Influence of Measurement Uncertainties on Results of Creep Prediction of Concrete under Cyclic Loading

During the last five decades, several experimental models and numerical models have been developed to predict the time-dependent deformation of concrete under cyclic loading. It is well knowntoengineers that creep accounts for a majority of structural deformation and failures. In plain and reinforced concrete structures, cyclic creep may lead to excessive deformation, excessive crack widths or structural collapse. The deformation in concrete under cyclic loading depends on a number of parameters, such as, the concrete composition, environmental condition, strength of concrete, mod- ulus of elasticity, the stress amplitude, the mean stress, number of cycle, wave forms and time under load. In this work, the quantification of uncertainty of creep models under cyclic loading is computed using four different creep models, (BP model, Neville und Whaley model, modified MC90 model and modifiedHyperbolic model) by considering different uncorrelated and correlated parameters respon- sible for cyclic creep. Four sources of uncertainty parameters uncertainty, model uncertainty, data uncertainty, and uncertainty of the creep phenomenon are considered for all of the models and com- pared showing significant differences. A general probabilistic method is developed for the prediction quality of creep models under cyclic loading. The Latin Hypercube Sampling (LHS) numerical sim- ulation method (Monte Carlo type method) was used. Further, global sensitivity analysis considering the uncorrelated and correlated parameters are used to quantify the contribution of each source of uncertainty to the overall prediction uncertainty and to identify the important parameters. The error in determining the input quantities and model itself can produce significant changes in creep predic- tion values. The variability influence of input random quantities on the cyclic creep was studied by means of the stochastic uncertainty and sensitivity analysis. All input imperfections were consid- ered to be random quantities. The Latin Hypercube Sampling (LHS) numerical simulation method (Monte Carlo type method) was used. It has been found by the stochastic sensitivity analysis that the cyclic creep deformation variability is most sensitive to the Elastic modulus of concrete, compressive strength, mean stress, cyclic stress amplitude, number of cycle, in that order. 1 Motra, H. B., Osburg, A. D. and Hildebrand, J.