DDWFTTW Turntable Test: 5 Min Video - Is It Conclusive?

[rcgldr]

If I was designing an outside cart I would definitely include a large sail area to help the cart get up to wind speed. The sail area would be removed at wind speed and beyond

A variable pitch propeller would achieve the same effect, you could start off with an effective pitch of zero (or even negative such as a thrust reversing prop). However if the goal is simply DDWFTTW, instead of maximum downwind speed, then a fixed pitch prop and an advance ratio << 1 will do the job.

Sorry but the cart actually tried to move BOTH torward the fan and away from it. That's why it didn't move at all.

Other than oscillations back and forth while stuck in a static position, the cart should not tend to move upwind. One exception would be if traction at the wheels were lost, and the prop quickly windmilled up to speed, and then traction at the wheels was somewhat regained, then momentum of the prop and wheels would allow the cart to move upwind, but upwind movement can't be sustained with an advance ratio < 1. The cart will follow the wind (tailwind or headwind) because the advance ratio < 1 ties movment of air flow through the prop with the direction the cart is moving. I really think you just need a stronger and/or larger fan to test the self startup case.

 

[swerdna]

A variable pitch propeller would achieve the same effect, you could start off with an effective pitch of zero (or even negative such as a thrust reversing prop). However if the goal is simply DDWFTTW, instead of maximum downwind speed, then a fixed pitch prop and an advance ratio << 1 will do the job.

Other than oscillations back and forth while stuck in a static position, the cart should not tend to move upwind. One exception would be if traction at the wheels were lost, and the prop quickly windmilled up to speed, and then traction at the wheels was somewhat regained, then momentum of the prop and wheels would allow the cart to move upwind, but upwind movement can't be sustained with an advance ratio < 1. The cart will follow the wind (tailwind or headwind) because the advance ratio < 1 ties movment of air flow through the prop with the direction the cart is moving. I really think you just need a stronger and/or larger fan to test the self startup case.

I think reduction gearing from the prop to the wheel would cause the cart to go upwind.

Would a one directional drive from the prop to the wheels improve start up? The prop would freewheel with the force going from the prop to the wheel but would drive with the force going from the wheel to the prop. Hope that makes sense.

 

[rcgldr]

I think reduction gearing from the prop to the wheel would cause the cart to go upwind.

True, if the advance ratio is > 1, you have an upwind cart. I was referring to a variable pitch prop, which would allow the advance ratio to be << 1 at the start, and then just < 1 (probably >= .5) after the cart was moving. Note an advance ratio < 0 is also << 1 and would work for startup; with negative pitch, the windmill direction of the prop would drive the wheels forward, but maximum speed would be < wind speed, this would work for startup, but to achieve DDWFTTW, the advance ratio has to be > 0 and < 1. The closer to 1 for the advance ratio, the higher the maximum speed, but only if there is sufficient efficiency to work with a high advance ratio.

Would a one directional drive from the prop to the wheels improve start up?

Nope, the one directional drive would only let the prop outrun the wheels, this only happens when the headwind is faster than the prop pitch speed, or the wheels are slowing down faster than the prop due to the prop's inertia. It would allow you to brake the cart without having to overcome prop inertia, but the free spinning prop would still be generating forward force as it free spooled down while stopping the cart.

 

[vanesch]

OK. Thank you. That is exactly what I expected. I think it would be safe to say that if we could place this cart out in the wind, with the wind blowing from the back, same direction as the fan and same direction as the apparent wind in your running test, that the cart would go nowhere. It would not go downwind, and it would certainly not go downwind faster than the wind. What we have shown here is how totally erroneous conclusions can be drawn by incorrectly matching reference frames. Do you agree?

No, I don't agree. What happens here is that the fan wind is mainly hitting the propeller, and not the cart or the arm (which acts as a sail). As such, this is as if the wind has now "more importance" on the propeller, and less on the cart which comes down to an effective change of gearing ratio between the wheel motion and the propeller coupling.

The test would be valid if you would put a BIG ventilator there that blows just as much on the arm and the cart as on the propeller. I refer to my post 136 in this thread, where we have that the cart velocity is given by v_B = p/(p-1) v_A. However, that was for a fixed mechanical gearing ratio with two wheels ; the coupling between the air and the propeller is not as rigid and especially, is not of exactly the same nature. It can moreover change "effective gearing ratio" at different speeds.

So what you have is rather a balance between the FORCES 1) between propeller+drag and air on one hand, and 2) between the wheel and the table on the other.

Let us model this crudely. With a velocity (wrt ground) v_cart corresponds:

a force by the air on the cart: F_air = A x v_cart + B x (v_wind-v_cart)
a force on the wheels: F_wheel = - C x v_cart

Positive signs are "downwind". A, B and C are model constants.

