I’m A French Aerospace Engineer And I’m Going To Teach You About Car Aerodynamics: Drag Coefficients

French Aero
ADVERTISEMENT

Bonjour fellow Autopians! I’m Manuel, your new French aerodynamics guy! I’m here to delight you with the physics that spin your wheels, and hopefully some sweet Gallic Domestic Market content in the near future.

First, a little intro: For years I’ve been following David and Jason’s shenanigans at the old site, so naturally when they asked for help with The Autopian, I volunteered my expertise. What expertise you ask? Well, I worked five years as an aerodynamics engineer in aircraft turbine design. After being done with turboshaft icing certification and air intake CFD analysis and the likes, I went and built robots that drill metal destined to take to the skies, because I wanted to find a job that’d let me turn wrenches between CAD modelling and FEA computations.

While that’s an OK-ish resume for a jack of all trades engineer, it doesn’t say much about my car cred, but rest assured, I’m one of you degenerates! My stable consists of a 19 year-old BMW Z4, a work-in-progress 280Z and a ’99 Renault Clio in cheapskate spec. I call this fleet setup “track, scrap and crap” and it keeps me busy in the garage. Between you and me, I blame David for this, as I didn’t give a damn about cars before reading about his Moab adventures and trench foot woes. It was a simpler time when I was only spending my hard-earned euros on flying lessons (hi Mercedes!) instead of going bankrupt trying to yank a Datsun from the grips of iron oxyde, or testing every brake pad on the face of Earth to keep fade from killing me before the end of the morning sessions.

But enough about me! This article was meant to look at how the good people of Hyundai managed to drag the Ioniq 6’s Cd down to the low low value of 0.22, but since I’m a sucker for a good equation, it ballooned into a bit of a science write-up. So buckle up, because today I’m gonna treat you to a free aero 101 lesson; a fairly basic but an important one!

Screen Shot 2023 03 21 At 9.37.04 Am

So, what’s a Cd in our music streaming age? If you’re perusing these pages, chances are you already know it measures the slipperyness of a shape; said shape probably has four wheels and makes glorious V8 sounds in your mind. The full name of that beast is “drag coefficient,” and we’ll dive deep into this seemingly simple concept. 

Drag Force

Drag (noted Fd) is the enemy. It is a force resisting motion and is proportional to the surrounding air’s kinetic energy, or more precisely its dynamic pressure (you may recall that total pressure is the static pressure in a certain space plus the dynamic pressure, which NASA defines as “a pressure term associated with the velocity…of the flow.” The static pressure is independent of motion). Drag is also proportional to our shape’s frontal area (the bigger the object, the bigger the force resisting its motion) and that shape’s ability to fend off the air rushing to meet it. That last element is defined by the Cd, which goes from 0 (but never reaches it, because life ain’t that good) up to about 1 (kinda) for the reeeeally blocky stuff and also downforce-oriented race cars. The resulting equation is the following:

Screen Shot 2023 03 21 At 10.36.43 Am

If that looks like a complicated mess, don’t worry, we’ll break all this gibberish down together.

1/2 Times What Now?

Every time we deal with air, things such as mass (noted m) stop making sense, so no one talks about the air’s kinetic energy (½.m.v²) as you’d do for a car because you can’t say that a car is driving through 100kg of air. The trick we aero people use in these cases is analyzing everything for a unit of volume (noted V), so our kinetic energy equation becomes (½.m/V).v². 

“m/V” defines the air mass for a given volume, aka density (noted ρ), so the equation becomes ½.ρ.v² and that’s the dynamic pressure I mentioned before. Fun fact! That means pressure is actually energy per unit of volume! Ain’t science amazing?!

Pitot Tiube 2
Credit: Murmakova/stock.adobe.com Another fun fact: What you’re looking at is a pitot tube that you’d find on an airplane. It’s a little tube that points in the direction of a plane’s flight path. It helps calculate a plane’s velocity using a pressure transducer to capturing the total pressure, which we know is the sum of the static and dynamic pressure. There’s another pressure transducer perpendicular to the flight path measuring the static/ambient pressure. Since we know that total pressure equals static pressure plus dynamic pressure, we can solve for the dynamic pressure. Then, since we know dynamic pressure is 1/2*density*v^2, we can come up with an equation for the velocity as a function of density  (which we know) and the total/static pressures (which we know). That’s how you find air speed.

A-Team

If you’ve made it this far, know that the hardest part is over! You can treat yourself to a well-deserved congratulatory glass of Saint Emilion and a Gauloise while day dreaming about the Bourguignon slowly simmering in the kitchen. 

As I was telling you, the drag force is proportional to the moving object’s frontal area. That means that for a given speed, if you were to make a ½ scale model of your favorite automobile, the drag force would be ¼ of the real thing since we divide both its width and height by 2. That’s rather intuitive but also means that you cannot compare the aerodynamic efficiency of a Matra Djet and a Facel Vega by only looking at the raw force they need to combat drag at a given speed, which brings us to the main topic of this article.

