UFAST: A novel Spectral Graph method for thermal simulation in Additive Manufacturing.

N. Piercy, L. Castro, P. Rao
University of Nebraska-Lincoln,
United States

Keywords: Additive manufacturing, simulation, thermal history, FEA

Summary:

We present a novel approach to model the thermal history of additively manufactured parts utilizing Spectral Graph Theory which reduces the computational cost when compared with Finite Element Analysis (FEA) software. In Additive Manufacturing (AM), thermal history dictates material characteristics, in-process flaws, and final part shape. By simulating the thermal history, we can predict potential design and manufacturing issues before a part is printed. This, in turn, can reduce trial and error during the development of AM parts resulting in significant cost and time savings. The approach is validated by comparing the simulated thermal history results with experimental data across multiple geometries, materials, and processing conditions.