Everything you ever wanted to know about auto aerodynamics.

[clarkspeed]

Well I am going down quite a rabbit hole now. I recently ordered Julian Edgar's newly released book. Very down to earth and all practical solutions to testing and tuning production car aero on a tight budget without CFD or wind tunnels. For example he uses a throttle stop to measure changes that affect drag. Stop the throttle at 10 or 15% and measure top speed at roughly 50 or 60 mph.  Make an aero change, then check top speed again.

 

Not cheap, but 500 pages packed full of info and lots of pictures. Highly recommended if you are looking to reduce drag and lift on your car.

Edited October 30, 2023 by clarkspeed
Spelling

[JMortensen]

I've watched a bunch of his videos on youtube and ended up buying a bunch of pressure pucks and five 1" manometers to do some track testing. Of course I bought all that shit and then put the car down and spent the next year doing other stuff. Still need to make a temporary dashboard to hold the gauges.

The one thing that he really bugged me with was his refrain that "you can't just slap a splitter, diffuser, and a wing on and make downforce" which he said quite a bit IIRC. I ended up linking in the youtube comments to AJ Hartmann's company, because that's exactly what he sells, and it obviously works. I haven't watched anything in a few months, but I did finally find one where he added I think it was a splitter and wing and was like: "HEY! That really worked!"

As you mentioned though, he has a ton of good info and practical techniques in his videos, I can imagine the book will be better. Did a lot of coast downs for drag measurement too as I recall. Anyway, thanks for the heads up. Will purchase.

[clarkspeed]

On 10/30/2023 at 3:28 PM, JMortensen said:

I've watched a bunch of his videos on youtube and ended up buying a bunch of pressure pucks and five 1" manometers to do some track testing. Of course I bought all that shit and then put the car down and spent the next year doing other stuff. Still need to make a temporary dashboard to hold the gauges.

The one thing that he really bugged me with was his refrain that "you can't just slap a splitter, diffuser, and a wing on and make downforce" which he said quite a bit IIRC. I ended up linking in the youtube comments to AJ Hartmann's company, because that's exactly what he sells, and it obviously works. I haven't watched anything in a few months, but I did finally find one where he added I think it was a splitter and wing and was like: "HEY! That really worked!"

As you mentioned though, he has a ton of good info and practical techniques in his videos, I can imagine the book will be better. Did a lot of coast downs for drag measurement too as I recall. Anyway, thanks for the heads up. Will purchase.

Thanks Jon,

This book seems to be everything he has done for 25 years all rolled up and adjusted to the latest technology of the last year or so. I'm about 1/2 way through but I can't find any aero mod you can do to a production car he doesn't cover in depth somewhere in the book. 

 

Yes, he emphasizes no aero modifications should be made unless you measure the effect. It may not be what you expect and may upset the aero somewhere else. I am trying to wrap my head around how much lift a car, any car, produces. So for a production body style, you are basically trying to reduce lift, not create downforce. Exception being major mods like splitters combined with diffusers and wings. And as expected, lift and drag are 2 completely different things. Air dams reduce drag, splitters increase downforce forward of front axle only which can really lighten up the rear axle due to the lever arm.

 

One example on an air dam he tested a 4" deep and a 9" deep. The 4" reduced drag almost as much as the 9" and left greater ground clearance. He tests plenty of bolt on parts in the book and proves they work as intented. So obviously there are bolt on parts designed for specific cars that benefit lift or drag. But if you have to make choices or optimize the part, better be ready to test. Obviously a rear spoiler works on an early Z. But how tall and what angle is best to optimize? 

 

[clarkspeed]

I jumped ahead and looked at the section on splitters. Seems he was a little late to the game. But when he tested a small one it ripped off his car. He then goes on to explain how it works and how to properly shape one. Seems a slight curved bottom is better than flat. And some profiles of leading edge are better than others.

[clarkspeed]

Well I am about 40% through this book now and it does not necessarily match what I have read in magazines, Internet, or anywhere for that matter.  Many of these concepts are not intuitive.  I will try to list out some of the salient points I am tracking.  I don’t think Mr. Edgar would mind since I am only scratching the surface and promoting his book. In my own words.....

