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!

Corner stress vs Center stress, which one shall be used ?

Status
Not open for further replies.

amec2004

New member
May 18, 2004
161
I am doing some FE analysis using plate element in the STAAD model.

For the analysis result I always use max Von Mise stress at center of plate (center stress) to compare with steel's yield strength.

Some other engineers spot that some plate's corner stress is 30% more than the center stress and exceeds material's yield strength.

Shall I ignore the larger corner stress value and only use center stress for comparison?

What's the relationship between corner and center stress in STAAD ?
 
Replies continue below

Recommended for you

The center stress is the most reliable but you cannot ignore the corner stresses. In many cases the maximum stresses will be at the edges of a floor/wall/region. For the corner stresses you should average all of the corner stresses at a node because they will not be equal in each plate. In effect the corner stress is an extrapolation based just on the displacements of one plate. So averaging the stresses from each of the plates will get the effect of the displacements of each of the plates attached to the node. StaadPro 2006 will draw the contours using the center stresses and the nodal average stress if you uncheck the "index based on center stress" option.
 
You've hit upon what I think is a huge weakeness in STAAD. If your mesh is too coarse you can miss high stress areas. Yes, this could happen in any program but STAAD is especially susceptible because of this averaging thing Rayc1 was talking about. STAAD does not report average nodal stressess, you have to do it by hand as Rayc1 described. This is just crazy to me.

Any other FEA program I've ever used allows you to get the average stress at a node, usually just by putting your cursor over the node or clicking on it.

And wait until you have to explain this corner stress to the client. "Oh Mr. Client, that stress is fictitious. You have to add the stress in the corner of each of the other three elements that share this node to the one we're looking at and divide by four."

Back to your question, no don't ignore it. Try to understand why it's happening.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor