General Mechanical

General Mechanical

Shell model contact trouble

    • ValiullinDamir
      Subscriber

      Hello, dear colleagues!

      Your help is needed: i cannot solve my problem with contact in shell model.

      I have a quite big assembly with bolt pretensions. At the first step, pretension loads are applied, and at the second step they are locked and the external loads are arise. All the contacts are frictional and all the details work together because of the bolt pretensions. The whole assembly consists of shells created by Midsurface option in SpaceClaim (i have solid model created in Solidworks). All the bolts are beams and they are connected to shells like @peteroznewman did it in this video:

      ","embedType":"youtube","name":"Shell Mesh Joint with Bolt Pretension","frameSrc":"https://www.youtube.com/embed/tQykfXMuQp4?feature=oembed&autoplay=1"}">
      ">
    • peteroznewman
      Subscriber
      I suggest you use Bonded Contact on one side of that detail piece to bond it to one flange. The other side of that detail piece can retain Frictional Contact with the other flange.n
    • ValiullinDamir
      Subscriber
      thanks for your reply! I made it as you suggested - changed some of the contact types to bonded. Now i have a good behavior in this zone, but two another contacts (created manually) dont work - i have big displacements:nAnd how it should be:nn
    • peteroznewman
      Subscriber
      Insert a Contact Tool under the Connections folder and Evaluate Initial Contact Status. You can remove the Linear (Bonded) Contacts and only look at the Nonlinear (Frictional) Contacts. They should all be Closed assuming you have no deliberately open gaps.nIf that is the case, then under Analysis Settings, turn on Automatic Time Stepping and set the Initial Substeps to 100. That may fix this problem. If it does not, try 1000.n
    • ValiullinDamir
      Subscriber
      I have checked Contact status in Connections folder - and all the Contacts are closed. Now i started the Solution with 100 Initial Substeps - it will take quite a long time, but Total Deformation plot tracker shows, that we have penetration at this zone:nSo, it is a really strange thing - when i chose Bonded type for manually created Contacts - everything is fine (there are no penetration and flying details), but when i chose Frictional type for them - such a thing occurs. n
    • peteroznewman
      Subscriber
      It's not strange at all, it is expected. nBonded contact is linear, which means it is easy for the solver to find equilibrium.nFrictional contact is nonlinear, which means it is difficult for the solver to find equilibrium.nIn complicated models, extraordinary steps are sometimes required to get nonlinear contacts working.nBut first, check that you actually have large penetration or is it in fact just an exaggerated displacement display scale? When you click on the Results tab, there is a displacement display scale area on the ribbon. The display scale defaults to an automatic value, which might be thousands of times larger than reality. Change that setting to 1.0 (True Scale). Did the parts return to their normal positions?nIf you were already showing True Scale plots, then make sure you have selected the correct side of the shell to make contact with. Maybe Bottom is wrong and it should be Top or vice versa.n
    • ValiullinDamir
      Subscriber
      thank you for help!nIn my case, it was penetration, not big displacement. The last picture i have attached was made with True Scale size. nChoosen shells sides were correct too. nThe problem was with small pinball radius. I increased it to the value that 0.1 mm bigger, than distance between midsurfaces - now solution goes right. n
    • peteroznewman
      Subscriber
      Are you saying a Contact Tool evaluated this contact as Closed before you increased the pinball radius? nOr did you just miss that this contact was Open and increasing the pinball radius made it Closed?n
    • ValiullinDamir
      Subscriber
      As i remember, all the Contacts before were closed.nThen i increased pinball radius, and now i have such a situation:nManually detected contacts are orange. They have small ammount of penetration, but i think, that this approximation can be done. nAnd as i understand now, the whole problem was with manually detected contacts - there are really small gaps at these zones, and because of these gaps contacts were not created automatically. n
Viewing 8 reply threads
  • You must be logged in to reply to this topic.