I have grain boundary (GB) disorientation based anisotropic GB energy in a Phase-Field coupled Crystal Plasticity based simulation in Open Phase to perform a uni-axial tensile test on a polycrystal of a pure FCC metal. The grain boundary varies between 0.05 - 0.56 Jm-2. A time step for the average GB energy did not work in the present case.
How the time step should be selected so that a stable simulation can be performed?
Dear Waseem Amin,
In general, you need to consider this: in every time step, the phase-field increment shall not be more than 0.1, i.e. dt*phidot < 0.1; best would be that you set the maximum increment to this condition so that you make sure it is then true all across the system; In OpenPhase you can check/output this numerically in the code to make sure that this is the case. Now, based on the double-obstacle potential that we worked out into the OpenPhase for a given GB with given mobility (\mu), energy (\sigma), and curvature (\kappa), the increment is as much as \mu*\sigma*\kappa*dt that must be set < 0.1. Note that if there are other driving forces (like plastic activities) those also need to be added to find the final phase-field increment. Since you have anisotropic energies, then logically the highest value of energy guarantees the safe value for dt to be chosen.
I hope these could be of help.
- Badges
- Science topic