Least fuel consumed from point A to B with no time constraint

[poiney]

Let's assume we have a generic, standard design ICE car, except that transmission is CVT (continously variable) and can be popped into neutral, since this simplifies the problem.

Assume car travels in a hypothetically-empty freeway at any speed and also assume that driver is optimizing fuel consumption only, not time-to-destination or any other parameter. I think that to optimize fuel, driver should do the following:

Starting from point A, accelerate while keeping engine RPM optimized for fuel engine efficiency to some maximum speed -- not so high as to cause excessive air resistance. At this maximum speed, the car is put into "neutral", essentially coast to zero speed, re-engage transmission and repeat the process until driver gets to point B. Let's assume that engine remains idling during coast phase, not turned off (although I don't think it affects conclusion).

Instinctively one might think it better to leave car at some constant, "optimum" speed (i.e., leave in cruise control) but I assume that coasting is better since it reduces the friction of the engine/transmission during coasting phase.

Is this logic correct? I realize this is overly simplified, but that is my intention. If so, how does driver choose the maximum speed before coasting?

 

[Danger]

Welcome to PF, Poiney.
I think that you're mistaken, but I can't back it up with numbers because I don't know any math. The one thing that you might be overlooking is the inertia that must be overcome with every new (positive) acceleration. A steady speed doesn't encounter that problem.
I'll defer to the real engineers around here, though, since they know how to work it out.

 

[poiney]

Thanks.

I'm viewing this from energy conservation only. At constant speed, since the car still needs fuel to overcome losses, I don't think that the inertia argument applies.

 

[ank_gl]

and do you know that fuel is wasted more in acceleration and declaration because of the richer mixture fed to the cylinders and for maintaining a constant cruise speed, optimum mixture is used?

 

[poiney]

Maybe you are assuming that this car has a regular transmission?

I assumed that engine has a CVT so it remains at optimum RPM and gearing as it accelerates, so to first order, I don't think that we need to worry that the car is not "tuned" at optimum RPM (although maybe you are referring to some other effect).

When car decelerates is just idling since it is popped into neutral.

 

[ank_gl]

CVT does not mean that the engine has to counter constant load, still the load is maximum at the start and decreases with increasing speed.
and the best efficiency is obtained at the cruise speed

 

[poiney]

IIUC, you're simply arguing that in the limit of no frictional losses, that it takes no energy at constant speed, so this gives the best efficiency.

Of course, it is less efficient to accelerate than cruise at constant speed. But it is also more efficient to coast than to cruise at constant speed.

There is still a load at crusing speed to overcome frictional losses, so I don't see anything special about maintaining constant speed.

 

[ank_gl]

. But it is also more efficient to coast than to cruise at constant speed.

and what after coasting, you again need to accelerate, don't you??

There is still a load at crusing speed to overcome frictional losses, so I don't see anything special about maintaining constant speed.

the load when you are accelerating is more