The term with A is the propeller acting as a propeller and it is driven by the speed of the cart.
The term with B is the drag of the wind, and also the effect of the difference between wind velocity and propeller.
The term with C is the force excerted through the gearing mechanism of the propeller (reaction from the fact that the wheels drive the propeller).

We get equilibrium when F_air + F_wheel = 0 or when (A - C - B) v_cart + B x v_wind = 0,

or when v_cart = B / (B + C - A) v_wind

Depending on the values of A, B and C (which are working-point dependent) we will find a solution with v_cart positive or negative, and greater or smaller than v_wind.

If A is very large compared to B and C, which means a high gearing ratio, then this is the case schroder has in mind, and the cart will move UPWIND (the propeller will act as a turbine). If A is very small, then the cart will move downwind, but somewhat slower than the wind (sailing cart). But if A is about in the right ballpark, compensating B+C, then v_cart can be larger in absolute value than v_wind, and it can even move upwind or downwind depending on whether A is a bit larger or a bit smaller than B+C.

This is of course a very crude model.

 

[zoobyshoe]

Hey, I just realized something. Considering this:

"The power from differences of speeds of surrounding media can be extracted rather independently of gearbox`s (=vehicle`s) own speed."

And this:

http://de.youtube.com/watch?v=9Yt4zxYuPzI (faster than the ruler)

I have realized that as soon as the ruler is moved at any speed the cart is instantaneously and automatically going faster than the ruler by a fixed ratio. As part of a gear train it is not at liberty to do otherwise.

At any time a non-slipping contact patch is established between the wind and gear # 1 of the gear box (the large wheel on the cart in the ruler video) the cart is automatically in a situation where it must regard all air in its direction of travel as a headwind. As soon as it begins to move it is traveling into a head wind. If we ascribe any speed we like to the ruler V=V the cart must automatically be going NV = faster than the ruler. Therefore, Flossie, the bear in the cart, always experiences the ruler as a wind blowing in her face. If you watch the video you can see this clearly: take Flossie's head as your inertial frame and you see the ruler is traveling toward her face. At any time the ruler moves "downwind" it is automatically a headwind to Flossie.

There is no acceleration period: any motion of the ruler means the cart is moving some multiple of the ruler's speed. It is instantaneously a headwind.

 

[atyy]

I don't see why DDWFTTW is a problem in this case. DDWFTTW is only not possible if you don't have any "reference" other than the wind. But if you have a ground, I don't see why this is a problem - I mean, why should it be impossible to go downwind faster than the wind ?

After all, the velocity difference between wind and ground allow you to extract some energy, and that energy can be used to drive something. If friction and all that are low enough, you can drive that something faster than the wind speed. Hey, you could have a stationary windmill which sends out microwaves, which are captured by an antenna and power an electrical motor of a car which can then drive as fast as it can.

Nice demonstration, BTW.

Is this a correct representation of your thinking?

P=windmill power
M=windmill mass
v=final car speed
m=car mass
k=drag coefficient

P=kv²
v=(P/k)^(1/2)

Centre of mass position = [m/(M+m)]x
Centre of mass velocity = [m/(M+m)]v=[m/(M+m)] (P/k)^(1/2)

We can ignore initial acceleration where car velocity is less than windspeed, as this will be negligible if the car maintains its final speed for sufficient time.

For large enough P/M, the CM velocity will exceed a given windspeed.

 

[swerdna]

I have made a couple of videos to demonstrate what I mean when I say a cart that wouldn’t normally advance against the turntable can be made to do so when it is “forced” up to speed by holding it against the turntable before releasing it.

Video 1 - The Little Cart that Couldn’t -

Video 2 - The little Cart That Could -

 

[rcgldr]

I have made a couple of videos to demonstrate what I mean when I say a cart that wouldn’t normally advance against the turntable can be made to do so when it is “forced” up to speed by holding it against the turntable before releasing it.

Those two videos mostly just prove that static friction is greater than dynamic (sliding) friction.

 

[schroder]

No, I don't agree. What happens here is that the fan wind is mainly hitting the propeller, and not the cart or the arm (which acts as a sail). As such, this is as if the wind has now "more importance" on the propeller, and less on the cart which comes down to an effective change of gearing ratio between the wheel motion and the propeller coupling.

The test would be valid if you would put a BIG ventilator there that blows just as much on the arm and the cart as on the propeller. I refer to my post 136 in this thread, where we have that the cart velocity is given by v_B = p/(p-1) v_A. However, that was for a fixed mechanical gearing ratio with two wheels ; the coupling between the air and the propeller is not as rigid and especially, is not of exactly the same nature. It can moreover change "effective gearing ratio" at different speeds.

