How to cope with a singularity in a report

[321GO]

Hi Guys,

In my assembly i have a stress singularity in the vicinity of the contact with another part of the assembly.

This location is of interest however, there is a decent amount of stress at this particular location, that i'm sure of. The FEA results though are a direct result of mesh size, stresses go to unreal value's.

Now the question:
What is the best approach fot such a situation in a final stress report:

a) mark the location as undefined? (this seems unacceptable to me though)

b) alter the geometry of the mate / parts, so that the boundry is less stiff? (this also seems unacceptable to me, since it would be a deliberate deviation from the actual assembly)

c) somehow do a handcalc to get some ball park value?


Would like to hear how you guys deal with such situations.


Thank you all in advance!

 

[corus]

Although the stress at that location is unrealistic, the question should be as to how you classify the stresses at that location. As it's equivalent to say a point load, or a zero radius fillet say, then you'd classify the stresses as a peak stress component in addition to your primary and/or secondary stresses. The stress limit depends on the classification, ie. a primary membrane stress might be limited to 2/3 yield. prminary membrane plus bending to yield, and the primary plus peak stresses are limited to fatigue considerations, if they apply.

Hoping to say Tata