Bridge Prestress Losses in Dry Climate

Abstract

A posttensioned, simply supported, box-girder bridge in southern Nevada was instrumented during construction beginning in June 1992. The study was triggered by concerns over the potential adverse effects of low relative humidity on prestress losses. The variation of prestress forces and deflection was monitored over a 24-month period. These measured data were compared with four loss prediction models: (1) American Association of State Highway Transportation Officials (AASHTO) specifications; (2) American Concrete Institute (ACI) Committee 209; (3) Prestressed Concrete Institute (PCI) Committee report; and (4) Naaman's time-step method. The measured data showed that the actual creep and shrinkage losses were less than those predicated by AASHTO by approximately 30%. Of the three time-step methods, namely, the ACI 209, Naaman, and the PCI Committee methods, the first two yielded creep and shrinkage losses that closely matched the measured data, whereas the PCI method underestimated the losses by 15%.