Cd, The Drag Coefficient

Up until now, everything we’ve been talking about are actual things you can see (frontal area), feel (force) or measure (dynamic pressure), but the Cd is a different beast. It’s a dimensionless number. If there’s one thing us air pushers love, it’s dimensionless numbers! We freakin’ love those! Reynolds, Prandtl, Mach, Nusselt, we can’t get enough! Since you aren’t here to become a button clicker in a big corporation R’n’D department, I won’t bore you with the details but know that the idea is always the same: when you fumble with equations and get a number with no dimension (i.e. no unit), it means it’s a ratio between two things that are comparable, like two forces for example. 

The concept can be declined ad infinitum; you can even make your own dimensionless numbers! Want to express how bored you are right now? Divide the fun you’re having reading this article by the fun you experience on a trackday while having sex and you can QUANTIFY how much you hate math! Isn’t that something you’ve been searching for your whole life?

But I digress, let’s go back to our drag equation and apply some aforementioned fumbling:

Screen Shot 2023 03 21 At 10.37.53 Am

This math trick is known as the “ol’ switcheroo.” Trust me on this, and please don’t look it up.

When you look at the equation like that, the Cd can be described as the ratio between the actual Drag Force measured on your object and the theoretical worst case scenario (essentially) you can achieve if you use the entire energy of the surrounding air to convert it in drag, and that’s why it’s a good measure of a shape’s aero efficiency. If you design a car with a Cd of 0.5 that means it’s TWICE better than a cube of the same size (technically a cube’s Cd is 1.05ish, but close enough)! Good job! If its Cd is 0.25 it’s 4 TIMES BETTER THAN A CUBE! Amazing!

How To Use All Of This When Designing Cars

Screen Shot 2023 03 21 At 11.58.26 Am

When singling out the terms of the drag equation, you quickly realize the aerodynamics engineer doesn’t have a lot to work with.

ρ: you can’t ask the car to only be driven on Mars where the atmosphere is thinner

v: you can ask people to slow down but where’s the fun in that?

A: making smaller cars isn’t exactly trendy right now

The only remaining lever we can pull to reduce aerodynamic drag, and therefore fuel consumption, is the Cd through the car’s shape, curves and aerodynamic devices. Doing so is done using well known rules of thumbs (teardrop all the things), past experience (round butts like on a 1st gen Audi TT are very very bad — abrupt tails to detach flow tend to be better), and a metric ton of CFD computations and optimization software. 

Making a car with a very good drag coefficient is actually rather easy. The 1939 Schlörwagen hit a Cd of 0.186, better than anything on the road today! However, just like anything in engineering, you’ve got to make compromises. Sure, there’s got to be a few fights with other departments (like the greedy rust loving cooling engineers wanting all the flow to keep the engine cool), but the constraints also come from the aerodynamics of the car itself. See, another aspect of vehicle aerodynamics is lift/downforce, which matters for planes and cars alike. But that’s a story for another time!

Top photo credit: Konstantinos Moraiti/stock.adobe.com

About the Author

View All My Posts

85 thoughts on “I’m A French Aerospace Engineer And I’m Going To Teach You About Car Aerodynamics: Drag Coefficients

  1. High CPM (chuckles per minute), good amount of nerding into relevant engineering details while staying relatable at the user level. This is top-level writing!

    Plus the cool euro theme!

    Love it.

    In addition to your aero coverage, also looking forward to some coverage of cars and culture across the pond 🙂

    1. Thanks man! We’ll see about that European car culture thing, I used to be in a great area for this (Toulouse) but now I’ve moved to the other side of France and it feels like the car culture is a little dead here. But all it takes is meeting a few cool people and that could change! Stay tuned 🙂

  2. Good stuff! I’ll be expecting an article about how the Cd is actually obtained via testing, both scale model and full-scale! Don’t be shy about doing some detailed calculations for scale model force/wind speed – I recall never being comfortable with that in school, didn’t feel intuitive.

    One more fun idea – perhaps do an experiment with an aero modification on one of your cars where you determine the change in Cd experimentally. We did that in school with a V-box type data logger and multiple coast-down runs in both directions. You have to make rolling resistance assumptions so the actual Cd is a guess but it’s OKish for comparison purposes if you make a big change and/or get lots of clean data.

    1. That’s in the works. I wanted to make an article about how wind tunnels work and how to use them to get the Cd but I realized that there’s a lot more theory I want to explain first.

      Basically, I’d like all of you guys to be able to do some aero design by yourself when I’m done!

  3. Welcome Manuel! Thanks for making this simple,even fun. My aero knowledge approximates that of a particularly smart sheep so every simplification helps.
    Ok i joke but not by much!
    Only last year,for the first time in my life i had the need to do actual aerodynamic pressure calculations – for a tiny house on wheels in case you’re wondering.Even there i cheated a bit.Online calculators FTW!

  4. I am very late to the party, but I have a question about the drag numbers on my personal car – a 1971 Alfa Romeo Spider. It seems that the Spider has a cd of .38.

