Very few long-span tied-arch bridges have been built recently in the U.S. due to concerns regarding redundancy of the structural system and the fact that cable-stayed systems are typically more economical. However, tied-arch systems can be developed to ensure sufficient redundancy and improve cost competitiveness. One such example is the new U.S. 20 bridge.
The new bridge will carry eastbound U.S. 20 over the Mississippi River between Dubuque, Iowa, and East Dubuque, Ill. There are 13 west and 23 east approach spans for a total bridge length of 5,635 ft. The estimated construction cost for the project is $160 million, which includes the main river crossing, approach roadway and other miscellaneous local roadway and bridge improvements.
The existing Julien Dubuque Bridge is listed in the National Register of Historic Places. When it was built, it was the second-longest span over the Mississippi River and longest continuous tied arch in the world. Its historic status requires that the new bridge meet the secretary of the interior's standards for rehabilitation. The new bridge must not destroy historic materials, features and spatial relationships that characterize the existing bridge. The span must be differentiated from the existing and be compatible with the historic materials, features, size, scale and proportion, and massing to protect the integrity of the existing bridge and its environment.
An extensive public involvement initiative and bridge type study was carried outto select a new bridge that was compatible with the existing historic bridge. A modern tied arch with an 845-ft span was selected. The rise of the new arch is only 105 ft, a rise-to-span ratio of 1:8, to match the lower chord profile of the existing arch. The adjacent approach span units will consist of 347-ft span continuous plate girders. The plate girders will be of variable depth and utilize high-performance steel.
The main span incorporates three major innovations that address challenging constraints and advance the state of the art in tied-arch bridge technology: 1) composite floor/tie system; 2) continuous deck; and 3) bolted, high-performance steel (HPS) tie girders. The end result is an efficient and durable floor system with a tie that has a high degree of both internal and structural redundancy.
The existing two-lane bridge is a three-span, 347 x 845 x 347-ft, continuous tied-arch bridge built in 1943 that will remain in place to carry westbound traffic.
The tied-arch structure has a main span of 845 ft to match the main span of the existing bridge. The arch rib follows the line of the lower chord of the existing arch, resulting in a rise of 105 ft and a rise-to-span ratio of 1:8. While this is lower than the optimal ratio of 1:4 to 1:5, it poses no unmanageable design conditions. It does, however, increase the thrust in the ribs and tension in the tie girder. The bridge consists of 21 equally spaced hangers at 38 ft 5 in. to match the existing bridge.
To accommodate the roadway and pedestrian walkway, the arch ribs are spaced at 56 ft center-to-center. The rib element is a 4-ft-wide and 7-ft-deep welded steel box. The rib is braced with a lateral system with an "X" configuration to match the bracing system on the existing bridge.
One of the most significant constraints imposed on the floor system was the necessity to minimize the overall depth to maintain the required navigational clearance and keep the two profile grade lines relatively close. The existing bridge has a total floor system depth of only 5.5 ft. This is because the spacing of the trusses is only 35 ft and the tie is shallow since the stiff arch trusses resist global flexure in the bridge.
A conventional floating stringer and deck system with the stringers running over the top of the floor beams requires a higher profile grade line that compromised the aesthetics and compatibility relative to the existing bridge. The solution was to develop and evaluate composite floor systems utilizing a relatively shallow tie girder.
Composite floor/tie system
The cast-in-place deck is 8 in. thick and consists of a mildly reinforced concrete slab that is composite with the tie girders, stringers and the floor beams. The deck is continuous over the length of the entire arch structure. The deck is placed with closure pours to avoid the induction of any dead-load tension into the slab. This also facilitates future deck replacement. Considering that the live-load thrust from the arch rib is small compared to the dead load, and the very large stiffness of the tie girder system, the live-load tension induced into the continuous slab is minimal.
The two tie girders are boxes located in the same vertical plane as the arch ribs. The boxes are built up from four plates joined using bolted angle connections in each corner. They are 2.5 ft wide x 7 ft deep. In all floor system alternates, the main tie girders transition to 10 ft deep x 4 ft wide at the piers (outside of the navigation channel) to match the depth of the flanking spans and the width of the arch rib, respectively.
