INTRODUCTION

Design of bridge piers and abutments is an important civil engineering task that will fall at some time to the lot of most practicing civil engineers. History is replete with many examples of substantial bridge works such as the piled foundation of the Roman bridge across the Rhine, and London Bridge, over the Thames. According to a third-century Roman writer, there was a bridge across the Thames just above its mouth as early as A.D. 43. On its arrival in A.D. 1014 to aid King Ethelred of England against the occupying Danes, the fleet of King Olaf (St. Olaf) "rowed quite up under the bridge and then rowed off with all the ships as hard as they could downstream (having secured ropes to the piles supporting the bridge). The piles were then shaken at the bottom and were loosened under the bridge," which gave way, throwing all the defenders ranged upon it into the river.The London Bridge so well-known through illustrations in history books appears to have been completed in the early part of the thirteenth century. The waterway was so reduced by this multiarched structure that swift rapids developed and many persons lost their lives in passing through. The old saying was that "London Bridge was made for wise men to go over and fools to go under. An act of Parliament h 1756 ordered all the buildings on the bridge to be removed and the two central arches rebuilt into one arch. This work inevitably diverted the main flow through the opening and set up serious scouring, which eventually led to demolition of the bridge and replacement with a modem structure. This is but one of the ancient bridges in which piers have caused trouble. Records are scarce, unfortunately, but it can safely be said that scouring out of the foundation beds adjacent to bridge piers has been a major cause of trouble in the past. The piers of ancient bridges rarely failed because of excessive loading on the foundation beds, if only because of the limitation of span length imposed by the structural materials available. The two defects mentioned can be regarded as the two main possibilities of failure to be investigated in the design of bridge piers. Both are essentially geological in character.

Figure 8.1 The ancient London Bridge supported by piled foundation.

Figure 8.2 Model of Caesars bridge across the Rhine (Museo nazionale della civilta romana, Roma)

IMPORTANCE OF BRIDGE FOUNDATIONS

However scientifically a bridge pier may be designed, the whole weight of the bridge itself and of the loads that it supports must ultimately be carried by the underlying foundation bed. Although piers and abutments may be relatively uninteresting to structural engineers, the careful consideration of foundation materials is as challenging as the determinate mathematical calculations relating to the arrangement of steel, reinforced concrete, or timber to be used for the superstructure. Sometimes it is assumed that the cost of foundations, compared with the total cost of a bridge, is relatively small. Actual cost records, however, show that the cost of foundations (piers and abutments) often almost equals the cost of superstructure, even on large bridges.It has not always been fully recognized that concern should always be given to the pier- and abutment-bearing surfaces and whether they can support the structure without fear of any serious movement in the future. Dr. Terzaghi once said that:On account of the fact that there is no glory attached to the foundations, and that the sources of success or failure are hidden deep in the ground, building foundations have always been treated as stepchildren and their acts of revenge for lack of attention can be very embarrassing.Of no group of foundations is this more true than of those for bridges.

SPECIAL PRELIMINARY WORKThe first considerations in bridge location are ge4erally those of convenience and economy. Foundation conditions usually take a subsidiary place, for the prime requirement of a transportation route is that it connects its terminal points by the shortest convenient route consistent with topography. For crossings of deep canyons, considerations of cost usually limit the choice to the site requiring the shortest possible structure. The bridge engineer must therefore often fit design to available foundation conditions. The limitation of site selectivity necessitates acquisition of the most complete geologic information possible.A still more compelling reason for obtaining full geologic information is that once the construction of bridge piers is started, their respective locations cannot be changed except in most unusual circumstances. More than the usual degree or certainty must therefore be attached to the design and anticipated performance of bridge piers and abutments. There is yet a further reason for this special care in preliminary investigations. Bridges, as a rule, are constructed to cross river or othervalleys-topographic depressions that generally exit because of departure from normal geologic structure. Terrain covered by glacial debris may now conceal an older riverbed or other depression well below the existing riverbed. Such conditions are common, and even if known in advance can have serious effects on design.Riverbeds contain many types of deposits, including boulders and if preliminary geologic work is not done carefully, an extensive boulder deposit can easily be mistaken for solid rock. A telling example is that of the Georges River Bridge, Sydney in New South Wales, Australia. Construction of a toll highway bridge to replace existing vehicular ferries was begun in 1923. Three possible sites were explored by an experienced drilling foreman. The borings at the site finally selected showed solid rock at depths below bed level varying between 10.5 and 14.1m (35 and 47 ft.) at regular intervals across a river section about 450 m (1,500ft) wide, the rock at the sides of which was known to dip steeply. On the basis of this information, a through-truss bridge of six main spans supported on cylinder piers was designed, and a lump-sum contract was awarded. During construction, rock was found at only two of the seven main piers. Additional borings taken to depths up to 39 m (130 ft.) failed to disclose any solid rock at all the other pier sites, and what is even more strange, they disclosed no stratum harder than "indurated sand". Construction had to be stopped and designs changed; in consequence, the bridge took five years to build instead of two and cost 27 .6 percent more than the contract price.Discussion of the paper in which this work was reported to the Institution of Civil Engineers naturally emphasized the rigid necessity of having borings most carefully watched by a trained observer. The absence of geological references in both paper and discussion suggests that neglect of geologic features may have been contributory cause of the trouble experienced.

