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 |
A Machine Learning Model to Predict Mechanical Property of Directed Energy Deposition Processed Low Alloy Steels |
| Author(s) |
Atiqur Rahman, Md. Hazrat Ali, Muhammad Arif Mahmood, Asad Waqar Malik, Frank Liou |
| On-Site Speaker (Planned) |
Atiqur Rahman |
| Abstract Scope |
Directed Energy Deposition (DED) was developed to manufacture high-performance steel components with intricate geometries. Nonetheless, identifying the complex relationships in alloy composition, cooling rate, thermal history, and mechanical properties remains a significant challenge. Machine learning (ML) algorithms are progressively equipped to address this challenge via interpretable, and data-driven modeling. In this work, an ML-driven model is introduced to predict high-accuracy tensile strength, yield stress, and hardness properties while offering insight into the underlying factors that drive mechanical performance. A dataset comprising 2430 analyses via JMatPro by varying low alloy steel compositions is used to train ML models. Multi-Variable Linear Regression (MVLR) and Polynomial Regression (PR) are then applied to extract critical insights from DED-processed steels. This paper demonstrates that the proposed ML algorithms enhance defect prediction accuracy of DED-processed low alloy steels, offering a valuable pathway to advance metal additive manufacturing. |
| Proceedings Inclusion? |
Planned: Post-meeting proceedings |