which results are more reliable or closer to the reality - averaged or unaveraged stresses?
How is averaged/unaveraged used and how can it help?
Which one are you using when you validate the resulst for von-Mises stesses?
As an example I attached a picture showing the von-Mises stresses of a region where a bolt pre-load is applied.
There is a Surface-to-Surface Contact including friction between the bolt-head (hidden) and the clamping region.
The greatest dicrepancy is found on the edge region which I find quite interesting.
Solved! Go to Solution.
In general, both results should be close . That means the mesh density is adecuate. Any large difference between node/element results means you have to remesh the area. But in your case the high stress is due to an edge?. Like a small "jump" at this location. This could represent a "singluarity" and FEM will never give you accurate results altough you remesh the area.