Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

Convergence issues for transient freezing (apparent heat capacity phase change)

Please login with a confirmed email address before reporting spam

Hello,

I am fairly new to COMSOL and am trying to solve a transient (over 600 seconds) freezing (phase change) problem in a gel (similar thermal properties to water). I am using temperature-dependent thermal conductivity and density (both implemented as cubic-smoothed lookup functions, f(T) ). I am simulating the phase-change with the apparent heat capacity method (also implemented as a cubic-smoothed lookup function, f(T) ).

Across the model, my temperatures vary from 140K to 300K, with fairly high temperature gradients near the cooling interface (order of 40K/mm).

I have a 2D geometry with phase change occurring over <1K, that solves without issue. However, I have a slightly more complicated 3D geometry that I am having problems converging. I can get the solution to converge, treating the phase change as a triangular function over >30K, but the solver gets hung up for any more aggressive phase change.

I would like to reduce the temperature duration over which my phase change is implemented (less than 10K). I am assuming a triangular “pulse” will provide the most gradual change for specific heat, but is this an incorrect assumption and will another function provide less of an issue for convergence? Is there any other settings I can look at to help with convergence? I am using the default settings for the transient solver.

Any assistance you can offer would be greatly appreciated.

Thanks!
Michael

1 Reply Last Post 15 mag 2012, 15:59 GMT-4
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 15 mag 2012, 15:59 GMT-4
Hi

I would rather propose to use a smooth step() or rect() function, than a sharp triangular one, as it is continuously derivable (or at least 2x)
And then probably tweak the solver to try a strict time stepping, with many small steps


--
Good luck
Ivar
Hi I would rather propose to use a smooth step() or rect() function, than a sharp triangular one, as it is continuously derivable (or at least 2x) And then probably tweak the solver to try a strict time stepping, with many small steps -- Good luck Ivar

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.