About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Aluminum Alloys: Development and Manufacturing
|
| Presentation Title |
Applying Object Detection for In-Situ Predicition of the Internal Temperature Distribution of Inductively Heated Semi-Solid Metal Billets |
| Author(s) |
Marco Speth, Mathias Liewald, Kim Rouven Riedmüller |
| On-Site Speaker (Planned) |
Marco Speth |
| Abstract Scope |
For semi-solid metal forming, metallic billets are usually heated inductively into the two-phase region between solidus and liquidus temperature. A homogeneous distribution of the liquid and solid phase in the billets is thereby crucial for the forming process respectively the manufacturing of high-quality components. Nevertheless, as the skin effect during induction heating leads to a thermal gradient, temperatures need to be determined both in the core and near the outer diameter of the cylindric raw material to achieve optimum heat-up control. Currently, the billet temperatures are measured via thermocouples, which need to be cleaned or even replaced after each heating process due to the high adhesion of the semi-solid metal material. For this reason, an image-based model is presented in this paper, which was first trained with data from various heating trials and, based on this, provides a high prediction quality regarding the liquid phase distribution already during billet heating. |
| Proceedings Inclusion? |
Planned: Light Metals |
| Keywords |
Aluminum, |