    71 Alfa Spider
    Here’s the table from which I’m drawing my numbers. The Alfa is # 7 on the list; the MX3 is #511.

    I’ve never understood how such a small car with such an aerodynamic appearance has such a high cd. Even the Mazda Miata MX 3 (which is very similar overall) has a cd of .31. Further according to the table the Alfa has a smaller frontal area than the Madza.

    Is all that drag in the Spider caused by the windshield? Nothing sticks down below the bottom plane of the nose which is as sloped as the top. How can a aerodynamic looking small car like the Spider take more effort to move through the than other larger boxier cars?

    I can’t question the math, but the reasons are just no obvious to me.

    Signed:
    Confused

      1. I’d say drop tops have a bad Cd. The fact that you have an open top means you create bigass recirculations and a whole lot of aero crap behind the windshield.

        Even with a cloth top, the rake of the rear glass is less corved, creating detachement at the roof level instead of the trunk, which increases the area with low pressure that’ll suck the car back.

        Also, the car being small has no impact on the Cd, but it impacts the drag force through the frontal area.

  5. All this theoretical and mathematical stuff is fine, but how about some real world help for my classic Mini that feels like the hand of god is pushing it backwards when I encounter a headwind? Not to mention wind noise…..I know that one of the biggest noise generators on my car are the external body seams, but I’m told those do very little to slow the car down aerodynamically and shaving them off creates weakness in the structure that requires a lot of metal and welding to gain back….I know it’s a tiny car shaped like a brick, but surely there’s something trick little things I can do?

    One of my friends put a small, curved dive plane on the left front below the parking lite on his Mini and said it dropped his engine temps 10* by pulling more air out from under the wheel well (the radiator vents out the left wheel well on these cars)

    That’s the kind of stuff I’m looking for. That and whether a spoiler off the back would help disconnect the body from the drag by creating turbulence behind the car? And if so, how big a spoiler and where does it need to be? Right at the roof lip? Lower? Higher?

    How does someone with no FEA ability (software or otherwise) figure this stuff out without spending huge bux paying someone to analyze it for them?

    Oh….. and welcome!!!!

    1. Thanks for the welcome!

      So, first off, aero acoustics is it’s own weird beast and I’m clueless about it so I’m not going to give any advice that I’d be pulling from my bum!

      Regarding the spoiler, I doubt it’s necessary as the roof has a seam that should be enough to promote detachement.

      I believe you when you say your friend’s scoop dropped temps. The cooling setup on Minis is just horrible from the factory so I guess any thing helps.

      The only realistic mod I would advise is lowering the car. A spoiler will slow you down, you could add some fairing to the rear wheels which is a trusted method to limit drag but it’s an acquired taste.

      Maybe you could install a tray under the engine too?

  6. While taking aero engineering at UC Davis a loooong time ago, we were to calculate the max velocity of a vehicle. I chose my ’67 VW squareback (which I still own) as my subject. using realistic values for power and Cd, we calculated the max velocity. Turns out to be 84 mph, which I then proved through practical testing on my drive home that night. Fun times.

      1. either you use actual data P = M*A*V from a Drag, or use publish P for the vehicle and work the equation backward. wind direction, vehicle state of tune, and added wind cheating devices are hard to account for without testing and tuning.

    1. I’m fairly certain that if the knights of the round table gathered a sufficient budget on mechanical calculators they could come up with solutions!

  7. Welcome Manuel. It it awesome the autopian was able to land you in what I assume was a very frantic bidding war. You do not have to prove your bona-fides to me or any of the Autopian readers. I have been reading multiple websites for years that have car owners bragging about their Manuel Transmissions I can’t wait to read about the career you have had. LOL
    Seriously Welcome aboard I look forward to reading your columns that I barely understand. (@@) We are watching you.

  8. Welcome Manuel!

    How does length affect the Cd and if it does so positively what is the ratio where you start seeing diminishing returns?

    For instance at top speed is a TGV with a single passenger car in the middle less draggy (and if so how much?) than a pair of front-to-back TGV power cars alone? How about 2 passenger cars? 10? Infinite?

    1. Trains actually have horrible drag coefficients, and exactly because of length!

      For a given design, adding length with a straight section will increase the “characteristic length” of the object, which affects the Reynolds number, i.e. it increases turbulence and losses.

      More simply put, it gives more time to the boundary layer around the moving object to grow, which creates more shear forces slowing the train down.

      Now I’d need to make some research to be sure but I’m fairly sure it’s better to drive a single TGV with 10 cars that 2 TGV with 5 cars, because even if you fight more drag individually, adding up the energy of driving the 2 small trains is more wasteful than a single big one.

      1. Interesting. I’d have thought the opposite, that each train car was essentially drafting the one before it for an aerodynamic free ride. But I suppose that there are no free rides.

        FWIW my dad was among other things also an aerodynamics engineer in aircraft turbine design. One of his early projects was the bypass doors on the J58s used on the SR-71. He was on the B70 too, a few early cruise missiles (e.g. snark) and *something* nuclear powered with wings (but not Pluto).

