Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
CFD time stepping: CFL vs implicit
Posted Sep 2, 2010, 7:48 a.m. EDT Fluid & Heat, Computational Fluid Dynamics (CFD) 2 Replies
Please login with a confirmed email address before reporting spam
Hi,
i'm struggling on which time stepping method for transient fluid-thermal-interaction i should use. (Comsol V3.5a)
The default setting for the stepping for the transient solver is the following:
Time steps taken by solver: "Manual"
Time Step: "5*dt_cfl_g1_chns"
So the time stepping is explicit based on the Courant Number. This results - in my case - to a time step size of about 0.0033s which is fairly low.
On the other hand, you can select time steps taken by the solver: "Free". This means implicit time stepping (?). This runs quite well and i get convergence.
So my question is: what is best: Time stepping based on "5*dt_cfl_g1_chns" or "Free"?
Any ideas? Suggestions? Doubts? Experience?
Thanks in advance!
i'm struggling on which time stepping method for transient fluid-thermal-interaction i should use. (Comsol V3.5a)
The default setting for the stepping for the transient solver is the following:
Time steps taken by solver: "Manual"
Time Step: "5*dt_cfl_g1_chns"
So the time stepping is explicit based on the Courant Number. This results - in my case - to a time step size of about 0.0033s which is fairly low.
On the other hand, you can select time steps taken by the solver: "Free". This means implicit time stepping (?). This runs quite well and i get convergence.
So my question is: what is best: Time stepping based on "5*dt_cfl_g1_chns" or "Free"?
Any ideas? Suggestions? Doubts? Experience?
Thanks in advance!
2 Replies Last Post Sep 7, 2010, 7:31 a.m. EDT