Contact errors

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Contact errors

mikegascoyne
Hi Everyone,

I have searched the forum and cannot get any answer. I am modeling contact between a rigid cylinder and a stent.  Boundary conditions are applied to the cylinder such that first decreases in diameter crimping the stent and then increases allowing for the expansion of the stent.

I am having contact difficulties with my analysis erroring out due to max penetration between surfaces. I have tried numerous things to improve convergence.

1. Contact definition: using Pressure - Exponential contact definition to soften contact between the surfaces.
2. Setting both the rigid surface and the stent as master and slave surfaces, effectively doubling up the contact definition.
3. Increasing the contact stiffness in contact controls.
4. Using node to surface instead of surface to surface

Here is the typical error I get at each increment.

MAX. PENETRATION ERROR 890.024E-03   AT NODE 1222869 OF CONTACT PAIR
   (SURF_STENT,CRIMPER_SURF)
   MAX. CONTACT FORCE ERROR 5.03376     AT NODE 7981 OF CONTACT PAIR
   (CRIMPER_SURF,SURF_STENT)
          PENETRATION ERROR TOO LARGE COMPARED TO DISPLACEMENT INCREMENT.

Any ideas would be very helpful.

Thanks

Mike

Reply | Threaded
Open this post in threaded view
|

Re: Contact errors

simon_f
Hi Mike,

I've had a few similar error messages and found the following steps have been helpful in the past - apologies if you've already considered these:

1.) Ensure the slave surface has a finer mesh density than the master surface
2.) Check element type and contact definition.  I tend to use C3D10 elements for most solids and use
*surface behavior, penalty=nonlinear
Then make sure the unsymmetric matrix is set to 'on' in the step command to allow the small deformations at the contact surface:
*step, name=Step-1, unsymm=YES
3.) At the risk of stating the obvious, it's worth double checking your surface definitions etc., especially if you have lots of different contact regions.  I have had cases where I've missed one region (e.g. a bolt head against a casing) and spent hours trying to work out why it won't solve!
4.) If you're using pressure-overclosure=exponential, you may need to specify a clearance (I've been advised by my much more experienced colleagues that a clearance just smaller than the distance value in the exponential definition is ideal) to improve convergence.

Hope that's a help, and good luck with the simulation.

Regards,

Simon
Reply | Threaded
Open this post in threaded view
|

Re: Contact errors

sandeep
This post has NOT been accepted by the mailing list yet.
Hello Simon,

I had the similar problem of contact. i had modeled a Dam structure with blocks. for contact between two blocks I used pressure overclosure = hard contact, but the problem wasn't converging. then I used pressure overclosure = Exponential with Po=1 & Co = 1,0.1,0.01, so on to 0.000001 and Po=0.000001 & Co=0.000001. With all these values the simulations were converging. But there were penetrations between the adjecent block.

1. Please let me know how to avoid this penetration

2. what exact values should be used for Po & Co. and

3. How can I measure the amount of penetration

Please clarify my queries. I would be really grateful to you.

With Best Regards

Sandeep