Tuesday, 1 April 2014

Flowing Fluids

THE PENULTIMATE BLOG!

Yeah, I just needed to get that off my chest. I mean, how often do you get to use the word penultimate? So for this penultimate blog I was trying to think of unique and fun topics to discuss. One certain professor recommended Implicit Surfaces and I won't lie, they made very little sense to me. I have yet to develop the necessary skills for working through the usual jargon you find in academic papers. So instead I'm going to talk about another topic: Fluid Dynamics!

So I had looked into Fluid Dynamics before, namely because when you have your prof saying that you shouldn't do something from scratch that's either a warning or a challenge...I chose to see it as a challenge (which I, regrettably, could not meet). So I've only really looked at some of the basics of Fluid Dynamics and as such my description won't exactly be up the level you would expect from something like, say, an academic paper. So to begin with, a simple way of looking at Fluid Dynamics is by looking at it in terms of a grid. Each block on the grid is a particle that is a part of the fluid and contains information such as colour, direction and magnitude. As the simulation runs through its updates, we can use the information in each particle to figure out what each particle should look like (i.e. should the colour become more vivid because more force is accumulating or is the particle untouched and so it should have no colour).

One way of sort of simplifying fluid dynamics is to think about it in terms of a scalar field (Field A) and a vector field (Field B). Field A is used to denote what colour each point in the simulation should be and then we can use bicubic interpolation between each one to figure out the colour on a per-pixel basis during the render process. This will result in a smoother colour blend for things such as smoke. Field B is then used purely for calculations and is ignored during the render step. In order to figure out how each particle changes we need some way of remembering the direction that the particle is travelling in as well as the force behind it in order to calculate the colour. This is important for things such as diminishing colour as the particle 'moves'.

This more or less my understanding of the very basics of fluid dynamics. There is the option to use more equations to create more realistic dynamics such as the navier-stokes equation which takes into account other factors such as pressure. Unfortunately, I will not be going into these at this point in time as I feel that I've still got quite a bit to learn in that area.

So yeah, that's more or less my understanding of Fluid Dynamics summed up in what seems to be one of my shorter blogs. I was kind of hoping that this one would come across as bigger and more interesting but I guess I didn't maintain the momentum from the last one. Oh well, they can't all be winners.

You can expect to see my final blog probably tomorrow/later today or tomorrow-tomorrow (as in April 3) but cheers for now!

No comments:

Post a Comment