As an RC aircraft modeller of 40 years, can I suggest you work more around wing profiles for downforce rather than angle of attack, high alpha aerodynamics , ie stalling the air over the airfoil are super draggy. That symmetrical airfoil you used with a rearward center of pressure is a poor one for downforce as you get too much separation, and separation is your enemy. I know this is just for testing, but if you test with a basic inverted Clark Y airfoil you would get more reliable and testable results as it is a forgiving low speed airfoil.
I agree, especially about the high angles of attack leading to flow separation/stall. However, I think an uncambered airfoil is better than a Clark Y or NACA airfoil for the proof-of-concept he has now. Separation is not necessarily your enemy for automotive purposes rather than aerospace. Active aero vehicles nowadays use it as an airbrake at high speed, since flow separation creates excellent drag. You definitely don't want this for a fixed wing of course but his video title indicates otherwise.
@@samfedorka5629 Clark Y is uncambered, bit yeah, I picked it out of my butt as it was the simplist with the most data available airfoil I could think of and from experience I know it works well on tiny RC airplane models, admittedly, the other way up. , I am sure there are way better, but anything is better than that truely aweful airfoil he was using. But he is running tests at such low speeds and scale i doubt much data he is getting will be useful in the slightest. You have to go up to 1/4 to 1/2 scale before air flows similar to the full scale vehicle. Certainly no use with an 1/18 scale BBurago toy.
@@ForeverNeverwhere1 Angle of Attack (AoA) is everything. If you want to live with a wide range of AoA, then you want something with a blunt-round nosed airfoil, like a Clark-Y. If you want low drag at high speed, you want a shallow AoA range, and a thin, almost flat, airfoil. Look at the leading edge of the F-16, in a sharp turn, and watch the Leading Edge Flaps hold a narrow AoA. An F-5 flys well with a flat wing, iff, the AoA is held within the limit of the critical angle.
@@DumbledoreMcCracken You're comparing airfoils designed for supersonic flight (F-16, F-5) to those designed for low speed. There's a big difference. Most cars seems to use thin but heavily cambered airfoils for downforce compared to the thicker, less cambered ones used on airplanes. Possibly this is because they are supported on both ends (or very short) on a car as opposed to cantilevered on aircraft? The speed range they are used in for car aero is generally lower too, corner speed in a car is lower than what most airplanes experience in cruise. If anything wingson cars most closely approximate an airliner wing in landing configuration.
A few thoughts from an aerospace engineer: -You want your flow straightener (straws) at the widest point of your intake nozzle and then a nice smooth contraction down to your test section. Both the straws and the contraction work to suppress turbulence in the intake. A contraction ratio of 10:1 (area-wise) or so would help a lot. -Your smoke rake should be airfoil-shaped to remove the turbulence generated over the back of the down-tube. Then the rearward facing tubes should be as small as possible to minimize their impact to the flow. Some 1/32" brass tubing from K&S would work great and could be pressed or glued into a 3-D printed part. -Optionally, you want some expansion (a diffuser) behind the test section. If you used 9 fans arranged in a square but kept the same test section size, that would be a good starting point. The diffuser should have a half-angle of no more than about 3-5 degrees. Can't wait to see what you come up with!
The contraction section is to accelerate the flow rather than smooth it out. Before creating a higher velocity wind tunnel, I'd rather assess repeatability of the set-up, which will probably be very, very poor. In order to improve it, it would be wise to remove the smoke section, which in real life is not used anyway, and also the test subject should be much smaller in order to prevent blockage and wall effects.
@@Agussg23 A larger contraction would only produce a higher velocity if you had the fans on the inlet and not in the diffuser. With them on the diffuser, all it does from a velocity standpoint is lower the velocity at the inlet.
Awesome video! One thing: "equal transit" explanation for lift is a very common misconception. No idea why this keeps getting taught to folks! Lift is mostly just Newtonian, that is momentum transfer, so with an angle of attack, planes can even fly upsidedown. The air does move faster over the top of the wing, but not due to some constraint on the time it must take to get there. It clings to the curved surface falling away from it due to the coanda effect. This serves to improve the efficiency of the airfoil mostly, but does not provide a significant amount of lift.
Yes. The equal transit theory is very flawed. Doesn't take much to prove that there is no equal transit. Basically, air takes longer to get to the trailing edge when it follows the upper surface rather than the lower one, which is intuitive.
@@youngtschakaloff That's true, at least it usually is, but the air _does_ move substantially faster over the top surface than the bottom. Not because of some magical "equal transit" arrangement of course (I love the idea of the air molecules slowing down to wait for their friends) but because the air above the wing is compressed into a narrower stream thanks to the air above _that_ not getting out of the way. Narrower stream means faster movement which means lower static pressure which means lift.
@@cooperised okay, reasonable. That way of explaining is a million times better indeed. I always keep in mind that there are multiple phenomena at work at the same time, and some of them are really hard to fully explain
@@youngtschakaloff Absolutely! You're thinking in exactly the right way. Multiple phenomena are at work at the same time, and are often the same phenomenon viewed from different perspectives. For example with a wing, it also deflects air downwards as it generates lift. I've seen people say "so it's the deflection of the air downwards that generates lift not the pressure difference", but that's not right. If the wing is generating lift then the pressure below _must_ be higher than the pressure above, that's what generates the force. But Newton's third law also applies: that upward force on the wing produces an equal downward force on the air. Or if you like, pushing the air downwards generates an upward force on the wing. Both things (pressure difference and Newton's third law) are true, and neither is "the explanation" for where lift comes from. They both just describe phenomena that can be observed around a wing.
6:58 - "All physics taught below Junior year college is a lie; and it's taught that way because it's funny" -Alice Caldwell-Kelly (Well There's Your Problem Podcast)
@@chrispollard6568 You missed the point. Being "within 3%" is by definition wrong, however, it is an adequate approximation of reality for most engineering purposes. Similarly, as explained in the link I gave, Newton's theory of gravitation (high school physics) is completely wrong because it goes against Einstein's theory of relativity, but again, Newtonian physics is far simpler and is more than adequate for the vast majority of purposes.
@@squiresuzuki It isn't "Wrong" because it works. My "high school" physics works to give people like you working products that you actually use. Your computer disk drive WORKS.
