America's Highways 1776–1976: A History of the Federal-Aid Program/Part 2/Chapter 7
Pre-1776 Bridges
History teaches that all sciences and arts, including bridge design and construction, benefit from the innovations, successes and failures of the past. This presupposes a tradition of learning not always available to the American colonists. The first bridges in this country were often built by untrained persons inexperienced with the difficulties and problems of constructing a bridge of adequate strength and reasonable durability. Instead, they substituted determination, native ingenuity, and more determination.
Economics and material availability generally limited the colonial bridges to timber or stone construction with timber being predominant. This was largely due to the greater time and labor required to quarry and transport stone. From this lowly beginning, the complex design, construction and material sciences have developed.
The existing bridges in the American colonies in 1776 were few and minor in size. At that time, all major cities and most towns and villages were located on navigable waterways, since the waterways offered the most practical transportation between populated areas. Barges, boats, canoes and fords were mainly used where roads crossed the waterways. As the population centers developed, some timber bridges were built over adjacent narrow waterways by local authorities to facilitate access by travelers and commercial goods.
Generally, the bridges had log beam spans and were limited to the length of timber available from local trees. Abutments at the stream banks were timber mud sills, wooden cribs, or dry stone masonry where stone was available. Where more than one span was required, timber pile bents, wooden cribs or wooden mud sills were used for piers in the stream bed. The cribs were usually braced and filled with rock or compacted earth. The life of these bridges was usually short due to the rapid deterioration of the timber and the washing out of the foundation structures. Some floating bridges were constructed, usually of large logs fastened together.
It is interesting to note that the Concord Bridge, at which the homespun New England “Horatius” fired “the shot heard round the world,” was a timber beam and pile bent structure very similar to the Sublician Bridge defended by his Roman prototype nearly 2,000 years before.
Since the waterways were the main and pre-existing arteries of travel and commerce, any bridges built over navigable waterways required openings and clearances adequate for the passage of the waterway traffic, whether sailing ships, barges, canoes or log rafts. Where the horizontal and vertical clearances of fixed spans were not adequate for such passage, movable spans were used. These were most likely single or double leaf bascules (hinged sections) operated manually. Floating bridges were provided with sections that could be hinged or pulled out of place to permit the passage of waterway traffic.[1]
Some short span stone arches were built, mainly in the colonies north of the Potomac River.
Reproduction of historic timber beam and pile bent bridge at the original site in Concord, Mass.
Timber Bridges 1776–1916
King-post or queen-post trusses (actually simple braced beams) were in early use for short spans. However, the very common pier washouts experienced by multispan beam bridges, plus the increasing knowledge of the society in general, led to the development of timber trussed arches that could span wide rivers. A New England millwright named Timothy Palmer built a series of patented trussed arches very like one of those illustrated in Palladio’s Treatise on Architecture (1570). The best known was the 244-foot span across the Piscataqua River near Portsmouth, New Hampshire, built in 1794.[2] The roadways were supported on the lower chords of the structure, resulting in steep grades ascending to the center from each end of the spans. These bridges being true arches, provision was made for their horizontal thrust to be transferred to the substructure.
The Upper Ferry Bridge over the Schuylkill River, Philadelphia, Pa., a covered timber trussed arch created by Lewis Wernwag in 1805.
Besides Timothy Palmer, two other men, Louis Wernwag and Theodore Burr, stand out as the first professional bridge builders in the United States. These men shared several things in common—they were, of course, contemporaries, their structures were all highly indeterminate combinations of trusses and arches, and they most likely did not have the theoretical knowledge to analyze such structures. They were virtually carpenters building bridges in accordance with their own experience or that related by others, and they were among the group that produced the distinctly American bridge characteristic, the covered bridge. Their picturesque practice was, of course, a sound and practical measure that extended the life of the wooden structure immeasurably.
Camp Nelson covered arch-truss over the Kentucky River during razing in 1933. The three trussed 240-foot span bridge was built by Wernwag in 1838 and closed to traffic in 1926.
The first of many covered bridges in America was built in 1800 by Palmer at Middle Ferry, Philadelphia, Pennsylvania.[3] The covering protected the bridge members from decay to the extent that a properly maintained bridge would give many years of service. The Waterford Bridge over the Hudson River, a covered wood truss bridge built by Theodore Burr in 1803–4, had a service life of 105 years (it was destroyed by fire in 1909).[4]
The steep roadway grades of the trussed arches were objectionable, and by the beginning of the 19th century, a combination truss and arch was developed with the roadway supported by the lower chords and having only a slight longitudinal camber. The height or rise of the timber arch was about equal to the depth of the truss, with the arch fastened to the truss web members at all intersections. The Susquehanna River bridge at McCall’s Ferry, Pennsylvania, built in 1814—15 by Theodore Burr, contained a truss-arch with a 367-foot span said to be the longest then built in America.[5] Burr’s bridge represented a small, but significant, change—the arch was added to the truss.
Inside view of typical Indiana covered bridge, built in 1900, showing arch-truss details.
Covered timber truss bridge over the Connecticut River between Cornish, N.H., and Windsor, Vt. Built in 1866, this 460-foot long, two-span Town truss structure is said to be the longest timber bridge remaining in this country.
In 1820, Ithiel Town patented a timber lattice truss, although lattice trusses had been built in Vermont as early as 1813,[6] and in 1829 Colonel Stephen H. Long developed a panel truss with double diagonals, similar to a Howe truss.[7] These trusses could stand by themselves, and arches were not necessary. Thus, the truss bridge appeared in a recognizable modern form.
The web members, diagonals and verticals of the earlier trussed arches and truss-arches were made of timber. Some of the later web systems used iron rods for tension verticals, diagonals and counters in the web system. There were many other variations of the truss besides those mentioned. Due to intensive promotion, the various types were identified by the name of the developer. When iron and steel trusses were developed, the timber truss designations were given to the corresponding geometric metal trusses, i.e., Howe, Pratt, Fink, etc. Connections of members were made by bearings of timber, forgings, iron bolts and spikes, mortises and tenons, and hardwood dowels.
The layout and relative position of the trusses and arches, the proportioning of the members, the methods of support of the floor system, and type and magnitude of the connections were determined by the experience and preference of the builder. There was no published method of stress analysis available to the builders for proportioning members and connections until Squire Whipple’s publication in 1847, An Essay on Bridge Building, and Herman Haupt’s General Theory of Bridge Construction in 1851,[8] although it appears that Colonel Long may have used mathematical theory in the design of his bridges.[9]
While the timber bridge had evolved into an almost determinate, fully utilitarian structure, it was not free of problems. Covering the truss with a roof and siding had retarded deterioration, but there were still many failures of covered timber trusses due to lack of maintenance, fires, floods, overloads, and inadequate design. Nonetheless, the covered bridge era might have lasted until the coming of the automobile were it not for a combination of events that led bridge designing into the modern era. The newly developed truss and the fledgling analytic methods found themselves a new client, the railroad. Trains being unable to ford even small streams or tolerate the sinuous alinement and steep grades used by pedestrians and wagon traffic, a great many bridges were needed, and not only more bridges, but stronger and more durable ones. This accelerated the growth of design technology. As far as durability was concerned, early attempts at increasing the life of timber in contact with earth and water achieved little success. Methods used included dipping, soaking or brushing the timber with creosote or salts, such as zinc chloride or mercuric chloride. These methods were ineffective due to the lack of penetration and leaching out of water soluble preservatives.
While the railroads were especially desirous of finding a satisfactory method of prolonging the life of the great number of ties and timber structures on their systems, it was not until 1865 that the first pressure creosote treatment plant was constructed at Somerset, Massachusetts, primarily to treat timber track ties.[10] Other plants were constructed soon thereafter for treatment of ties, piles and timber.
With the advent of timber treatment plants to supply the railroads, treated timber piles and lumber became available for use on highway structures, but by then the use of metal in trusses was becoming more common.
The general use of timber trusses for highway bridges continued to the 1880’s, but gradually decreased until by 1916, there were practically none. However, such construction has continued in isolated areas where timber was readily available and in scenic areas as tourist attractions.
Cast-Iron, Wrought-Iron and Composite Wood Bridges and Metal Bridges
The use of iron for incidental connecting parts was introduced early in the construction of timber bridges since it facilitated construction, improved the rigidity of the structure and reduced maintenance problems. Wernwag used iron rods for the light web diagonal of his Upper Ferry Bridge, popularly named the “Colossus,” in 1806.[11]
Soon after the railroad era started in 1827, bridge engineers realized that the timber truss-arch bridges were not serving satisfactorily under the speed and weight of railroad traffic and sought various methods to introduce iron members into the truss arch. Such structures were called combination bridges.
One of the early combination bridges was a through timber truss-iron arch bridge, a 133-foot span, carrying the Pennsylvania Central Railroad over a canal. The bridge was so constructed that the timber truss would receive the load from the ties, transfer it to the arch, and provide lateral support for the arch ribs which were the main load carrying members. However, should the arch fail, the truss was adequate for the full load. The engineer, possibly at the owner’s request, tested the completed structure by passing a 23-ton locomotive over it several times.[12]
In 1840, William Howe patented the “Howe” truss. This was a true truss with timber chords and compression diagonals but with iron bars for the vertical hangers. Iron members had been used before, but they had always played a minor role.
His patent also permitted prefabrication of parts, permitting manufacture away from the site. The ability to tighten up the hanger, via an adjustable nut should it become loose, was an added benefit. The parallel chord and X diagonal pattern soon became a familiar part of the countryside since the Howe truss was the most common of all timber trusses. However, this truss represented high tide for the timber truss and for the carpenter bridge builders, of whom Howe was the last well-known one. What had already happened in Europe was happening in America. Stress and strength were about to join stability as the criteria in bridge building.
The stone arch had been the main bridge type for nearly 2,000 years. A masonry arch, if stable enough to stand, was adequate for any conceivable load. For this reason, stress analysis had been unnecessary. With the advent of timber (and metal) beams and trusses, it could no longer be assumed that the bridge’s existence was proof of its strength. The possibility of overload and failure due to overload now existed, especially with the heavier loads of the locomotive.
The timber trusses and combination timber-metal trusses still suffered from the difficulty in joining timber—the tensile strength of the joint was always less than the strength of the timber member. This made these trusses especially susceptible to falling apart at the joints. On March 4, 1840, a Town lattice truss over Catskill Creek, New York, came apart dropping a train of boxcars into the water, resulting in the Nation’s first railroad bridge fatality.[13] The need for a metal bridge had arrived.
Other methods were used in replacing wood with iron in the arches of truss-arch railway bridges. While limited construction of short-span combination trusses with timber compression and wrought iron or steel tension members continued until the end of the 19th century, it steadily decreased.
The engineer’s first choice for a bridge metal was cast iron, and 1836 saw the first cast-iron bridge in this country, an 80-foot arch span, built over Dunlap Creek in Brownsville, Pennsylvania, by the U.S. Corps of Engineers.[14] Shortly afterward, in 1840, the first iron trusses came into existence when two highway bridges were built over the Erie Canal by Earl Trumbull and Squire Whipple. Trumbull’s truss spanned 80 feet and had a wooden floor system. It featured a parabolic bottom chord of wrought iron bars. Squire Whipple’s 72-foot bridge was the first example of the famous Whipple bowstring truss, so called because of its curved upper and horizontal bottom chord. The tension members were wrought iron and the compression members were cast iron.[15] While most of Whipple’s bowstring trusses have vanished, there are still some examples of his other design, the trapezoidal truss, in service.
In 1844, 4 years after the first two iron bridges, Thomas and Caleb Pratt patented the Pratt truss. This parallel chord truss with tension diagonals and compression verticals in the web system was well suited for metal trusses and, together with its many modifications, became the most popular type of truss for short and intermediate span trusses to the present day. Among its better known variations are the Baltimore, Parker, Pegram, Pennsylvania and Petit trusses.[16] 
First cast-iron bridge in the United States, over Dunlap Creek at Brownsville, Pa.
The second cast-iron arch in this country, completed in 1860, was the Meigs Bridge over Rock Creek in Washington, D.C.[N 1] Like the Dunlap Creek Bridge, it was also built by the Corps of Engineers. Its two cast-iron pipe arches supported the deck of the highway bridge and carried the water supply of the city over the creek. The 51-inch outside diameter pipes spanned 200 feet. The bridge deck and supports were removed in 1916 and a new bridge was constructed over and independent of the pipes still carrying water.[17]
Ironically, the chief factor in the decline of the cast-iron bridge was its success. The increased use of wrought iron and cast iron for bridges, rails and other related uses caused a boom in the iron industry and created an incentive to develop new processes for producing iron and steel.
- ↑ Not to be confused with the Meigs Stone Arch over Cabin John Creek in Maryland which was also built under the supervision of Capt. Montgomery C. Meigs when he was with the U.S. Army Corps of Engineers.
The original Meigs Bridge over Rock Creek in Washington, D.C., built in 1860 by the U.S. Corps of Engineers.
Although the Bessemer converter and open-hearth furnace made the steel bridge inevitable, several bridge failures also increased distrust of iron bridges. The most notable was a failure of an iron Howe truss at Ashtabula, Ohio, which took 65 lives in 1876, the worst rail disaster in America so far. Surprisingly, many highway bridges also failed, despite their relatively light live loads. This was due to the pressures for economy put upon county officials who lacked technical expertise and fly-by-night bridge salesmen and promoters, who, sometimes involved in political and business corruption, provided cheap and inadequately designed structures.
From the 1850’s, companies were formed primarily for constructing patented truss bridges, either under their own patents or as licensees. There being very few bridge engineers with a working knowledge of stress analysis and truss design in the early years of iron superstructures, these “bridge companies” became firmly established. Bids let for short- and medium-length span truss bridges allowed the bidder to furnish the superstructure to his own plans. While many safe and satisfactory bridges were obtained by this procedure, there were also many unsatisfactory bridges. In some cases, the companies furnished weak or inadequate bridges to compensate for underbidding or to make excessive profits.
The Eads Bridge over the Mississippi River was opened to traffic in 1874. Alloy steel was used for the first time in this country for the members of the three arches of this double deck bridge.
Due perhaps to a combination of poor engineering, poor material, and fraud, the numerous failures of iron and combination bridges brought about the complete abandonment of cast iron and led to the acceptance of wrought iron and steel, the flowering of modern truss design, and the rise of the civil engineer.
The first American all steel spans built were the five main spans of the Missouri River Bridge at Glasgow, Missouri. Bessemer steel was used for the 315-foot long trusses. The remaining trusses and trestle work were made of wrought iron.
However, the Glasgow trusses lagged by 10 years the first extensive use of structural steel in a bridge. It was in 1868 that construction commenced on the first of the two “great bridges” of American history,[N 1] the Eads Bridge across the Mississippi at St. Louis, Missouri.
Until 1855, the “Father of Waters” had never been spanned. Thirteen years later, James Eads, a hero of the Civil War, began building three alloy-steel arches of 502-, 520-, and 502-foot spans, respectively, across the river. This bridge is noteworthy for another reason: The first use of pneumatic caissons in the United States and, sadly, the first death from the “bends” or “caisson disease.” (In all, 13 died.) It is interesting that Captain Eads had never built a bridge before.[18] Despite the magnificence of the Eads Bridge, which is still in use, the future of the steel bridge was with the plate girder and the truss.
- ↑ The second of the “great bridges” in America—the Brooklyn Bridge—is discussed under Suspension Bridges.
After Squire Whipple’s essays on bridge building, the increasing knowledgability of the engineer led to further refinements in the metal truss as iron yielded to steel. Among the changes were the use of bridge rollers and the perfecting of the pinned connection.
The improved metallurgy permitted the evolution of machined steel pins from the old iron trunnions and the use of the now famous eyebar tension member. The eyebar became the trademark of the American truss bridge in the latter half of the 19th century. In this area, the American engineers deviated from European practice, which was turning to the more rigid riveted gusset plate connections. The economy and determinacy of the pinned truss was well suited to the American bridge engineer’s interest in structural analysis. While the riveted joint was finally accepted, the practice of using deeper trusses and longer panels was retained this side of the Atlantic.
Interest in determinate structures led to the use of cantilever bridges. Eventually, cantilevers were so widespread that Europeans referred to them as the “American” bridge.
In 1876, Charles Smith built the first such bridge over the Kentucky River for the Cincinnati Southern Railway with three 375-foot spans. Other long span cantilever bridges were the Monongahela River Bridge at Pittsburgh, Pennsylvania, built in 1904 with a maximum span length of 812 feet, and the Queensborough cantilever designed by Gustav Lindenthal in 1909 in New York City. Its maximum span was 1,182 feet.[19]
The Washington Street Bridge over the Harlem River in New York City.
