Construction History Rocky River Bridge, Rocky River Ohio
The design for the new bridge, prepared by Cuyahoga County Bridge Engineer A. M. Felgate under the direction of County Engineer A. B. Lea, called for a concrete structure consisting of a central twin-ribbed open spandrel arch with a 280-foot clear span, five approach arches with 44-foot clear spans, and end abutments with flanking, curved retaining walls.
The bridge was designed to carry a 40-foot roadway with 8-foot sidewalks on each side. The sidewalks were widened to 12 feet for observation platforms at the main arch piers. The roadway was designed to carry two 60-ton electric interurban cars on two tracks and two lanes of automobile traffic. This plan for the bridge at Rocky River was very similar to the Walnut Lane Bridge at Philadelphia (1906-1908), whose central arch it exceeded by 47 feet. Thus, at the time of its design and construction, the Detroit Avenue viaduct over the Rocky River was the longest concrete arch bridge in the United States.
The site for the new bridge was particularly favorable to the construction of a concrete arch. The borings for the foundations showed hard shale with no seams, and the stone piers of the previous bridge proved the bearing capacity of the shale was quite high. "These conditions were particularly favorable to the construction of an arch, as the horizontal thrusts could be provided for very satisfactorily," A. M. Felgate later wrote. The elimination of bridge piers in the river was especially desired in order to avoid the flooding problems created by spring ice jams; the long span allowed the piers to be placed where the current could not affect them. The choice of a single-span arch was favored, too, by preliminary plans showing the cost of a two or three-span structure to be approximately the same as for a single span. Finally, in Felgate's words, "It was also possible with a single span to evolve a design of a monumental character, the dignity of which would be commensurate with the requirements of the locality."
The Rocky River Bridge is of the type known as Luxemburg construction, named after the Luxemburg Bridge (1903) over the Petrusse River in Germany. The essential feature of the type is a twin method of construction. Two comparatively narrow bridges are placed side by side; the space is then bridged over by a roadway. This parallel twin construction is carried through the main arch span, piers and approach spans. The piers are made in two halves and connected with a curtain wall, forming a portal. Each half is made hollow in order to save material.
According to bridge engineer Felgate, the Rocky River Bridge included three features marking an advance in the design and construction of concrete arches: the design of the main arch with a span of 280 feet, then the longest concrete arch of its type in the world; the method of raising the compressive strength of the concrete by the introduction of embedded stone slabs; and the use of steel (rather than timber) centering during the erection of the main arch.
The 280-foot arch consisted of a pair of plain concrete ribs, each 22 feet wide by 11 feet thick at the springing line and 18 feet wide by 6 feet thick at the crown. The ribs were designed so that the central line of pressure resulting from the dead load followed closely the center line of the linear arch. Accordingly, these ribs were not reinforced, as they were used to resist compressive stresses only, and under no circumstances could tensile stresses ever be introduced.
The central arch ribs were "reinforced" by embedded slabs of stone, placed radially and separated by a space of about 6 inches. Felgate estimated that approximately 35 percent of the entire volume of the main ribs was composed of these derrick stones, or "floaters," whose purpose primarily was to increase the compressive strength of the concrete. Some of the slabs were over 6 feet long by 4 feet wide by 1 foot thick, and considerable time was required to place them correctly. According to Felgate, the use of the radial slabs proved cheaper than, and as effective as, steel for reinforcement.
Wilbur J. Watson, consulting engineer for the contractor, evolved the design for the steel centering, the essential feature of which was a three-hinged steel arch. (Centering is the temporary framework, usually wooden, used in arch and vault construction; it is removed, or "struck," when the mortar has set.) The use of steel centering, according to Watson, constituted a "radical departure" from the usual method of building concrete arches, that of timber centering. (The arch ribs of the Walnut Lane Bridge were cast by means of timber centering.) The contractor for the Rocky River Bridge decided that it would be cheaper to use steel centering and that, in addition, its use would allow a clear waterway and thus eliminate the danger of destruction of the arch centers by ice in the river. Steel centering also offered greater facility of movement since, upon completion of the first arch rib, it was necessary to move the centers into position for work on the second one. This could be done easily by means of rollers. Finally, all of the 20-inch I-beams used in the centering as cross beams, stringers and joists were later to be incorporated into the bridge as floor beams for the street railway tracks.
