Traffic Modeling via compressible, supersonic CFD-- has it been done?

[arkie87]

I was thinking the other day how traffic could be modeled using compressible, supersonic CFD.

In subsonic flow, flow accelerates through a nozzle, whereas in supersonic flow, flow decelerates. In traffic, when the number of lanes reduce, traffic decelerates, so it appears that traffic flow is "supersonic". However, what does that mean exactly? Supersonic flow means the flow is moving faster than the pressure wave can communicate boundary conditions/information, which causes molecules to pile up and shockwaves to form. In traffic, it therefore means that the cars are moving faster than the drivers can react to what is happening upstream. Since most of the time, cars can only see whats happening a few cars ahead of them, this analogy makes sense (and even implies ways to improve traffic flow by allowing the information to travel faster than the speed of the cars--incidentally, computer driven cars should be able to do this).

In addition, "speed limits" could be simulated using a "moving wall" boundary condition with V=speed_limit, and the left most lane would be zero shear boundary condition. Traffic can be spawned via volumetric sources or inlet/outlet boundary conditions etc...

Thus, has any work been done to simulate traffic using compressible, supersonic CFD? Other thoughts?

[Armchair Rocket Scientist]

Check out this paper: the abstract doesn't say whether it uses supersonic flow equations, but it is using CFD to model traffic.

http://link.springer.com/article/10.1007%2FBF03325741

[arkie87]

Check out this paper: the abstract doesn't say whether it uses supersonic flow equations, but it is using CFD to model traffic.

http://link.springer.com/article/10.1007%2FBF03325741

"Shockwave propagation". Sounds like it's supersonic.

Thanks for the paper!