 

1. Cd, coefficient of drag, is not an exact measurement.  Almost impossible to measure without only the best wind tunnels.  Most manufacturers quote a number from less accurate wind tunnels and no one can prove or disprove what they post in their marketing material.  Coast down testing is a disaster at best, not repeatable.  Way too many variables come into effect at speeds below 60mph and a perfect consistent environment is impossible to achieve.  Edgar proposes just measuring the change in drag rather than trying to quantify it.
2. Vehicle aero is complicated.  More complicated than CFD analysis can fully cover.  Even if you can simulate the flows over a car, the flows under the car and side winds (yaw) cause unplanned effects.  Even the best OEM aero specialists have to experiment, measure, and rent wind tunnels.  And for OEM’s, style often takes precedence over aero.
3. Every standard vehicle shape creates lift.  Air speeds up when it is required to change directions quickly so the transition from grill to hood and window to roof speeds up the air and creates low pressure.  The amount of lift depends on the area of these surfaces.  Some lift is created in the rear transition from roof to rear window.  So most all cars without any major aero features create lift. The lift/downforce can be different at any point location front to rear on the top and bottom surfaces of a car. The overall lift is the sum of all the points. Really the only exceptions are supercars with splitters, flat bottoms, and diffusers. Not Camry’s.
4. Since most all cars create more lift the faster you go, then obviously it is not such a bad thing as long as it is not excessive.  Counterintuitive, more lift in the front than rear gives a feeling of stability.  Yes, you read that right. 
5. Drag and lift are related but totally different.  Decreasing lift usually comes at the expense of increasing drag and visa-versa.  Race cars obviously want as little lift as possible or even downforce.  The performance gains usually outweigh the increase in drag when done correctly but you have to be aware of both.
6. I will not go into all the aero tricks available to reduce drag and increase lift.  Edgar goes into pretty much any mod you can think of with pictures and suggestions to make more effective. However he emphasizes, modification to the underfloor, adding a flat bottom, is the best place to start and gives the most benefits. He has plenty of examples on how this should be done. Air dams are basically a crutch to prevent as much air as possible from going under the car and hitting all the "stuff" hanging down there. A flat bottom is more effective and can make negative lift. Most newer OEM designs exploit this. 
7. Even small splitters can make large downforce but it is in front of front axle, which creates lift in rear.  Same problem for rear diffusers and spoilers.  They create lift in front because they are behind rear axle.  You have to be careful with these mods to get a balanced car.

[calZ]

I don't know if I've shared this before on the forum, but my day job is an aerodynamic analyst/designer. I work on planes more than cars, but if there is anything in the book that seems like it doesn't make sense or isn't covered, I could probably help decipher it.  

 

Some comments:

1. Absolute value of Cd is useful when comparing different vehicles, but I agree with deltas being the most useful when designing or modifying something. As long as your modeling/testing is equivalent, changes in Cd are accurate and reliable, and that's all that matters. You know how your car performs at whatever its current Cd is, so whether it's going to improve or not is what you need to know, not the true value of Cd. We're kind of now getting to the point where we'll fly UAVs for smaller programs based solely off CFD, but anything big or that carries people gets multiple wind tunnel tests during development. 

 

2. Steady(well-behaved) flows are pretty easily modeled very accurately in CFD, but unsteady, messy flow is tough. You get into needing dynamic simulations which are very expensive computationally. It's trivial to run yaw simulations, so I'm not sure I agree that yaw is something that will surprise you if you're doing things on a computer or in a tunnel rather than just track testing. 

 

5. Rear wings will always increase drag. I actually worked on a case where an SAE car gained a bunch of downforce with a front wing at zero drag penalty due to the upwash from the wing keeping a ton of air from hitting the front tires, but that's a niche situation. Air dams, flat floors, and diffusers are essentially "free" performance. You can lose both drag and lift if done properly. 

 

6. I don't know if I agree with calling airdams a crutch, but an airdam without a flat floor is leaving a lot on the table. The less air you let under the car, the easier it is to manage and the easier it will be to make your diffuser effective. That's true with or without a flat floor. If you put a flat floor on but leave your front end the same, you're still going to generate a bunch of lift on the bottom of your front bumper for anything shaped like a Z. Modern cars are pretty good at not having much upward-facing area, but old cars with the diving board bumpers really need airdams. 

 

7. Yep, adding aero to just one end of a balanced car is an easy recipe for an unbalanced car. Adding just a mild wing to a car prone to oversteer or a small splitter to one prone to understeer would be pretty safe, but you always have to be aware of the whole system. 

[clarkspeed]

6 hours ago, calZ said:

6. I don't know if I agree with calling airdams a crutch, but an airdam without a flat floor is leaving a lot on the table. The less air you let under the car, the easier it is to manage and the easier it will be to make your diffuser effective. That's true with or without a flat floor. If you put a flat floor on but leave your front end the same, you're still going to generate a bunch of lift on the bottom of your front bumper for anything shaped like a Z. Modern cars are pretty good at not having much upward-facing area, but old cars with the diving board bumpers really need airdams. 

