The variable valve timing can make a big difference on engine efficiency, leading to the unusual "gets better MPG at higher speeds" observation.
Aerodynamic resistance increases with the square of vehicle speed as was previously noted. That combines with tire rolling friction and mechanical friction in the engine/driveline. The mechanical friction is a function of the speed of moving parts so getting the engine RPMs low helps that --> thus modern multi-speed transmissions that have big overdrive ratios so the engine RPMs are really low. That also keeps the transmission input RPMs really low, reducing losses/drag on both the engine (piston rings, crankshaft main & rod bearings, valve train bearings, oil pump losses, other "parasitic" losses from the alternator bearings, water pump & bearings, serpentine belt idler
wheel bearing, etc) and transmission (oil pump, bearings on input end of tranny, etc.) Increasing to 70 MPH should increase those losses (compared to 55MPH) since the tranny is in the high gear for both 55 or 70MPH (so engine RPMs and thus engine drags are still higher at 70MPH) but a few engine tuning tricks probably more than compensate.
Tuning tricks:
Remember, modern engines are tuned for both fuel economy and EMISSIONS. The emissions tuning is what drives engine tuning today more than anything. Almost all modern engines could be re-tuned to get better MPG and still perform as nicely/smoothly as they current do... except they'd make much more emissions. Specifically, the NOx (oxides of nitrogen) emissions. Modern gas engines run at the chemically ideal air to fuel ratio (stochiometric) to minimize emissions. Modern fuel injection systems became the only way car makers could control the air to fuel ratio precisely enough to satisfy emissions laws... that's what killed carburetors.
Remember, the purpose of burning gas is to heat a bunch of air; this heat makes that air expand. In the cylinders, this expansion creates a lot of pressure... that pushes on the pistons to make power. At stochiometric, basically all oxygen is consumed by the burn: there are just enough oxygen molecules for the fuel molecules. If the engine were re-tuned just a little bit, to inject slightly less fuel (leading to a "lean" air:fuel ratio) the burn would actually be hotter - making more pressure. (why? Because gas molecules are "heavy" and absorb a lot of heat; a little excess oxygen makes the burn happen quicker and hotter) So a lean engine would actually make as much as, if not more, horsepower for a given amount of fuel --> better MPG. However, a too-hot burn makes the natural combustion product NO2 molecules separate into NO + O2 molecules. NO is a nasty smog emission. It comes from a hot exhaust... heat energy in the exhaust pipes is what separates a couple NO2 molecules into NO + O2 molecules. The EGR valve (Exhaust Gas Recirculation) on many engines is specifically intended to reduce NO emissions: it routes some exhaust gasses back into the engine to be "burned again." Since this exhaust is already burned (and thus has virtually no oxygen in it) it contributes no oxygen to the burn - it's just extra molecular mass absorbing heat from the fresh air+gas being burned. That lowers the total combustion temperature, and thus the exhaust temperature, to reduce NO emissions. It sounds odd - adding fairly hot exhaust gasses to cool the burn - but that's what happens. By the time the hot EGR gasses actually get to the cylinders, they've cooled down quite a bit... well below the "burn" temperature.
With modern variable valve timing, and the precise control of the fuel quantity via modern fuel injection, many cars have been able to delete the EGR valve. The variable valve timing is used to simulate EGR though at really low RPMs typically to get good NO emissions... thus the valve timing is no longer "optimal" for best MPG.
Engines also have "pumping losses" - i.e. the energy needed to draw air into the engine (past the restrictive throttle plates) and to shove it out the tailpipes. That energy is supplied by the pistons of course - burning air+gas. So anything that reduces pumping losses leads to more efficient operation.
Aftermarket exhaust systems, low restriction intake systems and air filters (e.g. the K&N filters) cost a fair bit of money but do provide some improvement. For the car manufactures, the "bang for the buck" isn't too good - they design "adequately performing exhaust and intakes" that don't cost nearly as much to mass-produce. 80% of the performance for 50% of the cost type of tradeoffs. Anyway, the length and diameter of the intake and exhaust manifolds can help the pumping loss issue: imagine the air molecules drawn in to the cylinder, and then shoved out of the cylinder, for each cycle of the piston. It's basically a "slug" of air moving. Air has mass, so getting it to move means momentum is changing --> ergo it takes energy to move, stop, and move again that slug of air. Now picture the intake manifold (pipe leading from the throttle to the cylinder) That's a slug of air... if the manifold diameter and length are sized right, the air in it moves to the cylinder... when the intake valve closes, the remaining air in the manifold has nowhere to go - but it has momentum. It's going to "pile up" against the valve, and actually bounce back. And bounce again on the other end of the pipe. Now if the intake manifold pipe is sized right, this rebounded air will be bouncing back towards the cylinder & intake valve just as the valve opens - so the piston doesn't have to "suck" as hard to get the air - its own momentum carries it into the cylinders. It's a type of resonance. This resonance happens at some particular engine RPM based on the length & diameter of the intake manifold pipes. The same type of thing occurs in the exhaust: a "slug" of air from one cylinder is shoved down the exhaust header... as that slug passes the "collector" (where the pipes from each cylinder merge into one big exhaust pipe) the momentum of that moving mass of air will induce a vacuum in the other pipes leading to the collector as it passes the collector. If the length of the pipes going to the header is just right, that vacuum will occur just as the next piston's exhaust valve is opening: the vacuum will help draw exhaust out of the cylinder so the piston won't have to work so hard shoving it out. This is what
aftermarket headers do, especially for engines with cheap "log" exhaust manifolds from the factory that don't have individual pipes.
Hyundai may have "tuned" the intake and exhaust systems to be most efficient at the engine RPMs corresponding to 70-75MPH. Combined with the variable valve timing, the engine can be far more efficient at 1700 RPM than it might be at 1400 RPM (or whatever 70MPH and 55MPH match up to). Emissions and "pumping loss" tuning needs are the factors the Tau engine designers had to work with.
mike c.
