Remember the debate about downwind vehicles

[spork]

We've had a couple of heated debates here about whether one could build a wind powered vehicle that would go directly downwind faster than the wind. So we built one...

First run: http://www.vimeo.com/10476453
Second run: http://www.vimeo.com/10477373
Last run: http://www.vimeo.com/10476216

 

[mender]

Hopefully, large scale physical evidence of DDWFTTW travel will help moderate some of the extreme anti invective and allow a calmer discussion of this fascinating phenomena.

 

[spork]

Hopefully, large scale physical evidence of DDWFTTW travel will help moderate some of the extreme anti invective and allow a calmer discussion of this fascinating phenomena.

Actually, the resulting discussion goes beyond calm - all way to non-existent.

 

[russ_watters]

...which then begs the question: where does the discussion go next?

 

[spork]

...which then begs the question: where does the discussion go next?

Ratification of an official DDWFTTW speed record (stay tuned), and then perhaps the overall land speed sailing record(?)

 

[rcgldr]

Perhaps the overall land speed sailing record(?)

DDWFTTW vehicles require a ground force that is backwards relative to the direction of travel (wrt ground) in order to provide the power to drive the prop. However the ground force on a sailcraft is perpendicular it's direction of travel (wrt ground).

Perhaps you mean the component of speed directly downwind? DDWFTTW vehicles require all of the ground force to be upwind, while sailcraft only require a fraction of the ground force to be upwind, depending on the sailcraft's heading. I'm pretty sure that ice boats will have a faster component of speed directly downwind, but land (sand) craft have more drag and I'm not sure.

 

[spork]

DDWFTTW vehicles require a ground force that is backwards relative to the direction of travel (wrt ground) in order to provide the power to drive the prop. However the ground force on a sailcraft is perpendicular it's direction of travel (wrt ground).

Perhaps you mean the component of speed directly downwind? DDWFTTW vehicles require all of the ground force to be upwind, while sailcraft only require a fraction of the ground force to be upwind, depending on the sailcraft's heading. I'm pretty sure that ice boats will have a faster component of speed directly downwind, but land (sand) craft have more drag and I'm not sure.

What you're saying is correct of course, but the numbers suggest that a DDWFTTW vehicle can take the existing overall land sailing speed record (not just the downwind component). Keep in mind the DDWFTTW vehicle gets a better break on aero drag too.

Greenbird is going after the ice record now. So I don't know what numbers they'll achieve. I'm pretty sure we're not interested in the ice record any time in the near future.

 

[rcgldr]

Greenbird is going after the ice record now.

Aren't they already in the 5x to 7x wind speed range? I recall something about 10 mph winds being ideal because the ice boats are designed with that in mind, achieving speeds of 50 mph to 70 mph? Beyond that and drag becomes an issue, so achieving the same ratio with a 15mph wind requires a different design (larger?), and the locations where they can use the ice boats are limited in size.

 

[spork]

Aren't they already in the 5x to 7x wind speed range? I recall something about 10 mph winds being ideal because the ice boats are designed with that in mind, achieving speeds of 50 mph to 70 mph? Beyond that and drag becomes an issue, so achieving the same ratio with a 15mph wind requires a different design (larger?), and the locations where they can use the ice boats are limited in size.

I haven't followed the Greenbird really. I believe they set the land record at 126 mph in winds that peaked at 47 mph. I believe they tried to outfit it for the ice, and found out they needed to basically start over from the ground (ice) up. But I haven't looked at the wind or grounds speeds for ice records.

 

[rcgldr]

I found an article posted at several websites. Apparently a Skeeter can achive 6x wind speed in light wind, less than the 10 mph I thought it was. In a chart of a 2 lap race, top speed was about 65 mph with 15 mph wind, but the goal there was to run a 2 lap course as opposed to a speed run.

The fastest wind powered devices are probably radio control gliders while dynamic soaring. At youtube there's a video showing a run of 375 mph, and later that day they recorded a speed of 392 mph, but didn't video the run, just the max speed on the radar gun.

http://www.youtube.com/watch?v=WaQB16ZaNI4&fmt=18

update - make that 399 mph as of 2009-12-22, same model, same location:

http://vimeo.com/8356047

 

[spork]

I found an article posted at several websites. Apparently a Skeeter can achive 6x wind speed in light wind, less than the 10 mph I thought it was.

