Why Don't Cars Have Long Tails?

The drag force cars create as they move through air is generated largely by a pressure imbalance. Cars are bluff bodies as opposed to streamlined; that is, airflow separates over them and, in modern cars, this separation usually occurs most prominently at the very back and creates a large wake (even fully streamlined shapes like wings have wakes, too, but they tend to be a lot smaller). This large wake results in negative gauge pressure acting on whatever body surface is exposed to it which, since it’s the back of the car, tends to be a large vertical surface (we call this the base)—so the low pressure (base pressure) results in a force acting almost directly backward.

 
One way to reduce the drag of cars is to make this wake area and its resulting pressure drag smaller by tapering the rear part of the car in a long tail; this allows the flow to recover pressure so long as it does not detach prematurely and reduces the vertical surface area exposed to low gauge pressure. Alternatively, other devices can raise this base pressure by extending the rear part of the car, including box cavities, base plates, tapered cavities, or strakes. Redesigning the rear ends of cars and making them longer has a lot of potential for reducing their drag and has been the subject of a lot of research, both in the past and ongoing. So why don’t we see this happening on production cars?

The lowest-drag cars ever offered for sale or lease—such as the VW XL1, the GM EV1 above, or the first-generation Honda Insight—have been hamstrung by the fact that they seat only two people, a consequence of the tapered tail not allowing sufficient interior volume for extra passengers.

Test cars, such as this GM Impact (concept predecessor to the production EV1) land speed record holder, have demonstrated the value of tail extensions. With the extension in place, the "Yellow Ferret" achieved a maximum 183 mph over one mile versus 177 mph without it—about a 6% change in drag. (Image credit: GM).

Styling
 
Probably the number one answer to this is styling. We are conditioned from an early age to have a certain expectation of how cars look (think of how early you started playing with Hot Wheels cars, or Tonka trucks, or the various Barbie cars—something my Sociology 101 professor long ago used as an example of gender norm reinforcement from early age but which also influences our perception from early childhood of what a “car” should look like), and this expectation plays into our buying choices even as adults. These choices are further influenced by the slow evolution of style changes. Cars that are radically designed or stray too far from the norm tend not to stick around for very long.
 
The overwhelming importance of looks and customer expectation is evident in the high priority given to styling by automobile manufacturers. Styling dictates almost everything else at an OEM; if people don’t like the way a car looks, it will negatively impact sales and may impact brand perceptions for years if not decades (the most famous example is probably the Edsel, which was responsible for the demise of its sub-brand, but more recent ones include cars such as the Pontiac Aztek. Scroll to the comments on any blog post about a new model or concept car and you are sure to find people who hate it and are quite vocal about their dislike, especially if the design is out of the ordinary in any way). And the reverse is true as well; journalists have fawned over the 5^th generation Prius’ styling, despite that styling resulting in a car with 10% higher aerodynamic drag than its predecessor among other drawbacks.
 
Drag Reduction Through Detail Optimization
 
Examples of recently updated or refreshed production cars show that reduced drag is still possible through detail optimization without changes to the shape as a whole e.g. Tesla’s Model S, which has a drag coefficient C_D = 0.21 currently, down from 0.24 when the car was introduced in 2012. How much more remains to be gained through detail changes is unknown, but the majority of production cars, if not all of them, could undoubtedly benefit from this approach if low drag is made a higher priority.
 
Radical or large shape changes are largely out of the question when drag can be reduced more easily and without affecting a car’s appearance through detail optimization. Appendages at the back of the car to increase base pressure and reduce drag, whether plates, boards, or body extensions, do not add to usable interior space, lengthen the car with no other benefit than reduced drag, add cost and weight, detract from practical usability (e.g. parking, hatch/trunk access), and are likely to be rejected by consumers for style reasons, many of whom erroneously believe that a bluff tail is as near the lowest-drag shape as to make no difference (the so-called “Kammback” in popular consciousness). Cars with shapes that allow drag coefficients less than 0.20 have so far failed to sell in large numbers when available (e.g. Lucid Air, which falls just under that benchmark at 0.197, or the Mercedes EQS at 0.20) or been deliberately limited in number by their manufacturers (e.g. Volkswagen XL1, General Motors EV1). Further, drag depends not just on drag coefficient but reference area and, as cars have gotten larger with each redesign, drag forces have sometimes increased or remained steady despite lowered drag coefficients.
 
