Work from 2 magnets > 19% output 2

[lumen]

@ Vidar

In the next two tests I have planned, I will take many more data points to remove any question of high angle transitions that could have been missed.

A graph of the last test data (not yet posted) does not show any indication of rapid transitions that could have amounted to 13.5% so I am reasonably convinced that there is something going on that could be useful in the design of a real operating machine.

Once I find the point of maximum attainable gain I plan to proceed to build a machine that can function within those parameters.
I still need to test floor's configuration because he states an even higher gain than what I have seen.

[Low-Q]

@ lumen


Looking forward to see the results. I hope you pull this one off

[lumen]

These are the results of the second test that I wasn't going to post but just in case someone is interested.
The magnets used were 1/8 x 1 x 2.

In the test the magnets are pushed together against the large faces but rotated 90 degrees to each other.
At zero distance the gap was .1 between the magnets and then they are rotated to be parallel and in full repulsion.
They are then moved apart at the exit stroke which seems to provide 13.5% more work than the other two mavements.

The step size is .05 inch and is the same in all the steps even the rotation which is shown as degrees at a force radius of .5093 and 5.625 degree steps is making a movement of .05 inch.

All of the next tests will have a reduced step of .01 inch to provide assurance of it's accuracy.

[Floor]

@ lumen

Nice presentation.

The only changes in force that could be easily missed, are those
occurring at very close range,  (domain flipping).

Other wise, the curves are pretty predictable.

I think one could  skew the ratio of RO to SL by differing the
number of samples taken on one side or the other.

How dramatically the law of diminishing returns is affecting the two sets (RO and SL)
is the only question I have at this point.

         floor

[lumen]

@ floor

Thanks, good to see you made it back.

I thought plotting the curves might show some problem area but it looks like nothing could explain 13.5% error.
Once I modify this setup to take some stronger forces I'm thinking of running this same test over using smaller steps and recalculate the results.
If it ends up about the same results, I suppose that would indicate we are on the right track anyway.

I first need to finish another project so it might be another week before I can do the tests.
Interesting stuff!