Iris wheel

[jim_mich]

One of our new members here, racer270, has an idea. He was having trouble posting (his scanned sketch is to big) and asked me for help. I made a drawing of his idea which looks interesting. It looks like an iris from Stargate SG-1. As drawn here it is unbalanced and would rotate, how far I don't know. Also I don't know where the weights on the lower left would hang, so I just guessed for this drawing.

[Jonathan]

At first I thought that this wouldn't work, but it looks to me like all the weights on the left would work to lift the bottom most weight up prematurely. I don't know if that would make it work but it is an interesting effect.
EDIT
After looking again, the lifting of the bottom weight would cause it to hit the weight to its right, and so on, causing almost all the arms to be forced inward, except maybe a few around 2 o'clock. It really is an interesting idea, not easy to model.

[racer270]

Guys, this is my first post. I used an 8 weight configuration at 45 degree intervals it does rotate no external circle let the left side wieghts extend to the right side laterally, I used a 90 degree stop on all the lever arms so that the weight at the 3 oclock position was at the same height as its hinge point on its respective arc arm it picks up the bottom weight just passed the 6 oclock position , it starts on its own whenever stopped . The only thing required for a restart is to pull out one of the weights at the 2 oclock position with all the other weights extended on the right side the device actually accelerated its rotation lifting the two oclock weights sooner and sooner as it accelerated. The only problems i encountered was the left side regathering in lateral alignment for the "pinching in effect" it will need what I am considering guide plates on the sides of the arc rods and bearings accurately arranged for pivot points of the arc arms. I am going to build a 2nd model with these things I have learned with, left side radius equal to one, and the right side radius at 4 times the distance from the radius but it actually does turn as crude as it is. gordy

stand by this model took about 4 hours 8 ping golf balls some convenience store dislay hangers curved in an arc and 8 holes drilled through the 20 inch bicycle rim, forks still attached exact cost $0.00 dollars
i find that the inside circle diameter the lengths of the arc arms and the extension of the right side weights to be the keys, along with the piching and capturing effect of the left side ascending weights. As far as numbers of weights i am considering 12 or 16 wiehgts instead of 8 and them an equally blalanced fly wheel for extra weight. this crude machine can be built it does work. what do you say we kick it around and make it happen? replies and critiques are welcomed.

Gordy

[racer270]

oh yeah jim thanks for your hehp; i think 12 weights and no outside containment ring needed just limit right side weights 90degrees stops gordy

[Gravmaster2000]

The picture is VERY interesting, I may try one as time permits.
It would be fun to try different arm lengths (different overlaps)

The way Isee it now: The weight at 12:00 presses down 100% on the
arm of the next wieght, that leverage gives 50%of that force to it, +its own weight, 150%... I am curious to see what happens below 90 deg on the left....the action would get weaker pretty fast...

I have a neat idea I scheched out-will post soon..

[grim]

Welcome to the real world! Good Job!

The bottom uplift is always the torque, the upper is the kick and speed. You've got it down!

One suggestion, you might want to have your thrown (past 2 o'clock) weights aid in your upper (2 o'clock) lift, similar to what you did at the bottom.

Timing is the key factor, it's just like adjusting a car distributor. At the 2 and 7 o'clock position it will tend to lope, which is probably good. At 6 and 12:30 it turns into a monster.

Best regards


grim

[Jonathan]

Grim, what are you talking about?

[grim]

See "Racer's" post

[Jonathan]

Sorry grim, I reread your post a few times and it makes more sense now, I just couldn't follow it before.
I guess I should mention that I built a model this evening, and though I do see the pinching effect, it isn't at powerful as I first guessed. I've asked Gordon for more specific info, I probably don't have the right curve or something, but it is 4am there so I doubt I'll get an answer really soon.

