Magnetism experiments....

[John Hill]

This is my first post to this forum..........


I have been trying to make a solenoid that had reasonable and consistent force over a long stroke, as you will know most common solenoid actions develop the most force near the end of their stroke.

I took two cylindrical magnets and stuck one to each end of a steel bar (stainless steel, but thats what I had on hand. The steel is magnetic). But I reversed one magnet, that is I applied a like pole to each end of the bar. There is a pole induced near the mid point of the bar, I presume it is actually two poles close together. These induced poles are opposite polarity to those at the ends of the assembly.

I put this bar with the two magnets attached in a solenoid coil and used a set of digital kitchen scales to measure the force which I found to be almost uniform over the length of the stroke. The magnets are bigger diameter than the bore of the solenoid which limits the stroke to the length of the bar minus the length of the solenoid, the bar is about twice that long.

My theory was that it would be useful to have a bar magnet with a like pole at each end and an opposite pole near the middle of the bar which would put a concentration of armature flux right in the middle of the solenoid. I believe my assembly of magnets and steel bar did this.

I made a little engine to prove my ideas by incorporating them in a "solenoid engine"...

http://flic.kr/p/az9wqs



... I dont know why the audio is so loud as the engine is really not quite that noisy..

As you can see this is a two "cylinder" machine and each cylinder is double acting. That may seem an odd way to arrange things but the double magnet and pole piece that makes up each armature is rather heavy and having two like this tends to balance things somewhat.

The engine can run quite slowly, just a little slower than in the video and it will even run without the flywheel. I think this demonstrates the long power stroke achievable with this arrangement of the armatures.

It is running on about 24V in the video and it goes quite a bit faster if I wind up the voltage to the maximum of my power supply which is 36V. The solenoid coils are quite happy and show no sign of heating at 100V which I suspect would be beyond the capability of the very light weight crankshaft I have made in this engine.

I tried to find some online software to map the magnetic circuits involved but I am afraid they were all well beyond my capabilities..

[chrismb]

Hello John,

Welcome!

That makes a lot of sense, what you did, to solve for a long steady stroke.

There are plenty of ways of magnetising magnets and what you have described I have seen advertised for sale commonly enough. I do not know what their industrial purpose is (whether the solution you've come onto has already been done) but one application is to present a pair as 'joke' magnets where it appears both ends repel!

Magnets are often magnetised with a series of regions on them, for encoding purposes in rotary or linear actuation systems. I'd imagine there might be a way to have not just two domains 'facing' each other along the length of the bar, but a series of them, and it can then be tuned by fitting some magnetic shielding around the regions where the domains go the 'wrong' way, compared with the regions going the way you want which are left unshielded.

Certainly food for thought there. Thanks for the intro to your test.

[John Hill]

Thanks Chris, can you suggest how the pattern of the flux lines would look inside the solenoid? Thanks.

[chrismb]

Hi John..

..quick sketch.. just an idea.. something like...

[John Hill]

Thanks Chris, I had not considered Mu metal mainly because I dont have any having taken all old transformers to the scrappies when we moved house.

I have been thinking that maybe I can get the same useful effect (as I get with my pseudo 3 pole bar magnet) by having two solenoid coils wound to simulate the magnetic pattern of my bar. Something to think about during my morning walk.

[chrismb]

Hi John,

Do you ever get that feeling you're rushed something out, looked ok, then later realise it wasn't what you meant to do!

That's as with the diagram above. Might sound bizarre, but I thought I'd drawn what I wanted to, but it dawned on me last night that it wasn't.

What I was thinking was to put the shielding around the shared domains. I'll try drawing it again. Maybe it won't look nice/work out when I draw it again... more laters...