Meet the Clockmaker: Cracking Bessler's Code

[spinner361]

I had been making models in a convenient size at the default zoom level. I have zoomed it in and made a new model with a main wheel that is two feet in diameter. Much better. I found the default spring force of 50k to be too strong, so I lowered it to 10k. I, too, use a push ball to start things off. I also now set the accuracy to "accurate" and then customize it with an overlap error of zero. After tinkering with the rebuilt model for about 30 minutes, this one is also self starting from a complete stop with no push ball. I have not seen if it continues to accelerate, but I believe if it did not accelerate at all it would stop before one full rotation.

Air resistance question: This is without air resistance. When I set the air resistance to the default "low", the thing barely moves at all, which does not seem quite right to me, because It acts like it is underwater. Perhaps someone has a tip about this.

[spinner361]

I let his rebuild run for a while and it has a noticable acceleration curve on the rotation graph. This is not enough for me yet - I am not totally convinced. A lot depends on what I learn about the air resistance in W2MD. I have taken the current configuration as far as I can for now, so I am going to try and figure out how to pound on it with a hammer.

[Jon J Hutton]

Thanks for the direction on where to get a "sample" of wm2d.

I started working on it right away. Very simply to use. After working some time I discovered that with the sample download you can't save your work.....like a dummy I should of checked that out first. So I lost about 3 hours of work. I checked on the price and they are proud of their program....so back to paper and pencil, or my faithful share ware architect program.

[ME]

There is an alternative... a Saving script. These are posted before (somewhere on this forum).

I'll give you a script (a combination of my own and Amateur), but it doesn't save Colors and Collisions.. but it's better then a complete rebuild.

Hope it'll work for you

[Jon J Hutton]

Thank,

I hate to say this but your going to have to walk me through this. I programmed using BASIC in college...its been a while. I copy and paste it where and how. Did you buy your wm2d or are you using the download demo.

[ME]

I use the Demo version

You don't need to understand the code to use it.
(although you can -and should- scan it to see i'm not reformating your harddrive with it or do some other nasty things)

For using the script, I can't explain it better then this:

[ME]

oh, and ofcourse:

When you have saved it, locate it on your harddrive (my saving location is the Mechanism directory of WM2D)

To restore the model, open the saved script (rebuild.wbs or whatever name) just like the saving script, and hit run.

[spinner361]

I was really dissappointed when I turned on the air resistance in W2MD to "low" and my device hardly moved. It just did not seem right at all. Out of frustration I performed a quich W2MD test which renewed my hope.

With tests I had done in my garage about a year ago, I counted approximately 70 swings of a pendulum before it came to a dead stop (or very close to a dead stop). I made the pendulum with a 3 foot 1x2 and old bicycle wheel bearings.

As a little test of W2MD's default low air resistance at 0.200 kg/m-s, I took a rectangle shape with a pin joint at the top and lifted it up to 45 degrees from BDC as a starting point, which only produced 5 swings before coming to a dead stop. The same thing happened when I started it from 90 degrees. Now I may have something to go by!

I will redo a real-world pendulum test again and count the exact number of swings until it stops, then build a W2MD model of the exact size and set the air resistance to match the real-world results. Could someone confirm if this test would be adequate, or if there is a better way?

[spinner361]

With a little testing, I figured I would try an air resistance of 0.035 (the default is actually .300). My device does not work there, so I tried 0.025, then 0.005. Failure. I think I will take a break and work on my Java certification.

[jonnynet]

unfortunately, novus's "clockwork" mechanism is definitely balanced. it's possible to realize this mechanism without any connection to the center axis or an external fixed structure but the center of gravity is always the big wheel's axis :-(
the animation below will demonstrate this "free-flying-mechanism". I think no more words are necessary to understand how it will work.

[ken_behrendt]

jonnynet...

Your animation (not a simulation...big difference) shows the weighted circles as always keeping their weights pointing downward. I found from the simulation I did of this design that this does not happen in practice. The weighted levers attached to the weighted circles will cause each one to counter rotate until the composite CG of each weighted lever/weighted circle is as low as possible. After that happens, the device will, for all practical purposes, be as motionless as a Ferris Wheel whose seats each hang below their point of attachment to the wheel.


ken

[jonnynet]

I know there is a big difference between an animation and a simulation! my animation was made from a simulation and because the wheel don't run by itself I've attached a hidden motor to show the operating principle. the weights combined with the bigger circles are much heavier than the "hands". their weight is 2000 gram and the "hands" weight is only 50 gram. the heavier weights hang almost down - they swing a bit to the left - and in practice they would swing to the right too (when the "hands" go from 12'o clock to 6'o clock position!). here is the wm2d file from which the animation was made...