• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Down wind faster than the wind

Cute! That makes sense - first gear adds the annoying extra turn = +2, then the second subtracts 1.5. So one of those sets for each vane could just be fixed to the actual wheel (excepting the central stationary gear).

I wonder how the planetary gears would compare. If the planets on their carrier (which again can just be the wheel itself) engage with a stationary annular gear, say, you'd have to gear the sun gear from the wheel rotation by another train, I presume, coaxial with the wheel. You're not reducing much in the way of gears, and adding a large annular ring would produce more drag beyond windspeed, perhaps.

I just downloaded freeCAD yesterday (there's also info at this site). I was looking for something simpler that would allow me to do 2D (or 3D) sims with automatic handling of gear meshing, etc., but I reckon it will be worth learning this, and it looks fairly user-friendly. I'm starting to think about studying mechanics or engineering with the Open University. Gulp. It models inertia, gravity, aerodynamics and impacts, etc., and is more general-purpose. It says that it automatically removes superfluous properties and constraints, so it might be quite easy just to fix circles in 2d with curve-on-curve connections to represent toothed gears, when I'm not bothered about how my windcart might levitate and such. ;)

Here's the basic spec in case anyone's looking for something like this. (What's "six DOF"?)

ETA: The only problem being the darn thing keeps on crashing every 5 mins.

"freeCAD is capable of full 3D pan, zoom, tilt and rotate on an assembly of parts in wireframe or rendered graphics. Parts can be solids or DXF file imports. Available solids are extrusions of rectangles, circles, ellipses and polygons. The parts can be positioned and oriented exactly in space, as are markers on the parts. Exact specification of linear and angular velocities of parts in space are also possible. Mass and inertia properties can be user specified or automatically computed based on uniform density. Available joints are spherical (ball), revolute (pin), translational (slider), cylindrical, planar, fixed, universal, point in line, point in plane, parallel, perpendicular, no rotation, constant velocity, rack pinion, screw. Both open and closed 3D loops are permitted. The curve-curve contact allows liftoffs and collisions based on coefficient of restitution. Available actuators are rotational, translational and full six DOF. Their motions are user prescribed functions of time. Forces and torques are user prescribed functions of time, displacements and velocities in all three components or along connecting markers. Example formulas for spring, damper, bushing, beam, aerodynamic, inverse square law and other forces and torques are given. Available functions are sqrt, exp, ln, lg, sin, cos, tan, arctan, arctan2, spline, spectral density. Users can specify constant gravity of arbitrary magnitude and direction.

freeCAD can compute kinematic, quasi-static or dynamic solutions for any interval of time going forward or backward based on the assembly and user requests. It does redundant constraint removal automatically and nondimensionalizes the equations for improved accuracy and stability of models that are microscopic or gigantic. Simulation progress is animated and the simulation can be stopped any time. After simulation, the computed solution can be used for animation or frame by frame analysis. Full 3D pan, zoom, tilt and rotate is available during simulation and animation.
Users can obtain engineering data in the form of plots and tabular output. XY plots can be zoomed and set to equal scales. Data series available include linear and angular displacements, velocities, accelerations, forces, torques, momenta and kinetic energies. Acceleration data include transverse, centripetal and Coriollis accelerations. Users can view forces and torques from joints, constraints, actuators, springs, dampers, applied forces and inertia. Fourier tranforms of all times series are available. Individual parts can be save into files and reinserted into any assembly repeatedly. Assemblies can be saved in binary or human readable, tab delimited, text format with notes and simulation data for later reload. The text format allows pre and post processing of assemblies by other programs, especially spreadsheet programs. Other specific text formats are for MOSES, PDMS Review and POV-Ray. freeCAD runs on Windows, Linux PC and Mac OS/X. Assembly data are unchanged across platforms."
 
Last edited:
Relative to the main wheel, the center sprocket has a rotation of -1, the middle sprocket has a rotation of +1 and the vane sprocket has a rotation of -1/2. Keeping the same reference, to drive the vane sprocket with a common ring gear, it would have to rotate at +1/2*rv/rgo.

A planet gear fixed to the main wheel would translate a sun sprocket rotation of -1 into +1 * rs/rgi. So all we need is a set of gears such that rv/rgo * rgi/rs = 1/2 (Note that if rgo=rgi, the ring gear disappears and it looks just like my image above).

With 3 or more vanes, this replaces the set of pinion gear with one ring gear and one planet gear. It's not a great savings in this application.


Planetary gears are neat. Somewhere around here I have plans I designed for a planetary gear with a 60 to 1 ratio with a very small tooth count. I thought it would be cool to make a minimalist working clock with all wooden parts. :cool:
 
John Freestone said:
Here's the basic spec in case anyone's looking for something like this. (What's "six DOF"?)
Click to expand...

Six degrees of freedom.

