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 The information provided may be out of date.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.
Moving geometry due to pressure load and contact between components.
Posted Apr 7, 2011, 10:46 a.m. EDT Geometry, Mesh, Structural Mechanics Version 4.1 3 Replies
Please login with a confirmed email address before reporting spam
Hello everyone,
I would like your comments regarding the simulation I'm trying to do.
(See Schematic in attachment).
- 2D axisymmetrical model;
- Pressure load over an elastomer moving over a fix aluminum hemisphere (both separated by an air gap);
Goal: To get the pressure distribution over the hemisphere (i.e. to simulate the movement of the elastomer contacting the hemisphere due to pressure load).
Reading from the forums, I tried to simulate in the way Ivar was suggesting: considering the air gap as a "gelly-solid" with very low resistance but still considered a solid in the Solid Mechanics study. I don't know if it is valid in my case since the "air gap" should end up being entirely crushed and replaced by the pressurized elastomer.
My questions are:
- The displacement of the elastomer should cover the gap and contact the aluminum hemisphere. How to I simulate this? Do I really need to use the ALE moving mesh for that purpose?
- I don't need to know what's happening in the "air gap". Is there a way to remove it from the simulation but still being able to get the elastomer contacting the hemisphere? (I tried to use a contact pair but my model didn't converge)
Any comments/recommendations on how I should jump-start this again would be greatly appreciated.
Thanks!
I would like your comments regarding the simulation I'm trying to do.
(See Schematic in attachment).
- 2D axisymmetrical model;
- Pressure load over an elastomer moving over a fix aluminum hemisphere (both separated by an air gap);
Goal: To get the pressure distribution over the hemisphere (i.e. to simulate the movement of the elastomer contacting the hemisphere due to pressure load).
Reading from the forums, I tried to simulate in the way Ivar was suggesting: considering the air gap as a "gelly-solid" with very low resistance but still considered a solid in the Solid Mechanics study. I don't know if it is valid in my case since the "air gap" should end up being entirely crushed and replaced by the pressurized elastomer.
My questions are:
- The displacement of the elastomer should cover the gap and contact the aluminum hemisphere. How to I simulate this? Do I really need to use the ALE moving mesh for that purpose?
- I don't need to know what's happening in the "air gap". Is there a way to remove it from the simulation but still being able to get the elastomer contacting the hemisphere? (I tried to use a contact pair but my model didn't converge)
Any comments/recommendations on how I should jump-start this again would be greatly appreciated.
Thanks!
Attachments:
3 Replies Last Post Apr 13, 2011, 1:37 a.m. EDT
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Is this too exotic?
No one has ever done that before? Or at least something similar?
I tried to work based on the "Hyperelastic seal" model example but even if I change my geometry to start with a point of contact, the model does not converge.
I know that contact pair are better if they are 2 convex shapes but I can't believe that COMSOL cannot deal with a concave enclosure?
Any helpful comments are welcome!
No one has ever done that before? Or at least something similar?
I tried to work based on the "Hyperelastic seal" model example but even if I change my geometry to start with a point of contact, the model does not converge.
I know that contact pair are better if they are 2 convex shapes but I can't believe that COMSOL cannot deal with a concave enclosure?
Any helpful comments are welcome!
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi, Dominique!
I don't know that if I got your idea. Did you mean that the air gap is far from necessary?
I haven't deal with such problem, but I think that you can creat contact pair betwen AL and elastomer. You can set a initial displacement value in the boundary setting of contact pair when you consider model with air gap.
With or without air gap is not crucial in my opinion. Your problem is a contact model, "no convergence" may be the key point. The mesh, the initial condition and the parameters of solver may cause errors.
I am dealing with a 3D contact model, the error "no cnvergence" has bothering me for quite some time.
Wish this may help you!
Good luck!
Yajuan
I don't know that if I got your idea. Did you mean that the air gap is far from necessary?
I haven't deal with such problem, but I think that you can creat contact pair betwen AL and elastomer. You can set a initial displacement value in the boundary setting of contact pair when you consider model with air gap.
With or without air gap is not crucial in my opinion. Your problem is a contact model, "no convergence" may be the key point. The mesh, the initial condition and the parameters of solver may cause errors.
I am dealing with a 3D contact model, the error "no cnvergence" has bothering me for quite some time.
Wish this may help you!
Good luck!
Yajuan
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
I believe the main issue you are facing is that in FEM with COMSOL (as well as most other FEM programmes) you cannot "fully crush" a domain, as you need some space to keep the existence of the mesh, else you are changing your topology which normally means renumbering of interfaces etc. Hence using 2 distinct models, one before contact, one after
Then you cannot set/define, in COMSOL, contact between internal boundaries (separated by a meshed volume in the same physics). You must at least use the "assembly mode and select the boundaries accordingly, then you need to maintain a minimum gap thickness (small but finite) to avoid the inverting of the mesh structure of your air.
In you case is the air gap really representative ? even to start with ? and is it a closed volume ?, in which case you need some additional equations to cope for the physics, rather than the "solid approach" that is too restrictive w.r.t physics.
You probably need the ALE for the elastomer I believe as you have often very large deformation of almost flow type there
--
Good luck
Ivar
I believe the main issue you are facing is that in FEM with COMSOL (as well as most other FEM programmes) you cannot "fully crush" a domain, as you need some space to keep the existence of the mesh, else you are changing your topology which normally means renumbering of interfaces etc. Hence using 2 distinct models, one before contact, one after
Then you cannot set/define, in COMSOL, contact between internal boundaries (separated by a meshed volume in the same physics). You must at least use the "assembly mode and select the boundaries accordingly, then you need to maintain a minimum gap thickness (small but finite) to avoid the inverting of the mesh structure of your air.
In you case is the air gap really representative ? even to start with ? and is it a closed volume ?, in which case you need some additional equations to cope for the physics, rather than the "solid approach" that is too restrictive w.r.t physics.
You probably need the ALE for the elastomer I believe as you have often very large deformation of almost flow type there
--
Good luck
Ivar