Unrealistic high stresses (linear FEA)

Question

Unrealistic high stresses (linear FEA)

321GO (Automotive)

(OP)

23 Mar 14 13:32

I'm doing a FEA on a clamped pipe assembly (bolted with some kind of straps) and i'm somewhat puzzled why the stresses on the pipe are literally of the scale (1000-2000 Mpa).

The pipe would probably have some / minor plastic deformation due to the clamping but still...
The constraints, loads, materials are ok and the deformation also seems legit, so i don't think it's due to the setup per se for the moment.

Could it actually be that all those of the scale location would yield and would that then mean that the stresses on those locations would be the yield stress? Or would it be some lower value after the yield has settled?
I know from experience that the clamping has a decent faitgue life so i'm very skeptical for the stresses being that high in real life.

Basically i don't know what's going on and how to interpret this.

Greatly appreciate the feedback

corus (Mechanical) · Answer 1 · 2014-03-24 04:15:13

High stresses may be due to a geometric discontinuity, or point load/contact between bodies where stresses would tend towards infinity. In real life these stresses are likely to be localised and limited to yield. If there is little stress range during repeated cycles then fatigue damage may not be of such concern.

321GO (Automotive) · Answer 2 · 2014-03-30 14:42:19

Thanks corus,

We also ran the simulation non-linear in Ansys and the results are similar in that the overal deformation and stress regions are the same but the stresses are much more realistic, i.e. close to or somewhat below yield.

I'm just kind of suprised that the stress was extremely high for the non-linear solution.
For instance the locations with yield (or close to it) in the non-linear solution where 2-3 times yield for the non-linear solution!

Perhaps i should also mention that it is an assembly with numerous contacts

corus (Mechanical) · Answer 3 · 2014-03-31 13:59:01

The value obtained from a linear FEA at this discontinuity are immaterial as they'll be mesh dependent, ie. the finer the mesh then the higher the stress. Of course with high stresses above yield then there is the danger of the pipe cracking as stresses reach UTS at some point.

Come Join Us!

Posting Guidelines

Unrealistic high stresses (linear FEA)