Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Problem with contact convergence...

Status
Not open for further replies.

vyg29

Mechanical
Nov 23, 2012
1
Hey everyone,

I'm a mechanical engineering grad student and for my research I'm working with ANSYS Classic (version 12.1). My research involves letting a structure of beams (a compliant mechanism) deform due to contact with another object. To get familiar with using contact elements in ANSYS I'm working on a small problem where I model just one straight beam with two contact points, and two target points on a circular arc. The beam is clamped in one end and has a point load applied to it on the other end.

After working on it for quite a while (weeks, unfortunately...) I was finally able to get it to converge and work perfectly, except that it takes far too long (>30 minutes for 1 run). I have to use this model for optimization, and right now it takes too long to do that. I need to have information (for example contact forces, deflections) over the entire range of motion, which is why I'm using a do/for loop.

So, my question is: could you perhaps take a look at the code, and see if there are things you'd recommend me to do differently? I feel that I'm missing something, but I just can't figure out what.

What I've tried so far:
- Use only solve instead of load steps (faster, but does not converge)
- Use solve but have NSUBST increase a bit each loop
- Change the contact stiffness and penetration tolerance (both higher and lower, no effect)
- Decrease the total number of load steps (by increasing P_step)
- Decrease NSUBST

Now I know that usually one comment is that when you're using contact elements you should not have a gap, as that leads to rigid body motion and that can give problems. However, I know of quite a few people who have used ANSYS in a similar method and that worked perfectly, so it should be doable.

I hope this is a clear explanation, if not please let me know! Thank you for your time!
Valerie
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor