General Mechanical

General Mechanical

Seal compression convergence issue.

    • Chinmay

      Dear All,

      Greetings of the Day.

      I am doing a project where I wish to find out the seal compression and the contact pressure between seal and housings. Please find attached images for reference. I have used 2 plane symmetry thus only quadrant is shown below.


      All except Seal have material as steel. Seal is Silicone rubber. Behavior of all parts is selected as flexible. No NLAD meshing used.

    • peteroznewman
      You can save time waiting for the solution by changing all the steel parts to Rigid behavior.
      Using Frictionless contact hurts the ability of the solver to converge and is unrealistic. You are better off using some small amount of friction. Very high friction is also difficult to converge.
      Here is the solution with the plunger 5 mm down.
      You were going for 7 mm of displacement, but nothing changes, except without friction, the seal wants to jump vertically.
      Attached is a model that uses friction of 0.1
    • Chinmay
      Thank you for your valuable help, I have few follow-up queries for this:
      1) I inserted pressure in contact tool as shown below, so my question is if this contact pressure is seal pressure (Amount of pressure required to overcome the sealing) ?
      2) In your APDL commands, you added 'neqit, 100' ; I guess default is 25 ?
      So how did you decide how many equilibrium steps you need to solve this particular problem ?
      3) Can and should we use NLAD meshing for seals in general ?
      4) In total deformation, I need complete seal deformation, thus I made changes in symmetry settings accordingly but unfortunately the result is not as expected.
      and lastly 5) You added "nd001_MXUP_Elements" & "nd002_MXUP_Elements" in the solution information, I did not find this on any online resource available, could you please tell me what was the purpose for doing this and what conclusion do we get from it or how do we infer the result obtained in this step ?

      Thank you and sorry for so many queries Chinmay
    • peteroznewman
      1) Pressure in the contact tool is contact pressure. That is the normal force of the seal pressing against the rigid part per unit area. That is not the air pressure that the seal would support before leaking. The question of how to determine the air pressure that a particular seal will support is worth its own separate discussion.
      2) Yes, the default maximum iterations is 26. It doesn't hurt to put NEQIT, 100 into the model. If a substep converges in 7 iterations, it increments the load.
      3) I never use NLAD. I have not found it necessary and I have not found it helpful.
      4) I don't understand what you changed and what went wrong.
      5) Click on the Solution Information Folder and see in the Details window the is a line to put in N-R Force Residual Plots and a line to put in Deformed Elements. When you put a non-zero number on this second line, Mechanical automatically creates a group of the Highly Distorted Elements found during the solution. If the elements get too highly distorted, the solution will stop with an error. Other times, the solver might bisect the increment and try again with smaller substeps and be able to keep going.
    • Chinmay
      Thank you for this wonderful explanation & solution and also spending you valuable time helping all of us here on this platform. You are awesome. :)
      P.S. I will make a new discussion on "how to determine the air pressure that a particular seal will support" if there is none here or other internet sources.
Viewing 4 reply threads
  • You must be logged in to reply to this topic.