Another aero engineer here. First off, great work so far and im excited to see whats next. I getting into arduino myself now, so ill be very keen to follow your project. A couple of points I wanna mention: 1)be careful about the blockage ratio, thats the ratio of the objects frontal area vs the area of the wind tunnel's cross section. Most I've see were around 10% and beyond that the wind tunnel's walls might start influencing your results. I realise it might become cost prohibitive though. 2)try adding a mesh of sorts at the start of your wind tunnel which should help to make the incoming flow less turbulent and thus lead to more accurate results 3)if possible make the airfoil holder from metal or design it to be as stiff as you can, so that there's no force loss in the flex. I work on racecar aero for a living so if you need any advice or cfd simulation help, feel free to shoot me an email.
The flex is an interesting point- if you notice the metal beam across the bottom was from results of flex on the bottom wall. I also want to calibrate it with and without the massive servo
@@WesleyKagan the way its usually done is you run your setup without the test subject first to get the "error" data and then you run it with your wing. At the end you subtract the first from the latter to get your real values. Id also suggest on getting a pitot tube in there to get the velocity of the flow (place it in front of your wing). That way you can get coeff of lift (C_L) or (C_L*Area) and the Reynolds number which should help you validate your data against other sources. Reynolds number is very important, flow will act differently at different Re numbers. Lower Re flows will tend to separate sooner, which will lead to lower peak C_L.
you should put your test subject on a sled to measure drag force. From there you can calculate downforce-efficiency. You can also measure air speed with a pitot tube
Ever since the new Alfa Romeo Giulia came out with the active splitter at the front ive been wanting to replicate it, so the next video is right up my alley
I love that this dude got Noctua fans for a wind tunnel test. If you don't know anything about PC building and cooling, Noctua fans are regarded as being the top of the line fan for their smoothness cooling properties and silence. I just think its hilarious he went all out for the wind tunnel build, but kudos to him, this dudes a genius with what he is doing and has done including incorporating free valve tech into a predator engine lol.
Please upload more🙏🏼 you’re a bona fide genius, like Mensa level! Okay okay enough of that, your channel is kinda awesome though. Thank you for your entertaining projects.
7:00 equal transit theory is an incorrect assumption made pretty often. In reality, the flow over the top of a cambered airfoil or symmetric airfoil at an angle of attack is much faster due to the decrease in pressure, reaching the trailing edge quicker than the flow over the bottom of the airfoil. There are a lot of subtleties that I am not sure of myself. Cool project, looking forward to your next upload.
Hey man, if you have neighbors with kids 10 through 20 you should really ask the parents if they would like their kids to come and join you in experimenting with this kind of stuff. Kinda like Casey Putsch's genius garage. Growing up I didn't have anyone around me doing stuff like this, but I always stayed late in chemistry and mechanics class to learn more. You could be a mentor and teacher for an aspiring racing engineer or something! I absolutely love this stuff, really nice work.
Holy crap, just as I start designing an active aero setup for my build... You drop this video series, how insane is this? I look forward to seeing what you come up with! You might see a "duplicate" on my car at some point if it will fit!
13:41 this is exactly what I’m trying to do too hopefully you keep up the effort, the video is 2 years old I’ll still be tuning in to see updates great job
Good stuff. You will run into the limitations of a miniature tunnel so I would say your best bet is your full size algorithmic approach otherwise you will end up having to do that anyway because of correlation issues. I built something similar many years ago with a broken vacuum cleaner which I then ended up starting, but not finishing, building a head porting flow bench. You will get better laminar flow if you have a vacuum chamber pulling though straws as well as a straw entry plenum away from the front of the pitots. Good luck
Mike Patey, who you can watch on youtube, built a wing with flaps and slats that may have interesting information to add to the conversation. It is a different level of complexity but airfoil design and engineering is getting some pretty cool visibility recently. Thank you for sharing this with the world!
Wesley, I thought of a great little input you could run for the active aero project. Snag a steering angle sensor (whether cable or touchless or otherwise), and anything UNDER 25* steering angle you drop it down as long as Brake Pedal is not enabled. That way, it'll work when you're going straight and we only have a total of 100% Grip at a tire, so once we have over a certain steering angle, and not on the brakes, we would likely be accelerating. I think it's a real simple sensor configuration and a great start on a budget given your Arduino capabilities. Good luck on the journey!
could potentially even be a linear relationship between steering angle and wing angle, or some other simple function (log, root, square,....?) depending on what fits
A soldering iron and some light machine oil on a rag creates great smoke for PCV system leak tester, too. Can use Johnstone's baby oil instead. I suggest the use of a 3d printer hotend heater in a glass jar so you can set exact temp and even a mini hotend fan to slowly blow the smoke though. Also recommend using blunted syringe needles (from a printer ink refilling shop).
Air ejector nozzle. Dealt with something similar to what you made using the venturi in one of my engineering internships. High pressure steam is run through an ejector nozzle and create a vacuum behind it in order to draw NCGs out of steam turbine condenser.
Always excited to see your videos! If I may offer one small but of advice, you can acquire a set of Bluetooth quarter scales used for rc drag racing and measure balance and changes made to the model in the tunnel in real time.
I think plugging the holes in the smoke distributor except one would help visualize the airflow better. Also, since hot air is less dense than cold air, maybe you can play with temperature to help get a more accurate set of data points. Very cool project. Keep going.
Just like you said, without watching the video I knew it was glikf to be awesome so liked and commented already. Now onto enjoying your enginuity and scientific approach.
It seems like my comment has vanished, but anyways. Wind tunnels and aerodynamics, awesome and can't wait for part 2. Watching this I was immediately reminded of a set of videos made by Yoshi Suzuka probably 10 or so years ago and the discussion around them. Back then he was working on, I think it was something like an office small scale wind tunnel. Really interesting and worth checking out for reference on how far it's gone since then. Also, at the time he was doing it with 1/24 I think scale models and had it down to getting repeatable results and trends, though not necessarily representative or completely accurate to full scale numbers. I remember that wind tunnel at the time he could measure lift by the corners of the car and drag. IIRC Suzuka did the aerodynamic development for the Nissan GTP cars for a large portion on a 1/7 or 1/8 scale wind tunnel and models. I vaguely remember Suzuka may have said that with the 1/7 (or 1/8) scale wind tunnel, he was able get 2% drag and 4% lift correlation to full scale. Anyhow, you're definitely on the right track to making something that is close enough. Oh yeah, there's a whole bunch of test data (or at least was time I checked...several years ago) you've already found for various NACA profiles and even "race car stuff" wings where it should be possible to try and repeat the test with the same profile and see how the results correlate.
