|About this Abstract
||2022 TMS Annual Meeting & Exhibition
||Cast Shop Technology
||Assessment of a neural network based approach for application as a digital twin in billet DC casting.
||Kjerstin Ellingsen, Mohammed M'Hamdi, Knut Omdal Tveito
|On-Site Speaker (Planned)
Hot-tearing is a severe defect limiting the productivity of DC-casting of aluminum billets. Hot-tearing indicators integrated in process FEM models have been shown to be valuable for assessing hot-tearing tendencies for different casting conditions. However, these FEM tools are not efficient enough for a quick screening and selection of process parameters before casting or adjustments during casting to avoid hot-tears. In this work, we assess the possibility of using an artificial neural network (ANN) to predict sump depth and hot-tearing tendency in the center of billets. A FEM DC-casting model is used to build a database for training and testing the neural network. Configuration of the ANN model is discussed as well as assessment for application as a digital twin in billet DC-casting.
||Planned: Light Metals
||Aluminum, Modeling and Simulation, Machine Learning