$10 000 Physics Wager Settles The Debate On Sailing Downwind Faster Than The Wind

By now, many of you have seen the video of [Rick Cavallaro]’s Blackbird, the controversial wind-powered land vehicle that can outrun the wind. The video has led to a high-profile $10 000 wager between [Derek Muller] aka [Veritasium] and [Alex Kusenko], a professor of physics from UCLA. [Veritasium] won the wager with the help of a scale model built by [Xyla Foxlin], and you need to watch the videos after the break for some excellent lessons in physics, engineering, and civilized debate.

After seeing [Veritasium]’s video on Blackbird, [Professor Kusenko] contacted him and said the performance claims and explanation were incorrect. After a bit of debate [Veritasium] proposed a wager on the matter, which [Professor Kusenko] accepted, and it was made official with a written agreement witnessed by [Neil deGrasse Tyson], [Bill Nye], and [Sean Carrol]. From the start, it was agreed that the entire debate would be made public.

[Professor Kusenko] made a very thorough and convincing argument, backed by calculations, against the claims in the video. He claimed the observations were due to a combination of gusty winds, a vertical wind gradient. He was convinced and that the vehicle would not be able to maintain a speed higher than the wind, directly downwind. By [Veritasium]’s own admittance, his original video could have contained more details and proof of performance claims of the Blackbird vehicle. He added these to the latest video and included two model demonstrations. The model that brought the concept home for us is at 13:46 in the video, and substitutes the propeller for a large wheel being driven by a piece of lumber being bushed across it. The second model, built by [Xyla Foxlin] was designed to demonstrate the concept on a treadmill. The 4th version of [Xyla]’s model was the first to be successful after she found out from [Rick Cavallaro] that the key design detail is the Vehicle Speed Ratio, which must be 0.7 or less. It is the pitch of the propeller divided by the circumference of the driven wheel, assuming a 1:1 gear ratio. All the 3D files and details are available if you want to build your own downwind cart.

In the end, [Professor Kusenko] conceded in light of the evidence, and [Veritasium] will use the money towards funding more STEM education videos. What stood out for us is the humble and civilized manner the debate was handled. As [Veritasium] says, scientific agreements are not problems, but rather an opportunity for everyone to learn. If all disagreements were handled in this manner, mankind would be a lot better off.

95 thoughts on “$10 000 Physics Wager Settles The Debate On Sailing Downwind Faster Than The Wind

  1. His original explanation always bothered me because it implied “bluff body” provides the force for the wheels. As far as I can figure that was 100% wrong. The second one he’s sort of clearer and it’s better, but he’s not clear that there is a force on the fan axially that is pushing the vehicle ahead, while the wheel torque is extracted back. I want to put strain gages on a model to show this because I think his new video and Xyla Foxlin’s videos still don’t make this clear.

    1. This sounds *way* more complicated than it needs to be…

      Putting a force Fg on the ground moving at -V (relative to you) generates power Fg*V.

      Exerting a force Ft on the wind moving at V-W takes power Ft(V-W). Set the power generated equal to that consumed, and Ft is greater than Fg. Everything else is just efficiency and gearing.

      Doing it in a vehicle’s cool for drama, but you could just do this with force gauges and a fixed power input.

  2. The prof kinda lost me at his wind gusts theory. The gust would have had to peaked incredibly high to have averaged out to get to a constant forward velocity of as high as it was.

    I just cant imagine betting 10k on something as silly as this. Maybe im just jealous i dont have this kind of disposable income.

    1. Frankly, before putting down $10k, I would have done the research better. It’s been well explained before, and it was obvious from some of his comments that he didn’t understand that the blades were pushing against the wind. Hubris? Doesn’t play well with intellectual rigor.

      I’ll also give up plus-points for Veritasium for admitting that he didn’t explain it well in the previous video, because he really didn’t. He did a lot better in this one. So that’s a win. Probably worth that dude’s $10k. :)

      1. I mean, the basic idea’s been explained for like, what, 40-50 years? The funny thing is that *this* version (downwind faster) hadn’t been shown because the *original* version (*upwind* faster) was much more interesting.

        Which is why it’s amazing this guy bet anything. There have been *plenty* of papers on this.

    2. It has certainly become the darling of the YouTube science explainers (several really done poorly). In a week or 10 days a dozen or more have jumped on it. And wasn’t it done to death maybe 10 years ago?

