About this Abstract |
Meeting |
2025 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2025)
|
Symposium
|
2025 Annual International Solid Freeform Fabrication Symposium (SFF Symp 2025)
|
Presentation Title |
Convolutional Neural Network Approach for Predicting Melt Pool Dimensions in Laser Powder Bed Fusion Process |
Author(s) |
Oluwapelumi Oluwaseyi Adejumo, Nayan Pundhir, K Chandrashekhara, Cesar Ortiz Rios, Misak Heath |
On-Site Speaker (Planned) |
Oluwapelumi Oluwaseyi Adejumo |
Abstract Scope |
Laser powder bed fusion (LPBF) is an additive manufacturing process that uses a high-powered laser to selectively melt metal powder and build parts layer by layer. In this study, a computational fluid dynamics (CFD) model that simulates the LPBF process has been developed using Flow-3D software to generate melt pool images. These images were used to train a convolutional neural network (CNN), a deep learning model well-suited for image-based data, to predict the melt pool dimensions. Using the mean absolute error (MAE) and root mean square error (RMSE), the performance metrics for validating the model built were evaluated yielding values of 0.0275 and 0.0371 for MAE and RMSE respectively. The model demonstrated predictive capabilities in estimating accurate melt pool dimensions directly from the melt pool images without relying on process parameters like laser power or scan speed, highlighting its potential for data-driven monitoring and quality control in LPBF processes. |
Proceedings Inclusion? |
Planned: Post-meeting proceedings |