Deformation through movement through small surface

  • 121 Views
  • Last Post 08 August 2019
AnsysKIT1 posted this 05 August 2019

Hi there,

I'm new at the forum, so at first hello to all.

 

My question is how I can map the following problem:

Simplified I have an elastic geometry/volume that I drag through a smaller surface (cylinder).




I'm interested into the deformation and the stress during the procedure.
(The bigger sphere will deformed during the movement though the cylinder.)
 
I currently have no idea how to image that at Ansys.I tried explicit dynamics but I only see the possibility to have an inertial speed, but I need a continuous speed. In addition, the part I want to drag through the cylinder ricocheted.

I also added the mechdat file (in a zip folder) for an import, if you want to have a look. I hope that someone of you can help me that would be awesome.


Many thanks for all help!


Best regards Patrick

Attached Files

Order By: Standard | Newest | Votes
jj77 posted this 05 August 2019

Just use an enforced displacement (Supports/Displacement) on one end, and pull it through - define a table say with 1 mm per second (which gives a constant speed of 1 mm/s). You can of course choose what ever you want so you get the correct speed. 

  • Liked by
  • AnsysKIT1
peteroznewman posted this 05 August 2019

Adding to what jj77 said, in addition to using enforced displacement, you might consider moving this from Explicit Dynamics to Static Structural if you are only interested in compressing the sphere until it is in the center of the ring.  If you want to see it shoot out the other side, then it's best to stay with a transient solver.

If you want to do Static Structural, there are a few changes to be made in CAD.  You want the sphere in initial contact with the edge of the cylinder, so that the contact between them is closed.  Another CAD modification is to put a blend radius on the sharp edge of the cylinder. That will provide a surface to surface contact definition which is much more likely to converge than an edge to surface definition.

Finally, this is an axisymmetric problem, so orient the parts so the axis of the problem is aligned with the global Y axis, then slice the model though the X and Z planes so you can have a surface that is a radial slice on the +X side of the geometry. Suppress for Physics all geometry except for this surface. In Workbench, you have to set the Geometry property to 2D, then drag out a fresh Static Structural block to drop on the Geometry cell.  In Mechanical, there is a setting under Model for the 2D analysis type, set that to Axisymmetric.

  • Liked by
  • AnsysKIT1
AnsysKIT1 posted this 05 August 2019

Hi guys, thank you for the fast response!

This is just a simplified geometry to check if it works. My main part more difficult and not symmetric.
For my case, it is necessary to see it shoot out.

Now I know how to map the speed, but I don't get a deformation when the sphere passes the cylinder.

It slides through the cylinder like a ghost: What could be wrong?



Btw. In the final model, I also have a friction between the parts when it passes, how can I map that?

 
I've set the parts close together and selected a friction contact between a blend and the sphere. But when I set a friction contact it doesn't run.


Again many thanks for your help!
Nice to have a place for such questions.

Attached Files

jj77 posted this 05 August 2019

To attach models see here (your attachment is not of any use):

 

https://studentcommunity.ansys.com/thread/saving-sharing-of-working-project-files-in-wbpz-format/

  • Liked by
  • AnsysKIT1
AnsysKIT1 posted this 06 August 2019

 

Hello

 

I've tried to solve the problem with "Explicit Dynamics", "Transient Structural" and "Static Structural". But I believe I've done something wrong...

 

I've uploaded/attached it in the correct way, thank you for the information.

 

Many thanks.

 

Best regards Patrick

 

 

 

Attached Files

jj77 posted this 06 August 2019

In explicit you have few cycles so it can not run the 10 s. 

 

Just to see that it works I reduce the final time to 0.01 s, and then increased the number of cycles to 1E8 cycles.

 

With that and making the part that you pull shorter, because other wise that part will just stretch, it solves fine in explicit.

 

  • Liked by
  • peteroznewman
  • AnsysKIT1
AnsysKIT1 posted this 06 August 2019

It sadly dosen't work for me with that settings...

Do you have changed something else?

I only reach 35 mm by 2,358e-003 secounds. The end time should be 0,01 and the Maximum Number Of cycles is 1e+08.

I attached the new file, with the changed settings.

Many, many thanks again.

 

Attached Files

jj77 posted this 06 August 2019

I suppressed the contacts and used only the explicit body interaction / frictional- I do no think the frictional coefficient is so high at least when it slides so I reduced that I think to 0.2 and 0.1

  • Liked by
  • AnsysKIT1
AnsysKIT1 posted this 08 August 2019

Hi

do you (or someone else) get a realistic solution?

I've shorten the part that I pull very much, but it only run through, when I change the minimum step time to a very low number.
But the results make not much sense.
They look like that:

 

Or it stops with the error 'time step too small' at that point:


Could you maybe attach your file, that I can compare the settings and run it?

Many, many thanks again!

Best regards

Attached Files

Close