      1. I was watching it unfold online ~15 years ago, but that’s the thing: there’s a whole generation grown up for whom this is new and really freaking cool.

        https://xkcd.com/1053/

        (And yes, the idea is older still. I’m sure some old codger was saying that they’d settled this in the late 60s, back when I was getting confused by it in 2005.)

        All of these “revelation” videos could be made a lot shorter by starting out with “the propeller pushes against the wind”. Think airplane, not boat. Airplanes fly faster than the wind using propellers all the time. (But then where does the power come from? The wheels. Done.)

          1. Nice find. I agree about new generations. But I’m surprised by just how many of the physical science related channels rushed something out. I really shouldn’t be as that is the nature of social media.

    1. I about never gamble.
      I’d accept the bet*, properly formulated and mutually agreed to.
      ie * I maintain that a globe earth model is a far far far better representation than a “flat earth” model.
      Or, did you mean the other way around? :-)

      Russell
      rmcspam@gmail.com <– yes, t's real.

    2. Is a 3D sphere “flat” when embedded in a 4D space, in the same way a plane is “flat” in 3D? Is “flat” just the quality of an object having one degree of freedom fewer than the space in which it’s embedded?

      B/c then I’ll take that bet, if you’re willing to take this hit of acid…

    1. I haven’t watched the newer videos yet, but, yes.
      Iirc the concept is essentialy that the angle on the blades puts it into a rotating tack (sailing at angles upwind) when it is faster than the wind. Using the rotation to drive the wheels.

        1. The energy is derived by the difference in speed between the wind and the ground. Essentially the device is “pushing the sail backwards” as it progresses. The small geared cart is a good example of how this works, another one is a thread spool that can be pulled faster than the thread it is pulled by while rolling over a table. It’s still the thread doing the pulling, and the ground providing the anchor point it is pulling against.

          1. Asking “where is the energy coming from” is always silly in these situations, because it just depends on your point of view. Is the wind moving, or is it the earth? Are you speeding up the Earth or slowing down the wind?

            Either way’s correct. You’re just pulling energy from the difference in speeds.

      1. The cart is travelling down wind but the power is coming from the blades that are travelling at an angle to the wind. So is it really travelling down wind?

      1. What made it click for me was the visual of a boat sailing at an angle around a cylinder. Suddenly it seemed trivially obvious: if a sail can go downwind faster than the wind on a straight path, it can do it on a spiral path, too. “That’s a prop!” indeed.

        1. The difficulty is that most people don’t know why a boat can sail downwind faster than the wind. Heck most people don’t know how sails work *period*. You’ll have people think your downwind speed can’t exceed wind, which is like thinking you can’t sail upwind. It’s just that people don’t understand sails generate forces in *two* directions: both in the direction of apparent wind and perpendicular to it.

          The funny thing is… this has nothing to do with why you’re able to do this, really. You don’t need a prop for this – it’s just easy and efficient. The reason why you can do this is just that there are two moving surfaces/fluids of different speeds.

          You can *always* extract energy when there are two surfaces/fluids of different speeds. It’s easy. A load Fd on the faster moving (at V) surface generates power Fd*v. That power generated thrust Ft on the slower surface (at V-W) of Ft = Fd*v/(v-w). Ft is greater than Fd, so you speed up. The challenge is in the inefficiencies. In an ideal case, this is trivial.

  3. For me, this highlights the importance of experimentation in physics, and the damn outright naivety of trying to teach it (or further the study of it) without any. My first choice for study at tertiary level was physics, but after 12-months of scribbling interminable notes, and not seeing or touching a single teaching apparatus, I completely lost my mind and changed subjects. Nowadays I do my own experiments, yet I’m still completely clueless as to how to go about that with any form of scientific rigour. My conclusion…time (and loads of money) wasted!

    1. Okay, I’m confused. Is it like an aerofoil, when you push it along (in this case rotate it) with a certain amount of drag-opposing force, and you then get a lifting force quite a bit larger than the initial impetus? If so, then I’m still totally confused as to how both aeroplanes fly, and Derek’s go-kart work.

  4. The level of debate reminds me of the “plane taking off from a treadmill moving in the opposite direction” myth that Mythbusters smashed early in their career. Bu the “faster than the wind downwind” debate was even more controversial because it seems so backwards.