The floor system is made up of W30 stringers that frame into floor beams located at each hanger location. Stringer connections to the floor beam at deck closure joint locations allow axial movement to prevent any accumulation of tension in the stringer and deck during concrete placement. Therefore, all dead-load tension is carried by the tie girders.
The critical aspect in the design of a tied-arch structure is to provide redundancy in the event of a tie girder fracture and failure. A redundant structure is considered to be one that upon failure of a member or element, the load previously carried by that member can be redistributed to other members or elements temporarily have the capacity to carry additional load without causing collapse of the structure. The general numerical approach is to analyze the structure with the failed member discounted at the failure location under full dead and live load and ensure that stresses remain within the elastic range.
The three measures of redundancy in a structure consist of internal, load path and structural redundancy. Internal redundancy is related to the ability of a single member, portion of a member or material to redistribute load around a damaged region. Internal redundancy is generally achieved by using a damage-tolerant or tough material, such as HPS, or by building up a structural steel member using bolts instead of welding. Welded members tend to propagate fractures into adjacent plates; whereas, the discontinuity created at bolted connections will arrest the cracks.
Load-path redundancy is achieved by having adequate parallel load paths and capacity to accommodate elastic redistribution of forces in a structure in the event of failure of a single member. A multiple (three or more) girder bridge is an example of a structure that is considered to have load-path redundancy as long as the deck and diaphragms have sufficient load redistribution capacity.
Structural redundancy is achieved by having adequate serial load paths. Continuous or statically indeterminate structures generally tend to have structural redundancy due to continuity. That is, the ability to redistribute forces to adjacent spans or members in the event of a failure of a member. In general, simple span structures and the end spans of continuous structures are not considered to be structurally redundant.
In the U.S. 20 bridge, flexural load-path redundancy is achieved through load redistribution in the opposite tie girder, hangers and ribs. In addition, the tie girders are structurally redundant in that the hangers provide continuity that allows load redistribution in a damaged girder. By using HPS and a bolted main tie girder, the tie system achieves a significant degree of internal redundancy.
The continuous, composite deck provides an additional load path to resist a portion of the tie force in the event of failure of a tie member, although this mechanism has been neglected in establishing redundancy. Redundancy of the ribs is not an issue since they remain in full compression under all load cases.
Bolted HPS tie girders
The recent trend in tied-arch design is to use bolted tie girders instead of welded girders, primarily to achieve a higher degree of internal redundancy. Even with a complete fracture of a single plate in the bolted tie for the U.S. 20 bridge, the stress in the tie remains well within the elastic limit under full dead and live loading. Internal diaphragm plates also are bolted to the webs and flanges to maintain internal redundancy. Bolting also reduces fabrication costs since welding in confined spaces is not necessary.
Discussions with fabricators also revealed that they generally prefer bolted tie girders because it eliminates the need for the distortion control measures associated with a welded box. However, they still prefer to use a welded section for the arch ribs due to the complications associated with bolting together a curved box.
This is the first bridge in the U.S. that the authors are aware of to use HPS tie girders. HPS provides an additional measure of internal redundancy due to its improved toughness characteristics compared to traditional bridge steels. The use of the higher strength grade 70W steel also is beneficial in resisting the higher tension force due to the small rise-to-span ratio and increase in flexural stresses due to the structure depth constraints previously mentioned.
The tie girder is made composite with the deck using welded shear studs. The deck strip over the tie is not placed until after casting the rest of the deck in order to minimize dead-load tension accumulation in the concrete.
Standing by history
The new U.S. 20 bridge will serve as a modern landmark and vital link for the cities of Dubuque and East Dubuque. However, the design pays respect to and complements the existing historic bridge. The project is one of the few major bridge projects in the nation that is a success in terms of gaining broad community stakeholder support, allowing the planning and design of the project to continually move forward.
Numerous constraints were imposed on the new bridge due to the historic nature and configuration of the existing bridge. This necessitated that creative yet technically sound solutions be developed, evaluated and implemented.
The use of a composite floor/tie system allows the use of a much shallower superstructure to maintain the required navigational clearance and keep the profile grade lines of the separate bridges relatively close. A continuous concrete deck eliminates the added cost of relief joints and stringer bearings and results in a much more durable system. The use of bolted, HPS tie girders provides a high degree of internal redundancy.