How to determine the area of stress focus ?

[Lirock]

Hi,everyone

We know that if a force is applied on a node, there will be a stress peak in the result. And this stress must be ignored. But I wonder how to determine the area influnced by this peak ? Or how to deter the stress value around the peak point?

 

[ansysfreak]

Hi

You can do this very easy with non-uniform contours.

grtz garry

 

[Stringmaker]

Here's a few thoughts which pop to mind:

A) Use some sort of RBE3 force distributed constraint or couple the node with several others making where you want to apply the force the master node.

B) Is it possible to unselect the few elements showing the fictitous stresses while post processing so those results will be omitted?

Or as Ansysfreak suggested you may be able to tweek the contours to filter the stress out.

Best,
-Brian

 

[cbrn]

Hi,
or also post-pro your results "by hand":
choose at least three concurrent directions (in the 3D space, three non-colinear direction however oriented form a base of this space) which include the node you want to filter out (the best way is by defining three paths, if you can, so you'll have the results equally spaced; if not and if you have to select nodes manually, then you'll have to measure the distance between them). In Excel, plot the three value-vs-distance and get three least-squares curves: at your "peak point", the least-squares curves won't rise up to the peak value, and they will give you an estimation of the "real" local stress increase due to the concentrated load. The three different directions serve to give you indications of the anisotropy of the stress increase (this may or may not be useful in your case, however...)
I don't know if I've been confusing enough... ;-) I hope not...

Regards