    What really “unlocked” it in my brain was realizing that the cushion of air behind the vehicle becomes part of the dynamics of the vehicle itself. Airspeed an Groundspeed are even more easily misunderstood than the treadmill myth.

    1. Adam Has talked about this. Apparently, from what he’s gathered, their are still a fair amount of people that believe it won’t take off, citing that the experiment was flawed for some reason. This experiment sure could use some more sensors and data to help make sense of it all. Would like to see a graph of the TQ on the wheels and prop for an entire run. Also, would love to see it described by an energy point of view. Where is the energy coming from (wind, duh) and where it’s going to, and where are the loses – so that you get Energy(in)=Energy(out)???

      1. I re-watched that episode recently, and their explanation of the decoupling of airspeed/groundspeed was really good. With some basic research, one can realize that this is proved further by the fact that with sufficient airspeed (in the form of a strong headwind) it is possible to take off without *any* groundspeed.

        I wonder if the deniers are also flat-earthers or another similar ilk.

        1. Hey, I’m one of the deniers. My complaint with mythbusters isn’t “there were tiny flaws in the experiment that made it invalid”, it’s “they tested something completely different to what the debate was about”.

          In hindsight what’s interesting about the plane treadmill debate is, it’s two groups of people who can’t agree on what the question is and are both correct about their interpretation of it.

          One group thinks the question is “can a plane reach take off speed relative to the surrounding air even though it’s on a treadmill moving at take off speed in the other direction”; this is what mythbusters tested and they were correct that it can.

          The other thinks the question is “can a plane take off if it’s moving at take off speed relative to a treadmill but stationary relative to the surrounding air”, and we were correct that it can’t. Mythbusters didn’t test this, their experiment doesn’t disprove it, and Adam Savage agrees that it can’t in his recent video on the controversy.

          So really, the reason I’m on the “it can’t take off” side of the debate is not because of any scientific principle, on which we all agree, it’s a semantic issue of what the original question was supposed to mean. It seemed obvious it was trying to describe a situation where the plane was stationary, and IMO the other question is not interesting or worth discussing.

          Either way it has nothing to do with science denial.

          1. But the “other” group got it wrong because that was never the question that got asked.

            While the answer to the second formulation of the question is obvious, it’s irrelevant, because it would amount to tethering the plane with a rope to stay stationary above the treadmill and then cranking the treadmill up to speed. Notice the hidden assumption: the airplane is held stationary by some external force, which already answers the question. It is this assumption that is being added to the question by the second group of people, which is why they’re unwittingly making a tautology: a plane held in place will not move.

          2. In the second case the power applied by the airplane is only sufficient to make up for the drag of the friction in rotating the wheels, which is really low, and not accelerating the plane.

            So, sure, a plane won’t take off at bare engine idle in still air conditions.

            As Dude points out, if the plane has a tether to prevent it from moving regardless of engine power then how would it ever take off?

            Like the Answer to Life, the Universe, and Everything, one must first understand what the question means before one can understand the answer.

    2. The debate over Marilyn Vos Savant’s claim about the Monty Hall Problem (Let’s Make a Deal) was monumental and almost entirely on logical grounds when it was easily checked by experiment or simulation to exhaustion. 15 minutes with a coin to flip and a piece of paper was convincing.

      It is an easy trap. In his Cosmos remake Niel Tyson makes a ridiculous basic blunder about the Moon and brightness in a narration that must have been written, re-written, reviewed, and edited then recorded and edited and previewed. Plus the concept being fundamental to astronomical measurements made it more glaring. I guess it is like misreading your own writing errors over and over.

        1. No discussion I know of, it is just something that got my attention. I actually did not watch the rest of the show after that. Something about the Moon being brighter when it was closer, which isn’t how light from an illuminated object works. If it did work that way, you would be vaporized as the brightness became infinite if you tried to land on the Moon or any planet. The Moon is a Sunlit object and the same distance from the Sun as the Earth. To take a picture use the same exposure as daylight on the Earth. It doesn’t matter how close you are. An astrophysics guy should have this in the blood.

          1. I don’t think you’re explaining it well. The surface brightness of *any* object is independent of distance. Doesn’t matter if it’s illuminated or if it’s radiating itself.

