Inaccurate prediction of long-term camber in precast prestressed concrete beams (PPCB) creates discrepancies in bridge construction which results in increased costs and construction challenges. Although emphasis is placed on erection and long-term camber, these challenges have been found to originate from the difference between the estimated design and measured camber at the transfer of prestress and will continue to increase with time. The estimate of the instantaneous camber is influenced by factors such as the method used for determining the camber as well as the assumed modulus of elasticity and initial prestress, while the measured camber can be affected by the measurement technique, quality of formwork and construction, and uniformity of concrete. To systematically study and eliminate the construction challenges associated with camber of precast girders, a research project was undertaken to examine the differences between the measured and designed camber of Iowa girders from fabrication through erection.

This thesis focuses on the causes of error found with instantaneous camber by examining the designer's prediction method and variables used for design, the construction practices used for producing a PPCB, and the precasters camber measurement method adopted by the precast industry. While the discrepancy between the measured and designed camber was found to be dependent on the girder type, the designed camber was underestimated due to inaccurate variables used for material properties and the design prestress force. In depth analysis was conducted to investigate different methods of camber prediction and to determine the effects that material properties have on camber. Evaluation of the historical camber measurements that were previously recorded using the current measurement technique revealed that camber is over predicted the majority of the time. Further examination of historical data exposed insufficient data that did not accurately capture the instantaneous camber. For this reason, precasters fabrication procedures were observed and instantaneous camber was measured by the project team from over 100 girders. Observations and independent camber measurements taken by the research team led to the identification of discrepancies between predicted and measured instantaneous camber.

Results found that friction, bed deflections, inconsistent top flange surfaces due to the finish that is applied, and inconsistent top flange surfaces due to the beam depth not being uniform along the length of the beam currently produce an average error of 17.6%, 2.8%, 5.2%, and 6.8% respectively. When predicting camber using the AASHTO LRFD (2010) modulus of elasticity, the agreement with the predicted and measured camber produced the best results at 98.2% ± 14.9% agreement. Accurately measuring camber to account for bed deflections, friction, and inconsistent top flange surfaces and predicting the material and fabricated properties of the girders, such as the modulus of elasticity, prestress force, etc., will improve the accuracy of instantaneous camber predictions.