| description abstract | Metals-additive manufacturing (MAM) is enabling unprecedented design freedom and the ability to produce significantly lighter weight parts with the same performance, offering the possibility of significant environmental and economic benefits in many different industries. However, the total production costs of MAM will need to be reduced substantially before it will be widely adopted across the manufacturing sector. Current topology optimization approaches focus on reducing total material volume as a means of reducing material costs, but they do not account for other production costs that are influenced by a part's structure such as machine time and scrap. Moreover, concurrently optimizing MAM process variables with a part's structure has the potential to further reduce production costs. This paper demonstrates an approach to use process-based cost modeling (PBCM) in MAM topology optimization to minimize total production costs, including material, labor, energy, and machine costs, using cost estimates from industrial MAM operations. The approach is demonstrated on various 3D geometries for the electron beam melting (EBM) process with Ti64 material. Concurrent optimization of the part structures and EBM process variables is compared to sequential optimization, and to optimization of the structure alone. The results indicate that, once process variables are considered concurrently, more cost effective results can be obtained with similar amount of material through a combination of (1) building high stress regions with lower power values to obtain larger yield strength and (2) increasing the power elsewhere to reduce the number of passes required, thereby reducing build time. In our case studies, concurrent optimization of the part's structure and MAM process parameters lead to up to 15% lower estimated total production costs and 21% faster build time than optimizing the part's structure alone. | |
