Please use this thread for discussion on virtual modelling & calculations of the Abeling wheel.
The old thread suggested usage is general discussion.
Thanks
AZ
Just a simplified model in VB to calculate, give a rough idea, about the torque involved.
Straight slots, no "hockeysticks" at the end.
The track geometry, is almost the same, as my model.
One weight is calculated, the other 8 positions are added and stored, so after 1 turn, the combined torque can be "plotted".
So it is easy (for me) to change the weights, or the number of weights.
What is clear, is that as it is a static model, is that friction plays a big role. This is supported by the patent of Abeling, where he shows some attempts to minimize the role of friction in figure 6.
So in NO way this is a perfect model, it is full of holes. It shows to me that practically it is almost impossible, because of friction, to get the wheel to run on static forces alone. For me proof that it will not be a self starter for sure.
Quote from: eisenficker2000 on May 18, 2009, 11:27:07 AM
Just a simplified model in VB to calculate, give a rough idea, about the torque involved.
Straight slots, no "hockeysticks" at the end.
The track geometry, is almost the same, as my model.
One weight is calculated, the other 8 positions are added and stored, so after 1 turn, the combined torque can be "plotted".
So it is easy (for me) to change the weights, or the number of weights.
What is clear, is that as it is a static model, is that friction plays a big role. This is supported by the patent of Abeling, where he shows some attempts to minimize the role of friction in figure 6.
So in NO way this is a perfect model, it is full of holes. It shows to me that practically it is almost impossible, because of friction, to get the wheel to run on static forces alone. For me proof that it will not be a self starter for sure.
Cannot seem to run this:
MSCOMCTL.OSX is missing?
Anything I can do in WM2d while I wait for materials for another project?
Would be happy to give another go at a few models.
AZ
Wow! That is a nice little program. I envy you your skills with VB!
M.
@AquariuZ try it after installing this: Vbrun60sp6 from
http://www.microsoft.com/Downloads/details.aspx?familyid=7B9BA261-7A9C-43E7-9117-F673077FFB3C&displaylang=en
Quote from: eisenficker2000 on May 18, 2009, 05:47:22 PM
@AquariuZ try it after installing this: Vbrun60sp6 from
http://www.microsoft.com/Downloads/details.aspx?familyid=7B9BA261-7A9C-43E7-9117-F673077FFB3C&displaylang=en
It works, thanks.
Impressive program... I am still thinking on how a spring could connect two sets of dumbbells and release its force on the left top side to launch the upper part of the set.
Comment by Cloxxi:
Quote
Springs of course press as hard back as do press forward. Would you propose using a "locked" position in the rise column to lose less on the back end than is won on the front? And would the second weight then be dragged along?
I may be mis-understanding oyu, but it sure is interesting.
I've done some sketching, based on non-connected weight, and found that the slot configuation can make vital differences in how the weights and wheel behave.
My current understanding is, that a flow can be devised, where velocities, heights and potentials all work in harmony. It's a game of phases and their counter-phase. Getting the most from the lower ramp. For instance: high release velocity at start of bottom ramp. The wheel can't help past this point, and IMHO shouldn't. Just let the weight roll on on its own some. To get optimal velocity at the lowest point, the counter weight will better not be taking too much energy from the wheel at that moment. So either it's just being slammed in its hook, or it's at least not draining energy from the lower weight anymore.
Next idea that's becoming more persistent in my mind, is that the "shot put" action might have to come from the wheel's inertia, not from the other weight. The wheel may need to have some positive mass to it. In case of 8 weight, perhaps this factor will be less of a problem.
HOWEVER if I would be correct, that we need optimal speed at the lowest point, then Abeling's explanation makes sence : "2 weights, one doing the work to raise the other", or something along that line. In a system of 1 wheel and 2 weights, when on it off the hweel, the other weight and the wheel together can do some serious work for a period of time, thus for a number of degrees. The speed variance that we see with just 2 weights MAY be what's making this possible to begin with. The wheel's weight vs. that of the dumbells may need to be very well balanced to get the desired effect.
Oh, about phases. The timing of fases will be hugely different for a Dusty-style thin part-of-circle slots, versus "spokes", and let along "offset spokes".
My sketches and visualizations seem to now point out that we want the 6-8:00 weight to get BEHIND. The getting behind saves serious energy, allows the wheel to store that, and release it when the slot has the raising weight play catch up, aided by a counter weight in optimum pulling phase, and a wheel ready to give a nice nudge in the back.
