Hi, Can anyone help me out how to solve a thermo-fluid-structure interaction problem in COMSOL. Which modules need to be selected and how should I set the boundary conditions and domain settings? Is their any coupling variabe needs to be used? Please reply soon. Regards,

krish

Krish, At what stage are you? Have you gotten both geometries into COMSOL such that you can select boundaries (faces)?

you'll need the multiphysics module. Under this you'll find "conduction/convection".

Before I stopped using COMSOL, I remember the setup to have 2 domains. One you would setup as a water "body" and the other a solid. That setup should be fairly simple dependant on what kind of fluid and material. Boundary conditions will depend on your problem. Do you have anymore information I could use? You'll need coupling variables. I could never get a good answer on this subject, however. Thats all taken care of in the $$$ training sessions. -M

Hi Wooster, Thanks for your response. I am trying to simulate a contact problem in a pipeline carrying high pressure hot steam of around 5 bar and 400 degC. I have taken an axisymmetric case where two splitted pipes are connected by a small connector by some thread joint. So there will be a thermal expansion due to conduction and convective heat flow will be there also as the inner steam is flowing inside. I have modeled the domains, but finding difficulty in choosing the application modes. Sometimes it is giving duplicate variable, sometimes it is giving improper initial guess also. I am not able to fix those problems. I was tried various combinations of application modes like, thermo-structural, thermo-fluid, fluid structural etc. please give me your suggestions. regards,

Krish

hmm... What do you mean by duplicate variable?

Also, check your inlet conditions units. COMSOL is horrible about figuring units and sometimes what may seem to be off is merely some unit change. As for conduction/convection, make sure your convective heat transfer coefficients are correct (material properties I think) and thermal expansion properties are sound.

In the end (and I'll check with my colleague who still uses it) we had to scratch the convection/conduction model and just apply a conduction model with convective heat transfer coefficients at the boundary.

To be honest, I've never dealt with steam interactions and thermal expansion in COMSOL, so take what I say with a bit of salt. I'll ask the people who still use it if they have any ideas though.

-M

Thnx. Please check it with your colleagues and suggest me some way out.

regards

krish

Sorry it took so long.

I've spoken with the girl who does our COMSOL stuff and she was out of ideas. The best thing I can think to do is apply some expected heat to the material surface. That's about what we do except we've managed to input a convective heat transfer coefficient in the cool tubes to mimic the effect of water.

Even after training classes (with support) we were unable to get the models working unless we did the above.

Very Sorry -M

no problem. thank u very much to think over my problem. i've done it same way as u've suggested. i took it as only thermo-structural coupling. regards,

krish