A word is in order here with regard to the computation of stresses in the main arch. Felgate has written that these were determined by the graphical analysis known as "Gain's method," then checked independently by algebraic methods and found to agree. Tables published in the journal Engineering-Contracting illustrate the methods of deriving the stresses- in the 280-foot main arch ribs under different conditions of loading.
In the assumptions made for calculating the stresses, the concrete in the main arch rib was assumed to weigh 160 pounds per cubic foot, and all other concrete 150 pounds per cubic foot. The safe compressive strength of the concrete was assumed to be 600 pounds per square inch. The 60-ton street cars were assumed to produce, for each half of the bridge to one side of the center line, an equivalent uniform live load as follows: for the first 10-foot width of roadway on one side of the center line, 270 pounds per square foot; for the next 10 feet of roadway, 100 pounds per square foot; and for the 8-foot sidewalk, 80 pounds per square foot. (Eccentric live loading was not considered.) The wind load was assumed to be 30 pounds per square foot of vertical surface.
Determination for stresses was made for the following conditions: (1) dead load stresses; (2) dead load plus live load stresses; (3) dead load plus one-half live load stresses; (4) dead load of arch ribs, spandrel piers and walls; (5) temperature stresses; (6) stresses due to shortening of the arch from thrust; and (7) wind stresses.
According to Felgate, the combined stresses for all possible conditions produced their greatest effect at the springing line of the main arch, where the unit stress was 566 pounds per squaxe inch. Numerous tests of the concrete made during construction showed that after 30 days 6-foot cubes developed an average compressive strength of 2,400 pounds per square inch; thus a safety factor of 4 was obtained. As the age of the concrete increased, Felgate estimated the safety factor would probably be increased to 5.
Centering for the main arch was constructed of steel in the form of a three-hinged arch. Wilbur Watson, the Cleveland consultant who designed the centering, described it in an issue of Concrete Engineering:
"These steel centers are composed of two parts which form a three-hinged spandrel braced arch with curved upper chord. The curved upper chord will carry cross beams composed of two 20-in. I-beams at each panel point. The cross beams will carry the stringers which are composed of four lines of 20-in. I-beams and these stringers in turn will carry the joists which run transversely to the axis of the bridge and carry the timber bows. There are four of these cross beams in each panel and the bows vary in size from 12 x 12 at the springing to 3 x 12 at the crown. These bows will carry the 2-in. sheathing on which the concrete will be laid. The timber bows will be so constructed that they will be able to carry the tangential component of the weight of the arch without assistance from the steel truss.
The entire vertical weight of the arch is carried to the shoes through four sets of steel wedges, one for each shoe, each set composed of four wedges. These wedges are of annealed cast steel planed on all sliding faces and provided with a powerful screw for holding them in position and for lowering .... After the construction of the arch, the wedges will be lowered by unlocking and turning this screw and will then be jacked over into position for the construction of the second rib."
At the close of his discussion, Watson remarked that a centering very similar to his design was then being used for the Delaware Bridge on the Delaware, Lackawanna & Western Railroad at Portland, New Jersey, and that "this will probably be the standard method of construction for arches of this character in the future."
