Considering the Statistical Distribution of Dynamic Fracture Toughness Data and the Actual Loading Rate at Fracture Initiation When Applying ASTM E1921 at Elevated Loading Rates

Abstract

Abstract ASTM E1921-15a describes the determination of the reference temperature T0,X for the loading rate X for ferritic steels in the transition range, where X is the order of magnitude of the average loading rate of the tests evaluated. Because this rate is defined for the linear-elastic part of a test only, the definition of the loading rate, valid for tests showing not only linear-elastic but also plastic behavior, is important. In a research project investigating the correlation of dynamic crack initiation and crack arrest funded by the German government, tests at −20°C on specimens of 22 NiMoCr 3 7 steel (A 508 Cl.2) were performed with different specimen geometry and loading devices in the range from 105 MPa√m s−1 to 3 × 106 MPa√m s−1. Evaluation according to ASTM E1921 shows differences of up to 30 K in T0,X obtained from linear-elastic test series and T0,X obtained from elastic-plastic test series with comparable dK/dt. New results were obtained in the current follow-up joint IWM-MPA project from test series at a loading rate of 5 × 105 MPa√m s−1. Results from tests with 1T C(T)-specimen at various test temperatures in the transition region are presented, analyzed, and reference temperature T0,5 is evaluated. The statistical distribution of the determined dynamic fracture toughness values KId resp., KJc,d and the determination of the characteristic loading rate in the case of elastic-plastic tests is discussed. This analysis shows the need of modifications of the standard evaluation method in ASTM E1921, if used for tests at elevated loading rate. The assumed statistical distribution has to be adapted and the actual loading rate at fracture initiation has to be taken into account. Suggestions for the considerations and requirements for the evaluation of data determined at elevated loading rates are presented.