            The confusion is because the casual idea of “brightness” isn’t the same as any more technical term (like you would use for photography, astronomy, or physics). The total light coming from the object obviously increases, but the surface brightness (power per unit solid angle) does not, so long as the object is resolved (obviously once it becomes unresolved, its brightness drops with distance because its angular size doesn’t change anymore). As you get closer to the object, the total power increases because the object’s angular size increases, obviously up until the point at which it fills your entire field of view.

            When the moon is closer, it appears bigger, and therefore provides more total light, but it does not actually have a bigger surface brightness.

            “An astrophysics guy should have this in the blood.”

            Well, I mean, he’s not a particularly *good* one, and his work is solely in communicating to the public, so using a “common public” idea of brightness (if the Moon is bigger, provides more light, therefore “brighter”) makes sense. A supermoon, for instance, absolutely washes out more stars in the sky (just like a full moon washes out stars more than a crescent moon).

            That being said, he’s incredibly arrogant and condescending, so, y’know, nitpicking him’s fine by me. But that’s not his *worst* physics mistake, by far. Worst one’s gotta be thinking that an unpowered helicopter falls, which any kid who’s watched helicoptering seeds knows is utter garbage. That was hilarious.

    3. “the cushion of air behind the vehicle”. When it’s traveling slower than the air downwind, maybe. When it’s traveling faster, that is in the past. Don’t think of air like a hand that pushes it along, but as a jet of fluid it moves through. The craft is pulling itself faster by grabbing some of the air ahead and accelerating it backwards.

      Energetically this works because it is reducing the speed of the air relative to the ground, which it is geared to.

    4. Airspeed/groundspeed confusion is a strong one. Search for “downwind turns” and you’ll find that even airplane _pilots_ get confused by this when they get close enough to the ground, even though they turn just fine when they’re high up in the air.

      But the airplane treadmill one is simply people thinking that airplanes accelerate like cars, by turning wheels against the pavement. Instead, they accelerate by pushing against the air with their jets, and the ground speed is basically irrelevant.

      If you asked, could you take off in a hang glider running on a treadmill, well then.

  5. I am not convinced that the treadmill model is really proof. The fact that the model has to be held in place on the treadmill until enough energy is stored in its moving parts is analogous to pushing the original vehicle into motion until it gets up to speed, perhaps momentarily exceeding the current wind speed. A more convincing proof with the model would be to show that, after having reached the initial startup speed, that the model does indeed continue to move forward on the (extended) treadmill for, in principle, and indefinitely long period of time. Once again, the method of proof is NOT
    well established. I am not saying the concept of using wind to travel faster than wind is not possible, I am only stating that the so called proof is not convincingly rigorous. I go back to my criticism in the original video, that the refusal to do the simple task of repeating the test but with the streamer flag mounted WELL ABOVE (at least twice the height) of the propeller to avoid the effects of the fan action, smells of a lack of scientific rigor.

    1. I’ll probably get roasted for saying this, but lack of scientific rigor as you’ve labeled it doesn’t make him wrong. This is why experimenting on our own, using our own method of scientific rigor that satisfies our own requirements (curiosity?) is so important.

    2. You should watch the video, at some point they show the vehicle with 4 flags, one well above the propeller, two on the sides and last one at the front. What’s more important is the end of the video where they show acceleration of the vehicle consistently increasing while the wind speed is actually slightly decreasing.

      1. What really needs to be done is to gather a proper budget and let proper engineering be let loose on a better full-scale prototype, one that will not be limited by fear of shaking itself apart…it could easily prove that no matter the wind stability, this can go faster. Way faster.

        1. Why? The math behind this is well known. It’s not really useful as a land vehicle after all. It’s like the solar powered car contests: they’re fun engineering challenges, but not practical.

          “Way” faster is an exaggeration, by the way. Powered by a prop you’re limited to around 2.5x wind speed theoretically.

          1. // “Powered by a prop you’re limited to around 2.5x wind speed theoretically.”//

            Nope. Top speed is limited merely by how low you can make the total drivetrain losses and that number is way above 2.5. The Blackbird itself has been recorded still accelerating while over 3x the speed of the wind, and the official world record is ~2.8x. All this with a hand carved propeller and cobbled together bicycle/go kart parts.

            With a good budget multiples of 4-5x are certainly attainable.

