Identification of dynamic properties using acceleration by the Virtual Fields Method

Abstract

In this study, the dynamic property at high strain rates was identified using acceleration information by the virtual fields method (VFM) without using load data. Numerical simulations and experiments with two different equipment were conducted and finally the Johnson-Cook parameter which represents strain rate sensitivity was identified by the VFM. Material used for this study was DP980 steel with 1.16 mm thickness. The VFM has an advantage of identifying dynamic property with two approaches. One is with the load data, and the other one is only with the acceleration data. Previous study numerically proved that identifying dynamic property by the VFM is possible using acceleration not with the load data. However, it was not proved experimentally. Therefore, numerical simulations using the ABAQUS, were conducted to figure out the minimum acceleration magnitude. The velocity boundary condition was given in simulations. For the velocity boundary condition, the velocity was extracted from experiment and applied for reproducing the result close to the one from the experiment. Only the total time of deformation was reduced in sequence maintaining the amplitude. Finally the minimum acceleration magnitude needed was identified with the simulation. The split Hopkinson tensile bar (SHTB) and impact frame high speed (IFHS) tester were used to achieve high strain rate and acceleration. Unfortunately, with the SHTB, even though specimen was modified to increase the acceleration magnitude, the acceleration was insufficient. Then, with the IFHS tester, it was modified to increase the acceleration. The diameter of an energy frame bar which is storing strain energy was reduced to increase the acceleration magnitude. Acceleration magnitude increased after the modification, the load and acceleration data were measured during the experiments. With the experiment results, the VFM was applied in two approaches. The results from each, one with the load data and the other with the acceleration data were compared and discussed. Lastly, an additional study regarding the solution of abnormal behavior with the Lim-Huh model observed in a previous study. A certain condition was applied to find proper upper and lower bound value for each parameter of the Lim-Huh model. Parameters were identified with the modified bound values and the results are discussed.