 

Thanks for your comments!  Cool job. 

I know virtually nothing about vehicle aero but i am a mech engineer so I can grasp the concepts. Even with my meager knowledge, Edgar's book seems to have some funky conclusions.

 

I just reread the flat floor section again last night.  What he recommends is a slightly rounded panel roughly starting at bottom of the front bumper to a smooth rounded transition knee point in front of front axle.  Then totally flat section back to where the diffuser starts.  And the diffuser being the minimum of another flat panel angled upward at 10 degrees or more to the rear bumper.  The start of the diffuser section seems to like being as far forward as possible.  With this arrangement, you want as much air as possible flowing under the car and no airdam.  So no extra drag and net downforce like a reverse wing.  I am still struggling with this.  It seems like what interference you do have with suspension arms and maybe wide wheels would screw this up.  Also it seems ride height would be a critical factor. For an S30Z I assume this would look like a short full width airdam, with the curved floor section attached at the bottom lip.

[JMortensen]

4. This doesn't seem counter-intuitive, as more lift on the front end will tend to make the car understeer, and that's the more inherently stable tendency.

5. I'm sure I read in a book somewhere that lift or drag translates directly to work. If the airflow is doing work it will show up as force on the car. So as calz says, if you clean up the airflow under the car so that it isn't turbulent, you can reduce drag and lift, as it has to do less work to get under the car - acting less like a parachute. If you have a lot of lift, that's a sign the the air is doing more work. Closing the huge hole in the front and ducting to the radiator prevents all that air from packing into the front and pushing up on the hood. This will reduce lift and drag. A spoiler on a Z reduces lift and drag as well (to a point, after which it will reduce lift and increase drag).

7. The further away from the axle the more leverage you have too. So a rear wing can be mounted lower and further back and make more downforce than the same wing mounted in cleaner air and higher in some circumstances.

Thinking back to what might be the same video about splitters, I think his view on the shape of a proper splitter is a little simplistic. Most of the time on a race car you're rule limited, or you have practical constraints that limit the shape, but in the more unlimited classes you see the upturned nose of a splitter or a raised tunnel entrance in the front. There are also interesting front wings on open wheel cars that we can learn a little bit from where they have unusual profiles: relatively high in front, flat and close to the ground for a bit, then sloping up at the back. Not unlike McBeath's drawings of what an undertray should look like, but with a more wing-y looking top side. 

Most low level race cars use flat splitters, but diffuser on the splitter kind of aim for the same sort of shape in the back as the wing. If you're going to design a splitter I'd suggest a close look at what AJ Hartman is doing. He's working with high level aerodynamicists, does lots of CFD and wind tunnel testing, and makes good stuff. Then he makes youtube videos telling you how to DIY it. Pretty amazing, really. Note the front lip shape. Sharp on top, rounded on bottom.
 

 

[JMortensen]

Here's the video with the CFD on the splitter diffusers: 

 

Edited November 15, 2023 by JMortensen

[calZ]

2 hours ago, JMortensen said:

5. I'm sure I read in a book somewhere that lift or drag translates directly to work. If the airflow is doing work it will show up as force on the car. So as calz says, if you clean up the airflow under the car so that it isn't turbulent, you can reduce drag and lift, as it has to do less work to get under the car - acting less like a parachute. If you have a lot of lift, that's a sign the the air is doing more work. Closing the huge hole in the front and ducting to the radiator prevents all that air from packing into the front and pushing up on the hood. This will reduce lift and drag. A spoiler on a Z reduces lift and drag as well (to a point, after which it will reduce lift and increase drag).

 

100%. A lot of people think diffusers are only to help create downforce, but the reason they're on basically every road car now is for drag. When you're driving along, assuming no wind, you're going through perfectly still air. Any air that is not perfectly still once you've passed has had work done to it. Diffusers allow higher speed, lower pressure air below the car slow down and regain pressure, getting closer to ambient conditions. In a perfect(impossible) world, it would leave the back end of your car with ambient pressure and zero velocity. That would equate to no drag. Blunt rear ends just dump all that air into the freestream without recovering any energy from it. 

 

 

2 hours ago, JMortensen said:

 Most of the time on a race car you're rule limited, or you have practical constraints that limit the shape, but in the more unlimited classes you see the upturned nose of a splitter or a raised tunnel entrance in the front. 