    3. The notion that this is analogous to pushing the car forward until it reaches the wind speed is correct, but it doesn’t make the experiment flawed. They just replace the force that the tailwind provides with a stronger force that can accelerate the vehicle to the speed of the treadmill faster. Allowing the vehicle to accelerate on its own would require an extremely long treadmill.

      You are essentially now saying “yes, the experiment demonstrates that it’s possible to go faster than the speed of wind when going downwind, but it doesn’t demonstrate that it is possible to reach the speed of wind when going downwind”. I think it’s generally accepted that it’s possible to reach the speed of wind when going downwind. It’s very easy to demonstrate as well, with e.g. a balloon.

      1. PS The wheels are driving the propeller, not the other way round. The model on the treadmill is therefore an accurate representation of what is happening. When she got the gearing between the wheel diameter and the prop pitch correct it worked. The wide part of the cart is at the front, and the prop is blowing behind it.

    4. Why the heck do we need “scientific rigor”?

      The math and physics here is easy. The equivalent problems on water have been done for decades. These are toy problems, like, I dunno, building a generator out of Legos or something.

      Just measure the drivetrain efficiency and get the prop diameters and such and plug it into a simple simulation. This isn’t like cold fusion or the EM drive where people are proposing something (nonsensically) beyond normal physics.

  6. The obvious way to prove it works would be to put in in a wind tunnel and show it moving towards the fan from a standing start. I’d even let them get a boost from a rear wall to get going, but I bet it would just get pushed back.

    The other way would be to attach a smoke bomb behind the car, if the smoke goes sideways or overtakes the car the experiment fails.

    1. You mean like how you know that siphons cannot work and it’s cheating to first fill them with fluid? Shoot, my car engine won’t run without someone turning it first, so that’s proof engines don’t really work.

  7. NOT proven! All that is happening here is that the (unknown) kinetic energy in the treadmill is being transferred through the models drivetrain to the propeller while the model is restrained. When the model is released, the flywheel energy which is stored in all of its rotating components plus the thrust from the propeller is sufficient to overcome the model’s inertia and the friction between its wheels and the treadmill. If the treadmill had sufficient length, the model would eventually stop and then be carried by the treadmill at the same velocity minus the slight reduction created by the model’s air drag and wheel friction
    I consider that there is a fundamental error in the experimental setup.
    While the experimenters idea of *reversing* the situation by creating a moving surface instead of the airflow maybe valid, this would only hold true for a condition where the moving surface was inputting to the model the same level of energy as it would have done from the expected airflow that the moving surface is setup to represent!
    If the energy was inputted to the model purely by means of a moving airflow, any propeller would only be able to extract a (very optimistic) maximum of 90% of it, the drivetrain and wheel friction would further reduce this energy and so the model definitely will travel downwind but at a lower velocity than the airflow.
    A sailplane can only travel downwind faster than the windspeed by converting altitude to speed, while birds and powered aircraft can also use their power as well. All 3 examples will eventually deplete their stored energy and cease to travel downwind. In fact they’ll just travel down!!
    Mechanical systems cannot exceed parity!

    1. I thought of the fly wheel example myself (which may have some effect in the small scale).

      However, the concept was easier for me to understand when I consider the generic problem from a different angle:
      the air and ground have a difference in velocity, and if you reduce that difference in velocity you can extract energy. (That’s how windmills work, right)

      In the treadmill example, the ground is moving while the air is still. The real life one the air is moving but the ground is still.

      All you have to do is engineer a moving windmill that efficiently transfers KE between the ground and the air such that the velocity difference reduces and you harvest the energy (in this case the energy moves it forward).

      With the treadmill, it’s pushing against the air to maintain it’s position, while it extracts energy from the moving treadmill, so, it actually works the same if a hand is holding it in place or the air is holding it in place.

      The energy always comes from the difference in velocity of the two, it’s just about clever engineering to extract that energy when it seems like you can’t.

      If you had a treadmill and a wooden plank parallel to each other, could you design a gear that would use the treadmill energy to move itself in the opposite direction of the treadmill?

    2. Imagine watching this video, the host of other videos demonstrating the effect, reading the Drela FTTWDDW paper, and still arguing from ignorance.

      Don’t feel bad, you’re in the company of at least one physics professor. But like him, just because you don’t understand something doesn’t mean it’s wrong.

