Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here


Mesh refinement Question

Mesh refinement Question

Mesh refinement Question

Hello guys!!
Just joined the family, and i would like to ask a question. Im doing a small project, and i am writing a fea program to make a mesh on a plane plate.
i have a problem with the stress concentration factor. Im calculation this factor from the geometry (peterson, or http://www.amesweb.info/StressConcentrationFactor/... ) and as a result i get 2.24. then when im trying to refine the mesh , i use Kt=Smax/Snom, and im calculating higher factors than the theoritical (2.8 etc). Is it possible to find higher factor than the theoritical. the Fem method is supposed to simulate some close to reality results,and my guess is that the theoritical factor , is just a aproximation.
Here is a picture of the one quarter of my plate.
Thanks in advance

RE: Mesh refinement Question

I have created many FEM's to determine stress concentration factors. When done correctly, the FEM will often closely match the Peterson solution (more on this on the second paragraph). For the simple problem of a plate with a hole, it is not too difficult to get within a few percent error (and less than 1% error is not difficult to achieve). A plate with a hole is probably the simplest problem.

For some problems, it may be difficult to accurately match the stress concentration factor. Peterson gets data from a variety of publications (numerical methods, FEM, lab experiments). An example is a pin loaded hole. The Kt is a function of clearance, assumed load distribution, relative stiffness of the pin compared to the plate, load magnitude, etc. In order to match the Kt from Peterson, you would need to use a similar set of assumptions.


RE: Mesh refinement Question

The skewness at the expected peak stress location appears poor. Try improving the mesh quality.

Kind regards,

RE: Mesh refinement Question

Thanks a lot for your fast replys! I know that my mesh is poor, but 1)does the mesh out of the peak stress location affect the results? 2) the poorer the mesh at the peak stress location, the lower the Smax=Kt*snom should be. My theoritical smax is 92Pa and i while im adding nodes or condesing the triangles, im getting results from 90 to 108+. is it possible that the error is over the Theoritical Kt. 3) ESPcomposites, as far as i can search about these assumptions, i use the same with petterson, but i have googled it and ive found that maybe its just an approximation. Also ive found that if you have theoritical Kt=3 on a plane plate with a whole, you might find Kt=3.03 because of the assumption that the plate is finite. but this error is very small.
(Attached picture: load on plate)
thanks in advance

RE: Mesh refinement Question

1. The mesh is most important at the stress concentration (number/quality/type of elements). It becomes less important the farther away you go, but you still need to have a reasonable mesh.

2. I believe the closed from solution to the finite width correction factor is an approximation. But it is relatively accurate one, at least for practical W:D ratios (width to diameter). If you are using a very small W/D ratio (maybe 1.5-2.0 or less) the solution may start to diverge. In that case, you may want to try a W/D of at least 3.0 or 4.0 to make sure you have agreement first. Don't forget the finite width correction factor has been used for many decades. If there was a significant accuracy issue, it would have been modified as new methods/computational power became available. I just had a look at an old paper I did where I used "boundary collocation" and it is very accurate compared to FEM models for finite width plates (and the classical equation). Usually less than 0.1% difference. I think you need concentrate on what you may be doing incorrectly.

In general, I don't think you have a very good mesh. What I have done in the past is make a "ring" around the hole (outer diameter about 110% of hole size). Within this ring, the mesh density/quality is very good (nearly ideal). Outside of the ring, the mesh quality is not as good, but that is OK. You want to force the distortion of the elements to occur away from the area of interest. If I was interested in the highest accuracy, I would create 2 or 3 rings of increasing size to allow the mesh to slowly distort.


RE: Mesh refinement Question

brian Thank you very very much!!
Greg im using the correct boundary conditions.

RE: Mesh refinement Question

why TRIAs and not QUADs ? TRIA6 I hope (not TRIA3 which are an awful element).

why such an odd mesh around a hole ? What features are driving this ??

just because you think you've got the correct boundary conditions doesn't necessarily make it so. If you showed us then we'd be able to agree that they are correct (or not).

what loads ?

a bad mesh makes for bad results (not higher or lower, just bad).

What code ? NASTRAN, ANSYS, ...

another day in paradise, or is paradise one day closer ?

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close