Suppose we have two identical permanent magnets. We install them at different positions relative to each other and measure the acting forces.
We want to find out what forces will act between the magnets when they are positioned arbitrarily relative to each other.
We change the angle between the magnets, move them horizontally and vertically. Once the magnets are at a position
we're interested in, we rigidly fix them, make a pause so that everything settles down, and take measurements.
Then we move everything to a new position, repeat measurements, and so on. Finally we'll have a big table describing
the forces for the two chosen magnets differently positioned.
Strictly speaking, you don't have to conduct such experiments in all cases. Everything depends on your problem and goals.
You can just write down Maxwell's equations and solve them for our magnets. You will know the magnetic inductance at
each point and calculating any forces from that will be a purely formal exercise. You can also describe the magnets in any
electromagnetic FEM simulation software and let it solve the same equations and compute the forces for you.
However, everything depends on your goals, and Maxwell's equations, while they apply with a tremendous success to a broad
range of electrical engineering needs, may not apply as well to your overunity research project.
Classical electrodynamics is just a theory, and no theory can describe the nature completely. For example, the attraction
between magnets is a little stronger than the repulsion (
http://www.supermagnete.de/eng/faq/Is-the-attraction-between-magnets-as-high-as-the-repulsion).
Now try to verify this statement in a simulator, and it will say the attraction and the repulsion are equal in magnitude.
Why? Maxwell's equations are built for a uniform space, and the difference between the attraction and the repulsion
forces is explained at the level of magnetic domains in permanent magnets. And the necessity to account for the magnetic
domains, the structure of crystal lattice, and molecular properties of magnetic materials doesn't make us any happier, right?
For springs, we have Hooke's law. It's useful and usable from the engineering standpoint. It's simple, yet yields great results
for the linear portion of deformations and for widest variety of springs. For permanent magnets, we don't have anything like that.
Our choice right now is complex overunity-ignorant Maxwell's equations vs physical experiments.
But once we've measured the forces, we can even find highly accurate analytical expressions by curve fitting. Such empirical
expressions will be both simple and accurate enough for the overunity purposes. The experimental data can serve as a test
bench for different electrodynamic theories and simulation software. You will instantly know whether certain simulator is good
enough for your problem.
There are minor purely technical challenges to accurately measure permanent magnet forces in practice. Most force sensors are
one-dimensional, and magnetic forces act in many dimensions, and often create torques. We want to separately measure all force
components, for which we have to isolate different degrees of freedom by using high quality linear guides. Permanent magnets
interact strongly when near each other, so we have to preclude significant displacements by using a rigid construct. We have to
reliably fix the magnets before the measurement. We also have to accurately read positions. The magnets should only interact
with each other, which excludes the use of ferromagnetic materials in the vicinity of the magnets. And the force sensors should
not affect/be affected by the magnetic field, which requires to place them at a significant distance from the magnets. Such
requirements call for a custom-designed test stand.
This is the experiment we crazily want to conduct. We're
crowfunding for this right now.