No Von Mises results with SimSolid

PBerry
PBerry Altair Community Member
edited February 1 in Community Q&A

I'm pretty sure when I started this project, I would get a normal looking "Percent of yield" and von Mises. After several runs, now it gives unreasonable results for both. The % of yield here is only about 1%. It should be closer to 60%. I ran the analysis "for stress". The materials are taken from the native database. Is there something in the setup that I might have missed to cause this?

YieldPercent.PNG

January 25th update:

I reset the spectrum scale, and the yield results are present, but they are a couple of magnitudes off. Hard to believe anything with that much obvious error. Thoughts?

YieldPercent2.PNG

January 26 update:

I tried starting completely over with a new session, and imported the parts, created contacts, applied support and load. The results are the same. Something just seems way off on the scale. Help anybody?

Tagged:

Answers

  • PBerry
    PBerry Altair Community Member
    edited February 1

    Comments relocated to first posting

  • PBerry
    PBerry Altair Community Member
    edited February 1

    Comments relocated to first posting

  • rdominguez
    rdominguez
    Altair Employee

    Hi PBerry,

    Do you have any comparison study where those conditions are causing 60% of Yield?

    Could it be possible that the pressure is not enough to cause such damage? By what I see in the pictures, the pressure is around 0.24 MPa.

    For what I see you're using the SimSolid solver, are you using Adapt for Stress when running the structural analysis?

    Let me know your comments and we can found a way to help you :)

    Best,

    R.

  • PBerry
    PBerry Altair Community Member
    edited February 5

    Thank you so much for the response! Undoubtedly this is my error, but I'm just not seeing it.

    <Do you have any comparison study where those conditions are causing 60% of Yield?>

    No, I have been iterating the design, and not keeping all of the old results. It would not have been on this exact geometry, or even materials. Honestly, there is room for me misremembering. It seems like I had seen 'normal' results while working through the design. My 60% approximation is arrived at by simply applying twice the design pressure pressure (1x design pressure = yield) to a structure with a safety factor of 3. 2/3=0.666

    <Could it be possible that the pressure is not enough to cause such damage? By what I see in the pictures, the pressure is around 0.24 MPa.>

    The intent of the 5000 lbf/ft2 was to apply 5000 pounds to every square foot. Did I do it wrong? This should be roughly 500 GPa total. Enough to significantly load it.

    <For what I see you're using the SimSolid solver, are you using Adapt for Stress when running the structural analysis?>

    Yes SimSolid. The geometry is far too irregular to make meshing it practical according to the tutorial that I watched, which had to break a simplified rocket nozzle into four (or was it eight?) parts to make it mappable and mesh it. Iterative designing for this geometry would surely be impractical. If you mean did I change the default "for stiffness" radio button to "for Stress" before running the analysis, then yes.

    Regards,

    Paul

    February 1st update:

    I set some of the options that I had changed back to default, and ran some more tests

    Just going by how the results spectrum had been appearing, I increased the pressure by 500 times, and then the % of yield results spectrum scale appeared closer to what I would expect (at 1/500th of that applied load). Of course, these results are unusable since that load is invalid.

    Also, I included buckling analysis on that run, and the buckling load factors appear that they might be in the realm of possibly at the higher pressure. For three modes, the BLFs were 0.011, 0.011, 0.085., which i think would correspond to a BLF of about 5 if the correct pressure were applied, and simple multiplication is a valid method of comparison.

    The Max Shear Stress results are very unexpected. The stress is at maximum when the model is not deformed, and at what appears to be zero when the deformation is at maximum. This seems backwards, and I don't know what I could have done to cause that. I'm starting to think that this is a bug.

    Another odd thing, is that where the pressure is actually applied is showing no stress. However the stress does transfer across the bonded contacts to the frame, which is showing stress, and causing the expected deformation.

Categories