Optimizing material strength constants numerically extracted from taylor impact data

Advanced design requirements have dictated a need for the mechanical properties of materials at high strain rates. Mechanical testing for these data poses a significant problem for experimentalists. High-speed testing machines have a limited capability at rates approaching 10²/s. The split Hopkinson pressure bar is the most reliable alternative for rates approaching 10⁴/s. Plate impact experiments are capable of generating strain rates of 10⁸/s and higher. The Taylor impact test occupies a place of particular importance by providing data at strain rates on the order of 10⁴/s–10⁵/s. The issue at present is extracting the data. This paper provides a method for obtaining dynamic strength model material constants from a single Taylor impact test. A polynomial response surface is used to describe the volume difference (error) between the deformed specimen from the Taylor test and the results of a computer simulation. The volume difference can be minimized using an optimizer, with the result being an optimum set of material constants. This method was applied to the modified Johnson-Cook model for OFHC copper. Starting from a nominal set of material constants, the iterative process improved the relative volume difference from 23.1 percent to 4.5 percent. Other starting points were used that yielded similar results. The material constants were validated by comparing numerical results with Taylor tests of cylinders having varying aspect ratios, calibers and impact velocities.