The crossarm is part of the test apparatus. Specifically, it provides the centripetal acceleration which enables the test cart to stay on the circular turntable. Unless you plan to keep the crossarm attached to the cart whenever you perform an outdoor test, it is not part of the cart. If this was an “Official” acceptance test being performed before a customer, and you claimed the crossarm was “part of the cart”, you would be laughed out of business. If you need to go to such bizarre lengths to try and prove your point, it is pointless for me to continue with this. I unsubscribe but I am looking forward to the video of the outdoor test, with the cart carrying the crossarm DDWFTTFW! Thanks for an interesting discussion just the same.!

 

[rcgldr]

cart velocity is given by v_B = p/(p-1) v_A.

You've inverted what was being called advance ratio (ar), p = 1/(ar). For a geared situation, you
Vc = Va p/(p-1)
Vc = Va (1/ar)/((1/ar) - 1)
Vc = Va / (1 - ar) (in an ideal 100% efficiency situation)

which gets back to the forumla that spork and I were using with respect to advance ratio, the effective prop pitch speed / cart speed.

The crossarm is part of the test apparatus.

The crossarm isn't needed. The cart can be drag free, with only the prop resisting the air flow. If the advance ratio is sufficiently less than 1, the prop diameter is sufficiently large (to generate sufficient aerodynamic drag to overcome rolling resistance), and there is no slippage of the drive tire(s); the cart will still self start if there's sufficient wind to overcome static friction in the moving components. For a cart at rest in a tailwind, the aerodynamic forward force on the prop of a typical DDWFTTW cart will be greater than the opposing backwards reactive force from the ground due to the forward force at the contact patch at the drive wheel due to the windmilling torque produced on the prop.

Again, with my previous analogy, note that for each revolution, a perfect prop would displace the air 6 inches backwards with respect to the cart, while the cart advanced 10.5 inches. The net result is for each revolution, the prop would tend to move air downwind 4.5 inches while the cart moved downwind 10.5 inches. The cart responds to a tailwind by moving downwind because that's the direction that corresponds with downwind flow through the propeller. You need an advance ratio > 1 in order for the prop to move air downwind while the cart moves upwind.

As a side note, for this cart, the prop diameter is 12 inches and the pitch is 6 inches. At the prop tips, the AOA = atan(6 / (12 pi)) = 9.0 degrees. At 6 inches from the hub, the AOA = atan(6 / (6 pi)) = 17.7 degrees. 75% of the area of the prop exist between 6 inches from the hub and the prop tips with the AOA transitioning from 17.7 degrees down to 9.0 degrees. If the prop was attached to linear and angular strain gauge, you could measure the linear force and torque on the prop, but these can't be directly compared. You need to compare the forward force on the prop against the backwards force from the ground that's opposing the torque induced forward force on the tires. If the advance ratio is sufficiently less than 1, the forward force is greater than the backward force and the cart would accelerate forwards.

 

[schroder]

Yes yes, mathematically, theoretically, if such and such a parameter and on and on. But the bottom line is, when the fan was blowing on the propeller, it tried to turn the other way, the wheels tried to move towards the fan, upwind, not downwind! This is confirmation enough for me that this cart cannot even go downwind at all, let alone faster than the wind. When on the turntable, it is getting all of its driving force from the wheel which in turn gets its driving force from the motor which is turning the table. The demonstration has absolutely Nothing to do with a cart being driven by a tailwind as we have seen a tailwind will not budge the cart! This makes the turntable and treadmill demonstrations totally invalid and cannot be considered as proof of anything! With today’s technology, radar speed guns, windtunnels and all, the ONLY acceptable proof is to let the WIND drive the cart, time it and show conclusively that the cart can exceed wind velocity. The rest of this is just spinning wheels both literally and figuratively. I have wasted all the time I am willing to waste on this. If anyone wishes to believe these turntable or treadmill tests actually prove some thing that is entirely up to them. That is all I have to say on this.

 

[vanesch]

The crossarm is part of the test apparatus. Specifically, it provides the centripetal acceleration which enables the test cart to stay on the circular turntable. Unless you plan to keep the crossarm attached to the cart whenever you perform an outdoor test, it is not part of the cart. If this was an “Official” acceptance test being performed before a customer, and you claimed the crossarm was “part of the cart”, you would be laughed out of business. If you need to go to such bizarre lengths to try and prove your point, it is pointless for me to continue with this. I unsubscribe but I am looking forward to the video of the outdoor test, with the cart carrying the crossarm DDWFTTFW! Thanks for an interesting discussion just the same.!

