PDA

View Full Version : particles to simulate snow drifts: where to start?


Breravin
11-17-2009, 06:11 PM
Hey Guys,

I've spend a couple days trying to track down some info on this and have come up short.

How would I go about using particles to simulate drifting, windswept snow?

I already have a basic scene blocked out that could act as the foundation for the simulation. I'd then want to simulate particles blowing and piling up. From that simulation, I'd like to generate a (not too dense) mesh.This simulation would not be for an animation, but instead to get the end result of the piled snow.

Sadly, I have only very basic dynamics experience (really, it could be considered "zero"). I'd appreciate if anyone could point me in the right direction.

Working in either Maya 9 or 10.

ps. I've tried nCloth with decent results, but it doesn't feel quite natural enough, and doesn't have that windswept, directional feeling.

YourDaftPunk
11-17-2009, 06:43 PM
Self-colliding nParticles and stickiness- you can have different stickiness settings on your geo and nParticles.

Breravin
11-17-2009, 06:51 PM
Self-colliding nParticles and stickiness- you can have different stickiness settings on your geo and nParticles.

Cool, but I'm going to need a little more info than that. What would my setup look like? Am I using an emitter? How am I affecting things like wind? Where is stickiness even set? What do I do to generate a mesh from that?

I've never set a scene like this up, so as my post stated, I don't even know where to start.

KidderD
11-17-2009, 06:59 PM
This might help a little. How much of this is simulation, just the wisps on the crests?

http://www.the-area.com/tutorials/making_snow_in_maya_2010

Breravin
11-17-2009, 07:10 PM
This might help a little. How much of this is simulation, just the wisps on the crests?

http://www.the-area.com/tutorials/making_snow_in_maya_2010

Hey, that's pretty interesting, but the result is kind of lumpy. "Half pipes" and "drifting" are critical for my area.

So this kind of stuff
http://media.photobucket.com/image/city%20snow%20drifts/kjev71/2005winter.jpg

http://www.southlakes.addr.com/userreq/SnowDriftsatBootleStation1900.jpg

http://www.islandnet.com/~see/weather/graphics/photos0708/blizzard_1888h.jpg

Ideally, the entire 'snow cover' is created via the simulation. Or at least the basic shape of it (i.e. if I could get the particles in the general shape and placement, I'd like to generate a mesh from that)

YourDaftPunk
11-17-2009, 07:21 PM
The nucleus node has its own built in wind/turbulence (and ground plane!) and you can use that in combination with the wind shadowing attributes on the nRigid node of your collision geometry.

For example- if the nucleus wind is blowing the particles against a flat wall, the particles will stick to the wall and not fall as a result of gravity. They will form a sheet of snow rather than a drift. The combination of wind/stickiness will hold them there unrealistically.

By turning on wind shadowing on your collision object, when the particles reach that wall they will no longer feel the force of the wind and begin to respond to gravity. This means that the particles will tend to pile near the base of the wall like a real drift.

-shawn

Breravin
11-17-2009, 09:02 PM
The nucleus node has its own built in wind/turbulence (and ground plane!) and you can use that in combination with the wind shadowing attributes on the nRigid node of your collision geometry.

For example- if the nucleus wind is blowing the particles against a flat wall, the particles will stick to the wall and not fall as a result of gravity. They will form a sheet of snow rather than a drift. The combination of wind/stickiness will hold them there unrealistically.

By turning on wind shadowing on your collision object, when the particles reach that wall they will no longer feel the force of the wind and begin to respond to gravity. This means that the particles will tend to pile near the base of the wall like a real drift.

-shawn

I will give it a try. If anyone has the time, I'd be extremely grateful for a test/demo scene with a basic setup. Someone with a little experience might be able to set it up in a few minutes, while I'm still fumbling.

Thanks for the responses!

edit: Actually I'm not seeing wind, turbulence, or a ground plane in the emitter nucleus.
edit2: Found it.

Breravin
11-17-2009, 11:16 PM
Alright, starting to get desirable behavior.

Now, what can I do to make the particles behave more like snow? With peaks, wisps, ridges, etc. And what can I do to make the generated mesh look more like smooth snow as opposed to lumps of spheres.

The hurdles I currently have are:




Making the snow less lumpy (more particles? does a particles exist that ‘flows’ into the next particle?)




Varying up the wind in a way that creates cool snow patterns (I could ‘trial and error’ this, but I’d rather have some sort of algorithm drive wind speed and noise randomness)




Having the snow stick to the “tops” of objects more (I was thinking I might just end up making a separate collision mesh for the ‘tops’ with a higher stickiness)

neurobasics
11-24-2009, 12:55 AM
drive them with fluids.

CGTalk Moderation
11-24-2009, 12:55 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.