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.

Residual stress in a membrane

Please login with a confirmed email address before reporting spam

I had post a discussion on this but now I cannot find it, it just vanished. So I try again:

I am trying to model the effect of residual (tensile) stress in a fully clamped membrane. The membrane is deflected by a torque applied to an inclusion at the center. Because of symmetry, I modeled a quarter of the membrane and used parametric solver to see the effect of the variable stress on the mode shape (please see the attached file). I entered the value for sigma_x and sigma_y in "initial stress" page of the subdomain settings. Now

- As you can see (postprocessing menu>Domain Plot Parameters>Line/Extrusion> plot the z-deflection for boundaries 2 and 8) the deflection curves are irregular at the boundaries. Why is that?
- It seems the membrane is getting more flexible which is counter intuitive. If the value of the sigma is positive does it mean a tensile or compressive stress in Comsol?

--------------------------------------------------------------------------------
I got this hint before but it does not help to solve the problem:

Unfortunately, it is not really physically accurate to have a constant initial stress. What happens is that the initial stress is not consistent with the boundary conditions and the interface between the softer and the stiffer material. This causes the solution to be somewhat strange. The solution is also mesh dependent; You will see that if you refine the mesh, the solution will be different.

Also, the deformation due to the initial stress is of the same magnitude as the deformation due to the torque. Thus, the increased deformation of the pre-stressed situation is not due to softening, but rather a superposition of the two deformations. You can see this by removing the torque completely and solving for just the initial stress.
--------------------------------------------------------------------------------

Does anyone knows what should I do to model the stress in the membrane?
Thanks :)


2 Replies Last Post 27 août 2010, 03:07 UTC−4
COMSOL Moderator

Hello Nima Izadi

Your Discussion has gone 30 days without a reply. If you still need help with COMSOL and have an on-subscription license, please visit our Support Center for help.

If you do not hold an on-subscription license, you may find an answer in another Discussion or in the Knowledge Base.


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26 août 2010, 15:49 UTC−4
Nima,

I am not answering your questions but just listing some of the counter-intuitive things that I have seen so far:

I am almost new in using COMSOL so may be it is me that I do not have much experience properly using it.

1- I applied an initial strain to a 2D slab (located in xy plane) made of an isotropic material. The initial strain was in x direction. It correctly solves for the deformation, which is the same as the analytical one i.e. delta= P*L/A/E. However, the numerical solution for the stress is completely wrong (400 Pa analytical vs. 1e-11 numerical) which is odd. In addition the stress is not uniform and changes significantly across the area that the strain was applied despite a uniform strain.

2- Then I deactivated the initial strain and activated the stress (the stress that I calculated analytically but did not get numerically). In this case, amazingly when I added a positive stress (outward) the slab underwent compression and when I applied a compressive stress the slab stretched !! (although the magnitudes of the deformations were correct).

3- Apparently, the mode of Plane-Stress or Plane-Strain does not affect the above observations.

4- When I switched from isotropic material to a hyperelastic material with large deformation option (or even w/o it) nothing happened as if there were no initial strain or stress.

Manuel
Nima, I am not answering your questions but just listing some of the counter-intuitive things that I have seen so far: I am almost new in using COMSOL so may be it is me that I do not have much experience properly using it. 1- I applied an initial strain to a 2D slab (located in xy plane) made of an isotropic material. The initial strain was in x direction. It correctly solves for the deformation, which is the same as the analytical one i.e. delta= P*L/A/E. However, the numerical solution for the stress is completely wrong (400 Pa analytical vs. 1e-11 numerical) which is odd. In addition the stress is not uniform and changes significantly across the area that the strain was applied despite a uniform strain. 2- Then I deactivated the initial strain and activated the stress (the stress that I calculated analytically but did not get numerically). In this case, amazingly when I added a positive stress (outward) the slab underwent compression and when I applied a compressive stress the slab stretched !! (although the magnitudes of the deformations were correct). 3- Apparently, the mode of Plane-Stress or Plane-Strain does not affect the above observations. 4- When I switched from isotropic material to a hyperelastic material with large deformation option (or even w/o it) nothing happened as if there were no initial strain or stress. Manuel

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 27 août 2010, 03:07 UTC−4
Hi

I have some issues when following your model:
1) you apply a force and not a torque on the cenral one
2) for me you cannot simplify the initial stress only as a X and Y stress components, because the stress tensor should change with the correct relation between each other
3) you have an antisymmetric boundary that I do not catch why to
4) your mesh is too coarse in the thickness, you should use either a surface shell, or use the advanced feature and expand the Z direction by a factor of 20-50 to get more elements in the Z direction (or use the sweep mesh)

Try to study the stress distribution in your system when you apply the different stress values, that would probably give you some clues of what is going wrong (not that the von mises hardly changes (combination of all), but the different stress components do change in different ways. But take care with the automatic plot rescaling when you change the parameter setting, check also carefully te absolute values, not only the rainbow = % relative values

--
Good luck
Ivar
Hi I have some issues when following your model: 1) you apply a force and not a torque on the cenral one 2) for me you cannot simplify the initial stress only as a X and Y stress components, because the stress tensor should change with the correct relation between each other 3) you have an antisymmetric boundary that I do not catch why to 4) your mesh is too coarse in the thickness, you should use either a surface shell, or use the advanced feature and expand the Z direction by a factor of 20-50 to get more elements in the Z direction (or use the sweep mesh) Try to study the stress distribution in your system when you apply the different stress values, that would probably give you some clues of what is going wrong (not that the von mises hardly changes (combination of all), but the different stress components do change in different ways. But take care with the automatic plot rescaling when you change the parameter setting, check also carefully te absolute values, not only the rainbow = % relative values -- 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.