    3. Uh, what? Where is the energy being stored: in the incredibly tiny elasticity of the driving connections?

      Flywheels work because they’re spun up by a force and then *clutched* to spin freely. There’s no clutch here. The prop’s geared to the wheels.

      How would your idea even work? The prop’s speed is directly connected to the wheels. You release, and if stored energy is being released, the prop *slows* and the wheels *slow down*. It wouldn’t accelerate.

      Nothing’s going “over parity” here. The treadmill has effectively infinite energy available (as far as the car is concerned), and the air gives you a constant “wall” to push off of.

  8. As a longtime sailor I can attest that direct downwind sailing is inefficient and only is available about 10% of the time when cruising. In the experiment the blade angles are converting downwind to a reaching wind angle…much more efficient. Still, the friction components of converting vertical to horizontal force makes it impossible to develop a force greater than the originating (wind) force. In other words a perpetual energy machine!

  9. I don’t even know what the debate was about. If the professor had just watched the original video by Veritasium he’d have seen that as the wind tail was pointing backward, he also rode right past a guy holding a wind sock pointing in the opposite direction, and it was a larger windsock so it clearly wasn’t suddenly NOT windy as it was being held fairly steadily. That’s what 100% sold it for me at the time of that video. The follow-up did include a better description though. I’m going to attempt to use this type of tech for my boat to see if I can move faster than the wind with it.

  10. I am from New Zealand.
    We made an America’s cup Mono-hull which sailed faster than the wind 30 years ago.

    Then we made radical foiling boats. They can sail 3-4 times faster than the wind up and downwind.
    We experimented with hard sails which flexed in the middle and achieved speeds faster than the wind.

    Scientists who desire to be known as scientist should be made to read every PhD ever created. That will stop the do overs of discoveries already known and in the book.

    1. Yes, the ability of these boats to sail faster than the wind were covered in Vertasium’s first video about this. The difficulty is in mapping the physics of those boats to the land vehicle. Dr. Kusenko seemed aware of the prior work but one of his arguments against was that the equation equating the power taking from the wheel to the power exerted by the propeller seems to “blow up” or approach infinity as the vehicle approaches the speed of the wind. Unfortunately, that equation was an approximation and the more complete equation doesn’t have that singularity.

    2. They cannot sail faster than the wind on a directly downwind course and at no time can the portion of their speed directly downwind be faster than the wind. The cross-wind velocity component can do so, not the downwind velocity component.

      1. Yes it can! Of course it can. How else would a boat sail upwind at an angle? It’s the exact same question.

        Suppose I’m sailing downwind at an angle, and the downwind component of my velocity is wind speed. So I have velocity (vw, vp). My velocity generates effective wind (-vw, -vp) (but not in my frame, for simplicity) and adding real wind (vw, 0) apparent wind is (0, vp) – again, not in my frame, for simplicity.

        This means, to me, the wind direction’s coming at an angle toward me: I’m sailing into the wind at an angle. Can I speed up in the +x direction? Sure: turn the sail to generate mostly lift. Apparent wind is in the +y direction, so lift is in +/-x (depending on sail angle) direction, since sail lift is perpendicular to wind.

  11. It’s very simple, the propeller blades are not normal to the wind direction, therefore the wind flows around the surface of the blade generating a lifting force.[1] The lifting force does work when it is used to move the vehicle.[2] A relatively slow wind can move a passive winged object many times faster than the wind itself. A dramatic example of this is Dynamic Soaring.[3]

    * References:

    1. Lift (force)

    https://en.wikipedia.org/wiki/Lift_(force)

    2. Work (physics)

    https://en.wikipedia.org/wiki/Work_(physics)

    3. Dynamic Soaring (Video)

    https://hackaday.com/2020/09/24/dynamic-soaring-545-mph-rc-planes-have-no-motor/#comment-6281089

    1. There’s no lifting force needed. It’s just the simple fact that it’s easier to exert a force on a slower moving fluid than a faster one. This is just fundamental: power is force times velocity.

      If there was no wind, using the wheels to power a prop would do nothing: energy in, energy out. You’d just lose out due to inefficiency.

      With wind, the fan can exert more force for equal power than the load on the wheels.

