Stochastic fracture-mechanical characteristics of concrete based on experiments and inverse analysis

The use of stochastic nonlinear computational mechanics in real-world applications faces a fundamental obstacle - the lack of detailed information about the stochastic properties of material parameters involved in the problem. The current paper describes the results of an extensive experimental program which focused on determining fracture-mechanical parameters and their stochastic models of concrete C25/30. The testing program consisted of compression tests on cubic specimens, three-point bending tests on beams with notch and wedge-splitting tests on cubic specimens with notch. In the case of the three-point bending tests, along with the standard evaluation of fracture-mechanical properties according to code specifications material parameter identification, artificial neural network based inverse analyses were carried out. In order to quantify the influence of the consistency of freshly mixed concrete on its fracture-mechanical properties, two concrete mixtures of the same strength class were tested: (i) mixture with a slump value of 45 mm and (ii) mixture with a slump value of 70 mm. In addition, the time development of fracture-mechanical parameters and their variability was studied. All results obtained from individual tests are presented, compared and discussed here. Stochastic models of selected parameters of the analyzed concrete for stochastic nonlinear FE-model analysis will be recommended.