Some suggestions if you're looking for maximum results: As mentioned, for better similitude between your model and reality, you need to try to match the Reynolds number. For a smaller-than-reality model, you can do this by increasing air velocity and increasing air density. You probably can't get either high enough for a perfect match with inexpensive equipment, but you can get closer. Just make your airspeed as high as feasible, and consider using cold air if it's available or can be manufactured. (Got an A/C unit?) You can improve airspeed and reduce turbulence in the airstream by using a choke. Basically use a larger size air intake and flow conditioner, then smoothly decrease the tunnel size into your test chamber. This will increase airspeed and shrink any remaining turbulence after the conditioner. Try not to overfill your test chamber. If the model is too big, it starts compressing air against the walls of the chamber, which will affect your results. If you are doing full vehicle modeling, be aware of ground effects and how they are different for a stationary car vs. a rolling car. There are various techniques for dealing with this, but I've gone on long enough and you can research those on your own. Anyway, even without being obsessive about it, it's a really cool project and I look forward to your results! It's something I've thought about a lot (clearly) but never got around to doing myself, so I can live vicariously through you. Also, don't discount the redneck solution: strap your model to the roof of your car and drive down an empty road.
I found that to form dense smoke you can't mix it with air on the way, but you can attach the compressor to the input holes on the vape or even put the vape within a sealed tube with just the output poking thru and an inlet hole, then you can pump smoke out that way. it comes out much more dense. then, the tree that distributes the smoke needs long thin tubes to produce laminar tubes of smoke
I've been wanting to work with some diy aero on my car, looking forward to seeing you actually doing it! I'm also working on a degree in mechanical engineering to catch up with you :)
It makes me so mad the fact that you are a terrific engineer who doesn't really document your projects - at least not publicly. These could be *hours* long videos but instead they are only 10 minutes long. It is like watching a cooking channel where they show the ingredients and bang - "here is the cake! Try to figure out by ourselves how the it was made!" I keep coming to your videos but I always end frustrated by how brief they are. Do you know Mustie1? Combining the two of you would result in the bestest engineering channel of the whole TH-cam! Anyways, congrats for being this awesome regardless!
I've actually been thinking about a diy active aero project controlled by arduino/servos; don't really have the money to pull it off. Excited to see how it goes!
@@WesleyKagan i'm not selling myself short :) i can see how much prep and thought went into that! Griptopia (matt) i did Casey P's intro graphics if you need any 3d (i know you are more than capable) hit me up, fun working with other engineers :) I'll need to pick your brains at some point, i'm putting a v12 merc engine in a GTLM at some point.. was thinking of using triumph throttle bodies.. then saw your idea.. 955i's bodies go for cheap too.. The brass welding is tig practice for when i do my exhaust manifolds and other things. Great work dude..
Very exciting stuff!!! I would like to see an AND gate that sees input from the brake light and a sensor in the steering column to return the wing to the lowest drag coefficient when neither braking, nor turning.
With DRS in F1, the raised rear wing element returns to its base position as soon as the driver brakes but the downforce doesn't return to base levels until a moment later after the airflow has stabilised. If you're running a 3 pedal car, detecting when you lift off the throttle may give it a little more stability by the time you're braking.
@@WesleyKagan That would be far more intuitive to program, since you could easily increase down force with negative values and decrease them with positive ones along one axis, and increase down force with any value greater or less than 0 along the other. Perhaps you could incorporate a transducer and take any repeated sharp input (such as a line of potholes or other harsh road) and have the wing come up to push your car down more when your traction is upset. Up here in Michigan, 90 mph and a washboard of potholes is a real white knuckle experience! 😅
I love your videos man, great work. Although as an aero engineer I always get nervous when someone I like tries to explain lift. I'm always disappointed when someone uses the "equal transit time" explaination.
There is so much happening behind the scenes when it comes to you Wes! Do you have a project list or some other social media where we can follow? I wanna know more about the Freevalve Miata, V12 conversion of the 60s F1-car, etc 😎😍
It’s a whole lot behind the scenes surprisingly haha. There’s about 120 hours into this, with around 4 hours of footage edited down to 15 minutes. It’s getting more efficient but it’s still a lot
lol, Phx. That’s what high efficiency hvac is for, even garages, have done a few in the fire bird valley for just the same reason as you Wesley…wanting to hang out in the garage of Funland and not light on fire 🔥.
Wesley, I am building wind tunnels and making experiments as a job. I could probably answer some of your questions. Good work, but there are many weak points. 4 fans make a swirling flow and are too close to your model. There is a dead spot in the middle between the fans, flow velocity from your smoke device should match the flow velocity in the tube, inlet should not have straight corners because you will have a separated flow, at least you can match a Reynolds number for such experiments. You can use oil smoke wire for visualization or apply oil paint to a surface and see where the flow separates. you can also glue tiny threads to your model and watch behavior of the thread. It will be aligned with the flow or just chaotically moving in a separated zone. you can also install pressure ports and monitor static pressures on your wing surface. with oilvis you will know approximately where the separation occurs. And so on. Good luck!
Thanks! Nice build! You mentioned I thought that there was a list of materials used on the sidebar, but I did not see any there. Maybe you mentioned same in video but I don’t recall too many (for example) part numbers (although, you can pass on the straws ;-). Nice touch on the straws though.
Man you think just like me. I've wanted to make a DIY wind tunnel for years but you are taking it to the next level. I've always liked the idea of active aero too. I've been thinking about how to do a spoiler for a long time. 12v linear actuators are slow, and air actuators need a compressor etc. My only suggestion is your smoke tube thing. Maybe make it in Y style. So 1 Y's into 2, 2 into 4, 4 into 8 etc. In an effort to split the smoke equally . I'd also make the holes a little smaller. And maybe try make it more aerodynamic itself to keep the smoke lines cleaner. (think the shrouds around aeroplanes wheels with stationary wheels, the typical teardrop shape) Gagging to see the next episode.
