The summary of my latest studies

[path_finder]

Dear rasselasss,
Many thanks for the advices.
I'm not a bike mechanician, but a PFM (Public Finance Management) expert.
I use what is in my hand and at my level.
So far don't wait on too much miraculous results...

Hereafter a shot of the modified first chain: a tensor has been inserted in the loop, located on one end of a swinging rod, where a spring assumes the tension of the chain.

The first experiments have been made with the crankshaft disconnected. The result is not smart, noisily and chaotical.
In addition this assembly is NOT bi-directional: There is a privileged sense of rotation (due to the presence of the tensor), and if we reverse the rotation some rings of the chain are accumulated between the two gears (the tensor don't make its job) like shown in the second shot (red circle).

The next experiments were negative as soon the crankshaft has been reconnected (the constraints are extremely strong).
Even with the inactive primemovers (the both weights attached to the center) I was never able to move the whole assembly.
IMHO its not a question of friction (yes there is still one) but a problem of transfer of the torque.
For a reliable operation the torque must be applied at the outer rim, and NOT very close of the center like here with these gears
The use of some gears, even manufactured with the best care, will NOT solve the nightmare of the linkage with the crankshaft.
On the other hand I'm pretty sure there are another ways to drive independently each primemover.

[mickegg]

Hi path_finder

Do NOT use a spring to tension the chain!

Take up the chain "slack" and then rigidly fix the tensioner sprocket at that position.

Provided the sprockets run true you should not have any chain binding problem.

Regards

Mick

[path_finder]

Dear mickegg,

Take up the chain "slack" and then rigidly fix the tensioner sprocket at that position.

I used this way with the second chain, but with some controversial results.
In any case many thanks for the recommendation.

[path_finder]

In view to improve the mechanical linkage, I think having found an acceptable (cheap but reliable) solution: the VTT chain/gear technology.

The first shot shows one of the four freewheels. Each one has 18 teeth (this is the standard value of the market). Cost: 6 USD (Decathlon).
On the right side of the shot is a 'VTT 10 speeds cassette', under blister, and including a set of ten plates with a variety of teeth.
I took this cassette because it was the only available with a 36 teeth plate (the double of 18).
For the professional VTT market there are some 32 teeth plates to be used in conjunction with the particular 11 teeth freewheel, but much more expensive.
Price of my cassette: 85 USD.
But for this price we have a second doubler (see on the second shot, the 22/11 couple).

The third shot shows how I got the final 36 teeth part, destroying the five fixation pins (yellow arrows) with the rest of the crone.

The chain is a standard set for bike. The length must be adjusted: for that purpose I purchased two 'quick-rings' for the linkage (4.5 USD each).
I hope this new materials will help me to solve the remaining mechanical problems.

[raj]

Dear Path_Finder,

This is exactly what I have used for my last ' Speed Doubler '

See picture below.

Raj

[raj]

(ii)

[path_finder]

dear raj,
This part is only a minor aspect of the problem.
The most important job is still remaining to do.
The major point with your kind of design, is the linkage with the crankshaft.
Then the integrity of this crankshaft (I don't think your's will survive more than a half turn), the stability of the four primemovers synchronization, and finally the dancing behavior of the whole assembly during the rotation.
Good luck anyway.

After review I prefer to focalize my energy around the design where the four primemovers are located in the same plane, like explained in a previous animation, see here: http://www.besslerwheel.com/forum/files/raj4x.gif

edited: the same design of order three:
http://www.besslerwheel.com/forum/files/raj3x.gif

Another version is shown below, where the four small rollers are the four freewheels, the central gear is grounded, and the linkage is made by a chain like in the second drawing.
Click on the drawings for an enlarged view.

[path_finder]

For my own birthday anniversary I offer here a shot of my latest wheel (first shot below).
For some newbies this unusual shape cannot be a wheel: but it is! even if it seems to be a cross inside a square.
For those members following my path and interested with my numerous attempts, this curious frame is the practical translation of the idea presented previously here: http://www.besslerwheel.com/forum/viewt ... 679#100679.

The second shot shows some details on the way I used for the 36 teeth central gear, which is fixed within a disk, itself sustained by four groved rollers. The hole at the center is for a pass-through shaft, held in place by a bearing (not installed yet on the shot).

