STEEL/CONCRETE COMPOSITE DESIGN FOR LONG SPAN BRIDGES Steven T. Hague, P.E., S.E. Associate Vice President HNTB Corporation 715 Kirk Drive Kansas City, Missouri 64105 [email protected]

ABSTRACT Composite construction has been used in bridges and buildings for many years, and for buildings and short span bridges the construction techniques are well developed. However, for longer spans, especially at restricted access sites such as major river crossings and environmentally sensitive areas, some unconventional approaches to composite construction are required. The focus of this paper will be the design of steel/concrete composite structures in long-span bridge applications. BACKGROUND Although bridges and buildings constructed of concrete floors supported on steel beams and girders have been in use since the early 1900s, the design of beams as composite sections is a relatively recent development, following closely with the development and acceptance of welding as a structural fabrication technique. In a conventional sense, the top flanges of steel girders are fitted with shear lugs, typically welded studs, to provide a positive means to prevent slip along the interface between the steel beam and the concrete deck. Conventional construction of a composite beam would be to erect the structural steel, complete with splices, connections, and other attachments as may be required to provide a stable framework for deck placement; installation of the deck formwork (either cold formed steel decking, prestressed concrete deck panels, or conventional plywood forms) and shear studs; and placement and finishing of the concrete deck. Generally, this method is quite satisfactory for building construction as there are no real obstacles such as deep streams or ravines, fast-flowing rivers, or major transportation routes to cross which require minimal disruption to underpassing traffic. However, in bridge construction where there is always an obstacle of one form or another construction methods are not always practical. THE PROBLEM As one can imagine, providing adequate shoring to support a structure with a span of several hundred feet would prove difficult at best. Add to that the excess compression flange area necessary to support not only the girder, but also the dead load of the concrete deck and additional compression flange bracing, and it is easy to see that the advantage of composite

1

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

construction quickly disappears. Except for live load, the entire structural system may as well be a non-composite section or elaborate shoring systems developed to support the girder while waiting for the deck to cure. Over the past twenty-five years or so, as analysis routines have become more sophisticated, universities have added to the knowledge base of structural behavior, and construction methods have been developed which utilize all components of a structural system, medium and long span bridges have become more efficient and economical. By beginning the design with a basic assumption as to the construction method and sequence, the engineer can develop a design which fully utilizes all components in place at a given stage of construction, without excessive materials or overstressing individual components.

Project Location FIGURE 1. Location Map - Bill Emerson Memorial Bridge

NONCOMPOSITE, COMPOSITE, NEITHER OR BOTH? As an example, consider the Bill Emerson Memorial Bridge over the Mississippi River now under construction at Cape Girardeau, Missouri (Figure 1.) Crossing the Mississippi River is a formidable undertaking in the best of circumstances, and a location clearly unsuitable to conventional bridge erection methods. The conceptual and preliminary design process for the Cape Girardeau bridge set the main span length at 1150 feet and determined that a cable-stayed bridge, of either concrete or steel/concrete composite construction, would be the most cost effective structure type for this location. Figure 2 shows an elevation view of the cable-stayed unit.

2

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

FIGURE 2. Elevation - Cable-stayed Spans

In order to effectively design a steel/concrete composite cable-stayed bridge, the engineer must include the anticipated construction technique in the design of the bridge. This includes the determination of the types, weight and location of construction equipment to be used in the construction, the anticipated construction schedule, and the pieces that will be shop-fabricated vs. the components to be installed in the field. Failure to accurately predict any of the above items will result in a structure that is either overdesigned, or requires the contractor to add steel to the sections in order to meet stress requirements during erection. An analysis of the conditions during erection will demonstrate that the bridge will be called upon to work in combined compression and bending during every stage of construction. Additionally, parts of the bridge will be acting as a composite section for bending and a different composite section for axial loads. This in turn effects the "locked-in" dead load stresses during construction. The conservative approach to the design of a steel/concrete cable-stayed bridge is to utilize a very detailed erection analysis in order to accurately capture the step-by-step locked-in stresses, then through superposition consider the superimposed loads the structure may be required to carry. During each step of the erection procedure the contractor will perform the following steps: 1. One field section of structural steel framing will be lifted and bolted into place. This section will be comprised of two steel edge girders and two floorbeams. The approximate weight of this field section is 70 tons. 2. Stay-cables will then be fabricated and attached to the tower and each edge girder, then stressed to an initial load of approximately 10 percent of final cable dead load force.

3

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

FIGURE 3. Superstructure Framing Detail

3. The contractor will then lift and place precast concrete deck panels, spanning from floorbeam to floorbeam and half the width of the bridge. Closure strips will then be cast atop the floorbeams and allowed to cure prior to longitudinal post-tensioning of the deck. 4. Once the post-tensioning has been completed, additional concrete strips will be cast above the edge girder flanges.

FIGURE 4. Superstructure with Deck Panels in Place

4

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

5. The stay-cables for that field section will then be stressed to the intermediate, or final erection force. Some minor force adjustments may be necessary after midspan closure to achieve final roadway profile. Upon close examination of these general construction steps it becomes obvious that the superstructure begins as a noncomposite section and with the placement of the closure strips the floorbeams and then the edge girders become composite sections.

SECTION CHANGES STEP BY STEP As each steel field section is erected it will be bolted into place and cantilever as a noncomposite section 11 meters (35 feet) beyond the previously completed section. The bridge is now ready for the installation of the precast deck panels. These panels, 13 meters (43 feet) in width, 5 meters (17 feet) in length and 280 millimeters (11 inches) thick, weigh approximately 50 tons each. With four panels per section, the dead load alone generates a great deal of bending in the floorbeams and edge girders as well as increasing the cable load and thrust component of that load. These forces all act on the noncomposite section of the current field piece and the composite section of earlier sections of the bridge. As can be seen in Figure 5, the previously erected section is partially composite and fully supported by stay cables. The effective width of composite concrete slab has two values; one for axial loads and another for bending. The effective concrete section at the leading edge of the composite slab is assumed to be zero, increasing at an angle of 26.5 to 30°, depending upon the reference one would choose to accept, until the effective width for axial load is equal to the width of the bridge. Since the basis for bridge design in the US is the American Association of State Highway and Transportation Officials (AASHTO), 30° was used for the design of the bridge at Cape Girardeau. This variable distribution of load into the deck somewhat increases the complexity of the computer model.

5

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

FIGURE 5. Distribution of Axial Load

In addition to the variable width of the slab effective for axial loads, as the section is erected, bending in the section acts fully on a noncomposite section. Then, as the stay cable is stressed, even to the minimal load of installation, the horizontal component of the load is carried through the noncomposite section to the previously completed field section. Then the load acts on an increasing larger composite section until the bridge section is fully effective. This force is additive to the "locked-in" forces from the erection of preceding components. Next, the panels are made composite with the floorbeams with a closure strip which is cast on the top flange between the panels. Once cured, the panels are post-tensioned in the longitudinal direction, then additional closure strips are cast on the edge girder flanges to make the edge girders composite. It is at this stage that the deck panels begin to benefit the structure. Although the full section may not be effective until two to three additional field pieces are installed, the deck begins to participate in carrying load.

CONCLUSIONS As can be easily recognized by the above description, the construction of steel/concrete composite bridges in the medium to long span range can be somewhat of a challenge to bridge designers. The complex analyses to consider non-composite and composite sections, various stress levels which may be "locked-in" to the non-composite section by the addition of a composite deck, and the combination of both bending and axial stresses which act on variable sections are the key to producing a safe, economical design. By careful consideration of possible erection methods and equipment which could be used and its location during construction ...

6

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.