In for a penny, in for a pound.

[agor95]

Hi All

Has anyone else tried simulating this device or built one?

It's would help to clarify the dynamics of the system.

Two case scenario:

1. The pendulum is extended and prevented from dropping.
This results in a balanced device.
2. The pendulum is a little longer in length and operating as 1.
This represents an unbalanced device.

The question out there is;

What angle does the Main mass start to dominate the smaller mass?

A 2D physics software could rattle out answers around this subject.
It would be quicker and better than a hand calculated or dodgy software approach.

Regards

[Robinhood46]

Agor,
I can't simulate a mechanism that functions, as a simulation, to see how everything interacts.
The last few videos i have shared are with stops. This is fine for getting a simulation to function correctly, the problem is that it isn't the desired mechanism. Stops, which determine the extent to which the lighter weights can swing, allow the weights to be held at their desired position to create the change in balance which is needed to cause movement. Stops also create the backward and forward swinging of the lighter weights, which is not desired.
Without stops we need a mechanism (style MT51) to coordinate the swinging of the lighter weight with either the wheel or a slight movement of the heavy weight. I can't simulate an MT51 mechanism with Algodoo, so i don't have the ability to play with all the variables that could be having an effect on the forces in play.
I have just received an email that the one way bearings, or clutch bearings, have been delivered so i will be playing around with them very soon. The problem, as you well know, playing around with things in the real world generally takes a hell of a lot more time than in the wonderful imaginary world of simulation.
I thought about replicating a mechanism style MT51 to achieve the desired effect and decided that one way bearings should do the job just fine, so why bother fannying around when i can buy cheap bearings. I don't have any experience with these bearings so i do not know if they will work or not, time will tell.
"What angle does the Main mass start to dominate the smaller mass?" This can be adjusted to where you want it to.
This is exactly one of the changes that i made recently because of my very interesting discussions with THX4. My previous attempts were with the weights either neutral (balanced) with the light weight fully in or fully out. Where things started looking really interesting is when the neutral position is with the light weight neither in nor out.

[agor95]

You have my sympathy for I share your pain.

Stretching what we can do with what we have is painful but inevitable part of learning. Running tests in sections are better than not having any tests.
At lease it takes some of the load from our imaginations.

I also look out for some MT51 like modern replacements.
I came across Mini Ratchet Wrench possibility.

Creating maths formulas to drive simulations are also painful, but one has too learn.
However the observation from a member that a pendulum moving in the x,y and z can be projected onto a 2d x-y or 1d y plain is a gift.
Why go thought all the pain of calculating the y position with 3d rotation when a simple 1d formula will do.

Getting back to the topic:

A balanced 4kg to 1kg at 1unit to 4unit length balanced beam.
Well we all know the device does not keel as it's balanced.

So any momentum will cause the 4kg to go over the top of it's arc path.

When scenario 2 is present then there is keeling and the severity of that keeling is related to the amount of in-balance present.

This is commonly shown by the centre of gravity searching for the lowest point.
Well that lessen 101 for 'Quester'.

That is an interesting observation about the light weight being neither fully in or out.
A one way ratchet can cause helpful bias but at the loss of the bias caused by it's repeated swing.

Best to start with a ratchet mechanism; so long as you do not loss too much via friction.

I always watch out for collisions, fast switch transitions and high speed motions.
They are the main cause of calculation errors.

P.S. Could you keep us updated on the one way bearings?

All the Best in your 'Questing'

[Robinhood46]

The ratchet is a good idea and if the one way bearings don't work i'll probably give that a try before making a mechanism. George Kunstler showed a wheel a while back with numerous socket wrenches fixed to it. The one way swinging is exactly the desired movement and i wouldn't have thought it too difficult to fix weights to the handle ends of the wrenches. The difficult part might be fixing the connecting rods. That being said, fixing a plywood, metal or plastic sheet to the wrench would be easy and then everything else is just done on the sheet.
"That is an interesting observation about the light weight being neither fully in or out.
A one way ratchet can cause helpful bias but at the loss of the bias caused by it's repeated swing."
I may be being a bit too fussy here but i prefer continual swinging as opposed to repeated swinging. The videos with the light weight swinging against a stop repeatedly is not the same as the weights swinging continually in the same direction. We just need to find the right coordination so as to cause a continual battle between the light and heavy weight to dominate. My guess would be that the unidirectional wheel can never achieve balance, even when stopped and the bidirectional wheel can find balance, but only when stopped.
I will keep you updated.

[agor95]

The ratchet is a good idea and if the one way bearings don't work i'll probably give that a try before making a mechanism. George Kunstler showed a wheel a while back with numerous socket wrenches fixed to it.

I remember the image; I searched and placed a posted asking him to say hello about 2 weeks back.

What I have attempt to do with the ratchet was a synchronized swinging, So all 8 ratchets are forced to do the swinging in one direction.

I have attached a picture so that you can see that all the ratchets are in one direction, they form a circle, look at the red parts.
You can also turn the picture by some degrees, then you can see the point when there will be no torque.
The picture and the experiment was from March 2018

That is to test the AP Wheel within Wheel device.

P.S. Well a continually swinging thing is a spoke of a perpetually rotating wheel; And we all want that.