The erection of the contractor's plant and the excavation were begun on 29 October 1908, The existing high-level bridge was continued in service while the new bridge was under construction parallel to and immediately downstream from it. Contractor for the construction of the Rocky River Bridge was The Schillinger Bros. Co. of Columbus, Ohio. Their plant was established at the east end of the site and included an office, repair shop, boiler house, blacksmith and carpenter shops, saw mill and storehouse, cement house, storage bins, concrete plant, derricks and cableway. The cableway was a 900-foot Lidgerwood on movable towers 65 feet high. The concrete plant was located to the north of the cableway tower. Storage bins were built over the mixer, and the sand and crushed stone were delivered to the bins by a 1-yard "Williams" clamshell bucket operated by a 60-foot boom derrick. The bins were arranged so that the measuring hoppers were filled by the use of a lever which opened and closed the hopper gate. They also were provided with a series of steam pipes for heating the materials in cold weather. The concrete mixer used was a pugmill type with a capacity of 200 yards per day. The concrete was delivered from the mixer in 1-1/2-yard buckets on flat cars on a standard gauge track running to the cableway.
Excavation for the footings of the piers was carried down to the solid shale rock about 19 feet below the water line. The first concrete was placed on 5 December 1908. The average run of concrete was about 125 yards per day. According to one engineer's log, by late spring of 1909 all of the footings, the west abutments, twin piers No. 1, one pier No. 2 and one pier No. 5 had been poured, and the large piers Nos. 3 and 4 had been filled to about 5 feet of their height.
Special attention was devoted to attaining a finish on the exposed surface of the concrete that would be uniform and "in keeping with the monumental character of the structure."
Concrete as a material for bridge construction was in 1908-1910 still relatively novel, and the results it yielded were frequently marvelled at by the Rocky River engineers in their field notes.
Work on the first rib of the main arch was begun on 6 August 1909. Notes kept by the resident engineer for the County show the timetable for work on the first rib: "First concrete in main arch rib placed Aug. 6. Voussoirs completed Aug. 30. Started running keys Sept. 3. Final key placed Sept. 9. Centering struck Sept. 28."
The deflection of the steel centers required that the arch ribs be placed in large voussoir sections, separated by key spaces of 4 feet. Concrete struts were built in these key spaces with sufficient strength to carry the tangential component of the load. This was to prevent the concrete from sliding down the centers. The struts in keys A, B, C and D were 30 x 36 inches, 24 x 30 inches, 18 x 24 inches and 12 x 18 inches respectively, and were placed on the neutral axis of the rib, each key space containing three struts. The concrete struts were reinforced with 1-inch and, toward the crown, 3/4-inch square bars cut in 6-foot lengths.
The concrete in the voussoirs and struts was 1:2:4 (1 part Alma Portland Cement, 2 parts sand and 4 parts crushed limestone). Into this were placed the derrick stones ("floaters"), which averaged 1x3x5 feet in size. The large stones were placed with their edges toward the lagging, their width in as close to radial lines as possible.
In placing the concrete, the crown sections were loaded first; the crown deflection of the steel was about 1 and 3/4 inches, with quarter points rising proportionately. The placing of the succeeding voussoir sections, beginning with the haunch segments and moving toward the crown, overcame this condition and caused the quarter points to recede to their true position. After the voussoir sections were completed, the key spaces (except for key F at the crown) were filled with a concrete of the proportion 1:1:2. It was made very wet to insure that the struts between the sections were thoroughly surrounded. The rich mix was designed both to develop strength early and to make it equal in strength to the concrete containing the derrick stones. The lowest two spaces were filled first, then the second pair, and so on until the condition was exactly the same as a stone arch ready to receive the keystone.
The concrete for the key at the crown was mixed comparatively dry and thoroughly rammed into place. The top of the key was built up somewhat higher than the extrados of the adjoining sections and loaded with flat slabs of derrick stone to retain the dense mixture secured by ramming. This load was removed the day after the key was cast and the concrete was then cut down to the line of extrados.
Tests of cubes of the two concrete mixtures showed that the 1:2:4 concrete had a 30-day strength of 3,100 pounds per square inch; the 1:1:2 mixture in the keys had a 7-day strength of 3,200 pounds per square inch. Following this, it was decided to lower the centering and construct the second arch rib before the cold weather set in. This would allow the river to be entirely cleared of falsework and the dangers of the January floods and ice jams could be avoided.
