| About this Abstract |
| Meeting |
2023 TMS Annual Meeting & Exhibition
|
| Symposium
|
Additive Manufacturing: Length-Scale Phenomena in Mechanical Response
|
| Presentation Title |
Toward Developing Processing-Microstructure-Property Prediction to Enable Digital Twins of Additive Manufacturing Process |
| Author(s) |
Mohsen Taheri Andani, Veera Sundararaghavan, Amit Misra |
| On-Site Speaker (Planned) |
Mohsen Taheri Andani |
| Abstract Scope |
Metal additive manufacturing (AM) promises to benefit significantly from using digital twins (DTs). This is due to the chaotic nature of the AM process, which results in poor reproducibility. Nonetheless, a DT in a supervisory capacity may inject assurance into the process by actively enforcing the process's boundaries with real-time control instructions. Developing physics- and data-driven methodologies for investigating processing-structure-property interactions is a crucial step in developing the DT for AM components. This paper presents an unsupervised machine learning technique for predicting the relationship between manufacturing process parameters, such as laser scan strategy and building orientation, microstructure features, such as crystallographic texture and grain boundary type, and the mechanical performance of metal AM components. The approach uses sensor information from the printer to improve upon thermal cycling predictions, leading to control of microstructure. Applications of the method to controlling texture, grain boundary character and phases are presented. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Additive Manufacturing, ICME, |