Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Unexpected changing of the geometry in ANSYS Workbench

  1. Jan 31, 2016 #1
    Hello!

    I started recently to use ANSYS Workbench 15 to solve a stiffness problem of a three parts assembly. I have a general knowledge about FEM. I create the geometry in the Design Modeler by inserting the assembly from Solidworks 2010.

    It is a three parts assembly like a bicycle's wheel. Two parts are cylindrical and concentric, like the outer rim and the inner component where the axis is. The third part is a few linking thin plates, that link the two abovementioned components, like the spokes do in a bicycle wheel. The scope is to set the outer rim steady and by revolving the inner cylindrical component, force the "spokes" to buckle (the assembly is designed appropriately and the plates should buckle).

    But during the solution step, I have an unexpected behaviour. The inner cylindrical component is changing its size by getting bigger and bigger as it is revolving. The plates are buckling as they should do and the outer rim stays steady. But of course the solution is useless. The results that I am looking for is the stiffness behaviour of the buckled plates.

    Thanks in advance!

    Vaggelis
     
  2. jcsd
  3. Feb 1, 2016 #2
    Hello,

    my first guess would be some problem with element distortion. Is there any similar message in your error file ?
    Maybe denser mesh could help. If not, then boundary conditions can be the problem. How is rotation applied on inner cylinder ? If you have rotation prescribed on its axis, try to apply it on cylindrical surface instead.
     
  4. Feb 18, 2016 #3
    I am not familiar with the "element distortion". But no, as I remember no message like this came up. Rotation is applied by Remote Displacement on cylindrical face. As you wrote, boundary conditions should be a matter to investigate. I did add some frictionless supports, I had a different result, with less unexpected deformation of the cylindrical part, but still it was deformed and the thin plates were not buckled as they should. Ansys2.png

    As I search through tutorials and videos, I start again new projects to solve my assembly with some small changes every time. In last attempt I changed the Large Deflection On. The solution didn't complete, due to convergence error: "The solver engine was unable to converge on a solution for the nonlinear problem as constrained. Please see the Troubleshooting section of the Help System for more information".

    Any ideas?
     
  5. Feb 19, 2016 #4
    Hello,

    if I see correctly, you did not use any mesh control. This can be ok for some cases, but if there are problems with convergence, I would try to refine mesh and /or reduce time step.

    I suppose one of frictionless supports is for inner cylindrical face. If that is the case, revolute joint from body - ground category may be better.
    What is the second frictionless support for ?

    Before that convergence error, were there any warning messages ?

    It seems like all your thin plates are held only by frictionless contact. I am not sure if this is the appropriate condition, and it may be the source of some problems. What about bonding them to outer ring or using body - body joints ?
     
  6. Feb 28, 2016 #5
    Gobi,

    Thank you for your time!

    I did follw your advice and substituted half of the frictionless contacts of the thin plates with body - body joints with the outer rim. Although I don't understand the difference between these two types of constraints. I also removed the frictionless support for the inner cylindrical part and defined a revolute joint as body - ground. The other frictionless support that you see is set to define no displacement on the Z axis.

    I was trying to figure out the suitable properties of the meshing control. I still don't know what method would be appropriate for the thin plates. I chose the tetrahedrons but it didn't work. Sweep method also didn't complete, due to convergence error.

    I leave in Greece, Athens. How could I find someone here to discuss about ANSYS, do you have any idea? I did search on web but nothing important came up.
     
  7. Feb 29, 2016 #6
    Hello,

    In your case, i would be more concerned about mesh sizing than method, since you have long and thin parts, potentially leading to elements with bad aspect ratio. But rhat is just one possibility. There are several diferences between frictionless contact and joint, for example, this kind of contact only restrains movement perpendicular to contact surface and only in inward direction. Part only constrained by this can just fly away, under some circumstances. Also, contact is more computationaly demanding.

    Anyway, I am sorry I was not able to help.

    About ansys counselling in Athenes, there is one company named SimTec, which, if I undestand correctly, is local Ansys provider.
    There should be also at least two engineering universities in Athens, and Mechanical engineerinfg departments are very likely to have someone familiar with FEA.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook