- Developing a Data Driven Method to Predict Overheating in Powder Bed Fusion
- Project Year:
2023
- REU Student (s):
Ethan Regal | Gannon University PA
- Student 1 Institution:
Gannon University
- Project Mentor:
Weihong 'Grace' Guo
- Project Mentor Area:
Industrial and Systems Engineering
- Project Abstract:
Powder Bed Fusion (PBF) is an extremely lucrative additive manufacturing technique that allows for highly customizable parts, increased resource efficiency, and reduced cost. However, PBF is prone to overheating which results in part defects. One possible solution to this deficiency is to employ a machine learning based, data driven method to predict overheating before it occurs. Such a method must predict process behaviors with high accuracy, provide explanations and clarification for the prediction, and obey physics principles. In this paper, methods of meeting these three conditions are explored, and a resulting technique is proposed.