The use of plate girder bridges for short spans began during the Civil War. While these spans were limited to moderate length by the size and length of the wrought-iron plates and shapes then being rolled, longer spans were built as the available size of wrought-iron and steel plates increased. The first American railroad girder bridge with a 100-foot span was built in 1887. Just 7 years later, a 182-foot span was constructed, a major span even by today’s standards. The use of wrought-iron and steel I-beams for short spans was initiated by the rolling of deeper beams.
Details of a 200-foot Finley chain link suspension bridge.
Still, early plate girder highway bridges were limited to short spans because of the inconvenience and cost of transporting and erecting long members except when the bridge site was near railroad transportation.[20] Thus, long spans still belonged to the truss, the suspension span, and the arch.
The first American plate girder arch bridge was the Washington Street Bridge over the Harlem River in New York City, built in 1886–88. It had two 510-foot arch spans with plate girders 13 feet deep.[21]
By the early 1870’s, the size of I-beams rolled was large enough to use for stringers in the floor systems of short paneled truss bridges. This, together with the development of plate girders for floor beams, led to the general discontinuance of the use of wooden stringers by 1875. However, wooden stringers were used on short- and medium-length spans until 1890 and later.[22] During the period 1874 to 1890, the maximum depth of American rolled I-beams increased from 10½ inches to 20 inches.
Suspension Bridges
This most fascinating of all bridges first appeared in America in 1801 when Judge James Finley built a 70-foot chain link suspension bridge over Jacobs Creek near Greensburg, Pennsylvania. The chains were made from 1 inch square wrought-iron bars. The links varied from 5 to 10 feet in length so as to match the distance between floor beams. The timber floor system was stiff enough to distribute live loads to several hangers and to resist deformation, undulations and vibrations from the live load and wind loads. The bridge failed under a six-horse team in 1825 but was repaired. A number of bridges were built under the direction of Judge Finley or his licensees. The maximum span lengths were probably about 150 feet. The cable supports were usually timber towers on stone masonry piers. A suspension bridge at Lehigh Gap, Pennsylvania, apparently the last Finley type in use, was replaced by a modern structure in 1933.[23]
Suspension bridge built by Charles Ellet in 1849 over the Ohio River at Wheeling, W. Va.
One of the largest of the chain link structures was the Point Bridge over the Monongahela River at Pittsburgh, Pennsylvania. Built in 1875–77, it was a stiffened chain suspension span 800 feet long. The main and backstay chains were of wrought-iron link bars 20 feet 6 inches long.[24]
The first highway bridge in America with wire suspension cables was designed by Charles Ellet and completed in 1842. It replaced Wernwag’s wooden trussed-arch Upper Ferry Bridge over the Schuylkill River at Philadelphia.[25] As would be the case for all American suspension bridges until the Brooklyn Bridge, drawn wrought-iron wires were used for the cables.
Roebling opened this suspension bridge over the Ohio River between Cincinnati, Ohio, and Covington, Ky., in 1866.
The Brooklyn Bridge over the East River in New York City.
Subsequent to 1840, several wrought-iron wire bridges were built in America by Charles Ellet, John A. Roebling, Thomas M. Griffith, Edward W. Serrell and others. However, the suspension bridge was still in its infancy. It began to come of age in 1849 when Ellet spanned the Ohio at Wheeling, Virginia, with a 1,010-foot suspension span. When completed, this was the longest bridge span in the world. While it was damaged by wind in 1854, the bridge was repaired and is still in service.[26]
John Roebling had built nine suspension bridges by 1855, one of which was the 821-foot span combination highway-railroad bridge over the Niagara Falls rapids, a bridge most engineers thought doomed to failure. When Mr. Ellet’s temporary suspension bridge was built at the site in 1848, the rope for pulling the first cable across the river was pulled across by a cord flown over the gorge by a boy with his kite.[27] In December 1866, Roebling opened his recordbreaking Ohio River suspension span at Cincinnati. This structure, 1,057 feet between towers, is also still in use.[28] During the flood of 1937, this was the only highway bridge open across the Ohio between the Mississippi River and Sciotoville, Ohio.
There were at least two factors that Roebling had considered in his bridges: the quality and protection of the cable and the bracing of the structure against aerodynamic loadings. Attention to these problems, plus the other design aspects, made his bridges successes.
These major bridges, although unprecedented achievements, were only the prototype for his crowning task, the second “great bridge” of America: the Brooklyn Bridge over the East River. This bridge, its 1,595-foot main span 50 percent longer than the previous record span at Cincinnati, linked Brooklyn and Manhattan and made New York’s expansion possible. It was also the first bridge to use galvanized steel cable.
Building the East River bridge was probably the most dramatic verse in the saga of bridges. Its creator, John Roebling, died of tetanus as a result of an accident during the early stages of construction. His son, Colonel Washington Roebling, who had survived the battle of Gettysburg, was crippled by the “bends” and had to direct operations through his wife, Emily. But in 1883, 14 years after construction started, man had conquered the East River and signaled the beginning of the great age of bridge building.[29]
Two other notable suspension bridges were built over the East River at New York City around the turn of the century. The Williamsburg Bridge, built in 1903, has a 1,600-foot main span. The main span of the Manhattan Bridge, built in 1909, while 130 feet shorter, yet is considered by many as the most graceful cable arc of any of New York’s bridges. The rising tempo of bridge building may be noted from the fact that the Williamsburg Bridge, although 5 feet longer than the Brooklyn Bridge, was just a long bridge. The long-span title had gone across the seas to the Firth of Forth’s 1,700-foot cantilever span in 1890.
Movable Bridges
Bridges with movable spans to accommodate water traffic also date back to colonial times. The early timber bridges were opened and closed by the only available power—manpower. As a rule, they were either bascule (draw) spans or swing spans which rotated to open the channel to marine traffic. This, of course, meant that only short span openings could be used. Fortunately, the pace of life, as well as of river traffic, was sufficiently slow so that lengthy bridge opening times could be tolerated.
Some early spans were a wooden drawbridge between Boston and Charlestown over the Charles River, built in 1785–86,[30] and the Haverhill Bridge over the Merrimack River at Haverhill, Massachusetts, built in 1794 with a wooden bascule drawspan. The 30-foot drawspan was raised by means of levers elevated on a post on each side of the draw.[31] The Tiverton Bridge over a tidal inlet near Howland Ferry, between Portsmouth and Tiverton, Rhode Island, built in 1795, contained a sliding drawspan.[32]
The advent of the railroads and steam-powered boats made the span weights heavier and the opening time of the movable spans more critical. Fortunately, as wrought iron, and then steel, became the material for the bridge members, steam became the motive power for opening railroad and main highway movable bridges over busy waterways. For minor waterway crossings and during power failures, movable spans were still opened manually.
As the larger rivers and tidal estuaries became more frequently bridged and river craft became wider and longer, wider channel openings were required, necessitating longer spans for movable bridges. Of the movable bridges constructed during the last half of the 19th century, two were notable. The first was a railroad bridge located between Rock Island, Illinois, and Davenport, Iowa, providing a swing span with two 120-foot channel openings. Built in 1854–56, it was the second bridge to cross the Mississippi River.[33] The other was the Utica Lift Drawbridge, a 60-foot x 18-foot movable deck, suspended by rods from the lower chord panel points of an overhead fixed truss. The vertical lift was 11.5 feet. Opening time, using preset weights, took 10 seconds. This was one of several unusual vertical lift bridges which Squire Whipple designed and built over the Erie Canal in New York in the 1870’s.[34]
This sketch of a manpower-operated bascule on the Pongo River Bridge in Virginia probably exaggerates the difficulties of such operations. From the Harpers New Monthly May 1858.
The development of modern movable bridge spans started in Chicago when the channels of the Chicago River and related waterways were improved at the turn of the century. Many of the existing movable bridges were manually powered swing spans, with center piers obstructing channels. These were replaced with bascule and vertical lift spans. This era was begun by two famous bridges. The Halsted Street Lift Bridge over the Chicago River in Chicago, Illinois, 1894, lifted a 130-foot truss span with a 34-foot roadway and two 7-foot sidewalks using steampower. Maximum vertical clearance was 155 feet above low water. Two light longitudinal, laterally braced, trusses connected the tops of the towers. The design of this bridge introduced another famous name in bridge engineering, J. A. L. Waddell.[35] The other bridge, the Van Buren Street Bridge also over the Chicago River, was opened in 1895. This 115-foot span, double leaf of the Scherzer type, was the first rolling lift bridge.[36]
From then on there was a series of new movable bridge types: the simple trunnion, or Chicago,[N 1] and Strauss bascules, the Rail rolling lift,[N 2] and various vertical lifts. The swing span, despite its economy and minimal power requirements, fell into relative disuse since it required a pier in the middle of the stream and blocked the channel more during opening and closing operations.
