This presentation was made at CAASE18, The Conference on Advancing Analysis & Simulation in Engineering. CAASE18 brought together the leading visionaries, developers, and practitioners of CAE-related technologies in an open forum, to share experiences, discuss relevant trends, discover common themes, and explore future issues.

Resource Abstract

Powder bed fusion based metal additive manufacturing (AM) processes form parts by successive melting and solidification of metal powder, scan vector by scan vector and layer by layer. Each AM machine manufacturer has developed unique scan strategies, resulting in machine-specific thermal histories in the parts produced on their equipment. In addition, when adding a new geometry into the build file processor software on a machine, the scan vectors that are produced can change based upon part location, orientation, and user parameter selections. As such, a single part design can have significantly different thermal histories depending upon what machine is used to build that part, what orientation the part is built in, and what parameters the user changes for that part.

In order to accurately predict additive manufacturing process effects on a part, that part's unique thermal history should be taken into account. This requires consideration of machine-specific scan patterns. ANSYS/3DSIM can read in scan vectors from machine build file processors to enable this information to be considered in a simulation. Using machine/part specific scan vector information, detailed thermal histories can be predicted. That thermal history can then be used to predict thermally induced strains, distortion and stress, and their combined effects on part accuracy. The same thermal history can also be used to predict part microstructure and part properties based upon that microstructure. And that same thermal history can be post-processed to predict what different types of sensors would measure during fabrication of a part.

This presentation will show an end-to-end simulation architecture for additive manufacturing based upon a combination of technologies. As part of this architecture, users can predict part thermal history, distortion, microstructure, sensor outputs, and properties. These predictions are useful for part design and qualification. An overview of these tools and validation of their predictions will be presented in this talk.