Eric, simulation of this type is not very practical to do in Ae.
After Effects expressions have no memory -- they are evaluated anew every time a frame is rendered, and they have no way of knowing what has happened on previous frames (nor what will happen on subsequent frames). To have knowledge of a past event, the expression must calculate it from scratch.
To illustrate: if calculating the value for a layer's property for the second frame of a simulation depends on a value from that property on the first frame, it must calculate both the first and second frame. Then on the third frame, it would have to calculate the first, second, and third frame, and so on. The problem compounds when you add multiple layers, as in a traffic jam simulation, where each layer depends on the previous layers.
The entire simulation-to-date must be calculated on every frame for each additional layer. The longer the simulation or the more actors in it, the greater the likelihood of your expression timing out and your render failing.
(This is probably why Newton calculates its simulations outside of After Effects).
If you can think of way to express the position of a car in the traffic jam based on its time and index in the jam only, then you can write a good expression to help animate the jam. Otherwise, you're better off animating by hand or writing a simulation elsewhere and outputting Ae keyframe data.
As for Particle Playground, that ephemeral x-speed property is not what you want it to be -- it's for defining the instantaneous speed of the particle at the position already defined by the simulation; it does not modify the simulation itself to reposition the particle. If you turn motion blur on, you'll see how the particle's speed, but not its position, can be affected by this property.
Walter Soyka
Principal & Designer at
Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog - What I'm thinking when my workstation's thinking
Creative Cow Forum Host: Live & Stage Events