All the Best

[Robinhood46]

I can't find the thread with George's wheel. The ratchets swinging in the wheel, if i remember correctly, was the source of the OOB of the whole wheel, at least this is where George was trying to find it. I too have tried to find it in this manner without any success. I think it boils down to the same problem that George found with the octagon, that it to say that trying to cause an OOB of the whole wheel doesn't work. This means that when Bessler says that with one crossbar the wheel rotates only just, but with more crossbars it rotates much better, he is actually saying that if only one crossbar is causing the wheel to rotate, then the wheel rotates badly, but with more crossbars causing the wheel to rotate it rotates much better. If each and every crossbar is connected to the wheel then they cancel each other out. A bit like you can't get a whole octagon to shift over, or get the accumulation of the swinging ratchets to cause an OOB of the wheel.
The only way to find any energy to exploit is to accumulate the infinitesimally small amount that can be obtained from each crossbar individually, and the only way to do this is to ensure that the crossbars are not connected (fixed) to the wheel. They obviously interact with the wheel but they are not connected directly.
Therefore, the continual swinging of the ratchet in the same direction, causes the swapping of the domination between one end and the other of the crossbar and the punctual interactions with the wheel causes the rotation and assists the synchronisation of the swinging.
Well, that's the plan, and other than absolutely everything, i can't see what can go wrong.

[WaltzCee]

https://besslerwheel.com/forum/viewtopi ... 45#p182445

[Robinhood46]

Thanks for the link Waltzcee,
I was under the impression that the ratchets could rotate 360° without hitting each other. Maybe i'm thinking of a different wheel or my memory is playing up.

[Sam Peppiatt]

Robinhood46,
I really thought pendulums had a good chance of working but, I could never find a way to reset them. The one way bearings only "hold" for about 90 degrees of rotation, which was very disappointing. Ratchets add a lot of friction, a lot more then what you would think. I don't think it can be done; or, if there is a way, I don't know how to do it.

But, if you could find a way, (a way to reset them), they will turn the wheel-----------------Sam

[agor95]

Hopefully the bearings use a roller blocking method. Thus having less friction and noise than an actual ratchet type.

Keep in mind there are two pendulums the one at the end of the arm and the one pivoted at the hub.

P.S. I hope some of the 2d physics software members chip-in on your thread.

Regards

[Robinhood46]

https://www.cdiscount.com/auto/pieces-a ... 67366.html
i don't know if the link will work or not, these are the bearings i ordered, i'm hoping they will do the job.

[agor95]

https://www.cdiscount.com/auto/pieces-a ... 67366.html
i don't know if the link will work or not, these are the bearings i ordered, i'm hoping they will do the job.

Good it looks like the roller block type.

Note.

The reason for putting the bearing on the hub and not the pendulum.
This is to give the pendulum time too move to it's lowest position.

Using the guides there is no need for a ratchet on the pendulum.

Regards

[agor95]

Simplify is the name of the game.

To simplify I started thinking what too do to the RH46 based device presented.
So we can simplify the design and the maths.

The first thought; drop the pendulum rod and use string.
Therefore it is a pendulum on the way up and on the way down just let in roll in a straight line.

That will simplify the maths and it's the shortest distance from A to B.

Then wait a minute drop the string and use a grooved tracks!

In the presentation the downward track z-plane path dips down by 5 degrees to give it a head start.
I am mindful any drop is a loss of weight pressing on the radial arm it is interacting with.

Also when at slow rotation rates the mass will follow the dipped curve or it's 2d equivalent; another simplification.

At increased rotation rates there would be a loss to 'end of radial' collisions.
Could the path be modified to deflect the mass along the other path?

Would that need a ratchet to prevent rollback?

Also would that increase complexity in the build and maths?

Regards

[Robinhood46]

I haven't yet gone to get the bearings, so no update on how they perform, or if they give the desired result.
I thought about Agor"s suggestion of using string, as opposed to rods, and somehow found myself thinking about coordinating the apologia wheel with 2 rolling weights interacting with the crossbar(s) to cause the shifting of the domination in the manner of my recent videos. My thinking is that it may be possible by using this method of shifting the domination on the crossbar, which causes 180° of rotation, by the 120° of rotation of the apologia wheel, we could get each system to give the other the much needed nudge to get them to achieve a full cycle.
If there was only one weight in this video the movement of the rolling weight would be identical, it would also be less often, but the force needed to create the movement would be far greater.
https://www.youtube.com/watch?v=UHRpWJN_yzo

[agor95]

...
If there was only one weight in this video the movement of the rolling weight would be identical, it would also be less often, but the force needed to create the movement would be far greater.
...

Naturally we are looking at ground other have go over. However we are doing this in the open rather than alone.

When I considered the approach a few points came to mind. The first was the collision at the end of the curve.
Could it be finessed too roll up at the end along a z-plane curve. The path would then prevent is from rolling down back the main curve.

Also it would trigger the initial drop at the start.

The other point is the time and weight the mass has when it starts to drop and when it makes contact with the end of the curve.

The mass being lifted is generating counter torque most of it's time. Were the active mass is either dropping supplying less weight.
Or briefly pressing down on the end of it's curve for less time than the counter torque mass.

What do you think on this analysis?

Regards