[ashtonJ]

Thanks Glenn-It works for contractions up to 50%; however, for more than that, I receive negative element volume error again.
As suggested, I created two short sections shown by red color in the attached image, and used the unspecified motion for them.
I checked the deformation shape and realized that transition between the deformed part and red sections with unspecified motion is still like a step.

Do you know why there is still a step like transition?

[ghorrocks]

Can you post the surface mesh in that region and a cross section through the volume mesh?

[ashtonJ]

Surface Mesh for deformed artery and surface+volume mesh for non-deformed artery are attached. Thanks

[ghorrocks]

Is there an interface at the end of the transition bit? I am trying to understand why it is modelling a step.

[ashtonJ]

I don't understand it, what you mean by interface at the end of the transition bit?? can you please explain what you mean??

[ghorrocks]

Do you have any interfaces (GGI interfaces) in this model?

[ashtonJ]

Yes, there are 4 GGI interfaces as shown in the attached image.

[ghorrocks]

Well, that is the problem then. You are getting different motion on nodes on either side of the interface.

Why do you have interfaces there? I suspect you do not need them. Mesh this body as one continuous mesh. If you need a section to define the mesh motion domain in make sure you have a contiguous mesh over the interface so the nodes are shared and you can use 1:1 interfaces.

The the nodes are shared and the mesh motion will not have steps.

[ashtonJ]

The geometry is made of 3 parts; stationary part, moving part, and the transition part with unspecified motion. As seen in the attached image, CFX will consider the interface between these part as "artery default=wall" if I don't define GGI at these interfaces. Is this what you meant?

[ghorrocks]

You join meshes with interfaces. Interfaces can be GGI or 1:1. All I am saying is that the interfaces should have contiguous nodes so you can use 1:1 interfaces.

[ashtonJ]

I got it, thanks Glenn.

[ashtonJ]

Hi Glenn. As you said, I mesh the body as a one continuous mesh, but when I chose 1:1 mesh connection, an error is given. I also used face sizing on the interfaces to ensure that nodes are shared, but get the same error. Tried different meshing techniques but with no success
Do you know what meshing method needs to be used to have a 1:1 mesh on the interfaces? Thanks

[ghorrocks]

If you are meshing in ANSYS meshing then you need to make this into a multibody part. If you are meshing in ICEM then you just need a single surface between the volume for two different parts.

[ashtonJ]

Dear Glenn, I am using ANSYS Meshing. As shown in the attached image, the geometry consists of five parts. I meshed whole body together, but there was not 1:1 mesh on interaction between parts 2, 3 and 4. I meshes each part separately, but still there was no 1:1 mesh. Can you please let me know what else can be done from meshing point of view to have a 1:1 mesh on interface between parts 2, 3 and 4. Thanks

[ghorrocks]

You need to make it into a multibody part. Then the mesh will be 1:1.

[ashtonJ]

Actually I have no idea how to make it multibody. Can u please explain?

[ghorrocks]

This is covered in the Meshing tutorials. You can download these from the ANSYS Customer web site.

[ashtonJ]

Thanks Glenn.