On the moon it would work, and NASA did that at one point, but on Earth the air drag buggers things up. Specifically the drag/weight ratio. I believe that Mr. DaVinchi had to use two identical boxes with different weights inside each.
Take a car, and take the same car but with four people in it. The drag force is essentially the same on each (small increase in rolling resistance, but let's do this on the highway where air drag dominates). But the second car is heavier, and as we know F=m*A. Same force, different mass -> less decelleration -> longer 'glide'.
But, you'll use more gas getting up to speed so lighter is still more efficient (unless you have to move four people, then carpooling is good)
cfg83 - your point about picking up extra mass at a higher elevation is something I think I have noticed. I go by a junkyard frequently that is at a higher elevation than home by a few thousand feet. Sometimes I will add significant weight to the car at the junkyard ( ) and it *seems* that I get better round-trip economy when I do that. Bear in mind we could be talking about 100lb of metal in the trunk here. No hard data though, but the theory is sound as long as you don't have too many rolling hills to go through on your way to a lower elevation.
I had a 30 pound jug of kitty litter the other day on my downhill trip, so I was 33% there .
Wouldn't it depend on how the hills roll? Once you have perfected the route, I think that *if* you can go downhill without using your breaks (i.e. no brakes or fuel-injector-cutoff-engine-breaking), then the only thing that matters is net elevation drop.
Absolutely. I would think that it depends on the elevation change per hill. Obviously a constant grade down from start to finish would be ideal. For each car there would be some downhill slope where drag forces would cancel the forward forces from the down and you could coast indefinately using no fuel. Adding more weight would make that magic slope a bit flatter for any given car, everything else remaining equal.
On short, gently sloped hills you could probably use your momentum from the last down to get further up the next up. The taller the hills get, the more frictional losses you will have going down and thus get less of the way up the next up before having to feed the engine some gas. At some point the car will reach terminal velocity for the grade you are on and past that the extra down gains you no more speed to get up the next one. It all comes back to the usual equation for good fuel economy - a consistent steady speed will beat a varying speed any day.
Also, real-world considerations enter into this when you reach taller hills and coasting would take you far past what the cop at the bottom of the hill will let you get away with. I can also see problems with annoying other drivers depending on the number of lanes since I drive mostly on two-lane roads.