| About this Abstract |
| Meeting |
2020 TMS Annual Meeting & Exhibition
|
| Symposium
|
Additive Manufacturing: Materials Design and Alloy Development II
|
| Presentation Title |
Rapid Process Parameter Discovery for Functionally Graded Heterogeneous Materials Using Machine Learning and High Throughput Experiments |
| Author(s) |
Behzad Rankouhi, Salman Jahani, Ankur Kumar Agrawal, Gabriel Meric de Bellefon, Dan J Thoma, Frank Pfefferkorn |
| On-Site Speaker (Planned) |
Behzad Rankouhi |
| Abstract Scope |
In this work, we propose a fast and efficient method to discover the suitable process parameters for manufacturing heterogeneous functionally graded materials (FGMs) using selective laser melting (SLM) process. Specifically, a regression model based on a multi-variant Gaussian process is developed to correlate laser power, scan velocity, and hatch spacing with material properties. The training data for the algorithm is collected using a high-throughput experimentation method that allows for rapid measurement of material density, macro-hardness and surface roughness. Laser beam diameter, powder particle size, and layer thickness are kept constant at 80 µm, ≤60 µm, and 20 µm, respectively. Furthermore, a Pareto optimality solution is used to determine the optimal set of process parameters. Finally, a scan strategy is derived to manufacture SS316L-HastelloyX and SS316L-Cu heterogeneous FGMs. Results indicate that heterogeneous FGMs can be manufactured via SLM with relative densities above 95% using the proposed predictive framework. |
| Proceedings Inclusion? |
Planned: Supplemental Proceedings volume |