| About this Abstract |
| Meeting |
MS&T23: Materials Science & Technology
|
| Symposium
|
Additive Manufacturing Modeling, Simulation, and Machine Learning: Microstructure, Mechanics, and Process
|
| Presentation Title |
Predicting Material Properties in Additive Manufacturing Using Acoustic Signatures and Machine Learning |
| Author(s) |
Alec S. Mangan, Dan J. Thoma |
| On-Site Speaker (Planned) |
Alec S. Mangan |
| Abstract Scope |
This presentation will explore how acoustic signatures of Laser Powder Bed Fusion (LPBF) additively manufactured CoCrFeMnNi High Entropy Alloy (HEA) can help predict microstructural and material properties of a final part. Through high throughput experiments, a processing region was identified for an equimolar CoCrFeMnNi HEA wherein all processing conditions produced high density and high hardness samples. All samples in this region were nominally defect-free, but microstructural signatures and tensile properties had noticeable variations across the processing region. The acoustic signature of each processing condition was recorded in the 2-16 kHz range using commercial off-the-shelf equipment. These acoustic signatures were then used to train a machine learning algorithm to identify material microstructural signatures of the final parts. The results of this experiment and how this process can be incorporated into future high throughput experiments will be presented. |