Toad Elevating Moment

[MrVibrating]

I had a think, and realised we can strip the key dynamic down to its bare essentials.

So i did.

The results are pretty spectacular, even if i say so myself...

Now we're really getting a wriggle on... :)





Last one to build it's a sissy :P

[MrVibrating]

IIRC it was jim_mich. ETA: http://www.besslerwheel.com/forum/viewt ... 3254#13254

In the mean time, short of a rack & pinion appearing, which btw is really hard to build in WM [well, I couldn't do it anyway, using gears function], try selecting each element > Properties & increasing static friction & dynamic friction from 0.3 to 1.0 - see if this helps.

ETA: direction rectifier / ratchet in attachment.

Thanks mate, i had played with friction but wasn't getting much luck with reducing slippage, unfortunately.

It's a shame because i think some of the configs with high slip were/are significantly OU.

However we can now scrap the previous config; V2.0 cuts right down to the chase, eliminating all the excess baggage.


Still very preliminary but it's defo a symmetry break between input via one type of torque, vs output via another. Rolling and rotating, in alternate steps. Big flywheel radius to align the inertial curve and it's just a matter of tuning the pole weights and wheel mass to an optimum smooth speed.

It should be a cinch to synch a bunch of these on a rotor... you could have upteen power pulses per rev.


Happy days! :)


ETA: cheers for the ratchet BTW, that's bound to come in handy..

[Fletcher]

Looks like it might be a 'zero runner' to me.

See 3rd sim on right - changed out the center square pin for two circular pins - I often swap out parts [simplify] or build another way to cross-check - also try altering the position of the center square pin to somewhere else & see if it still behaves the same.

Just My Suggestions - not purposefully trying to be negative but realistic.

[MrVibrating]

Probably should've uploaded this one first - starting with zero input PE (minor detail)..

The previous one is unrealistically fast as it starts with significant input energy...

[MrVibrating]

Looks like it might be a 'zero runner' to me.

See 3rd sim on right - changed out the center square pin for two circular pins - I often swap out parts [simplify] or build another way to cross-check - also try altering the position of the center square pin to somewhere else & see if it still behaves the same.

Just My Suggestions - not purposefully trying to be negative but realistic.

Nope i appreciate your caution, and i do apologise for my lack of restraint, it's hard not to get excited tho... ;)

And don't take any config i post as a definitive demonstration of the principle - just tweak away at the margins, whatever your intuition says.. generally, if the radius gets bigger, so does the gain, so play with the inertial curve relative to the pole weights - for instance increasing the pole weights can increase or decrease the gain. Adjusting one variable changes others, but we're definitely looking at an asymmetry...

In v2.0.1 here, 20mj from the first bar, 50 by the second, then peaking cyclically at 1.9J, perpetually. From a standing start. Net mass 9kg. Without solving the slippage problem, yet.

In v2.0.2 i've set air resistance to high, so we're definitely doing work. I opened up the asymmetry a little by widening the flywheel. It's that simple - the gain is a function of radius, just as Bessler said. With air resistance disabled it accumulates >20mj gain per bar.

Either the sim's wrongly confirming my hypothesis again, or this is really happening...

Incidentally, it would now appear that the reason his weights were hollow lead cylinders is because they're strung together into a chain. And THAT's what A is in MT138-141..!

[Fletcher]

Yes, I'm cautious because many of us have had zero runner's in the past - few times, if any, have I read a fleshed out hypothesis to back up anomalous results in a sim, so I really do wish it is true & you are correct & it's not an artifact zero runner.

FWIW - If I get a sim that's marginal, besides building it different ways as a cross-check, I then go to adding a load - btw the Air Resistance Low is based on m^2 & High m^3 - that doesn't mean too much at very low rpm but it works very well [Low] for faster rotating sims with more surface areas in contact with the air drag.

Sometimes I introduce some pin frictions or just add a friction wheel to the circumference - if you need one I'll dig one up for you - if it still accelerates then you have a winner, a PM runner.

