9 Replies Latest reply on Aug 3, 2018 5:06 PM by Ryan Dark

Unrealistic stresses in "simple" analysis.

Josh Barnett Aug 1, 2018 6:12 PM

Good evening all,

I am trying to analyze a part, simulating wind pressure on a large surface.

I have an aluminum extrusion, which is attached to a foam board.

The aluminum has a hinge which attaches to another piece.

So, the hinge would be fixed and the rest of the part is free to move. I want to apply pressure to the board, and I am concerned about the aluminum hinge failing.

First issue, my mesh ratio is too high, and I'm not sure if it is causing the following issue.

I want to apply 30-40 psf to an 8'x8' board. Then trasfer the load onto the extrusion.

For example I applied a force of ~500lb (made it up), but this displacement is only .00106 inches.

The material is 6061-T6 and the thickness of that area is only 0.09"

So it seems to me it's incorrect.

Re: Unrealistic stresses in "simple" analysis.

Ryan Dark Aug 1, 2018 6:39 PM (in response to Josh Barnett)

Hi Josh,
It looks like the aspect ratio of you mesh elements is very high which would lead to a stress plot just like what you are seeing here now. Stress results for a solid mesh on a thin walled object start looking normal once you have 2 elements across the thickness. As you might expect this would make your analysis have many more mesh elements than it presently has and take much longer to solve. Just looking at the screen shots of your model you have a good candidate for using shell elements for your mesh. That would yield better stress results at the same speed (or, more probably, faster). This document goes over how to define a shell mesh onto your existing solid geometry (GoEngineer - Siimulation: Definte Shell by Selected Face) for reference.
Like Show 1 Likes(1)

Actions
- Re: Unrealistic stresses in "simple" analysis.
  
  Josh Barnett Aug 2, 2018 6:30 PM (in response to Ryan Dark)
  
  I've tried very small mesh element size with the same result. I can figure out how to apply the shell to the flat faces thanks to that quick tutorial. But how do I apply the shell elements to the hinge part of the piece?
  
  I'm lost at that point because I don't know which would face would be top and which would be the bottom face(s).
  
  What is the correct procedure for a geometry like this?
  
  Attached is a screenshot of the analysis w/ .01" min mesh size.
  
  Like Show 0 Likes(0)
  
  Actions
  - Re: Unrealistic stresses in "simple" analysis.
    
    Ryan Dark Aug 3, 2018 12:02 PM (in response to Josh Barnett)
    
    Josh,
    Is this an imported .STL? I ask because the model itself has a weird triangular texture on its surface?
    
    Did you put elements on the model to that were set to be half the wall thickness in size?
    
    Like Show 0 Likes(0)
    
    Actions
    - Re: Unrealistic stresses in "simple" analysis.
      
      Josh Barnett Aug 3, 2018 1:27 PM (in response to Ryan Dark)
      
      It was a dwg drawing of the extrusion and i converted the dwg to a solidworks sketch and then extruded the part.
      
      Like Show 0 Likes(0)
      
      Actions
      - Re: Unrealistic stresses in "simple" analysis.
        
        Ryan Dark Aug 3, 2018 1:33 PM (in response to Josh Barnett)
        
        If you can, attach the file up here and let us take a look at it.
        
        Like Show 0 Likes(0)
        
        Actions
      - Re: Unrealistic stresses in "simple" analysis.
        
        Josh Barnett Aug 3, 2018 1:37 PM (in response to Josh Barnett)
        
        Attached are the parts, if you would like to examine them.
        
        Part2b.SLDPRT.zip 105.9 KB
        
        Part10-4in.SLDPRT.zip 103.5 KB
        
        Like Show 0 Likes(0)
        
        Actions
        
        Re: Unrealistic stresses in "simple" analysis.
        
        Ryan Dark Aug 3, 2018 4:33 PM (in response to Josh Barnett)
        
        Hi Josh,
        I just reviewed your study setups and it looks like the 3D model has a mesh with some pretty extreme element aspect ratios. I might expect the results to look a bit weird. The 2D study you have setup appears to be setup like a 2D Simplification but not totally there. You are on the right path with that though. Given the loading conditions you have it is a good candidate for a 2D simplification and you will be capable of creating a vastly denser mesh for better results (like this: GoEngineer - Simulation: 2D Simplification). I made a mockup and the plot looks much cleaner in 2D simplficiation:
        
        ...as compared to the full 3D model:
        
        Like Show 0 Likes(0)
        
        Actions
        
        Re: Unrealistic stresses in "simple" analysis.
        
        Josh Barnett Aug 3, 2018 4:51 PM (in response to Ryan Dark)
        
        Could you show me the displacement for this model? Are you using 500 lb?
        
        Like Show 0 Likes(0)
        
        Actions
        
        Re: Unrealistic stresses in "simple" analysis.
        
        Ryan Dark Aug 3, 2018 5:06 PM (in response to Josh Barnett)
        
        This example is a simple 2 lb load. The displacements are not going to be super helpful but here they are:
        
        The values for a 500 lb load will change but I expect the plot quality will be the same.
        
        Like Show 0 Likes(0)
        
        Actions

Go to original post