Where do you guys do your A-B-A testing?

[Xist]

Somehow, I over-slept two nights in a row. On Sunday, I slept through my alarm, which should have come back to haunt me every five minutes, but I still managed to wake up after Church started, and then I did not get any sleep last night. I was thinking of all of the things that I could be doing. If my roommates are out-of-town, I could have vacuumed and cleaned the carpet. I would have bought a battery terminal brush, but I could not find any auto part stores open in the middle of the night, and Walmart does not seem to carry it. I just started Spring Break and now that I completed my accelerated Global Health class, school seems really slow!

I decided to A-B-A test my grill block:

Yes, that is fake, it is just part of some guy's campaign poster. I figured that it was about time that I figured out if it did any good. I had been planning on using a one mile stretch of road a few miles from here:

It seemed perfect because the street is only one mile long and I do not know why anyone would take it unless they lived in the neighborhood. There are some side streets to the west, but a railroad track to the east, and no crossings, and then halfway down, it crosses the freeway.

So, it turns out that there is a surprising amount of traffic at 3am. It took me several tries just to get two runs and then I gave up and came home.

I guess that the best thing to do would be to set up an Ultraguage for my car and just drive ten miles for each A, B, and A. Aerodynamics would play a larger role at fifty-five miles per hour than coasting from fifty to twenty anyway.

[MetroMPG]

The faster you can drive, the better your data (when testing aero mods).

The route I use is a dead straight, nearly level 1.6 km (1 mi) stretch between the start/stop markers, with about half that distance on either end for speed up / slow down. I wish it were longer, though.

The posted limit is 80 kph (50 mph), but you can go 90 (55) without raising an eyebrow, and 100 (60) isn't uncommon (though it would attract police attention). There's also very little traffic on it other than start/end of the day plus holidays.

So to summarize, I'd say an ideal route for repeated (e.g. AAABBBAA) runs would be:

1) close to home (easy to use)
2) little traffic
3) flat
4) long enough between start/stop markers for your MPG instrumentation to settle into a steady average
5) a high posted speed limit (or no negative consequences for driving above it)
6) bonus: sheltered from the prevailing wind (if any)

Don't forget to drive far enough to warm up the entire car before testing. Coolant at operating temp <> warm drivetrain/tires.

[KrautBurner]

ive been recording my commute (the reason I purchased the car)
its 60 miles each way
almost all 2 lane at 55-65mph
lots of elivation

[freebeard]

3) flat
...
6) bonus: sheltered from the prevailing wind (if any)

I don't have much for instrumentation—there's a gas gauge but it's off and so is the speedometer. So I was using a freeway overpass to do coastdown testing with ~30mph drop against a stopwatch, but traffic is the problem. So I'm thinking about a hill that kills cars from the local community college on the uphill side. I can crest the hill at the legal limit and then see if I gain or loose speed before a landmark [driveway] before one of two exits a half-mile apart. I think the grade would amplify the differences. The engine introduces a lot of variables, but gravity is inexorable. A hand-held aneometer out the window could quantify headwinds.

The best spot for shelter from wind would be the Van Duzer Corridor. There are long straight stretches of 2-lane blacktop with 80+ foot trees right down to the road on both sides. It's almost like a tunnel. Traffic and wind are minimum at 4am.

[mcrews]

I always use a freeway/highway. And always do 5 miles up and 5 miles back.
Reason: so I can set a baseline of 50 55 60 65 70 mph and the corresponding rpm and mpg

[Xist]

Heh. My next A-B-A testing will be before I change my oil pan and after!

[freebeard]

Today I went to test the spot I wanted to use for A-B-A testing. The turnaround at the top is an overpass where you can preview the uphill traffic, then a downhill onramp so it's easy to get up to raffic speed as you go over the top. Then there's an underpass and merge ramps going back uphill.

My speedometer is optimistic to the tune of 65 is really 55. I did an A-A test—65 at the top gave 80 at the bottom, then 55 at the top gave 75 at the bottom. This was on a weekday afternoon; I might go out some night and try from 25mph.

It's a nice circuit, but it's just too fast.

[MetroMPG]

The faster, the better for testing aero mods.

At low speeds, a small mod will disappear into the noise of statistical variability. But it will stand out at higher speeds.

[freebeard]

Most folks are still encouraging me to slow down.

Maybe the best solution is to pass the easy turnaround and let it roll out on the flat 1/2 mile that follows. That will slow it back down. Then use the demand traffic light instead of the half-cloverleaf.

I like the idea of potential-->kinetic energy as an input.