Overclosure Issue 1

[f1paulzy]

Dear all

To be honest I am not sure where my problem is in the simulation, I hope you can help me with this, its been 10 days.

Here is my problem:
1. I got 2 components in assembly, both were 3D scanned and CAD processed. Like a shaft into a sleeve, but instead it is a residual limb with prosthetic socket.
2. I defined contact between the two component and also the friction coefficient.
3. I use kinematic coupling to define the rotation of the assembly (coupled to a reference, with boundary condition applied to it)
4. The boundary condition is time VS radians which is experimental data. I got 100 data in 0-1.28s.
5. Hence in step 1, the time period is 1.28s
6. The time scaling factor I use is 0.1
7. I tried to refine the mesh, but never succeed

I am not quite sure about where the problem is and apologize for not specifically describe where the problem is.
I hope someone can help me with this.

Here is the inp file (attached). Thanks you guys in advanced

Regards
Jing

 

[mike2014]

There is an option in CAE for adjusting overclosures. It should be an option when you define the contact. Also, try making your time step smaller. If you go into the job-1.sta file while its running, it should show the iterations and how small it has to make the time step in order to close all overclosures. The .msg file should then show you how bad the overclosures are. (all my tokens are being used right now so im going off of memory.) If you look you can get a feel for how bad it is. When i had this problem i made the time step smaller and it fixed it.

I hope it works

 

[corus]

You're using C3D4 elements, which are well known to be useless in stress analysis as they grossly underestimate stresses. You need C3D10M elements which were specifically made for contact using tetrahedral elements.

 

[rstupplebeen]

Jing,

I Ran your model and I made 2 changes.
1. For constraint-1 I changed the slave surface to be the superior face only.
2. For Disp-BC-1 I constrained all DOFs.
With these changes the model starts to solve and then there is too much mesh distortion. A significant amount of work is needed to create a better mesh especially for the sleeve.

I hope this helps.


Rob Stupplebeen

http://www.optimaldevice.com