There doesn't seem to be a point to making an active aero splitter up front because aren't ground effects like that without downside really? I'm not sure if I'm remembering this correctly but don't they reduce drag but also suck the car to the floor. Maybe some active aero canards? That would be pretty cool.
You should keep in mind that when you simulate small scale models, you have to alter some parameters to get the same airflow as around the full scale object you are simulating. In your case, Ma-number is low, so your most important goal would be to to preserve Re-similarity by altering flow velocity and/or viscosity (temperature). Also your explanation of lift was not quite correct. The wing experiences increased pressure on the underside (thats why we call it pressure side) and the explanation why the air flows faster on the suction side is also incorrect. Still, i think its great that you try this project. I would really like to have such a mini wind tunnel for my self.
Next you will be reinventing that Bose active suspensions system. Always blew my mind how the same system for noise cancelling can be used to cancel bumps in the road
Idea: get a big AC unit, use its output to push air into a wind tunnel. Problem is that the air exiting the wind tunnel will cool your shop too, so you'll have to figure out a way to reheat the air back to 100+ to remain consistent with your other videos.
This dude building awesome shit in his parents garage, at 23. Me: Rebuilding a Briggs and Straton lawnmower engine in shop class, with a shit load of important looking parts left over.
It might be useful to get some real numbers from the real car: add suspension actuation sensors and use corner scales to work out the correlation between suspension load and actuation, then mount a pitot tube to the nose of the car and add an accelerometer. The last part needed is a way to measure live engine output. Not sure if there's a standard kit for this, but I've got an idea of how it would be done using phase detection of both ends of a driveshaft, either magnetically with hall-effect sensors or with optical encoders.
My boyfriend vapes and it's very funny to see that huge box mod vape on the desk without you hitting it (big respect btw) because that would drive him crazy to just have that there without him being able to hit it
Hi Wesley, A couple of weeks ago I designed an airfoil specifically for a rear wing of a car. It has the highest possible lift/downforce I can get, and it is miles better than anything you can find on the internet. If you want I can send you the data points.
What I learned building several school science fair wind tunnels and test fixtures: Reynolds Numbers will kick your ass every time you think you can scale anything without animal sacrifice (at least) before the Reynolds Number Gods.
Very interesting! Seen this (dynamic wind brakes, down force generation, etc.) on a recent supercar but for the life of me I canno remember where (though I remember it was on a yt chan) or what that supercar was... So it does work. Now, I guess you don't have the same budget as those developping that supercar. So, yeah, more power to you!
Surprised you didnt just build your own vape using a mosfet, 18650 batteries, and a user rebuildable dripping atomizer (RDA). That would have allowed you to build a really big high resistance coil that better suited your vapor output needs all without burning out those expensive prebuilt coil modules! if it would benefit you, let me know & I can build one & have it shipped to you within 4 days. (even less if you just want to use the atomizer setup on your current power device) I can even label it as anime action figures on the shipping so your parents dont find out lololol
Who use prebuilt coils in 2021? And anime action figures? They can trigger parents waaay more than a vape. Guess I'm just old. Been hiding my smoke habits from parents for more than a year. 20 years ago.
Epic. I love your method of throwing yourself at stuff until you figure something out.I take it you've seen Claymore's twin-engine volvo on RR with the active aero?
Yo this guy is genuinely revolutionizing diy car mods, this should be on every car guys recommended.
Agreed!
Now we just need freevalve on the LS and I'll be all in.
@@androiduberalles get to printing...😂 freevalve ls with active aero, something mid-engine mabye?
Yes, except if he's talking about aerodynamics, almost everything he did in this episode was more or less wrong.
Especially his explanation of lift.
@@Enthropical_Thunder i agree about lift
As an RC aircraft modeller of 40 years, can I suggest you work more around wing profiles for downforce rather than angle of attack, high alpha aerodynamics , ie stalling the air over the airfoil are super draggy. That symmetrical airfoil you used with a rearward center of pressure is a poor one for downforce as you get too much separation, and separation is your enemy. I know this is just for testing, but if you test with a basic inverted Clark Y airfoil you would get more reliable and testable results as it is a forgiving low speed airfoil.
I agree, especially about the high angles of attack leading to flow separation/stall. However, I think an uncambered airfoil is better than a Clark Y or NACA airfoil for the proof-of-concept he has now.
Separation is not necessarily your enemy for automotive purposes rather than aerospace. Active aero vehicles nowadays use it as an airbrake at high speed, since flow separation creates excellent drag. You definitely don't want this for a fixed wing of course but his video title indicates otherwise.
@@samfedorka5629 Clark Y is uncambered, bit yeah, I picked it out of my butt as it was the simplist with the most data available airfoil I could think of and from experience I know it works well on tiny RC airplane models, admittedly, the other way up. , I am sure there are way better, but anything is better than that truely aweful airfoil he was using. But he is running tests at such low speeds and scale i doubt much data he is getting will be useful in the slightest. You have to go up to 1/4 to 1/2 scale before air flows similar to the full scale vehicle. Certainly no use with an 1/18 scale BBurago toy.
@@ForeverNeverwhere1 Angle of Attack (AoA) is everything. If you want to live with a wide range of AoA, then you want something with a blunt-round nosed airfoil, like a Clark-Y. If you want low drag at high speed, you want a shallow AoA range, and a thin, almost flat, airfoil. Look at the leading edge of the F-16, in a sharp turn, and watch the Leading Edge Flaps hold a narrow AoA. An F-5 flys well with a flat wing, iff, the AoA is held within the limit of the critical angle.
@@DumbledoreMcCracken You're comparing airfoils designed for supersonic flight (F-16, F-5) to those designed for low speed. There's a big difference. Most cars seems to use thin but heavily cambered airfoils for downforce compared to the thicker, less cambered ones used on airplanes. Possibly this is because they are supported on both ends (or very short) on a car as opposed to cantilevered on aircraft? The speed range they are used in for car aero is generally lower too, corner speed in a car is lower than what most airplanes experience in cruise. If anything wingson cars most closely approximate an airliner wing in landing configuration.
Crazy how randomly reading a comment on a TH-cam channel like this help me solve a problem I was having from something years and years ago
A few thoughts from an aerospace engineer:
-You want your flow straightener (straws) at the widest point of your intake nozzle and then a nice smooth contraction down to your test section. Both the straws and the contraction work to suppress turbulence in the intake. A contraction ratio of 10:1 (area-wise) or so would help a lot.
