NUMERICAL STUDY OF THE INFLUENCE OF THE SPECIMEN GEOMETRY ON SPLIT HOPKINSON BAR TENSILE TEST RESULTS

Abstract

FINITE ELEMENT SIMULATIONS OF HIGH STRAIN RATE TENSILE EXPERIMENTS ON SHEET MATERIALS USING DIFFERENT SPECIMEN GEOMETRIES ARE PRESENTED. THE SIMULATIONS COMPLEMENT AN EXPERIMENTAL STUDY, USING A SPLIT HOPKINSON TENSILE BAR SET-UP, COUPLED WITH A FULL-FIELD DEFORMATION MEASUREMENT DEVICE. THE SIMULATIONS GIVE DETAILED INFORMATION ON THE STRESS STATE. DUE TO THE SMALL SIZE OF THE SPECIMENS AND THE WAY THEY ARE CONNECTED TO THE TEST DEVICE, NON-AXIAL STRESSES DEVELOP DURING LOADING. THESE STRESS COMPONENTS ARE COMMONLY NEGLECTED, BUT, AS WILL BE SHOWN, HAVE A DISTINCT INFLUENCE ON THE SPECIMEN BEHAVIOUR AND THE STRESS-STRAIN CURVE EXTRACTED FROM THE EXPERIMENT. THE VALIDITY OF THE BASIC ASSUMPTIONS OF HOPKINSON EXPERIMENTS IS INVESTIGATED: THE UNIAXIALITY OF THE STRESS STATE, THE HOMOGENEITY OF THE STRAIN AND THE NEGLIGIBLENESS OF THE DEFORMATION OF THE TRANSITION ZONES. THE INFLUENCE OF DEVIATIONS FROM THESE ASSUMPTIONS ON THE MATERIAL BEHAVIOUR EXTRACTED FROM A HOPKINSON EXPERIMENT IS DISCUSSED.