An idea that is worth your time.

[LustInBlack]

** Look at the screenshots later in this thread **

Explanation of the mechanism is below too .


In my opinion this should work, just have one problem ..

[jim_mich]

In my opinion using wm2d friction is not a reliable test. Instead put a motor on the wheel center and adjust the torque to counter the rotation of the wheel. From the motor torque setting and the wheel's RPM you (or I) can determine the horsepower of a working wheel. If the wheel slows down when torque is applied then the wheel isn't working.

Also note that under certain circumstances when objects bump into or roll upon polygons it can cause the objects to incorrectly accelerate rather than decelerate.

[LustInBlack]

Thanks for the idea ..

I've found that my problem will have to be resolve with something else than a brake.

There are 2 wheels, the two wheels want to turn faster than the other, because of the way they are made .

Actually, there are 36 arms on each wheel, and 12 weights ..

So, the arms must cross the weights boundary 3 times, those 3 times result in a pinched weight .. That's where I believe I should put a latch or something to bypass the weight ..

A kind of mecanical diode if you want . ..

But I'm tired, so Tomorrow it will be done ..


The motion of the device seem to be plausible right now ..
There are not many parts involved, but WM2D takes a lot of time to compute . . .

[ken_behrendt]

LIB...

Your design sounds a bit on the complicated side. With all of those parts, you will probably need a very long time to calculate even a few tens of seconds of motion of the model. I think that with most CAD programs the time to complete the simulation actually increases geometrically with the number of parts in the model. Thus, if you have, say, eight parts and need 30 minutes to complete a simulation, then, if you increase it to 16 parts, you may need far more than 60 minutes to complete it. In the past, I've had to give up on some models because I just could not get them to run in a timely fashion (like less than an hour!).

If you provide a screenshot of the model, maybe we can be of further help.


ken

[LustInBlack]

Yep, I think you're right Ken .

Anyways, I post a screenshot ..

But if it proves to be working, I hope my name will be remembered .. ;]

I call it, the "immersion wheel" .

I am making some modifications to the 36 arms to prevent a lock-up, I believe I can overcome the lockup .

The description :
- The blue Arms are free to fall about 8 degrees to a stop on the wheels .
- Red Arms are fixed.
- Grey weights, fixed on the blue arms as a test and because the simulation is just too long with them free, but I want them free in the physical construction.
- You can see the Cisor Arm effect in the middle of the 2 wheels .. Maybe that have nothing to do with the MT Toy, but it's interesting.
- The "hammers" are the blue Arms falling on the Transparent (36) arms of the other wheel .

The principle is simple, the Leverage is greater on the wheels because the weights are hammering the transparent arms at a greater distance, and it overcome the totality of the weights on the wheels .

In fact, I designed this based on my first idea, which was to guide falling weights outside a wheel .

Movement : The mid-section between the wheel can be thought as a perpetual waterfall, the weights disconnect from the wheel to the left, locks on the wheel to the right, at a different leverage Ratio and that creates more torque than the total weight of the wheel. Vice-versa, mecanism is symetrical .

Ratio of this one, about 3:1.

It turned to 10 RPM, and locked because the transparent arms desynchronized and struck on the top of the weights, that caused a major lock-up!

I'm sure if I take some time to calculate the length/number of arms and orientation, I can make it work ..


What do you think !!??

[Paul]

Hi LustinBlack, good idea!
I have not analysed in the detail your system however I have the feel you are on something. Maybe with the right proportions and the correct mechanics (latch, etc) your idea could work.
Good work

edit: in that way you have a different lever ratio but also a different speed between the long arm and the short weighted arm. I do not know whether it is feasible.

Paul

[LustInBlack]

Thx Paul !

Yes I had the same thought about the Arm/speed ratio, that's why the weights are free to fall [I mean, the length of their fall, since the falling weights are the Heart of the principle, without them falling, nothing would move], my idea was to change the length of the fall ..

I could set it to fall until it hits the arm, then I would need the Long Arm to be within reach of the fall .

But synching the speed of the two wheel and limiting them, to say, 10 RPM, put a latch mecanism on the tip of the long arms and use inertia to let the arm turn when not in reach .. (I know, with only inertia and not falling weight, it cannot do much work, that's why I try to find something else) .. Maybe I need some springs in there . . . .