Of course the cross arm is part of the cart, as it is an aerodynamically sensitive element.

We went in this discussion from "DWFTTW is an over-unity device and hence impossible", to "it is with available materials impossible to have a DWFTTW demonstration of principle" to "the demonstration given by the turntable is not valid by change of reference frame (this one I still didn't get)" to, now: if this is to be a commercial product, you will be laughed out of business.

I limited myself to two things: 1) there is no fundamental principle in classical physics which forbids DWFTTW and 2) even with some caveats, the demonstration given by the OP is rather convincing (the caveat being that the reference frame in which the DWFTTW is demonstrated is not 100% inertial because of the finite curvature radius of the turntable, which means that the reference frame has some rotiation to it, and hence coriolis and centrifugal effects, which might change the result a little bit).

I didn't know that the claim was now to have a viable commercial product that will make a lot of money for the owner of the license :-)

 

[vanesch]

The crossarm isn't needed. The cart can be drag free, with only the prop resisting the air flow.

It isn't needed, but it will change the airodynamic properties of the cart/air link, and hence change the effective "gearing ratio". So one shouldn't be surprised by placing the cart in an aerodynamically different situation, that the behavior is different.

 

[vanesch]

When on the turntable, it is getting all of its driving force from the wheel which in turn gets its driving force from the motor which is turning the table.

But really, what difference is there between a turning turntable and a "static" amount of air, or a static floor and an air flow over it ? This is what I don't understand you are contesting.

If I run along with the turntable, in my reference frame, I DO see a wind, a static floor (the turntable), and a cart that is blown back faster than the wind. So why is this not a valid test ? If you'd do this test in a train, such that the turntable is static with the outside track, in what way is this not a cart having a wind (the moving air in the train, as compared to the track), and being blown faster than the train (hence faster than the wind, which has exactly the speed of the train).

How can you object to this ?

The demonstration has absolutely Nothing to do with a cart being driven by a tailwind as we have seen a tailwind will not budge the cart! This makes the turntable and treadmill demonstrations totally invalid and cannot be considered as proof of anything!

No, the tailwind demonstration is not valid because you had a non-uniform airflow, and a different airodynamic condition.

With today’s technology, radar speed guns, windtunnels and all, the ONLY acceptable proof is to let the WIND drive the cart, time it and show conclusively that the cart can exceed wind velocity.

If that demonstration is given, say, in the Mid-West, can I claim that it is still not a valid test in, say, Europe, because that needs a change of reference frame ?
Is galilean relativity now not valid anymore ? And if the demonstration is given with a red cart, do you consider it still not valid for a blue cart ?

 

[vanesch]

Is this a correct representation of your thinking?

P=windmill power
M=windmill mass
v=final car speed
m=car mass
k=drag coefficient

P=kv²
v=(P/k)^(1/2)

Centre of mass position = [m/(M+m)]x
Centre of mass velocity = [m/(M+m)]v=[m/(M+m)] (P/k)^(1/2)

We can ignore initial acceleration where car velocity is less than windspeed, as this will be negligible if the car maintains its final speed for sufficient time.

For large enough P/M, the CM velocity will exceed a given windspeed.

yes.

 

[rcgldr]

The bottom line is, when the fan was blowing on the propeller, it tried to turn the other way, the wheels tried to move towards the fan.

Agreed, these are the reactions related to the windmill torque at the prop, ultimately the wheels apply a forwards force to the ground and it the ground that reacts with equal and opposite force trying to move the cart upwind. However this backwards force from the ground is opposed by the forward aerodynamic drag on the prop (and cart).

Clearly in this video, the cart had a smooth transtion from stationary to moving forwards due to a tailwind. This doesn't prove that there isn't some sub wind speed terminal velocity, but it does prove that the reaction of this cart to a tailwind was to move downwind.

http://www.youtube.com/watch?v=QTAd891IpRs&fmt=18

The other issue brought up earlier is if the prop is driving the wheels, or if the wheels are driving the prop. The answer is neither, similar to Newtons 3rd law, any torque from the prop ultimately results in a corresponding counter torque from the wheel. If the prop turns clockwise when the wheels move forwards, then the prop normally experiences a counter clockwise torque due to a tailwind and/or the generation of thrust. The counter clockwise torque at the prop is tranfered via the cable ending up as a equal backwards torque at the wheels since the gearing for this cart is 1:1. This backwards torque from the wheels generates a forward force onto the ground which reacts with a backwards force (Newtons 3rd law), also creating a forward torque at the wheels that opposes the backwards torque from the prop (again Newtons 3rd law, plus any angular acceleration versus angular inertia effects). The issue is if the forward force from the air is greater than the backward force from the ground. In the case of sporks cart in the above video, the forward force from the air is greater, because the cart smoothly starts moving forwards.

