meshing together seperate parts with different material types

[brianpaul]

Alright, I'll admit I don't know how to do something that should be very simple and basic for everyone else who knows how to use I-deas. I'm a newbie and I can't find this in the tutorials or help files.

I have two parts. A small piece of steel glued inside a bulky block of epoxy.

I want to FEM them together, but only seem to be able to mesh them seperately. And I trust that when I try to apply the load to the steel, it won't transfer to the epoxy, if the errors don't get me first.

And I tried meshing the entire assembly, but from there I don't know how to seperate out the different material types.

If this was in ANSYS, I would just use the GLUE command, and mesh seperately with different MATTYPE definitions, knowing the nodes would be coupled and overlapped when fremeshing.

How do you do this in I-Deas?

 

[umit]

I would first sperate the volumes or surface ( depending on if it is a 3D or 2D mesh), then I will mesh it seperately with different material properties.

I hope his helps.

umit

 

[Liro]

You could try to merge nodes at the interface. This works fine if the meshes are similar. Just enter a global search distance, 0.1 mm or something.

Another way would be joining the parts via partition. This produces two volumes sharing surfaces (I think umit suggested similar procedure). Volumes can be then meshed separately with different material props.

Liro

 

[joyem]

Yes, as Liro said, you could use "Partition" to separate a single volume on conditon that the two parts are joined together. That could guarantee the continuity of meshes on the interface of two parts which could also be meshed with two different materials. Or if you cannot join the two parts together into a single volume, or just use assembly to join them(meshes are not continuous on the interface), you could try to connect the two parts with "Join Dissimilar Mesh" with RBE2 or RBE3 elements under "Define Connector Elements" menu in Meshing task.

 

[brianpaul]

joyem and all,

Thanks for the reply. I'll try your and Liro's method when I get a chance. Earlier I tried picking the individual nodes and coupling their degree of freedom - which I think I did correctly. Then after applying the load, followed by what I thought were the b/c constraints, I expected to be able to run the solution. Instead I got an error message saying no boundary conditions (b/c) were applied.

I'll figure it out. Thanks for the advice.