any fill or structure 20 to 35 ft deep, wood piles 120 ft burned upper sections had been sawed In the past these clays have beer long were used for 11,000 ft at the off. responsible for the settlement of th, western end and 1,500 ft at the eastern The original fills in the lake, dumped fill, for slides and slip-outs, and fo; end. The remainder of the trestle rests on the unstable clay bottom, haw -­ erosion of the support ~d for slope on shorter piles, driven to refusal on required expensive maintenance. The protection materials. Failure of the,:, the salt stratum. The 12 miles of wood rock fills settled for years after placing. clays to support fill in 1902 resulte,

40 (fol. p. 846) December 1957 • CIVIL ENGINEE~( 31,000,000-cu yd fill

San Francisco, Calif.

In foil sampler, at left, thin metal foils from reels in head come in contact with soil as it enters sampler and move upward with it as sampler is lowered. This Model VII Swedish sampler is very similar to that used for work described. In view above, foil sampler is forced down into clay by a grip-hoist. Mecha­ nism of sampler includes dial gage, which shows tension on foils, controlled by hand wheel.

Original lake bottom FIG. l. Gravel blanket up to 400 ft wide helps support rock fill where lake bottom Limit beyond which no gravel may prqect is soft clay. On thick salt TYPICAL SECTION ON SALT FOUNDATION layer in center of lake, width of blanket is gener­ ally le11& than 200 ft. Fill ,. has top width of 38 ft and I Tarying depth-<1bout 70 to 75 ft in central section.

Limit beyond which no gravel may project TYPICAL SECTION ON CLAY FOUNDATION

IL ENGINEERING • Decembflf' 1957 (Vol. p. 847) '1 encountered by the piston. The re­ lationship between the results obtained by the vane and those obtained by the penetrometer at platforms No. 1 and No. 2 are shown in Fig. 2. The Sanborne recording instrument \Y:1s Yery susceptible to the salty atmos­ phere. Because of the operation dif­ ticulties of the electronic equipment, the 'CSBR Yane shear de,·ice was ~elected for most of the work. The undisturbed samples were ob­ tained by using a 3-in. Hvorslev piston ~ampler developed b~- the USBR. A ~oils laboratory was set up at the job site, where unconfined compression tests, quick triaxial compression tests, consolidation tests, determinations of water content, density and Atterberg limits were made from the samples. The relationship between the vane tests and Supposedly fire-safe trestle caught fire after work on new fill had started. Railroad the laboratory tests for shear strength and construction crews pooled their efforts to rebuild 645-ft section in one week. is shown in Fig. 3. After a review of the laboratory results, it was decided that unconfined compression tests of carefully taken undisturbed samples would be used for future shear-strength in construction of the costly timber salt was outlined. This deposit ex­ determinations of the foundation clays. trestle, which is susceptible to fire and tends across the center of the lake for Additional data were needed to earthquake .and unsafe for fast freight a di5tance of some 5 miles. assure that the most economically safe traffic. A hydrographic survey of the pro­ section would be selected for crossing Study of these underlying clays was po,..<>ed crossing was made by IECO in the soft clays. A Swedish development , , the first step in determining the prob­ 1955 to determine the economiclocation the foil sampler, was used by IECO able cost of a crossing. Borings were for the fill and the approximate quan­ in 1956 to take continuous undisturbed made through the existing fill at the tities of gravel and rock necessary. clay samples up to 50 ft long, of approaches to the trestle to determine Borrow sources of fill were located, the lake bottom in depths of water the depth to which the rock bad dis­ and topographic maps of the area pre­ up to 35 ft. placed the clay. Undisturbed samples pared to aid planning of construction of the consolidated clay were alf'o ob­ facilities. The foil sampler tained, using 2½ and 31/8-in. Shelby In 1955 the underlying salt layer tubes. Other samples were taken from wa.s cored from the trestle at 13 dif­ The principle of the new sampler the trestle in the center of the lake to ferent locations to determine its depth is to insulate the core and protect it provide representative undisturbed clay :rnd e~ient. Vane shear tests of the from the sampler wall by means of for the determination of shearing soft cla~•s ,rnre also made from the thin axial metal foils. The upper ends ;:trength. trestle. Two independent crews of the of the foils are attached to a piston­ This preliminary investigation indi­ Selby Drilling Corp. took the samples. brake mechanism above the core. The cated that the foundation clays were One crew used an apparatus designed lower ends unwind from 16 different «uitable for the construction of a by the U. S. Bureau of Reclamation, roll• which are held in the sampler ·'floating" rock fill. HmYewr, consoli­ consisting mainly of a gear-reduction he:id. _.\s the sampler head advances, dation was expected to result in 10 driYe tmit, N-casing, A-rods and the the foils are pulled through guides and · ft of settlement of the fill after con­ ,·ane. The other crew used an elec­ follow up along the perimeter of the ;;:truction. It was recognized that an tronic recording Yane which eliminated core. The foils not onlv eliminate the extensiYe investigation of the cla~· the necessity for a casing but required friction produced by sliding but retain would be required for design purposes the use of a Sanborne recorder. the physical properties of each soil if a rock fill were to be built without Five permanent platforms were con­ layer encountered. danger of severe overrun of quantities. ;:tructed across the lake on the pro­ Success of the sampling operation From these studies it "l"l"US concluded po;:ed centerline of the fill. It was depends upon the contact between core that the annual cost for reconstruction rle;:irous to work seyeral days at each and foils. Since the foils must enter and redecking of wooden trestles would location and it was found necessary to the sample tube at the same rate as be nearly as much · as that for the build permanent platforms. The~e the sample, tension on the foils mu~t construction of a rock fill. However :'tructures, each 30 ft square, consisted be applied systematically. This is done other factors, such as a fireproof · oi wooden piling, 12-in. X 18-in. string­ by adjusting the level of the piston­ structure, full-speed train operation, ers and 4-in. decking. The deck was 10 brake mechanism. Special measures and a permanent facility, made the ft above the water surface for pro­ may also be required for sampling in fill much preferable. See Fig. 1. tection from wave action. Vane tests, expansive soils, very soft or very firm · In 1954 a seismic sun·ey was made penetrometer tests, and undisturbed cohesive clays, or permeable cohesion- of the proposed crossing by Fisher ;:amples were taken at each location. less soils. · Research Laboratories for IECO. No The penetrometer used was an elec­ The sample tubes may be divided subsurface strata of rock "·ere found tronically recording device which pro- into 8-ft sections and transported to but the deposit of crystalline Glauber's . \;ded a contin11011f' reror

