Resumen:
To this end, full-field mesoscale computational modeling of statistical volume elements and calibration and validation experiments provide a pathway for establishing microstructure-property relationships in additively manufactured parts. However, the fidelity of these relationships is sensitive to the microstructure representation and grain scale constitutive model, especially in complex geometries, and must be adequately characterized and understood. In this talk, I will present results from two projects highlighting existing challenges, recent progress, and future directions for research in this field. Part I of this talk demonstrates our approach to choosing an optimized combination of microstructure and topology that works efficiently for a specific loading condition in lattice structures. The distinguishing feature of our approach is that the topological optimization is performed while accounting for the heterogeneous distribution of strut-level microstructure concomitant mechanical behavior leading to peak lattice structural performance. Part II examines the size effect in AM thin wall structures from room temperature to elevated loading conditions. We specifically investigate the anomalous high-temperature size effect in Nickel superalloys. I will also review the machine learning – differential evolution (ML-DE) crystal plasticity (CP) framework that we developed to accurately interpolate the tensile properties of AM L-PBF Haynes-214 alloy across a wide temperature range from ambient to 1000oC, with minimal reliance on experimental data.