I need opinions !

I need more experimentation with this as you see ..

But I hope someone can come up with something I didn't think of !

I'll Attach the model file in a short while . . .

Edit : .. By the way, I believe I could put 1 parallel wheel to each of the wheel in the model, and those secondary wheel would remove the weight when the model is not in synch .. That way, I remove the problem of Leverage/Speed ratio, that second wheel would also have torque, since there would be only 1 to 3 weights on the right side of that wheel, but that would cancel out when connected to the main shaft with the primary wheels ..

Confusing . .

[ken_behrendt]

LIB...

Well, I've looked at your design an am puzzled as to how it is supposed to work. I would imagine that, if both wheels separately, have an equilibrium position, then it would seem to be impossible for them to apply any torque to each other.

However, maybe the idea of your design is that each wheel will apply some torque to it neighbor before it reaches equilibrium and that will be enough to move the other wheel past its sticking point?

It's an interesting variation. I've tried designs in the past that would use the interaction between neighboring wheels to produce a chronic state of imbalance, but could never get them to work. Maybe you'll be luckier than I.

Below is attached an idea that I had for such a wheel. The idea is that when the telescoping weighted sections collide between the wheels, they will be pushed toward their respective wheel's axle and then each wheel will be overbalanced on the side away from the collision area between them. Result...continuous motion.

However, it takes energy to compress the springs or gas (green) that restores each telescoping weighted section to its starting position. Most likely, the imbalance in the wheels would not be able to supply that energy and the wheels would not turn.


ken

[LustInBlack]

Ken : Your idea is good too, but as you say, springs always win !

I found a way to unlock the wheel ..

I'll use "pacman" weights .. I hope simulation will not be too affected by the change, I'll post it later ..


Btw, to explain why my wheel doesn't reach equilibrium, you have to see the movement of the falling weights.. They "disconnect" from the wheel when they fall, because they are on a moveable pivot, and that pivot doesn't receive the pull of the weight because of the long arms under those specific weights [transparent arm of other wheel]. BUT, the other wheel, will "connect" with the previously unconnected weight, because the long arm is under the falling weight and receive the torque !

So, wheel right turns, and when it turns, the weight of the left wheel will fall a bit lower, at the SAME TIME, the left wheel does the same thing, but with the right wheel .. Two symetrical mecanisms .

That doubles the torque because the two wheels are now connected with opposite wheel connected weights and they turn at the SAME TIME ! .. The Ratio of this specific model is 3:1 .. The difference in speed results in a lock, and I think I found a way around it . The weights of the model are 30 Lbs, so it should do 3 times the torque, unless there are 3 weights connected, which can be controlled by moving the wheels toward each other so the arms "connect" with more weights .. In that case, the wheel is much more efficient .

It's hard to understand the way I explain it, I'll try to make that clearer a bit later tonight when the fix is ready ! ..

I'll be back !

[Fletcher]

Ken .. imo your dual wheel mech clearly shows all that is wrong with the symmetry approach to finding a solution to PM, as you point out. It has symmetry (at least how I think of it) in that there is always an equal amount of back torque in the system to balance the torque generated by the device.

There _appears_ to be no overcoming it. Stevin's points this out (see museum of unworkable devices) by saying that any device that is segmented & where the mech starts & ends its position, each segment, in the same place cannot achieve excess torque. Others would call it the conservation of energy law at work.

Your wheel design shows this very well. On another thread we briefly discussed the concept of a permanently oriented/displaced CoG related to individual mechs & whether they could provide torque to prove PM. Your comment was "they must do" providing you could mitigate the back torque issue that normally arises from the use of ramps, or in this excellent example, the compression of the telescopic rods. I look forward to discussing this further.

Soon, across on my 'symmetrically balanced systems thread', I will post a design that will show you that even with no apparent physical back torque evident & a permanently displaced CoG, it appears it will still not work as you might expect. I'll build the pictures over the next day or so & post them up for discussion. I think the concept should generate some interest & perhaps some lively discussion about this whole concept of overcoming torque symmetry, or not.

LIB .. I too am also having trouble visualizing your design & its workability. Perhaps it will become clearer to me as the thread develops further.

