Stress Analysis Color Bar: wrong reading?

[pedrontts]

Hi,

I ran a simple stress analysis on Inventor'20. In my case I simulated a welded metal. Since I was interested in the geometry only, I assumed that the weld ideal, so the material properties are homogeneous across the base metal, the heat-affected zone and the fusion zone. Then I applied a tensile load (corresponding to 10% YS of the material) perpendicularly to the welded beam, as shown in Figure 1.

The results obtained from the simulation, when the color bar is set to show me results from 0 to 100 MPa, the base metal appears very green, indicating stresses between 40 and 60 MPa (Figure 2). When I changed the scale of the color bar to show me results from 0 to 200 MPa, I expected the base metal to become light blue, since that is the color for stresses between 40 and 60 MPa. However, the base metal became deep blue, which is associated with nearly 0 MPa (Figure 3). 

Why is that?

 

Figure 1.

Figure 2.

Figure 3.

[admaiora]

Can you attach the part?

[swalton · Accepted solution]

Have you used the probe tool to see the stress on the base part?  

Looking at the graphic, I would expect the stress would be between 30 and 40 MPA.  

You might also change the view from smooth shading to contoured shading.  Contoured shading can give more detail about the stress result.

[pedrontts]

Thanks! After using contour shading, the stress distribution made a lot more sense and changed the colours completely.
I still don't understand why, though. Here, for example, are two images showing the Stress XX after bending:

On the left with smooth shading, on the right with contour shading.

If we look at the the welded beam, it looks fairly compressed with smooth shading, and not at all compressed with contour shading. In both cases, however, the probe tool indicates +9 MPa in the welded beam, confirming the contour shading results.

[swalton]

I find that the contour shading result make it easier to understand the stress results over the whole surface.  The smooth shading results seem to be optimized to emphasize the peak results.  

I typically use the contour shading.