[ajelahe74]

Hello!



I have seen the tutorials for the punch and sliddingFrcitionBall cases, as I understand there are different patches for every boundary condition: the contact boundaries (both master and slave), the zero solid traction, the valued solid traction and the fixed displacements. As if every geometry has a thickness defined by this patches and the symmetry planes proposed there.



I'm trying to simulate a "very simple" case of solid contact. It mainly consist of an sphere located inside a cylinder. The idea is to "inflate" the sphere from within and see how that affects the cylinder wall. So, I have four patches: inlet, outlet, sphere and cylinder.



My question is, can I use several boundary conditions in a single patch? Say... A solidTraction pressure inside the sphere patch (to "inflate it"), and at the same time it would have to be the contact master BC. Analogously, the slave contact as well as the traction free would be applied to the cylinder wall.


Is that possible? Or should I have different patches for each BC, made with offsets of the original surfaces, as a thikness?


Not sure if I wrote the question clearly. But, thank you beforehand!

[bigphil]

Hi Alejandra,

A boundary can only have one boundary condition. So if you can logically split the patch into two separate patches, then that may be the easiest solution.

From the description of your case, it sounds like the outer surface of the sphere will come into contact with the inner surface of the cylinder; is this correct?
In that case, you could set the master as the cylinder inner surface, the slave as the sphere outer surface, and apply a pressure via solidTraction on the inner sphere surface.

If I misunderstood, maybe you could show a schematic of the case?

By the way, I suggest taking advantage of the symmetries in this case. It will help avoid rigid body translation as well as being faster.

Philip

[ajelahe74]

Originally Posted by bigphil

Hi Alejandra,

A boundary can only have one boundary condition. So if you can logically split the patch into two separate patches, then that may be the easiest solution.

From the description of your case, it sounds like the outer surface of the sphere will come into contact with the inner surface of the cylinder; is this correct?
In that case, you could set the master as the cylinder inner surface, the slave as the sphere outer surface, and apply a pressure via solidTraction on the inner sphere surface.

If I misunderstood, maybe you could show a schematic of the case?

By the way, I suggest taking advantage of the symmetries in this case. It will help avoid rigid body translation as well as being faster.

Philip

Indeed, Philip, here I attach the geometry I have. Also, if every patch has to have one BC, I would have to make like offset patches from what I have in order to apply the BC correctly, if I understood your explanation. I made an sketch of this, maybe you could tell me if I got it wrong.

[bigphil]

I am not sure I understand your schematic.

In the first image, it looks like a solid cylinder which has a sphere cut from it. In that case, no contact conditions are required as there is just one solid domain.

Whereas in the second image, it looks like a hollow sphere inside a hollow cylinder. In that case, you can apply a pressure on the sphere inside and use contact conditions on the sphere outside and cylinder inside.