-Your smoke rake should be airfoil-shaped to remove the turbulence generated over the back of the down-tube. Then the rearward facing tubes should be as small as possible to minimize their impact to the flow. Some 1/32" brass tubing from K&S would work great and could be pressed or glued into a 3-D printed part.
-Optionally, you want some expansion (a diffuser) behind the test section. If you used 9 fans arranged in a square but kept the same test section size, that would be a good starting point. The diffuser should have a half-angle of no more than about 3-5 degrees.
Can't wait to see what you come up with!
These are great insights, thank you for sharing.
What did you think about his molecule covering a greater distance in the same amount of time? I thought we were past that in aerodynamics?
The contraction section is to accelerate the flow rather than smooth it out.
Before creating a higher velocity wind tunnel, I'd rather assess repeatability of the set-up, which will probably be very, very poor. In order to improve it, it would be wise to remove the smoke section, which in real life is not used anyway, and also the test subject should be much smaller in order to prevent blockage and wall effects.
@@anidiotinaracingcar I think he was just trying to explain that air (and air molecules) moves faster when where is a decrease in pressure
@@Agussg23 A larger contraction would only produce a higher velocity if you had the fans on the inlet and not in the diffuser. With them on the diffuser, all it does from a velocity standpoint is lower the velocity at the inlet.
I'm sorry, did you say 23? Wow. You're even more of a genius than I thought. Your career of insanity has just begun!
I had the same thoughts. Kid is insane. I say kid now as your way younger then I thought possible. Haha.
@@huddleberryfin I'm 30 and he is making me feel like I should have actually done something with my life. 😂
I'm 22 and regretting college, and that's all my achievements outside of hobbyist writing
I got caught up on that too. All the thing you do at 23... wow!
I know I was like, but how??? :)
Cool windtunnel; if you want representative results, you’ll need faster flow to match the reynolds number of your actual project.
I was thinking the same thing, some dimensional analysis may be in order to get proper values for the scaled tests
Awesome video! One thing: "equal transit" explanation for lift is a very common misconception. No idea why this keeps getting taught to folks! Lift is mostly just Newtonian, that is momentum transfer, so with an angle of attack, planes can even fly upsidedown. The air does move faster over the top of the wing, but not due to some constraint on the time it must take to get there. It clings to the curved surface falling away from it due to the coanda effect. This serves to improve the efficiency of the airfoil mostly, but does not provide a significant amount of lift.
Actually, most of the lift is provided by the upper surface
Yes. The equal transit theory is very flawed. Doesn't take much to prove that there is no equal transit. Basically, air takes longer to get to the trailing edge when it follows the upper surface rather than the lower one, which is intuitive.
@@youngtschakaloff That's true, at least it usually is, but the air _does_ move substantially faster over the top surface than the bottom. Not because of some magical "equal transit" arrangement of course (I love the idea of the air molecules slowing down to wait for their friends) but because the air above the wing is compressed into a narrower stream thanks to the air above _that_ not getting out of the way. Narrower stream means faster movement which means lower static pressure which means lift.
@@cooperised okay, reasonable. That way of explaining is a million times better indeed.
I always keep in mind that there are multiple phenomena at work at the same time, and some of them are really hard to fully explain
@@youngtschakaloff Absolutely! You're thinking in exactly the right way. Multiple phenomena are at work at the same time, and are often the same phenomenon viewed from different perspectives. For example with a wing, it also deflects air downwards as it generates lift. I've seen people say "so it's the deflection of the air downwards that generates lift not the pressure difference", but that's not right. If the wing is generating lift then the pressure below _must_ be higher than the pressure above, that's what generates the force. But Newton's third law also applies: that upward force on the wing produces an equal downward force on the air. Or if you like, pushing the air downwards generates an upward force on the wing. Both things (pressure difference and Newton's third law) are true, and neither is "the explanation" for where lift comes from. They both just describe phenomena that can be observed around a wing.
"Aerodynamics can be a challenge, no that's too nice. Air is dumb and I don't like it." I feel like this should be on a t-shirt.
Merch? 👀
It irritates me how clever and capable you are at 23🤣
You are like the zoomer equivalent of Bad Obsession Motorsport.
You are literally my hero. Everything I wish I had time to do/don't make the time to do, you do! And share with the rest of us!
Thanks for the kind words!
6:58 - "All physics taught below Junior year college is a lie; and it's taught that way because it's funny" -Alice Caldwell-Kelly (Well There's Your Problem Podcast)
Not true. Most engineering is done with high school level physics.
@@chrispollard6568 All models are wrong, but some are useful. en.wikipedia.org/wiki/All_models_are_wrong
@@squiresuzuki My models are not wrong. They are mostly within 3% You use the results of those models.
@@chrispollard6568 You missed the point. Being "within 3%" is by definition wrong, however, it is an adequate approximation of reality for most engineering purposes. Similarly, as explained in the link I gave, Newton's theory of gravitation (high school physics) is completely wrong because it goes against Einstein's theory of relativity, but again, Newtonian physics is far simpler and is more than adequate for the vast majority of purposes.
@@squiresuzuki It isn't "Wrong" because it works. My "high school" physics works to give people like you working products that you actually use. Your computer disk drive WORKS.
Another aero engineer here. First off, great work so far and im excited to see whats next. I getting into arduino myself now, so ill be very keen to follow your project. A couple of points I wanna mention:
1)be careful about the blockage ratio, thats the ratio of the objects frontal area vs the area of the wind tunnel's cross section. Most I've see were around 10% and beyond that the wind tunnel's walls might start influencing your results. I realise it might become cost prohibitive though.
2)try adding a mesh of sorts at the start of your wind tunnel which should help to make the incoming flow less turbulent and thus lead to more accurate results
3)if possible make the airfoil holder from metal or design it to be as stiff as you can, so that there's no force loss in the flex.
I work on racecar aero for a living so if you need any advice or cfd simulation help, feel free to shoot me an email.
The flex is an interesting point- if you notice the metal beam across the bottom was from results of flex on the bottom wall. I also want to calibrate it with and without the massive servo
@@WesleyKagan the way its usually done is you run your setup without the test subject first to get the "error" data and then you run it with your wing. At the end you subtract the first from the latter to get your real values.