        He also worked on the turbo pumps of various rocket engines including the Titan II and the space shuttle’s main engines. When US aerospace hit a slump in the 1970s he jumped to nuclear engineering. Oops! Then back to aerospace in the early 1990s just in time for that to slump yet again.

        I used to try to talk with him about stuff but his engineer speak was wildly different from my scientist lingo. Even when I was taking coursework in his forte of thermodynamics he was incomprehensible. It took me a while to realise our two disciplines’ approaches were really that different even on the very same material. I ran into the same problem a few times even within my own discipline (notably mechanism oriented physical organic chemistry vs. rote memorization synthetic organic chemistry). It’s worth keeping in mind as you write your articles that some of your readers may not learn the way you teach so please, be patient with us.

  9. Thank you for this. I’m no engineer, more of a duct tape and filler guy, however I can see we are just touching on the edge of so much here. I want to know more. I been confused by cars, trucks and motorcycles at various points in my life. I’m not looking for total enlightenment at this point, but just a few “holy shit that’s what was going on”, moments. (-;

  10. My issue with Cd is that it does not account for frontal area. For example say you have two cars that are both the same aerodynamically except one is 8ft wide and the other is 8000ft wide, both would have the same Cd, but the Cda is magnitudes higher for the 8000ft wide one than the 8ft wide one and the 8000ft wide one would have magnitudes higher drag than the 8ft wide one.

    For that reason I don’t care about Cd really at all. I care about performance numbers external dimensions, and internal dimensions.

    1. I see your point, which is valid from a consumer stand point, but as an aero designer, you need to be able to assess how good your design is, especially with regard to what we call “the state of the art”.

      If a Cd of 0.19 can be achieved for a car being produced in 2023, you should aim for that number, even if your car is twice smaller or twice bigger.

      The reason dimensionless numbers are so big in aero is because it’s hard to compare stuff, yet it’s the only way to capitalize on past experience. An entire sub set of aerodynamics is what’s called “scale model theory”. When you make a scale model, it’s impossible for it to be perfectly comparable with the 1:1 thing, because if you scale the model, and scale the wind speed accordingly, you do manage to keep the same Raynolds number (i.e. quantity of turbulence), but you mess up your Mach number (which defines how compressible a flow is).

      Finding a good criteria that allows comparison is fantastic from a design stand point, otherwise every aerodynamics engineer would be like “my work here is done, if you want progress, make a smaller car”.

      1. I can agree with that to a degree. I’m sure there are plenty of people who design new awesome airfoils but then said airfoil ends up on a bad airplane and there are plenty of inefficient airfoils on amazing planes.

        Generally speaking the most practical automobile has the worst aerodynamics possible as it’s a cab forward or COE layout granting amazing driver visibility, with a giant box or bed in the back and plenty of ground clearance. It holds the most stuff in the smallest amount of space with the smallest overhangs, plenty of ground clearance, has the longest wheelbase for its size, and it turns very tight.

        The most aerodynamic automobile has the driver fairly far back due to the aerodynamic design, with extremely slanted A pillars creating massive blindspots, almost no usable storage space due to the teardrop shape and the simple fact most humans use boxes for transporting stuff and a teardrop shape does not fit many boxes, due to the teardrop shape the rear wheels cannot be at the end of the vehicle so there is a massive overhang in the rear and a large one in the front as well, it has no ground clearance to minimize air going under the car which creates drag, and due to the shape, overall size, and design the turning circle is dramatically compromised.

        I first learned of the term Cda from the Ecomodder forums which are made up of people who add aero and increase the efficiency of existing automobiles to be more eco friendly/efficient. Time and time again automakers flout their low Cd and people have cars that are 20+ years older that get much better performance with much worse Cds. I know someone who is getting over 200hp and more than 70 MPG from a first gen Golf with mechanical fuel injection.

        Don’t get me wrong, I believe there is a place for aerodynamics but that’s without compromising practicality. For example you have all these automakers putting those stupid electric retracting door handles on their BEVs to improve the Cd ever so slightly and now you can’t get into your vehicle when the 12v battery is dead, which sucks if your portable jump starter is in the car. Not to mention they get stuck inside when snow freezes to the outside, and they work much less often in general than mechanical door handles. Subaru solved this issue years ago with the Subaru XT. They’re mechanical door handles but they’re flush with the door (look it up).

        On the other hand you got the Citroen 2CV which is an amazing vehicle in terms of ride, simplicity, durability, reliability, etc. but it has surprisingly bad aerodynamics in spite of how curvy it is, and because of that it cannot do modern highway speeds. The 2CV is a great example of a great vehicle that suffered a good deal mainly due to its bad aero.

        In the area of sailplanes (gliders) modern ones are so focused on minimizing drag they build the wing internal controls in one half of the wing then put the other half on top and seal it. When something breaks your only two options are to either buy whole new wings (EXTREMELY EXPENSIVE) or to take a hole saw to the wing, make the repair, and put a removable cover there. Would it not have been better to just put a repair panel there from the factory that is optimized for aerodynamics while not compromising the ability to maintain and repair the controls.