--12 (Vol. o . R--IR) December 1957 . • CIVIL ENGINEERING . 0 PLATFORM 1 PLATFORM 2 PLATFORM 2 PLATFORM 3 0 PLATFORM 3 I 1 . -lo 0 0 0 0 0 0 ! i -+ 0 . 0 I I ,; 0 • 0 I; 5 + ;; • + +0 I • • 0 I • t' +O • + • •+ +0 10 • . ... • • 0 .. ~ -·+· i ,; • • a:. .5 . +· +< 'o +• 0 :, . • + E 15 . + 3: t + 0 . .; • .a • • + + .r: • + . a... ,. ➔ C 20

..,• Legend

+ legend Unconfined compression - tests: 11 + ! + Vane shear (USBR) + By 3 piston sampler • Unconfined compression test o By foil sampler Legend .• i + I o Qu ick triaxial test • Vane shear (USBR) • . I ... 30>----+--+-<>--- o Penetrometer (electronic), where P/A•i,'l''+900c I 0 2 4 0 2 4 6 0 0.1 0.2 0 0.1 0.2 0.3 0 0.1 0.2 Shear strength, psi Shear strength, tons per sq ft Shear strength, tons per sq ft

FIG. 2. Shear strengtha of clay obtained by FIG. 3. Shear strengths of clay obtained by FIG. 4. Two sam­ U. S. Bur. of Reclamation vane shear device vane shear device are compared with the pling methods yield are compared with those by penetrometer. results of laboratory tests. similar test results.

