PDA

View Full Version : Tornado of cloths


ZBUG
03-30-2010, 07:32 PM
Hi guys,

I have to design a tornado of cloths. The cloths should have a slender behaviour.
I'm trying with some ncloths and maya fields but i'm not achieving the greatest results.
Any ideas?

Aikiman
03-30-2010, 08:06 PM
Probably the easiest way would be to use instancing and a tornado rig. Create a 100 frame (or less) mesh sequence from an animated nCloth mesh. You can get a mel script that will do this for you. I have created a tornado rig which you can get from my site, there are probably better ones out there but its a good start. Once you have it in place then instance your sequence to them offsetting the start number.

My rig allows you to change the shape of the tornado, funnel shape, width, height and thickness. Its not the most user friendly rig but for basic shots is very workable and doesnt need any fields as the particles are controlled with expressions.

HowardM
03-30-2010, 08:43 PM
I think he means a 'whirlwind' of cloth, as in a bunch of slender silky cloth pieces spinning in a tornado like wind... correct?

Aikiman
03-30-2010, 08:56 PM
If your purpose is to actually learn nCloth rather than construct a working, easy to control FX shot then ignore my post.

YourDaftPunk
03-30-2010, 09:03 PM
Are you using a Volume Curve Field with Trap Inside and Around Axis force? I just tried and it's whipping up some nCloth objects nicely.

Aikiman
03-30-2010, 09:41 PM
Are you using a Volume Curve Field with Trap Inside and Around Axis force? I just tried and it's whipping up some nCloth objects nicely.

Damn I never get to play with that field because Im on 2008 at work, can you upload a playblast of your results?

azshall
03-31-2010, 02:42 AM
Damn I never get to play with that field because Im on 2008 at work, can you upload a playblast of your results?

Ugh.. I feel your pain. I too am still on 2008 at work :(

YourDaftPunk
03-31-2010, 05:28 AM
Here's a video and a test of 30 really simple shirt-like objects. The field lets you rotate, pull or push objects from the curve and shape the volume (you can taper the ends, bulge the middle, etc.).

http://vimeo.com/10573168
http://www.shawnlipowski.com/forumFiles/cloth_tornado_05.0072.jpg
scene file (http://www.shawnlipowski.com/forumFiles/cloth_tornado_cgtalk.zip)

The boundary defined by the radius doesn't act like a hard wall, which is nice.
-shawn

Aikiman
03-31-2010, 06:48 AM
Here's a video and a test of 30 really simple shirt-like objects. The field lets you rotate, pull or push objects from the curve and shape the volume (you can taper the ends, bulge the middle, etc.).

http://vimeo.com/10573168
http://www.shawnlipowski.com/forumFiles/cloth_tornado_05.0072.jpg
scene file (http://www.shawnlipowski.com/forumFiles/cloth_tornado_cgtalk.zip)

The boundary defined by the radius doesn't act like a hard wall, which is nice.
-shawn

very nice Shawn, Im gunna take a closer look at that field when I get some time. It must work quite well with fluids too then?!

noticeus
03-31-2010, 08:47 AM
I agree, the sim looks really nice.

I tried volume axis and volume curve fields on fluids with trap inside on but gave some weird results. it seems that the fluid creates velocities outside of the field which result in a strange behavior. so i turned off trap inside and everything was back to normal :)

YourDaftPunk
03-31-2010, 02:51 PM
I tried volume axis and volume curve fields on fluids with trap inside on but gave some weird results. it seems that the fluid creates velocities outside of the field which result in a strange behavior. so i turned off trap inside and everything was back to normal :)

Two things are happening here:

1) A field affects only the particles/polys in the field's volume (with exceptions). No velocity is added to an object if the object is outside the bounding volume of a Volume Axis Field for example. But, in a fluid, all space is connected. If you have a bounded field affecting only the bottom voxels of a fluid, that velocity will still propagate upwards to the rest of the voxels. It's more natural- energy is conserved.

2) Now imagine the strange case of a Volume Curve field- you define a tube like shape where the field has a default effect- velocity around the tube shape in my example. It also has a trap radius with it's own force. That force says that if an object leaves the volume, add a certain amount of force to push it back in. I think that the further an object leaves the volume, the stronger the return force that is applied. In the case of fluids, the voxels outside the volume will still interact with those inside and the voxels outside will experience a huge push in velocity as if they were individual objects violating the boundary. The way the field should work is that trap inside velocity should only be applied based on density that leaves the Volume Curve's boundary. Maybe then the fluid won't explode, but you'll still get a containment effect.


To sum it up:
1) voxels propagate forces (no perfect containment in a field)
2) voxels outside Volume Curve boundary receive velocity even if they don't contain density

HowardM
03-31-2010, 03:09 PM
The boundary defined by the radius doesn't act like a hard wall, which is nice.
-shawn

cool stuff!

I did a test bending the volume field and found that the wall did act like a hard edge...

as pieces fly upwards and smack into the wall they kinda stick...

tried a few things to get rid of it and couldnt figure it out... thought attenuation would help.

will do more tests...

Aikiman
04-01-2010, 02:23 AM
To sum it up:
1) voxels propagate forces (no perfect containment in a field)
2) voxels outside Volume Curve boundary receive velocity even if they don't contain density

To sum it up....bugger.

CGTalk Moderation
04-01-2010, 02:23 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.