Computational framework for microstructure identification and hardness prediction in ultrahigh-strength steel components fabricated by wire-arc directed energy deposition

This work predicts multiple reheat thermal histories and resulting microstructure and hardness of wire-arc directed energy deposition (WA-DED) fabricated components through development and validation of a computational modeling framework. The framework combines a finite element analysis (FEA) process model, a calibrated dual double ellipsoid Gaussian (DDEG) heat source, a thermal history-based microstructure identification procedure, and thermal history-property relationships. The framework was demonstrated and validated through accurate predictions of measured thermal histories, characterized microstructures, and measured hardness of a single bead per layer ultrahigh-strength steel WA-DED build. It was predicted that WA-DED process generates a transient temperature field along the build height. Deposition of the final layer remelts the previously deposited layer, forming as-solidified fresh martensite in the top two layers. The layer below is supercritically reheated, forming an affected zone (HAZ) with coarse- and fine-grained fresh martensite. The fresh martensite in the top three layers has hardness of 500–530 HV 0.5 . The material below the three top layers is tempered by subcritical reheats. The latter generate cyclic hardness variation along the build height from 370 HV 0.5 , in tempered fine- and coarse-grained HAZ regions, to 450 HV 0.5 in tempered intercritically reheated coarse-grained HAZ regions. This work represents the first application of a calibrated DDEG heat source in a WA-DED FEA process model, thermal history-based HAZ microstructure identification, and HAZ-specific tempering response relationships, for microstructure and hardness prediction in a WA-DED fabricated build. The demonstrated computational framework provides a basis for component property optimization by predicting the distribution of microstructures and hardness as a function of WA-DED process parameters.