MoI discussion forum
MoI discussion forum

Full Version: Flipped normals on export on a solid

From: Billabong
3 Aug 2024   [#1]
I have a solid I'm trying to export out and import into Max. You can see in the first image that it is in fact a solid. Just to be safe I did try to select any naked edges on the model and none appear. I have tried exporting out as an obj, fbx, step, and iges. Nothing works. I also tried importing into max as a body object. No luck there either. I also tried in exporting it out in V4 and V5.
In the second image you can see all the parts of the model that are not showing edges are flipped. Yeah I can go in and run through the export hundred times, see which ones are flipped and type "Flip", but that would be a nightmare and I always run the risk of not getting them all. The last two screengrabs show you what it looks like in Max after import. On first import it looks fine, but as soon as you add a material to it, it shows the flipped normals. I'm posting the 3dm as well if anyone wants to take a look.
Thanks for any help
-B









Attachments:
Cockpit_Plipped.3dm

Image Attachments:
Flipped_Normals.JPG  Max_1.JPG  Max_2.JPG  Solid.JPG 


From: PaQ
3 Aug 2024   [#2] In reply to [#1]
Hello, I have no idea why it's a mess like this.

A temp "fix" is to export your model as .igs and reload it.
While the model will not be a solid anymore (they are a bunch of naked edges that I have no clue how to correct), at least you should get a better .FBX export when it comes to flipped surfaces.

An other oddity is that separate and rejoin all the surfaces in MoI also break the solid state (you end up with an joined srf model), but for some reason the "naked" script don't show any open edge :/.
From: Michael Gibson
3 Aug 2024   [#3] In reply to [#1]
Hi Billabong, I think it has something to do with this area here:



If I separate out that piece then it gets a consistent orientation for everything else.

I'm still narrowing it down to try and find what is going wrong in that spot.

- Michael

Image Attachments:
billabong_flipped1.jpg 


From: PaQ
3 Aug 2024   [#4] In reply to [#3]
Hi Michael, how do you manage to isolate the problematic area ? was it just a guess ?
From: Michael Gibson
3 Aug 2024   [#5] In reply to [#4]
Hi PaQ, I tried testing some isolated chunks of the model and saw that they were making consistent normals ok and with repeated testing narrowed it down to that area.

Now I'll see if I can narrow it down further yet, there is probably a damaged trim boundary in there somewhere.

- Michael
From: Michael Gibson
3 Aug 2024   [#6] In reply to [#1]
It appears to be these 2 faces here, if these are removed the orientation is ok:


- Michael

Image Attachments:
billabong_flipped2.jpg 


From: Michael Gibson
3 Aug 2024   [#7] In reply to [#6]
There's an edge in there that's squashed down to to a point.

- Michael
From: Michael Gibson
3 Aug 2024   [#8] In reply to [#1]
Hi Billabong, please try with this attached version. I removed the squashed edges (using untrim) that were on those 2 faces.

- Michael

Attachments:
Cockpit_Plipped2_3dm.zip


From: Billabong
3 Aug 2024   [#9] In reply to [#5]
Hey Michael,

Sorry for the late reply. Thank you so much for fixing this. If you have a min. can you explain in some more detail on how you found the error? Did you just export out sections and bring that into 3D package to see if the problem?

@ PaQ,
Thank you for trying as well. It so much appreciated.

-B
From: Michael Gibson
3 Aug 2024   [#10] In reply to [#9]
Hi Billabong,

re:
> Did you just export out sections and bring that into 3D package to see if the problem?

Yes I broke it up into chunks and noticed that some pieces were exporting unified normals ok. So I took those pieces and gradually made them larger.

- Michael
From: Billabong
3 Aug 2024   [#11] In reply to [#10]
Understood. Thank you again