Vertical lift bridge over the Cape Fear River at Wilmington, N.C. The 400-foot lift span has a vertical clearance above mean high water of 65 feet in the closed position and 125 feet in the open position.
Lake Washington Pontoon Bridge, Seattle, Wash., is a concrete pontoon bridge completed in 1940. The floating portion consists of 25 units 350 feet long, together with retracting sections to provide a 200-foot opening for large vessels. Passage for small boats is provided near each shore under truss spans. The pontoons are fixed laterally by long steel cables fastened to submerged concrete anchors.
The George P. Coleman Memorial Bridge over the York River near Yorktown, Va. The double 500-foot swing spans are the largest tandem swing spans in the world.
Cabin John Bridge over Cabin John Creek in Montgomery County, Md. The aqueduct for the Washington, D.C., water supply is in the spandrel fill under the roadway.
The great size of today’s structures, which evolved from the short span crossings of the Charles River and Merrimack River, is due to the growth in technical knowledge, experience, better steels, steam and electric power, and a near perfect use of the old-fashioned counterweight.
Stone Arches
Stone arches have not been a major element in highway bridge construction in this country. However, numerous short-span stone arches have been built in areas where stone and skilled masons were readily available. These bridges are frequently seen alongside present highways where they have been left for their historic and scenic value when highway alinement, width, or grade was improved. Long- and short-span stone bridges have been built in parks and large cities to be compatible with their surroundings.
One of the most notable stone arches was built during the Civil War as a combination aqueduct and road. The 220-foot filled spandrel Meigs Arch, which carries Washington, D.C.’s water supply and MacArthur Boulevard over Cabin r Creek in Cabin John, Maryland, was the longest span of its type when built. It is still in service despite its narrow roadway.[37]
Where stone was readily available, stone arches of nominal spans and stone arch culverts were frequently used in railroad construction before metal superstructures were developed and accepted. Some of the noteworthy stone railroad bridges still in existence are:
- Carrollton Viaduct over Gynn’s Falls on the Baltimore and Ohio Railroad, near Baltimore, Maryland, was built in 1829 of granite ashlar masonry. The center span is 80 feet and the overall length is 297 feet.[38]
- Thomas Viaduct over Patapsco Creek on the Baltimore and Ohio Railroad, near Relay Station, Maryland, was built in 1835 of granite ashlar masonry and is 612 feet long and has eight full-centered arch spans.[39]
- Starrucca Viaduct over Starrucca Creek (a tributary of the Susquehanna River) on the New York and Erie Railroad, was built in 1847–48 with stone masonry made of stone quarried locally. The viaduct is 1,040 feet long with a maximum height of 110 feet above the creek.[40]
- The Morgan Bulkeley Bridge was constructed across the Connecticut River between Hartford and East Hartford with funds raised by subscription from all the towns around the area in 1905. This beautiful structure of pink granite has 11 elliptic arches, the largest being 119 feet long. In 1962, it was widened to carry Interstate 84.
The Morgan Bulheley Bridge over the Connecticut River at Hartford, Conn.
Reinforced Concrete Bridges
Although concrete bridges are part of the modern era, the early use of concrete was limited. Natural cement entered into bridge construction as mortar for stone masonry and for unreinforced concrete footings and substructures as early as 1850. Although many cement users switched to the more uniform and superior portland cement after its first manufacture in this country at Allentown, Pennsylvania, in 1871, natural cement has continued to be used to a lesser extent even to this day.
The Clefridge pedestrian underpass was built of concrete in Prospect Park, Brooklyn, New York, in 1871. This structure, said to be the earliest concrete arch constructed in America, is an arch that follows the design of stone arches and has a radius, of 10 feet.[41]
The first reinforced concrete building in this country was built by W. E. Ward in New York State in 1875, and the first reinforced concrete bridge in this country was a 20-foot span built in Golden Gate Park, San Francisco, California, in 1889.[42][43]
Other early reinforced concrete bridges were the Eden Park Bridge, Cincinnati, Ohio, a 70-foot arch span built in 1894–95; the five-span arches over the Kansas River at Topeka, Kansas, built in 1896 with a maximum span of 125 feet and a total length of 539 feet; and the 36-foot span stone-faced reinforced concrete arch built according to the Melan method near Rock Rapids, Iowa, in 1894. This latter bridge was moved to a roadside park when the bridge was replaced in 1964.[44][45]
The first 10 years of the 20th century saw a phenomenal growth in the use of concrete structures on both the highway and railroad systems. Many long-span concrete arch bridges with either plain or reinforced arches were built in populated places, probably because of the architectural improvement over the truss and trestle bridges of the time. The use of reinforced concrete for deck girders, culverts, slabs for steel truss and girder spans, and bridge abutments and piers soon became commonplace. Of special importance was the use of reinforced concrete slabs, still in use today, for bridge decks instead of timber, steel, or cast iron.
Construction to 1916
Through the years, the methods, materials and equipment for bridge construction had developed and improved with experience and invention. Steam-powered equipment had supplanted men and horses for heavy lifting and excavation. Deep foundation excavation in open cofferdams, cellular cofferdams, open caissons and pneumatic caissons had been developed and successfully used. Steam piston and pulsometer pumps had supplanted the chain and ship pumps of colonial days for draining foundation excavations. Improved timber sheet piling,[N 1] such as the Wakefield type, and steel sheet piling were developed which resulted in relatively watertight cofferdams.
Many steel fabricating shops had become established with power equipment for handling and fabricating the steel bridge members. Power shears, punches and drills were employed. Heavy hydraulic and pneumatic riveters were developed for driving tight rivets. Relatively efficient and structually satisfactory standards for riveted and pinned connections and other structural details had been developed and were in use, by 1916.
Efficient and safe methods had also been developed for erection of steel superstructures. Steam-powered cranes, derricks and gin poles[N 2] were available for lifting heavy members. Relatively light pneumatic riveting guns were developed for driving field rivets so that only small and minor structures had hand-riveted or bolted connections.
Most important of all, these construction improvements were documented in the literature, and engineers and technicians well versed in their application were available for bridge construction.
About 1870 engineers began developing shear and moment analyses to determine stresses for structural design. Before that, structures were “proof loaded” before acceptance, that is, highway bridges were subjected to carts loaded with pig iron or stone, while railroad bridges were subjected to two locomotives in tandem. While this primarily tested the quality of workmanship and protected against the most gross errors, it did not provide for overloads or fatigue failure. These proof-load requirements persisted in the specifications until the turn of the century.
However, engineers were aware that the proof-load was no substitute for a rational design and, as their knowledge grew, began to write design specifications. Whipple had already recommended a design load of 100 pounds per square foot, equivalent to the whole roadway area covered with men.[46] In 1875, as a result of a series of bridge failures, an American Society of Civil Engineers committee recommended live loading values for both railway and highway bridges. For highway bridges, loads varying from 40 to 100 pounds per square foot were given, depending on span and type of usage.[47] While railroad and “bridge companies” issued their own specifications for highway bridges, a concerted effort for a specification of national scope had to await the Operating Committee on Bridges and Structures of the American Association of State Highway Officials in the 1920’s.
In 1910, the Office of Public Roads established a Division of Highway Bridges and Culverts. This new division, upon request from a State or local authority, assisted in bridge design and construction and reviewed and advised on bridge plans and specifications prepared by States, local authorities or bridge companies. It also prepared and published bulletins on highway bridge and culvert design and construction, including typical plans for reinforced concrete culverts, abutments and piers and discussions on the use and design of steel truss and plate girder bridges.
In 1913, the Office of Public Roads issued Circular No. 11, Typical Specifications for the Fabrication and Erection of Steel Highway Bridges. This circular specified a live loading of interurban electric cars or a 15-ton road roller plus a uniform live load on the portion of the bridge deck not occupied by the roller.
The Era of Great Bridges
It was said earlier that the opening of the Eads and Brooklyn Bridges was the beginning of the great bridge building era. As noted, it continued slowly with the Williamsburg and Manhattan Bridges.
