AnsweredAssumed Answered

Linearized Stress

Question asked by Amit Gupta on Sep 13, 2013
Latest reply on Oct 16, 2013 by Jared Conway

Like • Show 0 Likes0
Comment • 5

Hi All,

I am trying to use Solidworks Simulation to analize a design that I have already validated by hand calcs and physical testing. Basically I have a flat rectangular plate that has a uniform pressure on one side. The opposite side is supported by a "picture frame" structure and a no-penetration constraint such that the center of the plate is free to deflect. The edges of the plate are not constrained in the lateral direction. Basically this arrangement should allow the plate to be in pure bending with minimal membrane stress. My Von Mises Stresses and Stress Intensity (Tresca) values are what I expect. However, if I create a pressure vessel analysis and then try to linearize the stresses at the center of the plate, I get huge membrane stresses. Additionally, the stress intensity values listed under the linearized results are way higher than those reported in the stress intensity results plot. I've attached some pictures to help explain. Am I misinterpreting these linearized results? The pressure vessel study was set up to use the results from only one analysis. I am thoroughly confused by this result. Can anyone help? I appreciate any pointers!

Please let me know if anything is not clear.

Attachments

Von Mises.JPG62.2 KB
Linearize.JPG157.5 KB

No one else has this question

Outcomes

Jared Conway Sep 18, 2013 2:11 AM

Haven't played with this functionality much. Can you post your model?

Also have you checked the solidworks kb? What version and sp are you using?

I think your investigation method is good, check if the other plots are working. Maybe it is just a display issue. Have you tried using the plots to come up with an estimate manually based on the results coming from the software? What values are you expecting?

Actions

Amit Gupta @ Jared Conway on Sep 18, 2013 8:27 AM

I'm using SW 2012 x64 SP5. Unfortunately, my company is not on subscription so I don't have access to the KB. My model and results files are much too large to post.

I have verified that the results I'm getting from the linearization tool are typical for the tool. We have an advanced Solidworks Simulation training book in the office and the linearized results in the book are 10x higher than the results from the default stress plots.

So I guess basically, I'm asking this: What is Solidworks Simulation doing that the linearized von Mises and Intensity would be 10x higher than the highest von Mises or Intensity reported by basic stress plots? What am I missing about how these stresses are being reported?

Actions

Jared Conway @ Amit Gupta on Sep 18, 2013 11:08 AM

The help and the training manual cover that topic pretty well and I think there are a couple of good articles in the kb that cover it as well.

Could you make a simplified model that you can post, no results needed. We can always rerun.

Then it would be helpful to see what you are expecting, what you output manually and compare it to linearize.

Actions

Amit Gupta @ Jared Conway on Oct 16, 2013 10:54 AM

So I think I solved the issue. My results file was somehow corrupted and re-running the analysis just continued the corruption. I had to manually purge all simulation data associated with my model (mesh, results, loads... everything) and then re-run. Now my linearized stresses make sense. I don't know what caused this issue but this seems to have fixed it... for now.

Actions

Jared Conway @ Amit Gupta on Oct 16, 2013 11:01 AM

thats where the test file helps out, unlikely that a new part will come out with corrupt results. sometimes just a new study will fix those problems as well. if they both come out bad, i'd look at your storage location..etc.

Actions

5 Replies

Jared Conway Sep 18, 2013 2:11 AM
Mark Correct
Correct Answer
Haven't played with this functionality much. Can you post your model?

Also have you checked the solidworks kb? What version and sp are you using?

I think your investigation method is good, check if the other plots are working. Maybe it is just a display issue. Have you tried using the plots to come up with an estimate manually based on the results coming from the software? What values are you expecting?
Like • Show 0 Likes0
Actions
- Amit Gupta @ Jared Conway on Sep 18, 2013 8:27 AM
  Mark Correct
  Correct Answer
  I'm using SW 2012 x64 SP5. Unfortunately, my company is not on subscription so I don't have access to the KB. My model and results files are much too large to post.
  
  I have verified that the results I'm getting from the linearization tool are typical for the tool. We have an advanced Solidworks Simulation training book in the office and the linearized results in the book are 10x higher than the results from the default stress plots.
  
  So I guess basically, I'm asking this: What is Solidworks Simulation doing that the linearized von Mises and Intensity would be 10x higher than the highest von Mises or Intensity reported by basic stress plots? What am I missing about how these stresses are being reported?
  Like • Show 0 Likes0
  Actions
  - Jared Conway @ Amit Gupta on Sep 18, 2013 11:08 AM
    Mark Correct
    Correct Answer
    The help and the training manual cover that topic pretty well and I think there are a couple of good articles in the kb that cover it as well.
    
    Could you make a simplified model that you can post, no results needed. We can always rerun.
    
    Then it would be helpful to see what you are expecting, what you output manually and compare it to linearize.
    Like • Show 0 Likes0
    Actions
    - Amit Gupta @ Jared Conway on Oct 16, 2013 10:54 AM
      Mark Correct
      Correct Answer
      So I think I solved the issue. My results file was somehow corrupted and re-running the analysis just continued the corruption. I had to manually purge all simulation data associated with my model (mesh, results, loads... everything) and then re-run. Now my linearized stresses make sense. I don't know what caused this issue but this seems to have fixed it... for now.
      Like • Show 0 Likes0
      Actions
      - Jared Conway @ Amit Gupta on Oct 16, 2013 11:01 AM
        Mark Correct
        Correct Answer
        thats where the test file helps out, unlikely that a new part will come out with corrupt results. sometimes just a new study will fix those problems as well. if they both come out bad, i'd look at your storage location..etc.
        Like • Show 0 Likes0
        Actions

Linearized Stress

Attachments

Outcomes

Related Content

Recommended Content