Extrusion is accomplished by pulling shallow holes where continuous samples sampler was found to be the mo~t out the foils (and core) into a trough. are not required. The average rate, successful device used during the in­ A thin slice is then peeled off the core under the same conditions, for a vestigations to evaluate the strength the entire length of the sample - for shallow uncased hole using the H vorslev characteristics of the clays. IECO's purposes of soil classification. piston sampler was 10 ft a day. The policy is to use engineers to supervise The advantages of the foil sampler need for a drill rig to take deep sampling. The foil sampler is regarded are obvious where a complete soil pro­ samples increases the cost of con­ not as a contractor's drill machine, but file is desired. All strata, including void ventional piston sampling. Laboratory as an engineer's tool of the same im­ spaces, are disclosed to the soils tech­ tests of samples obtained by the use portance as the devices used to test nician at their exact depth below the of the foil sampler are compared with and evaluate the sample itself. ground surface. Permeable layers or those obtained with a piston sampler Development of sampling technique hard crusts can also be accurately in Fig.4. progressed from the elementary to the measured. Disturbed areas sometimes The test data indicated very low technical. Special advisers and con­ are found in soft layers immediately bearing values for the clay soils. But, sultants were employed as follows. below a hard layer as a re,,tlt of the by removing some 25 ft of valueless David Piertz, University of California, resistance of the hard layer to the mud by dredge and then placing a aided in the design of the electronic cutting edge. With the new samples, sand-and-gravel blanket 25 to 40 ft devices; and William Holtz, USBR, such disturbed material can be avoided deep over a width of some 400 ft advised on vane shear testing tech­ · ·for testing, and representative samples (Fig. 1), a center rock section could niques. Consultants for testing tech­ selected elsewhere. be put down that would support the niques and requirements were Dr. A. The quality of the samples proved railroad embankment. The fill is being Casagrande, M. ASCE, Harvard Uni­ the feasibility of the sampler, and the constructed from both ends over clays versity; Dr. F. H. Kellogg, M. ASCE, number obtained provided its economic of low bearing value toward a 5-mile University of Mississippi; R. R. justification. No permanent towers center section which is underlain by Philippe, Chief, Soils and Cryology were required for the foil-sampler op­ the firm, deep bed of Glauber's salt. Branch, Corps of Engineers; Stanley I eration as the sampling at each location This will provide time for the fills at Wilson, A.M. ASCE, Shannon & required only a few hour;;' work. A each end to consolidate the underlying Wilson, Seattle, Wash.; and T. Kall­ 11 portable pipe tripod, 35 ft high with clays. Little consolidation is expected stenius, of the Swedish State Geotech­ ,:I two working decks, was built. After under the salt layer, and a wide fill nical Institute, one of the inventors I the sampler tower was attached, one section is not nec~ry to prevent dis­ of the foil sampler. continuous sample, 35 ft long, was ob­ placement in this area. The men responsible for the design tained at each setup. The equipment Foundation investigations at Great and construction of the rock fill were used included a barge with A-frame Salt Lake have been very successful William M. Jaekle, M. ASCE, Chief . and hoist, a tugboat, and the tripod. in providing the engineer with the Engineer, Southern Pacific Co., and C . Undoubtedly use of the piston-type reliable information he needs for a P. Dunn, M. ASCE, President, Inter­ sampler is economically justified for safe and economical design. The foil national Engineering Co., Inc.

CIVIL ENGINEERING • December 1951 (Vol. p. 849) 43 Great Salt Lake Crossing $49,000,000 proiect for Southern Pacific

W. M. JAEKLE, M. ASCE Chief lnelnHr, Souther11 hdflc Company, San Francisco, Calif.

FIG. 1. In new croBBing, three old fills are fully utiliz­ ed. These are Bag­ ley Fill. Saline Fill, and Rambo Fill. placed in 1902. All sand and grav­ el comes from one area on Promon­ tory Peninsula. Rock comes also from westerly side of !alee.