      Doesn’t have to be a prop. Could be anything that pushes air. But props are pretty efficient and easy to build, making it practical. But there are plenty of “thought experiment” options that will work with “sails” purely perpendicular to wind flow. They’re just not practical due to inefficiency.

      1. Also, to be clear, dynamic soaring is different than this. Dynamic soaring is actually how particles are accelerated to massive energies at a shock: the glider gains energy at the boundary, then losslessly turns around (in shock acceleration, this is caused by a magnetic field) and re-encounters the shock.

        The difference is that you don’t need two different surfaces to extract energy from. The “still air” portion of dynamic soaring is unimportant. You could do it in space, for instance, using solar wind and an occulting planet or something.

  12. I remember the disagreement between Mehdi Sadaghdar (aka ElectroBOOM) and Dr. Walter Lewin. I would have loved to see a video by Dr. Lewin about the mistakes Mehdi made in his video. But instead Dr. Lewin acted super arrogantly and made fun of Mehdi.

  13. Pat said: “There’s no lifting force needed.” Well, there MUST be force, without force there can be no work provided the target is displaced (which is the whole purpose of the exercise). As for “lifting” we’re talking about aerodynamic physics here – the norm is to refer to aerodynamic force generated by an airfoil as “lift”, but not always as in a vector relative to the surface of a planet, like Earth.

    “It’s just the simple fact that it’s easier to exert a force on a slower moving fluid than a faster one.” The airfoil does not exert a force on the fluid, it’s the opposite way-’round, the fluid exerts a force on the airfoil. But it’s OK if you choose to change your frame of reference. After all, Electronic Engineers go through their whole lives thinking charge goes the wrong way around an electrical circuit ;-) Also, it’s not whether the fluid is moving faster or slower, it’s the density of the fluid which is controlled by it’s interaction with the airfoil which matters, and that depends on a lot more than just simple speed, because the fluid (air) is compressible.

    “If there was no wind, using the wheels to power a prop would do nothing: energy in, energy out. You’d just lose out due to inefficiency.” You’re stating the obvious. 2nd Law of Thermodynamics, Entropy if you will – or more simply put you can’t get something for nothing.

    “With wind, the fan can exert more force for equal power than the load on the wheels.” Again, stating the obvious. This has nothing to do with what the force of the wind does to the wheels versus the propeller blades. Such an analysis would be useless.

    “Doesn’t have to be a prop. Could be anything that pushes air. But props are pretty efficient and easy to build, making it practical. But there are plenty of “thought experiment” options that will work with “sails” purely perpendicular to wind flow. They’re just not practical due to inefficiency.” Obviously. In this case my description is general, a surface that is not NORMAL to the fluid flow. As a rough order of magnitude, a well designed propeller at STP might be 70% efficient – roughly depending on wind speed.

    “Also, to be clear, dynamic soaring is different than this. Dynamic soaring is actually how particles are accelerated to massive energies at a shock: the glider gains energy at the boundary, then losslessly [sic] turns around (in shock acceleration, this is caused by a magnetic field) and re-encounters the shock. The difference is that you don’t need two different surfaces to extract energy from. The “still air” portion of dynamic soaring is unimportant. You could do it in space, for instance, using solar wind and an occulting [sic] planet or something.”

    You took the bait – hook, line, and sinker. I threw Dynamic Soaring in there to make a point. The idiotic click-baity “can something blown by the wind go faster than the wind” of the video is just plain insulting to any clear thinking person. Then to hear the dumb interpretations much less the lame attempts at a free-body diagram and mathematical explanations just makes it worse. So yeah, I’m put Dynamic Sailing in there as tongue-in-cheek example. I’m surprised the video producers didn’t get to it first.

    1. Yep, this post (above) was a reply to a thread with participant @Pat from above. But unfortunately, WordPress Jetpack has decided this post belongs at the BOTTOM of the entire thread stack. Dumb. Please HaD – GET RID OF Jetpack!

    2. Sails/airfoils have lift (perpendicular) and drag (parallel). You don’t need lift here, you could do it with just drag, if it wasn’t for mechanical inefficiencies (props are very efficient).

      The inefficiencies here are *not* “energy conservation” issues. They’re mechanical issues. If I put a 10 N load on a ground surface moving at 10 m/s, there’s 100W of power. Some of that goes to heat, some goes to sound. I then have 100W of power to generate more than 10N of thrust. If the air is moving at 1m/s in my frame, fundamentally exerting 10N of thrust takes 10W of power, minimum. I don’t care *how* that thrust is generated, lift or drag. Doesn’t matter.

