That just indicates my failure to communicate my idea :p Maybe tomorrow, if I finish my homework - I'll make a fake figure map to explain better :p

I think the math would show that FE is linear WRT Cd, only at infinitely high speeds.

At normal speeds, rolling resistance is large WRT wind resistance, so Cd improvements get dilluted.

Each car is different, of course, most most cars hit the 50% against aero, 50% against RR, ratio, somewhere between 40 and 60 mph or so... so at highway speeds, one can show that a 50% reduction in Cd would yield a 25% reduction in fuel used.

All said, a 10% reduction in Cd is large.... and 20% reduction in Cd is huge... a 50% reduction is a different car :-)

Yes as a percentage of overall drag on the vehicle .i.e a truck has more losses in its engine and drive train than a car so the aero improvement in a truck would result in less FE improvement than the same aero improvment in a car. In the case of my xB the major drag is the aero because of its size and shape and already small efficient engine and drive train has little further effect since I already reduced it's friction.

RR is, for the most part, constant while in motion, but lets explore the differences with respect to each other.
Here's a measured number that's pretty close to accurate (Here's where I'm getting my numbers from.

RR= CRF*Normal
AeroR=.5*(rho)*v*v*cD*A


RR=.0106*1000(kg)*9.81(m/s^2) = 103.986(kg*m/s^2)
AeroR=.5*1.2(kg/m^3)*v*v*.370*2.3(m^2)

So...
103.986(kg*m/s^2)=.5*1.22(kg/m^3)*v*v*.370*2.3(m^2)
therefore
v= 14.1533 m/s

or 31.66mph

That is, RR dominates until 31 - but that doesn't mean AeroR doesn't play a role until then. 103N = 23lbf

So lets say we're going 40 (perfectly reasonable, normal speed).
That 165.98 N = 37.31 lbf

@55 (gassaver's highway speed :p)
313.81N=70.549lbf

@70 (Highway speed limit for most of everyone else - as if people went that slow :/)
508.33N=114.27lbf

From 31-55... Not even doubling our speed - we more than triple the aeroR. From 31-70 -- A speed increase of ~2.25X, we nearly quintuple our aeroR.

All within normal driving speeds -- really most cars live between 0 and 80. We could say people on here don't drive that fast... But unfortunately, we're not a good sample of the driving populace :/

This is why aero is such a big deal and what allows a bike to go 81mph under human power o.0

Even if you halve a car's rolling resistance and halve its drag, fuel consumption will not go down by half with a gas engine. The reason is that engine efficiency goes down as load is reduced.

If you want to halve the gas used, you must reduce all three, RR, drag and engine size.

wait wait wait engine size is quite an over simplification. if you were to cut the rpms in half via the transmission with the same exact engine then in really simple theory you would halve the fuel consumption. also what if you were to use the same displacement with other engine mods like more aggressive cams or a radically different exhaust system? you could definitely make an engine less efficient if you are going for full throttle power with the same displacement. if that is true can you also engineer mods to work the other way? dont get me wrong though, less reciprocating mass(which can be strongly related to displacement) does matter.

there are a lot of other variables though. as a generalization, i think its safe to say that some solid aerodynamic mods on most cars can net at least a 10% increase in efficiency even if its only reducing the coefficient of drag by 1/3.

i dont mean to call you out or anything, i just think its important that we over simplify things.

Actually, I don't mind being called out.

If you look at engine maps, there is usually a sweet spot where brake specific fuel consumption is lowest. It's usually between 2500 and 3500 rpm in production engines. Above the sweet spot and, in this case, below the sweet spot, efficiency suffers and fuel consumption rises per horsepower.

So, gearing an engine to low rpms, while it does help mpg, does not help as much as would a smaller engine.

Interestingly, a similar effect happens with displacement. Reducing displacement by using smaller cylinders doesn't help as much as reducing the number of cylinders.

If automakers want to get serious about fuel economy in an affordable (i.e. non-hybrid) car, they should be looking at single cylinder or twin cylinder (for better smoothness and balance) engines.

I should add that even though a larger engine at high load will be less efficient than a small engine at high load, the difference isn't nearly as much as each of those engines at low load, at around 10-30%. While differences in load at a given speed can drop efficiency by 100-200%. Offset cranks and other friction reducers also help reduce the drop in efficiency compared to engine speed, as is seen in a few newer engines. This could allow for a relatively small drop in efficiency compared to speed and reduce the penalty, even though it's fairly small already, of having a large displacement engine run slower at the same load.