A mountain of material is being eco­ These are R!pplemented by side-dump Promontory Peninsula forced builders nomically moved into Great Salt Lake rail cars working from the westerly to resort to a 12-mile wood-pile trestle to put the Southern Pacific Railroad shore of the lake. to complete the crossing. This crossing crossing on a completely dependable Great Salt Lake, which covers a shortened the line between Lucin, roadbed. The project entails the 2,000-sq mile corner of northern Utah, Nev., and Ogden, Utah, by 43 miles, dredging of 16 million cu yd of muck has long been an obstacle to transpor­ eliminated 1,500 ft of adverse grades, from the lake bottom and the dump­ tation. In frontier days, wagon trains and cut out 4,000 degrees of curva­ ing of 31.5 million cu yd of rock, took a long, tortuous detour to the ture. sand and gravel. When completed in north. When the rails of the first Freight and passenger trains of the 1960, the new embankment will meas­ transcontinental railroad moved closer Southern Pacific have been operating ure 12.68 miles long and will have an to a junction from east and west, its across the lake since 1904. But while overall height of 70 to 75 ft and a builders elected in 1869, to veer north the lake was bridged, it always re­ width of 38 ft at the top and up to to finish the line rather than try any mained a problem. 400 ft at the bottom. short cut across the soft clays com­ Southern Pacific's present Great All forms of modern earth-moving prising the bottom of the lake. Salt Lake crossing project calls for equipment are used. First a 15-in. and The Southern Pacific did tackle the construction of a new rock fill cause­ an 18-in. hydraulic dredge take out job of crossing Great Salt Lake in way across the deeper part of the lake, soft and weak lake-bottom sediments 1902, some years after it had taken spanned now by trestle. The crossing along the line of the new fill. Electric over the pioneer Central Pacific from utilizes all the existing embankment. shoYels of 8-cu yd capacity load mixed San Francisco to Ogden. Under Wil­ The new fill is being built 1,500 ft sand and gravel on 27-cu yd bottom­ liam Hood, M. ASCE, Chief Engineer north of the wooden structure to keep dump trucks. These trucks feed the of the Southern Pacific Co., railroad construction equipment from bump­ world's greatest (in terms of tons per construction ' crews set out to build a ing into the old trestle and possibly hour) belt conveyor to deliver up to rock fill as far as - practicable across interfering with train traffic. The new 4,200 tons per hour to storage. In a the two ma,jor northern arms of the crossing will be nearly indestructible separate operation, additional big lake. The easterly arm, Bear River which will greatly strengthen thiE shovels load trucks that transfer rock Bay, was spanned only after consider­ transcontinental line-the Overland to bottom-dump and flat-deck scows able t rouble. The exceedingly soft bot­ Route-as a defense artery. or haul direct to the core of the fill. tom between the west shore and the Maintenance costs on the wooden jj (Ynl. n . 850) December 1957 • CIVIL ENGINEERING Harbor and material supply center for Great Salt Lake crossing are located on Promontory Point. Belt conveyor at upper right brings 4.200 tons per hour of sand and gravel to storage near waterfront. Barges secure rock from trucks loaded at principal quarry in foreground. Deck barges are loaded at dock at lower left. Ways where barges were assembled are at upper center~ and pier for general service is on far side of basin.

trestle have been mounting steadily cost of the project will be about $49 All equipment is moYed in by rail. under heavier loadings, the wear and million. However, all construction Six of the largest bottom-dump barges stress of the elements, and fifty years equipment, valued at $15 million, be­ ever constructed, each capable of hold­ of use. To reduce impact and assure longs to the Southern Pacific and will ing 2,000 cu yd, were fabricated for • safety, trains ha,·e been restricted to be sold or salvaged for other use. the project by Kaiser Steel Co. These "slow-order" speeds. Fire or possible A construction camp was established were shipped on railroad flatcars in 32 sabotage has been an ever-present two miles north of Promontory Point, sections, weighing 10 to 30 tons each, possibility for the wood structure. The close to supplies of rock, sand, and for assembly at the site by Chicago unexpected vulnerability to fire . of gravel. Included are storehouses, work­ Bridge & Iron Co. Five flat-deck wood exposed to salt spray for years shops, dining hall, barracks, facilities barges were built by the Hammond was demonstrated in May 1956, after for 300 family trailers, a supermarket, Iron Works at Provo, Utah, to carry work on the new fill was well under clothing store, restaurant, post office about 1,000 cu yd each. way. At that time, 645 ft of trestle and school. Six tugboats, each equipped with was burned out in the first major fire The first major job was the dredg­ twin 500-hp engines, were built by in the history of the Salt Lake cutoff. ing of 1,000,000 cu yd of lake shore Gunderson Engineering Co. at Port­ Planning for the new work has been to build a half-mile-wide harbor and land, Ore ., then halved down the cen­ under way since 1950. Actual con­ a channel to deep water 3 miles long terline and moved by rail to Great struction started in June 1955, with and 15 ft deep. Docking and barge­ Salt Lake for reassembly and launch­ the dumping of some fill material loading facilities were installed by ing. Additional 600-hp tugs handled from railroad cars. Work ,vas acceler­ Ben C. Gerwick of San Francisco. A the flat-deck barges while 300-hp tugs ated in March 1956 with the award of 44-kv power line was brought in from supplemented the larger vessels. a $45-million contract to Morrison­ Ogden by Trowbridge & Flynn. Material for the fill is of two kinds. Knudsen Co., Inc., of Boise, Idaho. A pocket flotilla of barges, dredges, Gravel, available at lower cost and Installation of railroad track and tugs, and workboat.5 is bearing the better for contact with the soft clays, "gnal equipment-to be done by brunt of fill construction. It is fed by is used for the major part of the uthern Pacific ·s own forces-will an unusual aggregation of shovels and fill. Quarry-run crushed stone is used st about $2 million. With such other trucks, a major part of the material in the center section of the fill to best penditures as those for exploration, being moved by a most unusual belt distribute the load with a minimum gineering, testing and overhead, the conveyer. of settlement. Selected large rock from

