My new way of measuring on-road drag

[freebeard]

Separation line marked on outside of aerofoil:

It looks like the next iteration needs an angled or tapered chord.

You're limited by the width of the extrusion. Were it deeper the back edge could be maintained while the leading edge reaches further forward. It would look constant from the front, but catch air moving sideways from the stagnation point. Free air?

[Snax]

Frankly, I had given up on attempting to measure efficiency on my i3 until now. Either the battery thermal management was skewing results, or countless other factors that I could not really control for with how the car manages energy were. In an effort this evening to burn off some charge to verify that my DCFC system actually works (since the only fast charger within 50 miles was just completed last week 2 miles from my house), I decided to do some very informal testing with a throttle stop to that end.

Obviously, pedal travel limiting is the only way to do this, so I used a stack of washers taped under the pedal to limit my top speed to around 70 mph on the flats.

Fortunately the display provides relatively accurate feedback about throttle percentage, so I could see that it would peak and hold at the limit of travel. Doing so, I was able to verify window down drag at about +6% over a distance of just a couple of miles. Frankly, I was surprised to see my top speed drop by 4 mph just from that. Now I can get a little more serious with trying to improve the car instead of attempting to compile annualized data which can swing hugely with varied weather conditions.

I'd thought about doing a throttle stop before, but for whatever reason, just never bothered since nobody else seemed to talk about it.

This is a game changer.

[JulianEdgar]

Originally Posted by freebeard

It looks like the next iteration needs an angled or tapered chord.

You're limited by the width of the extrusion. Were it deeper the back edge could be maintained while the leading edge reaches further forward. It would look constant from the front, but catch air moving sideways from the stagnation point. Free air?

Interesting idea. Or maybe a twisted angle of attack, a la propellor blades? Anyway, it will do me as it is. I am happy it is working, and law of diminishing returns applies for ultimate tweaking.

(And it was on a windy day.)

[JulianEdgar]

Originally Posted by Snax

Frankly, I had given up on attempting to measure efficiency on my i3 until now. Either the battery thermal management was skewing results, or countless other factors that I could not really control for with how the car manages energy were. In an effort this evening to burn off some charge to verify that my DCFC system actually works (since the only fast charger within 50 miles was just completed last week 2 miles from my house), I decided to do some very informal testing with a throttle stop to that end.

Obviously, pedal travel limiting is the only way to do this, so I used a stack of washers taped under the pedal to limit my top speed to around 70 mph on the flats.

Fortunately the display provides relatively accurate feedback about throttle percentage, so I could see that it would peak and hold at the limit of travel. Doing so, I was able to verify window down drag at about +6% over a distance of just a couple of miles. Frankly, I was surprised to see my top speed drop by 4 mph just from that. Now I can get a little more serious with trying to improve the car instead of attempting to compile annualized data which can swing hugely with varied weather conditions.

I'd thought about doing a throttle stop before, but for whatever reason, just never bothered since nobody else seemed to talk about it.

This is a game changer.

Excellent news. Windows up / windows down I think is a critical evaluative test for anyone trying to quantify drag changes using any technique.

[JulianEdgar]

So I tried some rear Edgarwits to see if I could reduce wake size. These were additional to the front ones, and taller. GOE222 true aerofoil profile.



I did some throttle stop testing with the following results (km/h at 30 per cent throttle, low-moderate gusty headwind):

• No rear Edgarwits: 95
• 1 finger rear gap: 93
• 2 finger rear gap: 95
• 3 finder rear gap (ie about parallel): 96
• 4 finger rear gap (ie diverging): 96.5

That would make 4 finger vs std about 3 per cent improved in drag.

4 finger gap:



For people interested, but having other cars, don't forget Insight boat-tails a lot with its narrower rear track.

Interesting to see the better performance at the rear with the diverging duct - opposite to what occurred at the front.

I'll do some more testing on a still day, but to me this result is borderline. I'd need to see a big jump in drag reduction on a still day to go in this direction.

[M_a_t_t]

One of the points you make is to have a tps readout. How import would you say that is? My car is a 95 and so it has OBD1, so no obd reader. I installed an old racheting hand brake lever as a hand throttle so I don't believe it will vary any if at all. I feel like it would be fine, but I would like a second opinion.

[JulianEdgar]

Originally Posted by M_a_t_t

One of the points you make is to have a tps readout. How import would you say that is? My car is a 95 and so it has OBD1, so no obd reader. I installed an old racheting hand brake lever as a hand throttle so I don't believe it will vary any if at all. I feel like it would be fine, but I would like a second opinion.

That would probably be fine. Or you could just use a multimeter to monitor TPS output.

[aardvarcus]

I don't actually think the throttle position sensors are varying their output, rather the ecu/ computer is reinterpriting what that output means. Some of the newer chevy sensors have multiple potentiometers built in for redundancy, so the ecu is comparing those results and "deciding" the result.

Thus you would want to monitor the ecu output if possible. Not exactly sure on 95, my experience is with TBI (approx 94 back) and gmt800 (approx 99+).

[JulianEdgar]

Originally Posted by aardvarcus

I don't actually think the throttle position sensors are varying their output, rather the ecu/ computer is reinterpriting what that output means. Some of the newer chevy sensors have multiple potentiometers built in for redundancy, so the ecu is comparing those results and "deciding" the result.

Thus you would want to monitor the ecu output if possible. Not exactly sure on 95, my experience is with TBI (approx 94 back) and gmt800 (approx 99+).

At this stage, this technique is not endorsed for those cars with electronic throttle control or CVTs. So I am referring to mechanical throttle cars, where I found using a throttle stop, and then monitoring throttle blade position, best for consistency.

[aardvarcus]

Yes, i was thinking his 95 was a vortec but it looks like they were likely still tbi. ( 1995 was a Frankenstein year for some components, splitting from 94 down and 96 up). Thus nevermind.