In for a penny, in for a pound.

[agor95]

Hi Robinhood46

I have tried numerous methods to get the wheel and the small weight to interact to achieve the reset, they have all failed.

The speed of the arm rotation is impressive. This is similar to well balanced wheels being rotated by a dropped
pendulum that then runs free from 6 o'clock. I think that last one show 20 rotations before stopping.

It's like building a device out of a mirage.

Regards

[johannesbender]

You could try a stationary downward ramp to deflect and guide the mass as it comes down on the right side , which should guide it lateral (if theres enough KE left) .

[Robinhood46]

You could try a stationary downward ramp to deflect and guide the mass as it comes down on the right side , which should guide it lateral (if theres enough KE left) .

I have tried that already JB.
That is one of the first things i tried, because of how easy it is to miraculously place objects, that just hover in space, and only interact with some objects, but still allow others to pass through them, in the sim world.
It is because of all my attempts, that i think the problem is the small weight's participation in rotating the crossbar. It stops participating as soon as you apply any force to change it's trajectory. It is as if we need to find a way of changing it's trajectory, without taking away the effect it has on the counterweight. This is why i think an additional mechanism is needed, either to stop the counterweight from becoming much heavier (it's effect on the crossbar), or avoid reducing the effect the small weight has on the crossbar. We are playing with a tendency change, the light weight's effect on the crossbar is critical for the tendency to swap from one end to the other. Take the effect of the small weight out of the equation and the heavy weight becomes much more (a lot, loads, exaggeratedly) heavy compared to the light end. We need the domination to be hovering from one end to the other, no end should be considerable more dominant at any stage of the operation.

[agor95]

Hi Robinhood46

We need the domination to be hovering from one end to the other, no end should be considerable more dominant at any stage of the operation.

I was thinking the exact opposite by going full trebuchet.

Which means looking at pulling the pendulum arm back and letting it swing.

You may have noticed that the trebuchet projectile is place behind the main arm.
This results in the projectile being pulled along and swings like a flail up and over.

Well we want the over for that is the reset.

So think on the arm with the main mass at the top just as it drops to give that fast clockwise rotation down.
During this fast first quarter rotation we could rotate the pendulum anti-clockwise and rebounding off the main arm.

Then it would be moving clockwise with extra gravity assist and propelled by the trebuchet arm up and over.

Regards

[Robinhood46]

Hi Robinhood46

I was thinking the exact opposite by going full trebuchet.

Regards

I think the only way to go full trebuchet, is to make the trebuchet totally independent of the wheel. So that it throws weights onto the wheel and then by some mechanism the wheel arms the trebuchet. But i can't see it.

[agor95]

I think the only way to go full trebuchet, is to make the trebuchet totally independent of the wheel. So that it throws weights onto the wheel and then by some mechanism the wheel arms the trebuchet. But i can't see it.

I am keeping it simple. That means no wheel. So after the pendulum has been flailed up and over into the reset position.
Then the arm is balanced and rotates around for the next accelerated rotation cycle.

Regards

[agor95]

You could try a stationary downward ramp to deflect and guide the mass as it comes down on the right side , which should guide it lateral (if theres enough KE left) .

Well there is no reason why the pendulum could be lifted by an upward ramp along the bottom.

Regards

[Sam Peppiatt]

One last thought, FWEIW.
Add some counter weight to the pendulum @ 90 degrees. At 3:00 it would have an "L" shape with the short leg of the "L",(counter weight), pointing down, if you haven't tried it all ready--------------Sam

ETA, This would make it easier for the bell-crank to reset it.

[Robinhood46]

I have only just seen your comment Sam, i don't know why, maybe i clicked on "mark all posts read" without seeing it.
I am not entirely sure what you mean, but it sounds something like when i tried to have three arms, two with counter weights and one with the swinging weight. My thoughts were based on the apologia drawing, i couldn't get that to work either.