        Perfect is the enemy of good. For me I think one should make the most practical automobile then work on improving the aerodynamics while not compromising those practicalities.

        1. You’re inverting the direction of the aerodynamic teardrop – the fat end faces the wind, the point faces aft. Ergo, the best shape would be quite cab-forward, mid-engined, and have very little in the way of crash structure ahead of the occupants.

          1. Not really, rather that unless you make the whole front of the car out of a clear plastic you still have visibility issues and you’re still fairly far back for a cab forward/cabover design. I’d gladly drive around said teardrop, they just haven’t been mass produced. Sadly basically every production ready ultra aerodynamic car wasn’t or isn’t cab forward by any stretch of the imagination.

        2. If I get what you’re saying, your gripe seems to be more about gadgets/simplicity/repairability than aerodynamics.

          I’m a big proponent of being able to fix your car yourself, that’s why I don’t want anything more modern than my bimmer, but it has nothing to do with Cd vs CdA, nor how important aero is for fuel economy.

          Regarding your perfect automobile (which sounds more like a lifted Combi!), a cabover design is problematic regarding crash protection anyway, I think that’s the bigger issue there. But you have to keep in mind that everything is an engineering compromise. If you want to make the most practical automobile, you basically end up with a van, which is fine for workers but not exactly necessary for a single person. If you make the most aerodynamic one, you sacrifice usability, but also style. If you want to make the easiest to repair, you can’t pack the engine bay to the brim so you lose storage …

          There’s no such thing as the perfect car, it all depends on the use case. Maybe yours call for a lifted Combi but mine calls for a 2 seater drop top 🙂

          1. My gripe is with the whole ‘aerodynamics no matter what’ motto that many vehicle manufacturers adopt.

            Cabover can be an issue crash testing wise but with proper engineering (like that which went into the second gen Smart Car) you can get away with hardly any car in front of you.

            I wasn’t saying that the most practical car is the perfect car. My Perfect car doesn’t exist and likely won’t ever exist because I have somewhat odd preferences for vehicles and specifically when referring to automobiles I like a lot of specific things from various eras and I’d like all of them in one car. For example I love the VW Teijo, but I hate hydraulic brakes. Sadly VW never put cable brakes on the IRS VW transaxles but they did do so for the swing axle VW transaxles, but since the Teijo is FWD swing axles are not useable on the front so short of a completely custom brake setup (which is possible albeit not cheap or practical). Also even if you could add a driven axle to the rear to give it 4WD you’d have no way to send power there from the engine as the transaxle is in front of the engine but I digress.

            I agree that everything is a compromise. There are no solutions, only tradeoffs.

            I agree the “perfect car” does not exist. The point I was making that short of performance figures the most practical automobile is antithetical to the most aerodynamic vehicle. The question is what is one willing to compromise in the name of aerodynamics?

  11. Yes, more of this! Intro to aerodynamics was one of the most eye-opening classes I took in college. The first lesson that I took to heart was that all cars are essentially blunt bodies, aerodynamically speaking – they just aren’t long enough to get the aspect ratio down to something you can actually streamline. That means that making the overall shape swoopy (which was the trend back in the 90’s) doesn’t really do much for you. Drag reduction is all in the details, eliminating turbulence. There was a figure in one of my textbooks that showed some very boxy early 80’s car (might have been a Golf) before the aerodynamics team got a hold of it, directly from the styling department, and the Cd it had then (0.45 or so, if I recall) and the after picture, with the new Cd (something like 0.32) – it was hard to tell the difference! Little tweaks to manage flow around the tires, under the body, changing some lip radii, etc. was all it took. Neat stuff, and increasingly important in the age of electrification.

    1. Oh I WILL give you more of this! Aerodynamics is awesome, but it’s easy to be put off by the math. However, in my experience you basically stop using complicated equations altogether once you start working and do what we call “aerodynamics with your hands”

      I have never worked on a car’s aero, but I can tell you that designing a good air intake is all about making very very subtle change to delicate curves which can have massive consequences.

    2. On this subject, you can streamline a box into something with roughly 1/3 the drag of a box. Consider the Cd value of the 1st generation Scion xB is 0.35. I think this particular car is an excellent illustration of this.

    3. I’d like to sneak in some thermodynamics too at some point, but as a turbine specialist I’d have to make some extensive digging before that can happen.

      1. Which area of thermodynamics? That’s one of the few I probably can help with, along with CFD. Amazing how you need that kind of thing when you’re cramming over 4kW into 4U. Which also lends to sixteen hour long dissertations on properly engineering systems for performing continuous batched or parallel FEA using LS-DYNA.

        Extremely related; if you’ve ever had FEA or CFD jobs execute on a system with a large bird silhouette on the front panel or a cabinet with a bright red and cyan “DANGER: DO NOT OPEN FRONT DOOR DURING OPERATION” label? I built those. (And seriously, don’t open the front door. It’s integral to the noise baffling and those are over 85-95dBA under load.)