The Hell Gate Railroad Bridge in New York City, built in 1916, and designed by Gustav Lindenthal, spanned 977½ feet between support pins and was the longest and heaviest arch in the world. When it was built, it was considered an engineering marvel because of the location conditions and the long span. It was to have a profound effect on steel arch design.
At almost the same time, the long-span continuous truss had finally come to America in the form of two continuous spans of 775 feet over the Ohio River at Sciotoville, Ohio. (Strictly speaking, a continuous truss had been used in the approaches to an earlier structure, but Sciotoville was the first use of it for main river spans.) Coincidentally, the longest simple span truss, 720 feet, was built over the Mississippi at Metropolis, Illinois. Both of these were railroad bridges.
Not long after, a major cantilever truss with two cantilever spans of 1,100 feet, designed by David B. Steinman, crossed the Carquinez Straits near San Francisco, California.
The long-span records continued to fall through the 1920’s and 1930’s. Detroit’s Ambassador Bridge, 1,850-foot suspension span surpassed the Quebec cantilever’s 1,800 feet. In 1931, Othmar Amman’s arch over the Kill van Kull between Staten Island,New York, and Bayonne, New Jersey, set a new arch record of 1,652 feet.
In that same year (1931), the George Washington Bridge was opened. This bridge across the Hudson River in New York City reached 3,500 feet between towers. At one bound, it had virtually doubled the Ambassador’s span. Not only that, but it was one of the strongest bridges ever built, with the greatest capacity—eight highway traffic lanes on the upper deck and provision for future rapid transit below (ultimately six highway lanes were added instead) when the stiffening trusses would be added. While the increase in technical knowledge was one factor, the most important reason behind this great leap in span length was, as Roebling had realized years before, the great improvement in the quality of the steel wire.
Since then, the George Washington Bridge has been surpassed several times in span but not yet in strength and capacity. The three longest spans are the 4,200-foot center span of the Golden Gate in San Francisco (1937), the 4,260-foot span of the Verrazano-Narrows (1965) in New York City and also an Amman design, and the 3,800-foot center span of the Mackinac Straits Bridge which joins Michigan’s upper and lower peninsulas.
The George Washington Bridge. The upper roadway has eight traffic lanes; a lower deck with six lanes was added later.
San Francisco’s Golden Gate Bridge.
The Mackinac Bridge over the Straits of Mackinac, where Lakes Huron and Michigan meet. Completed in 1957, this bridge is nearly five miles long.
During the 1930’s, many other noted long suspension bridges were built, among them two beautiful spans, the Bronx-Whitestone in New York City and the Oakland Bay Bridge, and that fateful structure, the Tacoma Narrows Bridge over Puget Sound in Washington.
The Tacoma Narrows Bridge, like the Whitestone, was built with a plate girder instead of a stiffening truss. In addition, although 2,800 feet between towers, the cables were only 39 feet apart. Just over 4 months after being opened, it failed, literally twisting itself to pieces. The shallow stiffening girder, combined with the light dead load of the relatively narrow bridge, was unable to resist the aerodynamic forces that developed because of the steady winds of Puget Sound. This failure led to considerable research and improvement of future structures, such as the Mackinac and Verrazano-Narrows Bridges, as well as stiffening of existing bridges. Since the Whitestone Bridge showed signs of serious oscillation, a half truss was added to the top of the original girder. It is noteworthy that the George Washington, despite the absence of a stiffening truss, showed few symptoms of instability, possibly due to its great weight and width, as well as to the less constant wind patterns of New York Harbor.
Current technology seems to make it less necessary to build record spans. However, even today, a new steel arch with a record span of 1,700 feet is under construction at the New River Gorge in West Virginia.
These “great bridges” were all toll facilities built by bond issues rather than Federal-aid financing. After enactment of the Federal Aid Road Act of 1916, the Division of Highway Bridges and Culverts, established in 1910, became the Bridge Division of Public Roads. The immediate task of this Division was to set standards for design and construction of bridges to be constructed under the 1916 Act. Under the Act, roadway and bridge planning became a co-operative undertaking with the States initiating, planning, designing and constructing the projects and the Bureau of Public Roads (BPR) advising, approving, committing Federal-aid matching funds for satisfactory plans and specifications and paying such funds upon successful final inspection of the completed projects.
This cooperation was facilitated by the formation, in 1921, of the Operating Committee on Bridges and Structures of AASHO, known popularly as the AASHO Bridge Committee, which was composed of the bridge engineer of each State highway department and a designated bridge engineer from Public Roads.
The Tacoma Narrows Bridge, shouting oscillation 1 hour before its collapse.
Architect’s drawing of the New River Gorge Bridge in West Virginia. The hinged arch has a span of 1,700 feet, the longest of its kind in the world. Construction began in 1973.
The AASHO Bridge Committee’s continuing function was to develop policies and specifications for design and construction of highway bridges. Public Roads bridge engineers served as chairman of the AASHO Bridge Committee from 1921 until 1953 and as secretary from then until the present time.
The AASHO Bridge Committee gradually developed the bridge specifications and issued the first printed edition of the AASHO Standard Specifications for Highway Bridges in 1931. These “Standard Specifications” were not just specifications, but served as a standard or guide for the preparation of State specifications and for reference by bridge engineers. They combined design criteria and policies with detailed specification guidelines. Primarily, “Standard Specifications” set forth minimum requirements which were consistent with the current practices. The same objectives prevail in the current edition.
A notable innovation in the specifications was the use of a truck system of live load instead of road rollers. The loadings, designated as H-20, H-15, and H-10, specified basic two-axle design trucks of 20, 15, and 10 tons, respectively. The H-truck loading was a basic truck in each lane of the bridge, preceded and followed by a train of trucks each weighing ¾ as much as the basic truck. An equivalent “lane loading,” consisting of a concentrated load plus a uniformly distributed load for each lane of loaded structure, was provided.
During the “great bridge” era between 1900 and World War II, there was a phenomenal increase in the number and size of lesser bridges as well as a growth in technical knowledge. The biggest new development was reinforced concrete. The short-span timber bridges were being displaced by concrete slab or concrete I-beam bridges. The reinforced concrete box culvert also became common. But aside from this, most of the growth consisted of further developments and refinements of existing bridge types. Riveted joints in steel trusses replaced the earlier pin connected joints, and stiffer and more substantial truss members developed. Steel rolled beam and built-up “plate girder” structures became common. The small arch brige changed from stone to concrete and began to grow in span length. Wider bridge roadways were being used.
There were three new, or at least different, developments that were significant beginning in the 1920’s. The first was the construction of a “parkway system” in Westchester County. This progressive county began building roads with some control of access and separation of cross traffic. This kind of road meant bridges wherever traffic was to be separated, even where there was no river to be crossed. Usually, these grade separation structures were not very large but did need eye appeal. The result was the rebirth of the stone arch (or stone-faced at least) and its variation, the rigid frame. The Westchester Parkway used many stone-faced reinforced concrete arches and frames for the Bronx River, Cross County, Saw Mill, and Taconic Parkways. Arthur Hayden was the bridge engineer at Westchester, and his book on the Rigid Frame Bridge has become a classic. The parkway and freeway concept spread to Connecticut where the Merritt and Wilbur Cross Parkways were built, again with emphasis on esthetics. Every bridge on the Merritt Parkway is different. This road is still one of the most delightful and scenic in the United States.
The second development was the almost exponential increase in engineering theory and application. While there were many whose studies made these new frames and arches possible, there was one outstanding contribution. In 1932 Professor Hardy Cross of the University of Illinois presented the method of moment distribution for determining the moments and shears in continuous beams, arches, and frames in a paper to the American Society of Civil Engineers. Prior to this time, the design of continuous frames was a tedious, intricate and time consuming process which discouraged the use of indeterminate structures. Moment distribution came at a very favorable time—at the beginning of the widespread use of highway grade separation structures for which rigid frames were especially appropriate.
The third significant factor was the growth of Federal interest and activity in the highway field. Public Roads was still a relatively small agency, but it helped the States through the transition to heavier automobile loadings. While the designated Federal-aid system at the start of the Federal-aid program was in most States essentially a system of county roads located and designed for preautomobile traffic, the quality of bridges increased dramatically during the 1920’s because of the cooperation of the States and Public Roads, increased experience, and improved criteria, specifications and guides. The example of the improved highways and bridges created a demand for similar improvements on other State highways. In general, bridges on the Federal-aid system in this period were of short to medium spans and of moderate cost. The Federal -aid allocation and the State highway funds were not sufficient to finance high cost structures and at the same time to construct other highway facilities in the State. Consequently, high cost bridges and tunnels were frequently built by bridge or tunnel authorities or private interests as toll facilities.