A profile of the lagging taken prior to the removal of the centering showed that the intradosal curve corresponded quite closely to the theoretical curve; the center point at the crown was exactly right, the average variation elsewhere in the arch was less than 1/2 inch. Intradosal and theoretical curve lengths checked to about 1/2 inch.
The steel centering was jacked over into proper position for the construction of the second (north) rib. The labor of rebuilding the forms for the rib, which were left standing during the move, was thus saved. Procedures and methods for building the second rib were exactly the same as those employed for the first one, but the rate of progress was somewhat greater due to the experience gained in building the first rib and to having larger and better shaped derrick stone. Progress was as follows: the first concrete was placed 9 October; voussoirs completed 31 October; started running keys 30 October; final key placed 6 November. The centering was struck sometime around mid-November (no exact date is available).
The spandrel arches were built next. There were four spandrel arches on each side of the main arch ribs to carry the floor system of the roadway. These were identical in arrangement to the five approach spans. The spandrel arches each consisted of two ribs with their piers connected by a curtain wall. The ribs of each arch were also connected by a 6-inch reinforced-concrete slab that acted as a strut between them.
Like the main arch, the five approach spans consisted of two ribs, built in pairs. The concrete for each arch rib was started at both springing lines simultaneously and carried toward the crown at an equal rate in order to keep the weight on the centering balanced. The ribs of the approach arches were 5 feet wi4e and 3 feet thick at the springing line, 4 feet wide and 2-1/2 feet thick at the crown.. These ribs were reinforced over the extrados at both haunches and at the intrados for 9 feet on each side of the crown by six 1-inch bars.
The centering and forms for the 44-foot approach arches and the 21-foot spandrel arches over the main ribs were constructed of wood. The forms, prepared in a framing yard adjacent to the work, were bolted together so that they could be removed 24 hours after the concrete had been placed to allow for the washing and brushing of the concrete surface.
A subway, or hollow passageway, 3.25 feet high by 11.5 feet wide, was provided under both sides of the roadway for water, gas, sewer pipes and public service wiring connections. The floor of each subway was made by the 6-inch reinforced concrete slabs that acted as struts between the pairs of ribs of both the approach and spandrel arches. The roadway slab was reinforced over the subways with steel rods, but between the arch ribs it was carried by transverse 20-inch 65-pound steel I-beams embedded in concrete (those same I-beams that had been used as part of the steel centering for the main arch). The 100-pound T-rails of the street car tracks were attached directly to these beams. A three-ply felt waterproofing was applied over the concrete floor of the bridge. A 1-inch layer of sand was applied next, and the roadway was paved with brick. The joints of the brick were filled with tar.
No accounts of the finishing work on the bridge,the construction of the walks, balustrade and light poles, are available. Final cost of the structure was $224,850.00, nearly $30,000 less than the estimated expense. Bronze plaques bearing the names of the County Commissioners and Engineers, the Contractor and Mayors of Lakewood and Rocky River were installed at both approaches of the completed structure.
In 1910 Cleveland celebrated the County Centennial the week of 10 October "witji yells of Chippewas, flag raising by pioneer citizens, daring aerial features, and salutes from lakecraft." The week's program also included the dedication of the Denison-Harvard and Rocky River Bridges. At the Rocky River crossing, County Commissioner William F. Elrick presided over the ceremonies from a stand in the center of the bridge, Band music and a parade of 1500 decorated automobiles marked the occasion. According to newspaper accounts, Ohio State Senator Thomas P. Schmidt declared the new bridge to be "typical of the inventive genius of the age." Another speaker asserted that the bridge "demonstrated that no longer did American engineers have to go to Europe for ideas on bridge construction. In fact Europeans weres coming to America . . .
The Rocky River Bridge -- one of the most handsome and convenient bridges to be found in Northern Ohio," a "marvel of beauty and efficiency," in the words of one Lakewood writer-" was opened to traffic following its dedication on 12 October 1910. The burgeoning suburban population, many of whom earned their living in Cleveland, and the increasing number of automobiles would in the next twenty years earn for the bridge its dubious distinction as "one of the worst bottle-necks in Cuyahoga County."