Partial Kammback prototype tested - preliminary results

[MetroMPG]

Originally posted: 07-16-2006, 07:13 PM

Spent about 45 minutes playing this afternoon. No, it's not tested yet, and was more of a design exercise than anything else for the moment. I may reinforce and test it if the right weather comes up in the next day or 3 (before it falls off/gets wet). Otherwise, it'll come off having served its purpose as a learning tool for how V 2.0 should look.

Constructive feedback welcome!







Above: what this crappy picture from my antique 1 MP camera intended to show was that Metro95 was right: it doesn't seriously obscure rear vision. The boat-tail cuts of 1/5 of the rear window view, and though you can't tell from this photo, the horizon on level ground is just below the "roof", so full visibility is retained of following traffic.

[MetroMPG]

My thoughts while making this:

- compound curves are hard to get right (I got the "roof" of the tail eventually, but not the sides right).

- I need to round the transition from roof to sides; right now it's a sharp angle (unlike the car's shape ahead of it).

- The side-of-hatch to side-of-tail transition is messy. I really need to start the boat-tail sides at the forward edge of the hatch sides. Also, you can't tell in the photos, but the right side is tapered further than the left side (at first I was focussed entirely on roof taper, and didn't taper the right side as much as it looks like I could have).

- I tried to project the existing tapering lines of the bodywork into the tail without any big changes, but I could possibly be a little more aggressive. Only tuft-testing would say for sure (looking for clean flow, right to the back of the tail)

- Ryland: I had the same thought about getting a junk hatch. I know where there's a '95 sitting in a farmer's field about 20 minutes away.

- Materials? If this is worth making a "good" copy, I was thinking: fibreglass over high density foam (messy, complicated, time consuming, little experience with the materials); junk hatch with attached sheet metal, faired & painted (also time-consuming, but I'm more familiar with the materials & methods); lightweight plastic or aluminum frame covered with sheet plastic/sheet metal... Plexiglass? Might be hard to form into compound curves (the stuff I've used in the past was fairly brittle. Also, how to join the top & sides in a rounded transition?)

- Not sure about dimples or VGs though. Since I'm an amateur, I'm just trying to stick to the basics: reducing the size of the trailing wake by tapering the shape as much as possible without causing flow separation. Also, I recall reading that dimples don't work for automotive scale/shapes.

- a finished version which I leave permanently on the car could serve as an attachment point for an extended version for highway use.

[MetroMPG]

Public reaction...

I've never had too much unsolicited public response to the wheel skirts, but the cardboard partial boat tail drew a small crowd. A question and answer session ensued in which at one point my sanity was debated.

[MetroMPG]

I'd prefer a more rounded top-side transition if possible. Sharp angles can trip smooth airflow into unwanted vortices, according to people smarter about aero than myself.

Fiberglass... I did some searching on how people make custom body kit add-ons, and it's a pretty complicated process:

- Build the prototype on the car out of whatever combo of materials you want (wood, metal, high density foam)

- make a mold by coating the prototype with fiberglass

- use the mold to make your good copy...

Those are the basics, as far as I can tell, anyway.

[MetroMPG]

Possible construction techniques:

One-Off Construction Using GRP/Urethane Foam Composite
http://www.rqriley.com/frp-foam.htm
(for making the car - trike - body shown below)



Also:
http://www.airfieldmodels.com/inform...cowl/index.htm

[MetroMPG]

Did a non-rigorous test - A-B only. (Cardboard/duct tape prototypes don't lend themselves to removal & reinstallation, so sorry, no A-B-A.)

That said, conditions were PERFECT. Next to no wind.

Results @ 55 mph (88.5 km/h)

- with partial boattail: 61.23 mpg (US) avg.
- without boattail: 59.85 mpg (US) avg.
- partial boattail gain: 1.38 mpg (US) / 2.3% above "stock"

Conditions:
Observed at: Grenadier Island 18 July 2006 4:00 PM EDT
Temperature 28°CPressure/ Tendency 101.5 kPa↑Humidity 63 %Humidex 36 Dewpoint 20°CWind calm DRLs, alternator ON, fan ON @ 3

W E

Partial boattail
59.8 61.6
60.2 63.3
59.4 63.1

61.23333333 avg

Non-boattail
57.2 61.7
58.7 61.8

59.85 avg.

Observations
This is about in line with what I predicted, based on the gains seen from grille block & wheel skirts (5.7% combined).

The difference between this mod and those, is this is a complex shape which can be optimised with tuft testing. The others were essentially no-brainer gap-fillers.

[MetroMPG]

For testing, I run alternator ON.

And today I ran DRLs (2-lane stretch of highway, oncoming holiday traffic), and the interior fan at "3".

Also, I normally test with cruise control to remove my bias & right foot from the experiment and leave it all up to an impartial and more consistent microprocessor. All of which is to say: no DWL. My cruise sucks at getting good FE, even on mostly level roads (speed corrections, even minor ones, are not "gentle").

[MetroMPG]

I should have tested at 85 km./h, not 55 mph.

Why? Because all of my other tests have used metric speeds, and I can't compare today's numbers with past tests done at 85. (55 = 88.5 km/h).

I did it at 55 mph because I was barrelling towards the "start" line when I suddenly remembered I hadn't reset the SG to report in metric. So I just set the cruise to the nearest multiple of 5 and kept going. Lesson learned.

[MetroMPG]

I posted a fresh article at MetroMPG.com about the "Kardboard Kammback" described in this thread.

(And yes, I realize it's not technically a Kammback )

• MetroMPG.com: Kardboard Kammback: testing a partial boat tail prototype

In it, among other things, I outline the reasons for grafting cardboard on the *** end of the Firefly in an attempt to mimic the shapes of these slick hatchbacks:





FYI, you are forgiven for not recognizing the car on the right. It's the 99-05 Audi A2, which of course was never available in North America. With its at the time world-leading Cd for a production car equal to that of the Insight, the TDI version of the A2 was rated at 3L/100 km, or about 78 mpg (US).