Toad Elevating Moment

[MrVibrating]

A slightly optimised version:

http://www1.zippyshare.com/v/28934344/file.html

It likes inertia and a smaller inner axle, apparently. Around 36kg here for ~1J on the rotor so crap energy density but hey, early days eh.. and this is still without pumping iron! It's just a frame that sits on the axle, tries to roll off it and accelerates it in the process, before flipping 60° and continuing, ad infinitum. Which is nice.

So where to place the weights? Looks like about a dozen possibilities, maybe, to try out.

And this is just one mechanism... MT137 has four of these!



ETA: Try tweaking the inner axle width up to .120 to.130 - 0.125 seems to be around the sweet spot, for a 1.5J peak output.

Eventually.

Yes, it's slow. But hey it's unloaded. And it's only day 1. Besides, as noted earlier, direct drive systems like this have a preferred speed that resists further acceleration... and anyway it's not supposed to move in the first place... i'm still not entirely sure why it does. I just drew the frame as illustrated, bolted it together and off it went. Crazy eh. Yet somehow i wasn't surprised.. Still, whatever the reason, we've obviously barely scratched the surface... Bessler's must've been putting out 20W or more..?

[jim_mich]

Mr.V, why don't you upload the files directly to the forum? That way they be available as long as the forum exists. When you put them on external servers, they might not be available at some point in the future, thus making the links to your files go dead.

Also, I noticed your WM2D file is very large (4.85 MB) and the forum has file size limits. To make the file smaller, simply modify the WM2D assembly by adding a pin to the background then removing it. Any change to the assembly will wipe out the history of all the motions. Then save the WM2D file and it will be very small because its history is empty.

[MrVibrating]

Ooh didn't realise that was an option.. And i'd been wondering why the files were so large! Thanks for that... i'll look into it, cheers..

Edit: ah got it, ta..

[Fletcher]

Hi Vibrator ..

I was expecting to start by saying "nah, sorry not this time" so as to not overly discourage you, especially after practically pulling an all nighter.

Then I zoomed in on your axle & watched in fascination as the mech settled down - it was mesmerizing - suddenly I was having thoughts of gravity waves flexing the long arms, momentum & stiction etc, much like you I suspect.

So, I went back & changed the accuracy to 50 fps & it continued to rotate at about the same pace but settled in the opposite direction to 20 fps - no big deal, just better accuracy at higher frames.

Then I decided it was time to add some dissipative forces - the quick & dirty way to do this is just turn air drag onto LOW - this introduces some frictions or a pseudo LOAD even if it is small at low rpm.

Well done, keep tweaking & maybe you really will have something there - I love its simplicity & elegance & it sure would be a major face palming moment for me if it were as simple as this.

[ruggerodk]

---anyone fance to upload a screenshot, please?
wm2d is not on every pc (and mac) ;-)

regards
ruggero ;-)

[unstable]

Hi MrVibrating, your design is very clever and I like it. Maybe this is the one. It remind me the image (in Bessler documentation ?) of the compass and the ruler (90°) overlapping. Even the image "and you still not understand?" have similitude with your design. In my opinion you are on the right path !!
good luck !

Claudio

edited, I attached a screenshot of MrVibrating's idea for ruggerodk and others.

[rlortie]

---anyone fance to upload a screenshot, please?
wm2d is not on every pc (and mac) ;-)

regards
ruggero ;-)

I would if I could, but your probably already aware that my "Windows Explorer" has crashed putting a crunch on all picture management.

Ralph

[ruggerodk]

I know, Ralph - hopefulley you solve the problem soon ;-)

Maybe Fletcher could help?
He's pretty good on wm2d...but got a slow net-connection ;0

[MrVibrating]

Hi Vibrator ..

I was expecting to start by saying "nah, sorry not this time" so as to not overly discourage you, especially after practically pulling an all nighter.

Then I zoomed in on your axle & watched in fascination as the mech settled down - it was mesmerizing - suddenly I was having thoughts of gravity waves flexing the long arms, momentum & stiction etc, much like you I suspect.

So, I went back & changed the accuracy to 50 fps & it continued to rotate at about the same pace but settled in the opposite direction to 20 fps - no big deal, just better accuracy at higher frames.

Then I decided it was time to add some dissipative forces - the quick & dirty way to do this is just turn air drag onto LOW - this introduces some frictions or a pseudo LOAD even if it is small at low rpm.

Well done, keep tweaking & maybe you really will have something there - I love its simplicity & elegance & it sure would be a major face palming moment for me if it were as simple as this.

Cheers mate (and everyone else!). Obviously when i said 'unloaded' i mean unloaded with moving weights - with only a few microjoules per rad it's not going to handle much dissipative loss at the moment.

One clear limitation here is that the jack is currently not 'jacking', much. Presumably there IS some jack action, somewhere - a small amount of unintended flex, perhaps, and this is resulting in OU, however Bessler was clearly claiming 4x OU and had the performance to back it up.

So, somehow the 'perpetual motion structure' here has to be somewhat more animate. It has to 'give' somewhere, extending in some sense, perhaps compressing a spring in the process (per the Toys page).

Also, this jack action has to facilitate the exchange of a pair of weights from center to rim. Why paired weights, i do not know.

Presumably it will become apparent that hollow cylindrical weights are best suited to this, but for now i can only guess why that might be.