In that state I did not made any test: there are still some changes and adjustments needed.
First I must modify the fixation of the four freewheels.
The geometry of the chain has been solved: the chain don't jump anymore. The bikers know the difficulty to adjust the chain with Three points (Plate, freewheel and derailleur). You can imagine the difficulty with FIVE gears. But this has been solved by a strong respect of the distances, and now the five gears are in exactly the same plane.

But the chain is tied too much strong, and the motion of the chain is difficult because of a too much high friction: I must install all freewheels on a mobile axle, held in high tension by a spring (a single tensor will not be convenient, especially for the bi-directional test)

Then a difficult but important job is still to be done: the adjustment of the different parts.
1. The correct 90 grades dephasing between the primemovers (shifting the chain).
2. The reset position of each primemover at 12:00
3. The position of the chain for each primemover in relation with the central gear linked with the ground.
As you can see this is not an easy job.

[path_finder]

Bad news...
The test of the chain after the modification has been successful. You can see the video below, where the central gear is in the reverse rotation without jump of the chain. At the end of the video you can see the attempt to let rotate the central gear in the right direction: the wheel is locked, this is normal, the two bottom primemovers being on the floor.

But the next step has been disastrous. After one turn of the wheel, one of the four rollers holding the central gear has been lost, the support of the rollers being broken, as shown in the second shot. The torque is too much strong (may be the motion of the primemovers being also too much chaotical). Anyway I have to think an another mechanical design.
This don't surprise me. I knew the same accident with another promising attempt in the past
(remember here: http://www.besslerwheel.com/forum/viewt ... 3479#93479).

[Furcurequs]

Hey path_finder,

I'm a bit late, but happy birthday!

I also wanted to say that I enjoy seeing your builds.

With your creativity, building skills and quite prodigious output, though, I sometimes worry that one day I'm going to peek into your thread and see one of my own designs that you've just come up with and whipped out in a matter of minutes.

...and that I've been thinking about and laboring over for months - if not years.

...lol

Anyway, good luck.

Dwayne

[path_finder]

Dear Furcurequs,
Many thanks for the kind words.
I do not pretend to be original, and for sure there are several solutions.
IMHO the difference is between the idea discovery (almost usually in front of the mirror at the morning during the shaving time), and the practical building of this idea. Unfortunately we are few.

[path_finder]

Like suggested earlier the solution for the 'speed doubler' should be any independent mechanism, ideally included in a discoidal volume, and able to drive by itself the two rotating parts.
This configuration allows the building of a kind of battery, just by the repetition of the basic cell.
An analogy with this concept is the diaphragm, where a discoidal cell can do by itself a particular job, and in addition where a superposition of such as cells, opened and closed within the right timing, can translate a ring of air pressure from one end to another (a kind of multi-stage stato-reactor).
If this mechanism is found, and if several such as cells have been manufactured, the remaining question is 'how to synchronize the mechanisms together?'. The answer is easy: by a clever dephasing of each outer rim. So far during the building a particular care must be taken for a correct adjustment in the positioning of each cell.

The drawing below shows this kind of mechanism.

The disk A (in rosa) is the largest part. Its outer rim will be linked to the outer rim of the next cell. The red circle (with a reference radius R) is the support of the 12 yellow pins. These pins are used as reference position during the rotation of the ring B.

The ring B (in light yellow) has two particular dimensions:
1. The diameter of the outer rim circle is the 5/6 of the previous red circle diameter on the disk A.
This outer rim support 10 half circular holes where the 12 pins of the disk A take contact, assuming a mutual rotation of the two A and B disks. This solution has been preferred to a double full teeth gear system, much more difficult and expensive to manufacture. For sure any double gear assembly can be also used.
2. The diameter of the inner rim circle is the half of the previous one: like the outer rim the inner rim has 10 half circular holes for the contact with the next disk C (see below).
This ring B rolls on the inner rim made by the 12 pins of the disk A, in the same direction of rotation and with a speed equal to 6/5 of the disk A rotation speed.

