I stumbled across this which shows the simulated effects of vehicles joining roads, hills, merges etc. Fascinating and provides hours of harmless fun......

It is interesting - it's noticable that simply removing the conditions that caused congestion doesn't always cause the congestion to go away...

It's interesting but not particularly realistic. For example, in the "lane closing" simulation, once you cause a jam the blocked lane hardly moves when we all know that the blocked lane moves faster than the clear lane.

Originally Posted by saga_lout

It's interesting but not particularly realistic. For example, in the "lane closing" simulation, once you cause a jam the blocked lane hardly moves when we all know that the blocked lane moves faster than the clear lane.

Couldn't find it on the web the last time I searched, but there used to a be a site with an article comparing traffic merging at M-way junctions with turbulent flow analysis of liquids where two pipes join.

Suggesting was that you get the initial turbulence as traffic joins, then a bit later a further 'sort out' as other people change to the lane they really want to be in - suggestion was about 1 mile later.