| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Advanced Real Time Imaging
|
| Presentation Title |
Study of Cracking Initiation and Evolution Dynamics with In-situ X-ray Imaging and Automated Machine Learning |
| Author(s) |
M Matsive Ali , Nick Calta, Christopher Tassone, Sen Liu |
| On-Site Speaker (Planned) |
M Matsive Ali |
| Abstract Scope |
Laser powder bed fusion (LPBF) process has made in-situ alloying of advanced materials with rapid speed and lower cost than traditional design process. However, it is difficult to print crack-free Al alloy due to fast cooling rate. To understand the cracking formation during LPBF, we have utilized high-speed synchrotron X-ray imaging to characterize LPBF Al alloy at sub-ms and µm resolution. In-situ X-ray video along with automated image processing are used to quantify crack initiation and propagation dynamics. The cracks are distinguished from other types of defects with machine learning (ML). The effects of different laser power and scan speeds on crack formation are discussed. The output of cracking segmentation gives insight into how a higher cooling rate leads to residual stress accumulation that can drive crack formation. The efficient ML-based quantification give an insight into crack evolution physics and how to suppress the crack formation with optimized process parameters. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Additive Manufacturing, Characterization, Machine Learning |