Another Glitch?

[David]

I have built a W2MD that shows some promise. It would be expensive to build, but if proved positive would output a great deal of energy. I know your thinking ...ah oh another one... but before I make a complete idiot of myself here, is there a way to confirm the model?

1) It uses gears. I read somewhere that there is a problem with them.
What kind of problems are there?

2) It has parts that over lap. The colisions are set to not collide.
Is there a bug in the software that will ignore this?

3) After the model reaches a certain speed the pins begin to red line, almost as if the contruction is falling apart.
What do these lines mean?
If it has to do with the greater force applied to them is there a way to prevent this?

4)I have the demo version of the program.
Besides the most obvious differences, what are the least, if any?

I applogize for not posting the design directly, but because of the above
problems with W2MD, I would like to verify first. If it turns out to be a glitch, I may post it as a novel failure.

[jim_mich]

Hi David,

I see you've been lurking here for a long time without posting.

As far as I know wm2d has no problems with gears.

Individual objects can be set to collide or to not collide. Circles, squares and rectangles collide and rebound OK. Polygons have a problem with collisions. Flat sided polygons work ok sometimes. Curve sided polygons seldom collide properly. The problem is polygons are constructed from line segments and once an object overlaps one of the polygon's line segment the program see the polygon as being on the other side. The collision forces then get calculated backwards. Such a model will speed up continually.

When the pins red line it means that the model is falling apart. The program has determined that the forces on the pins are too great. There is a glitch in the program that makes the parts suddenly fly apart for no apparent reason. The only solution to this glitch is to reduce the time increment for each step, reduce the overlap value, and cross your fingers. Mr. Woo from Working Models doesn't seem to recognize or see this glitch.

As far as I know the only differences between the demo and the full version is the save features.

Some people initially think they have a PMM solution when their model starts turning and keeps turning. If there is no load on the wheel (no friction, etc.) then an initial imbalance will start the wheel turning and without anything stopping or slowing the wheel it just keeps turning and turning. There are a number of was to put a load on a wheel to see if its coasting or actually working. For a quick check you can set air friction higher than normal. For a more accurate check add a constant torque motor and adjust the torque to see how much work energy the wheel is producing.

[ken_behrendt]

David wrote:

After the model reaches a certain speed the pins begin to red line, almost as if the contruction is falling apart.
What do these lines mean?

It's possible that your model is tearing itself apart due to the increase in CF (centrifugal force) on it parts! You might try reducing the masses of the parts using the Properties window and see if that allows the model to run longer before it breaks apart.

Yes, welcome to the board and I hope we can see more of your efforts...they sound very interesting.


ken

[David]

Thanks.

I will let you know in a few days if the model tests negative. If not then I will be needing some special equipment/parts in my shop to prove that it will work in the real world.

[David]

The Picture is out of focus so I will explain this here.

1) Large circle rock, mass=12.566KG
2) Small circles std, mass=.283KG
3) weights attached to large circle steel, mass=.48KG
4) pendulms attached to small cicles steel, mass=.320KG (does not have to be steel)
5) All object set not to collide except the starter weight steel, mass=6.16KG
6) All circles connected with gears.

Operation: The starter weight falls on the offset upper weight of the larger circle.
This displaces the two pendulms (operating in a near vacume condition)
The middle pendulm increases velocity until the larger circles velocity becomes 0 and reverses direction.
Then as the velocity is transfered back to the larger circle, the velocity becomes larger than the original input value.

Please let me know your thoughts on this problem. Thanks David

[David]

Hi all,

Was just wondering if anyone had a chance to set this one up? Any thoughts?, questions? problems? Let me know. Glad to help, anytime.

David

[David]

Again sorry for the awful picture here is a better one.
This one was done with slightly lighter pendula.

[ken_behrendt]

David...

Ahhhh...that's much better. Now we can see what you have constructed.

However, I'm still a bit puzzled at what I am looking at. Are the two smaller circles supposed to form some sort of jointed pendulum whose parts are connected together by gears and which then swing around the large circle during its rotation?

You mentioned that it is in a near perfect vacuum. Does that mean you switched off the air resistance? If so, then you should know that practically any motion in a WM2D model in the absence of air resistance and pivot friction will continue ad infinitum. However, your rotational velocity graph seems to show acceleration.

Well, I might try to do a quick WM2D model to verify what you have, but I am going to have to understand more about how the parts in your model are supposed to move with respect to each other. Chances are, however, that the increase in velocity is due to yet another glitch in WM2D. But, then again, maybe not...it's seem worth further analysis.


ken

[David]

Ken,

Please add anohter step to the setup.

Each circle has a pin joint to the back ground behind the gear.

This setup was tested with multiple settings on the air friction.

It will work only lower than a certain value after that the losses are to great to build up any energy in the system, or so it seams.

David

[Jon J Hutton]

David,

You rediscovered my frustration (see the posts in sorry guys its a glitch) with wm2d. Until they send out a revision or a fix it will be a waste of time to trust the results of a pendulum model. I have contacted them many times but they are not returning my email. If you have bought the program or are using the demo you deserve a response too. Click on their website...under contact us, and don´t give up till you have a reply. I should of seen this ahead of time.....why. It seems ( I read on a web site) a group of college students accepted a challenge that was. Use any two programs to simulate real life physical events. They made it all the way to the semi finals but lost when they were tested if they could....using a pendulum throw a 30 pound basket ball through the hoop at something like 30 feet. They simulated it on.... you guessed it, broken model, sorry, I mean Working model and another program. Using the same specs as the computer model, the pendulum was set up and they put a new meaning to the phrase air ball. I have decided their is no test like an actual test.......Good Luck

JJH

[ken_behrendt]

Jon...

Yes, WM2D is not perfect. However, I suspect that neither are any of the other far less frequently used CAD programs. If we were to question those making a lot of models on any of them, I'm sure we would hear about one glitch after another. CAD programs are very complex programs and, unfortunately, "bugs" can pop up in them after their release that the original programmers and testers never anticipated. Such is life...

However, I still maintain that if a person wants to get into CAD modeling as quickly as possible with what has got to be the most "user friendly" program in existence, then their first choice should be WM2D. The work that I have already completed with it in the last year probably has saved me a decade of labor in the shop that would have been wasted chasing unworkable designs.


David...

Well, I still have not had time to work on "verifying" your model...maybe someone else here can give it a try.


ken