The summary of my latest studies

[Jon J Hutton]

edited out

[murilo]

Here is a product that I ran across that is great for making prototypes out of plastic. They sell chemicals to make reusable cheap molds and also sell the different hardness plastics, foams and rubber. It is a 2 part mix that depending on the job usually sets up in 10 minutes. If you are inventing or need to replicate a part on the wheel many times over make the original if needed out of modeling clay or whatever, make the mold then cast the part out of hard plastic. The molds can be used many many times over. If interested Write me a private message. I am not selling this product simply passing along a real time saver and great for those headache projects. The price is around 21 cents U.S. per cubic inch.

Jon

Jon,
thanks for your initiative and offer!

Your idea has a lot to see with any kind of buildings, everywhere! Believe me!

(You got just now a sample of European Springtime Spirit... this is a private thread, DEAR Jon! Start your own!)

Best regards!
Murilo

[path_finder]

I have been always surprised by the bad result obtained with the design based on the ratcheted pendula (ref: http://www.besslerwheel.com/forum/viewt ... 3802#93802).
In fact the abnormal point was the motion of the pendula, still continuing to swing all around the rotation of the main wheel although the presence of the freewheel in the fulcrum.
So far I decided to clarify this mistake. My explanation is shown in the drawing bellow.

The part in yellow is one of the arm of the main wheel's cross where the pendulum (in rosa) is connected. The freewheel is represented by the red circle.
When the main wheel rotates (supposed clockwise) the arm moves clockwise like indicated in A.
Meanwhile the reference axle of the freewheel (in B) rotates also (for one turn on itself every turn of the main wheel).
The swing of the pendulum is represented by the red path.
The wished effect obtained thanks the freewheel is a lock in C, just immediately after the full extension.
But in the reality in my built wheel this lock appears in D. Why?
The reason is simple: during the swing to right and then back, the reference axle B has rotated and the freewheel has not be able in this short time to recover the introduced offset.
What we can conclude from this analysis?:

-1- the size of the wheel has some importance.
Translation: with a small wheel there is no chance the swing can be in time.
On a practical point of view: more larger the size of the wheel, more slowly will rotates the reference axle (allowing an earlier lock by the freewheel).

-2- the size of the pendulum must be not to large (because the slow swing).
The optimum design shall use some short numerous pendula all around the rim, instead few long pendula in the middle of the main wheel.
A relative high frequency of the pendulum will allow it to overpass quickly the C locking point. But not to much early, loosing in torque efficiency.
A low frequency will give a swing to much slow for overpassing the rotation of the reference axle.
IMHO (to be completed) there is an optimum couple of values for obtaining the C locking point exactly at the maximum of the swing extension.

These comments seem to confirm some ideas developed earlier in this forum.

[raj]

Dear Path_Finder.

I believe the freewheel(in B) will rotate counter-clockwide.

Raj

[path_finder]

Dear raj,
The B part (in yellow) is an axle soldered to the arm (in yellow), being the same body.
The freewheel authorizes the pendulum to rotate anticlockwise around this axle, but locks the pendulum as soon it tries to rotates clockwise, creating the torque.

On the other hand: the animation is in progress. Be patient.

[path_finder]

As explained above, the challenge now is to control the rotation of the freewheel internal shaft.
The minimum level of control is to oblige this shaft to be immobile (no rotation at all, named earlier 'the polar position').

But somewhere I was thinking why not clearly reverse the rotation?
That was the purpose of this attempt shown in the shot bellow:
D is the cross, main frame of the rotating wheel, and where the axles A are fixed: so far the first gear B (fixed to the cross arm) rotates with the arm within one turn on itself for one turn of the wheel.
The second gear C is fixed to the internal shaft of the freewheel E, and rotates in the reversed direction (with a speed much more higher due to the used gear ratio). F is the pendulum.

Without any further experiments cancelling my position, unfortunately this concept seems to be not successful.
In fact I cannot tell why, the position of the internal shaft of the freewheel being obviously always in advance on the pendulum.
The pendulum still continue to swing, but this time around the gear B instead around its own fulcrum.

For sure the best solution should be based on an assembly of some large gears, linking the ground (through the hollow central shaft) to the freewheel internal shaft..
But I still be convinced the gears are easy for the small parts but difficult to adjust for the large dimensions.
I have in mind another clever solution inspired by the MT138 drawing.

[path_finder]

In this new concept the four shafts (passing through the freewheels) are now self-rotating (not still tied on the cross arm).
On the both ends of each shaft is fixed a rod (the yellow circles in the shot), swinging with the shaft rotation.
All these eight rods are connected to a squared frame (the small green crosses on the shot), assuming all four shafts are now aligned on the same way.
I didn't use another two crosses for connecting the rods because the center of these crosses will be in mechanical conflict with the axle of the main wheel.
But by using a square shaped frame, I discovered another problem: as you can see on the shot the square cannot be saved, the frame being transformed very quickly into a lozenge.
I'm working on correcting this trouble.
Then the next point is to force the position of the squared frame in the lowest position while the full turn of the wheel, or -may be why not - to let rotate the two squared frames counterclockwise together.

[path_finder]

Finally I found the missing link.
An old topic had no resonance at its time: remember here http://www.besslerwheel.com/forum/viewt ... 9440#79440
and specially this video: http://www.youtube.com/watch?v=1v5Aqo6PaFw, in particular the last sequences.