Is that the NALSA article? I always thought it was more like 4x-5x

 

[rcgldr]

Is that the NALSA article?

Yes.

Skeeter ... It can sail about 6 times the wind speed in light winds and tops out a little over 60 mph.

nalsa_article.pdf

I think that the DDWFTTW carts are a bit more surprising and impressive that they could work at all; outrunning the wind that propels them.

 

[spork]

Interesting. I bet you found that in places that JB or I posted references to it. I never noticed that line. I've been saying 4X - 5X for a long time.

Interestingly, that article was written by Bob Dill. He's on the NALSA B.O.D. and was our contact and host at the Ivanpah event. We're working with him now to ratify a category to set a downwind record in.

 

[Subductionzon]

So where are all of the deniers to this topic? Now that spork has extremely strong video evidence of the cart going faster than the wind have they all run away? Check out the last run where the cart drives through a cloud of dust that is being blown by the wind. I think schroeder especially owes spork at least an "I'm sorry".

 

[DJM111188]

Spork, at your request, I have posted my questions below. Anyone with insight please feel free to chime in. Also I have now a better understanding of what the term perpetual motion means. At the time when I contacted you I hadn't taken into account that A)the wind is energy. But now I have revised the original question.

Hello I am not a frequenter of online forums but I recently was on the site and came across the "DWFTTW" machine and I felt the need to contact you. I have watched some of your videos on youtube. One of those was where you raced the turbine shaft vs prop shaft carts. I was wondering if you could clear some things up for me.

1) Are both carts using a 1:1 drive?
2) In the video you said the turbine shaft cart accelerated much quicker then the prop shaft cart correct? If this is true then would changing the gear ratio allow the turbine cart to have a much faster speed?
3) If your answer to number 2 is "Yes", then would the turbine cart be geared high enough to where it could keep up with the prop car? Or is the only reason a vehicle is able to travel faster down wind then the wind itself because of the wheels powering the prop?


I am just trying to learn so please take no offense to my inquiries. I would love to know how exactly this works. It seems to me that if 20mph of wind is applied to the turbine cart then it will never ever be able to reach a faster speed, even taking into account a gear ratio change, because when it would get to the 20mph threshold there would be no more force to turn the wheels. Am I correct in this assumption?

Also can you explain to me the end result in my hypothetical test? Say you had one of the carts that had a steering servo hooked up to the front wheels and you were able to find a stretch of 5 miles with absolutely flat and even surface. Let's say wind-speed could be controlled at exactly 20 mph. If you were to put the cart down and get it to travel the whole 5 miles directly downwind in a striaght line, is there any way to determine how fast it could actually go? More specifically would it have some magic number as a threshold top speed or would it continue to gain speed indefinitely as long as it had the room? If this description is too vague please notify and I will try to clear it up as best as possible.

Thanks for your time.

 

[spork]

Spork, at your request, I have posted my questions below.

Thanks. I'll do my best to answer your questions.

One of those was where you raced the turbine shaft vs prop shaft carts. I was wondering if you could clear some things up for me.

1) Are both carts using a 1:1 drive?
2) In the video you said the turbine shaft cart accelerated much quicker then the prop shaft cart correct? If this is true then would changing the gear ratio allow the turbine cart to have a much faster speed?
3) If your answer to number 2 is "Yes", then would the turbine cart be geared high enough to where it could keep up with the prop car? Or is the only reason a vehicle is able to travel faster down wind then the wind itself because of the wheels powering the prop?

1) Those particular carts don't use a 1:1 drive. I don't recall the gear ratio off-hand, but I think it was 13:16 in both cases. Of course the gear ratio is only part of the story. To get the "vehicle speed ratio" (speed through air divided by speed over ground) you also have to know the wheel diameter and prop pitch. The difference between the two carts isn't the gear ratio per-se, but the sign of the gear ratio. The gear ratio for the turbine cart is -13:16. So the turbine turns the opposite direction from the prop of the prop cart.

