Learning Forum

[Khatibo]

 

Hi,

I am having a hard time simulating the elastic deformation of a sphere press fitted into a plastic cylinder. I have used a frictional contact between the surface of the sphere and the inside of the cylinder. I have only changed the behavior under the 'Definition' settings to asymmetric. I have set an displacement on the cylinder so it moves into the sphere. The sphere has only a frictionless support on its surface. I really do not know how to simulate this simple movement. The results i get always end up end some sort of explosion.

Thanks in advance!

 

[peteroznewman]

Here is a close up of your photo attachment.

I recommend you create a 2D axisymmetric model. That means take a radial slice through your geometry.

The axis has to be the Y axis and the slice has to be drawn on the +X side of the Y axis on the XY plane. You will end up with two surfaces: one C-shape for the ball and a rectangle with the large radius corner.  Your 3D geometry has the axis of the tube along the Z axis, you you will have to rotate that to be along the Y axis.

      

Create a new Static Structural model.  Pick the Geometry cell and RMB to get the Properties. Set the Analysis type to 2D.  Draw those two surfaces. Open Mechanical and click on the Geometry branch and set 2D Behavior to Axisymmetric. I assume you are setting Large Deflection On.

What are you using to pull the ball into the tube?  If you split the ball in half using a plane parallel to the XY plane, you can apply a down force to the half of the ball that is "inside" the tube.  Using only a force makes the simulation more difficult. It is better to have a displacement input to control the position of the ball.

You might benefit from using Hyperelastic material properties for the ball. When you do that, there is also a setting that can help. Set the Brick Integration Scheme to Reduced Integration.

[Khatibo]

Thanks for the reply!

I forgot to mention that the sphere is already defined as a hyperelastic material, rubber. I am also importing the geometry from an assembly made in Creo, it is a simplified model of the model i really want to simulate. Because of this I would rather try to solve the case in 3D (if possible?). For the pulling, I am only using a displacement to move the plastic cylinder into the rubber ball, the other way around where I fix the cylinder and move the rubber ball result in an excessive deformation as the ball "explode". I am also having trouble defining the constraints for the ball when moving the cylinder into it.. Bear in mind that I am new to ANSYS workbench and have only worked with APDL.

That said is the method still applicable? 

[peteroznewman]

In my 2D model, the pipe had zero Y displacement. The axial cut edge of the sphere had a Y displacement and I applied a force downward on the bottom half of the sphere.

It's fine to have a displacement of the pipe, but what is holding the ball in place?  Is there a surface on the opposite side of the ball from the pipe that the ball is pushing against when the pipe moves toward the ball?  That would be fine, but it is going to squash the ball and make it harder to get into the pipe.

Do you need to see the ball entering the pipe or are you only interested in the ball inside the pipe with the contact resolved?  It's more computation to move the ball into the pipe than just resolving contact of a ball that is already completely inside a pipe.

You can solve in 3D but you will pay a price in solve time. What feature of the 3D model is not axisymmetric?

[Khatibo]

It was also the initial idea that the ball would be pushed into the tube, by a displacement and fixing the tube in its end, but I could not make it work as the ball seemed to explode or that the ball just move away although it is constrained... Yes, the reason for the simulation is to see how the ball deform while it is getting pushed into the tube. The solution also need to be in 3D. To the last question, nothing that was a mistake made.