Improving the Model 3 Aerodynamics (for range)?

[TexaCali]

The Model 3 is already an incredibly aerodynamic car, but I wonder what improvements might be possible? Unplugged Performance claims a 21% reduction with their mods (front and rear spoilers, lowering), but I'm highly skeptical (eps since it has been available for a few years and I don't see any folks talking about it).

How much could you reduce drag while remaining road legal and practical? Removing the side mirrors would be an obvious starting point (if it was legal), but how much would you gain? How much could you gain from a front spoiler? Are there rear spoiler designs that would provide meaningful gains without extending past the rear bumper (i.e. road legal)? Other improvements besides lowering (which isn't practical for road use)?

To be clear, my interest would be range improvement (not cornering or track use) with simple, user installed mods that are reversible, and this is mainly just academic curiosity. "Miracle" aerodynamic gadgets you stick on your car to boost mileage have been sold by snake oil salesmen for decades, so I'm instantly skeptical of kits like the Unplugged Performance mods, but still curious what might be possible.

 

[MLXXXp]

I wonder if covering the rear fenders would help (like the original Honda Insight)?

 

[RickCRNA]

Why don’t you satisfy your “academic curiosity” and pose your questions to the Tesla automotive engineers? If they thought the aerodynamics could be improved to increase range, don’t you think they would done it by now?

 

[TexaCali]

Why don’t you satisfy your “academic curiosity” and pose your questions to the Tesla automotive engineers? If they thought the aerodynamics could be improved to increase range, don’t you think they would done it by now?

There are lots of factors that influence car design, not just aerodynamics. I can think of several reasons Tesla engineers would selected a less than perfect aerodynamic design:

1. Appearance. A raindrop like blob would probably be the most efficient aerodynamically, but who wants a car that looks like that? Same might be said for some fairings and spoilers - they could potentially limit the market appeal.

2. Cost. Some aerodynamic additions might not be deemed worth the extra manufacturing cost.

3. Safety. Some designs might reduce visibility, compromise structural integrity, or complicate other aspects of the safety systems.

4. Regulatory (i.e. removing mirrors and using cameras instead).

5. Practicality. Lowering the car is an excellent example.

I have no doubt the aerodynamic engineers at Tesla know how to reduce the drag of the car significantly. I'm also certain the tradeoffs and compromises for doing so were debated extensively. Changes that were rejected for appearance/market acceptance, or cost reasons would obviously be the most interesting. I also would not expect Tesla to engage in a public discussion on these design choices. Hence the thread here in hopes that some forum members have experience with aerodynamic mods, or maybe work in computational fluid dynamics and can contribute to the discussion.

 

[TexaCali]

I wonder if covering the rear fenders would help (like the original Honda Insight)?

Interesting question, esp since the aero wheel covers already reduce some of the turbulence over the rear wheel though obviously not as much as having the wheel covered by body panels. The Insight also had a teardrop tail design that was chopped off (because a big long teardrop tail would have obviously been impractical and not road legal).

BTW, here are a few Cd numbers for consideration:

Honda Insight - 0.25

Tesla Model 3 - 0.23 (there are several BWM, Mercedes, and Audi ICE cars with Cd's this low or lower)

BMW 5 Series (2017+) - 0.22 (look at all those grills, intakes, and front trim work plus lack of aero wheel covers...and it's still lower than the 3!)

Mercedes EQS - 0.20

2021 Tesla Model S - 0.208

GM EV1 - 0.19 (also has covered rear wheels like the Insight, and a more pronounced teardrop tail)
VW XL1 - 0.19 (again, covered rear wheels and a very teardrop tail. Also note cameras instead of side mirrors)

 

[TexaCali]

Interesting article on the future of aerodynamic sedan designs:

In the shift to electric, the three-box sedan is obsolete: Here’s why

The era of the boxy sedan doesn’t make sense in the era of the electric car—but curvaceous kammbacks do.

www.greencarreports.com

 

[MountainPass]

Some ideas:

• reducing ride height
• increasing tire pressure to 51psi (or the maximum cold pressure of your tire)
• completely flat wheel covers
• taping up all the gaps in the bodywork
• removing side mirrors
• extended rear spoiler
• making sure the brakes are not dragging at all