I'm wondering if the toy page doesn't come into play, with the woodchoppers "being used differently", or whatever it was he said.
Agor mentioned a while back, maybe the moving the weight back out, to reset it, is done in a different way than just "swinging back" the same way it swung in but in reverse. He suggested the weight being on a rope or chain, and then moving on runners or ramps, if i remember correctly. I too thought about something along those lines, but have a lot of problems when i try to sim ropes. I can get one arm to function with ropes but multiple causes bugs or glitches.
The woodchoppers having their axes on the end of a rope could explain the difference Bessler was talking about, but I'm more inclined to go with some kind of variation of the linking rods.
Maybe, instead of a counter weight, there is the opposing light weight, that is shifted back out in a manner that shifts the weight, the rod, and the pivot. As apposed to simply trying to reset the weight alone. This would give us the swinging in of the weight after 6 o'clock, as seen in my recent videos, which causes the change in domination. Instead of trying to reset the weight with the pivot staying put, the weight is shifted to the rim and the pivot is brought into place for the reset.
This would give the weights a path similar to the teeth of a circular saw. Which is what i was trying to do on the "Gravity wheels with a fundamental difference" thread, when i first started it.

[Sam Peppiatt]

Robinhood,
I was referring to the little short arm at the very end of the long swinging arm. It looks to be maybe six to seven inches long with a small double weight at the very end of it. When things come to a stop it hangs more or less straight down, like a pendulum.

Instead of straight, make it "L" shaped, with the pivot point at the corner of the "L". Add a small weight to the short leg of the "L", (the short leg hangs down). This will cause the long leg of the "L" to swing upwards to say about a 45 degree angle, instead of hanging straight down. This will make it a lot easier to reset it, with the bell crank. If; if there aren't too many bad side effects, that is.

Every thing would have to be re balanced because of the added weight to the pendulum---------Sam

[agor95]

Hi Robinhood46

I'm wondering if the toy page doesn't come into play, with the woodchoppers "being used differently", or whatever it was he said.
Agor mentioned a while back, maybe the moving the weight back out, to reset it, is done in a different way than just "swinging back" the same way it swung in but in reverse.

I have a few 'being used differently' with the Toy Page and the AP wheel.

However what was in my mind was three of your trebuchet like units. Where the long and short sections are rods.
When a unit is straight they are balanced. But when the unit is acting like a trebuchet it's 4:1 ratio.

The projectile section gets pulled up a chute along the bottom like most trebuchet have.
With a slight ramp up. The ramps purpose is to assist in getting the unit straight when the main mass is at 6 o'clock.

A joint stop keeps it straight until the projectile end circles around to the bottom chute ramp again.

Regards

[Robinhood46]

A drawing would be nice.
A simulation even better.
If we could get in each other's head, i think we would have solved this puzzle years ago. We often explain very simple mechanisms, interactions or modifications, the only reason they are simple, is because we can see exactly what we a talking about.

[agor95]

A drawing would be nice.
A simulation even better.

I agree with all your points. Just need to get myself motivated to build an animation to illustrate.

Regards

[Robinhood46]

This is getting more and more promising. At least i think it is.
If we consider the force with which the lighter weight is raised, once it has folded, because the heavy weight end of the arm wants to go down.
We can see that when i add additional arms, they all stack up against each other, because the heavy weights seek 6 o'clock. They can't find it because of the arm in front of it being in the way. The heavy weights pile up trying to go down and totally fill up the down side of the wheel.
I can't put ropes on them, because they will not work, but an MT9 style connection between the arms, might be enough to get the light weight to shift over where desired.
The stacking up of the arms can be ridiculous, especially if the weights are narrow cylinders. It is possible to literally pile them all on top of each other and totally fill one side of the wheel. The accumulated force of all the individual arms, seeking equilibrium, only needs to shift one light weight laterally, for it to take it's turn to apply force to the axle.
If a rope system with pulleys is pulling the weight over, then there needs to be a net dominance of the light weighted end when it is fully extended, to cause the wheel to turn and shift the light weight laterally.
If the accumulation of the arms all stacked up is shifting the light weight over, it needs a mechanism that causes only the desired weight to be shifted, not each and every weight. This would imply that all the arms are held apart when stacking up and only allowed to close fully to push the desired weight over.
I can't post a bloody image again. LOL

[Robinhood46]

I can't post a bloody image again. LOL

Here is a video which shows what i was talking about.
https://www.youtube.com/watch?v=RD5NeGFCo1A