The Mechanical Engine

[Inquorate]

Marxist

Correct on all counts. Heavy pendulums don't bounce. I'll figure out a way around that soon.

[marxist]

need for mass to achieve flywheel effect

Hi Inquorate,

A thought:
Maybe it would be beneficial to put the weight of the pendulums at their ends i.e. use a light stick and fix a weight at the end, instead of using simple rectangles for the pendulums. (I tried it in the simulation but I always get an error message, maybe due to the outdated program version I am using)

Beams are a virtual 100kg each; pendulums are a mere 5kg each. It's 4meters in diameter, all virtually..

YouTube - final software version of gravity mill - Ben Brandwood

So you are again employing a relatively heavy wheel and relatively light pendulums, as you did before.

My thought about this:
The ratio of how much mass is oscillating/swinging (i.e. total weight of pendulums) compared to the amount of rotating mass (i.e. weight of the naked (fly-)wheel without the pendulums) seems to be quite important.
If the weight of the (fly-)wheel is to small (relatively), it will have too little inertia (set in motion by the first falling pendulum) to pull the first few pendulums over the yellow bump at 12 'o clock and consequently the wheel will stop after 1 or 2 pendulums were pulled over.
True?

On the other hand, if the (fly-)wheel is too heavy, it will not even start to move/rotate.
True?

Now the usual thinking is, that Bessler built a very lightweight wheel, which presumably was not suitable as a flywheel, as it had not enough mass.
I agree to this thinking.
So how did he build up the necessary inertia during the starting sequence to pull the the first pendulums over the 'yellow bump'?

I think he achieved this by means of his external pendulum. Instead of a flywheel he used the external pendulum(s) for a flywheel effect.

In addition to that, his external pendulum also worked as a kind of a 'load-balancer'. I mean to say, that if the machine was not loaded or the load varied, this external pendulum just swung higher (at times when the load on the machine was light) or it swung less high, when there was more load on the machine - I guess.

edit: The above turns out to be a wrong statement, as Bessler's external pendulum was connected to the wheel with a con-rod, like a crank shaft to a piston. So the pendulum could not have swung higher and less high. It must have always swung to the same hight.

And as for the rest of it: I hope you will come up with a suitable link between the 'internal pendulums'.

[Inquorate]

A few thoughts about achieving optimum tradeoffs in my design

Ok, so important things happen at 6 o'clock, 7 to 9 o'clock and at 12o'clock.

At 6, the sprung lever/s is/are either pushed over to the left by the pendulum, taking energy from the swing, or more ideally, the pendulum swings over the sprung lever and swings back to the right momentarily, pushing the lever against it's stopper - which transfers the energy from the pendulum's swing into torque for the lever arm.

That only stops the pendulum from swinging briefly however, and by 7 o'clock it's back to having high kinetic energy.

So, I've been using the pendulum's impact against an arm braced with springs and dampeners (like what stops your screen door from closing abruptly) to again take energy from the pendulum, returning a fair portion.

I have to damp the pendulum's energy from 7 to 9 because an attempt to immediately stop the pendulum's swing will just transfer left movement to right movement.

Here there is a tradeoff. If one takes only a little bit of energy from the pendulum, then it still swings, still creating a downward force that we don't want on that side of the wheel (as it will detract from the downward force being given from 1o'clock to 6o'clock.)

But, take all the energy from the pendulum at 7 to 9o'clock, an when the pendulum gets to 12o'clock (11:50) it will not bounce off the bouncing pad as the end of the pendulum falls behind the lever arm's rotation. If the pendulum doesn't have enough energy here, it will try use torque from the lever arm, and we don't want that.

We need the pendulum to bounce at 12o'clock because the arm will swing it out to the right, setting up the oscillations from 1 to 6 that we need.

Now, my thoughts are this; why take energy from the pendulum at 7 to 9 AND NOT USE IT?

I'm currently mulling over ways to get the energy from the pendulum at 7 o'clock and transfer it directly to the pendulum currently at 12o'clock.

That way there's little pendulum swing from 7 to 12o'clock, but we still get outward swing after that.

Something that takes the leftward movement at 7, and turns it to upward movement at 12o'clock.

Something simple, because it will still theoretically work as is, and gravity is free so...

Love and light
Ben

[marxist]

Hooray

Hi Inquorate and all,
it is early morning where I live and I had the honour to be the first to watch it on youtube.
Hooray indeed.

coat of arms of German town Fuessen:


coat of arms of Sicily:


coat of arms of Isle of Man:

[Inquorate]

Final software version of gravity mill

Beams are a virtual 100kg each; pendulums are a mere 5kg each. It's 4meters in diameter, all virtually..