Again, this only covers the startup condition. Spork's video where the cart is pushed back on a moving treadmill, but recovers and returns to forward motion shows that at near wind speed the carts can transition from below wind speed to above wind speed. However it doesn't prove there isn't some lower speed barrier that can't be crossed. swerdna's second video with the block to hold the cart in place on the turntable and then lifting the block via a fishing pole and line attached to the block, releasing the cart, shows the case where the cart starts up, and continues to accelerate past wind speed, to reach a terminal velocity well above wind speed.

http://www.youtube.com/watch?v=k4owZkoeGAU&fmt=18

 

[A.T.]

This makes the turntable and treadmill demonstrations totally invalid and cannot be considered as proof of anything!

The treadmill demonstrations are perfectly valid, if done properly.

With today’s technology, radar speed guns, windtunnels and all, the ONLY acceptable proof is to let the WIND drive the cart, time it and show conclusively that the cart can exceed wind velocity.

Wind tunnels allow to test objects in a stationary position that go upwind
Treadmills allow to test objects in a stationary position that go downwind

I have wasted all the time I am willing to waste on this.

You could have used all this time to learn about reference frames.

 

[atyy]

vanesch and Jeff Reid, thanks! OK, I am convinced that DDWFTTW is possible in principle.

swerdna, it's fantastic what you've made! and Spork and co. earlier too! I have some questions about the demonstrations.

The first one is not an inertial frame because of the initial angular acceleration of the turntable, before it reaches the final angular velocity. I don't know if this matters. I am not worried about the other noninertiality that vanesch brought up due to the rotation. ().

In the second demonstration, the angular velocity is constant, so that may be more of an inertial frame. But are the wheels rotating when you lift the block, or does the block also prevent the wheels from rotating (ie do we really start only when constant turntable angular velocity is reached)? ()

 

[rcgldr]

In the second demonstration, the angular velocity is constant, so that may be more of an inertial frame. But are the wheels rotating when you lift the block, or does the block also prevent the wheels from rotating?

The wheels aren't rotating. If you cycle through the first part of the video and pause carefully, you can see that the prop remains in the same position, inner blade slightly downwards, as the cart approaches the camera from the right.

 

[vanesch]

The first one is not an inertial frame because of the initial angular acceleration of the turntable, before it reaches the final angular velocity. I don't know if this matters. I am not worried about the other noninertiality that vanesch brought up due to the rotation. ().

Well, of course during speedup it is not an inertial frame, but that doesn't matter, does it. What matters is the steady state after some time, because that simply comes down to specific initial conditions.

I have no idea, but given the non-linear relationship between velocity and force on a propeller, it is not entirely inconceivable that there is more than one stationary solution to the equation of motion (different attractors in phase space if you want to say it in a fancy way). It is not inconceivable that there is a different solution, for instance, upwind, with the same device, if launched upwind initially. I'm not saying this is so, I'm simply pointing out the possibility. It could be tested on the turntable by launching the axis holding the cart faster forward than the turntable, and let it go then.

 

[zoobyshoe]

Jeff: did you see my post #215?

 

[A.T.]

There is no acceleration period: any motion of the ruler means the cart is moving some multiple of the ruler's speed. It is instantaneously a headwind.

Well, in reality there is always an certain acceleration period, due to material deformation and sliding. Especially if you replace the gear & ruler by propeller & air.

 

[swerdna]

Those two videos mostly just prove that static friction is greater than dynamic (sliding) friction.

Yes, and that’s all it was meant to demonstrate. It doesn’t prove that the cart with block and Pikachu can’t travel DDWFTTF (in fact the second video obviously shows it can). It shows the conditions of holding a cart against the motion of a moving surface are different than the conditions of not doing this (that’s it). Didn't the addition of Pikachu indicate it was just a piece of fun?

 

[swerdna]

I have wasted all the time I am willing to waste on this.

Does this mean you have given up on our step by step evaluation process? I would be very disappointed if so. I don't think it's a waste of time and I'm very keen to continue if you are. You don’t have to respond to the posts of others in the process. Please let me know if you wish to continue or not - thanks.

 

[swerdna]

vanesch and Jeff Reid, thanks! OK, I am convinced that DDWFTTW is possible in principle.

swerdna, it's fantastic what you've made! and Spork and co. earlier too! I have some questions about the demonstrations.

