We don't care the Bessler wheel, the most important is to build a working wheel...

[path_finder]

step #11 of my explanation,

After review I can explain why I displayed this rodx3.gif animation in my last post.
Remember: the most important point was to ROTATE THE ROD BETWEEN 4:00 AND 6:00 o'clock...not for the fun, but for two reasons:
- If the weight must be located on the inner circle at 6:00 , it obvious that the 4:00 position is inverted versus the 6:00
- The weight MUST ABSOLUTELY FOLLOW THE HORIZONTAL PATH (yellow in the realpath_unidir.png picture, page #1 of the topic).
So the challenge is to ROTATE THE ROD DURING 1/6th OF THE WHEEL and SIMULTANEOUSLY OBLIGE THE WEIGHT TO FOLLOW A STRAIGH LINE.

How can we define the motion parameters? very simple:
the vertical distance between the weight and the horizontal line passing at 9:00 and 3:00 is constant, and equal to the half radius of the wheel.
This definition is simple but how to build a mechanism respecting this constant distance.
The first idea was to use an horizontal slide, to insert a bearing inside and to hang up the weight axis with a cord.
This bar could remain horizontal by a linkage (passing through the main axis of the wheel) assuming that the left part of the bar is really FIXED (near 9:00)
See the picture 'coulisse.jpg' below.
If the principe seems to be attractive, it appears quickly hard to build (see the 'point_fixe.png' picture)
In addition, we can not use a single slider for all weights (there will be some collisions)
So we need so many sliders than weights, wich supposes a battery-like design (with some parallel cells) and a thickness more important for the wheel.
This solution was bad, with one exception: during the experiments I have observed that one moving bearing can drive TWO weights.
In fact I had forgotten the symmetry: what is true at the bottom of the horizontal line is true at the top also.
This is the reason why I kept the design with the lozenges (rhombs) wich has two advantages:
- the number of moving points is divided by two
- there are no cords tangle anymore
We now obtain this new animation: rodw4.gif

In the next post I will explain how to fix the left point and/or obtain the horizontal line.

[path_finder]

step #12 of my explanation,

Instead to use a sliding bearing, we can use a very common mechanism 'the linkage of Watt', wich has the property to allow the middle of the center bar to follow a straight line.
This linkage on the animation 'rodx3b.gif' consists of three equal segments.
Their lenght is equal with the excentricity distance of the sliding point.
In this animation the lozenge (rhomb) is attached to the middle of the bar and we obtain the whished motion.
But take care: the wheel rotates, that means that the both attachment points of the Watt linkage must be remain fixed.
Alternatively we can say that the white circle of the picture must rotate counterclockwise (reversed versus the wheel) with the same rotation speed.
This movement can be obtained with a serie of gears: do you have any ideas?
Do you recognize the scissor-jack, the moustache of the workers and the AB toy...
In the next post I will show you a complete wheel with eight weights.

[genmurphy]

Hi Path-finder,

I do not have much to say. I am just very curious to read the next post...

[AB Hammer]

path_finder

It is funny that I had already drew this same path as rodw4.gif yesterday after my posting. Since I am a builder, your point_fixe.png is very similar to what I had figured for dividing up the sections needed. I do believe I am following what you are doing. But I still have to be skeptical. No offense intended.

[jupter]

/This movement can be obtained with a series of gears/
some planetary gears ?
the last post has thrown me on where your going but i like it
nice work path finder

jupter

[greendoor]

I don't see any similarity to Chas Campbell (CC)... the CC wheel is doomed to failure. The operating principle is based on lifting balls at an inner radius, and dropping them the same distance at an outer radius. Certainly there is more leverage & torque at the outer radius BUT there is a much faster peripheral velocity at the outside radius too.

The CC wheel is doomed to fail, because it runs out of time. There is no way the ascending balls can keep up with the descending balls - which is why he can't demonstrate a running wheel and talks about 'synchronising problems'.

This Time issue will always be a problem with perfect circle orbits. But in nature, do we see perfect circle orbits? I like path_finder's principle, because it is displaying eccentricity, which I believe is vital to making a gravity wheel work.

I will try to attach a GIF that shows an important principle ... stay tuned

[path_finder]

Dear greendoor,
This is the Euler law (sweeped surfaces by unity of time).
The ellipse could be an alternative to my ideal path (only approximative, but theorically working).
I tried to use this way, like the picture below shows.
But I had so many troubles with the cords (too much tangle).
Many thanks for you comments.

note: for some reason the displayed picture is not correct.
Please click on the picture to open a new window. It will be better.

[path_finder]

As promised, an eight weights wheel.
But don't be mistaken.
Sometime you will see only, Five, or Six or Seven weights: this is because the twin weights appear as a single one (due the 2D view).
Be carefull calculating the COG...

[path_finder]

step #13 of my explanation,
I'm sorry. I apologize to have forgotten the solution for the first design above (not symetrical) with some cords.
There are several solutions for rotating the rods between 4:00 and 6:00.
Instead to give you immediately the solution, I will submit to you several ways you can experiment and/or think about.
The first one I tried was not really efficient., but after improving could be usefull for the skate-boarders...
In that design you need two rails on the inner rim of the wheel. Why not?
One problem was the fact that the trolley want to leave on the left side.
I was obliged to implement a plank (not represented on the picture) at the bottom of the trolley, left free and acting like a cliquet (ratchet) with teeths located on the inner rim
(I can't translate the french word 'crémaillère').
This can explain the 'cracking' noise heart by the widnesses (just a suggestion)...

[jim_mich]

http://babelfish.yahoo.com/translate_txt
crémaillère = toothed rack ??