 

[Gauss Guzzler]

BTW, here are a few Cd numbers for consideration:

Honda Insight - 0.25
Tesla Model 3 - 0.23 (there are several BWM, Mercedes, and Audi ICE cars with Cd's this low or lower)
BMW 5 Series (2017+) - 0.22 (look at all those grills, intakes, and front trim work plus lack of aero wheel covers...and it's still lower than the 3!)
Mercedes EQS - 0.20
2021 Tesla Model S - 0.208

Note that the drag coefficient is a meaningless and is inherently misleading number which often means the opposite of what you might think it does. The term "coefficient" means that it has to be used in a formula with at least one other number to have any meaning. That number in this case is frontal area.

So the reality is that a gigantic car like the BMW 5 (Cd=0.22) might have 30% more drag than a normal car like the Insight (Cd=0.25) despite the 12% lower coefficient.

And if you made a huge BMW-sized Insight but kept the mirrors and wheels the same, the Cd would likely drop into the teens simply from being bigger - that's why Cd values taken out of context are often opposite of what you'd expect. It's easy to make a big pickup truck with a low Cd but hard to do on a compact car.

 

[TexaCali]

I know Kyle (Out of Spec Motoring) lowered his Model 3 before his Cannon Ball run, and I've read many statements about how lowering improves efficiency, but I've never seen any numbers. For sustained highway travel at 70mph and up (western US), how much improvement does one see with lowering? And how much lower do you have to go to see significant improvement? 1"? 2"?

I have a 2022 Model 3 RWD (60kWh LFP battery, 272mile stated range, Aero Wheels) and will be driving multiple very long road trips this summer, so looking for any advantage I can get.

 

[jackmott]

The closest thing I've seen to numbers is from Unplugged Performance but it is just CFD (software simulation) which came up with ~8% reduction in drag from their lowering springs, which would amount to something like ~4% better range on the highway. Trying to measure whether that 4% range increase is real with casual real world tests isn't really possible. Would have to wind tunnel it or at least get on an empty oval and do a few runs before and after and hope wind is steady.

 

[GtiMart]

Did unplugged performance also have a front lip or bumper to go with that? Or am I mixing this up with another test?

 

[jackmott]

Did unplugged performance also have a front lip or bumper to go with that? Or am I mixing this up with another test?

They did lowering, front lip, and bigger rear spoiler, but the 8% drag reduction was from the lowering alone. Again, all software simulation figures.

 

[TexaCali]

They did lowering, front lip, and bigger rear spoiler, but the 8% drag reduction was from the lowering alone. Again, all software simulation figures.

Yep, their white paper does a nice job breaking it down, but it is all simulation. I realize real world testing is very difficult to do, but simulation only tells you so much.

FWIW I am planning to install the UPP lip and high efficiency spoiler, and may or may not lower the car. I'm planning some before and after real world testing. Here is my proposed methodology:

Fully charge and precondition car before testing to assure warm battery, check tires are at max cold PSI
All testing done around 3AM to avoid other vehicles skewing test results
All testing done with HVAC OFF - so we are only measuring what the car uses
only measure consumption while car is established at speed (plan to use 70mph) - will not include portion while slowing down/turning around/speeding back up
measure both directions on same road and average the two (correct for wind and elevation changes)
measure on multiple nights to establish repeatability and understand error bars

I will at least do this with the base car and again with the UPP lip and rear spoiler. If I lower I will measure with all 3, but won't be pulling the mods off to test each individually, unless someone wants to pay me LoL. I also realize this methodology isn't perfect, but I think it is about the best one can do without a dedicated test track, wind tunnel, and hooking a bunch of test equipment up to the car. Thoughts?

 

[jackmott]

It is possible to do aerodynamic testing in the field, but that probably isn't enough. For instance out and back doesn't just cancel out the affect of wind if the wind is different. Even humidity/air pressure differences could be bigger than the expected difference of the spoiler/lip etc.

When cyclists do this kind of thing (measuring small aero differences in the field) they pick a time and place with minimal wind (tree cover, early morning) are carefully tracking power and velocity so they can do statistics on it later, not just looking at average consumption for each trip. There are data logging tools for the Tesla that would let you get that into it if you wanted to.

