Combining Structural Optimization and Process Assurance in Implicit Modelling for Casting Parts

• Published in: Materials Vol. 14; no. 13; p. 3715
• Main Authors: Rosnitschek, Tobias, Erber, Maximilian, Hartmann, Christoph, Volk, Wolfram, Rieg, Frank, Tremmel, Stephan
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 02.07.2021; MDPI
• Subjects: Assurance; Cantilever beams; Casting; Die casting; Geometry; implicit modeling; Investigations; Iterative methods; Knowledge; Manufacturability; Manufacturing; Mechanical properties; Metamodels; Optimization; Performance evaluation; process assurance; Product development; Proposals; Reconstruction; Simulation; Solidification; structural optimization; Topology optimization; virtual product development
• ISSN: 1996-1944; 1996-1944
• DOI: 10.3390/ma14133715

The structural optimization of manufacturable casting parts is still a challenging and time-consuming task. Today, topology optimization is followed by a manual reconstruction of the design proposal and a process assurance simulation to endorse the design proposal. Consequently, this process is iteratively repeated until it reaches a satisfying compromise. This article shows a method to combine structural optimization and process assurance results to generate automatically structure- and process-optimized die casting parts using implicit geometry modeling. Therefore, evaluation criteria are developed to evaluate the current design proposal and qualitatively measure the improvement of manufacturability between two iterations. For testing the proposed method, we use a cantilever beam as an example of proof. The combined iterative method is compared to manual designed parts and a direct optimization approach and evaluated for mechanical performance and manufacturability. The combination of topology optimization (TO) and process assurance (PA) results is automated and shows a significant enhancement to the manual reconstruction of the design proposals. Further, the improvement of manufacturability is better or equivalent to previous work in the field while using less computational effort, which emphasizes the need for suitable metamodels to significantly reduce the effort for process assurance and enable much shorter iteration times.