IL ENGINEERING • December 1951 (Voi. p. 851) -45 At left: Stacker builda atodrpile 70 ft high O'Hr two reclaiming tunnela. Stacker in tum is fed by conveyor 7,500 ft long, which delivers 4,500 tona of sand and gravel per hour.

the quarry is used for a b!:lnket onr the gravel fill and for riprap on the slopes. Sand and gravel come from ;1 rang<' of hills some two miles from the lake and about 400 ft above it. Blasting is not necessary. Three 8-cu yd electric shovels load a fleet of 11 bottom-dump trucks for a short haul to a durn ping station over the high-capacity belt conveyor. The sand and gravel, with plus 8-in. material scalped out and crushed, is loaded on the 54-in.-wi

46 (Vol. p. SS!' December 1957 • CIVIL ENGINEERING trench in which the material is to be placed. A wide area is excavated and replaced with gravel fill, requiring some 400 to 500 cu yd per lin ft . I Dredges work some 2,000 ft ahead of .••,· • .:-::-. I the fill operation, about the minimum .. ·--: :.: i.; . ~ . ~ ! distance permissible with equipment that can place as much as 50,000 cu yd per day. The bottom-dump barges are posi­ tioned over the trench and dumped b~- remote control from the bridge of the tug. Each barge has seven hoppers that can be opened all together or indi,;dually. Profile of the underwater fill i.s checked daily, or after each dumping if necessary, by an electric fathometer mounted in a small boat. Th!> bottom-dump barges haYe a draft of 11 ft so cannot be used when the · fill comes within 12 to 15 ft of the Above: Rock is dumped down skids which are adjustable for height to prevent surface. For this work the flat-deck damage to fiat-deck barge. Tractors shove rock off barge to build upper part of fill. barges are used; they are unloaded Below: Work train of air-dump cars unloads rock at westerly end of fill project. Next, by tractors that push the material bulldozers will move in to push rock ahead and out over gravel blanket. Blanket was l_ over the side. This procedure is sup­ placed first, by bottom-dump barges, to within 12 to 15 ft of water surface, in trench plemented by direct truck-haul oper­ provided by dredging. ations at the east end of the project and by train haul with side-dump cars from the western end. The fill is being placed from both ends toward the middle section founded on the rock salt formation. Two years are expected to elapse be­ tween completion of the end and mid­ portions of the fill. There is much work yet to be done before trains can maintain full speed across Great Salt Lake. Unit costs of fill are given in Table I. Technical phases of the operation are being handled by International Engineering Company under the di­ rection of President C. W. Dunn and Chief Engineer T. Mundal, M. ASCE. H. V. Anderson, J. M. ASCE, has directed field sampling, testing and development of engineering data and design. Construction Division Engi­ neer H. J. Willard and Engineer In­ Since 1904, Southern Pacific trains have passed over this 12-mile-long timber spector J. C. Strong, J. M. ASCE, trestle on their way across Great Salt Lake. High salt content of lake water have been active in coordinating the preserves the wood and damages iron fastenings leas than would sea waler. work of the contractor with that of the International Engineering Com­ pany for Southern Pacific.

TABLE I. Unit cost of fill across Great Salt Lake

Operation cost only, not including purchase of equipment, plant installation, and overhead

1Tn1 C OST PEil CU YD Dredsm& ...... 10.5 cen h Gravel fill by bottom-dump bal'&• . 39 cenb Rock fill by bottom-dump barge . . • 75 cents Hock fill by flat-deck bar&e . . . . . 90 cenls Truck haul to fill, rock ...... 90 cents Train haul to fill , rock • . . . . • . . 67 cents

CIVIl. ENGINEERING • December 1957