Id also suggest on getting a pitot tube in there to get the velocity of the flow (place it in front of your wing). That way you can get coeff of lift (C_L) or (C_L*Area) and the Reynolds number which should help you validate your data against other sources. Reynolds number is very important, flow will act differently at different Re numbers. Lower Re flows will tend to separate sooner, which will lead to lower peak C_L.
you should put your test subject on a sled to measure drag force. From there you can calculate downforce-efficiency. You can also measure air speed with a pitot tube
Ever since the new Alfa Romeo Giulia came out with the active splitter at the front ive been wanting to replicate it, so the next video is right up my alley
I love that this dude got Noctua fans for a wind tunnel test. If you don't know anything about PC building and cooling, Noctua fans are regarded as being the top of the line fan for their smoothness cooling properties and silence. I just think its hilarious he went all out for the wind tunnel build, but kudos to him, this dudes a genius with what he is doing and has done including incorporating free valve tech into a predator engine lol.
Please upload more🙏🏼 you’re a bona fide genius, like Mensa level! Okay okay enough of that, your channel is kinda awesome though. Thank you for your entertaining projects.
Thanks! I’m trying to get a bit more consistent with uploads!
I absolutely love your vids man. Truely an inspiration for my own projects that im in the process of researching.
Glad to hear it! Thank you!
Looking forward to seeing more. Like that you dont know if it works but think it will so giving a go. I think it will too. Good job on the wind tunnel
Love the ‘anything is possible’ problem solving attitude. Real smart!
7:00 equal transit theory is an incorrect assumption made pretty often. In reality, the flow over the top of a cambered airfoil or symmetric airfoil at an angle of attack is much faster due to the decrease in pressure, reaching the trailing edge quicker than the flow over the bottom of the airfoil. There are a lot of subtleties that I am not sure of myself.
Cool project, looking forward to your next upload.
Hey man, if you have neighbors with kids 10 through 20 you should really ask the parents if they would like their kids to come and join you in experimenting with this kind of stuff. Kinda like Casey Putsch's genius garage. Growing up I didn't have anyone around me doing stuff like this, but I always stayed late in chemistry and mechanics class to learn more. You could be a mentor and teacher for an aspiring racing engineer or something! I absolutely love this stuff, really nice work.
Holy crap, just as I start designing an active aero setup for my build... You drop this video series, how insane is this? I look forward to seeing what you come up with! You might see a "duplicate" on my car at some point if it will fit!
Fair enough! Yeah it’s kinda uncharted territory for me but I’m excited to tinker with it
13:41 this is exactly what I’m trying to do too hopefully you keep up the effort, the video is 2 years old I’ll still be tuning in to see updates great job
If you can get some of this stuff you're tinkering with onto one of your cars and working, this channel is going to the moon!
Good stuff. You will run into the limitations of a miniature tunnel so I would say your best bet is your full size algorithmic approach otherwise you will end up having to do that anyway because of correlation issues. I built something similar many years ago with a broken vacuum cleaner which I then ended up starting, but not finishing, building a head porting flow bench. You will get better laminar flow if you have a vacuum chamber pulling though straws as well as a straw entry plenum away from the front of the pitots. Good luck
Mike Patey, who you can watch on youtube, built a wing with flaps and slats that may have interesting information to add to the conversation. It is a different level of complexity but airfoil design and engineering is getting some pretty cool visibility recently. Thank you for sharing this with the world!
Wesley, I thought of a great little input you could run for the active aero project. Snag a steering angle sensor (whether cable or touchless or otherwise), and anything UNDER 25* steering angle you drop it down as long as Brake Pedal is not enabled. That way, it'll work when you're going straight and we only have a total of 100% Grip at a tire, so once we have over a certain steering angle, and not on the brakes, we would likely be accelerating. I think it's a real simple sensor configuration and a great start on a budget given your Arduino capabilities. Good luck on the journey!
could potentially even be a linear relationship between steering angle and wing angle, or some other simple function (log, root, square,....?) depending on what fits
Ahh, what a friday treat! Thanks Wesley
👌🏽 thank you for taking us with you on this journey.
A soldering iron and some light machine oil on a rag creates great smoke for PCV system leak tester, too. Can use Johnstone's baby oil instead. I suggest the use of a 3d printer hotend heater in a glass jar so you can set exact temp and even a mini hotend fan to slowly blow the smoke though. Also recommend using blunted syringe needles (from a printer ink refilling shop).
It was 11° C here (Switzerland) today, raining.
Great ideas and experiments, thanks!
20 here in switzerland (you're on the wrong side of the gotthard obviously :)
@@odl21 very well observed.
I really like your mini wind tunnel, with the close up it looks legit.
Pretty sure XF motorsports is watching this :-)
Air ejector nozzle. Dealt with something similar to what you made using the venturi in one of my engineering internships. High pressure steam is run through an ejector nozzle and create a vacuum behind it in order to draw NCGs out of steam turbine condenser.
Always excited to see your videos! If I may offer one small but of advice, you can acquire a set of Bluetooth quarter scales used for rc drag racing and measure balance and changes made to the model in the tunnel in real time.
Bit^
I think plugging the holes in the smoke distributor except one would help visualize the airflow better. Also, since hot air is less dense than cold air, maybe you can play with temperature to help get a more accurate set of data points. Very cool project. Keep going.
Just like you said, without watching the video I knew it was glikf to be awesome so liked and commented already. Now onto enjoying your enginuity and scientific approach.
Thanks so much!
That Mercedes is just beautiful.
They've lost their way lately. Your's is a keeper.
It seems like my comment has vanished, but anyways. Wind tunnels and aerodynamics, awesome and can't wait for part 2. Watching this I was immediately reminded of a set of videos made by Yoshi Suzuka probably 10 or so years ago and the discussion around them. Back then he was working on, I think it was something like an office small scale wind tunnel. Really interesting and worth checking out for reference on how far it's gone since then. Also, at the time he was doing it with 1/24 I think scale models and had it down to getting repeatable results and trends, though not necessarily representative or completely accurate to full scale numbers. I remember that wind tunnel at the time he could measure lift by the corners of the car and drag. IIRC Suzuka did the aerodynamic development for the Nissan GTP cars for a large portion on a 1/7 or 1/8 scale wind tunnel and models. I vaguely remember Suzuka may have said that with the 1/7 (or 1/8) scale wind tunnel, he was able get 2% drag and 4% lift correlation to full scale. Anyhow, you're definitely on the right track to making something that is close enough.
