Tesla Cybertruck Aerodynamics - Exclusive 3D scan reveals flow secrets through CFD simulation

[@AverageJoe928]

Amazing detail in both the model and your CFD analysis. Thank you for sharing!

[@AirShaper]

Thanks for the feedback, glad you liked it!

[@boostav]

Love the in-depth analysis, much appreciated. Worth noting the aero covers on the wheels are missing, that could explain part of the discrepancy - would be interesting to see how much of a difference they make.

[@AirShaper]

Ah very true, we missed those!

[@satoshimanabe2493]

@@AirShaper The manufacturer's drag coefficient could also be based on all-season tires. The vehicle was delivered with all-terrain tires, so I would think these were used in the scans. Great episode, looking forward to future videos!

[@w0ttheh3ll]

@@AirShaper for good reason, they chew up the tire sidewalls so nobody uses them. running a simulation with tire-eating covers would be very optimistic

[@war6193]

@@w0ttheh3ll the old ones or the new ones that are supposed to correct for this?

[@w0ttheh3ll]

@@war6193 ah, I didn't notice there were new ones. thanks for mentioning it

[@marquisdaily7715]

Thank you for sharing. Impressive job by Tesla for a truck that looks like a door wedge :)

[@w0ttheh3ll]

Thanks for calculating CdA and stating that's what really matters. Would've been nice to see a direct CdA comparison with some popular other trucks and cars.

[@wrighty338]

Just found Airshaper today after catching up with Robin Shute's channel and Im hooked! I have a build in progress il certainly drop into the app for analysis and see where improvements can be made.

[@AirShaper]

Ah fantastic, looking forward to see your design! You can connect directly with me at [e-mail] if you like, perhaps you can already drop me your contact details?

[@rpkruizinga9599]

Great video, very insightful and interesting! I learned something as a car designer!

[@AirShaper]

Thanks a lot!!

[@yagoa]

removing the mirrors would be very interesting as well! legal and easy in many situations

[@gavinbetteridge7524]

Excellent work, thanks.

[@O-cDxA]

This really proves what you were saying about the importance of accurate 3D models that are provided to AirShaper ! Regarding the design of the truck, I wonder why they simply didn't just add ( low poly ) beveled edges to the sharp edges. This would have kept the angular look, but also provide a more radiused edge for the airflow.

[@AirShaper]

Very true, small bevels could mimic a radius to a large extent, from an aero perspective, and reduce local flow separations. I think design aesthetics are still one of the most dominant drivers for a car's design, overruling aero in many cases.

[@kagander8619]

Great video ❤❤❤❤❤

[@AirShaper]

Thank you!!

[@TechnicalLee]

Do you have the data on the open vault cover vs. closed?

[@AirShaper]

We haven't run that comparison yet, would be interesting!

[@runem5429]

Coming here from the previous millenium and seeing *this* is possible I am amazed..

[@HolbrookAerospace]

Excellent video!

[@AirShaper]

Thanks a lot Austin!! Good to see you here :)

[@JB30G]

Thanks for the video Wouter! How did you take care of the exchanger's porosity? Did you use viscous and inertial coefficients from Tesla data?

[@AirShaper]

In this case we didn't model the radiator. But we can do so at AirShaper, using porous elements and 2 points on the pressure drop curve. Just drop me an email at [e-mail] if you have more questions!

[@635574]

I think teslas values didnt have rhe mirrors

[@TeslaElonSpaceXFan]

👍

[@Dpmcalexander]

Very nice! I am about to add the Cybertruck Co Molle rack to my Cybertruck, would be interested to know how it affects the airflow. I have heard a 10% reduction in efficiency, but im more curious to know if there is a wind deflector I could add to help. I have a theory that I could direct the air to flow over the top of the rack and across the surface of a rooftop tent that sits 8 inches higher than the peak of the roofline. Similar to the airflow of a Suburban. The setup is a "Cybertruck Co" Molle rack plus an "Intrepid Camp Gear Geo 3.0" tent.

[@O-cDxA]

This really proves what you were saying about providing accurate models to test in AirShaper ! The difference is quite large for even minor variations. Regarding this truck design, I have to wonder why it was that the sharp edges did not simply have a ( " low poly " ) bevel applied along the edges. I understand the desire to only have a boxy look, and this would have kept the angular look, while giving the air ( what it sees as ) a curved surface to flow around.