(The "getting behind" feature would totally complement my idea of pivoting rods. I proposed a setup before, where a pivoting rod, at roughly 75% of it's length, the weight being at the very end, would allow the weight to comply to the circle, and catch back up when so required to. Weight would not really be "off" the wheel, just not taking energy from it on the first part of the upswing. If the rod IS a spring, or incorperates one, there's some more explosive energy to be stored and put to good use.)
Anyway,
Short : the slots might need to start at 0-3-6-9:00, and at the hub be well off-center, CCW (back). The part of circle shape is nice, gives dramatic slinging at the top, but may be "over the top". And, get the timing wrong, and it will just fall apart.
I wish I had the skills to draw and simulate this like you guys.
Regarding the weights, I'd suggest using roller around the bearingss to achieve greater diameter. Slot will need to be wider too. Friction will be greatly reduced. Much less spinning and bouncing. Bearings themselves will work more efficiently, too.
Cloxxi,
Imagine a round spring with no ends and two weights attached to it where the distance between the weights is half the length of the spring.
Now imagine an egg, place one weight on the top of the egg and one on the bottom, coming sideways, now start turning the weights CW.
This is the movement I am thinking of, where somehow the contraction of the spring between 9 and 11 o clock gives an extra boost to the top weight, effectively shot putting it into the curve which ends the holding shaft.
Bit hard to explain.
Something like this.
The dumbbells are connected in pairs with a spring through the shafts
The dumbbells are pulled around the egg
The spring releases (contracts) and ejects the dumbbells
8 dumbbells with a hole in the shaft
4 springs
One egg
Omitted the wheel with the hockeysticks but yes, this is still part of the setup.
Acceleration starts after 9 as the spring contracts and is at full contraction at 11.
Abelings words:
"The weights are applied two by two: one weight is pushing/falling, the other one has to be lifted. Due to the invention of the dual lifting system , the falling/pushing weight will hardly be hindered by the weight that has to be lifted. In the topleft of the system the weight is accelerated (like with shot put). The weight is moving faster than the system, and as the system catches the weight it is propelled forward."
I say the ommitted part from the patent is a simple spring system.
What if there are eight springs with hooks, one hook connected to a ring on the shaft, one hook connected to the shaft of the dumbbell?
Thats even better.
Now how do I start to simulate this.
Hmmmmmmmmm
Initial crummy model attached. 8)
Just an Abeling alike, the old concentric circle model, which shows very nice the effect from the angle between the slot and the track.
As for launching and springs. I think the less external forces, like track and ramps and the more internal..Forces by smart systems or energy collectors like springs, are more likely to be successful.
I think of Bob Kostoff. Although in the end the needed energy comes from an electric compressor. I doubt if the needed energy to run that is delivered by Kostoff's concept.
At ironhumper (lol) nice model
wm2d and springs are still a mystery to me but I will figure it out eventually.
Just a reminder "The weight of the bodies together with ... (intentionally omitted) and the rotational velocity determine the amount of energy that can be generated."
Could the blank be: the flexibility of the springs???
I hope I can ask Mondrasek some questions about the use of springs in wm2d when needed....
AZ
Fire away. But I must warn you, they are one of the most difficult items to use correctly as far as I have seen. And they kill even the fastest CPUs.
Also, I am by no means a WM2D expert. I can only comment on what I've already experienced and theorize based on some experience with other simulation programs. But I'm happy to give my opinion and will try to note it as such when it is not actually "fact".
M.
@AquariuZ:
I like you spring ideas. Right now I can't choose between springs hooking to weight or to the central shaft. In a totally symmetrical system, the effect woudl eb about the same I think.
Nice about the springs, if those would be the key, is that they'd make it a self-starter most likely.
I start to wonder about the ramp layouts then though. Forces might often be inward.
Quote from: Cloxxki on May 23, 2009, 09:55:43 AM
@AquariuZ:
I like you spring ideas. Right now I can't choose between springs hooking to weight or to the central shaft. In a totally symmetrical system, the effect woudl eb about the same I think.
Nice about the springs, if those would be the key, is that they'd make it a self-starter most likely.
I start to wonder about the ramp layouts then though. Forces might often be inward.
If I take Abeling´s hints literally, it would be two opposite dumbbells connected to each other via their respective shafts... Egg and ramps as in Fig 2, 3,4.
No working model I can present yet...
ONE IS PUSHING ONE IS PULLING
Release is near 11 o clock (eject) where it is caught in the ramp and where according to him the extra energy comes from.
Definite self-starter...