Although this is an unusual and possibly exceptional example, the construction of the Georges River Bridge is a telling reminder of the supreme importance of preliminary geological information in bridge design and of the vital necessity for professional supervision of test boring work.Another reason for devoting unusual care to geologic investigations at bridge sites in all cases of river crossings is the fact that so much of the ground surface involved is hidden below the water. The results of the underwater borings must be correlated with geologic observations secured at the adjacent shores. Where sound rock is encountered, this calls for no unusual attention, provided the exposed surfaces of the rock show no signs of weathering or frequent fracturing, but if any part of the foundation bed consists either wholly or partially of clay, then it is desirable-in most cases imperative-to obtain samples of the clay in as undisturbed samples of clay and other unconsolidated materials, even through great depths of water.The cities of San Francisco and Oakland are separated by the entrance to San Francisco Harbor. Yerba Buena Island stands in the center of the harbor and divides it into the East Bay and the West Bay. For many years, transportation across the harbor was restricted to ferries, but a bridge reached the construction stage in 1993, being officially opened on November 12, 1936. Early in their planning, the engineers decided upon a program of borings and soil testing to enable them (1) to determine the nature of the subsurface materials, (2) to ascertain the most desirable location for the center line of the bridge, (3) to determine the best location for individual piers, (4) to select basis for the design of the piers. Preliminary jet borings provided the basis for contouring the top of the rock surface of the harbor. With the aid of additional wash-pipe borings and diamond core borings into the rock, they prepared a final design for the West Bay crossing. Piers were located and designed; all were founded on solid rock and constructed by means of caissons, the behavior of which could be accurately foretold.The East Bay crossing presented quite distinct problems. Since rock was not found by borings at practicable depths, it became necessary to rely on the overlying unconsolidated material. Cores were obtained and hermetically sealed in the sampling tube, right on the deck of the drill barge; they were soon tested at the University of California. When the containers were opened, perfect cores were generally found, although in some cases a slight swelling was noticed, possibly due to the change in internal pressure in the sample as it came up to the surface. Material was obtained in this way from depths of 82m (273 ft) below water level.This unique example still has many ordinary features of preliminary investigations. Adequate borings, not only along the line of the selected bridge site, but also on either side of it; careful study of core samples; and correlation of this information with the geologic structure of the adjoining dry ground should present a reasonably accurate structural picture of the foundation beds. This information will enable the designing engineer to locate and design accurately the bridge abutments and piers.Finally, the necessity of taking all borings deep enough below the surface of solid material (and especially of unconsolidated material) must be stressed. Loadings from bridge supports are always relatively concentrated and often inclined to the vertical. It is therefore doubly necessary to be sure that no underlying stratum may fail to support the loads transmitted to it, even indirectly, by the strata above.An interesting example of trouble due to this cause is the failure of a highway bridge over the La Salle River at St. Norbert, Manitoba. Thebridge was a single reinforced-concrete arch, with a clear span of 30m ( l00ft.); the spandrels were earth filled. The roadway was about 9m (30ft) above the bottom of the river, and the height of the fill placed in each approach was about 6m (20 ft.). The bridge abutments were founded on piles driven into the stiff blue clay exposed at the site and thought to overlie limestone bedrock, as shown by preliminary auger borings and the record of an adjacent well. Failure occurred by excessive settlement. The north abutment dropped l.2m (4ft), and bearing piles were bent and broken. Subsequent investigations disclosed the existence of a stratum of "slippery white mud" (actually bentonitic clay) about 7.5 m (25 ft.) below the original surface; this material failed to carry the superimposed load.Local soils were formed in an ancient glacial lake and usually overlie lodgment till, under which is limestone carrying subartesian water. The existence of this water complicated the underpinning of the bridge foundation, but the work was successfully completed, and the bridge was restored to use. The bentonitic clay was previously unknown in the vicinity, and illustrates the uncertainty of glacial deposition. The occurrence has a special interest for engineers; although the bearing piles were driven into the lodgment till ("hardpan"), settlement of the abutment occurred as the result of failure of soft material underlying this.