The disk C (in green) is centered on the same axis than the disk A, but not linked with. A bearing included at the center of the disk A can support the axle of this disk C, which is free to rotate thanks the 6 pins (in light blue) taking contact with the inner rim of the previous B disk. There are 10 holes on the disk B inner rim, and 6 pins fixed to the last disk C, giving a ratio of 3/5. The speed of the disk C is therefore 6/5 x 5/3 = 2 times the rotation speed of the disk A.
The overall dimension of the green disk C is much more greater than this 6 pins pseudo gear because we need a support for the elbow of the primemover mechanism situated at an half radius of the disk A red circle.

Thanks to the particular dimensions of the three disks, the ring B assumes a permanent contact between the A and C disks.
The weight of the ring B must be important enough for a good contact due to the gravity field. This extra weight can be obtained by a population of cylindrical masses located around it: this will provide an useful inertial wheel with an important momentum very efficient for the speed regulation.

This is the principle.
Note the behavior of this mechanism, which follows exactly the hypocycloidal path given by the famous drawing of Apologia.
(for the new members, here: http://www.besslerwheel.com/forum/viewt ... 4342#94342)
In addition a major improvement is possible for this mechanism.

N.B.: the weights and the associated linkage has no been included in this drawing (cf: http://www.besslerwheel.com/forum/viewt ... 8398#98398)

[mickegg]

Hi path

I understand you wanting to fabricate this with what you have and are used to working with; but I honestly think you will be wasting your time with this.

Unless it is mechanically held, I think you'll have great problems trying to retain the yellow "idler" disk in position.

Lateral movement would allow the disk to tilt ..disengaging the drive pins.

Prevention by "sandwiching" between the red and green will give friction problems tending to lift the
yellow out of mesh.

I also think that under load conditions it will be far easier for the yellow disk to ride out (cam action of the pins in the half-hole drive ) than it will be to transmit the torque and rotate.

Regards

Mick

[path_finder]

Dear mickegg,
Many thanks for your attention.

I think you'll have great problems trying to retain the yellow "idler" disk in position

Your comment is pertinent, but I can give to you two answers.

1. The contact between the yellow ring and the pins of the disk A can be improved by an excessive weight of this ring B.
This is the reason why I have planned to supply a population of heavy weights all around this ring, in view no only to assume a good contact but also to create a significant flywheel for the speed regulation.
Remember the plenty of 'weights' (named such as) put in the container by Bessler when the wheel has been relocated: for sure some of these weights did not take place in the primemover mechanism, but were used as masses in a flywheel somewhere.
Note also the very low excursion of the ring from the vertical position due to its automatic relocation during the rotation of the main wheel (in fact it should be an alternative motion across the main position itself fixed by the requested torque).

2. IMHO most difficult to assume is the contact between the inner rim of the B ring and the small 6 pin C disk.
Nevertheless there is an easy solution: a lozenge (rhombus) with two summits connected together with a spring, the two other summits being connected to the axis of the C disk and the B ring (like a kind of jack).
Another way is obtained by the use of a second C disk as counter part on the outer rim of the B ring.

The purpose of the drawing was just to explain the principle. It was not a blue print for any immediate manufacturing.
Please consider I cannot put everything in the same drawing, the global view becoming not understandable for the majority of the members.


Edited:
Another comment on the drawing: the final number of pins is not the same, but a high multiple of the given figures (12, 10 and 6).
This can be in relation with the number of square teeth depicted in the key-stone of the bookstore in Karlshafen (96, 80, 24).

[path_finder]

Before to start a time consuming (and perhaps expensive) building, I decided to check the above concept with a demonstrator.
Remember the previous old video (here: http://www.besslerwheel.com/forum/viewt ... 8602#98602) where the pseudo-polygonal mechanism has been checked.
In this old video the assembly was in motion in a horizontal plane (confirming the chosen geometry and the motion of each part).
Now in the new video the same mechanism has been tested in a vertical plane.
Although the comments of mickegg (see above) I can tell it works pretty well and the friction remains minimal: as you can observe, there is no additional weight on the intermediate ring B (which is a very light part made in polycarbonate). The result should be more better when some masses could be inserted, assuming a better contact.
Anyway the big question now is: what happens after the addition of the two pendulas (raj Balkee's linkage)?