Structural performance of Texas U-beams at prestress transfer and under shear-critical loads

The Texas U-Beam standard designs were released in the 1990’s and have been used increasingly in bridges across the state since. While prototypes of the 54-in. deep prestressed concrete beam were built during the design phase, no full-scale load tests were performed.
This study of the U-Beam had five goals: (i) determine the magnitude and location of stresses induced in reinforcing bars in the end region of the beam at prestress transfer, (ii) measure concrete curing temperatures in square and skewed end blocks, (iii) establish the vertical shear capacity of the standard section, (iv) evaluate interaction between behavior at prestress transfer and performance under shear-critical loads, and (v) identify design and detailing improvements and make recommendations. Eight full-scale Texas U54 prestressed concrete beams were fabricated to achieve these goals. Load testing of the first four of these beams revealed a critical weakness along the bottom flange-to-web interface of the beam. The weakness caused failures that occurred at loads well below the calculated shear capacity. Given the horizontal sliding observed, the failure mode was called horizontal shear. The next two beams were fabricated to test three modifications to the end-region design, two of which were deemed successful. The final two beam sections tested contained the recommended new standard reinforcement and concrete geometry. A method to evaluate the horizontal shear demand on and capacity of the bottom flange-to-web interface of prestressed concrete beams was developed. The calculations were formulated using the theories of beam bending and shear friction. This method was calibrated and verified using the U-Beam test data, a series of small-scale specimens, and results of shear tests in the literature.
Stresses induced in reinforcing bars at prestress transfer met expectations set by existing codified equations. No modifications to the current U-Beam standard design are needed to manage these stresses. The induced stresses did not influence vertical shear behavior, and no interaction between the two is believed to exist for U-Beams. This dissertation contains the specifics of the beams tested and the data collected, and provides the details of recommended changes to the Texas U-Beam standard drawings.