| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Verification, Calibration, and Validation Approaches in Modeling the Mechanical Performance of Metallic Materials
|
| Presentation Title |
An Open-Source Framework for Data Augmentation and Emulation: Application to Process Optimization in AM |
| Author(s) |
Amin Yousefpour, Sanaz Zanjani Foumani, Ramin Bostanabad |
| On-Site Speaker (Planned) |
Ramin Bostanabad |
| Abstract Scope |
Data-driven methods are increasingly used to design advanced materials by, e.g., optimizing their composition or manufacturing process parameters. In many applications, the success of these methods relies on their abilities to build accurate and interpretable predictive models which can also quantify the uncertainties that ubiquitously arise in materials design (e.g., lack of high-fidelity data and noise). In this presentation, we will introduce a learning framework based on Gaussian processes (GPs) for data fusion (i.e., learning from multiple sources of data), inverse model calibration, and probabilistic metamodeling. This framework is implemented in Python and is freely available on GitHub under the name "GP+". |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Mechanical Properties, Additive Manufacturing |