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.

Wrong temperature profile time dependent study

Please login with a confirmed email address before reporting spam

Hi!

I am modeling the temperature distribution inside a layer of biological tissue. The geomtry is a cube. I put a temperature BC of 78°C on one of the face and a temperaure BC of 37°C on the opposite face. The others face are insulated. I am using the mode BIotHeat equation to model the heat removing by the blood circulation, so I put 37°C to the blood too.

I am writing this post because I think my temperature profile is wrong, indeed I obtain a minimum temperature inferior to 37°C inside the cube and it is physically impossible! The temperature inside the cube should be between 78°C and 37°C and the minimum 37°C. But I remarked that the problem disappear when I use Stationnary equation form in the tab Heat Transfer in Solids. I put in attachment the temperature profile of the cube.

It would be great if somebody could help me!!

Thank you!

Have a nice day!

Tim


4 Replies Last Post Jul 24, 2015, 10:53 a.m. EDT
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: 9 years ago Jul 23, 2015, 2:15 a.m. EDT
Hi

As you state a temperature below 37 is not physical, I agree with this, but you have a curve that is typically an overshoot/ringing shape with a "small" error.
Those effects appear typically when you interpolate with higher orders over few spiky points, and/or that you run a FEM model without having respected a minimum mesh quality compliant with the physics/equations and BCs you set up.
You have a very steep gradient to the left, do you have enough mesh elements to resolve that temperature gradient ? are your mesh density, time stepping and the heat diffusivity alpha of your material set up to respect the minimum criteria for correct convergence ?

For the interpolation overshoot, you can set the plot "quality" such to use linear interpolation instead of the default second order, that might reduce the overshoot, but still check carefully your mesh, for diffusion equations (HT and chemistry) the correct resolving of the gradients of the dependent variables is crucial

--
Good luck
Ivar
Hi As you state a temperature below 37 is not physical, I agree with this, but you have a curve that is typically an overshoot/ringing shape with a "small" error. Those effects appear typically when you interpolate with higher orders over few spiky points, and/or that you run a FEM model without having respected a minimum mesh quality compliant with the physics/equations and BCs you set up. You have a very steep gradient to the left, do you have enough mesh elements to resolve that temperature gradient ? are your mesh density, time stepping and the heat diffusivity alpha of your material set up to respect the minimum criteria for correct convergence ? For the interpolation overshoot, you can set the plot "quality" such to use linear interpolation instead of the default second order, that might reduce the overshoot, but still check carefully your mesh, for diffusion equations (HT and chemistry) the correct resolving of the gradients of the dependent variables is crucial -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jul 23, 2015, 7:01 a.m. EDT
Hi!

Thank you a lot for your answer, it is very helpful! The problem was due to the time step too short. I change it and now I have a nice curve with convenient temperature!!

Thank you a lot and have a nice day!

Tim
Hi! Thank you a lot for your answer, it is very helpful! The problem was due to the time step too short. I change it and now I have a nice curve with convenient temperature!! Thank you a lot and have a nice day! Tim

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: 9 years ago Jul 23, 2015, 9:04 a.m. EDT
Hi

Still "time step too short" means you were hitting the

(alpha_th*Delta_time_step < h^2) limit

(h=average mesh size).

Often its worth to double check your results with and without Plotting smoothing, in "raw" mode see "Plot - Quality - Smoothing: none, Resolution: no-refinement"

--
Good luck
Ivar
Hi Still "time step too short" means you were hitting the (alpha_th*Delta_time_step < h^2) limit (h=average mesh size). Often its worth to double check your results with and without Plotting smoothing, in "raw" mode see "Plot - Quality - Smoothing: none, Resolution: no-refinement" -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jul 24, 2015, 10:53 a.m. EDT
Thank you so much for your answer Ivar! It is been few months I am visiting COMSOL forum and your answers are always very helpful!!

Have a good day!
Tim
Thank you so much for your answer Ivar! It is been few months I am visiting COMSOL forum and your answers are always very helpful!! Have a good day! Tim

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.