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 The information provided may be out of date.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.

"Normal" Reynolds number ; Constant temperature through tubes

Please login with a confirmed email address before reporting spam

Hello

Model can be found: www.dropbox.com/sh/mvgwp1qbc0dcwa6/YhJhuLItkj

My model of a shell and tube heat exchanger has a constant temperature-profile.

It's possible that steady state inside the model is reached and overall temperature difference is made.
But this is not what I wanted of course.

How to solve this?

Thanks,
Adriaan De Bolle


6 Replies Last Post 27.02.2014, 10:59 GMT-5

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 24.02.2014, 09:48 GMT-5
Maybe a picture will be better, if dropbox wouldn't work...
Fluid is flowing through 3 tubes and over it. There should be heat transfer though the fluid and solid tubewalls.

Everything stays at constant beginning temperature.
Maybe a picture will be better, if dropbox wouldn't work... Fluid is flowing through 3 tubes and over it. There should be heat transfer though the fluid and solid tubewalls. Everything stays at constant beginning temperature.


Frank van Gool COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 25.02.2014, 04:22 GMT-5
Dear Adriaan,

You are solving two heat transfer interfaces, nitf and ht. The first computes the temperature "T", the second "T2". These have no influence on eachother when they are not manually coupled.

The easiest solution is to remove the heat transfer in solids interface, and add a "heat transfer in solids" node in the non-isothermal flow interface. This makes sure that all is computed in one interface.

I would propose to take a look at the "shell-and-tube heat exchanger" in the model library.

Best regards,
Frank
Dear Adriaan, You are solving two heat transfer interfaces, nitf and ht. The first computes the temperature "T", the second "T2". These have no influence on eachother when they are not manually coupled. The easiest solution is to remove the heat transfer in solids interface, and add a "heat transfer in solids" node in the non-isothermal flow interface. This makes sure that all is computed in one interface. I would propose to take a look at the "shell-and-tube heat exchanger" in the model library. Best regards, Frank

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 25.02.2014, 09:37 GMT-5
Hello Frank,

Thanks for tip! I tried it before, but I think my problem then was that I just updated the solution and didn't re-solve it. With the all-in-one module "conjugate" I already got better results.

Thanks,

Kind regards,
Adriaan
Hello Frank, Thanks for tip! I tried it before, but I think my problem then was that I just updated the solution and didn't re-solve it. With the all-in-one module "conjugate" I already got better results. Thanks, Kind regards, Adriaan

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 25.02.2014, 21:10 GMT-5
I updated the model with correct computing of results.

How can I determine the "normal" Reynolds number for my case?

Comsol gives the cell reynolds number depending on the mesh element size. This gives values much lower than theoretical Reynolds values.

In fact the heat flux looks also too small: nitf2.ntflux.

Thanks everyone (also for previous problem)
I updated the model with correct computing of results. How can I determine the "normal" Reynolds number for my case? Comsol gives the cell reynolds number depending on the mesh element size. This gives values much lower than theoretical Reynolds values. In fact the heat flux looks also too small: nitf2.ntflux. Thanks everyone (also for previous problem)

Frank van Gool COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 27.02.2014, 08:14 GMT-5
Dear Adriaan,
The overall Reynolds number should be computed by hand, as the dimension for the input is not known to COMSOL. You can use the "measure" button in the geometry tab to get a quick idea of dimensions in your model.

Regarding the heat flux, make sure that your physics are correctly defined. It can be a good idea to create a 2D (axi) model to get an idea on what to expect. And this also helps in getting a first approximate result.

If you cannot find the mistake yourself, you may post the model, and also your expected results (and why you expect these). I (or someone else) can take a look at it.

Best regards,
Frank
Dear Adriaan, The overall Reynolds number should be computed by hand, as the dimension for the input is not known to COMSOL. You can use the "measure" button in the geometry tab to get a quick idea of dimensions in your model. Regarding the heat flux, make sure that your physics are correctly defined. It can be a good idea to create a 2D (axi) model to get an idea on what to expect. And this also helps in getting a first approximate result. If you cannot find the mistake yourself, you may post the model, and also your expected results (and why you expect these). I (or someone else) can take a look at it. Best regards, Frank

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 27.02.2014, 10:59 GMT-5
Hello Frank,

The updated model can be found in this dropbox folder: www.dropbox.com/sh/mvgwp1qbc0dcwa6/YhJhuLItkj

For the variables taken these are by hand calculations:
Re,cold=2238
Re,hot=8270
Dh,cold=0.0087m
Dh,hot=0.01m
experimental:
Q, cold=3500 W
Q, hot = 3700 W

Comsol:
cell Re,cold = 89
cell Re,hot = 302
I understand that these are local numbers (dx) and depending on the mesh element size h=0.00345m.
My suggestion was Re = ( cell Re / h ) * Dh, but then it's 10 times smaller.
Re,cold = (89/0.00345)*0.0087 = 224 ...
ntflux = 1218 W
Flux depending on overall heat transfer coefficient (depending on Re and Pr) is also smaller probably because of cell Re, so also doing by hand, but then Re should be correct :)

This is my first Comsol experience, I'll try later the 2D model.

Thanks a lot,
Adriaan
Hello Frank, The updated model can be found in this dropbox folder: https://www.dropbox.com/sh/mvgwp1qbc0dcwa6/YhJhuLItkj For the variables taken these are by hand calculations: Re,cold=2238 Re,hot=8270 Dh,cold=0.0087m Dh,hot=0.01m experimental: Q, cold=3500 W Q, hot = 3700 W Comsol: cell Re,cold = 89 cell Re,hot = 302 I understand that these are local numbers (dx) and depending on the mesh element size h=0.00345m. My suggestion was Re = ( cell Re / h ) * Dh, but then it's 10 times smaller. Re,cold = (89/0.00345)*0.0087 = 224 ... ntflux = 1218 W Flux depending on overall heat transfer coefficient (depending on Re and Pr) is also smaller probably because of cell Re, so also doing by hand, but then Re should be correct :) This is my first Comsol experience, I'll try later the 2D model. Thanks a lot, Adriaan

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.