Hi Michael,
I've sent the files to you. I think may be interesting to investigate.
The dimensions of the assy is not so big: less than 400 units.

Notice (in the imported iges) that when you select the long cylinder the 'ghost glued' parts will flash for a moment.
Delete the long cylinder and then undo. Now select again the cyl and the 'ghost glue' become clearly visible.
This is not the only assy with this kind of problems. I will try other experiments and ther report them to forum.
For now I am using the trial version.

Thanks for the support, this forum it is a great 'feature' of MoI!

Gianrico

New tests.






In this image the ring on the back is fully decoupled (in Alibre) from the assy but mantain the union with the other parts.
This happen ONLY if i click on the ring; the other parts selection is correct.









This is the strange thing when i click on cyl to select it. In this case the cyl drives the selection. NOT the other parts.




I deleted the ring, then undo and move: the other parts are magically duplicated!

Gianrico

Hi Gianrico, there are several touching parts all included inside of this one file.

Right now MoI will get confused by this type of structure, it isn't able to understand how to join these pieces together properly.

So for instance in the collar here:



If you slide that over, you can see it has an exactly overlapping part on the other piece:





Above you can see that there are some edges from each part that overlap on top of each other. MoI is getting confused and making attempts to join edges like that, it does not understand how these surfaces are supposed to be separated into individual parts at this point.

It looks like there is some kind of simple grouping information in the IGES file (IGES entity 402). I guess that is supposed to give a kind of unordered listing of surfaces that are supposed to belong to each part? Right now MoI does not look at that entity 402 group information so it doesn't know to only try to join together surfaces within the same group. Probably PolyTrans is looking at that information and that's how it knows not to just try and join everything.

It looks like I may be able to fix this up in the future.

In the meantime, there are a few different choices for how to proceed.

#1 - it is possible to have solids inside of an IGES file - these are called IGES entity 186. If it is possible for you to generate the IGES file using this type of object, then MoI will be able to identify individual parts properly and things should come in joined correctly.

#2 - turn off auto-joining in the moi.ini file, and then select sets of surfaces that you know belong together and join them manually.

#3 - break this into several files exported from the source program, so that each file only has one part inside of it instead of the whole assembly in one file.


Right now you'll have to use one of those methods, please let me know if none of them works for you.

- Michael

Hi Michael,
thank you for your time!
If you read my last post you'll find that the coincident edges is not always the problem.
I can understand the connection between coincident parts but it is not my case. The body of the object it is in three parts but it is comprehensible their union. Other object do no have coincidence thru them!
I'll try to manipulate the IGES file to find some 'trick'.

Gianrico

Hi Gianrico, once the joiner tries to merge edges that connect to more than 2 surfaces, the results can be pretty unpredictable.

There is a process at the end of joining that tries to separate out different volumes and break "non-manifold" pieces back into more normal pieces, but in this case something appears to be going wrong with that final processing.

There seems to be some kind of bug in that final processing in this case, and that is why you end up with that strange disconnected piece.

But it's not really a great case for me to focus on fixing, since the joiner just does not have a chance to process this situation properly anyway, due to the coincident edges. If you use one of the 3 methods above, it should give the joiner a chance to work.

It just doesn't have a chance to work right when you have any coincident edges in there...

Maybe I have not understood your message ... ?

- Michael

Ok Michael, away from me the idea to ask fixing something for me!
Otherwise I must to buy at least ten licenses of MoI! :-D

I will continue to investigate on it (inside my possibilities) just for learn something more.
If some idea comes out I will inform the forum.

Thanks again
Gianrico

I have looked at the different file format but none seem to be for Maya.

these are the programs I use

*Maya 8.5 Personal learning version
*Wings 3D
*Autocad 2008
*MoI.

I use to have MoI Beta but had to update :(

I really do not want to loose my designs as I enjoyed making them.
also I want to make them look cool by texturing them along with doing the plans view of them.

Hi Alphapack - I believe the Maya Personal Learning Edition does not have the full set of file support that the regular Maya version has.

One thing that should work is export from MoI to .3ds format, then use the 3dsin command in AutoCAD.

Does that do the job for you?

Another possibility is to export from MoI to .obj format, using the "Output: Quads & Triangles" option, then use a converter program like 3DWin to convert it into DXF format, which is supported by AutoCAD as well. 3DWin supports a lot of formats so with this method I would think you should be able to find something in common with your other programs.

- Michael

Hi Gianrico,

> Ok Michael, away from me the idea to ask fixing something for me!

Well, the true fix in this case would be for MoI's IGES reader to look at the group elements in the IGES file and only attempt to join surfaces from the same group.

I would like to fix this in the future, I have added it to my list of stuff to work on.


The other part that you are asking about, the disconnected object, is related to difficulties in processing the big pile of surfaces with coincident edges. Even if the area is not immediately in the same spot as the coincident edges, it is caused by side effects from those in other places of the model.


Are you able to use one of the methods above to get this model finished up properly?

Let me know you if you need any help finishing this, I could do one of those steps for you and send you the completed model if you are having difficulty finishing this up, just let me know.

- Michael

Don't worry, Michael, I have no problem with this model.
It is only a test to understand what's happen.
I have other models with the same problem with completely different structure but for now I use a little trick:

1 - I look for the part that 'drive' the group
2 - Cycle thru select until it is the only selection (two or three clicks, strange but real)
3 - Click on 'separate' command'. Now the group is ungrouped but all the parts in the group (except the driver) are duplicate.
4 - Delete the duplicate parts.

Unlimited thanks
Gianrico

Hi Gianrico, I'm glad you have a solution!

I you have other sample files that have this same kind of behavior, can you please e-mail me those if possible? (moi@moi3d.com)

That way I will have more samples to test with when I get a chance to work on improving the IGES import.

- Michael