Anchorage-Controlled Shear Capacity of Prestressed Concrete Bridge Girders

As part of the ongoing research on shear at the Phil M. Ferguson Structural Engineering Laboratory (FSEL) located at The University of Texas at Austin, the anchorage-controlled shear capacity of prestressed concrete bridge girders was in this research studied in two distinct ways, experimentally and analytically. The results of this research are an important step towards improving understanding of strand anchorage-related issues.

For the experimental program, two full-scale Tx46 prestressed concrete bridge girders were fabricated at FSEL. The Tx46 girders were topped with a concrete, composite deck. Both ends of the two girders were instrumented and tested. For the analytical program, a new Anchorage Evaluation Database (AEDB) was developed, by filtering and expanding the University of Texas Prestressed Concrete Shear Database (UTPCSDB), and then evaluated. The AEDB contained 72 shear tests, of which 25 were anchorage failures and 47 were shear failures.

The results and analysis from the experimental and analytical programs generated the following three main conclusions:

1. A reasonable percentage of debonding in Tx Girders does not have a marked impact on girder shear capacity calculated using the 2010 AASHTO LRFD General Procedure.

2. The AASHTO anchorage equation is conservative but not accurate. In other words, this equation cannot be used to accurately differentiate between a shear failure and an anchorage failure. In regards to conservativeness, anchorage failures in AASHTO-type girders may lead to unconservative results with respect to the 2010 AASHTO LRFD General Procedure.

3. The 2010 AASHTO anchorage resistance model and its corresponding equation do not apply to Tx Girders. Because of the Tx Girders’ wider bottom flange, cracks do not propagate across the strands as they do in AASHTO-type girders. This fact yields overly conservative results for Tx Girders with respect to AASHTO Equation 5.8.3.5-1.

In summary, this research uncovered the short-sided nature of the AASHTO anchorage design method. Given its short-comings, there is an obvious need for a validated, comprehensive, and rational approach to anchorage design that considers strength and serviceability. To appropriately develop this method, additional full-scale experimental testing is needed to expand the AEDB, as currently there are not enough tests to distinguish major, general trends and variables. Any future additional research would be expected to further validate and expand the significant findings that this research has produced and so take the next step toward safer, more-efficient bridge designs.