2) No, the turbine cart can never reach wind speed. It uses the tail wind to turn the turbine, which in turn turns the wheels. If it reached wind speed it would no longer feel any relative wind, and would thus have no torque at all on the turbine.

3) There are several approaches to traveling directly downwind faster than the wind. But all require that the vehicle exploit the energy of the wind relative to the surface - not relative to the vehicle. The BUFC does this by gearing the prop to the wheels such that the wheels do the work to turn the prop. A turbine cart can never do this.

I am just trying to learn so please take no offense to my inquiries.

On the contrary. This is what I'm here for. I enjoy these kind of questions.

I would love to know how exactly this works. It seems to me that if 20mph of wind is applied to the turbine cart then it will never ever be able to reach a faster speed, even taking into account a gear ratio change, because when it would get to the 20mph threshold there would be no more force to turn the wheels. Am I correct in this assumption?

You're exactly right.

Also can you explain to me the end result in my hypothetical test? Say you had one of the carts that had a steering servo hooked up to the front wheels and you were able to find a stretch of 5 miles with absolutely flat and even surface. Let's say wind-speed could be controlled at exactly 20 mph. If you were to put the cart down and get it to travel the whole 5 miles directly downwind in a striaght line, is there any way to determine how fast it could actually go? More specifically would it have some magic number as a threshold top speed or would it continue to gain speed indefinitely as long as it had the room? If this description is too vague please notify and I will try to clear it up as best as possible.

Not too vague at all. In fact it's quite well posed.

For a given VSR (vehicle speed ratio) the cart will have a theoretical top speed as a function of wind speed. For example, if the cart is trying to advance through the air at 1/2 the speed that it advances over the ground, it will never go faster than twice wind speed. As you change the prop-pitch and/or gear ratio such that the geometric VSR becomes closer and closer to 1.0, you'll get a higher theoretical multiple of wind speed - but at a price. To achieve more than about 4 times wind speed requires a cart with ludicrously efficient transmission, prop, etc. That being said, there is no theoretical bound to the vehicle's speed as a multiple of wind speed.

 

[ThinAirDesign]

I think that the DDWFTTW carts are a bit more surprising and impressive that they could work at all; ...

Then you will be *really* surprised when we take Greenbirds record of 126mph and smash it in a DDWFTTW vehicle.

JB

 

[rcgldr]

I saw the youtube discovery channel video and they missed the key point of the DDWFTTW carts, which is they rely on the difference between wind speed and ground speed, by using the wheels to drive a prop to generate an upwind thrust that only needs to be somewhat greater than the apparent headwind speed in order for the cart to work. Using an 2x example, with a 15 mph tailwind, and 30 mph cart speed, the prop thrust speed might only need to be 18 mph in order to compensate for all the opposing forces, only 60% of the ground speed, allowing the thrust to be 1.6 times as great as the opposing ground force if there were no losses.

 

[robinson]

This is fantastic!

 

[spork]

This is fantastic!

Thanks. We did some testing yesterday with the vehicle in its final configuration (finally!). We got to 2.85X wind speed. I think Pi is in sight.

 

[rcgldr]

When trying to optimize for speed, how do you choose between pitch and gearing? I'm assuming diameter is fixed since it's just one prop. The maximum speed is related to the overall reduction factor from ground speed to prop wash speed, but the forces involved are related to the prop thrust and opposing ground force used to drive the prop, and I assume that any force greater than the minimum required to overcome drag and rolling resistance increase losses in the drivetrain, but I don't know if those losses are linear.

Also assuming the prop doesn't have a one way clutch, how diffcult is it to slow the cart down via the brakes (how much of the momentum is in the prop)?

 

[spork]

When trying to optimize for speed, how do you choose between pitch and gearing? I'm assuming diameter is fixed since it's just one prop. The maximum speed is related to the overall reduction factor from ground speed to prop wash speed, but the forces involved are related to the prop thrust and opposing ground force used to drive the prop, and I assume that any force greater than the minimum required to overcome drag and rolling resistance increase losses in the drivetrain, but I don't know if those losses are linear.

Also assuming the prop doesn't have a one way clutch, how diffcult is it to slow the cart down via the brakes (how much of the momentum is in the prop)?

