Light Vehicle Efficiency Over Time

In the first chapter of Aerodynamics of Road Vehicles you will find, regardless of edition, an excellent overview of the history of passenger vehicle aerodynamics written by Wolf-Heinrich Hucho. In it is this chart, showing the general trend of drag coefficients over time:

This is from the most recent edition, published 2016.

This gives us an idea of progress in reducing drag coefficients (at least in European cars, but I imagine a chart of American or Asian cars would show the same), which you can see sometimes moves in fits and starts before plateauing for a while. Generally, the trend over time is that drag coefficients get smaller, and it is assumed that cars become more efficient as a result. But is this really what’s happened? Sure, cars are more efficient now—but how much more efficient? And is all of that attributable to a decrease in drag coefficients? I found myself wondering this over the last several weeks as I’ve played around with various models simulating road load and efficiency.
 
Data
 
Unfortunately, unlike OEMs, I don’t have access to independently measured drag data for a bunch of cars. Instead, I have to rely on what manufacturers publish, usually buried in press releases for new model launches. Because of this, I’ve limited this analysis to just four models out of necessity: the BMW X3, Toyota Prius and Corolla, and Mercedes-Benz S-Class.
 
Dimensions, curb weights, drag coefficients, and EPA fuel economy ratings are available for each of these models going back multiple generations. Plotting drag coefficients shows the same trend as Hucho’s chart indicates:

Aside from a small increase in the current Corolla and somewhat larger increase in the most recent Prius, drag coefficients for these four models have fallen over time—quite rapidly in the case of the X3, which is a relative newcomer to the market. But this doesn’t give the whole story; as I’ve written before, drag coefficients aren’t very meaningful without knowing reference areas. Multiplying these drag coefficients by estimated reference areas based on the known dimensions of each model and generation changes the picture somewhat:

Now we can see that a model like the Corolla actually has a larger drag area (and thus more aerodynamic drag at any speed) than it did 30 years ago!
 
Then, remember that power requirements and efficiency depend not just on aerodynamic drag but also weight. A chart of weight over time for these same models shows an unfortunate but expected trend:

Up, up and away. These weight increases negatively impact efficiency by increasing the force required to accelerate the vehicle and the power required at any speed proportionally.
 
Efficiency
 
Despite the above, light vehicles have gotten more efficient over time—in large part because they have been required to by government regulation (CAFE in the US, (EU) 2019/631 in Europe). Combining aerodynamic and rolling load into one metric—power required at cruise—and plotting it along with EPA highway fuel economy ratings makes apparent a curiosity. While some cars, such as the X3 and S-Class, have gotten more efficient roughly in step with reductions in aerodynamic drag (which, in these models, more than offsets their respective weight increases):

…other cars, such as the Corolla and Prius, have gotten more efficient in spite of flat or increased aerodynamic drag:

How can this be?
 
Thermodynamics
 
It all comes down to drivetrain efficiency: if the applied load has stayed the same or grown over time, any increase in efficiency must be attributable to improvements in how the engine extracts energy from its fuel source and how much or little of that energy is lost in the drivetrain before it can be applied to the road at the tires as a traction force.
 
This is best illustrated in the comparison of the absolute values of fuel economy of the Corolla and Prius (or, if you want, the conventional Corolla on these plots versus its hybrid variant, which has fuel economy ratings similar to the Prius). Although they are of similar size and similar required power, the Prius is rated 10-15 mpg higher than the Corolla in every generation. This is probably because its hybrid CVT (which is completely unlike a conventional transmission) allows the engine to run closer to peak thermal efficiency in more conditions and the engine has been tuned for higher thermal efficiency overall since the electric motors can add torque in conditions where such tuning would otherwise result in less-than-acceptable performance (the Miller cycle engine Toyota uses in its hybrids has different valve timing from its other engines, a feature that alters the effective pressure ratio over different parts of the stroke). Other small changes may contribute as well: the Prius routes engine coolant through the exhaust system to warm up the engine faster and has, since 2010, incorporated a supplemental resistance heater element in the cabin HVAC system; recommended oil viscosity has been lowered in each generation of Prius, from 5W-30 in 2004 to 0W-20 in 2010 to 0W-16 today; the best brake-specific fuel consumption (BSFC) "island" has been expanded so that the engine operates at lowest fuel consumption over a wider range of RPM and torque output; the mechanically-driven water pump was replaced starting in 2010 with an electric pump, etc. All these have added up to significant increases in efficiency from one generation to the next.
 
What a Drag
 
So, while it may be that drag coefficients are trending downward, it might not necessarily follow that drag forces have fallen similarly. At least some of the increase in EPA fuel economy ratings is driven not by reduction of aerodynamic drag and certainly not weight (which continues to go up for each new generation of nearly every model, with a few notable exceptions such as the 2014 to 2015 Ford F-150) but by improvements in engine and drivetrain efficiency. This suggests that there is room for improvement still if we can focus vehicle development on meaningfully reducing both weight and aerodynamic drag (the easiest way to do this being, by far, to simply make cars smaller). Of course, this is something you can experiment with on your own car; I’ve taken around 75 lb out of my Prius so far and reduced aerodynamic drag on my truck by more than 10%. At the same time, one person modifying a couple of vehicles doesn’t have much bearing on the American car market as a whole; OEMs have to step up too, and buyers—who largely drive the product mix and future development objectives of OEMs—must prioritize these attributes in their purchasing decisions.

Given how many Americans are freaking out right now over “drone” sightings that are actually airplanes or stars, I’m not hopeful we can make rational decisions on the whole.