[path_finder]

Jim_Mich
Many thanks for the translation wich is showed here:
http://en.wikipedia.org/wiki/Rack_railway
There is a big difference in my picture: the teeth are not symmetrical
I modified the picture with more details (the teeth are shown only inside the blue circle but there are present all around the inner rim of the wheel).
There is another design with the same effect (see MT51) where the trolley is attached with the tool at the most right part of the drawing)
Many thanks for your attention.

[path_finder]

Just to confirm the pertinence of this design, you can find herafter a picture where the weights are not lift up by some cords, pulleys, trolleys, gears, etc.
Here we use a very old principe: Archimedes.
In this example there are only SIX weights.
Prepare your diving palms, mask and tuba...
In view to minimize the water resistance, the weights must be not cylindrical but discoidal.
You can use any other method, some magnets par example (I did not tried).

[jim_mich]

Divide the circle up into pie shaped pieces and calculate the sum of the torques.

During section 'A' the torque is W×R×Cos(aº) with aº being the angle to rim connection point and 'R' being the rim radius.

During section 'B' the torque is W×R×Cos(aº) with aº being the angle to the center of the weight and 'R' being the radius to the weight.

During section 'C' the torque is W×R×Cos(aº) with aº being the angle to the center of the weight and 'R' being the radius to the weight and 'W' being variable depending on how much the weight is immersed.

During section 'D' there is no torque.

During section 'E' the torque is W×R×Cos(aº) with aº being the angle to the center of the weight or the rim connection point (they are the same angle) and 'R' being the radius to the weight and 'W' being variable depending on how much the weight is immersed.

During section 'F' the torque is W×R×Cos(aº) with aº being the angle to the center of the weight or the rim connection point (they are the same angle) and 'R' being the radius to the weight.


Note that during all of section 'D' the weight produces no torque and during section 'A' the torque start at zero and progresses to maximum. During section 'F' the torque starts at about half of maximum and digresses to zero.

A computer program could easily sum the incremental torques of a weight as it travels full circle and I'm willing to bet that the wheel will, on average, exactly balance.

I might add that the weight will not follow the expected path during section 'A'. As it enters section 'A' it will need to speed up in order to move outward but inertia will hold it back. This will cause the weight to not drop as quickly and the rope will be swung to the left. Then gravity will cause the weight to start swinging out. Depending on the speed of the wheel, the weight will then swing out beyond the expected path and oscillation will develop, which will end when the weight rests on the rim.

Static analysis is always different from dynamic analysis. Looking at static over-balanced wheels is rather easy but the laws of physics always make it unproductive. Looking at dynamic wheels is much harder but it might offer a loop hole.

[path_finder]

Dear Jim_Mich,
The calculation is not so simple you say. What you say is true on a statical point of vue. If for some positions the balance is true, you can also observe some other positions really unbalanced.
The difficulty is coming from the fact that the transition phases are dynamic, and you can't apply a static calculation. It's specially true when a weight is entering inside the water or is exiting. On the animation the level of water is too high (there is no need to enter entirely the weights in the water). When the internal level in a weight reaches the main level (when the both levels are equal), the weight is immediately reset to zero. So the force needed for lift up just the rod (supposed much more light) is low, and few millimeters of level difference will be sufficient.
This is the reason why the main level should be adjusted few millimeters above the middle of the 6:00 radius.
The exact calculation must take in account the following actions (see the picture below) and I give hereafter two specific positions where there is no doubt.
(the weight are in black and the momentum in violet)
Another point is the number of used weights: if you increase this number, the unbalance will increase too.

[path_finder]

step #13 of my explanation,

I return back about the various ways to lift up the weights between 4:00 and 6:00.
Various indeed there are. Let me show you some examples:

1. weights lifted by some cords:
If you attach one end of the cord at the weight's axis, where to attach the other end?
If you attach it at the opposite location of the wheel circle (following the MT137 par example) a simple geometrical calculation will show that the lenght is constant with the rotation of the wheel.
We need a variable lenght and any attachment on any point of the wheel will give nothing.
So we need a mechanism were at minimum one moving part, desynchronized with the big wheel, bringing a variable delay during the rotation.

1A. The planetary gear:
The most common mechanism doing this is the sun-planets geared assembly, where the attach points are located in the circumference of the planet.
an animation showing the principe is here: http://www.angellabsllc.com/video/animation.xls.
Also see the Kauetz and Vimel engines at the exceptionnal web site of Douglas Self:
http://www.dself.dsl.pipex.com/MUSEUM/P ... &mouse.htm
showing an exceptionnal number of engines and forgotten old mechanisms
For these readers not familiar with such as mechanism, see our Wikipedia friend at: http://en.wikipedia.org/wiki/Sun_and_planet_gear
And for these people still encountering some difficulties by finding any usefull mechanism, I recommend the Del's engine at:
http://www.youtube.com/watch?v=u7ziwuIp ... =1&index=7
wich summarizes most of usable mechanisms. A mandatory!.. I liked it...
For the Bessler fans: I will be surprised if he used a planetary design. Nevertheless I still have a doubt (somewhere there is a mention to the sun and the planets).

1B. the revolving cams:
One cam is fixed (linked to the ground through the central axis.
The second cam is linked to the rotating wheel.
The bearings are connected together with some springs.
Depending of the number of used weights the shape of the cam has three or four (or more) lobes.
As you can see on the animation, you can attach the second end of the cord at one of the rolling bearing.
Who talked about flowers?
This design was much more efficient (and easy to build) than the previous planetary.
Try yourself!...

some other usefull data on the next post.