BMW i3 - Low Hanging Fruit (improving aerodynamics)

[California98Civic]

Originally Posted by Snax

... didn't go any lower than 30 psi because the BMW spec is 33f /40r and it was already complaining about 35 /35. Regardless, it seems pretty clear that exceeding the maximum spec of 50 psi for the tires is likely not productive. Much over 45 may or may not be of measurable benefit. Dropping below 40 however seems to have a clear penalty based upon the averages.

3-4% over OEM pressure specs is significant. That's measurable extra range.

Originally Posted by Snax

Even so, I have decided to keep pressures at 45 / 50 for now as I like the way the car feels there. Lower pressures just feel mushy.

That's exactly what I thought looked ideal in your data. Roughly 10 psi over recommended as the ceiling.

[Snax]

Originally Posted by California98Civic

3-4% over OEM pressure specs is significant. That's measurable extra range.

I've come around to agreeing with that as I have spent way too much time trying to mold numbers from the online calculator here into working with a pure EV more directly in miles per kWh in my own spreadsheet.

I figure that my i3 uses approximately an additional 150W for every mile per hour faster it travels. If the vast majority of that can be attributed to tire rolling resistance and a Crr guess of 0.007 at BMW recommended pressures, bumping it up to near max and the lowest rated rolling resistance of about 0.006 for the EP-500s nets about 21 watts saved per mile per hour.

Put another way, that's an extra 13% range at 30 mph.

The mirror delete I think I have figured a little more accurately now, likely saving about 1.0% at 65 MPH. It's closer to 0.7% at 30 MPH and 2.2% at 75 MPH if it does not improve Cd at all. I am betting however on at least a 0.01 improvement there which will pay a larger dividend.

All combined in a best-case calculation, it should be about 11.5% better at 65 MPH.

But it's time to quit estimating and making educated guesses. I need to get some serious baseline numbers with the stock mirrors before I rip the door panels off to remove them.

[Snax]

After about 30 miles of baseline testing, I'm having trouble validating my OEM dynamics. Between crosswinds and traffic, getting good data is a nightmare!

Even so, it looks like I have way overestimated the baseline rolling resistance which is closer to 82W per mile per hour. Cd was looking closer to 0.34, but with a sustained 90 degree crosswind kicking 10-15 mph right now.

Apparently I'm going to have to go out at 3 AM to get good measurements.

[Snax]

1:30 AM; Close enough!

I'm beginning to think that BMW has put a bogus number out there for the Cd of this car, the frontal area, or the mi/k-watt meter in the car is low by 14%. Their claim is 2.38 square meters and Cd of 0.30, but I can't make that work at higher speeds without bumping it up to an assumed 0.34 for the numbers to fall where they should on the curve.

At least I am seeing consistency with them. I was able to reproduce nearly the same numbers in 22 degree cooler weather with calm air, so at least there is good reason to hope that anything I change will produce a repeatable measurable result if it changes anything.

For the purposes of testing my mods, I am just going to bake the 2.38 and 0.34 into base curve calculations and work from there.

[aerohead]

Originally Posted by Snax

1:30 AM; Close enough!

I'm beginning to think that BMW has put a bogus number out there for the Cd of this car, the frontal area, or the mi/k-watt meter in the car is low by 14%. Their claim is 2.38 square meters and Cd of 0.30, but I can't make that work at higher speeds without bumping it up to an assumed 0.34 for the numbers to fall where they should on the curve.

At least I am seeing consistency with them. I was able to reproduce nearly the same numbers in 22 degree cooler weather with calm air, so at least there is good reason to hope that anything I change will produce a repeatable measurable result if it changes anything.

For the purposes of testing my mods, I am just going to bake the 2.38 and 0.34 into base curve calculations and work from there.

Wouldn't their numbers come from the composite EPA dynomometer testing protocols,plus using a crosswind-averaged drag coefficient,fully-warmed,at EPA test weight (300-pounds added to front seats)?

[Snax]

Originally Posted by aerohead

Wouldn't their numbers come from the composite EPA dynomometer testing protocols,plus using a crosswind-averaged drag coefficient,fully-warmed,at EPA test weight (300-pounds added to front seats)?

Stuff I have little idea about . .

It's probably more fair to go with telemetry error and leave it there, though I do have 15mm wheel spacers installed which bring them out flush with the fenders. I'd be surprised if they bump the Cd by 0.04, but the 's' version of the car has a 0.33 Cd with 20mm wider tires and flares, so that certainly seems possible. I was not aware of that difference until now and will likely test it in the next day or so.

[Snax]

Lowest of the fruit . .

[redpoint5]

That even looks cool...

of course the lowest fruit is driving slower, but I understand that the question pertains to mods of the machine, not mods to the operator.

[Snax]

Thanks. Will have to see how well the execution can reflect the idea though.

I'm going to try fiberglass for the skirts if cardboard shows good results there.

Update on the Cd issue after some preliminary testing this afternoon: Wheel spacers suck!

They were apparently responsible for the 0.03 Cd difference from what I just recorded with them removed. At 65 mph, that works out to nearly a 9% hit! I'm going to repeat my testing in wee hours of the morning again to be sure. There was a fair amount of traffic, but after multiple relatively clear stretches the numbers seem valid.

I'm going to leave them off from now on except for autocross and hillclimbs. (Planning on getting some much wider wheels and rubber for that which will not stay on the car much.)

[Snax]

Table and graph matching the data before and after removing the spacers.

Edit: Just noticed that my column headers were incorrect where it says "mi / kWh Mech". Both columns should be "mi / kWh Aero".