Quote from: AquariuZ on May 23, 2009, 10:25:20 AM
Release is near 11 o clock (eject) where it is caught in the ramp and where according to him the extra energy comes from.
Not sure you are right:
I believe Abeling rightly said the extra energy was build up from lowest left and spend at top left....?
ruggero
Quote from: ruggero on May 23, 2009, 12:07:14 PM
Not sure you are right:
I believe Abeling rightly said the extra energy was build up from lowest left and spend at top left....?
ruggero
Exactly what I am seeing, release at 11 buildup just after six...
Stay tuned.
So, at 11 the spring start to unload by entending after compression? Would the neutral length of the spring be between the shortest and longest distance the 2 weight vary, or is the tension one-sided?
[Semi off-topic.] Supposing springs are the secret and solution to PPM here, would then not also a horizontal plane setup be possible with just a magnetic pull from one side?
Or better: a "gravity wheel system" mounted flat on a spinning disc, having centrifugal force provide the "pull" on the entire system? Or would gravity have a unique feature to it allowing to "cheat" the laws of physics as we so elimentary seem to understand them to date? When I think of it, if we can reduce friction, putting a gravity powered perpetuum mobile in a centrifuge should increase its power output? Would running the centrifuge not be an extremely efficient "power source"? Once at speed, the centrifuge can sustain a "gravity like" power output of multiple G's at only the cost of friction to be overcome. Running on magnets, and being an aerodynamic shape with smart final coating, I see possibilities there. People have been trying to harness CF for as long as trying to harsess gravity alomst I suppose. When we get the latter, the former could become a no-brainer to boost things.
CF from a cetrifuge I suppose is like gravity, but being spat out from the center of the rotation. Slightly differnt vectors than with gravity, but with a small system on a large wheel, this could be reduced. Or, one could figure out a gravity wheel where it's actually convenient to not have "gravity" "pointing" the same direction everywhere. Let me know if this line of thinking is worth its own topic.
[/semi off topic]
Centrifugal force is unlikely to be the source of the free energy.
A mathematician will tell you there is no such thing as centrifugal force: what you perceive as centrifugal force is the force needed to provide a acceleration/deceleration to alter the momentum of the object to change it's direction of motion.
Or to put it a simpler way, the centrifugal force is energy you put in somewhere else and is stored as momentum, until it is returned. (Like a spring, except without fatigue and losses to heat.)
The good news is, just like the guides and slots, as long as you can determine the actual path of the weights, you can ignore the actual causes of the object following that path, and so calculate still calculate the torques supplied/required - a force moving due to "centrifugal force" (or lack of centripetal acceleration as a mathematician would say)
Posting error, initially a redundant post.
I'll just add that I feel there's enough to be improve on the basic springless, single dumbell weight design. IMHO, slots that see the weight get behind it's rim position on the rise, are less likely to work than the opposite, getting ahead on the rise. I posted more on this in the other threads, and if there ignored I may post a seperate thread for it, to propose foremost the most simple improvements to the Dusty/Eisenficker2000 designs, and then offer some further option to go about trying to get this wheel to turn.
Quote from: stgpcm on May 25, 2009, 05:43:39 AM
Centrifugal force is unlikely to be the source of the free energy.
A mathematician will tell you there is no such thing as centrifugal force: what you perceive as centrifugal force is the force needed to provide a acceleration/deceleration to alter the momentum of the object to change it's direction of motion.
Or to put it a simpler way, the centrifugal force is energy you put in somewhere else and is stored as momentum, until it is returned. (Like a spring, except without fatigue and losses to heat.)
The good news is, just like the guides and slots, as long as you can determine the actual path of the weights, you can ignore the actual causes of the object following that path, and so calculate still calculate the torques supplied/required - a force moving due to "centrifugal force" (or lack of centripetal acceleration as a mathematician would say)
Then can you explain why most trap throwing equipment uses a accumulated CF force sling method. The trap is set in the middle when the arm is at rest, then thrown out when the arm is at center forward.
You could block the end, then put the trap on the end and stop it at a point to let it fly off at a tangent, but it doesn't go as far.
Better example is a test centrifuge, except put the chair at midpoint on a rail that lets it slide out to the edge. Bring it up to an rpm that a subject could handle if the chair was at the outer position. Release the chair and after the impact scrape the subject off the chair. :D Any volunteers?
It is the
accumulated acceleration caused by CF from middle to edge that makes the impact much higher then just the CF force at the edge.
Regards, Larry
Springs suck. They operate by deforming under stress. That deformation wastes some of the energy put in as heat When compressing the spring you increase the reaction forces in the components driving that compression, increasing friction.