On U.S. route 5 in Vermont, large vehicles were forced to straddle the pavement centerline in crossing this narrow bridge with its sharp-curved approaches.
During these years, the AASHO Bridge Committee continued to develop the bridge design specifications. Basic two- and three-axle truck design loads were substituted for the truck train loading in 1941, leaving the concentrated truck loading for short span designs and the conventionalized lane loading for longer span designs.
This continuous steel truss bridge over the Allegheny River on I-80 in Pennsylvania was completed in 1968. The 68-foot wide concrete deck easily accommodates four lanes of traffic and a concrete median barrier.
The design formulas for concrete floor slabs were revised in the 1941 and the 1961 editions of the AASHO Bridge Specifications to conform to the results of slab tests.
New revised editions of the specifications continued to be developed and issued at 4- or 5-year intervals up to the present day. The revisions keep the specifications up-to-date on new design concepts, safety policies, use and development of materials and construction practices. The AASHO Bridge Specifications have become a valuable addition to the bridge building profession, serving as guides to State highway departments, cities, counties and foreign countries in the design of adequate but economical structures of all types.
A new design concept evolved during the 1930’s—that bridge alinement should conform to the overall alinement of the highway. Previously, bridges, which cost much more per unit of length than roadways, were generally located for minimum cost reasons, with little regard to approach road curvature. The new concept has resulted in much safer approach alinements.
Developments After World War II
The Federal-Aid Highway Act of 1944 authorized the National System of Interstate Highways. Initial standards for construction of the Interstate System were developed jointly by Public Roads and AASHO in 1945. The high standards for the Interstate System to meet present and future traffic needs also necessitated changes in the character of highway bridges. Standards for traffic capacity, load capacity, safety and appearance of structures were given careful study in order to provide for the safe and free movement of vehicles over and through bridges.
The field of the bridge engineer was broadened because of the number of large bridges required over waterways and because of the greatly increased number of grade separations.
Bridge railings were studied and led to the use of rail curbs and streamlined railings. Full width roadways (surface and shoulders) were advocated for safety on short bridges, and deck type rather than through type bridges were emphasized. Particular attention was given to horizontal clearance at underpasses, with greater clearances to sidewalls and center piers than had been customary.
Design with continuous spans was advocated. In anticipation of the increased use of three-span continuous reinforced concrete structures for urban and Interstate bridges, Public Roads prepared a report, Computing Moments for Continuous Concrete Bridges, including design procedures and moment charts, which was published in the January–February–March 1944 issue of Public Roads.
I-480 exit ramp in Omaha, Nebr. This was one of the first major interchanges where curved composite steel box girders were used.
While the use of AASHO H20 loading was quite general, the new and heavier H20-S16 loading was adopted for bridges on routes that would probably be included in the Interstate System. There was a general trend toward the use of heavier loading design of bridges on all highway systems.
With the toll turnpike era starting again after the war, good architectural treatment of structures was encouraged, especially where improvement could be made without substantial increase in cost. Consulting engineers and architects were retained for design of these toll road structures and for other major bridges on the Federal-aid system.
In the 10 years following World War II, the march of new developments in the bridge field resumed. One of the most widespread was the use of composite steel beam bridges, which enabled the deck slab to work with the steel beams as a main load-carrying member.
Another development was the use of welded bridges. Welded steel bridges, except for minor details, had not been permitted on the Federal-aid system until after World War II because of the lack of toughness (and resultant welding inadequacy) of most structural steel used for bridges. At first, welding on bridge members was limited to welding flange cover plates to beams and web to flange connections on plate girders. At BPR’s urging, steel producers developed a weldable steel for bridges which had sufficient chemical controls to produce a tougher steel. The American Society of Testing and Materials (ASTM) adopted the Specifcation for Structural Steel for Welding (A373) in 1954. ASTM A373 steel was supplanted in 1960 by ASTM A36 steel which had the same toughness with a higher yield point. The availability of these steels rapidly expanded the scope and volume of welded structures and eliminated the riveted plate girder save for very large bridges and railroad structures. More versatile designs and economy of weight were obtained with welded designs.
At about the same time, high strength bolts were developed for use in place of rivets in the connection of structural members. These bolts, tightened to their proof load, clamp the plates tightly together and transfer the stress in joints by friction rather than by bearing and shear on the bolt.
However, the most significant development of this period was the use of prestressed concrete in which highly stressed steel wires introduced compressive forces into the concrete to offset the tensile stresses caused by normal service load. In the early 1950’s, reports on European use of prestressed-concrete bridges had aroused the interest of engineers and cement producers in this country. The first American prestressed bridge was the Walnut Lane Bridge in Philadelphia, Pennsylvania, built in 1951. Subsequently, in 1955, Public Roads published the Criteria for Prestressed-Concrete Bridges, the only such publication in the United States at the time. With the help of this pamphlet and the standard plans published by the cement industry and Public Roads, the use of prestressed-concrete for bridges rapidly developed and greatly expanded the span range of concrete structures.
The Red Bridge over the Seekonk River between Providence, and East Providence, R.I., a welded plate girder bridge of composite construction.
The desire for separation of grades led to extensive use of long sections of viaduct construction in urban areas on the Interstate as well as on other freeways. Divided roadways, liberal bridge roadway widths and increased pier and abutment clearances under overpasses became normal safety practices.
The mainline and ramps of the Mission Valley Bridge in San Diego, Calif., are prestressed concrete box girders. Architectural treatment gives the piers an old “Spanish flavor.”
The National System of Interstate and Defense Highways
The Federal-Aid Highway Act of 1956 provided for construction of the 41,000-mile Interstate System with an accompanying source of funding. This System, which was to be a modern, safe, limited access, divided highway, set highway and bridge design engineers to work developing ideas for facilities which would be adequate for estimated future needs. The tremendous number of structures required to maintain access control encouraged continuous study of structure types and construction methods. Coincident with this was the dawn of the computer age. The capability of these electronic machines to perform lengthy mathematical operations permitted more refined analysis of increasingly complex structures in less time. Just as moment distribution gave impetus to continuous frame analysis, the computer made indeterminate analysis commonplace. The elimination of the computational drudgery brought a return to more basic and theoretical solutions and away from the approximation methods of the second quarter of the century. The computer also created new approaches, such as the finite element method.
High strength steels were used for the welded arch truss members of the Sherman-Minton Bridge over the Ohio River on I-64 between Louisville, Ky., and New Albany, Ind.
Completed in 1973, the Sitka Harbor Bridge in Sitka, Alaska, is the first cable-stayed highway bridge in the United States. The cable-stayed welded steel box girder design was selected to make the structure as visually unobstructive as possible from Sitka and Castle Hill, the site of Baranof Castle, from which the Russians governed Alaska prior to 1867, when it became a U.S. territory.
The Poplar Street Bridge over the Mississippi River at St. Louis, Mo.
This pedestrian bridge over U.S. 41 in Menomonee Falls, Wis., was the first cable-stayed girder bridge in the United States.
Built in 1973, the pedestrian overpass of the 47th Street Bypass in Boulder, Colo., is a cast-in-place concrete box girder.
As happens so often, a breakthrough in one area seems to lead to one in another area. Prestressed concrete has already been mentioned. New structural steels were also developed to permit larger, more heavily stressed bridges. In 1959 and 1960, a new high-strength low-alloy structural steel for riveted construction, ASTM A440, and a companion steel for welded construction, ASTM A441, were produced to supersede the then-standard silicon steel.
In 1961, a temporary specification for a 90,000–100,000 pounds per square inch yield strength steel for long-span bridges was approved by Public Roads for use on Federal-aid projects. After 3 years, an ASTM specification for this steel, High-Yield-Strength, Quenched and Tempered Alloy Steel Plate, Suitable for Welding, ASTM A514, was issued, and the temporary specification was discontinued. This same procedure was followed in 1966 with the high-strength columbium and vanadium steels. ASTM issued a specification for the steel, A572, without notch toughness (impact testing to insure steel will resist fatigue cracking) requirements. BPR accepted the ASTM specification for use on Federal-aid projects with welded steel only if a special provision covering notch toughness was included in the project specification. This was done, but the use of the special provisions was discontinued in 1974 when ASTM included notch toughness tests for all structural steels.