 

11 hours ago, clarkspeed said:

 

I just reread the flat floor section again last night.  What he recommends is a slightly rounded panel roughly starting at bottom of the front bumper to a smooth rounded transition knee point in front of front axle.  Then totally flat section back to where the diffuser starts.  And the diffuser being the minimum of another flat panel angled upward at 10 degrees or more to the rear bumper.  The start of the diffuser section seems to like being as far forward as possible.  With this arrangement, you want as much air as possible flowing under the car and no airdam.  So no extra drag and net downforce like a reverse wing. 

 

This is all basically trying to make the whole car a wing. Capture some air, force it to accelerate under the car, then allow it to slow down and exit the back of the car without separating. Look at F1 sidepods, for example. Using the whole car as a lifting body is how you're going to maximize downforce, but it's so hard to do if you don't have a bunch of time and money for R&D. 

 

 

12 hours ago, clarkspeed said:

It seems like what interference you do have with suspension arms and maybe wide wheels would screw this up. 

 

Yep, suspension and driveline is a huge pain in the ass for aerodynamics. FWD cars are easier because you don't have the diff and axles hanging down in the back. 

 

12 hours ago, clarkspeed said:

And the diffuser being the minimum of another flat panel angled upward at 10 degrees or more to the rear bumper.  The start of the diffuser section seems to like being as far forward as possible. 

The sooner you start your diffuser, the larger you can make it and the more you can get out of it. Being longer also means you can use a less aggressive angle and still get the height out of it you want. The angle is tough and dependent on your speeds. Too sharp and the air is just going to separate, rendering the diffuser useless, but too shallow and you aren't maximizing its effectiveness. Better than a flat plate is to have a smooth transition with a gradually increasing angle. It's really not that much harder to fabricate either.  

[calZ]

Something else that's just kind of assumed in all this is that your underbody is isolated from the outside air. If you do a great job of shaping the underside of the car but don't have it sealed off, you're just going to draw in air below the doors and it will ruin all the low pressure you worked hard for. F1 achieves this using strong vortices to "seal" the sides of the car, but sideskirts are a lot simpler. 

[clarkspeed]

I think I posted this on Hybrid before.  Here is a good video on how F1 cars use vortices to seal the sides and do lots of other things.

 

 

[JMortensen]

15 hours ago, calZ said:

The sooner you start your diffuser, the larger you can make it and the more you can get out of it. Being longer also means you can use a less aggressive angle and still get the height out of it you want. The angle is tough and dependent on your speeds. Too sharp and the air is just going to separate, rendering the diffuser useless, but too shallow and you aren't maximizing its effectiveness. Better than a flat plate is to have a smooth transition with a gradually increasing angle. It's really not that much harder to fabricate either.  

Here's something that really surprised me when I read it: diffusers actually work best when the air starts to separate:
https://www.racetechmag.com/2017/08/willem-toet-explains-motorsport-diffusers/

[calZ]

39 minutes ago, JMortensen said:

Here's something that really surprised me when I read it: diffusers actually work best when the air starts to separate:
https://www.racetechmag.com/2017/08/willem-toet-explains-motorsport-diffusers/

 

That was a good read. Thanks for the link. 

 

That makes sense to me. A small separation bubble means you're stressing things to the max. You see the same thing on the top side of a wing just before stall. Max performance, but you're right on the ragged edge. You stall as soon as the bubble starts expanding, like his pictures show. To me, a device being "separated" means it's completely blown up. I did a poor job of communicating that I think. 

Edited November 15, 2023 by calZ

[JMortensen]

I think tube80z shared that with me a few years ago. Made me regret that I had installed my fuel cell at a 7* angle because when I did it 7* was considered the max angle you could run without separation. He's talking 20* in there though. That's way outside of what I would have thought would work. I think when I do build one the focus will be on making it as long as possible and I may try to change the angle along the length. Been looking at AJ's gigantic one, and it appears that maximizing the angle in that particular space is the focus, rather than sticking to a particular angle. You can see it kinda steps up past the axle, then flattens out a bit under the gas tank, then steps up again at the end.
https://ajhartmanaero.com/modular-diffusor/

[calZ]

He also mentioned speed, but didn't show the data. The faster you go, the less angle you can get away with. You usually run autocross, correct? You'll be able to get away with more angle than a road racer.

 

You could always just make an adjustable flat plate diffuser with a flexible transition and then dial in the angle with tufts and a camera. Once you know where your limit is, you could build something more permanent. 

[JMortensen]

Autox but will be doing track days hopefully next summer. Track should have me doing 140 - 150 at the end of the straight, from what I hear.