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 The information provided may be out of date.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.

Mesh continuity at boundaries

Please login with a confirmed email address before reporting spam

Hello,

I'm starting with Comsol, and I have this question, that may sound really basic to many of you.

I have defined a contact pair between two parts of my geometry - an assembly. Now I would like to know whether there is an automatic way to fit the edges of the elements belonging to either sides of the bondary. Or do we have to work it by hand for each case ?

Thank you for your help !

Greets,

Astrid.

7 Replies Last Post 26 avr. 2013, 02:17 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 19 mars 2013, 18:08 UTC−4
Hi

when you have complex geometries, with only a few boundaries you want to use i.e. for contact, then its worth to group the objects by explicit union(s). COMSOL will apply "continuity" on all internal boundaries of your union and apply "assembly" (duplicate boundaries) for all parts between union objects

In this way you have far less identity boundaries to pick by hand.

one can also ask for default continuity on pairs, then change these if only a few need specific contact selections
--
Good luck
Ivar
Hi when you have complex geometries, with only a few boundaries you want to use i.e. for contact, then its worth to group the objects by explicit union(s). COMSOL will apply "continuity" on all internal boundaries of your union and apply "assembly" (duplicate boundaries) for all parts between union objects In this way you have far less identity boundaries to pick by hand. one can also ask for default continuity on pairs, then change these if only a few need specific contact selections -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 20 mars 2013, 10:01 UTC−4
Hello,

Thank you for the answer.
But in my case (two parallelepipeds made out of concrete in contact), I cannot form my geometry as a union, as I want to define an elastic thin layer between my two parts...
So I formed an assembly, created a contact pair between my two parts, and defined my elastic thin layer. The pre-defined mesh is not continuous at the interface, but I can do it by hand by adjusting the size of the elements.

First - does it actually matter to have the continuity of my mesh at the interface when i specify it as an elastic thin layer ?

Then, if it does, I want to know what I should do in this situations to have automatically a continuous mesh at the interface. I'll have later a far more complex geometry and I doubt I can fit the meshes by hand everytime.

Regards,

A. Billon.

PS : my file is attached
Hello, Thank you for the answer. But in my case (two parallelepipeds made out of concrete in contact), I cannot form my geometry as a union, as I want to define an elastic thin layer between my two parts... So I formed an assembly, created a contact pair between my two parts, and defined my elastic thin layer. The pre-defined mesh is not continuous at the interface, but I can do it by hand by adjusting the size of the elements. First - does it actually matter to have the continuity of my mesh at the interface when i specify it as an elastic thin layer ? Then, if it does, I want to know what I should do in this situations to have automatically a continuous mesh at the interface. I'll have later a far more complex geometry and I doubt I can fit the meshes by hand everytime. Regards, A. Billon. PS : my file is attached


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 20 mars 2013, 16:09 UTC−4
Hi

I'm not 100% sure I understand what you want to do. You need to have either one (but not both) of identity "exclusive or" contact pairs. And thin elastic layer can also be defined in union mode.

The main difference with assembly and union is that in assembly you have two distinct adjacent boundaries, one per domain and you apply physics between, hence these might dissociate / separate. While in Union mode the two common domain boundaries are "glued" and the thin elastic layer is "virtual" or infinitely thin

Contact pair physics allows for friction and normally avoids that the two layer interpenetrates, what a thin elastic layer does not restrict

In assembly mode the mesh might not be identical on both overlapping boundaries, COMSOL maps the flux across accordingly (with the usual numerical errors if the mesh are too different). In union mode there is only 1 boundary, hence they have the same mesh and mesh density

Last thing, do not forget that a contact physics is tedious to make converge (check the doc carefully for the mesh, and definitions required.

Finally your spring constant look very high, and check that you do not enforce numerical round-off issues

--
Good luck
Ivar
Hi I'm not 100% sure I understand what you want to do. You need to have either one (but not both) of identity "exclusive or" contact pairs. And thin elastic layer can also be defined in union mode. The main difference with assembly and union is that in assembly you have two distinct adjacent boundaries, one per domain and you apply physics between, hence these might dissociate / separate. While in Union mode the two common domain boundaries are "glued" and the thin elastic layer is "virtual" or infinitely thin Contact pair physics allows for friction and normally avoids that the two layer interpenetrates, what a thin elastic layer does not restrict In assembly mode the mesh might not be identical on both overlapping boundaries, COMSOL maps the flux across accordingly (with the usual numerical errors if the mesh are too different). In union mode there is only 1 boundary, hence they have the same mesh and mesh density Last thing, do not forget that a contact physics is tedious to make converge (check the doc carefully for the mesh, and definitions required. Finally your spring constant look very high, and check that you do not enforce numerical round-off issues -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 21 mars 2013, 06:33 UTC−4
Hello Ivar,

Thank for your last post, it answered all the questions I had.
This was just a test-programm to see how Comsol deals with interfaces and get used to the pairs.
My spring constant are arbitrary and tended to simulate a perfect bound between my two blocs.

Have a nice day !

Regards.

Astrid.

Hello Ivar, Thank for your last post, it answered all the questions I had. This was just a test-programm to see how Comsol deals with interfaces and get used to the pairs. My spring constant are arbitrary and tended to simulate a perfect bound between my two blocs. Have a nice day ! Regards. Astrid.

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 21 mars 2013, 06:49 UTC−4
Hi

I suspected so, just do not forget about numerical issues when you have more than 6-8 orders of magnitude betwen elements. COMSOL does scalings, but cannot take all into account

--
Good luck
Ivar
Hi I suspected so, just do not forget about numerical issues when you have more than 6-8 orders of magnitude betwen elements. COMSOL does scalings, but cannot take all into account -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 25 avr. 2013, 12:44 UTC−4

Hi

I work with a multiphisic problem, a square pzt and a host structure. Even if the geometry is set with 'form union' there is no continuity between the pzt and the host structure. Why? Is there a specific condition that I don't know between pzt and the host structure?

Thaks in advance

Vincenzo
Hi I work with a multiphisic problem, a square pzt and a host structure. Even if the geometry is set with 'form union' there is no continuity between the pzt and the host structure. Why? Is there a specific condition that I don't know between pzt and the host structure? Thaks in advance Vincenzo

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 26 avr. 2013, 02:17 UTC−4
Hi

these two volumes in 3D must be adjacent or have a common overlapping boundary, for the "continuity to work by default, perhaps they are almost overlapping due to some numerical rounding. Try also the resolution setting in the geometry

--
Good luck
Ivar
Hi these two volumes in 3D must be adjacent or have a common overlapping boundary, for the "continuity to work by default, perhaps they are almost overlapping due to some numerical rounding. Try also the resolution setting in the geometry -- Good luck Ivar

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.