Gamma prediction models for long-term creep deformations of prestressed concrete bridges

For long-span bridges as well as statically indeterminate frame structures it is essential to implement efficient and realistic prediction models for the long-term processes of concrete creep, shrinkage, and steel relaxation. In order to systematically study the main influential factors in bridge deflection measurements a probabilistic analysis can be per­formed. Due to the associated computational costs such investigations are limited. The predictions based on the highly scattered input parameters are associated with uncertainties. There is interest in alternative prediction models decou­pled from complex analytical and computationally expensive numerical models, using measured structural responses. A gamma process is an example of such an alternative method. This process is suitable for capturing evolving structural response quantities and deterioration mechanisms like crack propagation, corrosion, creep, and shrinkage, as reported in Ohadi and Micic (2011). The objective of this paper is to illustrate the use of gamma process approaches for the pre­diction of the creep and shrinkage performance of prestressed concrete bridges. The presented approaches incorporate uncertainties and make predictions more reliable with the help of structural health monitoring (SHM) data. The creep-shrinkage response of a prestressed box girder bridge serves for the calibration and evaluation of the considered gamma process approaches.