Massive air dam on '98 Geo Metro (Firefly), A-B-A coastdown test; Update: version 2

[anthonye81]

No matter what the results show, I'm liking this method of experimenting with something large and obvious! I know aero design is as much about optimisation through small details, but there's something satisfying about slapping on a big sheet of cardboard and seeing if drag is affected. It would be great to see the effects on CFD or smoke trails in a wind tunnel.

As for the 14.5mpg Pinto, maybe the journalists hobbled their car before the test thus making it easier to improve?

[Big Dave]

My air dam was massive and made of very tough rip-stop conveyor belting. I was very aggressive, allowing only 1.5 inches ground clearance on both the front air dam and side skirts.

The combo was worth an extra 0.75 MPG (3%).

But they did drag more than I liked. Maybe a little more clearnce...

[micondie]

I remember reading years ago in some hot rod magazine that when you calculate frontal area of a car you should measure from the ground up because the turbulence of the air going under the car contributes at least as much if not more to the wind resistance as the equivalent frontal area. This assumes no belly pan.

[Christ]

Originally Posted by micondie

I remember reading years ago in some hot rod magazine that when you calculate frontal area of a car you should measure from the ground up because the turbulence of the air going under the car contributes at least as much if not more to the wind resistance as the equivalent frontal area. This assumes no belly pan.

I don't think this is always true, but I do it this way anyway. This is why I say that lowering a car reduces more frontal area than just the tires and suspension components.

Of course, I also believe in tucking the exhaust system up closer to the body and insulating it as well as possible

[ NOTE from admin - discussion of exhaust insulation split to this thread: http://ecomodder.com/forum/showthrea...ust-10121.html ]

[MetroMPG]

Originally Posted by EVDRVR

My Solectria weights the same all the time.

Lucky you - I've driven a Solectria Force a few times. Very slick little EV!

Anyhow, I thought for your records if your not already doing so you might want to record fuel level and other things that might increase or decrease the weight if you want to compare these tests over time.

I haven't been tracking fuel level during these coastdown tests, mostly because I don't think tests done on different days are directly comparable. I feel only tests done on the same day in the same weather conditions are comparable with the smallest margin of error. EG. temperature changes have a large effect on rolling resistance, and to a lesser extent on drag (as air density changes).

But you're right - the changing mass of the car might show up in the results.

palemelanesian documented an effect on his fuel economy depending on the amount (changing mass) of fuel in his tank. See http://ecomodder.com/forum/showthrea...html#post27765

[MetroMPG]

Originally Posted by anthonye81

there's something satisfying about slapping on a big sheet of cardboard and seeing if drag is affected.

Agreed 100%!

Plus it's much harder for us to test small optimizations, vs. relatively large changes, since we just don't have the data resolution in coastdown testing or controlled-as-possible A-B-A instrumented runs (eg. based on ScanGauge MPG readings).

[MadisonMPG]

Would keeping less than a half tank be better if the gas station was out of my way?

[tomicdesu]

Originally Posted by anthonye81

As for the 14.5mpg Pinto, maybe the journalists hobbled their car before the test thus making it easier to improve?

If you read the article (available on the C & D website for free as a PDF; google search for the full exact title and it pops up) you'll see that they used a 5th wheel and some inline fuel flow meter. It's hard to measure low-rate liquid flow, especially when it's pulsating like a mechanical fuel pump is.

Assuming that the readings were at least self-consistent, the percent change is what I'd go with, rather than actual MPG numbers. Yeah, that car even driven badly shoudl get 20 on the highway.

[tomicdesu]

Originally Posted by MetroMPG

Agreed 100%!

Plus it's much harder for us to test small optimizations, vs. relatively large changes, since we just don't have the data resolution in coastdown testing or controlled-as-possible A-B-A instrumented runs (eg. based on ScanGauge MPG readings).

The only reliable method I've found to characterize changes is to average many tankfuls over the same route. Commuting to work is ideal for that, assuming it's substantially useful terrain.

For me, my commute is nearly ideal; Los Angeles to Irvine [yes I'd prefer public trans but Orange County is ideologically allergic to it], flat, smooth steady 55 - 60 mph 3 days a week.

I keep a record (spreadsheet) of every single fillup (every one) noting odometer, etc.

If I want to test a change, I make it then fillup the car, and make a note in the spreadsheet. Then I drive on that change for a couple of weeks, usually 600 miles. That way, fillup/topoff/expansion errors are minimized; even a 1/2 gallon error over dozens of gallons gets swamped.

[MetroMPG]

I'd argue that tank to tank testing is madness! Way too many variables.