The first one is not an inertial frame because of the initial angular acceleration of the turntable, before it reaches the final angular velocity. I don't know if this matters. I am not worried about the other noninertiality that vanesch brought up due to the rotation. ().

In the second demonstration, the angular velocity is constant, so that may be more of an inertial frame. But are the wheels rotating when you lift the block, or does the block also prevent the wheels from rotating (ie do we really start only when constant turntable angular velocity is reached)? ()

The cart is help “stationary” on the turntable and the wheel isn’t rotating (rolling on the surface).

 

[atyy]

Well, of course during speedup it is not an inertial frame, but that doesn't matter, does it. What matters is the steady state after some time, because that simply comes down to specific initial conditions.

Yes, I think one the initial conditions were one of the things swerdna wanted to clarify.

The wheels aren't rotating. If you cycle through the first part of the video and pause carefully, you can see that the prop remains in the same position, inner blade slightly downwards, as the cart approaches the camera from the right.

The cart is help “stationary” on the turntable and the wheel isn’t rotating (rolling on the surface).

OK, looks beautiful to me!

 

[rcgldr]

The power from differences of speeds of surrounding media can be extracted rather independently of gearbox`s (=vehicle`s) own speed.

http://de.youtube.com/watch?v=9Yt4zxYuPzI (faster than the ruler)

There is no acceleration period

Well the ruler doesn't change speed instantly, so the rate of acceleration of the vehicle is just a multiple of the acceleration of the ruler, the same multiple as the difference in distance moved or speed.

advance ratio and its effect

Note the diameter of the upper wheel doesn't matter. The upper wheel rests on the axle of the lower wheels, and the advance ratio is the ratio of the axle diameter / wheel diameter of the lower wheels. It's essentially the same thing as pulling a string that's wound under the axle of a yo-yo, but by placing a wheel on top of the axle, the string could go forward up and back over the bottom axle, then back, up, and forward over the top wheel, so that the string would be pulled from the top of the upper wheel. As mentioned before, the string could be replace by a thin ruler that contacted the bottom surface of the lower wheel axle, or with the second wheel on top the ruler can be placed on top of the uppper wheel. The upper wheels purpose is to simply provide a wheel that rotates in the opposite direction of the inner axle of the lower wheel, so a force applied to the upper surface of the upper wheel is the equivalent of applying the same force to the lower surface of the lower wheel axle.

The device should also help settle the debate about the initial start up reaction of a DDWFTTW cart. For the ruler vehicle, note that as the ruler moves to the right, it applies a right force at the top of the top wheel, creating a clockwise torque on the upper wheel, which applies a force to the left on the top of the lower wheel axles, creating a counter clockwise torque on the lower wheels, which in turn apply a right force to the ground, which responds with a left (backwards) force onto the vehicle. The torque response seems to imply that the vehicle should move left in response to the ruler moving right, but the vehicle's movment is determined by linear forces between ruler and ground, and not by the torques.

Because the advance ratio is < 1 (lower wheel axle diameter smaller than lower wheel diameter), the linear force applied to the ground by the lower wheels is divided while the speed is multiplied. This means that the right linear force from the ruler above is a multiple of the left linear force from the ground below, so the vehicle moves to the right, the top wheel ends up rotating counter clockwise, and the bottom wheels end up rotating clockwise. With an advance ratio < 1, the linear forces cause the vehicle to move in the same direction at the ruler, with the speed of cart = speed of ruler / (1 - advance_ratio), or more general, speed of cart x (1 - advance_ratio) = speed of ruler (to allow for advance ratio == 1).

If the advance ratio was > 1, with the lower wheel axle diameter greater than the lower wheel diameter (this would require the wheels to ride on a rails with the axle protruding below the rails, then the force at the bottom is multiplied and the speed divided, and the vehicle would move backwards, an upwind cart. This would be equivalent to a yo-yo resting on rails, with larger diameter inner "axle", and pulling on a string that unwinds from the bottom of the axle, with the string unwinding to the right as the yo-yo moves left.

If the advance ratio is 0, then the vehicle moves at the same speed as the ruler, the eqivalent of pulling a yo-yo with string that's slipping on the axis of the yo-yo instead of winding up on it.

If the advance ratio is < 0, then the vehicle moves slower than the ruler, the equivalent of pulling the string and unwinding it off the top of the axle of a yo-yo.

If the advance ratio is 1, then the ruler can't move, and the cart can move at any speed (cart speed x 0 = ruler speed), this is the same a string wrapped around a yo-yo where the axle diameter is the same as the yoyo diamter.

In all cases, speed of cart x (1 - advance_ratio) = speed of ruler.

