Jump to content


  • Content Count

  • Joined

  • Last visited

Posts posted by Felix-Th

  1. Hello everyone,


    I'm trying to set up a direct transient analysis of the beam you see in the picture below (simplified geometry):

    • The beam is meshed with tetra elements.
    • On top the beam is constrained in all dof.
    • In the middle of the beam is a hump and the mass of it has an influence on the eigenfrequency of the beam.
    • At the lower end of the beam a periodically enforced displacement in the range of the eigenfrequency (3 kHz < f < 4 kHz) of the beam is applied.


    With the FE analysis I want to get information about the following questions:

    1. What is the stress at the top of the beam? Later I would like to get information about fatigue.
    2. What is the eigefrequency of the beam?
    3. What is the exact displacement of the hump in the middle of the beam when a certain displacement is applied at the lower end with its eigenfrequency?


    So far I've tried to adapt the tutorial "OS-T: 1310 Direct Transient Dynamic Analysis of a BracketOS" to my purposes but I have problems with applying the periodically enforced displacement. Can someone tell me what the best way is to realize this problem with optistruct?


    Thank you in advance



    Link to the tutorial:


  2. Hi

    could you please tell me how you solved the problem? I'm getting the same error but I think I don't have any prescribed displacements in my model. I'm also using weighted compliance and vol Frac as responses. The optimization is running but as soon as I set a max stress I'm getting this error.

  3. Hi,


    Due to contacts in my model I wanna let my optimization run nonlinear. So far the linear analysis and optimization works. To let the analysis and optimization run nonlinear I created a load collector with the card image NLPARM. Additionally I created a non-linear quasi-static loadstep. The nonlinear analysis gives me results where as the optimization does not give any result even though I can let the optimization run without any error.


    Here I'm uploading a very simplified model that is compatible to the HyperWorks student edition. The contacts are not implemented yet. But there should be contact between the green, the yellow and the blue component. Maybe someone can tell me by which criteria I have to decide if the contact partner is master or slave.


    I'd be glad to get any advice how I can set up a nonlinear optimization with contacts in Optistruct. My question is: which steps do I have to carry out to turn a linear into a nonlinear optimization. Unfortunately I couldn't find that much information about it.




    P.S. The model shows a medical implant for connecting bones. Maybe someone is wondering what it is ;)


  4. Hi Prakash,


    that's the error I get:


    *** ERROR #   99 ***
      CBUSH           element     576893 references incompatible PBUSH.
      K2/M2/B2, K3/M3/B3, K5/M5/B5, and K6/M6/B6 on PBUSH must be zero for
      CBUSH with no G0, CID, and blank X1, X2, and X3.


    As soon as I create a vector and select it while crating the spring it works. But what does that vector mean?

    jker likes this

  5. Hello,


    in order to get a smooth stress distribution with no spikes, my plan is to constrain a surface with RBE3-Elements to an SPC. Since it is not possible to directly connect the dependent nodes of the RBE3 to the SPC I thought I could use a spring (cbush) between the RBE3 and the SPC. As long as K1 and K4 of the pbush property of the spring are rigid my simulation runs. But obviously the spring will be deformed when K2, K3 K5 and K6 are zero. Nevertheless I get a smooth stress distribution. Making all six Ks rigid gives me an error. Using an RBE2 directly connected to an SPC gives me spikes.


    So my questions are:

    - how can all 6 Ks of the pbush be rigid without getting an error?

    - how can I connect an RBE3 to an SPC (is there a better way than using a rigid spring)?


    The attached screen shot shows the problem.


    Thank you in advance.




  • Create New...