Data-Driven Optimization of Wire Arc DED Manufacturing Conditions for Improved Bead Shape Prediction

Описание: Presented by Stephen Price, a graduate student at Worcester Polytechnic Institute and 2024 participant of the ASM Student Speaking Symposium (S3). To learn more about this contest, please visit https://bit.ly/2WXL3WV



Full Presentation Title: Data-Driven Optimization of Wire Arc DED Manufacturing Conditions for Improved Bead Shape Prediction


Combining advancements in additive manufacturing and predictive modeling, this study explores the impact of processing parameters and material selection on Wire Arc Directed Energy Deposition (Wire Arc DED) bead geometry. Through collection and analysis of the largest dataset to date, this research revealed the dominant impact of Feed Rate and Travel Speed on bead dimensions. A higher Feed Rate increases bead size, whereas a higher Travel Speed reduces it. Material choice, while less influential, still notably affects bead shape. This research highlights a novel feature engineering approach to improve model performance. Specifically, with an average error of 2.4% and 1.2% for predicting bead width and height, models trained here are the highest performing on an external validation set. Additionally, through the inclusion of chemical properties, these models are the first that can be used on unseen materials, significantly reducing the costs and time required at the start of new projects.