DESIGN OF BRIDGE PIERSGenerally speaking, there are four types of bridge-pier loading, one or more of which may have to be provided for in design: (1) vertical loads possibly of varying intensity, from truss or girder spans or suspension bridge towers; (2) inclined loads, again of varying intensity and possibly varying direction, for arched spans; (3) inclined tensions, from the cables of suspension bridges; and (a) horizontal thrusts due to the pressure of ice or possible debris, the flow of water impinging on the pier, and the wind acting on the bridge superstructure and piers. In earthquake regions allowance must also be made for seismic forces that may act upon the piers. Combinations of these several loads will give rise to certain maximum and minimum unit pressures to be taken on foundation beds. From considerations of these results and of the nature of the strata to be encountered, the type of foundation can be determined.Estimation of foundation load at the site of a bridge pier is generally similar to the same operation for other foundation work. Aside from concern for weak strata below the surface, there are two unusual features that may require reductions of the calculated net load on the base area. The first is the allowance for the natural material excavated and for the displacement by the pier of water; and the second is the reduction for skin friction on the sides of the pier because of the usually large surface area exposed as compared with the base are. These two factors are obviously dependent on the nature of the foundation strata. Estimation of the first is straightforward, but that of the second is generally a matter of experience or of experiment during pier sinking, tempered by the results of careful laboratory soil tests.Weaker strata may even dictate the use of hollow piers to reduce unit loads or of such unusual structures as the open reinforced-concrete framework abutment supports adopted for the Mortimer E. Cooley Bridge across the Manistee River in Michigan. This singularly beautiful bridge, consisting of two 37.5-m (125-ft) deck truss steel cantilever arms supporting a l5-m (50 ft) suspended span and balanced by two 37.5-m (125-ft) anchor arms, has its deck level about 18 m (60 ft.) above the level of the ground on either side of the river. The foundation of varying strata of consolidated materials was accurately explored and foundation loads were kept to a minimum through use of the open framework design.When preliminary investigations indicate foundation material of poor bearing capacity, consideration may be given to the use of artificial methods of consolidating such material to improve its bearing capacity. This is no new expedient. The account given by Leland (antiquary to King Henry VII) in 1538 concerning the Wade Bridge in England revealed that the foundation of certain of tharches was first sette on so quick sandy ground that Lovebone (Vicar of Wadebridge) almost despaired to perform the bridge ontylsuchtyme as he layed pakkes of wollefor foundation". Although this use of wool has been disputed, the record demonstrates that some artificial means was used to improve bearing capacity. Modern methods include grouting chemical consolidation, or leaving t[e steel piling of the pier cofferdam in place to confine the foundation-bed material and thus prevent lateral displacement. In this way bearing capacity will be increased to some extent.The foundations for the Tappan Zee Bridge that carries the New York Thruway across the Hudson River for a distance of 4.5 km (2.8 mi) between Nyack and Tarrytown, New York, provide an even more unusual approach to the problem of minimizing loads on weak strata. The bridge site selected through careful studies had to be accepted even though the bedrock drops off under the bridge to depths as great as 420 m (1,400 ft.) below water level-a depth too great to be reached by end-bearing piles. The approach spans are therefore carried on friction-bearing piles, driven into the silt, sand, and gravel that form the riverbed.The main piers carrying the 363.3 m (1,212-ft) cantilever main-channel span, however, are founded on buoyant, reinforced-concrete boxes, carrying about two-thirds of the dead load of the superstructure. The remaining part of the dead load, and the live load, are taken by 75-cm (30-in) concrete-filled pipe piles for the four main piers and by 35-cm (14-in) steel H piles for the four buoyant boxes; in each case piles and boxes are ingeniously connected together. The steel piles had to be driven to depths up to 52.5 m (175 ft.), but the concrete pipe piles went as deep as102.0 m (340 ft.) below water level, being driven through clay and then gravelly clay after the sand and gravel had been penetrated. The hollow piles were mucked out to full depth by water jet and airlift techniques and then grouted into preplaced aggregate. The grouting consolidated the sheared gneiss and decomposed sandstone bedrock.