The animation below shows the same principle applied to the 'virtual cam', which allows an excentricity centered at 3:00 (a very old quest).
The yellow rollers are fixed and grounded: in the practical way their axle is screwed into one side of the stand supporting the wheel's shaft.
During the rotation of the wheel they are just here such as an envelop for the belt.
The white roller is a tenser for the belt where the spring (in violet) is attached.
The other end of the spring is attached to a point acting on the pendulum orientation.
As shown in the animation the result is conform with our wishes: a difference of force depending from the side you are.
I'm pretty sure the best of the members will discover immediately what we can do with this for the 300th anniversary.

edited:
Nota bene: this single assembly can be multiplied in accordance with the number of pendula, each belt occupying its own sector (no mechanical conflict).

[path_finder]

If you observe carefully the balls inside a rotating bearing, you can say their centers are located on a circle rotating at a speed equal to the half summary of the both rings.
In addition this effect is absolutely independent from the size of the balls.
So far why do not use the same principle for obtaining our famous 'double speed disk.

The drawing below shows the suggested way.
The half left side represents a classic bearing. The outer cage is in red, the balls are in blue, and the inner ring is in yellow.
If the outer ring is grounded, the centers of the balls will rotate at a X speed, but the inner ring (in yellow) will rotate at 2X speed.
Now look at the right side: the balls have been replaced by a sector (in green) where TWO sets of rollers have been included.
The effect is the same like before: you get a sector rotating at a X speed, and the inner ring still rotates at 2X.

You don't see the benefit? Here is your speed doubler mechanism (see my signature).
This is exactly what the trilobed disk of Sabu is doing.

In relation with the topic sent to raj (see here: http://www.besslerwheel.com/forum/viewt ... 221#100221) I show hereafter a shot of the principle applied to the wheel.
At this step of the building only the rollers with the inner ring are installed.
In view to reduce the friction losses I selected a mechanism using THREE rollers, which is sufficient but reliable enough.
This is one of the feature of the 'flowerbowl' (trilobed disk of Sabu).
The three springs assume a correct contact in any case with the inner disk roller.

The next step (shot coming) is to include the second set of rollers in contact with the outer ring (in fact a hollow drum).

edited:
On the shot the rollers assembly is not easy to see: the frame is made with some transparent rods.
The principle is available here in these old topics:
http://www.besslerwheel.com/forum/files/flowerbowl5.jpg
http://www.besslerwheel.com/forum/files ... rbowl6.jpg

[path_finder]

An important improvement can be made in the previous design: the second set of rollers (between the green sector at a X speed) and the outer ring (in red, and grounded) can be located directly on the outer ring, like shown in the second drawing below.

What are the consequences ?
First in some circumstances we don't need the second set of rollers: if the rim of the green sector is a real circle (not only a cross), we have just to put the assembly inside the outer ring, and the green sector will just roll easily around the internal rim.
IMHO it was the way used in the ancient time for the flowerbowl, a reversed waterwheel, like suggested earlier here:
http://www.besslerwheel.com/forum/download.php?id=5996
where the 'double speed' wheel assembly acts as a hamster in her cage.

[path_finder]

The next step in the building of the wheel based on the concept of Raj Balkee.
This time the second compartment has been supplied with the second set of weights (there are FOUR weights in fine).
Note the new set of rollers, much more greater and much more efficient thanks a more stronger contact force (the first set will be replaced by the new one). The frame (made of transparent plastic rods) has been enhanced by a violet color.
Don't take in account the multiple holes inside the two polycarbonate disks, which are recycled from old attempts.
The next step will be the implementation of the set of rollers at the outer rim of the two disks.

[raj]

My very dear friend, Path_Finder,

Your work is so neat and so professional, that makes me overjoyed to see how you are proceeding with my gravity wheel concept.
It puts my own effort to shame.

I, sincerely, hope that the outcome of everyone's efforts will be beneficial in the learning process of all members on this forum.

Please, please keep it up till WE arrive to a conclusion.

Thank you ever so much.

Raj Balkee

[path_finder]

the first set will be replaced by the new one

Done.

[path_finder]

The primemover is now installed inside the outer ring (in fact only two couple of rollers linked to the ground thanks the stand).
The two last shots show the details: the outer rim of the two disks (rotating at a speed of 1X) is rotating inside two couples of rollers.
Even if some care has been taken for a good parallelism, a mechanical feature has been implemented in view to compensate any variation in the disk clearance: the rollers are able to shift along the shaft, this motion is limited by the two washers.

Now I know the most important question you want an answer to: what's happens?
The torque exists and the primemover has a real tendency to rotate, depending of the initial position of the four disks.
But I cannot obtain a continuous motion.
This can be easily explained: in that state the two inner disks (rotating at 2X) are not linked.
They must be physically linked together and mutually dephased of 90 grades (what is unfortunately NOT the case in my building). I was not concerned enough with this particular point.
So far I have to make some important changes in the geometry.
Basically we need a single 2X disk (like a pedalier) located in the middle plan of the stand (and not located at each side like now).

In addition I suspect a need for a third ring, totally free, between the rollers and the primemover.
This extra ring will allow a bi-directional rotation, but dividing the speed by two (this could explain the 52rpm/26rpm speed of the Bessler wheels).

[raj]

Dear Path_Finder,

I had seen it at the very beginning of my gravity wheel design but did not mentioned it before:
The design looks very likely to be a bi-directional gravity wheel design.

Raj