Numerical and experimental simulation of adiabatic shear localization in tantalum and Armco iron

The microstructure, mechanical properties and adiabatic shear localization of tantalum after deformation have been investigated together with those of Armco iron and AISI 1020 steel. Distinct adiabatic shear bands, extensive twins and voids were observed in Armco iron Taylor cylinders and EFPs (explosively formed projectiles). Tantalum was found to be more resistant to the formation of voids and twins and exhibited less strain hardening at high strain rates than the other two materials. Only diffuse shear bands were observed in tantalum although tantalum has the hardness and lower yield strength close to those of Armco iron.; A high strain rate punch device has been designed to produce adiabatic shear bands in AISI 1020 steel, Armco iron and tantalum with the aim to elucidate tantalum's adiabatic shear banding behaviour and investigate the mechanism leading to adiabatic shear banding. By correlating microhardness tests on Armco iron and AISI 1020 steel with experimental results reported in the literature, it was shown that with carbon present the transformed adiabatic shear bands had a higher hardness than that of its tempered martensitic surroundings, while in the carbon free Armco iron the shear bands were softer than the matrix. The hardness differences are due to different microstructural changes at high temperatures during the adiabatic shear banding process. Because of its high melting point and low shock hardening ability only deformation bands were observed in tantalum punch tests. Temperature, strain and strain rate were found to be important variables in the formation of adiabatic shear bands.; A general purpose constitutive equation considering thermal softening, strain hardening and strain rate hardening was developed and applied to tantalum. This equation quantitatively agrees well with experimental results obtained at strain rates from 10{dollar}sp{lcub}-5{rcub}{dollar}/s to 10{dollar}sp5{dollar}/s, temperatures up to 1293K/s and strains up to 1. This equation was shown to be useful in predicting the shapes of deformed Taylor cylinder impact test specimens and strains at maximum loads in tensile tests at strain rates from 10{dollar}sp{lcub}-4{rcub}{dollar}/s to 5 {dollar}times{dollar} 10{dollar}sp3{dollar} and temperatures from {dollar}-{dollar}78K to 293K.; The mechanism for the formation of adiabatic shear bands has been quantitatively verified by a one-dimensional thermo-mechanical model. This model successfully simulated the torsional Hopkinson bar tests reported in the literature. The simulation also showed that heating and cooling rates in adiabatic shear bands can reach 10{dollar}sp7{dollar}- 10{dollar}sp8{dollar}K/s, which explains the microstructural changes and small grain sizes observed in adiabatic shear bands. This simulation and the experiments suggest that the shear localization actually occurs well beyond the maximum stress point. This observation and the proposed physical model of adiabatic shear bands explain the poor agreement between experiments and models based on the maximum stress criterion. Therefore a localization growth criterion was used in the two-dimensional hydrocodes for modeling failure due to adiabatic shear banding. The two-dimensional simulation showed that ordinary quadrilateral elements with one Gaussian integration point and "crossed" triangle elements can capture the formation of shear localization and no special element arrangement is needed for modeling adiabatic shear localization. The only requirement is that the constitutive equation include thermal softening or strain softening. The constitutive equation developed in this study was used in large scale engineering hydrocodes to predict the experimentally observed adiabatic shear localization in punch tests.