Oh yeah, there's a whole bunch of test data (or at least was time I checked...several years ago) you've already found for various NACA profiles and even "race car stuff" wings where it should be possible to try and repeat the test with the same profile and see how the results correlate.
Some suggestions if you're looking for maximum results:
As mentioned, for better similitude between your model and reality, you need to try to match the Reynolds number. For a smaller-than-reality model, you can do this by increasing air velocity and increasing air density. You probably can't get either high enough for a perfect match with inexpensive equipment, but you can get closer. Just make your airspeed as high as feasible, and consider using cold air if it's available or can be manufactured. (Got an A/C unit?)
You can improve airspeed and reduce turbulence in the airstream by using a choke. Basically use a larger size air intake and flow conditioner, then smoothly decrease the tunnel size into your test chamber. This will increase airspeed and shrink any remaining turbulence after the conditioner.
Try not to overfill your test chamber. If the model is too big, it starts compressing air against the walls of the chamber, which will affect your results.
If you are doing full vehicle modeling, be aware of ground effects and how they are different for a stationary car vs. a rolling car. There are various techniques for dealing with this, but I've gone on long enough and you can research those on your own.
Anyway, even without being obsessive about it, it's a really cool project and I look forward to your results! It's something I've thought about a lot (clearly) but never got around to doing myself, so I can live vicariously through you.
Also, don't discount the redneck solution: strap your model to the roof of your car and drive down an empty road.
I found that to form dense smoke you can't mix it with air on the way, but you can attach the compressor to the input holes on the vape or even put the vape within a sealed tube with just the output poking thru and an inlet hole, then you can pump smoke out that way. it comes out much more dense. then, the tree that distributes the smoke needs long thin tubes to produce laminar tubes of smoke
I've been wanting to work with some diy aero on my car, looking forward to seeing you actually doing it! I'm also working on a degree in mechanical engineering to catch up with you :)
It makes me so mad the fact that you are a terrific engineer who doesn't really document your projects - at least not publicly. These could be *hours* long videos but instead they are only 10 minutes long. It is like watching a cooking channel where they show the ingredients and bang - "here is the cake! Try to figure out by ourselves how the it was made!" I keep coming to your videos but I always end frustrated by how brief they are. Do you know Mustie1? Combining the two of you would result in the bestest engineering channel of the whole TH-cam! Anyways, congrats for being this awesome regardless!
Absolutely cannot wait for part 2!!
I've actually been thinking about a diy active aero project controlled by arduino/servos; don't really have the money to pull it off. Excited to see how it goes!
I worked in F1 in the wind tunnel, a mind blowing system even back in the day using blown exhausts even this was simulated..
Nice video 👍
Very cool! Thanks!
Love the build, the video and your humor! Keep it going, curious how this evolves
You are amazing..now this is real influencer
i tried to weld two bits of brass together and you built a wind tunnel... awesome stuff..
Hey, brass is hard. Don't sell yourself short.
@@WesleyKagan i'm not selling myself short :) i can see how much prep and thought went into that! Griptopia (matt) i did Casey P's intro graphics if you need any 3d (i know you are more than capable) hit me up, fun working with other engineers :) I'll need to pick your brains at some point, i'm putting a v12 merc engine in a GTLM at some point.. was thinking of using triumph throttle bodies.. then saw your idea.. 955i's bodies go for cheap too.. The brass welding is tig practice for when i do my exhaust manifolds and other things. Great work dude..
Another great video. A smart concept and development.
Thank you!
The weight sensor is identical in design to those used in digital scales and goes to 500g usually instead of 1/3.
Of course there is a WK video about wind tunnels!
Very exciting stuff!!! I would like to see an AND gate that sees input from the brake light and a sensor in the steering column to return the wing to the lowest drag coefficient when neither braking, nor turning.
I’d actually like to set it up so it goes off accelerometer instead of sharp inputs.
With DRS in F1, the raised rear wing element returns to its base position as soon as the driver brakes but the downforce doesn't return to base levels until a moment later after the airflow has stabilised. If you're running a 3 pedal car, detecting when you lift off the throttle may give it a little more stability by the time you're braking.
@@WesleyKagan That would be far more intuitive to program, since you could easily increase down force with negative values and decrease them with positive ones along one axis, and increase down force with any value greater or less than 0 along the other.
Perhaps you could incorporate a transducer and take any repeated sharp input (such as a line of potholes or other harsh road) and have the wing come up to push your car down more when your traction is upset. Up here in Michigan, 90 mph and a washboard of potholes is a real white knuckle experience! 😅
The concept of adaptive/AI aero on a budget sounds so cool
I agree!
I love your videos man, great work. Although as an aero engineer I always get nervous when someone I like tries to explain lift. I'm always disappointed when someone uses the "equal transit time" explaination.
Your videos are great, man. Keep doing great videos.
There is so much happening behind the scenes when it comes to you Wes! Do you have a project list or some other social media where we can follow? I wanna know more about the Freevalve Miata, V12 conversion of the 60s F1-car, etc 😎😍
It’s a whole lot behind the scenes surprisingly haha. There’s about 120 hours into this, with around 4 hours of footage edited down to 15 minutes. It’s getting more efficient but it’s still a lot
lol, Phx.
That’s what high efficiency hvac is for, even garages, have done a few in the fire bird valley for just the same reason as you Wesley…wanting to hang out in the garage of Funland and not light on fire 🔥.
Excited for next steps!
Always amazing Wes. Very interesting.
You always manage to make me feel bad about my self - kudos
Keep it coming
Wesley, I am building wind tunnels and making experiments as a job. I could probably answer some of your questions. Good work, but there are many weak points. 4 fans make a swirling flow and are too close to your model. There is a dead spot in the middle between the fans, flow velocity from your smoke device should match the flow velocity in the tube, inlet should not have straight corners because you will have a separated flow, at least you can match a Reynolds number for such experiments. You can use oil smoke wire for visualization or apply oil paint to a surface and see where the flow separates. you can also glue tiny threads to your model and watch behavior of the thread. It will be aligned with the flow or just chaotically moving in a separated zone. you can also install pressure ports and monitor static pressures on your wing surface. with oilvis you will know approximately where the separation occurs. And so on. Good luck!
