About this Abstract |
| Meeting |
2027 TMS Annual Meeting & Exhibition
|
| Symposium
|
Advances in Ceramic Materials and Processing
|
| Presentation Title |
Complexity and Cost of Chemical Mechanisms in Modeling Silicon Carbide Syntheses |
| Author(s) |
Vimal Ramanuj, Pei Zhang, Reetesh Ranjan, Jamie Wilson, Luke Riggs, Ramanan Sankaran, Murat Barisik |
| On-Site Speaker (Planned) |
Vimal Ramanuj |
| Abstract Scope |
Silicon carbide composites synthesis is limited by slow deposition for highly dense parts. The goal of our work is to investigate characteristics of the deposition mechanism followed by exploring reduction strategies for potential adoption in computational studies. We present simulation results employing mechanisms of varying complexity. Beyond physical capabilities and limitations, we emphasize simulation cost to evaluate practicality. We begin with a detailed mechanism including gas phase and heterogeneous reactions. Next, mechanism complexity is reduced based on relative magnitudes of reaction rates. Third is a globally reduced description characterized by a single activation energy: the most widely used in simulations. Finally, we investigate a machine learning based reduction approach. Principal component analysis reduces the dimension (analogous to number of species) followed by a neural network reconstruction of state. Minimum dimensions for adequate accuracy are sought. We summarize by outlining key differences between these and potential use cases for each. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Ceramics, Other, Modeling and Simulation |