Here's the concept / pilot version

YouTube - demo of principle of gravity mill

Here's the beast herself

YouTube - final software version of gravity mill - Ben Brandwood

[Inquorate]

One for Woopy

It crashes itself if unloaded :-D

YouTube - Woopy version of Inquorate gravity mill.

[Inquorate]

Been working on this all night..

Woopy is right, using eccentric wheel there must be friction - wheels tend to make it run backward as the pendulums roll off.

So, I got rid of the eccentric wheel, of sorts. Just one or two bounces suffices, so one could make steps instead of a gradual incline.

While dampening the movement of the pendulums at 6 o'clock, I came across a method for using the energy in the swing to push the wheel around. By accident of course. You can see it happening once or twice; I've yet to get the timing of damping vs springing etc.

I'll upload Woopy's file as well. Coming soon :-)

[woopy]

Hello Peter

Thank's for your indications

i experimented your sugestions in the Working Model 2d soft and the result is that the wheel does not get energy on the contrary it slowly get to a stop or in the best case to a very slow running and i can not put any friction on the shaft. (Inquorate experimented the same on it's original set up with the damping at 7 o'clock even with zero air drag) and as soon as you increase the normal air friction and the minimal bearing friction it goes to a stop

So what seems to be (so long that and if Working Model 2d is accurate of course) is that the pendulum effect does not produce a single down force but produces an up and down very powerfull force and that on a very short track.

so to think is that we have to get the most and maximum swinging pendulum on the longest distance all arround the disponible circle which will give as long as possible the most effect on the ratchet and the most torque

For example if i setup my Milkovic 2stage oscilator with the lever exactly in equilibrium it simply oscilates up and down in a very short distance
Now if we admit that the pendulum is on the right side of the system i have to experiment that the left part of the system (in equilibrium) will not raise..or rotate around the main shaft.. but it simply stay almost horizontal and makes short and mighty up and down movement

so i think that Johan Bessler could be right with the combination of swinging pendulum added to 2 stage oscilating and with a incredible but necessary simplicity at this age that is to say a ratchet

if i can replicate what we are speaking here and what seems to be possible as per Working Model 2d ( so long and i insist ,if it is accurate of course)

i would be inclined to think that we missed something for a very long time shame on us Hu!!

waiting for the comments

good night

Laurent

[Peter Lindemann]

Minto Wheel

Hello all, I am new to this thread but I have been working for some time on a vortex/cetrifugal design which does not use weights as such but the change of liquid to gas and back to liquid again. Freon gas, where the weight change of gas to liquid and in reverse will take the place of the pendulum weights. This is not new as a slow turning wheel has been constructed and it used the sun to change the liquid to gas and on cooling it turned back to liquid. May be this could be used in this thread, if not tell me so and I will not post here.

Mike

Mike,

Thanks for asking. Your wheel sounds like a THERMAL/GRAVITY engine, similar to the Minto Wheel. If you would start a new thread to discuss this, I'd appreciate it. We are trying to focus exclusively on mechanical engines that run on centrifugal force from swinging weights, here.

Thanks,

Peter

[Michael John Nunnerley]

Inquorate,

Heavy pendulums is what you want. The "over speeding" problem disappears when you ask the machine to turn some "constant load" device. Bessler ALWAYS ran his wheels loaded with water pumps and stamping mills. This is part of the balance of the machine. It has to be able to "run away" unloaded if it is going to have any power to deliver to a load.

For your first prototype, a small electric generator running an incandescent light bulb will provide a constant load.

I am thrilled with your solution. It should be easier to build than my design, since it does not require a cam, or a latch and release mechanism. Plus, it gets more swings per pendulum per rotation.

Great work!!!

Peter

Hello all, I am new to this thread but I have been working for some time on a vortex/cetrifugal design which does not use weights as such but the change of liquid to gas and back to liquid again. Freon gas, where the weight change of gas to liquid and in reverse will take the place of the pendulum weights. This is not new as a slow turning wheel has been constructed and it used the sun to change the liquid to gas and on cooling it turned back to liquid. May be this could be used in this thread, if not tell me so and I will not post here.

Mike

[Peter Lindemann]

Very Clever!

Laurent,

Very interesting group of ideas. Now that you have a design that is running in the simulator, may I suggest a few modifications?

First, as soon as the "ratchet zone" is finished, just let the pendulums fall off a cliff, right there. There is no need for the central member to be circular. The sooner the pendulums start swinging, once they are passed the "top dead center" of the wheel, the more force the wheel will produce.

Second, move your damper down to around the 7 o'clock position. The sooner you damp the swing passed "bottom dead center" the better the wheel will run, because any swinging of the pendulums on the left side of the wheel produces a force that works against clock-wise rotation.

