Inverse computational method for constitutive parameters obtained from torsion, plane-strain and axisymmetric compression tests

In this work the authors seek to clarify two different requirements for torsion and compression tests in order to solve the problem of identifying constitutive parameters using an inverse computational method. Considerations are restricted in the present research to the proper use of available experimental data in conjunction with an existing constitutive model. For ease of description, a simple power-law constitutive equation has been chosen that can illustrate the different requirements in an elegant manner. First, it will be shown that the determination of a unique set of constitutive parameters during the experimental-computational inverse solution of the torsion test requires either sets of load-deformation and load-deformation-rate data, or a number of load-deformation measurements each with a different constant twist rate. These data are used to define the required cost function. It will also be shown that for the torsion test, the strain hardening index can be found accurately using a single set of load-deformation data. A new cost function based on load-deformation-rate data will be introduced for the torsion test for determining accurately the strain rate index. The load-deformation-based cost function will be used subsequently in conjunction with the new cost function for proper determination of all parameters during the inverse solution of the torsion test. It will be shown further that for the case of plane strain and axisymmetric compression tests, one set of load-deformation data alone suffices to identify the constitutive parameters uniquely. © 1998 Elsevier Science S.A. All rights reserved.