Send me a note:
darin AT metrompg D-O-T com,
Kardboard Kammback: testing a partial boat tail prototype
Posted Saturday, September 9/06 in Mods & Tests
One of the biggest aerodynamic problems with the majority of hatchback-shaped vehicles (like my Blackfly) is the size of the trailing wake. The large reduced pressure zone acting on the rear end can account for up to 1/3 of total drag.
That said, a hatchback configuration isn't automatically an aerodynamic death sentence. On the contrary, three of the most slippery production cars available in recent years were hatchbacks: the 2nd generation 2004- Toyota Prius (Cd .26), 1999-06 Honda Insight (Cd .25), and 1999-05 Audi A2 (Cd .25).
But these cars are exceptions to the rule. So what do you do if you're stuck with a conventional more or less van-shaped hatch? Just grin and bear the aerodynamic shame of it? No way! Get out the cardboard and duct tape, and go to work fixing what the stylists messed up!
Boat tail prototype overview ...
Hatch wake: size matters ...
I've only ever found one reference to the drag coefficient (Cd) of the 1995 and up Geo Metro, quoted at .34. That's not good, but it's believable, given the car's profile. It's kind of got a van-shaped rear end, where the air flow separates near the maximum roof height, causing the size of the trailing wake to be practically the same size as the vehicle's projected area.
Modern sedans - "3-box designs" - with trunks of sufficient height and length, normally have less drag than their "2-box" hatchback counterparts. This is easy to understand when the relative size of their wakes is revealed by a smoke test in the wind tunnel:
Teardrop shape reduces wake...
Aerodynamicists have known for a long time that minimizing the size of the wake decreases Cd and increases fuel economy. The shape that best achieves this is one which tapers at the rear in the approximation of a teardrop, both in profile (roof tapered down and floor tapered up toward the rear ), and in plan view (as seen from above, sides tapered back toward the centerline).
By keeping flow attached as long as possible, adverse pressure gradients are avoided and streamlines are shepherded gradually together behind the vehicle resulting in minimum pressure drag.
Chopped teardrop: the Kammback
Obviously, it's not practical to use a fully tapered teardrop shape in an every day vehicle - the back end would be too long, and would have limited usefulness for carrying people and cargo (one of the advantages of the van-shaped hatchback).
Fortunately, in the 1930's a German aerodynamicist named Wunibald Kamm discovered that when the ideal teardrop taper was applied to the shape of a car body, it could be cut off well before the tip with little change in drag - the air flowing off the rear still behaves as if the full teardrop shape is there. The Kammback was created.
Not surprisingly, all three of the low-drag hatchbacks shown above employ the Kammback design.
Boat tail retrofits: a known cure ...
"Fixing" a poor rear end design by retrofitting aerodynamic add-ons to minimize wake isn't a new idea.
Aerodynamic drag can be significantly reduced with trailer add-ons that reduce the wake and increase the base pressure. - Source: SAE 2000-01-2209
Kardboard Kammback prototype ...
While the seed of inspiration for trying something similar on the Blackfly was planted originally by the EVWorld article about Phil Knox' aero-modded Toyota pick-up, the final push to get out the cardboard, box cutter and duct tape to construct a drag-reducing shape came from a drawing made by Mark and posted on his web site (now defunct).
(I know, I know. Technically it's not a Kammback, since it doesn't quite reduce the rear area of the car down to 50% of the total projected area... I just like typing 'Kardboard Kammback'.)
To make a long story short, I adapted Mark's basic design in about 45 minutes, with the goal of simply extending the existing roof & side taper already designed into the car.
And despite the impression that the boat tail must obstruct vision out the rear window, the horizon was actually still visible in the rear view mirror, just below the end of the cardboard.
It sure looked interesting. But did it work? Quick! To the wind tunnel!
Testing the Kardboard Kammback ...
Much to my neighbours' entertainment, I left the cardboard creation on the car long enough to wait for a calm day to visit my usual test route for some before & after runs. It's always a risk making car parts out of cardboard: had it rained, the boat tail would have turned to mush and fallen off. Fortunately, I beat the weather!
I wouldn't call this a rigorous test: Cardboard/duct tape prototypes don't lend themselves to multiple removal & re-installations, so the preferred A-B-A method was not used.
That said, conditions were great:
Observed at: Grenadier Island 18 July 2006 3:00 PM EDT
- The experiment
This was a simple A-B comparison, where A = with cardboard boat tail prototype, and B = cardboard removed.
Speed was 55 mph (88.5 km/h) using cruise control, and it was hot, so the fan was on 3/4.
@ 88.5 km/h / 55 mph - in MPG (US)
A = boat tail on
A/B ... West ... East ... Average MPG (US) / L/100 km / MPG (imperial)
A ..... 59.8 ... 61.6 ... 60.70 / 3.88 / 72.9
B ..... 57.2 ... 61.7 ... 59.45 / 3.96 / 71.4
61.23 mpg (US) - average of A runs
1.38 mpg (US) - difference of between A avg. and B avg. runs
Interesting! On the one hand, I was pleasantly surprised to see a measurable result from a slapped-together prototype. But at the same time, I was also slightly disappointed that it wasn't a larger effect.
The difference between this mod and the mirror removal, grille block and wheel skirts, is this is a complex shape that needed to be planned with sound theory. Unlike those other aero mods, which were essentially no-brainer parts-removal and gap-fillers, this one had the potential to actually increase drag if not shaped properly.
I'll probably attempt to refine it a bit more. I paid fairly close attention to getting the top right: I stayed at a fairly conservative 10 degree taper, where the Prius and A2 are slightly more aggressive with angles around 12-13 degrees. Also, I didn't pay enough attention to the sides - they were different angles, and also the side/top transition should be radiused, not sharp.
The next version should also be made of something a little more durable than cardboard (I'm thinking foam core & fiberglass). One of the things that interests me most about making a permanent rigid "hatch tail" is that it could serve as a mounting point for a more extreme, extended version of itself - something to get the car even closer to that ideal teardrop shape for highway use.
But I think before I do that I'll move on to other mods first which - based on what I learned here - may have the potential to yield bigger results (smooth undertray comes to mind).
darin AT metrompg D-O-T com, or here