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.

Extrusion coupling variables

Please login with a confirmed email address before reporting spam

Hello,

I have a model that consists of two geometries: one in 1D and the other in 2D axial symmetry. For each time step, I want the values attained by solving for one of my variables from a line boundary in the 2D axial symmetry to be used to solve for a second variable in the 1D geometry. After this variable is solved for in the 1D geometry, I want these values used to solve for the first variable in the 2D axial symmetry geometry. I want this to occur repeatedly so that my model is continuously sending the last solved for values in one geometry to the other geometry. My line in 2D axial symmetry space is the same length as the line in 1D, so I believe I want to do a linear transformation with "extrusion coupling variables". I have tried to set up two linear transformations, but am not getting my model to work properly. Does anyone have any insight for this problem?

Thanks,
Andrew

4 Replies Last Post Mar 11, 2010, 4:22 PM EST

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 3, 2010, 7:10 PM EST
can you post your model?
could be helpful...
jf
can you post your model? could be helpful... jf

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 4, 2010, 11:13 AM EST
My model is attached. Geometry 1 is a 2D axial symmetry rectangle where I am modeling heat transport by conduction. Geometry 2 is a 1D line where I am modeling heat flow from water flowing down a pipe. I am using the Comsol Multiphysics, Heat Transfer, Conduction, transient analysis for the application mode of the 2D axial symmetry geometry and the PDE Mode, Coefficient Form, Time-dependent analysis for the application mode of the 1D geometry. I am using the PDE Coefficient Mode so that I could insert my own PDE equation to simulate heat flow down a pipe with a diffusion coefficient, an absorption coefficient, a source term, a damping/mass coefficient, and a convection coefficient.

I want my model to solve for T (block temperature) in the 2D geometry (rectangle) and use the T values along boundary 1 (or r=0) to insert into the equation to solve for w (water temperature) along the line in geometry 2 (T is in the source term for the PDE coefficients). Then I want the w values from geometry 2 to be used to solve for the heat flux boundary condition along boundary 1 (or r=0) in the 2D rectangle. I want both of these geometries to continually provide each other with updated T or w values. I believe I should be able to do this by setting up 2 extrusion coupling variables, but I have not been able to make it work yet.

Everything is set to 283.15 K (10 C) initially, and the input of my 1D pipe has a sine wave where temperatures should fluctuate between 273.15 K (0 C) and 293.15 K (20 C). I have assigned a constant water velocity (u=1 m/s) down the pipe so I shouldn't need to simulate fluid flow with this model.

Does anyone have any insight as to why I haven't been able to couple my variables correctly?

Thanks,
Andrew
My model is attached. Geometry 1 is a 2D axial symmetry rectangle where I am modeling heat transport by conduction. Geometry 2 is a 1D line where I am modeling heat flow from water flowing down a pipe. I am using the Comsol Multiphysics, Heat Transfer, Conduction, transient analysis for the application mode of the 2D axial symmetry geometry and the PDE Mode, Coefficient Form, Time-dependent analysis for the application mode of the 1D geometry. I am using the PDE Coefficient Mode so that I could insert my own PDE equation to simulate heat flow down a pipe with a diffusion coefficient, an absorption coefficient, a source term, a damping/mass coefficient, and a convection coefficient. I want my model to solve for T (block temperature) in the 2D geometry (rectangle) and use the T values along boundary 1 (or r=0) to insert into the equation to solve for w (water temperature) along the line in geometry 2 (T is in the source term for the PDE coefficients). Then I want the w values from geometry 2 to be used to solve for the heat flux boundary condition along boundary 1 (or r=0) in the 2D rectangle. I want both of these geometries to continually provide each other with updated T or w values. I believe I should be able to do this by setting up 2 extrusion coupling variables, but I have not been able to make it work yet. Everything is set to 283.15 K (10 C) initially, and the input of my 1D pipe has a sine wave where temperatures should fluctuate between 273.15 K (0 C) and 293.15 K (20 C). I have assigned a constant water velocity (u=1 m/s) down the pipe so I shouldn't need to simulate fluid flow with this model. Does anyone have any insight as to why I haven't been able to couple my variables correctly? Thanks, Andrew


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 11, 2010, 2:58 PM EST
well I looked at your model and the first thing that catch my eye is that your pipe is at r=o.
So the heat flux wil be coupled to a null area. I beloeve that your block should start at coordinate rp rp being your pipe radius.
Indeed axi symmetric has 2*pi*r term invlved and I dont know how it will handle a flux at r=0.
did you think of that..?
second
In your 1d geometry why did you make so many subdomain?
it is possible that it screw your extrudsio n process

SO in short I will recommend to use only One 1d subdomain
I will start my axisymmetric block at the radius of my pipe and not zero...
Once this is done if ti does not fix it we should look for somehting else...
good luck
jf
well I looked at your model and the first thing that catch my eye is that your pipe is at r=o. So the heat flux wil be coupled to a null area. I beloeve that your block should start at coordinate rp rp being your pipe radius. Indeed axi symmetric has 2*pi*r term invlved and I dont know how it will handle a flux at r=0. did you think of that..? second In your 1d geometry why did you make so many subdomain? it is possible that it screw your extrudsio n process SO in short I will recommend to use only One 1d subdomain I will start my axisymmetric block at the radius of my pipe and not zero... Once this is done if ti does not fix it we should look for somehting else... good luck jf

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 11, 2010, 4:22 PM EST
JF,

Thank you. The pipe at r=0 was indeed the problem. I moved the edge of the block to the pipe radius, and my model now appears to be running correctly. I overlooked the problem that this would create.

When I created the line in 1D, I didn't specify the number of subdomains. The default was 1000 subdomains, the length of my pipe. However, the edge of the block at r=0 was most likely the biggest problem.

Thanks again,
Andrew
JF, Thank you. The pipe at r=0 was indeed the problem. I moved the edge of the block to the pipe radius, and my model now appears to be running correctly. I overlooked the problem that this would create. When I created the line in 1D, I didn't specify the number of subdomains. The default was 1000 subdomains, the length of my pipe. However, the edge of the block at r=0 was most likely the biggest problem. Thanks again, Andrew

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.