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.

moving boundary

Please login with a confirmed email address before reporting spam

hi
i need to use operators 'up' and 'down' to calculate the normal heat flux in the two sides of a moving boundary.
i tried the expressions

down(fluxx_ht)*dnx+down(fluxy_ht)*dny
and
up(fluxx_ht)*dnx+up(fluxy_ht)*dny

but it gives an error :
Failed to evaluate variable jacobian
- Variable: flux_ht
- Geometry: 1
- Boundary: 7

note that i want apply these expression for the boundary n4 which is the moving boundary.
thank u for help.

4 Replies Last Post Feb 19, 2010, 8:20 a.m. EST

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Feb 15, 2010, 4:50 a.m. EST
Hi,

I think your problem is that the variable fluxx_ht is only defined on the subdomain, not on the boundary. I have a couple of suggestions that you might look at though, if you want to get a the two heat fluxes. First is to define two heat transfer modes (or use an assembly), and set the temperature equal at the boundary. You might need to use a non-ideal weak constraint though to avoid the flux from being coupled naturally as the reaction force.

Do you define the temperature at the boundary (Dirichlet type)? If so, then look up calculating the flux from the Lagrange multiplier (weak constraint). Since this is an internal boundary, the lm is calculated as the difference in heat fluxes normal to the boundary. Plus you get (supposedly) better accuracy in the flux calculations. I use this for the Stefan model of melting.

Hope this helps!
Hi, I think your problem is that the variable fluxx_ht is only defined on the subdomain, not on the boundary. I have a couple of suggestions that you might look at though, if you want to get a the two heat fluxes. First is to define two heat transfer modes (or use an assembly), and set the temperature equal at the boundary. You might need to use a non-ideal weak constraint though to avoid the flux from being coupled naturally as the reaction force. Do you define the temperature at the boundary (Dirichlet type)? If so, then look up calculating the flux from the Lagrange multiplier (weak constraint). Since this is an internal boundary, the lm is calculated as the difference in heat fluxes normal to the boundary. Plus you get (supposedly) better accuracy in the flux calculations. I use this for the Stefan model of melting. Hope this helps!

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Feb 15, 2010, 7:15 a.m. EST
hi thank you for asking,
in fact i defined two heat transfer modes, conduction in solid and convection in liquid, and i set the temperature equal at the interface = equilibrium temperature, and i used non-ideal weak constraint.
since you have already used the stefan model of melting, can u please tell me the melting rate expression that u used.
thank u
hi thank you for asking, in fact i defined two heat transfer modes, conduction in solid and convection in liquid, and i set the temperature equal at the interface = equilibrium temperature, and i used non-ideal weak constraint. since you have already used the stefan model of melting, can u please tell me the melting rate expression that u used. thank u

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Feb 19, 2010, 4:38 a.m. EST
Hi again, it's basically a heat balance across the boundary. The way I think about it is that the difference between the heat going into the boundary and coming out of it, is stored as the latent heat associated with the solid melting (ie: the boundary moving). The expression is then

Jin-Jout=Hfus*R (minding your units of course)

Now it gets a lot more fun if you try to add non-congruent melting. There are more details here if you want: www.mikewelland.com/publications/journalarticle/Welland_CSONCMOHUD_manuscript.pdf
Hi again, it's basically a heat balance across the boundary. The way I think about it is that the difference between the heat going into the boundary and coming out of it, is stored as the latent heat associated with the solid melting (ie: the boundary moving). The expression is then Jin-Jout=Hfus*R (minding your units of course) Now it gets a lot more fun if you try to add non-congruent melting. There are more details here if you want: http://www.mikewelland.com/publications/journalarticle/Welland_CSONCMOHUD_manuscript.pdf

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Feb 19, 2010, 8:20 a.m. EST
Thank you very much for informations and documentation.
Thank you very much for informations and documentation.

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.