 

[swerdna]

Just want to make it clear that I’m not saying there is anything wrong with using any means to get the cart up to wind speed (as there’s not) as long as it doesn’t store any excess energy in the process. The test isn’t if it’s possible to get a cart up to wind speed, it’s to exceed it sustainably only using energy from the immediate wind.

 

[zoobyshoe]

Well the ruler doesn't change speed instantly, so the rate of acceleration of the vehicle is just a multiple of the acceleration of the ruler, the same multiple as the difference in distance moved or speed.

Well, in reality there is always an certain acceleration period, due to material deformation and sliding. Especially if you replace the gear & ruler by propeller & air.

atyy has seen the problem: the wind in a practical situation is already in motion when you expose your propeller to it. It does not accelerate from 0 for you like the ruler does. There is a shock problem. It is like trying to directly mesh a moving driving gear with a stationary gear train. It's like putting your car in 4th gear, bringing your engine up to cruising speed, then letting go of the clutch. Your engine, powerful as it is, will just stall, unable to suddenly overcome all this inertia.

The models must also stall when exposed to the already moving wind. I wouldn't be surprised to find out that in the cases where people take their models out into a real wind that most of the initial speed of the model represents them simply being blown downwind with their wheels skidding over the pavement. When we see a propeller turning I wouldn't be surprised to find out the wheels of the vehicle are slipping over the pavement providing no positive drive. The only acceleration permitted by the configuration is an undesirable sort of "deformation and sliding".

-----------------------------------

If we keep our focus on this vehicle as one which harvests power from relative motion of surrounding media then we have also to contend with the fact that the relative motion of the surrounding media has two separate and distinct configurations to deal with. In the first the cart sees the wind as a tailwind and the ground as a head wind (provided the cart's in motion at all.) In the second the cart sees both the wind and the ground as headwinds. These two headwinds have energy to harvest, in principle, by virtue of the fact they are moving relative to each other at different speeds (according to MGrandin's description of the situation). However, they are still both headwinds, and will require an engineering solution specific to that situation, which must be different than the solution to the tailwind verses headwind situation.

I think it would be very useful to begin analyzing each design according to how it solves for two separate situations 1.) the Tailwind-Headwind (TH) and 2.) the Headwind-Headwind (HH). This would greatly clarify design strategies and discussions.

Personally I think that designing for the HH situation is the critical problem to solve. I would design for that and let the TH solution be direct blowing (DB) i.e. accept and allow for the startup to consist of the wind simply blowing the whole thing physically downwind. Otherwise you have to design transmissions or rotors that change configuration somehow.

 

[zoobyshoe]

Just want to make it clear that I’m not saying there is anything wrong with using any means to get the cart up to wind speed (as there’s not) as long as it doesn’t store any excess energy in the process. The test isn’t if it’s possible to get a cart up to wind speed, it’s to exceed it sustainably only using energy from the immediate wind.

See my post above: I think the whole design strategy should shift to how to take advantage of the speed difference from two "headwinds".

 

[swerdna]

atyy has seen the problem: the wind in a practical situation is already in motion when you expose your propeller to it. It does not accelerate from 0 for you like the ruler does. There is a shock problem. It is like trying to directly mesh a moving driving gear with a stationary gear train. It's like putting your car in 4th gear, bringing your engine up to cruising speed, then letting go of the clutch. Your engine, powerful as it is, will just stall, unable to suddenly overcome all this inertia.

The models must also stall when exposed to the already moving wind. I wouldn't be surprised to find out that in the cases where people take their models out into a real wind that most of the initial speed of the model represents them simply being blown downwind with their wheels skidding over the pavement. When we see a propeller turning I wouldn't be surprised to find out the wheels of the vehicle are slipping over the pavement providing no positive drive. The only acceleration permitted by the configuration is an undesirable sort of "deformation and sliding".

I made this video to show a cart stationary to the moving surface starting in a wind - http://www.youtube.com/watch?v=k4owZkoeGAU&fmt=18

There are other videos that clearly show this as well. But even if there is a “problem” with starting a cart in a wind so what? Simply don’t start it in a wind. Push it up to wind speed and let it go. "Problem" solved!

 

[zoobyshoe]

The wheel is driving the propeller. The thrust speed from the propeller is slower than the speed of the wheel. We've been calling (effective thrust speed / wheel speed) advance ratio, and it needs to be < 1 for a downwind cart, > 1 for an upwind cart.

This needs clarification in my mind. The "We've been calling..." means you have adapted this term from aeronautics to apply to the DDW vehicles? (Who does "we" refer to?)

