| Concrete fabricators, contractors and designers need accurate predictions of time-dependent deformations of concrete, such as creep and shrinkage. This is especially true for fabricators of precast, prestressed bridge girders, for which large differences in camber between adjacent girders can lead to significant contractual difficulties, increased cost, and legal consequences. In order to improve predictions of these deformations, good quality data and computationally convenient and well-calibrated models are needed.;Fabricators often use accelerated curing regimes to make the concrete gain strength faster and thus increase girder production rates. Little data currently exist on the creep and shrinkage characteristics of concrete for such a curing regime, and current models have not been calibrated using concrete of this type. In this research, an experiment was conducted to obtain data on the creep and shrinkage of high-strength concrete with an accelerated curing regime.;Current prediction models of creep and shrinkage all assume a constant stress history and deal with changes in stress by applying the principle of superposition, originally proposed in 1943. The validity of the application of the principle of superposition to creep strains has been greatly debated, and the experimental program was partly designed to test this assumption. Dealing with variable stresses through the use of superposition is not computationally convenient. A new one-dimensional rate-type model based on visco-elastic behavior was designed and calibrated using data from the experiments. This new model can deal with variable stresses without the need for superposition. Several configurations of the model were calibrated, and a good fit was achieved to creep data from concrete with a diverse set of loading and unloading histories. |