Quote from: LarryC on May 25, 2009, 09:39:29 AM
Then can you explain why most trap throwing equipment uses a accumulated CF force sling method. The trap is set in the middle when the arm is at rest, then thrown out when the arm is at center forward.
1. slingshot - moving the pivot point gives you some lovely mechanical advantage.
2. spin, that design puts a lovely spin on the clay, letting it frisbee.
3. by putting a block at the end you're either shortening the arm, (by the radius of the disc), or adding extra mass at the arm away from the pivot, soaking more energy to accelerate it
4. by putting a trap in to stop the arm, you're prematurely stopped it's operation - there was energy left in that spring.
Quote
Better example is a test centrifuge, except put the chair at midpoint on a rail that lets it slide out to the edge. Bring it up to an rpm that a subject could handle if the chair was at the outer position. Release the chair and after the impact scrape the subject off the chair. :D Any volunteers?
you certainly would go splat - all the acceleration you could take in a quater turn has just been delivered to you an instant. That hasn't change the total amount of acceleration, just how quickly it was delivered
Quote
It is the accumulated acceleration caused by CF from middle to edge that makes the impact much higher then just the CF force at the edge.
Most of us call accumulated acceleration "momentum", and yes, by allowing the chair to slide, some of the force is turned into angular momentum, but the rest into linear. at the moment you hit the rim all of the linear momentum is delivered as an impulse.
A half pound weight can sit on top of a tomato all day. If you lift it 5 meters above the tomato, then release, it accumulates acceleration (or would you prefer me to say it accumulates gravity?) which it then imparts to the tomato after about a second.
Regards, Larry
[/quote]
In a end of arm scenario, the centripetal acceleration is provided by the end stop, which applies the force perpendicular to the circular motion, so it is all dispersed as strain - in effect the only time 100% of the torque is being applied to accelerate the disk is at the moment it is released. So for a 90 degree swing you gain about 1/root(2) of the work you put in when you cocked the arm.
In the thrower, no centripetal force is applied, so you gain all the work you put in when you cocked the arm. Further, as you pointed out, the disk leaves thrower releases at around 135 degrees, so for the same cocking torque you can put an extra 50% of work into the spring. Now, in reality, the speed of the disk from the thrower is going to be less than the theoretical 3/root(2) times the tangential arm, but with a well designed system it should get close.
Quote from: stgpcm on May 25, 2009, 11:47:32 AM
In a end of arm scenario, the centripetal acceleration is provided by the end stop, which applies the force perpendicular to the circular motion, so it is all dispersed as strain - in effect the only time 100% of the torque is being applied to accelerate the disk is at the moment it is released. So for a 90 degree swing you gain about 1/root(2) of the work you put in when you cocked the arm.
In the thrower, no centripetal force is applied, so you gain all the work you put in when you cocked the arm. Further, as you pointed out, the disk leaves thrower releases at around 135 degrees, so for the same cocking torque you can put an extra 50% of work into the spring. Now, in reality, the speed of the disk from the thrower is going to be less than the theoretical 3/root(2) times the tangential arm, but with a well designed system it should get close.
Actually, the equipment in both of my cases would swing 135 degrees. For the tangent thrower it could still have a extended edge to cause a rotation. It still would not throw further than the accumulated acceleration CF sling version.
BTW, your
root(2) notation is unusual, did you mean sqrt of 2 or an array named root where you wish to divide by the value of the second member in the array?
Regards, Larry
Quote from: LarryC on May 26, 2009, 06:17:08 PM
Actually, the equipment in both of my cases would swing 135 degrees. For the tangent thrower it could still have a extended edge to cause a rotation. It still would not throw further than the accumulated acceleration CF sling version.
BTW, your root(2) notation is unusual, did you mean sqrt of 2 or an array named root where you wish to divide by the value of the second member in the array?
Regards, Larry
I meant the square root - so the non-centripetal thrower would (ideally) throw over twice as far as the tangential thrower.
A tangential launcher rotating 135 degrees will throw less far (for the same input work) than one rotating through 90 - for the first 45 degrees both components of the acceleration are in the opposite direction to the final throw, and need to worked against - hence are completely wasted. In the real world this has some benefits to reduce the losses inherent in a system where the arm and disk are not perectly inelastic, and the spring can't contract instanteously, but your 135 degree tangential thrower would throw the same distance as the 90 degree thrower for the same cocking force (but which would half as much again cocking work for the 135 tangential thrower than the 90 degree one)