The vector plot of Max and Min Principal stress in a shell model looks strange.

[fllunmei]

Hi There

I just modeled a deep girder by shell model to check the vector plot of max and min principal stress and compared the result with the theoretical result. The girder is with two ends fixed and superimposed distributed load on the top. 

The theoretical trajectory of max and min principal stress should be like:

2-Stress-trajectories-in-homogeneous-rectangular-beam-Arthur-H-Nilson-et-al-2010.png

As for the result from Inventor Nastran, the  max principal stress (tension)with vector looks right:

Snipaste_2021-05-11_14-24-15.jpg

However the min principal stress (compression) is strange, in some areas, the vector is out of the plane? What caused that to happen?

Snipaste_2021-05-11_14-27-57.jpg

The model is attached

Thanks

[imajar]

I don't know the answer, but I have a few comments and I'm interested in what the answer ends up being!

First:  your second plot is actually showing the minimum principal stress contour with max principal stress vectors according to the tag in the bottom left corner - so there is a mismatch there.  I've updated new plots below for reference.  

Second:  Just to be sure - min and max stresses are not defined as tensile or compressive - the min and max stresses can be either tensile or compressive depending on the state of stress.  You might already know this, but the way you worded your question below suggested that you were defining the principal stresses by stress direction.

It does seem strange to me that the compressive max principal stress vectors are out of plane while the compressive min principal vectors are in plane. . . 

MinimumPrincipal.JPGMaximumPrincipal.JPG

[John_Holtz]

Hi all.

This looks like a problem with the vector plot. (In other situations, defining the material axis solved the problem with the arrow direction. However, the material axis was defined in this model, so that is not the solution here.)