About this Abstract |
Meeting |
2022 TMS Annual Meeting & Exhibition
|
Symposium
|
Computational Thermodynamics and Kinetics
|
Presentation Title |
High Performance Cahn-Hilliard Solver for Advanced Microstructure Modeling |
Author(s) |
Kubra Karayagiz, David Montiel, Siamak G. Faal, Marcus Sarkis-Martins, Adam Powell |
On-Site Speaker (Planned) |
Kubra Karayagiz |
Abstract Scope |
A robust solver scheme based on the convexity splitting technique is presented to obtain unconditionally stable time-stepping discretization when solving the Cahn-Hilliard equation with a logarithmic nonlinear free energy. In this approach, the free energy is split into concave and convex parts that are treated explicitly and implicitly, respectively. The numerical implementation is carried out using the finite element method in deal.ii, a C++ software library. The exact same problem is also solved using a fully explicit scheme, serving as a benchmark. The results from 2-dimensional and 3-dimensional simulations are presented. The methods are compared by means of computation time and accuracy. 50 to 200 speed-up is achieved using the explicit-implicit scheme as compared to the fully explicit scheme. A method for splitting free energy function terms in ternary and multicomponent systems is also discussed. |
Proceedings Inclusion? |
Planned: |
Keywords |
Computational Materials Science & Engineering, Modeling and Simulation, |