        1. I’ve been doing some turbojet pre-design in my younger days. It was mostly excel stuff but it still blew my mind that you could predict how efficient a design was based on a few formulas and macros.

          I’ve never actually seen the cluster I was running on, so I couldn’t tell you how they look. And I was mostly a Fluent guy with a touch of elsA!

          1. Oh yeah, now extend that a lot, lot, lot further. Like say you want to crash test a car 50 times without building a single part. Or you need to calculate the precise pump rate for a given restriction setup at a specific temperature. 😉 Pretty diverse mix of Ansys, COMSOL, and Numeca. (And obviously LTSC pre-Ansys buyout.) I got seriously into it back when I found out Shelby was using LS-DYNA to crash test the Series 1 without destroying cars.

            Probably safe to say you weren’t on one of mine, sadly. Never known anyone to not show them off to employees for cost bragging if not performance, and you’d definitely know if you were on a Kestrel (since it says that on the fiber receiver and I think there’s only 4 or 5 still in operation.) But my claim to fame remains that I know FEA/CFD – nobody can or will compete on CPU clock. And I’ve got an over 500MHz per-core advantage.
            Also why my systems are exceedingly rare. I don’t do lease, and that 500MHz advantage ain’t cheap. (Nevermind when you start adding SG2+ FPGAs or GPUs.)

            1. Oh yeah, I had courses on crash simulation with Radioss, that was awesome!

              Regarding the cluster, I was working in a big conglomerate, chock-full of FEA and CFD engineers. Managers didn’t take the time with each of us to brag about the cluster!

            2. Are you writing in assembly? To get that large an advantage, I would imagine you are writing in something pretty close to the bare metal.
              I wish my career had taken me in that direction.

  12. Excellent introductory article. This area is where cars have the most room for improvement regarding efficiency, especially at highway speeds. Cd values of around 0.15 are possible for practical midsized sedans while retaining enough room for stylists to distinguish the vehicles from each other. Experimental vehicles that are road worthy have gone much lower, with solar cars from universities getting near 0.10. A Milan SL velomobile I own has a Cd value of 0.08, and I’d love to scale that shape up into a 2-seater electric sports car.

    A could definitely be shrunk. Everything has gotten too big. My 1969 Triumph GT6 has a frontal area of roughly 1.4 m^2. Most new cars are roughly double that. Consider that roomy sedans like the W123 and W126 Mercedes models are around 2.2-2.3 m^2, which is not an unreasonable size for a car, and couple that with a Cd in the mid 0.1X range, and it is conceivable that you could have 80+ mpg highway midsized sedans and sub 150 Wh/mile EVs without really giving up anything.

    At 0:49-0:51 in the video above regarding the Schlörwagen, you can see tufts of yarn taped along the body, in roughly equidistant spaces from each other. This is being done so that visual observation can confirm attached flow on that section of the car. Turbulence induced is an indicator that there is a pressure differential somewhere on the car, which in turn means there is a pressure vacuum induced, adding drag.

    The average new car today has a Cd value that is equivalent to the 1921 Rumpler Tropfenwagen. This value is 0.27-0.28 depending upon the article referenced. Consider that the Rumpler has the penalty of outboard wheels, which can easily double the drag coefficient value versus the wheels being flush with the body due to all of the induced vortices and flow separation all along the body aft of the wheels. Many modern vehicles have their wheels flush with the body or are even relying uoon techniques like induced air curtains, and are getting the same Cd figure as the Tropfenwagen, and often worse. Massive oversized grilles and large wheels are really bad for aero drag, among other things.

    1. While I agree that improvement is technically possible, the real challenge is getting people into really eco-friendly vehicles.

      In school we talked about the Mathis VEL 333, a car with 3 seats, 3 wheels that had a fuel consumption of 3l/100 (78mpg) in the 1940s! Yet the company didn’t find success despite how impressive the car was. Economy is but one of the parameters consumers look at, otherwise nobody would drive a V8 or an SUV today.

      The biggest hiccup with your theory on aero is the “without giving up anything”. While I believe lower drag coefficients and frontal areas are possible, I doubt it comes with no compromise. Engineering is all about compromise, and surprising stuff can come up and mess your plans (like style, practicality, certification and whatnot).

      1. Probably more difficult as illustrated by the Ionic 6’s polarizing design comments in other articles, is getting people to accept the designs that work aerodynamically.

        Simplified, the more something looks like a drop of water falling though atmosphere, the better. But the parts people miss is the things that spoil this. wheel opening are terrible for aerodynamics, undercarriages without Belly pans are are also affected, Massive open grills affect things. Spoilers are called spoilers for a reason, they in effect spoil the airfoil nature of a 4 wheeled vehicle shaped like half a waterdrop. this is necessary of course to avoid aerodynamic lift, but still.

        The size of the engine is somewhat moot if the drag is brought down enough. Most V8’s these days with decent Overdrives and aero tricks are barely idling at freeway speeds. So the Aero stuff just makes that better without sacrificing the low end grunt that many still desire.

      2. If you want to get people into eco-friendly vehicles, they must pose a value proposition that the competition does not.

