In this paper a new loading history for extracting the stress-strain curve as well as the viscosity and creep behavior from indentation experiments is developed. It is based on a simple model describing the viscoplastic spherical indentation with a power-law hardening rule and a velocity-dependent overstress. Using this model, patterns were generated consisting of load-depth data and corresponding material parameters. The loading history for the simulation of the patterns was considered as a variable combination of loading and creep processes. To compare the identification potential of different loading histories, the inverse problem of determining the viscoplastic material parameters was solved by using neural networks. The emerging loading history uses a multiple-creep process with equidistant load steps and allows an identification of material parameters with much higher accuracy than with single creep. It will be used for further work, where the identification method is generalized using more realistic finite element simulations for a finite deformation elastic-viscoplastic material behavior.
展开▼