Practicality
 
Styling is evolutionary and largely constrained by tradition (a sedan looks like this, a truck looks like this, etc.) but these traditions derive from practicality and not just vanity. A sedan looks the way it does because of its occupant, storage, and engine space packaging which have evolved over decades, and a practical size requirement which has been “built into” the system in the form of standard lane widths, parking space sizes, and typical garage dimensions. Cars don’t have long tails despite those being better for low drag not just because people would think they look weird but because they are functionally worse for everything a car has to do in the road transportation system as we have built it.

Parallel parking, for example, would be complicated by the presence of a 26" tail. This spot would still have been fine but often on school days I have to squeeze the car into smaller spaces.

I encountered this in my own tail development and my travels last summer, where such a device would have had an undeniable and mostly undesired impact on how I can use my car. Fitting a long tail would have precluded easy access to the hatch area and underfloor storage; it would have negatively impacted rear visibility; it would have made parking in the various cities I stopped in much more difficult (especially Seattle, where street parking is horrendous even at 11:30pm and I was lucky to snag a spot that just barely fit my Prius); it would have added mass to the car that would reduce its benefit somewhat, especially over the multiple mountain passes I went over; and it would only have saved about $40 in fuel costs in a best-case projection over a 5,000 mile road trip.

Mitchell, SD, however, would have been completely doable with a tail—there’s a large parking lot behind its main attraction, the Corn Palace. This is peak Midwest.

Would that be worth it? At this point, probably not; maybe in the future, if gas prices rise precipitously or I buy an EV where range becomes a more pressing concern and every mile counts, it will be. And I’m an outlier; I'm willing to put up with a lot more inconvenience than the average driver in the pursuit of efficiency. Normal drivers want, for the most part, normal cars. I am very much not a normal driver but even I am hesitant about adding a permanent tail to my car for the reasons above.

This is just fine for me; in fact, not only does it have just as much functionality as an external mirror but I think it improves the exterior aesthetic of the car. Most people, however, would not be happy with an internal mirror like this.

The Future
 
Years ago, I corresponded with a well-known and respected aerodynamicist. I was surprised when he wrote that he expected production cars would eventually achieve drag coefficients around 0.20 “and that’s it.” In retrospect, that seems like about the limit of low drag possible through detail changes without radically changing the shape of our cars, and even then it only really applies to long, low sedans; SUVs and trucks remain far higher for the most part (the lowest drag coefficient for an SUV, as far as I know, is the Tesla Model Y at 0.23, and for a truck the Rivian R1T at 0.31—much lower than normal vehicles in these segments. And, of course, their reference areas are quite a bit larger than typical sedans).

 
So, what then? Are we just doomed to a future where cars hit a soft limit of 0.20 and progress no further? Probably, although some concept cars, such as the Ford Probe series of the 1980s (especially the Probe IV and V—two of my favorite cars ever) show that extremely low drag coefficients are at least possible in cars that fit within conventional footprints. Achieving that in production cars will take an incredible amount of work and ingenuity, and will undoubtedly require some influence on consumer preference by stylists and advertisers to be accepted; see how a lot of EVs are already criticized today for their looks and a Jaguar executive recently making the absurd claim that EVs already spend “too much time in the wind tunnel.” Only time will tell if barriers like these can be overcome.

Some low-drag concepts look quite conventional, such as the Mercedes EQXX with its drag coefficient of 0.17. Will we one day have production cars with similar numbers? That remains to be seen. (Image credit: Mercedes).