Simulate Clamping a Tube Shut

- May 21, 2021 at 11:42 pm
  
  dtocol
  Subscriber
  
  Our model consists of a 12 inch hollow tube with a radius of 0.75 inch in between two rectangular bars 3 inches long by 0.5 inches wide. The tube models a fire hose with isotropic elasticity consisting of an elastic modulus of 60 MPa and Poisson ratio of 0. The goal is to simulate the upper bar clamping the hose shut with pressure in the tube, eventually bringing the two bars together while keeping them in parallel. To do this, we applied a 300 psi pressure normal to the surface in the tube and a displacement of 1.5 inches on the top face of the upper bar. The faces of the ends of the tube and the bottom face of the lower bar are fixed supports. However, we are experiencing an issue with the solver regarding boundary conditions and large deflection, and do not receive results. Our question is how to model the assembly with the correct boundary conditions for the simulation to run.
- May 24, 2021 at 1:00 pm
  
  1shan
  Ansys Employee
  
  Have you turned large deflection under analysis settings to 'On" ? Also, a zero Poisson's ration might create problems. Give a small value and try rerunning the analysis.
  Regards Ishan.
- May 24, 2021 at 11:37 pm
  
  dtocol
  Subscriber
  
  I set a Poisson's ratio of 0.25 and turned on large deflection, but that did not seem to work either. These were the new errors I was receiving after the change. Any help would be appreciated, thank you.
- May 25, 2021 at 3:09 am
  
  peteroznewman
  Subscriber
  
  I recommend you create a hose using a cylindrical surface instead of a solid. If you use a solid, it should have at least 4 elements through the thickness. If you use a surface, then a simple shell element will be used that has the assigned wall thickness.
  I recommend you add a large radius to the edges of the square bars, the radius might even be so large as to turn the square bar into a cylinder. I also recommend you split the hose at the bars and bias the element size so that many elements are used to wrap around that radius, at least 4 elements.
  I assume you have turned on large deflection. You also need to turn on Auto Time Stepping and change the Initial and Minimum Substeps to 100 and the Maximum to 1000.