        Consider why people drive V8s. They want something with some ass hauling capability.

        A $25,000 car that can accelerate, corner, and top out like a $250,000 one would be an excellent start. EV technology coupled with an extreme focus on size reduction and drag reduction makes this possible. Increasing the horsepower of an EV drive system 10-fold would not increase its cost proportionally, but by only a small fraction. The motor, inverter, and battery tech to allow 500 peak kW versus 50 peak kW has about a 20% difference in cost when produced in high volume, most of that difference being raw material costs. In the case of a battery, cost per kWh is a much more significant cost regardless of battery type, than cost per peak kW which heavily varies depending upon battery type. Most of the costs of motors/inverters are NRE costs and must see high volume production to get costs down, regardless of whether they are 50 kW or 500 kW. Building such a vehicle and selling it to the public would likely cannibalize the sale of much more expensive and high-margin products, which is probably why we will never see such a thing from the mainstream automakers(note how GM killed the Fiero when it threatened to out-perform its Corvette at a cheaper cost, or how Porsche never lets the less expensive Cayman exceed the performance its most expensive 911 in spite of the Cayman having more potential as a platform), or alternatively, the company will decide to mark it up passed affordability to pad margins and/or limit its sales numbers to prevent it from displacing the sale of more expensive vehicles.

        What I mean by “not giving up anything” is that an ultra aero car with good efficiency doesn’t necessarily need to be a cramped penalty box with no guts. A large bodied sedan made of inexpensive and conventional materials, with plenty of storage space and passenger room, with good viability for the occupants, good NHV, good performance, et. is not rendered impossible due to aero. However, some corporation emphasizing its brand identity with the corporate styling dujour WILL have to be given up, but that’s not really giving up anything of substance or impacting what the car can do or how it can be used, is it?

        The dilemma isn’t physics, or even practical constraints. It’s money.

        1. I am drawn to modern NA V8’s over Turbo 4’s and 6’s with similar HP and torque simply because I am not sold on longevity of the complicated additional items such as the turbo/wastegate systems. The EGR systems are also in question a bit at this point as it seems like every turbo benefits from a catch can, yet nobody installs one from the factory. My major Caveat is the Fuel and Timing management systems. some brand seem to be much better than others in this regard, so this effect V8 choices as much as anything as well.

          1. I prefer NA V8s as well. Less disposable engines than something more high-strung.

            The platform the vehicle is built on, and the resultant mass and CdA values as well as tire choice and gearing, combined can have a much larger impact on a vehicle’s economy than the engine displacement or cylinder count. Sure, pumping losses and friction losses increase with bigger, higher-cylinder count engines, but that can be more than mitigated with a greatly lighter, more streamlined design, provided the engine has features such as VVT to minimize the pumping losses.

            There really is no excuse why we can’t have a V8-equipped small car that gets 40+ mpg highway, other than the auto industry refusing to build such a thing.

            I like inline-6s even more than V8s. REALLY love the Barra 6 from Australia. That would go hard in a Triumph GT6 or Datsun 240Z. At least we can get the Cummins turbodiesel in the U.S., but unfortunately, none of the vehicles that come with it were designed with efficiency in mind. Were money no object, I’d love to build a mid-engined AWD streamliner of roughly 2,700 lbs around a Cummins tuned to 1,000+ horsepower. The Cummins would be almost half the vehicle’s weight, but make the vehicle slippery enough, and fuel economy could be phenomenal on the highway, and as long as you kept your foot out of it, and would be comparable to a modern 4-cylinder gasoline sedan in the city. Step on it, and there would be massive hoonage coupled with clouds of glorious soot! And think of the top speed possible with such a contraption…

            1. Wait,
              Isn’t the super slippery aero benefit only help when cars are uh,
              actually moving?

              Most Americans drive in cities we know. And I read somewhere years ago that the AVERAGE speed a car goes over its lifetime is something like 21 mph.?!?

              Not everyone can drive 75+mph all day long to gain the aero benefit.

              Don’t know how the aero helps big comfy engines in that scenario when most of the time cars are idling or moving slowly.

              I know. That’s why everyone is looking at other ways to cut energy consumption when cars just sit or crawl along. Aero is only one part of it.

    2. Also, another area where cars could improve is mass. You will always need to use brakes in a car, and that means dissipating kinetic energy that scales with mass.

      I think the trend where we see cars getting heavier and heavier (for a given size) is mainly driven by safety standards, which require more metal to protect us bags of meat. Again, it’s a compromise!

      1. I tend to see mass follow HP. to hit the warranty period the more powerful the car the beefier the components seem to be. And of course the tech to make them drivable by dolts.

        EV’s of course have similar issues as they are generally pretty stinking fast in many cases, but they also have to lug around massive batteries.

        I think the Hyperion Supercar might be a good thing for many to look hard and consider as an alternate in the future.

        1. Mass begets more mass. Build a small chassis with low frontal area, and its mass requirements for reliability will be greatly less with a more powerful engine than a larger car. AND the smaller, less massive vehicle won’t need as much power for a given performance anyhow.

