|Title||Assessment of grain structure evolution with resonant ultrasound spectroscopy in additively manufactured nickel alloys|
|Publication Type||Journal Article|
|Year of Publication||2020|
|Authors||Rossin J, Goodlet B, Torbet C, Musinski W, Cox M, Miller J, Groeber M, Mayes A, Biedermann E, Smith S, Daly S, Pollock T|
|Keywords||Additive manufacturing, Bayesian inference, Elastic constant inversion, Finite element modeling, Recrystallization, Resonance ultrasound spectroscopy|
Despite the advantages of metal additive manufacturing (AM), ensuring integrity and reproducibility for built components is a barrier to the implementation of AM components in critical applications. Component qualification necessitates Non-Destructive Evaluation (NDE), but existing NDE frameworks are insufficient for the rapid and cost effective screening of variable AM components. In this study, alterations in laser powder bed fusion (LPBF) AM process parameters were characterized using resonant ultrasound spectroscopy (RUS). Samples that were subjected to a Hot Isostatic Press (HIP) and Heat Treatment (HT) post-processing step exhibited changes in resonance frequencies that varied in magnitude and direction with the type of resonance mode. The initial build direction prior to HIP and HT had a negligible effect on resonant frequency changes after recrystallization. The change in resonant frequencies at each process condition was predicted using Finite Element Modeling (FEM) informed with Electron Backscatter Diffraction (EBSD) data. FEM identified that the experimentally measured change in resonant response between the initially textured state and the recrystallized state was dominated by grain orientation dependent changes in elasticity. The EBSD-estimated elastic constants and FEM results were validated using experimental laser vibrometry and RUS inversion of elastic constants. RUS Inversion by Bayesian inference and Hamiltonian Monte Carlo has not been used to characterize an AM nickel-base alloy prior to this work. These results demonstrate that RUS is capable of detecting part to part microstructure variability between built AM components.