[grim]

My bad, should've been more post-specific. Early here too, half a cup of mud only down.

regards


grim

[racer270]

its not that powerfull yet! I am going to try to make the right side radius 4 times the distance of the radius the pinch is about a 1/3 the length of the radius arm in overlap. I used 8 weights I am considering the next generation 16 weights..i.e.,221/2 degrees the key is to let the weights extend as far out as possible on the right side. I literally stopped the wheel a 3 oclock on every weight and adjusted the stop so it would not exceed the same vector/line of 3 oclock this allowed (using jim's 12 weight model) the 2 oclock the 3 oclock the 4 oclock and 5 oclock were all extended the 6 oclock was starting to lift at bottom dead center. Keep in mindthat this is the first concept the problem was a mechanical alignment of the left side pinching effect but it has achieved 3 rotations before a mechanical gremlin stopped the rotation..i.e.,weight being not captured or pinched, weight being captured under next lower arc arm or the weight catching on the bicycle fork. It will accelerate but guys this is still in the wright bros. type chronology. I can see it evolving into possibly a magnet to launch the 2 oclock weight, when centrifical force reaches a lift on all the weights from what I am seeing it would probably stop accelerating when it becomes balanced. (it's wright bros. now can you imagine what it will be like if we could turn it into the stealth bomber!!!!!)
Gordy....email me
p.s. remember to put a stop on the weights so they cannot fall any lower than their hinge point a 3 oclock position this will help also no outside ring let the weights on the right side extend as far as possible.

[Jonathan]

Oh, so my results aren't that bad. I thought you really had it going, but only three revolutions. I doubt I could get much better results because the plastic parts are so elastic that the flung arms bounce and shake the whole thing. I was expecting more pinching, I'll have to look but I think the 1/3 pinch to overlap ratio sounds right. Unfortunatly it seems to me that it is constant too, because if you increase the overlap you also increase the back leverage the 6 o'clock weight has on the others. I'll keep fiddling.
EDIT
In case I'm wrong, to what extend do you suggest overlapping?

[racer270]

1/3 over lap any more would pinch the 6 o.c. weight, the 7 o.c. has little force on the 6. (arm leanth ,inner hub size)and right side distence from axis all seam to be ie; 2/1 (going to try 4/1) and good alinement on the left to (pinch) the left side to a tite radios are all part of the big picture. bessler (lift 4lbs with 1lb) exc. i am rounding up parts to get past the exp. stage, understand it was a 4 hr exp incuding bild time. when i get to a working proto. that has no flaws i will show photos/vidio tell then a good simulation would help. ie; inner cercomfrence,1/3 overlap radious arm size and were to buy good parts for a (p. m.) machine are all good questions . ??? LOL gordy

[jim_mich]

I will attempt to write a computer simulation. I'm already thinking about how the program must work. It will need to be a 'solver' type program where it makes a couple guesses then checks which one is closest to being right then keeps narrowing down until the answer is very close to solving. Also should include inertia. The math and the trig will be complex. This is one of those cases where a good well built model might be faster. I expect it will take 30 to 40 hours of programing and debugging, maybe less if all goes well. But hey, this is what I love to do!!!

A good source of off the shelf parts delivered fast is http://www.mcmaster.com/ The price is fair. You won't get any bargains. The same item may be cheaper in your local hardware store. But the selection is super. They have materials such as plastic, aluminum, steel, pipe, tubing, etc. and materials such as bearings, screws, set collars, couplings. About anything you can think of. Their catalog is 2-3/4 inch thick and over three thousand pages. They deal mostly with businesses but anyone can order. Morning orders ship the same day. I usually receive mine on the second day out of Aurora Ohio. They have plants in Atlanta GA, Elhurst IL, SantaFe CA, and Dayton NJ.

[Jonathan]

I suggest home depot. Or K'nex on the off chance you have some lying around. I strongly suggest K'nex if it turns out that this device doesn't work, it allows reasonable quality rapid prototyping.
So there should be no pinching on the 6 o'clock weight? That's good, because I wasn't getting much of that. I've been fiddling for a few hours without much success. For a little bit I thought I was on the right track by restricting the swing angle to way less than 90, but I think that just made it turn longer because it was more like a flywheel and didn't waste all the energy in the banging. BTW, by 'turn longer' I mean after an intial push, I haven't gotten any overbalancing results yet.
I look forward to the outcome of the simulation jim.