I dream of making my own active aero, so I'll be eagerly watching these
Old video of a clay model Mercedes 190E in the wind tunnel is really cool👍
You’re telling me you’re 23! You keep getting more impressive
Thanks! Nice build! You mentioned I thought that there was a list of materials used on the sidebar, but I did not see any there. Maybe you mentioned same in video but I don’t recall too many (for example) part numbers (although, you can pass on the straws ;-). Nice touch on the straws though.
Love your videos dude! Keep up the GREAT work!
Thanks!
I thought I saw an Allante behind you and got a little excited
That wall mounted PC is sick
Man you think just like me. I've wanted to make a DIY wind tunnel for years but you are taking it to the next level. I've always liked the idea of active aero too. I've been thinking about how to do a spoiler for a long time. 12v linear actuators are slow, and air actuators need a compressor etc.
My only suggestion is your smoke tube thing. Maybe make it in Y style. So 1 Y's into 2, 2 into 4, 4 into 8 etc. In an effort to split the smoke equally . I'd also make the holes a little smaller. And maybe try make it more aerodynamic itself to keep the smoke lines cleaner. (think the shrouds around aeroplanes wheels with stationary wheels, the typical teardrop shape)
Gagging to see the next episode.
There doesn't seem to be a point to making an active aero splitter up front because aren't ground effects like that without downside really? I'm not sure if I'm remembering this correctly but don't they reduce drag but also suck the car to the floor. Maybe some active aero canards? That would be pretty cool.
thank you professor Kagan
You should keep in mind that when you simulate small scale models, you have to alter some parameters to get the same airflow as around the full scale object you are simulating. In your case, Ma-number is low, so your most important goal would be to to preserve Re-similarity by altering flow velocity and/or viscosity (temperature).
Also your explanation of lift was not quite correct. The wing experiences increased pressure on the underside (thats why we call it pressure side) and the explanation why the air flows faster on the suction side is also incorrect.
Still, i think its great that you try this project. I would really like to have such a mini wind tunnel for my self.
Great video 👍I'm sure Gayle Banks would share your kiss approach when building out your windtunnel etc.
Next you will be reinventing that Bose active suspensions system. Always blew my mind how the same system for noise cancelling can be used to cancel bumps in the road
VHS
You're killing me Smalls!
VHS joke was epic!
Didn't expect Nick Robinson to teach me fluid dynamics today
Amazing as always 👌
Dude you have at least a good bit of knowledge of two different engineering degrees displaying here, awesome stuff
Thanks!
Will there be updates for the "FreeValve" project? You've got many cool things but for me that is the one. Thanks for all the videos!
Videos like this make me feel so stupid... And it's a great feeling after all... Love it... 😎
Idea: get a big AC unit, use its output to push air into a wind tunnel. Problem is that the air exiting the wind tunnel will cool your shop too, so you'll have to figure out a way to reheat the air back to 100+ to remain consistent with your other videos.
3:15 Wait, you're my age??? Man, Im even more jealous of your work now
This dude building awesome shit in his parents garage, at 23.
Me: Rebuilding a Briggs and Straton lawnmower engine in shop class, with a shit load of important looking parts left over.
It might be useful to get some real numbers from the real car: add suspension actuation sensors and use corner scales to work out the correlation between suspension load and actuation, then mount a pitot tube to the nose of the car and add an accelerometer. The last part needed is a way to measure live engine output. Not sure if there's a standard kit for this, but I've got an idea of how it would be done using phase detection of both ends of a driveshaft, either magnetically with hall-effect sensors or with optical encoders.
Yeah I have a few different ideas for measuring speed, wheel speed, forces involved, exc.
My boyfriend vapes and it's very funny to see that huge box mod vape on the desk without you hitting it (big respect btw) because that would drive him crazy to just have that there without him being able to hit it
Hi Wesley,
A couple of weeks ago I designed an airfoil specifically for a rear wing of a car. It has the highest possible lift/downforce I can get, and it is miles better than anything you can find on the internet. If you want I can send you the data points.
How the hell are you 23. I'm 22 myself, but goddamn brother. Good shit, man. Keep it up!
Very cool. All hail the algorithm!!
What I learned building several school science fair wind tunnels and test fixtures: Reynolds Numbers will kick your ass every time you think you can scale anything without animal sacrifice (at least) before the Reynolds Number Gods.
Dude that R129 is awesome!! I need to know what you’re gonna do to it!
Very neat project, thanks.
Very interesting! Seen this (dynamic wind brakes, down force generation, etc.) on a recent supercar but for the life of me I canno remember where (though I remember it was on a yt chan) or what that supercar was... So it does work. Now, I guess you don't have the same budget as those developping that supercar. So, yeah, more power to you!
Awesome idea using the vape to generate the smoke.
lmao your explanation on lift finally got my stupid brain to wrap around the actual science of it lol
Surprised you didnt just build your own vape using a mosfet, 18650 batteries, and a user rebuildable dripping atomizer (RDA).
That would have allowed you to build a really big high resistance coil that better suited your vapor output needs all without burning out those expensive prebuilt coil modules!
if it would benefit you, let me know & I can build one & have it shipped to you within 4 days. (even less if you just want to use the atomizer setup on your current power device)
I can even label it as anime action figures on the shipping so your parents dont find out lololol
Who use prebuilt coils in 2021? And anime action figures? They can trigger parents waaay more than a vape. Guess I'm just old. Been hiding my smoke habits from parents for more than a year. 20 years ago.
Nice setup, I was thinking about making small air tunnel for long time. Btw, active aero are banned in most racing series at least in Europe.
Yeah, with most of my projects it’s not destined to be under any sanctioning body but mostly just for personal use
I spotted that metal sheet with the Spa Francorchamps cut out of it 😎
Damn that bernoulli joke was slick
Not sure how important smoke delivery is down low, but you should feed smoke into both ends of the nozzle manifold. It'll give better distribution.
Epic. I love your method of throwing yourself at stuff until you figure something out.I take it you've seen Claymore's twin-engine volvo on RR with the active aero?