Effects of Increasing the Allowable Compressive Stress at Release of Prestressed Concrete Girders

Over the last decade, an increase of the current allowable concrete stress in compression at prestress transfer (0.60f'ci) has gained considerable support within the precast / prestressed concrete industry due to several economic, safety, and efficiency benefits of relaxing the limit. To investigate the feasibility of increasing the allowable compressive stress at release, a research study funded by the Texas Department of Transportation was conducted at the University of Texas at Austin. The two main objectives of the project were to evaluate the impact of increasing the allowable compressive stress at release of 0.60f'ci on (i) the live-load performance and (ii) the initial camber of prestressed concrete girders.

In the live-load evaluation part of the current study, 36 static-load tests and 4 fatigue tests were performed. In the static tests, the cracking loads of the 36 test specimens were experimentally evaluated. Twenty-four of the specimens were scaled rectangular, tee, and inverted-tee beams. Twelve of the specimens were full-scale TxDOT Type-A girders. The maximum compressive stress at release for all of the test specimens ranged from 0.46f'ci to 0.91f'ci. The measured cracking loads obtained in the static tests were compared to predicted cracking loads. The effect of increasing the allowable compressive stress at release on the initial cracking of a pretensioned member was evaluated with the ability to accurately estimate the cracking loads of the test specimens. In the fatigue tests, two beams subjected to release stresses within the allowable limit (< 0.60f'ci) and two beams subjected to higher release stresses (~0.80f'ci) were tested under fatigue loads. The performance of the bottom-fibers of the member under the repetitive opening and closing of flexural cracks was evaluated in these tests. The results of the static-load and fatigue tests justified the increase of the allowable release stress in compression to 0.65f'ci for the girders tested in this study

In the initial camber evaluation part of the current project, an initial camber database was compiled with information from 223 pretensioned girders. Twenty-six pretensioned girders were subjected to a range of compressive stresses at release in excess of the allowable limit. The predicted and measured initial camber values of these girders were compared. The results indicated that increasing the compressive stress at release did not affect the ability to predict initial camber. However, in general, the magnitude of initial camber increased with increasing compressive stress at release. The remaining 197 girders in the initial camber database were conventional, full-scale girders that were used to address the current state of camber prediction for beams fabricated in Texas. The predicted and measured initial camber values for these girders were also compared. The results indicated that initial camber prediction can be greatly improved if local material variability, particularly the specific coarse aggregate, is accounted for in the initial camber estimate.