To make a very long story short, we do two things... we make research what kind of numbers we think we're capable of in terms of rolling resistance, prop efficiency, aero drag, etc. With those numbers we simulate the system and come up with the optimal parameters (pitch, gearing, etc.). From there we can vary the prop pitch on the fly to optimize a bit further.

If this were a high budget project with a real purpose, we'd re-optimize based on the real-world data, perhaps make a 2nd, 3rd, and 4th prop, change gear ratios, and even find the optimal parameters for any given wind speed.

As it happens, our initial parameter guesses were close enough to allow us to achieve better than 2.5X wind speed. So just adjusting prop pitch (which doesn't adjust TRUE prop pitch) is enough to satisfy us. When someone else breaks our record by a few percent, the next guys are going to have to do all that stuff to beat theirs - and we'll be long gone onto the next silly project.

Our prop does have a freewheel mechanism, but it's still incredibly effective in braking. It's the same as a helicopter in autorotation. No torque is put into the shaft, but it still slows the heli's descent drastically.

ETA: The momentum in the prop is a TINY fraction of the overall momentum in the cart. When we run the prop on the dyno, it slows down very rapidly when we stop powering it. It's trying to push a LOT of air.

 

[rcgldr]

rolling resistance, drag

Seems to me you could remove the prop, then get the cart up to speed using a vehicle tow, then measure the rate of decelertation to get an estimate of these factors.

The momentum in the prop is a TINY fraction of the overall momentum in the cart. When we run the prop on the dyno, it slows down very rapidly when we stop powering it. It's trying to push a LOT of air.

I was mostly wondering about the angular momentum versus the torque required to produce the thrust. Apparently the torque required to drive the prop and generate the thrust is relatively larger compared to the angular momentum of the prop. The prop must be relatively light for it's size.

helicopter auto-rotate

This is the opposite effect of what you'd want for braking in the DDWFTTW cart. Auto-rotation relies on a relatively high amount of angular momentum versus the opposing torque involved during the landing sequence under autorotation (prior to landing, negative pitch and/or forward speed during decent sequence is used to maintain rotor speed).

 

[spork]

Seems to me you could remove the prop, then get the cart up to speed using a vehicle tow, then measure the rate of decelertation to get an estimate of these factors.

Perhaps, but we've found it's a lot tougher to get accurate data than we'd hoped. Add to that the fact that some of the items are constant, some linear with speed, and some squared with speed, it might be pretty tough to tease out the various components. Then there's the dependency to changing losses when loaded (e.g. transmission). We've found that for the level we're playing for the moment, it's just as well to test for the actual results.

I was mostly wondering about the angular momentum versus the torque required to produce the thrust. Apparently the torque required to drive the prop and generate the thrust is relatively larger compared to the angular momentum of the prop. The prop must be relatively light for it's size.

I'm not sure how to compare angular momentum and torque other than to say the prop certainly doesn't freewheel for long - at least not when it has pitch. I would say it's definitely light for it's size - unlike the rest of the cart.

This is the opposite effect of what you'd want for braking in the DDWFTTW cart. Auto-rotation relies on a relatively high amount of angular momentum versus the opposing torque involved during the landing sequence under autorotation (prior to landing, negative pitch and/or forward speed during decent sequence is used to maintain rotor speed).

Auto rotation doesn't actually require any momentum at all - the final flare does however. In steady-state auto rotation the "gliding" blades slow your descent tremendously with no reliance on their mass. It'd be a bummer to land an auto without flaring, but you'd walk away (most likely).

 

[A.T.]

NALSA confirms 2.8 wind speed DDW

From http://www.nalsa.org/:

Get out your slide rules and physics textbooks...On July 2, 2010 on El Mirage Dry Lake, Blackbird sailed directly down wind at a speed of 27.7 mph in a 10 mph wind to set a first record for the ratio of Boat Speed to true wind speed of 2.8. BlackBird was designed and built by the Thin Air Designs team (Rick Cavallaro and John Borton) and sailed by Rick. Links to follow soon.

 

[spork]

 

[Curl]

Can anyone explain how this works? I've never heard about this before.

 

[rcgldr]

Can anyone explain how this works? I've never heard about this before.

