Nodal Solution vs Element Solution

[speed320]

Hey all,
This question might have been asked before but a search didnt bring it up.
Cutting to the chase...
I am doing a detailed analysis on a 18" pipe that we're using as a strut. Prelminary analysis gave me the forces and moments applied on the top of the beam. Now, I'm working on a detailed analysis on the lower end of said strut using shell63 elements. I run the analysis and look at the nodal and element solutions for von Mises. The element solution listing for SEQV top layer for element #1 says 12000 at node i. But I open up nodal solution listing and look at the von Mises there it says only 9000. So basically, my question is this - If the element solution listing says 12000 at node i, shouldnt the nodal solution listing also say 12000? The top and bottom layer results for the element in question are the same at 12000. What am I missing here?
Thx in advance
Speed

 

[bagarbilla]

For a quick check. Try to change increase the mesh density in the region under consideration. And then check that is result of nodal and element result is more close or not.
I hope this give you the right guess.

 

[engn555]

Nodal solutions are averaged whereas element solutions are not. Look at the contours, the nodal contours will be nice and continuous between elements, whereas the element contours will be disjointed.
Increase the mesh density until the element contours are continuous as well and the nodal vs element values should then be more or less equal.

 

[speed320]

Ok guys, thanks for all the replies and apologies for not responding earlier.
Well I took your sugegstions, and yes the solutions do get closer. But as one question is answered, two others come up. Can we safely say that if the nodal solutions are roughly equal to the element solutions, the model solution is converged, or at least close to it?
Next questions, why are element solution results different from those plotted on element table?
Once again, thanks in advance