Enhanced Characterization of Elastoplastic Properties of Additively Manufactured Specimens Using Stochastic Inverse Modeling of Indentation Data

Rapid characterization of mechanical properties and material structure of additively manufactured (AM) components via non-destructive techniques (NDT) is crucial for their wider adoption. However, accurate characterization of AM components using NDT remains a challenge. To address this, our work focuses on characterizing the elastoplastic properties of AM components from instrumented indentation measurements, addressing the inverse indentation problem. Previous approaches to this problem have limitations in generalization and in estimating the variability of elastoplastic properties. In this work, we explore a stochastic inverse problem (SIP) formulation, estimating a distribution over elastoplastic properties—Young’s modulus (E), yield strength (σy), and strain hardening exponent (n)—that aligns with observed indentation data. By implementing this methodology for AM components subjected to different heat treatments, we achieve predictions of n, E, and σy within 1.1%, 1%, and 5% of the actual values, respectively. The recovered distributions closely match those obtained from standard tensile tests, indicating our methodology’s accuracy in characterizing mean elastoplastic properties and their variability through high-throughput indentation measurements.