I have some difficulties correcting the non-manifold curves error I'm getting in prosurf/proamm. The geometry consisit of a pipe attached to one of the walls of a lab(which is the flow domain).

On importing the geometry created in IDEAS in igs format to prosurf, after running the automatic repairs, I still get this error of 3 non-manifold cruves which are all located on the pipe (the circular part attached to the wall and the straight part connecting the two ends).

I dont know how to solve this error and it's seriously retarding my progress. Could some one pls help!!! Thanks for your response in advance. Shariff

for a curve/surface to be manifold it should have 2 surfaces attached to it, more than this and it will be flagged as non-manifold. In some situations you may want a non-manifold curve where you have a separate region or an internal feature (baffle etc). If this is the case then you shouldnt have a problem providing you specify the internal feature/baffle type during meshing but if it is not the case then you need to analyse the curve further and work out why you have additional surfaces.

Thanks Ben, yes the internal feature (pipe in this case) is representing a baffle and the non-manifold description is desirable (base on what I read in the proamm user guide). However, I'm disturbed because it appears as an error.

Could you please be more specific as to what stage in the volume meshing in proamm I need to define the pipe as a baffle (or can I do this in prosurf)? Is it at the surface preparation stage? I get this error before the surface triangulation in prosurf and is carried on to the surface preparation stage in proamm. Thanks once again for your invaluable response. Shariff

In addition to the non-manifold description, proamm also descibed the same attached pipe(geometry with thickness)as multiple edges, duplicate cells and self-intersecting. Perhaps this additional info could be relevant in guiding solve this problem.

If I include the pipe of a given thickness hanging freely in space within the geometry, will it reduce the chances of getting such error description?

Will really appreciate your informed guide and suggestion.

It comes up as an error as most of the time you want to avoid non manifold surfaces and it can point to problems in your CAD. I am sure you would get people complaining if it wasnt flagged as an error and the mesher failed later down the line.

All you need to do from a volume meshing point of view is point out which cell type is the baffle type at subsurface generation (this will grow the sub-s on both sides of the shell not just one), you just need to make sure that your baffles are a different type (you can do this via surface prep in pro-amm)

Multiple edges should be the same cells as the non-manifold, check the cells that it flags to make sure they are in the region where the pipe is attached to the wall.

The self intersecting and duplicate cells probably arent anything to do with the non-manifold reagion and are a serious problem, you need to find out where they are and rectify accordingly.

Thanks Ben for the tips on non-manifold error which I solved by identifying the pipe as baffle, increasing its diameter and detaching the pipe from the wall (I still get this error plus self-intersection as long as the pipe is attached to the wall). I also get the self-intersection error at the point where the raised platform of engine test bed (solid block) is attached to the floor.

However, I believe there should be a procedure for avoiding the self-intersection error in proamm where one component of the geometry is attached to another, but don't know how (still a novice determined to learn star-CD). More so, I don't get this self-intersection error in IDEA where I created the geometry.

Every one out there, I need some valuable hints. Thanks

if you work with Ideas: In Ideas you can do an "interoperate", to get the part in parasolid (.x_t), this format is importable in star-design! In your problem this is not very useful...