mesh and mesh control issues

[ajroc]

Hello,

I am having the following problems with meshing an assembly in cosmosworks 2005:
1. When viewing failure diagnostics, while reviewing the part that failed, cosmos will point me to an edge on a different part.

2. After applying a control to a part, and leaving the global mesh settings the same, parts that previously meshed before no longer mesh.

What could be going wrong here?

Thanks in Advance,
Adrian

 

[cbrn]

Hi,
the problems you have are very common in meshing, but a bit difficult to explain.
1- The mesh on a component was successful and the mesher began meshing another part, but encountered a problem at the interface between those two components; consequently, it had to try and restart meshing the first component, which failed for some other reason. You encounter very easily this problem when you have several components that are mutually interconnected. Solution (if you can): uncouple the meshes of the two components (i.e. if you have a contact, set it to "surface" or "free" instead of "coincident node" or "bonded", and use a "connector" to restrain one component to the other)
2- it's almost the same reason as before: by introducing the control, you may make the mesher's life harder where it previously had no problems, undirectly.
As a general rule, when you have several components interconnected, you should avoid leaving the meshes "couple" with each other. In fact, CosmosWorks' mesher has the internal default setting to merge the coincident nodes of two adjacent meshes, whether for example ANSYS behaves the opposite (you have to explicitly merge the coincident nodes if you want them "jointed).

Hope this helps,
Regards

 

[ajroc]

thanks for the info - I'll check out your recommendations.

One quick thing - I didnt think mentioning this specifically was important but when I say that I am pointed to a different part, the part I am speaking of is not connected/mated to the part listed by failure diagnostics.

 

[cbrn]

Hi,
I see your point: it's a bit confusing but, in the case several components are "cross-linked" with one another, the mesh failure can occur in a component "far away" from the one which is currently examined by the diagnostics. I mean, mesh in Component 1 can cause remesh failure in Component 5 even if 1 and 5 are not directly touching, if they are "correlated" by a sequence of one or more other components. However, this can occur only if all the "chain" has "compatible" (or "coupled", call them as you prefer) meshes: if every mesh is independent from the others, then the only possible failure is due to own-component reasons (which doesn't guarantee you not to have any problem at all, of course!).

Regards

 

[ajroc]

Ok that goes along with what I am seeing.

Thanks!