General Mechanical

General Mechanical

Topics related to Mechanical Enterprise, Motion, Additive Print and more.

About convergence issue of simple 3D model applied large deflection option

    • JooJaehoon
      Subscriber

      Hello!

      I'm conducting compression test of quarter cylinder

      It is composed by upper press, lower press and quartercylinder

      Cylinder has 8 mm height and it compressed into 4 mm by upper press

      Cylinder has non-linear properties(attatched as picture) and upper, lower press has rigid body(set extreme high young's modulus, poisson ratio = 0)

      So, I think deformation has a large value so large deflection option must be checked

      Analysis started and it's convergence failed

      Error occured at convergence gauge at 16% and I attached error result picture

      I attached analysis condition picture too

      I raise initial substep into 1000 but it was failed too

      With unchecked large deflection option, convergence completed

      How can I converge this problem?

      regard

    • peteroznewman
      Subscriber

      The model had some converged substeps. Select the last converged subsetp to plot the deformation. That is the row above the End Time in the Tabular Data.

      Under Analysis Settings, click on the Solution Information folder, and in the Details window, select Force Convergence and reply with a screen snapshot of that graph.

      Then change from Force Convergence to Solution Output. Click ctrl-F to find the word error. Read the text around that word, Copy paste the text into your reply.

      Set the Initial and Minimum substeps to 1000 and solve that.  With plasticity, it is important to prevent the solver from taking large steps late in the solution time.

    • JooJaehoon
      Subscriber

      Hello! Peter!

      Thank you for applying!

      The first picture is force convergence graph

      And, the second and third are error message in Solution Output tab

      regards

       

    • peteroznewman
      Subscriber

      Was that force convergence plot from the run with the Minimum Substeps set to 1000?

      Next you should look at the last converged substep. We can see the time is about 0.11 which is the fraction of the 4 mm total displacement or 0.44 mm that the top plate moved before elements became too highly distorted to continue the simulation.  I suggest you change the displacement of step 1 to 1 mm and add a second for 2 mm, and so on.  Keep the initial and minimum substeps at 1000 for step 1. Does the simulation get any further down? 

      From the error message, you have an element number to look at. Create a Named Selection so you can isolate that element and watch how it deforms over the time steps that converged.

      In the unconverged displacement plot, I see some elements along the center axis seem to have buckled. Are the nodes staying on the two symmetry planes you defined? Change the contour option to Capped Isosurface so you can see the elements that had displacement > 0.1.

      • JooJaehoon
        Subscriber

        No, it's rest I've post first

        But, I try to analysis with applying substep as 1000, but result still have not converged with distortion of element.

        Now, I set displacement step as more than 1 and will conduct analysis.

        As soon as end of analysis, can I ask a question with new result?

        I'm appreciate of your help.

        Have a nice weekend!

    • peteroznewman
      Subscriber

      Yes, reply with new results when you get them.

Viewing 4 reply threads
  • You must be logged in to reply to this topic.