Improved Tub Girder Details

Sponsor: Texas Department of Transportation
Principal Investigators: Todd Helwig, Michael Engelhardt, Eric Williamson, and Patricia Clayton
Research Assistants: Stalin Armijos Moya, and Yang Wang

tub girder test setup as seen from above in warehouse

Figure 1 - Tub Girder Test Setup

Steel trapezoidal box girder systems, which consist of steel tub girders with a cast in-place concrete deck on top, are a popular alternative for straight and horizontally curved bridges due to their high torsional stiffness and aesthetic appearance. However, this type of girders possesses a relatively low torsional stiffness during transport, erection, and construction because of the thin-walled open section. Additionally, during the concrete casting, the top flanges of the tub girder are in compression in the positive moment region and are susceptible to lateral torsional buckling (LTB) and local buckling of the plates. Usually, top flange lateral bracing, in the form of a horizontal truss, is fully installed along the steel tub girder to prevent flanges from buckling and to increase the torsional stiffness of the girder. Additionally, tub girders are built with internal K-frames, which work as bracing to restrain distortion of the cross-section. This research study is focused on improving the efficiency of steel tub girders by investigating the impact of the girder geometry and bracing details on the behavior of the girders.

Two primary factors are under study in the tub girder cross section: offsets in the top flanges relative to the webs and flatter web slopes. By offsetting the top flanges, the top lateral truss members can have more space to be directly connected to the girder top flanges. This detail could ease construction by eliminating complex gusset plate details and flexibility in the connection plate, therefore improving the construction economy and girder performance. However, the impact of undesirable effects, such as increment of lateral bending in the top flanges, needs to be evaluated. Besides offsetting the top flanges, the impact of the web slope on the girder behavior is investigated. Current AASHTO LRFD design provisions limit the web slope to a ratio of 1H:4V (H-horizontal and V-vertical). The use of a flatter web slope can reduce the width of bottom flange, but global instability due to the decreased torsional and warping stiffness would be a major concern.

With regard to the bracing layout, the impact of different bracing configurations on the response of straight and horizontally curved tub girders is studied. The top lateral braces and K-frames are connected to the tub girders through bolted connections to facilitate the modification of bracing layouts between tests. Different bracing configurations can be achieved by removing (or adding) certain bracing members. For the top lateral truss, the number of panels (space between two consecutive struts or space between a strut and an end diaphragm) with top lateral diagonals along the length is a test parameter. Top lateral diagonals have been found to be more efficient in regions of maximum shear deformations for straight and mildly curved girders. Therefore, the top diagonals near mid-span are relatively ineffective and may be eliminated. Different internal K-frame layouts are tested as part of this research program.

The study includes large-scale experimental tests and parametric finite element analysis (FEA) studies. Elastic buckling tests are being conducted in three large-scale simply supported beams (84-ft long) which provide data to validate finite element models. These analytical models serve as basis to perform an extensive parametric finite element study of the tub girders with the proposed improved details.

After the initial phase of testing, shear studs will be welded to the top flanges and concrete decks will be constructed on the tub girders to obtain composite sections. The composite tub girders will be tested up to ultimate capacity. All the test specimens will be loaded to failure so that data can be obtained on the impact of each of the changes on the final response.

partial top lateral bracing as seen from above in warehouse

Figure 2 - Partial top lateral bracing

internal k-frame bracing as seen from inside the beam

Figure 3 - Internal K-frame bracing

lateral torsional buckling on control specimen, viewed within warehouse testing area

Figure 4 - Lateral Torsional Buckling on control specimen