General Mechanical

General Mechanical

Modeling straight-braid McKibben Muscle – Contact Issues

    • lwarnerx
      Subscriber

      Hi,


      I'm trying to model a McKibben muscle - or pneumatic artificial muscle - in Mechanical and I can't seem to figure out how to get the contacts correct between the parts in order for the FEA to run. Right now, there are 20 strands that run along the tube/bladder that is capped at both ends. There is an interior pressure in the tube with one capped end as a fixed support and the other capped end as allowing displacement only in the z-direction. The tube is a Mooney-Rivlin hyperelastic material, the caps are stainless steel, and the strands are carbon fiber. The idea is for the inner tube to expand and the strands cause it to shorten when it pushes against them. I'm not sure what settings I need the contacts to be in order for it to work. 


      Thank you!


    • Aniket
      Ansys Employee

      how are you modeling the strands? Do they have uniform crosssection? Are they modeled as beams?


      If yes you can manually define contact between the line bodies and Face of the bladder, This can be bonded initially, and once that is solved you can move to frictional which will be more likely in actual case.


      -Aniket


      How to access Ansys Online Help Document


      How to show full resolution image


      Guidelines on the Student Community


      How to use Google to search within Ansys Student Community


       

    • lwarnerx
      Subscriber

      They are uniform circular cross-sections. I'm not sure what you mean by being modeled as beams.


      I ran the simulation with the strands bonded and that worked and had no issue, but when I try to change the contact to frictional, I get warnings of contact status has abruptly changed, my bladder element has become highly distorted, etc. and the solution is terminated due to "Error in element formulation". I have Pure Penalty on with a 5e-5 tolerance to prevent penetration. Any ideas what is going wrong after changing it from bonded?

    • Aniket
      Ansys Employee

      are the strand modeled as beam elements or as solid elements?


      Also, are you properly fixing the ends of the strand? try to keep algorithm as default and use smaller factor for the normal stiffness of the nonlinear contact such as 0.01 along with smaller time steps and see if that helps?


      In 2020R2 (which will be available for student community shortly) has new improvement for contact detection that would help too.

Viewing 3 reply threads
  • You must be logged in to reply to this topic.