cluster deformer problem (normal flipped)

# 1 28-10-2014 , 03:33 PM

Registered User
Join Date: Oct 2014

Posts: 18

I am trying to simulate paper folding using cluster deformer. It is a simple setup.
At first I applied a diagonal fold to my mesh with less than 180 degree to avoid overlapping and then delete history. From this setup, I applied a cluster deformer to the single top vertex (Fig. A) and then rotated it along a locator point (Fig. B). I use double-sided shader by utilizing the normals (Fig. C).

Fig. D shows the initial state before applying a mountain-fold to it. Then I applied a rotation and it looks fine up to 89 degrees (Fig E).

Attached Thumbnails

Click image for larger version

Name: (C)-hypershade.jpg
Views: 370
Size: 151.9 KB
ID: 56253

Attached Images

Quote

# 2 28-10-2014 , 03:40 PM

Registered User
Join Date: Oct 2014

Posts: 18

The problem takes place when the rotation is exactly at 90 degrees. Somehow the upper part of the face normal flipped to show the backside shader (Fig. F). The face normal of the other side was also flipped (Fig. G).

Anybody know what's causing the issue? If yes, then I would like to know how to avoid this problem.

Thanks.

Attached Images

Quote

# 3 29-10-2014 , 06:10 AM

Registered User
Join Date: Oct 2014

Posts: 18

I just found out that the normals are o.k. The problem is starting at 90 degrees, the left layer sticks out of

the right layer. This is clearly shown on attached figs when I increased the angle of diagonal fold. This seems

an intrinsic problem of simulating paper folding. Only way to avoid this problem is to give a constraint of not

allowing the left layer passing through the right layer. Does Maya has this kind

of feature?

Attached Thumbnails

Click image for larger version

Name: (H)-wider gap with less than 90 deg.jpg
Views: 353
Size: 60.9 KB
ID: 56263

Click image for larger version

Name: (I)-wider gap with 90 deg.jpg
Views: 364
Size: 63.5 KB
ID: 56264

Click image for larger version

Name: (J)-wider gap with larger than 90 deg.jpg
Views: 338
Size: 65.3 KB
ID: 56265

Last edited by muonium; 29-10-2014 at 06:20 AM.

Quote

# 4 30-10-2014 , 04:25 AM

Super Moderator
Join Date: Dec 2006

Location: South FL

Posts: 3,522

So the color is flipping without the geometry interpenetrating?

- Genny
__________________
::|| My CG Blog ||::
::|| My Maya FAQ ||::

Quote

# 5 03-11-2014 , 07:47 AM

Registered User
Join Date: Oct 2014

Posts: 18

There is no flipping of color. As you said it, it is the intrinsic problem of geometry interpenetration when you apply multi-folding of paper.

Nobody has a solution for this problem?

Quote

# 6 03-11-2014 , 10:25 AM

The thin red line
Join Date: Aug 2009

Location: England

Posts: 4,472

What is the connection from the sample info node to the condition node..............dave

Avatar Challenge Winner 2010

Quote

# 7 03-11-2014 , 05:29 PM

Super Moderator
Join Date: Dec 2006

Location: South FL

Posts: 3,522

Could try giving joints a shot.

- Genny
__________________
::|| My CG Blog ||::
::|| My Maya FAQ ||::

Quote

# 8 06-11-2014 , 04:46 PM

Registered User
Join Date: Oct 2014

Posts: 18

The answer is: samplerInfo1.flippedNormal->condition1.firstTerm.

It just check if the paper is flipped or not and depending on this state, a shader is chosen.

This works flawlessly.

Quote