The Wind is Free!

Great!! There power of science lies in its ability to predict (while of course acknowledging those awkward details like unpredictability inside singularities!)

So here is a challenge - can your working out predict what the separation weight will be of the same test rig if the distance is increased from 20mm to 30mm? (for further than that I need to use longer bolts)

Humm,
The trouble with your magnets is that they are oblong. How am I going to account for that in formulas that are based on circular circumferences? Can you not buy some longer bolts, or maybe use some plastic cable ties?

I've collected a lot of info on Neodymium Magnets, and I'm trying to get my calculations to agree with some of the info supplied. So far, little fun. But I'm sure I'll get there.

OK, I worked it out.
The separation force at 30mm will be

At 30mm, 1055g+63g+121g = 1236g. Not bad! And there may be measurement inaccuracy - its not clear exactly when the plates separate as the strings are not exactly equal length. At 50mm, between 255g and 340g (uncertainty percentage is higher with smaller weights).

So it seems you have those equations whacked, even for non circular faces?

BTW no need for extra long bolts - I just added some spacers between the magnets.

Wow
At least I got something (approximately) right. Thanks for that windgat.

Well . . . . studying the subject I got convinced that there is no way in which I can replace these magnets with coil-wound equivalents. So . . I dropped that idea.

I'm still battling to get my formula approach to agree with some of the information given on the internet. I mean: what use would this be, if I can't correlate the test result to the grade of the magnet?

I've been on this for days now (enjoyed it) but . . . . need to take a break.

Catch you later
Regards
Cor de Wit

Yes its been fun, thanks! Be great if you scanned/photographed your workings out sometime and posted them here!

Oh yes I will. But at the moment I'm not sure if they're ready for peer-review. Just need to do a bit more research.

See you.

I finally managed to compute the force between two rectangular magnets using numeric integration on an Excel spreadsheet.
But I'm disappointed . . . . the results differ from windgat's measurements by more than 10% - mostly upwards. Last time I was - in fact - more accurate using a "fiddle factor".
The trouble is that I'm having to transform the magnets into a coilwound equivalent of similar dimension for the formulas to work. However, I can't figure out how "life-like" that actually is.

Hmmmmm . . . . . back to the books.

I think I got it right
I (mathematically) replaced the magnets by a combination of elementary mini magnets, each with their own magnetic moment and then calculated the forces between each of them. I then summed them up using Simpson's numerical integration method, and I got:
Force between the two magnets:
@ 20 mm - 2873g
@ 30 mm - 1100g
@ 50 mm - 275g

Next is an attempt to calculate the volts induced when using windgat's constructions.

I managed to calculate the flux density between the magnets that are

Well,
Seeing that you guys are so enthusiastically interested - - I am going to try out a novel way of generating elctricity, viz:



A conductor moving in front of the magnet will cause the generation of eddy currents that will create a magnetic field opposed to the magnet's field.
Using two slotted discs rotating on either side of the magnet will therefore cause field fluctuations inside the magnet from maximum to (almost) zero weber, and back again.
These field fluctuations will be picked up by the coil surrounding the magnet in which - as a result - a voltage will be induced.

What I like about this design is that you dont fling pieces of steel around. The moving parts are very light. What I envisage as a disadvantage is that the slotted discs might get very hot as a result of the eddy current losses.
But a trial will tell all that.