          It’s a sort of cascading effect, one which Amory Lovins of the Rocky Mountain Institute noted when he proposed various hypercar concepts(not in the sense as we know hypercars today, but as ultra-fuel efficient vehicles that met or exceeded the performance expectations of the cars typical of the 80s and 90s when he made the proposal).

          Note the GM Ultralite. It is a 4-seater compact sedan that can do 0-60 mph in under 8 seconds and top out at 135 mph, with all of 111 horsepower and a “fierce”, 1.5L 3-cylinder engine. This is because nit had a drag coefficient of only 0.19 and weighed 1,400 lbs. Fuel economy was 88 mpg.

          If one were to put in a more powerful engine, not all of the car’s mechanicals would need to more robust, just the ones that will see stress from the increased output. Most of the stress they will see is the result of the car’s mass coupled with its velocity(when you hit a pothole at speed, both the mass of the vehicle and the speed you hit the pothole determine the forces that can damage the components, but when you don’t hit anything and simply accelerate hard, you only need to worry about the durability of the moving parts which also see more stress with more mass). Thus, you could double or even quadruple the horsepower, but you might only need to add 20% to the car’s mass to do so without compromising its reliability. The more mass the vehicle has, the more mass that needs to be added in order to add more power without compromising the reliability.

          Reducing or increasing the mass of a vehicle has a cascading effect on the mass of most of the other parts of the vehicle.

          1. except the ultralight would suffer greatly on our modern potholed roads. the parts being lightened to that extreme almost always have long term durability issues

    3. Engineer curmudgeon here, but I get angry when people talk about drag coefficients. And dont get me started on how bad the basic idea of dimensionless numbers is (yes I’ve been fighting this battle for years, yes I know I’m not on the majority side here).

      It is not relevant even amongst vehicles of the same “class” since class “size” is not very comparable. There is so much gamesmanship that can be done on fudging your “frontal area” that cd by itself is nothing but a marketing gimmick. Tell me your product drag area (A*Cd) or get out of my way. As I’m sure the author knows, it is quite plausible to design against a high Cd instead and a more compact area, and by converse, make crappy designs look good in Cd by making it spread out (and worse).

      1. To me saying “our Cd is awesome” just means the aero department is good.

        If you worry about the A.Cd, it’s because you care about fuel consumption, and manufacturers provide these numbers.

        I really fail to see your gripe here.

        1. Nice! That reminds me, I think my father in law has a wedge shaped micro car rotting in a barn, maybe I should enquire about it someday.

              1. Go, man: go! Looks like you have a firm handle on the ferrous mites there-any thought of playing with its cd when done? Or, is it going full-stock when you’re done with the metal work?

                1. I’m not doing the metal work myself as I lack both the equipment and the skills to do so, but I’m taking care of the mechanical aspect of the resto!

                  It’s going to look bone stock except maybe for shorter springs, but I’ll do some massaging to the thermodynamics of it 😉

                  1. If you turned that Z car into a streamliner, the potential exists to double its highway fuel economy. With taller gearing, its top speed could go up greatly on stock horsepower.

                    1. I’m not gonna streamline it but I’ll improve fuel economy with a gearbox swap and higher compression ratio in the spare engine I’ll building up.

                      I have thought about adding un undertray though, that’d be a cool enhancement but I’ll about that when the car is back in running order

    1. Glad you liked it! I too like seeing what’s outside of my field. If I was more courageous I’d try and learn about electronics but seeing as EV swaps are basically illegal in France I have little to no incentive to do so.

      1. What kind of certifications and inspections do you need to do an EV swap in France? EValbum.com does show examples of converted EVs n your country, but I imagine there are a number of significant barriers imposed, which require money to overcome.

        I’m glad to live where I do in this regard.

        I’m an electrical engineer, and EVs are what motivated me to pursue this field.

        1. You need to be a certified shop to be allowed to install certified retrofit kits that are constrained by boring stuff such as “don’t increase” horse power nor mass. It’s not exactly diyer friendly.

          1. How would the government even enforce that rule regarding horsepower? Electric motor horsepower isn’t rated the same way as ICE horsepower. A 20 horsepower continuous electric motor could have a “20 horsepower” sticker on it, possibly pass scrutiny, and then be capable of over 200 horsepower peak. Then there’s the possibility of the electric motor matching the ICE’s output, but making ridiculous amounts of torque a the low end, and then with a transmission delete, still making more power at the wheels than the ICE did and readily outperforming it, even though the motor/controller/battery combo makes the same brake horsepower as the ICE it replaced.

            Do they want the car dyno tested? Does it have to go through a massive bureaucracy to certify that the EV horsepower doesn’t exceed the gasoline ICE horsepower?

            Government regulations are silly.

            1. The kit goes through a massive bureaucratic mess before being certified, and is only valid for a given vehicle.

              It’s a shame but it’s in keeping with the overall regulatory framework regarding car modification laws which can be summed up as “don’t mod your car”

Leave a Reply