[@edbrowne4114]

Any suggestions about where vortex generators could be placed to improve laminar airflow on the CT?

[@AirShaper]

Vortex generators can really help to keep the flow attached on an inwardly curved surface, for example. But here, flow separation is often abruptly forced by a sharp edge. Perhaps adding some ahead of the A-pillar could help to already divert the flow, but it would compromise visibility out of the window, for example (if they work at all).

[@Tore_Lund]

cD of 0.33 is late 80' aerodynamics for that kind drag coefficient, when boxy cars begun getting rounded corners. Conclusion is this is not a vehicle for highway driving. The giant wake from the transition from the grille to the side, is not a feature to get the front wheel to create less turbulence, it is far worse as it creates turbulent air for the whole side of the vehicle. Then there is the impossible transition from top of the windshield to the roof. The simulation you made, is only the boundary layers, not the complete air column. It would not be pretty!

[@creativepan9708]

I’m surprised with how much air stays connected over that sharp angle on top

[@AirShaper]

Yeah that's indeed impressive! There are even aircraft with similar shapes 🙂

[@mhsohel47]

Wheel cover will improve it.

[@AirShaper]

That's true, even though the tires themselves already impact the flow a lot before it reaches the rims.

[@webspeedracer]

I agree it’s an amazing vehicle, a combo of my Model X and a Ford Raptor. Could you comment on a couple things I experience with my 3XXX VIN CT? - I get significant shaking in the drivers side mirror (but passenger side is rock solid) at ~80mph (+/-5mph) and also significant door shudder (eg, shakes water bottles in the door storage) with a ~45°NW frontal crosswind. Tesla Service has been unable to solve. Could it be that the windshield wiper induces a separation of airflow at that speed that causes oscillating air pressure on the door skin at those speeds, but which doesn’t affect the passenger side due to lack of wiper on that side? - or, another theory for the mirror shake/door shudder: my frunk is more aligned with the front edge of the drivers side quarter panel than the passenger side. I asked Tesla service to move the frunk forward to be slightly prominent of the leading edges of the front quarter panels, they said they actually prefer the 1/4 panel leading f edges to be prominent for airflow. Does that make any sense from an airflow perspective? I used the adjustable bumpers under the frunk lid to adjust forward/upward as far as I could (which is why the drivers side frunk aligns closer than the passenger side), but I wonder if the airflow around the side of the frunk is actually submerged rearward of the 1/4 panel leading edges, such that a bulb of air would form at the front corners and create an “invisible” longer nose and rounded corner of airflow? Not sure I’ve described it clearly enough, hope you understand the visual, but I can post pics of my CT frunk/panels if that helps - why, if the wiper dramatically improves airflow on the drivers side, did Tesla not include a second wiper on the right side (other than cost)? Is the aero improvement not that significant in terms of overall drag (i see the increased pressure on the passenger side roof glass is significantly more from the vortices, vs the drivers-wiper-side roof glass)? - I’d like to begin modeling a DIY cap for my truck to smooth airflow over the family mountain bikes and flow air more smoothly to the 19’ camper I tow on our adventures. Which modeling software would you recommend, that would easily upload models to Airshaper to sort out how best to minimize drag losses with the new cap (FWIW, I have limited CAD/Sketchup skills but reasonable aptitude for learning new compute skills)? Any AI-infused modelers worth the time, that would bolster my old-dog-new-tricks learning curve? Thanks for your work!

[@AirShaper]

Thanks for the very detailed comment :) The Cybertruck is a solid car, so vibrations on just one side would probably come from a local source of vibration, and not the whole vehicle shaking? Aero "flutter" etc could be a cause, but then it'd have to be pretty intense to make the car shake locally! But yes, the wind screen wiper could explain why it is more prominent on one side, as the vortex is quite different. If it were symmetric, you could take it off and install it on the other side. But for now, I'd just try to take it off all together and see if the problem goes away. If yes, then we know the cause! If no, then we have to examine the other theories :) As for the modelling, happy to discuss this in detail - can you perhaps drop me an email at [e-mail]? Thanks a lot!

[@Tnga]

There’s a development video this week where the engineers said the windshield wiper will move slightly at speed to improve the drag. There’s also the wheel covers as well.

[@replica1052]

would a lower narrow body cybertruck withhout flared wheel arches and smaller wheels have had better aerodynamics

[@AirShaper]

Lowering a car usually lowers the Cd (and frontal area, as you cover the wheels more) up until a certain point. Removing the wheel arches could also expose the tires more, so not certain that would always be a win though!