Figure 8.3 The Tappan Zee Bridge over the Hudson River, New York, looking east, showing the spans which are supported by the special piers, described in the text.Details of the piers are admittedly an engineering matter, but it was geology of the site that dictated such bold design.Geologic information can be applied to predict settlement of loaded piers. What happens when uneven settlement does take place is well illustrated by the failure of piers 4 and 8 of Waterloo Bridge, London; the whole bridge had to be taken down and a new structure erected. Described by Canova as "the noblest bridge in the world worth a visit from the remotest corner of the earth", Waterloo Bridge was constructed from l8ll to 1817. Timber rafts on timber piles bearing on gravel were designed to protect the pier foundations against scour. Progressive settlement became serious in 1923; the total settlement of pier 4 exceeded 75 cm (2ft) and naturally caused an arching action between piers 3 and 5.

Figure 8.4 The old Waterloo BridgeSettlement may occur from one or more of the following causes: (1) displacement by scour, (2) lateral displacement due lack of restraint, (3) consolidation of the underlying material, or (4) failure of an underlying stratum. Only condition 3 can be controlled; the other three types are of a nature that may cause serious trouble on the structure. All types can be predicted on the basis of adequate preliminary geologic information.Provision against unequal settlement of piers has assumed considerable importance in recent years owing to the development of the rigid-frame type of structure, requiring "unyielding" abutments and uniform settlement of piers. Rigid-frame structures founded on clay require isolation of the bridge foundation from the bearing piles and load transmittal through a tamped layer of crushed rock (employed at a Canadian National Railways bridge at Vaudreuil in Quebec). Uneven foundation-bed loading, especially that caused by irregular construction scheduling, must also be carefully considered in design. During the 1932 construction of the Broadway Bridge, Saskatoon, Saskatchewan, concreting of the six arches proceeded in varying stages. As a result, the piers tilted when carrying the dead load of only one adjacent arch rib. Amaximum deflection of l5mm (0.6 in) was recorded as anticipated.Inclined tensions of the third type of loading are generally transmitted to anchorages in solid rock. This design approach provides for shearing resistance in the rock, which, together with allowance for the dead weight of the anchorage, will be sufficient to balance the tensile forces in the bridge cables. Some inclined tension in bridge cables is taken up wholly by concrete piers, as at the Ile d'Orleans (suspension) Bridge, in Quebec. The suspension span is carried into the long approach structures and secured in anchor piers, one of which is founded on rock, but the other on sand. The stability calculations for these piers had to keep the unit toe and heel pressures within the limit for the foundation-bed material. Frictional values of concrete-to-rock and concrete-to-sand provided the basis for this anchorage. Inclined H-beam piles wore driven into the sand underlying one of the anchor piers and were left to project outwardly in counterforte, into the concrete of the finished pier to give the necessary increase in stability.As a final example, Burford Bridge across the river Mole in Surrey, England, was designed to accommodate an unusual geological condition. The bridge is a single reinforced-concrete arch span of 24 m (80ft.), 30m (100 ft.) wide between parapets, with specially selected brick facing. The Mole Valley, located some 40 km (25 mi) to the south of London, flows through chalk formation, which is highly susceptible to groundwater dissolution. Underground cavities here are so large as to receive the whole normal flow of the stream. Borings were put down to see if any such "swallow holes" in the chalk were revealed. Two soft spots were located which proved to be dissolution channels having most vertical sides and filled with alluvial matter. Concrete domes were constructedover each of the holes, domes founded on circular ledges cut in the chalk around the tops of the excavated channels; the largest dome was 17.4 m (58 ft.) in diameter with a rise of 2.4 m (8 ft.). The holes were filled up to the undersides of the domes, and the filling was then covered with waterproof paper and used as the lower form for concreting the domes. Each dome was furnished with an access shaft con1ecting to a manhole at road level by means of which engineers may inspect the swallow holes from time to time to see that no dangerous undercutting or further erosion of the chalk is taking place.