How do I resolve solution results failing to read/write in my Explicit Dynamics model?

NicDMacNicDMac Member
edited December 2020 in Explicit Dynamics

Greetings all, see below for the .zip attachment of the program I am working with and its associated files. Also, see below for the error readout I am observing.

I am running a basic cutting tool model whereby a single-body "cutting head" rotates and translates through a single-body block of soil. I have ensured no geometry failures have occurred using DesignModeler repair tools, and all procedural steps have been cross-validated with simpler models. The errors I am seeing when I click "Solve" are as follows:

Things I have already tried, especially with help from related forums:

  • Turning off "Distribute Solution" in the Advanced Solve options menu
  • Ensuring the "Default Folder for Temporary Files" in Tools -> Options carried read/write permissions
  • Ensuring all result objective "End Time" values were set to "Last"
  • Increasing the number of result steps in the Analysis Settings to ensure convergence
  • Clearing Generated Data at different combinations of steps throughout the analysis procedure

I look forward to determining the root cause here. Thank you for your help in advance!

MODEL:


Best Answer

  • peteroznewmanpeteroznewman Member
    edited December 2020 Accepted Answer

    @NicDMac I opened your model.

    First observation is the ball is far away from the soil. Move the ball so that it is tangent to the soil. You don't want to waste time computing the ball approaching the soil.

    Second observation is the End Time of 20 s is very long for Explicit Dynamics. It will take days (or weeks) to compute that. For debug purposes, I changed the End Time to 0.01 s.

    Third observation is under Analysis Settings, the Output Controls, the Result Number of Points was left at the default of 20. I changed that to 2000.

    I didn't move the ball to be tangent, but with the other two changes, it solves just fine. After 3 minutes of computation, the ball is spinning with the angular velocity specified and slowly closing the gap to the soil.

Answers

  • @NicDMac

    This model has no load applied. Only a support.

    This model also has way too much detail in the cutter that creates unimportant small elements that will push the time step down to such small values that even when there is an applied load, it will take months to solve. Also the element size for the soil (light grey) is too large.


  • @peteroznewman After reading your suggestions, I went back and heavily simplified the cutter geometry (essentially, now it is a cylinder with a sweep block instead of cutting tools). I also decreased the mesh to the minimum resolution and applied a sizing to the block to make the mesh spacing smaller than that of the cutter. Keeping all else the same, I still received the same result output errors.

    I then decided, "Hey, let me see if I can run the most basic Explicit Dynamics model out there". Attached is a screenshot of this test model - essentially a steel ball crashing into a thin plot of soil. As you can see, I still run into the same result output errors. My node and element counts are silly low for my Academic license (sum of just above 10K), and the mesh generation runs lightning fast, so performance seems very far out of the question here.


    I'm at a loss for words as to why I'm getting these errors. Is there some sort of internal corruption to where others also have issues writing the result files when I have them open my sim?

    Thank you for the help!

  • @NicDMac

    If you delete the Equivalent Stress from the solution, and create a new one, does it show a result?

    If not, under Analysis Settings, Output Controls, set the Number of Results to 1000 instead of 20, does the solution run?

    If not, please use File, Archive in Workbench to save a .wbpz file and attach that to your reply so I can have a closer look.

  • @peteroznewman Tried both in the past and just now again to no avail. Archive file attached here. Thanks again!


  • peteroznewmanpeteroznewman Member
    edited December 2020 Accepted Answer

    @NicDMac I opened your model.

    First observation is the ball is far away from the soil. Move the ball so that it is tangent to the soil. You don't want to waste time computing the ball approaching the soil.

    Second observation is the End Time of 20 s is very long for Explicit Dynamics. It will take days (or weeks) to compute that. For debug purposes, I changed the End Time to 0.01 s.

    Third observation is under Analysis Settings, the Output Controls, the Result Number of Points was left at the default of 20. I changed that to 2000.

    I didn't move the ball to be tangent, but with the other two changes, it solves just fine. After 3 minutes of computation, the ball is spinning with the angular velocity specified and slowly closing the gap to the soil.

  • @peteroznewman It appears I get the same results as you with these changes. I will make the cutter tangent to the soil plot, thin the soil plot by a sizeable margin, reduce the runtime heavily, and maintain a boost in the number of points in the original model. After that, I'm okay with something along the magnitude of a 24hr run. Thank you for your help!

  • @NicDMac

    I reran with the ball starting tangent to the soil.

    Since the soil is not fixed, but floating in space, the impact of the ball with the soil pushes the soil in the -x direction. Since the ball only had an initial velocity, it bounces off the soil and ends up traveling in the +x direction.

Sign In or Register to comment.