[@replica1052]

@@AirShaper (pull the wheels inwards )

[@AirShaper]

@@replica1052 Yes, that sounds like something which could reduce drag (if the impact on stability, interior space, ... is acceptable).

[@replica1052]

@@AirShaper for an entry level cybertruck with less performance

[@robhills2613]

Thank you, very interesting and informative. A respectable Cd for the type of vehicle - but a ridiculously large truck nonetheless, which wouldn't be safe for other road users and pedestrians in Europe 🇬🇧

[@jamesengland7461]

You Europeans seem really scared. I wonder if you remember that there are big vans and commercial trucks roaming your streets? Not scared of them? They can't evade or brake to avoid wrecks like the Cybertruck can.

[@madmotorcyclist]

Conclusion: It's still a brick.

[@war6193]

Cd of a brick is about 1.05. It's not even close.

[@madmotorcyclist]

@@war6193 I'll remember that, thanks!

[@StudentLoanJustice]

The aero is crap. The shape is crap, so the flow/drag will necessarily be crap, and no amount of CFD trickery will make that not so. If you had ACTUALLY designed a through duct from front to rear, perhaps with some turbine driven acceleration, then it would be far, far, better. But I'm not seeing that.

[@StudentLoanJustice]

Why don't you spend a tiny bit of $$ and build an actual wind tunnel? CFD is crap, and will probably always be crap for the next 50 years at least for real, turbulent,high Reynolds number flow. David Wilcox, author of the K-Omega turbulence model, readily acknowledges that prediction of reattachment points, etc for even 2 dimensional flow is absolute crap (ie with something like 30% error or more in either direction), regardless of which model you're using. Much less 3d. He once calculated that to accurately predict a real, turbulent, Reynold's number > 10 million flow, down to the Lagrangian scale, you would need ONE MOLE OF CRAY YMP2 computers!! Granted those were slow, but it's pretty safe to say that we'll never get close to that in our lifetimes. LOL. Ask him yourself. INVEST IN AN ACTUAL WIND TUNNEL. This is not hard. It will serve you well for years to come.

[@AirShaper]

Reality is that both CFD and wind tunnels have flaws and benefits. Relying on just one would be a very risky strategy. CFD is unbeatable to get insights when your design only exists on a screen. And the wind tunnel is essential to have accurate real world validation.

[@alancollinge9136]

@@AirShaper Myeahhh....I don't think so. I think CFD is so heinously inaccurate when predicting reattachment points in a complex 3-d flow as to be nearly worthless. But fine. If it makes you feel better to use cheap CFD for insights, more power to you. But using CFD as a substitute for REAL, ACTUAL validation is just bad.

[@AirShaper]

I wouldn't see it as a substitute - CFD requires validation. Here are some validation cases: www.airshaper.com/validation Never perfect, but certainly very close and useful in many cases.

[@alancollinge9136]

@@AirShaper OK. Where can I find the wind tunnel validation write-up for the Cybertruck?

[@AirShaper]

@@alancollinge9136 There is no public report available for the Cybertruck unfortunately. That's why we try to build as many validation cases on other cars, airplanes, drones, ... as you can see in the link. In terms of this specific process of 3D scan to CFD, we did the same for the Rivian R1T and Rivian themselves endorsed our analysis as well (see kzbin.info/www/bejne/gqm7dIKKaq53aJI).

[@dominicanwar363]

Legodynamics from a company run by a child.

[@AirShaper]

Feel free to back up your statements with actual information rather than random claims 🙂

[@constantinosschinas4503]

Why do they even bother to scan this trashcar at this level of detail and you make the cfd analysis, is beyond comprehension.

[@ZipZoomZip]

It's a lot more efficient than a regular truck shape. If you had ever used one to get real work done, to drive on rugged roads, and to haul heavy loads, you would know it's not a "trashcar" it's a phenomenal driving/hauling machine that is smooth and quiet and responds instantly to your every command. There is nothing else like it.

[@NO3V]

*beyond YOUR comprehension

[@LaJuanHughes]

So many people will be shocked when they get to drive one. People hate on what they don't understand . And if you haven't driven a CT you won't understand.​@@ZipZoomZip

[@LaJuanHughes]

​@@NO3Vyou need to drive one so you can understand.