Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

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

conflict with ANSYS workbench and APDL using bolt pretension

Status
Not open for further replies.

dbhollan

Automotive
Jul 27, 2007
2
Good afternoon,

I've got a tightened nut problem that is giving me conflicting results in ANSYS. Both models have the same dimensions and material properties (not steel or aluminum, but linear). No other loads or BC's are applied, except friction contact between the nuts and members. The method I use for checking the preload is detailed below:

1.) Simple static calculations (EA/L) estimate that the total displacement (bolt stretch + member crush) required to develop the preload is 0.000331".

2.) I apply a bolt pretension load in ANSYS Workbench, equivalent to 8000 lbs. The solution comes back with 0.000501" required. 150% of my theoretical. However, I can believe this because the nuts and washers are also meshed and compressible pieces of the puzzle.

3.) I apply this same displacement, 0.000501, to an ANSYS APDL (aka "Classic") model. I do this by "modeling in" an initial penetration between the nut and the member, and letting the contact work out the rest.

To check whether or not the preload has been correctly applied, I look at the sum of the nodal forces (FSUM) on the surface of the nut. I put 8000 lbs in, so I expect to get 8000 lbs out, right? Wrong. The Workbench model says the FSUM is 10,000 lbs, and the APDL model says it is 4000 lbs.

What gives? Is it because the FSUM value is sensitive to mesh density or element type? Load distribution effects?

If this method isn't effective, what other suggestions does anyone have for determining the initial penetration to develop the required clamping force?

Thanks,

DBH
 
Replies continue below

Recommended for you

I found part of the problem...ANSYS workbench is sensitive to where you select on the surface for applying the preload. It puts the cut plane for the joint on that area. The point I had selected was inside one of the nuts. That's what caused my Workbench results to look so funny. Once that was fixed, the preload FSUM fell right into line (8000 lbs.) However, I'm still chasing down the problem in the APDL model. Just wanted to leave a word of warning for ANSYS workbench users.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor