Link to Latest Report: Coming Soon!
Precast pretensioned concrete girders offer low initial cost, low life cycle costs and excellent durability. Consequently, they are the most widely used girder type in the country for new bridges. When they were first introduced in the 1950s, the AASHTO Bridge Committee considered a provision to limit their span to a maximum of 60 ft. Today, the longest span that has been used is over 220 ft. This demonstrates the inexorable rise in span lengths over the years, driven by a combination of urban congestion and environmental limitations, which often prevents the placement of an intermediate column at the location best suited to the structural requirements.
Girder cross-sections have changed in response to the demand for longer spans. Typically, the depth has become greater, and the elements, thinner. While these changes benefit the in-service behavior, they simultaneously jeopardize lateral stability during handling. For girders shorter than about 125 ft, stability seldom poses a problem, but its influence increases with span length. The potential for instability is greatest during handling, because then the girder is free to undergo rigid body rotations, referred to herein as “roll”, about its “roll axis”, namely the line joining the two lifting points. The roll rotation causes a component of the girder weight to act in the weak direction, i.e. along the x-axis, with moments about the y-y axis. The resulting lateral deflection means that the girder weight is no longer directly below the roll axis, thereby causing further roll.
A need thus exists for longer girders, but the design requirements for them appear to be mutually exclusive; high strength and stiffness with respect to vertical loads (for in-service behavior), and good lateral and torsional resistance (to promote lateral stability), but low weight (for hauling).
UHPC is a candidate material, but the cross-section shapes that have so far been proposed (for example in PCI-sponsored research) appear to have focused on the in-service condition and to have disregarded stability.
Important parameters include both material and geometric properties. When a girder starts to roll, it risks cracking due to the lateral bending, and the subsequent behavior depends strongly on the material’s post-cracking characteristics. Conventional concrete essentially loses all tension strength after cracking, but fiber-based materials such as UHPC have the ability to use the fibers to bridge the cracks and to provide residual tension strength. This is likely to be very beneficial. On the other hand, UHPC typically achieves very high compressive strengths at, say, 56 days, but pretensioned girders typically need high strength at an age between 12 and 16 hours. At that age, the fraction of the 56-day strength achievable with UHPC is lower than with conventional concrete, especially if the latter is heat-cured. Thus, the full range of UHPC parameters will be investigated. This will be done from the studies available in the literature, including the extensive work at FHWA under the leadership of Graybeal, and the work done within the Center, e.g., Garber at FIU, and Floyd at OU.
The geometric parameters will again be selected from the literature, much of which has already been surveyed by the team in previous research. There, the work of Mast, Plaut and Moen (in the USA), Burgoyne (in the UK) and Lebelle (in France) stands out. However, few of these authors included cracking, so the team will likely make extensive work of their own previous studies on cracked sections. The goal is to elucidate the critical dimensionless ratios that determine the girder’s stability (or lack thereof). Ratios such as L/ryy, EIyy/GJ, etc., are obvious choices, but others will be sought as well.
This task follows from Task 1. The important characteristic is the complete stress-strain curve in tension, for which the best UHPCs typically maintain at least 80% of the peak tension strength out to approximately 1% strain. Graybeal has found this for proprietary UHPCs, but it is not immediately clear whether all non-proprietary UHPCs can match that performance. Girder producers are unlikely to use the proprietary material, so the question is important. Note that potential differences between the desirable characteristics identified in Task 1, and the available performance in this task may help motivate new research by materials experts.
The girder shape must include several important characteristics, such as:
- Space for a large number of bottom strands (i.e. a large bottom flange).
- High torsional strength and stiffness. This is usually concentrated in the web-flange junction, where the element thicknesses and fillets provide the majority of the section’s J value. However, a large solid region would add to the girder’s weight.
- A large Iyy, to inhibit lateral bending.
- Top flange dimensions that allow it either act as a precast, or to accommodate a cast-in-place, deck.
- Web dimensions that facilitate casting and that can accommodate prestressing strands. (While external tendons are theoretically possible, they pose many challenges which cannot be overcome within the scope of this project).
This step is seen as important. We will likely be in continuous contact with Concrete Technology Corporation (CTC), so their input may be in the form of an ongoing, continuous feedback rather than a one-off task. We also have a relationship with Van Dyke Trucking, who delivered the 223 ft (record length) girders, and who have a deep understanding of the torsional stiffness and other characteristics of the trucks needed for transporting long girders.
Task 5 – Combine the different constraints to optimize the cross-section
The foregoing constraints will be evaluated, and balanced, to provide an optimum girder design philosophy. We believe that a design philosophy, embodied in a set of design principles, is preferable to a single solution. For example, if shipping by water on a barge is possible, the transportation constraints, and the weight limits, are likely to be different from those applicable to truck-transported girders. The different constraints are likely to lead to different solutions.
Task 6 – Generate design examples with which to test the principles developed.
Design examples are important, not only for the purpose of illustrating the principles advocated, but also as a means for testing out the procedures that have been developed. (The latter can be seen in building codes. Several provisions come to mind in which the requirements are either unworkable in practice, or contain some significant internal inconsistencies, both of which suggest that no design example was run before the provision was adopted). For both reasons, design examples will be included as an important deliverable.
Principal Investigator: John Stanton
Co-Principal Investigator: Richard Wiebe
Research Assistant: TBD