Elasto-plastic TCD as a method of failure prediction

Development of the method for assessing the strength of engineering structures, considering the effects of the non-locality fracture in the area of stress concentrators, under different loading regime is one of the major scientific interests. In the present study the linear-elastic Theory of Critical Distances (TCD) is modificated for the case of elasto-plastic material behavior. The for plotting Johnson–Cook model was used for simulation of material behavior, and an elasto-plastic critical distance value was used in order to estimate the strength of the tested samples. The accuracy and reliability of the proposed design methodology was checked against experimental data: the cylindrical un-notched specimens and samples with stress concentrators of titanium alloy Grade2 were tested under tensile loading with different strain rate. Mechanical tests were carried out using a 300 kN electromechanical testing machine Shimadzu AG-X Plus and Gopkinson-Kolskiy's split bar. The obtained results showed that the use of the modification of the TCD based on elasto-plastic material behavior gives us estimates falling within an error up to 12% which means that the elasto-plastic TCD gives more accurate predictions than the linear elastic TCD solution (when using of elastic stress information the error interval is ± 15-20%).