The data points available for developing a Fracture Locus (FL) database for structural steel used in the Indian construction industry are extremely limited. The current study is conducted to determine the FL data points for three different grades of Indian structural steel, namely E250, E350, and E450. Uniaxial tests on notched dog bone specimens of three different specimen configurations are performed. The selected configuration is used to determine points for plotting FL corresponding to high-stress triaxiality (0.7 < T >1) and Lode angle (L) almost equal to 1. With the help of numerical simulation, the FL points are obtained and reported. The accuracy of numerical simulation is checked by precisely matching the load versus displacement obtained from the experiment. Six fracture prediction models are chosen for the present study. These six models are chosen using the following criteria, (1) only depends on stress triaxiality (b) depends on both stress triaxiality and Lode angle and (c) the number of coefficients used. The effectiveness of all the selected models in predicting fracture initiation across all three steel grades is compared, and the findings are reported.
EXTENDED ABSTRACT