Numerical approximation of time-dependent chloride diffusion model parameters via probabilistic Monte Carlo method

Correct evaluation of laboratory test data is an integral part of input data preparation for numerical models. In the case of the chloride diffusion model, the most important parameter is the diffusion coefficient which can be obtained, for example, from the measured chloride profile. Measured concentrations need to be approximated by curve most appropriate to the values; which can be done for instance by least-square method. Most of the currently used approximation methods do not consider the time dependence of the diffusion coefficient; thus, the results of the numerical simulation of chloride ingress can be inaccurate. The presented approximation based on stochastic Monte Carlo method includes the effect of concrete maturation. The diffusion coefficient, aging factor and theoretical chloride concentration on the surface can be determined. The selected procedure is demonstrated on measured data from laboratory tests.