      Also, last time I checked Newton’s third law (A exerts force on B means B exerts equal force on A), I have no idea why you think “occulting” is spelled wrong, and I have *no* idea why you would think I’m an EE.

    1. “winning would mean he defied newtons law of equal and opposite reaction”

      No, it really, really wouldn’t. Call the force exerted by the wheels onto the ground (and ground onto the wheels) Fd. Since the ground is moving at v, the power corresponding to that is Fd*v. Call the force exerted by the prop on the air (and air on the prop) Ft. Transfer all the power, times some efficiency, from the wheels to the prop. Since the air is moving at (v-w), and the power input is e*Fd*v, where e is the overall efficiency, the thrust force is Ft = Fd*e*(v/v-w).

      So long as e*(v/v-w) is greater than 1, the vehicle will accelerate.

      I have *no* idea why this seems at all surprising. You have an “infinite” source of energy in the relative speed difference between the two media, and you can push off of either one.

  14. Thanks still have trouble with this experiment. If this truly worked as proposed you do the following: take this cart in still air, spin up the wheel s so it travels at designed speed, then remove power and it will accelerate to faster speeds. If that’s the case then we don’t need gas or electric engines to power our cars. What is not accounted for in this experiment is the fact that we have a moving treadmill belt traveling against a stationary air mass. Against the belt the air mass is traveling at the speed of the belt. This air motion transfers energy into the surrounding air mass by viscous means. I propose that the prop is driven by the velocity gradient between the belt and stationary air. Incidentally this was proposed by the professor as well. To prevent this situation you’d need to isolate the prop from this gradient. It is easily done by suspending the cart loosely within a frame by loose strings. On the top of the frame would be a plate that is parallel to the belt and with a slot in it for the prop drive shaft. If this drives forward when placed on the treadmill then you’ve proven this concept. But if it drags backwards you’ve proven a shear gradient in the air above the belt is transferring energy into the prop.

    1. It is because the “experiments” only show half the system. What is missing is the part that occurs on the vehicle from rest. In that part the propeller acts as a sail and is only turned into a propeller/fan when sufficient ground speed is attained so the torque from the wheels to the fan can overcome the “reverse torque” from the pushing of the wind on its blades. This is achieved by a variable pitch propeller and gear ratios. It isn’t hard to understand but this mechanism is never well explained, to add to the “mystery”, I suppose.

  15. > take this cart in still air,

    It doesn’t work without wind relative to the ground. You need groundspeed > airspeed to produce thrust > wheel_drag. Look at the left screenshot from the video under the title above. The math is right there.

  16. It is not as it seems because the “propeller” changes pitch as required to transform itself from being a sail into a fan depending on ground speed. The small models on the treadmill are not really related to the concept in any way except to show that wheels can power a fan to push an object – this is only half of what the actual Blackbird does.

    The other half is accelerating up to the point so the system can work. The concept is quite simple but it absolutely depends on the wind not being able to turn the propeller the opposite way when the vehicle is at rest and at low speeds – lower than the “critical speed” is attained and the fan can begin to push the vehicle faster from the power from the wheels, and that is achieved by gear ratios and changing the pitch on the propeller. The maths doesn’t really account for this complicated change of system behaviour.

    As soon as i understood this then it makes perfect sense that the vehicle can travel the way that it does, but this part is not fully explained.

    1. >> the “propeller” changes pitch as required to transform itself from being a sail into a fan depending on ground speed.

      Nope. The propeller always acts as a propeller – never a turbine.

      >> The small models on the treadmill are not really related to the concept in any way

      The small models on the treadmill are in fact an instance of DDWFTTW and operate on exactly the same principle as the Blackbird.

      >> and that is achieved by gear ratios and changing the pitch on the propeller.

      There is no need to change gears or prop pitch. We have NEVER changed gears during a run, and we did not have a variable pitch propeller for our early tests. In all cases the vehicle would self start, exceed wind speed, and continue until the wind died or we hit the brakes.

      >> The maths doesn’t really account for this complicated change of system behaviour.

      That’s because there is no change in system behavior.

Leave a Reply to Edward NardellaCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.