On efficiency of identification of a stochastic crack propagation model based on Virkler experimental data

Abstract

In the paper we concentrate on one aspect of the experimental design: how the information coming from an experiment can be utilised for identification of a specific mathematical model. To express the consistency of the data and the model we need some quality measure, allowing to transform our intuition to numbers. As the mathematical tool we propose a version of the statistical procedure of cross-validation of the data. Then we verify the efficiency of the suggested method on the example of the Virkler experimental data of stochastic crack growth and the mathematical model of Paris-Erdogan of the fatigue crack growth.