General Mechanical

General Mechanical

Convergence of hyperelastic simulation

Tagged: 

    • andrea giordano
      Subscriber

      Hello, 

      I am running the static structural hyperelastic simulation of a soft wing, but I cannot reach convergence due to a "highly distorted element". Despite I have followed all of @peteroznewman on how to make hyperelastic simulations converge, it is difficult to mesh my domain with Solid185, and most of all my domain is not axis symmetric. THerefore, I cannot use Keyopt 6 or others. I tried many strategies, as forcing the mesh to use Solid185, changing my curved domain to a segmented one... However I still get the same error, despite the 100 substeps employed during simulation. Could anyone or maybe Peter help me? Thank you very much. I can share link with simulation archive:

      https://drive.google.com/file/d/18nqKTMepR87w9IYtF61Vuvs3WjBb2aeu/view?usp=sharing

    • peteroznewman
      Subscriber

      Hello Andrea,

      I’m looking at your model now, and thinking about ways to simplify the meshing.

      The Inextensible layer has many holes. What is the function of the holes?  Can the holes be eliminated from the geometry?

      The inextensible layer is very thin. Could this thin solid layer be replaced with a surface model?

      In pink are the cavities where the pressure is applied. In Static Structural, there is only pressure applied to the top side of the wing, is that to get it to change shape?

      I noticed Frictionless contact used. Friction can help convergence. However, that won’t fix the highly distorted element error. Better element shapes may be what is needed.

      One reason models fail to converge is due to mistakes in the material properties. Here is the Ecoflex 00-50 properties entered in the model.

      Those numbers seem way too small. Were they entered in the wrong units and are off by a factor of 1 million? This would be the cause of the immediate error messages of highly distorted elements.

      I found a Masters Thesis on this material. This has good information and you could use this Mooney-Rivlin data.

      https://www.politesi.polimi.it/bitstream/10589/187294/4/Thesis_Lavazza.pdf

      If you would like to discuss this further, I can set up a meeting where we talk and share our screens. I’m in the Eastern US Time Zone.

      Regards,

      Peter

    • peteroznewman
      Subscriber

      Hello Andrea,

      I decided to apply Standard Earth Gravity to the wing. You can see how the Eco-Flex part of the wing sags by 13 mm at the trailing edge when there is no pressure in the wing.

      When a pressure of 4.e-2 MPa is applied to the chamber on the top of the wing, the trailing edge goes down to 26 mm.

Viewing 2 reply threads
  • You must be logged in to reply to this topic.