The cart just like any sail craft, takes advantage of the fact that with a tailwind (or any wind) the difference between wind speed and ground speed for a sailcraft is constant, and independent of the sailcrafts speed. (If there's a 10 mph wind, then wind to ground speed is always 10 mph, regardless if a sailcraft is going 5 mph or 15 mph).

The cart is designed so that the wheels drive the prop, at some reduced ratio, so that relative to the cart, the prop thrust speed will be less than the ground speed. This effective reduction in gearing effect allows the opposing force from the ground driving the wheels, to be increased by the effective gearing ratio (minus losses) to allow a greater force from the prop thrust (but at a lower speed). This net greater force from prop thrust versus opposing wheel thrust allows the cart to accelerate until the net force is countered by opposing forces related to rolling resitance and aerodynamic drag. How much less the prop thrust speed needs to be depends on how effecient the entire cart is.

At start up, the prop just acts as a bluff body, a small sail, but it's enough to get the cart going. At some point, the prop wash starts to act as a bluff body, but one that is moving backwards away from the cart. This allows the cart to go faster than the tail wind, as long as the prop thrust direction relative to the ground is upwind.

The source of this power is the same as any wind powered device, the prop thrust slows down the tail wind, extracting the energy related to the change in momentum of the wind per unit time.

 

[kjl]

Can anyone explain how this works? I've never heard about this before.

Or, for a more visceral demonstration that may or may not give you that "aha" moment, see:



Pretend that the long toothed rod is "wind", and then the toothedRod->gear connection is basically just a 100% efficient prop. The vehicle moves faster than toothed rod in direction of travel.

Oh, and congratulations, Spork - I followed your crusade to not get shouted and laughed at with great interest :) 2.8X downwind is bad ***!

 

[Curl]

so what was the big debate all about?

 

[spork]

Oh, and congratulations, Spork - I followed your crusade to not get shouted and laughed at with great interest :) 2.8X downwind is bad ***!

Nicely done video. I got a little kick out of the notion that you think it's unfakable. You try and make something foolproof, and I can bring you a better fool *every* time.

so what was the big debate all about?

Physics, and why even some very highly educated people substitute their own faulty intuition and call it physics.

Critical thinking requires being skeptical but not closed-minded, and it requires using your intuition, but not in the place of rigorous math and physics. Learning requires not already knowing everything.

 

[Pythagorean]

Unfortunately, many people equate skepticism with intelligence.

 

[spork]

Unfortunately, many people equate skepticism with intelligence.

There's an element of truth to that though. The real problem is that many people mistake denialism for skepticism.

 

[sophiecentaur]

I want to understand this but my experience and intuition tell me it's wrong. I am not dismissing it out of hand and I would even be pleased to find it is genuine.
The movies are compelling but the explanations are not so compelling. There are statements like:
"But all require that the vehicle exploit the energy of the wind relative to the surface - not relative to the vehicle."
which worry me.
Energy is a scalar quantity and not 'relative' to anything so I can't see what is meant.
Some energy must be involved in turning the prop through the air and moving the cart against friction. If this is to come from the wind then there must be a Force X a Velocity to account for the energy taken from the moving air.

There seem to be two regions of operation. Firstly, at sub-wind speeds, the wind will turn the prop as you'd expect and also be pushing it forwards. Then, once above wind speed, where is the force coming from? Can there be any more 'forward' force if the prop is moving faster than the wind it is moving through?
A simplistic argument would say that the cart, moving faster than the air, would be pushing air ahead of it. That implies that the system just consists of a moving cylinder of air, pushing the cart. The cart wouldn't work in a simple tube - it would just end up going at wind speed.
However, there is a loophole argument, 'for' the system, which I haven't seen used in any of the above arguments. That is the fact that there is an almost infinite amount of moving air around the cart and prop. There is a lot of energy / momentum in this total body of moving air - much more than in the apparent cross section of the craft. I might suggest that the actual airflow through and around the prop gives it an effectively huge cross section. This is not unlike the way a thin wire dipole can gather a lot more RF energy than its thin cross section might suggest.
Have you (the builders / designers) ever considered a wind-tunnel type of experiment in which you put a smoke source at the centre of the prop or you could possibly drive the craft through a smoke plume and observed what is happening to the air around it? I suggest that a large 'cylinder' of air around the cart will be influenced and that this is where the energy is coming from. The prop will be producing a region of turbulence in which some of the slow moving air is, in fact, slowed down and some is sped up and it is that disturbance which produces the required Force times Velocity to power the cart at the high speeds.

