1
\$\begingroup\$

I want to create a solenoid to match the magneted strength of an n52 magnet of around 4-6 inches in length. A quick google search reveals an n52 magnet is about 14,000 gauss. Using online solenoid calculators I can get to this point at sort of sensible currents with around 15000 turns, but there's no wire that can go 15000 turns around 4-6 inches that can withstand that current without being incredibly bulky right?

\$\endgroup\$
  • \$\begingroup\$ Sorry I meant 4-6 inches in length. \$\endgroup\$ – coinmaster Jan 13 '18 at 2:40
  • \$\begingroup\$ still unclear ... what about 15000 turns? is that 4-6 inches in length also? \$\endgroup\$ – jsotola Jan 13 '18 at 3:03
  • 1
    \$\begingroup\$ An N52 magnet is defined by its energy per unit volume. In this case, \$52\times 10^6\$ Gauss-Oersteds, or \$\approx 4.14\times 10^5\frac{J}{m^3}\$. Magnetic moment is quantized (QM) at about \$\mu_e\approx 1\times 10^{-23}\:\textrm{A}\cdot\textrm{m}^2\$ for an electron. In thinking about possible core materials and the fact you probably cannot expect more than one electron orbital per atom to be involved, I don't think you can get from here to there. \$\endgroup\$ – jonk Jan 13 '18 at 4:28
  • 1
    \$\begingroup\$ The surge generator was 30 each 1.8 uF 60 kV capacitors with 3/4-10 threaded rod out the top. It was about 66 uF with all of them in parallel. Max safe charge voltage I set in software was 32 kVDC. Into a fat 1/4" thk by 2" wide bus bar it could sink 150 kA as a single pulse. I avoided large inductive loads like 1 mH (only 70 amps could pass through 6 awg)as they made a huge ring wave that could weld the contactor shut and splash the room with molten copper from the contacts. The noise was profound. My sound meter peaked out at its 150 dB limit. \$\endgroup\$ – Sparky256 Jan 13 '18 at 4:46
  • 1
    \$\begingroup\$ +1 Why the downvotes? He applied what knowledge he had and arrived at a seemingly nonsensical answer and asked for comment. The problem was not with his calculations but with the non intuitive nature of the real-world solutions - usually many kA's and few turns. \$\endgroup\$ – Russell McMahon Jan 13 '18 at 13:15
4
\$\begingroup\$

I want to create a solenoid to match the magnetic strength of an n52 magnet

  1. Not possible, or at least not practicle. With nickel steel alloy you can make a very powerful solenoid, but due to saturation it will not match the intensity of n52 as a raw material.

  2. Your first mistake is in thinking that turns alone will get you an intense magnetic field, when in fact it is 6 to 12 turns of 2 awg to 6 awg that work the best. The reason is that it is the intense current that does the work. Many thousands of amps. To push that much current through wire you need 20 kVDC to 30 kVDC from a huge capacitor bank.

  3. You also need a large air-driven contactor to apply the pulse once the capacitor bank has charged up. The capacitor bank should be at least 33 uF and rated for twice the surge voltage you apply. It will be about the size of a refrigerator laying on its back.

  4. The charger would be 3-phase ideally, but it needs to supply 150 mA to start with and will take a minute or more to charge up. The transformers we used were 208 VAC primary and a fused 14.4 kV secondary that went to a 3-phase bridge rectifier with 100 6A10 diodes in series on each 6 'legs'.

  5. You will have issues with coronal leakage you will have to deal with. You will need 40 kV probes to measure the charge and know the exact charge you need. Too much and your solenoid could crack under eddy currents that act as an angular force.

  6. Your second mistake is in thinking that there is a substitute for n52. There is not, though future NIB alloys may equal or better it. Neodymium is a unique metal in its properties and combined with iron-boron as an alloy it is the most powerful permanent magnet material known. It took maybe an intense 2 Tesla pulse of energy to magnetize it to 1.4 Tesla, far beyond what silicon or nickel steel can handle.

  7. Also solenoids per se do NOT retain a magnetic field of their own once the electric current is removed. A snubber diode shorts out the kick-back from the coil, so no magnetic field is left. To make a magnetic field permanent takes special alloys with no snubber diodes.

  8. You will need to heat the material to be magnetized to beyond its Curie temperature, then apply the surge, then wait for the magnet to cool naturally.

  9. In short you have a good idea that looks good on paper, but real-world physics puts it beyond your reach, unless you are a wealthy person and have the knowledge base to implement this device. Please read the comments by @jonk and others. They lead to the same conclusion.

  10. If you need a powerful and/or fast solenoid for practical purposes, consider using an air cylinder. For raw power they can beat any magnetic solenoid.

\$\endgroup\$
  • \$\begingroup\$ If I need to re-magnetize a tiny N35 disk-shape magnet, how long do I need to have the solenoid field applied? Don't I need to keep the field until the magnet cools down? \$\endgroup\$ – Ale..chenski Dec 12 '18 at 20:26
  • \$\begingroup\$ @Ale..chenski. The pulse of current is at least 2 giga-watts of peak power, so it can only last for 10 to 20 uS before things blow up. The material is heated to well below the temperature where it would lose its magnetic field, so it cools off maintaining most of the magnetizing field. Newer alloys can become very hot before losing their magnetic field. \$\endgroup\$ – Sparky256 Dec 12 '18 at 21:49
  • \$\begingroup\$ Here is an inspirational video how people magnetize raw magnets, youtube.com/watch?v=g9_JgAHv59k Looks like production, water cooled... \$\endgroup\$ – Ale..chenski Dec 12 '18 at 22:28
  • \$\begingroup\$ @Ale..chenski. Cool video. Notice the heavy gauge twisted pair of cables to power the magnetizing coil. It is a brief but very intense pulse of current. \$\endgroup\$ – Sparky256 Dec 12 '18 at 22:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.