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.
Solvers in COMSOL
Posted Sep 25, 2016, 6:52 a.m. EDT Fluid & Heat, Computational Fluid Dynamics (CFD), Results & Visualization, Studies & Solvers 2 Replies
Please login with a confirmed email address before reporting spam
Hello everyone!
I am a newbie who seeks help in COMSOL.
I have simulated a laminar flow in an annulus using direct solvers. The results match well with the problem (taken from Bird Stewart Lightfoot). I am trying to observe if changing the solver I have used will bring about any change in simulation time, easier convergence,etc. However, I am not sure where to look for these details.
Could you help me find where any changes, due to changing solvers, may be possible?
Also, I find a convergence plot laid out while using DIRECT solvers. It shouldn't be there, right? I have read that the solution is generated 'at once' using direct solvers using matrix method.
Thank you for your time.
--
Raja Barnwal
2 Replies Last Post Sep 28, 2016, 12:01 p.m. EDT
Please login with a confirmed email address before reporting spam
Posted:
8 years ago
The COMSOL blogs is a good place to start, for example this:
www.comsol.com/blogs/solutions-linear-systems-equations-direct-iterative-solvers/
The direct solver is used as a linear system solver. Navier-Stokes is a nonlinear problem per se. As such a nonlinear method is used around each linear step. That is the convergence progress you see in the graph.
Niklas
www.comsol.com/blogs/solutions-linear-systems-equations-direct-iterative-solvers/
The direct solver is used as a linear system solver. Navier-Stokes is a nonlinear problem per se. As such a nonlinear method is used around each linear step. That is the convergence progress you see in the graph.
Niklas
Hello everyone!
I am a newbie who seeks help in COMSOL.
I have simulated a laminar flow in an annulus using direct solvers. The results match well with the problem (taken from Bird Stewart Lightfoot). I am trying to observe if changing the solver I have used will bring about any change in simulation time, easier convergence,etc. However, I am not sure where to look for these details.
Could you help me find where any changes, due to changing solvers, may be possible?
Also, I find a convergence plot laid out while using DIRECT solvers. It shouldn't be there, right? I have read that the solution is generated 'at once' using direct solvers using matrix method.
Thank you for your time.
--
Raja Barnwal
Please login with a confirmed email address before reporting spam
Posted:
8 years ago
Thank you, Niklas.
You mentioned, "As such a nonlinear method is used around each linear step". This means linear and nonlinear step work together to generate the solution. Right? I have gone through the link you suggested. Could you throw some more light in this regard? Maybe some other comprehensive source?
--
Raja Barnwal
You mentioned, "As such a nonlinear method is used around each linear step". This means linear and nonlinear step work together to generate the solution. Right? I have gone through the link you suggested. Could you throw some more light in this regard? Maybe some other comprehensive source?
--
Raja Barnwal