too vague... the speed limit is only one limiting factor, you need the number of vehicles passing the entry of the sector at some time, their spacing, number of lanes etc. The minimum is definitely zero, the theoretical maximum would involve calculating the number from the average vehicle length, with no spacing, times number of lanes moving at your desired speed through a unit traverse length. Of course, this maximum is not possible, although the bumper-to-bumper scenario is seen in many municipalities on a daily basis... rush hour... The traffic volume moving through a traverse length, in that scenario, is minimal since the speed limit can't possibly be reached. I think you might want to qualify your conditions or, use data from other sources.
Velocity and speed are two different metrics. Speed is a component of velocity, the other component being direction.
For exampke Km/hr is the magnitude or measurement of speed: 60 km/hr. However 60 km/hr northbound is known as velicity, a vector quantity, as both speed and direction are provided.
I agree with Dan in that your question is a bit too vague and suggest you reconsider exactly what it is you are trying to model, and what additional data sources you may need to achieve your goal.
Consider this: My street is posted as 25mph, but it is not a through street, that is to say, if you don't live there, you probably won't drive there. It ties into another street that is posted as 25mph. However, that street feeds the entire neighborhood of multiple streets so traffic volume is considerably higher.
See my point? It's a stretch at best to correlate speed and volume.
I just googled "determining traffic volume" . There are tons of things to look at....