[ken_behrendt]

LIB wrote:

They "disconnect" from the wheel when they fall, because they are on a moveable pivot, and that pivot doesn't receive the pull of the weight because of the long arms under those specific weights [transparent arm of other wheel]. BUT, the other wheel, will "connect" with the previously unconnected weight, because the long arm is under the falling weight and receive the torque !

This disconnection process is probably what will render your design unworkable. As the weights on each wheel disconnect, that wheel will then become heavier on its other side. The effect on the two wheels is that they will want to rotate in the opposite direction from which you expect them to turn. Thus, the disconnection will create a CCW torque in your left wheel and a CW torque in your right wheel. I suspect that these counter torques will exactly equal the desired torque created in each wheel by the weighted arm of the opposite wheel resting on its extended arm.

You need to play around with the design for a while until you realize what is happening. I suspect that you will see what I am referring to after you solve the "lockup" problem.


ken

[LustInBlack]

Ken :

I believe that you are right if my weights were not falling .

When the long arms reach the weights, they must be a bit below it, so the weight have some space to fall before hitting the Arm . That happens both side at about the same time . That makes the two wheels turn in the desired direction ..

That will happen everytime one Arm reach the weight spaces ..

I don't think there is counter-torque there, because, the hit of the weight, with the effect of leverage, will make them more heavier for their opposed weights .. And, the contrary would cause the wheel to try to lift a heavier weight ..

If the wheels try to lift the weight that is falling on the long arms, I agree that energy will be lost, but, because of leverage effect, again I don't think it should happen, but that is left to verify ..

I have to modify one thing, it's the synchronization of the arms and weights, if there is idle time, when a weight doesn't fall on the arm, then, the wheel will move only on inertia, and that is bad for lifting a load .

[LustInBlack]

I post this simple model so that everybody can simulate and test it .

In Red, a Weight of 60 Lbs that simulate the weights that would rise around the wheel.

In Blue, the Arm-Weight System, 30 Lbs. It's action is to pivot from it's position -> to the Black Stop.

Wheels will turn according to the Green Arrows.

Brown Arms are attached to the rim of the wheels and create the Leverage.

The two weights hits the brown arms and rest on them causing the wheels to rotate.

When they pass the brown arms, you can clearly see that the 60 Lbs weights takes control and balance the wheel, that proves that the simple 30 lbs weights can and do rotate the heavier 60 lbs weights .

Simple Leverage ..

Now, if you take just one of the wheels, the 60 lbs weight will be heavier than the 30 lbs.

I believe, that, if we take ANY mecanism that we thought possible, for example, the simpler one that move the weights near the center on the ascending side, we would increase the efficiency of this principle, making it a better wheel..

Maybe that's why Bessler mentionned he could make his other doomed wheels work .. (That's what I've read here on the forum, I didn't read that from an official source).

Augmenting the distance between the wheels makes them more powerful.. According to Bessler: A bigger wheel will provide more power.

Put the mecanism in a bigger Wheel, and the diameter will be longer thus, bigger wheel.


I hope I'm onto something . . . .

[ The reference from Maschinen Tractate are just to show the similarity I've found . ]

Pease Understand this : The problem with Leverage is Speed vs Torque ..
BUT, with my wheel this problem is solve, the free falling weights on the right side of the wheel, are solving the problem, because they fall faster then the lifting weights .. Gravity at work .

[ken_behrendt]

LIB...

Well, I thought that making a WM2D model of your "Immersion Principle" would be a waste of time, but I'm happy to report that the results surprised me!

Each of my two large yellow wheels weighed 10 lbs and the small gray weights are 5 lbs each.

I started the simulation Running with the lever mounted gray weights each resting on the extension from the other wheel. The result was that I got nearly 3/4 of a complete rotation from each wheel!

You might be on to something with this!



ken

[LustInBlack]

:)

Ken, coming from you, that is very encouraging ! Thank you !!


I had an idea to fix the problem (- The sync issue -), my first solution wasn't all that good, Idle time between Connection arms, I may try a gear system .. But ideas are welcomed . .


You said you don't think this is the Bessler principle!? .. Why !?

By the way, I modeled exactly what you did, and I got more than 3/4 of a turn, the weights hit the arms at about 15-20 more degrees than 3/4 .

Changing the Leverage will probably make 1 complete turn or more .

If true, than it's even better .


EDIT : Well, you know what!? .. Even more than 1 turn ... If only I put a latch mecanism on the arm !