The primary reason for a velocity stack on a two cycle is to retain the fuel from carburetor stand-off. If you examine the carb in operation, you can see a mist of fuel outside the carb entrance. A stack captures this fuel. This can only be seen when there is dead air around the carb. One of the most obvious cases was on the McCulloch racing carts. They would wet a hand placed several inches from the carb.
DBryan,
That is not completely correct. Retaining fuel is a benefit from velocity stacks, but it is not the primary reason. Without going into a bunch of physics and junk, the purpose of a velocity stack is for two things, primarily to keep the column of air moving after the induction cycle stops, so that the next time it begins the air hasn't lost it's pressure. The second (and the reason stacks aree tulip shaped) is to funnel air in from a broader area without disruption of airflow (eddy currents, etc.) To expand on my first point: Air is actually a fluid medium, kinda similar to water (except it is compressable to a point). If you can imagine a bottle half full of water, tip it one way slightly, then tip it the other way slightly, what you will see is the water rushing from one end to the other and building up at the opposite end until you tilt the bottle the other way, and vise versa.
The comulmn of air kinda works this way in the intake tract. When the intake begins it moves a mass of air into the engine thru the intake manifold, carb, and port. Because this air has mass it also has velocity. When the intake valve closes (reed petal, piston port or rotary valve) it closes very quickly, and the air column that was flowing inward has nowhere to go so it bunches up and builds pressure right at the inlet point of the engine where the valve closed (like in the example of the water bottle). But just as in the case of the water bottle the pressure will quickly reverse direction and go the other way to try and even itself out to atmosphere. The idea behind tuned intake velocity stacks is to time this action so when the next intake event begins the air pressure behind the intake valve is at it's peak. This timing can be controlled by the length of the stack or port. The angle or shape of the intake velocity stack also has a play in this pressure by allowing a larger moment of air to effect the pressure build at the intake valve.
There is also a sound wave harmonic that is similar to a tuned exhaust at play here that also requires correct timing. You can begin to se how this can be quite difficult to calculate.
The reason the carb spitback is reduced with a velocity stack is because the incomming air charge is meant to be constantly under pressure (If the design calculated correctly) And directed only inward. When the intake is out of time, or the intake dwell is to large, or the engine is not running at the rpm that the intake was tuned for then the air column can crash against the closed intake valve and find its way backward thru the carb causing spitback. Usually intakes are tuned for peak TQ rpm, so any thing under that rpm the intake tuned length will be to short to be effective, alowing spitback to occur, and Vise versa, if the port is to long, to try and eliminate spitback then the intake length will be much to long to be effective for power production.
If your still with me congratulations! I tried to keep it simple!
BK