For a rigid body 6 degrees of freedom can fully describe its position and orientation. The rigid body can translate in X, Y, and Z; and it can rotate about each of these axes. But you should confirm this with humber to be sure. :D
 
Michael C said:
Neat. What would be the ratio between speed of the wheel along the ground and theoretical speed of the vane-propeller-thingy through the air?
Click to expand...

As a VERY rough approximation the wheel should approach twice the wind speed since the vanes appear to be attached at 1/2 the radius. However, the chord of the vanes is quite large and I would expect very significant losses from the actual rotation of the vanes as well as imperfect alignment of the "non-active" vanes with the air-flow. It would be interesting to run a simulation on this device.
 
spork said:
Six degrees of freedom.

For a rigid body 6 degrees of freedom can fully describe its position and orientation. The rigid body can translate in X, Y, and Z; and it can rotate about each of these axes. But you should confirm this with humber to be sure. :D
Click to expand...
Oh dear, silly spork. I checked with humber and it's a kite flying term, Drogue Optimized Flight - it's best when you have six, and you can let go of the string. Or a brick works fine. ;)

Thanks for that, BTW.
 
There's a guy going by "schroder" on this Physics Forum thread who is seeming more like Humber all the time:

http://www.physicsforums.com/showthread.php?p=2060743&posted=1#post2060743

Fast approaching a thousand posts of an entire group trying to help him get it right but he's off on some Humber like 'hop, skip and a jump' routine. He's calling it a "mechanical heterodyne" (invented term), but for sure not DDWFTTW.

JB
 
ThinAirDesigns said:
There's a guy going by "schroder" on this Physics Forum thread who is seeming more like Humber all the time:

http://www.physicsforums.com/showthread.php?p=2060743&posted=1#post2060743

Fast approaching a thousand posts of an entire group trying to help him get it right but he's off on some Humber like 'hop, skip and a jump' routine. He's calling it a "mechanical heterodyne" (invented term), but for sure not DDWFTTW.

JB
Click to expand...

Glancing at that discussion, I'd say he's still got a way to go before he reaches humber's level. He's doing his best, though. There's certainly the similarity of getting hold of the wrong end of the stick and then hanging on for dear life.

The "heterodyne" idea evidently comes from the fact that the turntable setup rotates: he's got obsessed by the idea that there is a relation between the periods of rotation of cart and turntable. He can't correctly translate between what happens on the turntable and what happens on the ground, but he has decided that other people are missing the essential point that the turntable represents a rotating reference frame:

"The TT is clearly a rotating reference frame while the earth is an inertial frame. Previously, I did not think this was important to the question but I now realize it is all important."

I knew that some doubter would cotton on to the fact that "real wind" doesn't blow in circles :)
 
My paddlewheel cart with fairing

Hi, all. New to the list but not to the obsession.

Here's a short video of the paddlewheel cart with a fairing that I finished today. On this list, I think you'd call it a tumbleweed with a cowl. youtube dot com /watch?v=TcSh0qqgYKc

I uploaded an image here too. forums dot randi dot org /imagehosting/326534a299af6d628b.jpg

I find it to be the easiest design to understand, especially due to its similarity with the wire-spool demo and the cotton reel in coolaun's delightful video "along the paper faster than the paper", youtube dot com /watch?v=E7vcQcIaWSQ

I can't see any fundamental reason why this cart won't go DDWFTTW except perhaps for its cheesy construction. Some people here expressed concern about churning of air under the fairing, but it seems to me that such turbulence could be all but eliminated by leaving no gap between the fairing and the air paddles. I haven't done that because, well, that's the least of this styrofoam cart's problems. Let's call it a visual aid, shall we, rather than a serious DDWFTTW attempt?

Cheers
Marc
 
Last edited:
Very nice effort. I believe your design can work in principle, but I expect it would be tough to make it work in practice. Not that I've run any numbers on it - it's more of a hunch. If you did get it moving faster than the wind I'd really love to see it. I expect it will be much easier to demonstrate on a treadmill.
 
One-sentence explanations

So far I've seen one good oversimplified single-sentence explanation:

It drives itself faster than wind speed because each moving blade behaves like a tacking sail.

Let me offer another:

It's confusing because the spinning blade secretly is performing two separate functions: it acts like a flat sail being pushed along by the wind, and it *also* acts like a reaction engine.

Perhaps our gut-level thought experiments would become simpler if we split these two functions. Use a cart with both a large flat sail held perpendicular to the wind, and also a smaller ducted fan driven by the cart's wheels.
 
wbeaty said:
It's confusing because the spinning blade secretly is performing two separate functions: it acts like a flat sail being pushed along by the wind, and it *also* acts like a reaction engine.
Click to expand...


When going downwind faster than the wind, the propeller acts exactly as any propeller does. It does not at that point act as a flat sail catching the wind in any way. It does perform that funtion to some extent when starting from a stand-still and then approaching wind speed.
 
You must log in or register to reply here.

Back
Top Bottom