I think MT41 and 138-141 hold the last remaining keys...

[murilo]

Hi MrVibrating, your design is very clever and I like it. Maybe this is the one. It remind me the image (in Bessler documentation ?) of the compass and the ruler (90°) overlapping. Even the image "and you still not understand?" have similitude with your design. In my opinion you are on the right path !!
good luck !

Claudio

edited, I attached a screenshot of MrVibrating's idea for ruggerodk and others.

No words to describe how much I like this clever design!!!
Thanks Claudio, thanks MrVibrating!
My regards to MrsVibrating!

[ruggerodk]

Appreciate the screenshot Claudio ;-)

Tell me: Are these rigid 3 x 90 angled 'jacks' pined or pivoted somewhere? Or just sort of laying on the axle?

regards
ruggero ;-)

[MrVibrating]

Good question - originally only the 90° angles were hard-pinned, and the inter-connecting angles pivoted. This evening however i've experimented a little with allowing these square angles to move a little, under sprung tension, as MT138-141 seems to be instructing (my interpretation here may be way off, but it seems to be working out so far).

This does offer some improvement - i can now get up to 2J with enough tuning; the ratio of the axle circumference to that of the Wankel (i haven't bothered to look up the proper geometric term yet) it tracks round seems to be important - it's highly sensitive to these adjustments. This makes sense as it dictates the travel distance of the direct-drive portions of the cycle, relative to the lengths of straight beam riding over the axle.

But adding these springs - and thus allowing the 90° joint to flex slightly - has also shown that there's an over-balancing moment. In the attached model, i've trimmed the wheel radius to match the length of the weight poles, so that you can see where they overlap.

Watch closely the top section, between 10 o'clock and 2 o'clock, and you'll see the overbalancing mass disappear, whack-a-mole style, from the upwards side, to re-appear 90° opposite at the downwards side. Hence, the overbalancing mass moves through the wheel, poltergeist-like, under the inner-axle and back up again, translated 90° out of phase. The poles have only shifted radially a few mm, but this is enough to transfer the wheel's center of mass to the opposite side of the wheel.

Increasing this overbalancing moment looks like a tangible goal.

Increasing the direct-drive moment currently comes up against stiction - if the axle is too thin, it can't ride over the 90° angles, and so slams into them and stalls. However even if it slips a bit (as current iterations do), we're still losing valuable torque.

If we simply make the axle larger to compensate, we sacrifice direct-drive torque. So smaller axles are better, but have their own attendant problems to be worked out.

I can't yet get to grips (sorry) with friction in WM2D, but tried adding little ramps to make the 90° bump shallower to negotiate, however this also decreases the wankel's track length, so there's diminishing returns there. A better option might be gearing or just a few little teeth bolted on, to prevent slippage. However looking back at MT138-141, those eyelets on the axle might imply some kind of semi-rigid coupling, perhaps obviating the stiction problem... something to look into... FWIW i've already tried square axles, and they didn't help, though perhaps another shape might be better, dunno, WIP..

I'm pretty sure we're on the home straight though. So there's two problems for folks to think about:

- how to reduce axle slip

- how to boost the OB moment

Hopefully these are eminently solvable design issues, and we have the form of the excess impetus nailed already.

And BTW, MT135 has just slotted into place - it's the same arrangement of square-intersecting weight poles, this time showing how they collectively overbalance.

It now seems increasingly obvious, given this evening's findings, that there are no additional discrete weights required to be placed anywhere. The cylindrical lead weights simply fit over the poles, to weight them. Pure lead poles would be too flexible, hence lead cladding on probably iron poles was the most logical alternative. The weights don't move on the poles though, only the poles themselves need to move.

Kebab time. Later dudes..

[MrVibrating]

Hi MrVibrating, your design is very clever and I like it. Maybe this is the one. It remind me the image (in Bessler documentation ?) of the compass and the ruler (90°) overlapping. Even the image "and you still not understand?" have similitude with your design. In my opinion you are on the right path !!
good luck !

Claudio

edited, I attached a screenshot of MrVibrating's idea for ruggerodk and others.

No words to describe how much I like this clever design!!!
Thanks Claudio, thanks MrVibrating!
My regards to MrsVibrating!

Cheers guys, exciting innit.. :)

[MrVibrating]

Lol just wasted half an hour staring at the zoomed-in axle over a beer - like Fletch says, it's mesmerizing..

The thing is, the sequence is perfectly causal. Perfect cause and effect. It simply must rotate.

Every step, perfectly consistent with CoE. It rotates because of it, not in spite of it! It's not like there's some magical, inexplicable trick in the sequence.

Surely this can't be violating anything? It satisfies everyone from Lorentz to Noether and Joule, Newton and Einstein. It's hard to even say exactly where the symmetry break lies... (?)



The only question is, where does WM2D get it's energy from?

[MrVibrating]

"these come now always two and two" - three pairs of two, specifically. "arranged one against another..."

Just watching the axle, it all falls into place... "..poles or parts sliding over eachother.."

8 thumps per cycle = 2 OB phases per jack for one full rotation of a four-jack wheel (ie. MT137), because of the 270°/360° power ratio split between the axle and jack.

The suggests the poles are prevented from over-extending by their collision with the rim... rather than springs alone, perhaps... maybe incidental, or not... something to investigate there...