The use of curved bridges produced other problems besides design, e.g., the great amount of scrap metal resulting from cutting the flanges to the curve. This required development of new fabrication methods whereby members were fabricated straight and then heated and bent to the desired curvatures, thereby eliminating the waste.
To achieve better quality control in welding bridge members, nondestructive testing by radiography, magnetic particle and ultrasonic testing were developed by Public Roads bridge engineers. Ultrasonic testing of welded groove joints was found to be a faster, more sensitive and cheaper method of locating weld defects than the methods previously used.
Among the many long-span bridges built on the Interstate System to cross wide waterways, the Poplar Street Bridge over the Mississippi River at St. Louis is one of the most notable structures. A truss or arch span was not acceptable since it was considered that its bulk would detract from the nearby Gateway Arch and the Eads Bridge. The 8-lane bridge, constructed in the mid-1960’s with a 600-foot center span, is the Nation’s longest box girder bridge and the first large structure in the United States to use orthotropic design. Orthotropic design, developed in Europe, consists of main girders and a stiffened steel plate deck welded together so as to act jointly in supporting the structure. The stiffened plate deck serves a fourfold purpose as bridge deck, stringers, and an upper flange for both the floor beams and the main girders, reducing the dead load of the structure.
Pedestrian crossings of freeways became necessary in some locations to avoid undue division of established neighborhoods. Topography and personal safety of pedestrians generally dictate the design of overpasses. Protection for the traffic below from objects dropped or thrown from pedestrian crossings led to the use of enclosures or high fences of closely meshed wires on the overpasses. Many unique and attractive structures have been developed for these overpasses.
Bridge Research
The AASHO Road Test near Ottawa, Illinois, discussed in Chapters 4 and 6, is a good example of Federal-State cooperation on research. Public Roads bridge engineers designed and prepared the plans for the 18 bridges in the program. The bridges were designed with high working loads so that fatigue failures could be expected. There were 18 steel and/or concrete beam spans representing contemporary practices.
At the conclusion of the regular test traffic in December 1960, 7 of the 11 surviving bridges were subjected to accelerated fatigue tests, and two were tested to failure by increased vehicle loads. The mass of test-to-failure data on the behavior of the bridges under repeated loading has proved to be of major assistance in subsequent studies for developing specifications and design revisions.
Seven years later, full size welded plate girders were fatigue tested at Lehigh University under the sponsorship of the Welding Research Council. Analyses of the results of these tests resulted in design and specification changes for welded and riveted plate girders and confirmed the integrity of properly designed and constructed welded girders.
Since the late 1920’s, research and development has been greatly broadened in the nonstructural areas, such as hydraulics. While bridges and culverts have been used for centuries to cross streams and rivers, the structural elements of these crossings have attracted most of the attention of engineers with less attention given to the bridge’s capacity to accommodate floods, except for very large bridges over major rivers. Progress in highway hydraulics and drainage was slow, and in the early days, designs were frequently based on judgment without well developed engineering technology or adequate rainfall data. Drainage structures, including bridges, were commonly sized by using empirical formulas developed in the 19th century.
Progressive engineers had long recognized the shortcomings of drainage design and had stressed the importance of estimating the magnitude and frequency of flood flows and the risk of damage. The lack of hydrologic data and the dearth of information on hydraulics of highway structures made it difficult to develop policy and design procedures.
Research on hydraulics and hydrology started in the 1920–1930 era, mainly encouraged by Public Roads and other agencies of the Department of Agriculture, and after World War II, extensive development in the application of hydraulic engineering principles to highway design began in Public Roads. During the 1950’s, an extensive program of research on the hydraulics of culverts, bridges, and storm sewers was undertaken, largely by contract with various universities and other Federal agencies.
The infamous Silver Bridge over the Ohio River at Point Pleasant, W. Va.
In 1959 a Hydraulic Branch was established in the Public Roads Bridge Division. The functions of this Branch were to initiate and participate in hydraulic research, to disseminate the practical application of hydraulic design to bridge engineers and to review hydraulic features of Federal-aid structures. Studies were also made on stream pollution from highway and bridge construction and guidelines were issued on reduction of such pollution.
One of the most important contributions in this area was BPR’s Hydraulic Engineering Circulars and Hydraulic Design Series Bulletins. These publications, still in use, provide practical methods for applying research results to designs and are widely used by Federal, State and consulting engineers.
Other areas of bridge related research and development included safety measures, with particular emphasis on:
- Elimination or neutralization of hazardous fixed objects along the roadside including bridge piers, abutments, parapets, and culvert headwalls.
- Strengthening of bridge guardrail designs and the new concept of a strong beam (rail) with a weak post that would yield partially and redirect the errant vehicle safely back to the road.
- Breakaway sign supports and lighting poles that break when hit by a vehicle, thereby reducing the injury to vehicle and occupants on collision. In collisions, the vehicle usually passes under the sign or pole before it falls. These concepts were developed at Texas A&M through State and Federal funding.
Gore[N 1] protection devices to protect out-of-control vehicles at the diverging point of ramp exits and branching roadways. These rapidly reduce the speed of an out-of-control vehicle by the absorption of energy by using springs, crushing of drums, displacement of water or sand or related methods.
New criteria for design of sign supports, including new recommedations for certain aluminum alloy design stresses, accepted and published by AASHO in 1961.
- ↑ The elongated triangular area where one-way roadways begin to separate.
The Scene Today
The State transportation agencies, the Federal Highway Administration, the highway industry and research groups have achieved a finely tuned, cooperative highway program. It is difficult to say that any one group is responsible for new developments, since each concerned group has played a part. This is the way it has been with the latest developments in bridge design, ultimate strength design (USD) for concrete and load factor design for steel. The Bureau of Public Roads supported these developments vigorously and was in the forefront of this movement when it published the Strength and Serviceability Criteria—Reinforced Concrete Bridge Members—Ultimate Design in 1966. This helped bring the bridge specifications up to the American Concrete Institute (ACI) Building Code Requirements for Reinforced Concrete which had adopted USD in 1963. A Tentative Criteria for Load Factor Design of Steel Highway Bridges, with commentary, was developed and circulated by the American Iron and Steel Institute in 1968. This criteria for steel design was similar to the strength and serviceability criteria for USD in that the designs are based on yield conditions and load factors. Basically, these methods represent a departure from designing a structure for service loads with an overall safety factor. Under USD and load factor design, magnification of individual safety factors are applied to the various types of loads. The magnified loads are then totaled and used to proportion a member assumed to be in a state of incipent yielding. These methods, with additional controls for overload, excessive cracking, deflection, vibration, and permanent set and for fatigue of material at working loads, have been incorporated in the AASHTO Bridge Design Specification criteria.
Comparisons of costs have shown an initial economy in material costs for both steel and concrete members proportioned by these methods over those proportioned by working stress design, especially for long spans.
The events chronicled here, aside from a few setbacks, have been a series of successes. One failure should be recorded to alert the bridge engineer fraternity that these is more to be learned about materials and fatigue structures. On the evening of December 15, 1967, a 39-year old eyebar suspension bridge quivered slightly and then collapsed, carrying 46 persons to their deaths in the icy Ohio River. This catastrophic failure of the Silver Bridge at Point Pleasant, West Virginia, shocked the Nation.
An extensive investigation, including mechanical and chemical tests of the failed structure, showed that the failure was initiated by fracture of an eyebar in the suspension chain. It was further determined that a small corrosion pit on the pin hole face of the eyebar started a small crack and that the crack reached critical size under the joint action of stress-corrosion and fatigue.
The C&O Bridge over the Ohio River between Cincinnati, Ohio, and Covington, Ky., was removed by explosives after inspection revealed it was unsafe for highway traffic. The replacement bridge was opened to highway traffic in 1974.
The disturbing part about the Silver Bridge failure was the contributing causes listed as:
- In 1927 when the bridge was designed, the phenomena of stress corrosion and fatigue were not known to occur in the classes of bridge material used under conditions of exposure normally encountered in rural areas.
- The location of the flaw was inaccessible to visual inspection.
- The flaw could not have been detected by any inspection method known to the state-of-the-art today without disassembly of the eyebar.
