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.

Soild-mechanics, loading due to gas pressure and contact (Comsol 4.0)

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

I think this might be easier to start by explaining what I'd like to do and how I was planing on implementing it and why it isn't working. Then hopefully someone can either help me get it working, or suggest a better way of trying to do it.

I have a 2d axially symmetric geometry composed of many small solid pellets contained in a solid tube filled with pressurized gas. The pellets are initially spaced such that the pellets are not touching each other. The tube is then heated and the pellets expand such that they will come into contact with each other. This would cause the sub-domains to become divided.

My implementation:
I have fully meshed the volume (both gaps, pellets and tubing). I have implemented heat transfer physics. I have added the solid mechanics module for the whole domain (including the gas). My plan was to include a very low young's modulus for the gas thus making it very easy for the pellets to expand into the space occupied by the gas, but include boundary loads (perpendicular to the boundary) to account for the gas pressure. This is easily implemented using an integration coupling variable to determine the volume of space occupied by the gas allowing one to calculate the approximate gas pressure. So far all is going well.
Now I have defined contact surfaces between the necessary pellet surfaces. My intent was to include a small offset distance between contact surfaces to prevent mesh elements in the gaseous region from being crushed. The solid mechanics module would move the mesh elements as necessary to prevent this. If the offset distance was small, it should allow for numerical stability and be a good approximation for contact between the pellets.
Problem: Since the solid mechanics module considers the boundaries internal (since I've meshed the gaps) it will not allow me to apply the contact surfaces.

Some thoughts:
-I don't see a good reason it wont let me use contact surfaces for internal boundaries from a mathematical standpoint provided I include sufficient offset. Is anyone aware of a reason?
-If I don't apply the solid mechanics to the gas in the gap then boundaries are not internal therefore I can apply the contact surfaces. But since I want to have the heat transfer module use the deformed mesh from the solid mechanics I need to have the gap meshed (Ie I don't see how I can make it work by including a second mesh).
-In comsol 3.5a it was possible to enable contact for internal boundaries.

Any ideas?

3 Replies Last Post 9 juin 2010, 12:26 UTC−4
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: 1 decade ago 31 mai 2010, 02:27 UTC−4
Hi

You have many questions there, and I do not have time now to answer that many, but one thing, your mesh (hence geometry) should not change topology (domains should not appear/dissapear) as this cannot be controlled by COMSOL, so you must find ways to express the phenomena without having gaps of "vacuum" suddenly dissapearing.

Now with V4 there should be more possibilities to go in this direction, if your geoemtry changes and you reapply a full meshing process in between, but you might well still have problems with the BC's as your wall numbering would change.

Do you believe you can inplement a surrounding sub-domain of your pellets, to "fill" all gaps, and in there pilot the material properties to mimic the expansion of your pellets, without changing the geometry/topology ? (ALE deformation is still allowed so long no volume tends fully to "0"


Have fun Comsoling
Ivar
Hi You have many questions there, and I do not have time now to answer that many, but one thing, your mesh (hence geometry) should not change topology (domains should not appear/dissapear) as this cannot be controlled by COMSOL, so you must find ways to express the phenomena without having gaps of "vacuum" suddenly dissapearing. Now with V4 there should be more possibilities to go in this direction, if your geoemtry changes and you reapply a full meshing process in between, but you might well still have problems with the BC's as your wall numbering would change. Do you believe you can inplement a surrounding sub-domain of your pellets, to "fill" all gaps, and in there pilot the material properties to mimic the expansion of your pellets, without changing the geometry/topology ? (ALE deformation is still allowed so long no volume tends fully to "0" Have fun Comsoling Ivar

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 1 juin 2010, 02:02 UTC−4
My plan to address the topology issue (splitting domains) was to prevent them from splitting in the first place. This would be accomplished by establishing an offset between contact surfaces which prevents them from actually touching while still applying the forces like they are in contact.

In this situation the subdomain would be linked by a narrow region where the surfaces appear to be in contact, but are actually slightly separated. My problem with this plan is that in COMSOL 4 I have been unable to enable contact surfaces on internal boundaries as was possible in comsol 3.5a. My question is, how do I accomplish the same goal without it?

I do not believe that I would need to pilot the material properties of the gap. I would use a low young's modulus such that it is easily deformable (ie near zero, but not for numerical stability), but apply the gas pressure via load on the boundary.

Is there a method of applying the contact surfaces between boundaries which are labeled as internal boundaries in comsol?
My plan to address the topology issue (splitting domains) was to prevent them from splitting in the first place. This would be accomplished by establishing an offset between contact surfaces which prevents them from actually touching while still applying the forces like they are in contact. In this situation the subdomain would be linked by a narrow region where the surfaces appear to be in contact, but are actually slightly separated. My problem with this plan is that in COMSOL 4 I have been unable to enable contact surfaces on internal boundaries as was possible in comsol 3.5a. My question is, how do I accomplish the same goal without it? I do not believe that I would need to pilot the material properties of the gap. I would use a low young's modulus such that it is easily deformable (ie near zero, but not for numerical stability), but apply the gas pressure via load on the boundary. Is there a method of applying the contact surfaces between boundaries which are labeled as internal boundaries in comsol?

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 9 juin 2010, 12:26 UTC−4
In case anyone else is having similar issues, I thought I'd post an update.

I have tried two suggestions by COMSOL staff, using assemblies, and disabling the structural mechanics in the gap domain. It is possible to do this using assemblies, however I ran into meshing problems (having to do with imprints) when working to assemblies that are making it difficult to apply it to my geometry of interest.

The second solution, is also not applicable because the structural mechanics module produces the deformed mesh which I require in the gap region for other physics modes.

I've managed to get my original idea working in COMSOL 3.5a (attached model file), but but not in 4.0.
I enabled the "internal boundaries" checkbox in the boundary settings tab, and then proceeded to use contact surfaces with offset in the normal way.

Now, if you try opening the working 3.5a file in COMSOL 4.0 it stops working. This is because you cannot enable internal boundaries in 4.0. Check for yourself and see that in 4.0, the contact surfaces in the solid mechanics module say 'not applicable' even though 'internal boundaries' was allowed in the 3.5a model. Is there any possibility that there is an internal boundaries feature in COMSOL 4.0? Any other suggestions?
In case anyone else is having similar issues, I thought I'd post an update. I have tried two suggestions by COMSOL staff, using assemblies, and disabling the structural mechanics in the gap domain. It is possible to do this using assemblies, however I ran into meshing problems (having to do with imprints) when working to assemblies that are making it difficult to apply it to my geometry of interest. The second solution, is also not applicable because the structural mechanics module produces the deformed mesh which I require in the gap region for other physics modes. I've managed to get my original idea working in COMSOL 3.5a (attached model file), but but not in 4.0. I enabled the "internal boundaries" checkbox in the boundary settings tab, and then proceeded to use contact surfaces with offset in the normal way. Now, if you try opening the working 3.5a file in COMSOL 4.0 it stops working. This is because you cannot enable internal boundaries in 4.0. Check for yourself and see that in 4.0, the contact surfaces in the solid mechanics module say 'not applicable' even though 'internal boundaries' was allowed in the 3.5a model. Is there any possibility that there is an internal boundaries feature in COMSOL 4.0? Any other suggestions?

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.