2014-03-22 20:47 GMT-03:00 Sebastien Sterling <sebastien.sterl...@gmail.com> :
> Is this "volume grid context" something that ICE can't deal with. or is it > just a matter of their not being a specific solver written to demonstrate > this behaviour, like Raff is saying for Lagoa. I'm not saying you cant do ice with ICE (:p). It is just that in the specific video you pointed to the guy specifically explains that they use particles to define the mass, but velocity and collision calculation happens on grids. ICE is great at dealing with particles, you can even build a grid with particles, but you dont have many tools for dealing with grids (which are often used in smoke simulatores) specifically, nor a native grid context (ie: no self.VolumePosition or GridPosition like we have PointPosition, VertexPosition, PolyPosition and so on...). I have no experience in trying to recreate such a thing in ICE, but I assume it is not easy to implement the nicest papers out there which describe dynamic and even adaptive ways to do this... emFluid5, for example, is a not only a nice fluid solver but also a tool for creating and messing with such grids. but vanilla ICE does not have that.