These revelations led to a program of immediate bridge inspections and bridge inventories directed by a committtee formed at the direction of the President. As a result of this committee’s actions, new inspection equipment is being developed and purchased and sophisticated devices are being employed for inspecting bridge members and for determining scour at bridge piers. Bridges deemed unsafe are being closed to traffic, and bridges are being reappraised as to load capacities.
The 1968 Federal-Aid Highway Act mandated national bridge inspection standards for the Federal-aid systems. These standards establish inspection qualifications and require biennial inspections. All States are required to maintain written inspection reports and a current inventory of all bridges on the Federal-aid systems.
The 1968 Act also required that an inspector training program for employees of the Federal and State governments be established and be kept current with new and improved techniques. In the Federal-Aid Highway Act of 1970, a bridge replacement program was required to replace substandard bridges over waterways and topographic barriers. This program is now in full swing. Many of the most deficient bridges have been or are being replaced. There are thousands of bridges on the Federal-aid system that are posted as having limited capability of carrying truck traffic. It is presently estimated that replacement cost would be approximately $2.3 billion.
The 60 years since the first Federal-aid act have been a challenge to the bridge engineer. The first goal was to set a plan of cooperation and encouragement between Federal and State organizations in developing bridge standards, then to devise a comprehensive and workable program for planning, designing, approving and inspecting Federal-aid bridges to meet the current and predictable future needs of traffic. With no early background for predicting traffic growth and facing an unprecedented increase in vehicle volumes, speed and weight, increased higher standards for width, alinement and strength of bridges were required, and many existing bridges became obsolete.
Continuous and careful attention to new developments was necessary to maintain current criteria and specifications for bridge planning and design. The first 60 years’ pursuits and accomplishments leave a background of experience and knowledge for continued accomplishmment of future comprehensive programs.
Highway Tunnels
As was the case in Europe, tunneling activity in this country began with the mining industry, followed by canal tunnel construction, then the railroads and finally highways. Most of the early canal and railroad tunnel work was done in Pennsylvania, but California holds several important “firsts” for highway tunnels.
By stretching the definition of a highway tunnel to include tunnels constructed for horsedrawn traffic before the automobile came into general use, probably the oldest highway tunnel in the United States is located in California. Built during the 1870’s, the tunnel pierces a high rock cliff on the Pacific Ocean about 6 miles south of San Francisco. Complete and accurate records concerning this tunnel are lacking, but historians speculate that the bore was built to permit local ranchers easy access to San Francisco along the beaches.[48][49]
The first tunnel of substantial length constructed to accommodate automobile traffic is also located in California. Completed in 1901, the Third Street Tunnel passes through Bunker Hill in downtown Los Angeles. Interestingly, the contractor made an unsuccessful attempt to use a tunnel boring machine on the project.[50]
The pioneer spirit and Yankee ingenuity of American tunnel builders was perhaps most evident in the 1920’s. During this era, some of the most famous highway tunnels in the United States were built. Early in the 1920’s a comprehensive research program developed the fundamental data for vehicular tunnel ventilation. The research was conducted in connection with the design of the Holland Tunnel.[51] Passing under the Hudson River between New Jersey and New York, the Holland Tunnel was the first major subaqueous vehicular tunnel in the United States. It was opened to traffic in 1927. Three years earlier the Liberty Tunnels were completed in Pittsburgh, Pennsylvania, and, at 5,800 feet, were the longest highway tunnels in the country at that time.[52] About the same time, the Posey Tube under the estuary between Oakland and Alameda, California, was nearing completion. When opened to traffic in 1928, the Posey Tube became the first highway tunnel in the world to be built by the “trench method.”[53] An American innovation, the trench method consists of sinking pre-fabricated tunnel sections into a prepared trench on the riverbed. A total of 13 highway tunnels have been built by the trench method, the latest being the Wallace Tunnel on Interstate 10 in Mobile, Alabama, opened to traffic in 1973.
The decade of the 1930’s saw most of the tunnel construction being concentrated in the western States. Many of these tunnels were located on the new access routes into the national parks. Probably the most notable accomplishment during the 1930’s was the Yerba Buena Island Tunnel connecting the San Francisco-Oakland Bay Bridges in California. Completed in 1937, it remains the only double-deck highway tunnel in the country.
Immediately following World War II, tunnel construction activity quickened, and in 1950 another record was established. The Brooklyn-Battery Tunnel under the East River in New York City became the longest highway tunnel in the country—a distinction it still holds today.
Since the early 1960’s, the bulk of U.S. highway tunnel construction has been within the Interstate System. To date 15 major tunnels have been completed on the Interstate, 4 are under construction and 13 others are under design. These facilities are designed to safely handle high-speed traffic, a far cry from that narrow passage carved through a cliff on a California beach. REFERENCES
- ↑ L. Edwards, A Record of History and Evolution of Early American Bridges (University Press, Orono, Maine, 1959) pp. 47, 48.
- ↑ Id., p. 197.
- ↑ Id., p. 40.
- ↑ Id., pp. 49, 50.
- ↑ Id., p. 50.
- ↑ Id., p. 56.
- ↑ Id., p. 61.
- ↑ Id., pp. 80, 82.
- ↑ Id., p. 63.
- ↑ G. Hunt, A Century of Engineering in Wood Preservation, Proceedings of Wood Symposium, September 3–13, 1952 (Timber Engineering Co., Washington, D.C., 1952) p. 66.
- ↑ J. Gies, Bridges and Men (Doubleday, Garden City, N.Y., 1963) p. 107.
- ↑ H. Haupt, General Theory of Bridge Construction (D. Appleton and Co., New York, 1853) pp. 243–247.
- ↑ J. Gies, supra, note 11, p. 112.
- ↑ L. Edwards, supra, note 1, p. 75.
- ↑ Id., p. 71.
- ↑ Id., p. 74.
- ↑ Id., p. 76.
- ↑ J. Gies, supra, note 11, pp. 155–177.
- ↑ A. Reichmann, The History of Bridge Development, Journal of the Western Society of Engineers Special Edition—75 Years of Engineering Achievements In the Midwest, Dec. 1944, pp. 60, 61.
- ↑ L. Edwards, supra, note 1, pp. 125, 126.
- ↑ Id., p. 204.
- ↑ Id., p. 122.
- ↑ Id., pp. 41–44.
- ↑ Id., pp. 202, 203.
- ↑ Id., pp. 162, 163.
- ↑ Id., pp. 88–90.
- ↑ Id., p. 167.
- ↑ Id., p. 90.
- ↑ Id., pp. 199,200.
- ↑ Id., p. 190.
- ↑ Id., p. 194.
- ↑ Id., p. 198.
- ↑ Id., p. 197.
- ↑ O. Hovey, Movable Bridges, Vol. I (John Wiley and Sons, New York, 1926) pp. 146–149.
- ↑ Id., pp. 151, 152.
- ↑ L. Edwards, supra, note 1, pp. 203, 204.
- ↑ Id., p. 191.
- ↑ Id.
- ↑ Id., p. 198.
- ↑ Id.
- ↑ Id., p. 200.
- ↑ F. Turneaure & E. Maurer, Principles of Reinforced Concrete Construction (John Wiley and Sons, New York, 1911) p. 2.
- ↑ L. Edwards, supra, note 1, p. 137.
- ↑ Id.
- ↑ A. Reichmann, supra, note 19, p. 64.
- ↑ L. Edwards, supra, note 1, p. 140.
- ↑ Id., pp. 144, 145.
- ↑ E. Drew, First Highway Tunnel, California Highways and Public Works, Vol. 35, Nos. 5–6, May–Jun. 1956, pp. 12, 13.
- ↑ F. Stangeb, An Interesting Historical Letter, California Highways and Public Works, Vol. 35, Nos. 9–10, Sept.–Oct. 1956, p. 37.
- ↑ J. Quinton, The Third Street Concrete and Brick-Lined Tunnel, Los Angeles, Cal., Engineering News, Vol. XLVI, No. 3, Jul. 18, 1901, pp. 34–37.
- ↑ Studies and Methods Adopted for Ventilating the Holland Vehicular Tunnels, Engineering News-Record, Vol. 98, No. 23, Jun. 9, 1927, pp. 934–939.
- ↑ Pioneer Highway Tunnels Modernized for Service, Engineering News-Record, Vol. 131, No. 9, Aug. 26, 1943, p. 68.
- ↑ T. Merriman & T. Wiggin, American Civil Engineers’ Handbook, 5th Ed. (John Wiley and Sons, New York, 1949) p. 1647.