Your enthusiasm & rapid development is fun.

[MrVibrating]

If A is strung weights, and B is obviously the axle, C and D are the square, and circle.

The hammers must be for punching yourself in the head.

Squares and circles. Circles'n'squares.

Not one or the other, but BOTH at the same time; a square and a circle.

No wonder it took 300 years!

So any takers? 20 grand, the pair. Tell you what, 'gis 10 grand for the circle and i'll throw in the square free, can't say fairer than that..

It's so simple it's deviant. We can drop one weight and lift 4, 10 or a hundred - how long's a piece of string?

So provided there's no imminent shortage of circles and squares (together, at the same time) i think we may be onto something...

[MrVibrating]

Yes, I'm cautious because many of us have had zero runner's in the past - few times, if any, have I read a fleshed out hypothesis to back up anomalous results in a sim, so I really do wish it is true & you are correct & it's not an artifact zero runner.

FWIW - If I get a sim that's marginal, besides building it different ways as a cross-check, I then go to adding a load - btw the Air Resistance Low is based on m^2 & High m^3 - that doesn't mean too much at very low rpm but it works very well [Low] for faster rotating sims with more surface areas in contact with the air drag.

Sometimes I introduce some pin frictions or just add a friction wheel to the circumference - if you need one I'll dig one up for you.

Your enthusiasm & rapid development is fun.

Cheers mate, however i think development is pretty much wrapped up, finished, done and dusted. It cannot physically, geometrically, get more refined. The whole system reduces to a gravitational / inertial interaction between a square, and a circle.

Nothing left to do but build.

[Fletcher]

Go for it - then for me, the fun part begins - retro-ing the sim to the build & seeing if they align - then optimize with the sim.

Best of British.

P.S. JB also sought a solution for squaring the circle lol.

[MrVibrating]

Ooh there's potentially another, better implementation - have a bigger square tumbling inside against the rim!

Or even have two sets, one on the axle and another, larger one in the drum!

This is harder to sim but i'll have a go tomorrow... this has actually been a hunch for a while, so if it works out i'll be gutted i didn't try it earlier...

[Fletcher]

That would be a hamster wheel analogue - one runner a day is enough ;7)

[MrVibrating]

Go for it - then for me, the fun part begins - retro-ing the sim to the build & seeing if they align - then optimize with the sim.

Best of British.

P.S. JB also sought a solution for squaring the circle lol.

Yep and i'll your bottom dollar this isn't gonna work in the real world, just no way. Nuh-huh. It's crazy.. but crazy enough?

Wild night... gotta work in a few hours, goin' bed...

[MrVibrating]

Last thought for the night; yep i thought the "squaring the circle" foray was a likely lead - and the actual form of the problem may yet speak to the asymmetry here - and perhaps, if so, Bessler had understood this. He wouldn't have known what a transcendental number is, of course, or that the squaring was impossible, but perhaps there is a causal relationship... i'm no mathemetician though...

[jim_mich]

Hey, I downloaded your Doh! file and ran it. The two circles are not quite centered on each other (0.001 off) and the four bars are not centered either. This is what gives it its initial rotation. I added a graph to chart of wheel velocity to see what was happening.

I have a very fast computer and it was taking a long time to run, so I checked and found that the program was set to limit running speed to real time. It would have taken about 27 minutes at the speed it was going to reach maximum cycle limit. So I switch off the real time and it took less than 1/2 minute to run until it hit the max cycle number (about 32767).

I copied the velocity graph and am attaching it.

It seems this has an initial OOB of the 'box', which starts it to rotating, then it simply coasts.

[Fletcher]

Thanks jim_mich - I was having a bite to eat & running the graphs also.

Note that in zDoh_2 I swapped out the rigid center square pin for two circular pin joints which usually clears up an anomaly.

I too thought dropping it was giving it some KE to get a start & then it coasts with no losses - I guess starting with some PE converted to KER is no different from a small impressed push start.

It's getting up to quite a clip.