Design a electromagnetic set-up for alternating motion of magnetic particles

[alwaysphoenix]

Hey,

I want to design an electromagnetic set-up with AC input using which I can have a liner alternating motion in magnetic particles under the application of the field. I thought of having the AC input because the alternating current in the electromagnet can produce quick polarity changes in the magnetic field and the magnetic particles will move with the alternating field, resulting in a linear alternating motion.

I want suggestions and advice with the design or any related issues I have to know for this.

Looking forward to have fruitful discussions soon!

 

[jsgruszynski]

Not so trivial. First you have to consider the mechanics of the particles and if the particles are suspended in a fluid, the fluid dynamics of the suspension. Naturally this also gives you additional information about the B-field spec needed.

Even if you apply the "ideal" magnetic field of a triangle wave, you'll not get linear motion because of things like particle inertia and viscosity of the fluid. What you'd likely need to do is determine these mechanical effects and then pre-compensate the B field accordingly. The current-coil-field-particle-motion system is like a 2-port.

This assumes you have a particular specification for linearity of particle motion/alignment. If you are simply assuming you need it, you are already making a mistake, most likely. The effort to create linearity is sufficient to justify knowing for sure first. Otherwise, don't worry about the linearity and just throw any B-field waveform you can generate at it. Only don't expect high frequency and high field at the same time.

The next issue is that electromagnets require current. Lots of current. Add to this that the electromagnetic coils are inductors and you quickly set a shockingly low upper bound to frequency. This is limited by L di/dt effects combined with the fastest possible slew rates you can get from high power linear amplifiers. A system I worked on had a maximum bandwidth of 10-15 Hz due to this effect. This was with state-of-the-art high power linear amps driving 300 A per coil at peak B field less than 1T.

I've worked on custom high linearity, high field ohmic electromagnets and the design challenges are significant even for someone like me (30 years EE experience).

 

[alwaysphoenix]

Thanks for the reply.

The magnetic particles will be attached to surfaces and the idea is to create a linear motion on the magnetic particles with an effort to make them separate out from the surface. Though, the motion can be non-linear too, its not important.

The main issue with me is designing the setup so that I can have the alternating motion on the particles by application of electric field, which will result in magnetic field. This won't be possible just by alternating the magnetic field, as the magnetic particles will always be moving towards the coil, and not to and fro.

Won't it be possible with multiple electromagnets and having a different wave function electric current pass through the magnets? This can result in varying magnetic fields in different directions across the magnetic particles at different timepoints.

What do you think?

 

[Bob S]

About 12 years ago, a group at General Atomics and LANL developed a ferrite magnet system using IGBTs that produced a ± 500 Gauss triangular waveform field at about 500 Hz. The IGBT circuit required about 1 kW of power, and produced about 30 kVA of peak inductive power that cycled between the magnet and a storage capacitor. See

http://accelconf.web.cern.ch/AccelConf/p99/PAPERS/FRA160.PDF
http://accelconf.web.cern.ch/AccelConf/l98/PAPERS/TU4089.PDF

The IGBT H-bridge applied a V₀ = ±300-volt square wave to the magnet, and the magnet current was a ±100-amp triangular wave form, based on the relation dI/dt = V₀/L.

This alternating polarity triangular waveform magnetic field could accelerate magnetic monopoles, but none have been found. This spacially-uniform field will not accelerate magnetic dipoles.

Similar time-dependent (actually 60-Hz sine wave) magnetic fields have been used in particle accelerators called betatrons (based on the Faraday induction principle) to accelerate electrons up to energies of many million volts.

Bob S