Work from 2 magnets > 19% output 2

[lumen]

@ Lumen

   some observations

In your photo of your RO setup..

It looks like you have to re-position the force gauge after
every  degree or two of rotation ?  this could account for
an apparent OU result ?

A rack and pinon gear would translate the rotational
to a linear motion with a consistant margin of error ?

or a nodding donky set up and a PULLING force measurement
could be used... but then chain / or string,  stretch could become
problematic ?


       regards
        floor

The rotary table rotates everything as a single unit.
There are bearings inside which the magnet mounts to, but they don't turn because the scale holds it stationary measuring only the force applied.
The entire setup is then rotated against the stationary SL magnet using whatever steps desired.

So the scale never needs to be moved because it rotates with the setup on the rotary table measuring the forces applied from a stationary external magnet (SL).
This way the scale always remains at 90 degrees to the force arm so the reading is not affected by a changing angle on the arm.

As far as the math?
I understand the force distance formula for work but what the problem amounts to exactly is what force?
In my case the distance is always the same so that part is easy. The force is at points and the change between any two points is not linear.
So thinking that the force to use is the average of two points over that distance could only be close but not exact, and if not exact, what could the error be?

The method I use is (average of 17 points) * 16 spaces = total force over .800 inch travel which is the same over all tests.

Just looking for an explanation for the energy gain that should not exist.

[lumen]

Another thought is that these tests are only a comparison of the work done in making moves between two magnets that should amount to zero but do not and that even with a small percent of error in the math, the same error would exist on all measurements and would have no real effect on the result.

If I ran the same tests using two round magnet and used the same method to find excess energy then the result would in fact show zero gain.

So it is then the magnet shape that impacts the result.

[Floor]

@ lumen

I'm commited else where for the next 4 days.

I think we are still on the same page.

We need to be certain of  the validity of our processes and  methods,
both meachanically and mathematically, or else change them.
I'm certain that in either case we can do so.

                     later
                         floor

[lumen]

@floor
No problem, I have a few things to do also before setting up this next test.

[Low-Q]

Collecting data can be helpful but I'm wondering about the math.
Because magnets are not a linear force is it reasonable to assume that using a linear calculation like averaging is in fact accurate.

It just seems that finding results of 12% and then 13.5% for a calculation that should in fact be equal ( 0% OU), appears it could be an error in the math process.
Is there anyone good at math that can verify the data indicates an actual gain?

Math never makes mistakes. It is the person who do them who fails. That is the platform we humans must build our experiences on.


If you want reliable measurements, you must take a high number of samples. Especially when the magnets are very close to eachother. A 1mm resolution sample when the magnets are 10mm apart is 100 times more accurate than the sample made 1mm apart. So you have to scale down the sample rate the closer the magnets are to each other. Maybe 0.1mm resolution when they are close, and 1mm resolution further apart.


It is easy to miss out major force or torque readings when you have a coarse sample rate.


If the math proves over unity, it is the samples that are wrong. Calculating an accurate result is very time consuming, but not actually rocket science. I would say closer to primary school math. The trick is to not miss out samples that have a major effect on the result.


Vidar