Data-Driven Framework for Mechanical Behavior Characterization from Instrumented Indentation

This study developed a novel data-driven indentation computation framework to characterize the indentation response and behavior by combining the reduced-order model (ROM), optimal technology, and indentation response curve. ROM was utilized to build a surrogated model to approximate the accuracy of the indentation response. The dataset was generated using the indentation test or physical model for ROM. Simplicial Homology Global Optimization (SHGO) was considered an optimal technology for searching for mechanical properties in inverse analysis. The developed framework was illustrated and verified using a numerical example. The results were compared with the actual value obtained by an indentation test. The results show that the developed framework characterizes the material mechanical property and indentation response well and agrees with the engineering practice. The proposed framework provides a feasible, scientific, helpful, and promising way to capture the material mechanical behavior and indentation response. Meanwhile, it also has essential reference significance for another engineering field.