3 ms·
Intuitively there feels like there should be an upper bound on this axiom. For example, imagine a highway 1 mile wide with 50 lanes going each way. That's 220 l
by tschwimmer 7y ago
Intuitively there feels like there should be an upper bound on this axiom. For example, imagine a highway 1 mile wide with 50 lanes going each way. That's 220 lanes in each direction. Assume dense traffic with each car taking up 25 of lengthwise- space. A 1 mile stretch of this megahighway would have ~60,000 cars on it. At this density, the 101 between SF and San Jose would be able to accommodate ~2.8 million cars. A quick search reveals that there are only ~2.5 million cars registered in the counties this highway would run through.
So while I don't doubt that this relationship ('traffic increases to fill the available road space') is true for some conditions, it's not a universal law. The real question is finding out where it starts to break down. Maybe there is some merit to building a super wide highway ;)
- whoopdedo 7y ago> A quick search reveals that there are only ~2.5 million cars registered in the counties this highway would run through. Actual number of cars should be lower after the highway is built because construction would require removing houses to make room for it.