Some inspiration:

A challenge to cycling aerodynamicists

Training and Racing with a Power Meter is THE book for Cyclists looking to take their training to the next level. Hunter Allen and Dr Andrew Coggan

www.trainingandracingwithapowermeter.com

https://www.youtube.com/c/JulianEdgar

 

[TexaCali]

Thanks! Great info! And yes, I realize wind and temp can change during the run, so I'm planning legs that are 5-10 miles - long enough to stabilize, but hopefully short enough that conditions remain relatively constant.

Also note, a side wind hurts you in both directions, as you are always steering into the wind slightly to keep straight. That's another reason for my 3AM sessions - I'm hoping to find some very calm nights. Obviously if it is windy or unsettled then that would be a bad night to test.

Just thankful this isn't a pedal car

 

[KenC]

Yep, their white paper does a nice job breaking it down, but it is all simulation. I realize real world testing is very difficult to do, but simulation only tells you so much.

FWIW I am planning to install the UPP lip and high efficiency spoiler, and may or may not lower the car. I'm planning some before and after real world testing. Here is my proposed methodology:

Fully charge and precondition car before testing to assure warm battery, check tires are at max cold PSI
All testing done around 3AM to avoid other vehicles skewing test results
All testing done with HVAC OFF - so we are only measuring what the car uses
only measure consumption while car is established at speed (plan to use 70mph) - will not include portion while slowing down/turning around/speeding back up
measure both directions on same road and average the two (correct for wind and elevation changes)
measure on multiple nights to establish repeatability and understand error bars

I will at least do this with the base car and again with the UPP lip and rear spoiler. If I lower I will measure with all 3, but won't be pulling the mods off to test each individually, unless someone wants to pay me LoL. I also realize this methodology isn't perfect, but I think it is about the best one can do without a dedicated test track, wind tunnel, and hooking a bunch of test equipment up to the car. Thoughts?

Why not let the ABRP app measure your efficiency at 65mph?

 

[TexaCali]

Why not let the ABRP app measure your efficiency at 65mph?

I knew ABRP could get data from your car, but I can't find anything on how it calculates efficiency at 65. Do you know how it works?

I still think the method I outlined above is better since we have more direct control over the variables, but the ABRP approach certainly is easy enough.

 

[KenC]

I knew ABRP could get data from your car, but I can't find anything on how it calculates efficiency at 65. Do you know how it works?

I still think the method I outlined above is better since we have more direct control over the variables, but the ABRP approach certainly is easy enough.

No, I don't know, but you can read their small explanation about measuring efficiency at different speeds; but I would think using it would eliminate some of those measurement variations, since its approach would be consistent between test runs.

 

[Rothgarr]

I was surprised that the body panels aren't plastic like a Saturn, or even fiberglass. I assume those are lighter than the steel and aluminum panels, no?

 

[oreobbq]

Yep, their white paper does a nice job breaking it down, but it is all simulation. I realize real world testing is very difficult to do, but simulation only tells you so much.

FWIW I am planning to install the UPP lip and high efficiency spoiler, and may or may not lower the car. I'm planning some before and after real world testing. Here is my proposed methodology:

Fully charge and precondition car before testing to assure warm battery, check tires are at max cold PSI
All testing done around 3AM to avoid other vehicles skewing test results
All testing done with HVAC OFF - so we are only measuring what the car uses
only measure consumption while car is established at speed (plan to use 70mph) - will not include portion while slowing down/turning around/speeding back up
measure both directions on same road and average the two (correct for wind and elevation changes)
measure on multiple nights to establish repeatability and understand error bars

I will at least do this with the base car and again with the UPP lip and rear spoiler. If I lower I will measure with all 3, but won't be pulling the mods off to test each individually, unless someone wants to pay me LoL. I also realize this methodology isn't perfect, but I think it is about the best one can do without a dedicated test track, wind tunnel, and hooking a bunch of test equipment up to the car. Thoughts?

You don't need fancy equipment. Just do this test a thousand times for robust data. Satisfying your own curiosity is payment enough, let's go!