Where did the design of the turbine come from? How much aerodynamic theory was involved and how much 'suck it and see'? I would be interested to know.

 

[cjl]

There seem to be two regions of operation. Firstly, at sub-wind speeds, the wind will turn the prop as you'd expect and also be pushing it forwards. Then, once above wind speed, where is the force coming from? Can there be any more 'forward' force if the prop is moving faster than the wind it is moving through?

You seem to misunderstand the method of operation. The prop is not acting like a wind turbine - it is never extracting energy from the air, so to speak. It is acting as a propeller, pulling the cart forwards against the air. The energy source is the wheels - they are rolling against the ground.

Basically, the wheels drive the propeller. This would not work in still air, as you could never gear it in a way which would result in a greater propeller thrust than the wheel drag. However, if the air is moving, the propeller pitch can be reduced such that the propwash is moving slower relative to the cart than the wheelspeed. This allows the propeller thrust to be greater than the wheel drag (assuming minimal frictional losses, so that almost all of the wheel drag is because of the power being extracted to drive the propeller). In still air, this would not move, since the propwash speed relative to the vehicle would be slower than the vehicle airspeed. If the air is moving relative to the ground though, thrust can be generated.

One of the easiest cases to visualize, at least for me, is the case where the vehicle is moving at the speed of the wind. The vehicle's airspeed will be zero, but its groundspeed will be the same as the speed of the wind. In this situation, any propeller will be producing thrust, regardless of its pitch, as the propwash will always be moving backwards (relative to the vehicle), while all of the surrounding air is stationary. All that is required for this to be a sustainable condition then is for the prop thrust to be equal to the wheel drag.

This is doable because the wheels can extract energy from the groundspeed, making the power generated by the wheels equal to the retarding force multiplied by the groundspeed. The power required for the propeller to counter this force is then simply the same force (assuming steady state) multiplied by the propeller's effective airspeed. If you consider this to be the speed of the propwash (a simplification to be sure, but it makes it easy to visualize), then the propeller requires less power to turn than is available from the wheels so long as the propwash speed is lower than the ground speed. With relatively small losses, this same balance works even if the cart is moving faster than the wind. The power required to turn the propeller will go up, as the propwash speed (relative to the cart) must always exceed the airspeed, but the groundspeed will always be higher than the airspeed, so there is always a potentially favorable energy balance.

A simplistic argument would say that the cart, moving faster than the air, would be pushing air ahead of it. That implies that the system just consists of a moving cylinder of air, pushing the cart. The cart wouldn't work in a simple tube - it would just end up going at wind speed.
However, there is a loophole argument, 'for' the system, which I haven't seen used in any of the above arguments. That is the fact that there is an almost infinite amount of moving air around the cart and prop. There is a lot of energy / momentum in this total body of moving air - much more than in the apparent cross section of the craft. I might suggest that the actual airflow through and around the prop gives it an effectively huge cross section. This is not unlike the way a thin wire dipole can gather a lot more RF energy than its thin cross section might suggest.
Have you (the builders / designers) ever considered a wind-tunnel type of experiment in which you put a smoke source at the centre of the prop or you could possibly drive the craft through a smoke plume and observed what is happening to the air around it? I suggest that a large 'cylinder' of air around the cart will be influenced and that this is where the energy is coming from. The prop will be producing a region of turbulence in which some of the slow moving air is, in fact, slowed down and some is sped up and it is that disturbance which produces the required Force times Velocity to power the cart at the high speeds.

Where did the design of the turbine come from? How much aerodynamic theory was involved and how much 'suck it and see'? I would be interested to know.

The airflow around it will be very much like the airflow around any propeller. As the cart moves, it will leave a cylinder of air behind it which is moving slower (relative to the ground) than the surrounding air. That energy gained from slowing the air down (again, in the ground frame) is what drives the cart. It's counterintuitive, but it works.