Integrated Computational and Experimental Methods for Additive Manufacturing

Abstract

Additive manufacturing (AM) holds the promise of revolutionizing current paradigms in engineering design, manufacturing, and material and structural performance.1-3 However, full realization of the possibilities of metal-based AM will require significant advances in both the process and performance modeling of additively manufactured materials and structures. [4][5][6][7][8] Modeling and simulation of AM processes is particularly challenging, since it involves multiple time and space scales.9 Furthermore, metal-based AM involves multiple physics, and is thus multidisciplinary in nature, requiring interdisciplinary research efforts. Fortunately, the unique manufacturing process of AM affords the possibility of tight integration of in situ measurements and computational modeling.10-12 Significant challenges remain, however, in material/structural qualification and development of process-structureproperties-performance linkages. [13][14][15][16][17] The recent 3rd Sandia Fracture Challenge focused on metal AM is an interesting example of the challenges involved in predicting the performance of AM components, even provided with significant a priori information of the as-built structure. 18With the goal of fostering communication and collaborations between the AM computational and experimental communities, a conference was recently held at the Colorado School of Mines, on September 6-8, 2017. 19 The conference brought together researchers from both the computational mechanics and materials science research communities involved in advancing the state of the art in process and performance modeling of additively manufactured (AM) materials and structures. There were over 31 invited speakers from academia, industry, and national laboratories, organized in a single-track session to foster communication. A central theme of the conference was integration of computational modeling and experiments, leveraging the unique manufacturing aspects of AM, to improve overall modeling predictivity and enable process optimization. This conference was jointly sponsored by the U. While the conference did not have published proceedings, presenters were invited to submit articles for a special topic in JOM. The following three articles have been assembled from this process. The first article by Roehling et al. describes the use of single-track laser-melting experiments to explore process parameters and their effects on microstructure. Experimental results were used to benchmark a phase-field model that simulated the rapid solidification in conditions relevant for AM. The second paper by Lindwall et al. models the creation of bulk metallic glasses using a powder-bed fusion process. The study evaluates computational approaches that can be used to increase the computational speed while maintaining accurate temperatures in critical regions. The accurate prediction of temperature history is particularly important in ensuring that reheated material does not become devitrified. The third paper by Donoso and Peters uses a combined discrete-contin...