These two modifications should make the design "run away" much sooner. But just let the simulation run. There will be a speed that it will not exceed because the pendulums will just swing out ONCE and stay out at the maximum position because the radial component of the Centrifugal Force will dominate over the gravitational component. At this point, the wheel will begin to slow down until an equilibrium is reached.

If you would be so kind as to run these modifications to your working design, I'm sure the group would be interested in your results.

Thanks,

Peter

[woopy]

Bessler code decripted?

Hello all

today is a good day i made a lot of test with Working Model and the file that Inquorate kindly sent to me.

After having activated the air resistance the wheel with its very light pendulum stopped after a few rotation. I was somehow desappointed but.

Than i removed the eccentric center guide and i noticed that if all the pendulum on the right side pendule together and the ones on the left side stood still nothing happens and the wheel does not rotate and that independently of weight of pendulum

I was almost completly desappointed as thePeter Lindemann swinging effect seems not working. but

than came back on my Milkovic 2stage oscilator and i could check that if the lever is completly in equilibrium when the pendulum swings, the lever stays nicely horizontal with small up and down movement and a lot of force

so the idea came if all the small pendulums of Inquorate's design are acting as a single bigger one of a Milkovic design perhaps we could get a very simple wayto harnes the extra Milkovic power on a rotating Wheel as Peter Lindemann thought in his article.

So i put back the eccentric centerguide and i covered it with a grippy surface (in my case caoutchouc surface) so when the pendulums arrive in the 10 to 13 o'clock sector they bump against the the grippy surface as against a ratchet and they use the Milkovic pendulum power to lift the wheel

And in Working Model 2d it works perfectly the wheel accelerates slowly until a speed of about 8 to 10 rpm. At this speed i had to put a damper to calm the swing and control the rotation

In the set up i put full air drad and some friction on the main shaft

here under a picture to make things more clear

i have e-mailed some working designs to Inquorate for testing so he can send it further to interested if it is OK for him (I have a very slow connection and the files are quite haevy)

I don't know how Working Model 2d is accurate has somebody experience with this soft?

And now at the workshop i hope that reality will not be too severe wit us

thanks to all and good luck

Laurent

[Vortex]

.......here is an animated gif of what I've gotten so far. It is all from the book "Machinen Tractate." It is easier for me to build the thing and then make the animated gif than to do WM2d. Besides, WM2d demo wouldn't let me save the project for some reason. I do like the idea of WM2d though!

Anyway, with this setup the hub center actually moves to x1,y1 for some reason and then moves back to x0,y0. I'm looking into that. FYI; I determined that the hub center moves from taking pictures and drawing where the center lines are in relation to the hub. For example, between any two weights there is a center bolt. Below that bolt will be another center bolt. Drawing lines through those bolts show where the center is. Alternatively, from the center bolt on the weight towards the axle hub is another bolt. Through these centers can be drawn a line to determine where the center is.

The link to the file is here; Bessler 3.22.09 picture by mrbreau - Photobucket

Warren
..

The book Warren is speaking about can be found here
and images of devices in that book are here.

images 1-143 (there's a menus to view the images)

[Vortex]

Save often

Inquorate,
Great stuff

Inquorate said: "The one I made (but didn't save - Aaaaargh!)"

I am a little late getting here to say this, but save historical
versions often .. what might seem like a "good-to-go" version
could end up as a dead-end and you will want to go back to
an older version to restart, maybe.

Your in the virtual world now .. and...
Software programmers know this and live by it and it should
serve you well to follow this habitual version saving approach.

Names of the files are important, but the date/time of the saved file
is your ticket home to the version that worked.

[marxist]

hot diggidy.wm2d arrived

Hi Inquorate,

I received the file. Thanks.

... It has to be able to "run away" unloaded if it is going to have any power to deliver to a load...

I agree and think this is is a very important point.

Just some features I discovered in my first attempts with 'Working model 2d': You can:
- create .avi video-files directly using File>Export.
- set air-resistance in the menu 'World'.
- select the central axle (Constraint 3 - pin joint) of the wheel (Circle 1). Go to menu 'Script' > 'Friction' and set a higher friction for this joint/bearing. You will at the same time need to define a diameter of this joint/bearing. This is useful to make the simulation more realistic, as the program makes its calculation with zero bearing friction when using unaltered standard settings.
Bearing friction is also useful to slow down wheels which speed up too much. It's similar to adding a load or drag to the wheel.
- select the wheel (Circle 1) and then use the menu 'Measure' to select values you want to measure. Such as torque or kinetic energy. The values will then be displayed in a graph. Very useful in order to see whether a wheel slows down over time.