This dictionary seems to be defining advance ratio differently:

advance ratio
n (Aeronautics)
1 the ratio of wind speed along the axis of a rotor or propeller to the speed of the blade tip
2 the ratio of forward flight speed to the speed of the rotor tip of a helicopter

http://dictionary.reverso.net/english-definitions/advance ratio

 

[zoobyshoe]

I made this video to show a cart stationary to the moving surface starting in a wind - http://www.youtube.com/watch?v=k4owZkoeGAU&fmt=18

There are other videos that clearly show this as well. But even if there is a “problem” with starting a cart in a wind so what? Simply don’t start it in a wind. Push it up to wind speed and let it go. "Problem" solved!

I'm confused by your description of the situation you offer as a solution. "Simply don’t start it in a wind." If there is no relative motion between your two "surrounding media" (air and ground) then there is no energy for the cart to harvest.

I have dial up (exceptionally primitive, I know) and am waiting for your video to load.

 

[rcgldr]

I'm not sure how to mathematically model a propeller or any thrust generating device and the air flow involved for a DDWFTTW cart. A mathematical model for a DDWFTTW prop would include prop pitch, prop diameter, prop angular speed and prop linear speed with respect to the air, which goes beyond the simple advance ratio concept. Ultimately, it's how much the prop slows down a tailwind versus the forward speed of a DDWFTTW cart (which the rate of rotation of the prop is tied to).

In a tailwind, at startup the issue is if the prop acts more like a bluff body than it acts like a windmill generaing a backwards torque on the connnected wheels, or more accurately, if the forward force due to aerodynamic drag is greater than the backward force from the ground related to the windmill torque effect of the prop.

Even with a fixed pitch prop, it's clear that swedna's cart is self starting at the speeeds shown in the videos. Sporks cart self starts in the mild breeze of the fan, and in a mild wind, but a wind gust was strong enough to break traction with the wheels and allow the prop to windmill in the "wrong" direction. However the cart still responded by going downwind, and eventually as it's speed increased, both drag and torque related forces decreased, and the wheels regained traction, and the cart returned to it's normal operating mode.

In order for a cart to respond to a tailwind by going upwind, the effective advance ratio has to be > 1, and I'm not sure if this is possible with a properly configured DDWFTTW cart.

In order for a cart to simply not move in response to a tailwind, the static friction forces have to be = the difference in the linear forces, increasing at the same rate as the difference in linear forces. Eventually, if the wind is strong enough, it will overcome the static friction forces, but at that windspeed, the cart may not be able to go DDWFTTW because of the rolling resistance factor at high speed. For example, in a 100 mph wind with the cart moving at 100 mph downwind, the apparent wind is zero, but the ground speed is 100mph, and the ground speed related drag factors would be high, even if the prop had a variable transmission or variable pitch.

 

[zoobyshoe]

I'm not sure how to mathematically model a propeller or any thrust generating device and the air flow involved for a DDWFTTW cart. A mathematical model for a DDWFTTW prop would include prop pitch, prop diameter, prop angular speed and prop linear speed with respect to the air, which goes beyond the simple advance ratio concept. Ultimately, it's how much the prop slows down a tailwind versus the forward speed of a DDWFTTW cart (which the rate of rotation of the prop is tied to).

In a tailwind, at startup the issue is if the prop acts more like a bluff body than it acts like a windmill generaing a backwards torque on the connnected wheels, or more accurately, if the forward force due to aerodynamic drag is greater than the backward force from the ground related to the windmill torque effect of the prop.

Even with a fixed pitch prop, it's clear that swedna's cart is self starting at the speeeds shown in the videos. Sporks cart self starts in the mild breeze of the fan, and in a mild wind, but a wind gust was strong enough to break traction with the wheels and allow the prop to windmill in the "wrong" direction. However the cart still responded by going downwind, and eventually as it's speed increased, both drag and torque related forces decreased, and the wheels regained traction, and the cart returned to it's normal operating mode.

In order for a cart to respond to a tailwind by going upwind, the effective advance ratio has to be > 1, and I'm not sure if this is possible with a properly configured DDWFTTW cart.

In order for a cart to simply not move in response to a tailwind, the static friction forces have to be = the difference in the linear forces, increasing at the same rate as the difference in linear forces. Eventually, if the wind is strong enough, it will overcome the static friction forces, but at that windspeed, the cart may not be able to go DDWFTTW because of the rolling resistance factor at high speed. For example, in a 100 mph wind with the cart moving at 100 mph downwind, the apparent wind is zero, but the ground speed is 100mph, and the ground speed related drag factors would be high, even if the prop had a variable transmission or variable pitch.

What about the notion of a vertical shaft rotor such as we see on anemometers which can accept power from wind in any direction?