Magnetizer questions 2

[JerryRoy]

Howdy, new guy here,
I tried posting this earlier today, but it didn't seem tp go through, so I'll try again.
We are re-purposing an old transformer to use as a magneto magnetizer.
We started with a transformer that has a two piece armature. The bottom was one piece, in the shape of an I, 1.75 in thick by 3.5 in wide, made of laminations approx .015 in thick. The rest of the armature was in the shape of an E , with the two outer legs measuring 1.75 in X 3.5 in, and the center leg being 3.5 in X 3.5 in.
There were three coils on the center leg, with approx 400 turns total of 6 ga CU wire. At 200 amps, that will give 80,000 amp turns.
Our plan is to wire them in series, and leave them on the center leg.
We will cut one leg off, surface grind the outside edge of it, and the outside edge of the remaining leg, and put them back to back to get another 3.5 X 3.5 leg, with what now becomes a flat bottom U.
The bottom of the U will be ground, as will one side of the I piece, and the U will be set on it to get a 3.5 X 3.5 U.
When we dis-assembled the transformer, the bottom I piece was held on by some side straps, and there were some iron shims between it and the ends of the E legs, along with some sort of metal filled plastic like substance, which looked as though it was a paste when applied, and hardened after the transformer was assembled. Is there something used in magnetics, which transfers flux, that is analogous to heat sink compound used in electronics? We were thinking about using epoxy filled with ground ferrite to bind the pieces together, but didn't know if that would degrade the flux transfer more than just the ground surfaces in contact with each other.
I heard some place that one pole of a magnet is slightly stronger than the other. Since the coils will be installed on one of the legs of the U, will the flux on that pole be stronger, and thus bebetter used to magnetize the weaker pole of the magnet being magnetized?
Does a laminated core perform better with a direct current magnet than a solid core?
I will have more questions as this progresses, so thank you in advance for your help and patience.

JR

 

[EdStainless]

lamination does not help in DC, but having better steel (higher B, lower H) sure does.
Have you built one of these before?
Be careful when you switch the power off, you will get a huge reactance, and a lot of current.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[JerryRoy]

Thanks for the warning Ed,
No, this is the first go round. Doesn't seem to be a lot of Web information that I can use, so that explains my presence here.
I am looking for some big diodes to handle the reactance. I suppose what I'm looking for is a BIG MOV.
Another question I have is, in a given cross section of iron, and a given flux density, if the cross section were reduced for a portion of the length, would the flux density be greater in the smaller cross section? (Can the flux be "focused" by reducing the area of the pole ends?)
Once again, all thoughts appreciated.

JR

 

[MagBen]

Yes, the epoxy with ground ferrite will degrade flux tranfer due to 1. epoxy acts like air which will decrease permeability, increase reluctance. 2. ferrite has less Bs than other normal soft magnets (why not just use iron powder). when ferrite is saturated, it will also act like air.

The leg with coil(s) will have a stronger magnetic field than the other leg without coil due to 1. the flux leakage, 2. the reluctance of the core material.

A transformer would not operate with a DC. when connected to a DC source, it will produce a short output pulse. The lamination is to reduse eddy current losses which will be converted to heat. If the heat/temperature arise is big concern, you should not use a solid core. Note the flux leakage could even heat up the nearby conductive support structure, let alone the core that "see" the change of majority of flux.

 

[EdStainless]

If you put tapered pole pieces on it then you will get higher flux densities, as long as you don't surpass the Bs of the pole.

MOVs are a good idea, you also want interlocks of some kind so that when you shut power off you can't reapply power too soon. Let the pulse decay first.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[JerryRoy]

Thanks for the input, guys.
Looks like two ground surfaces cheek by jowl is the best way to mate the two core pieces.
And that leaves out the magnetic paste, so that question is moot.
This is a one shot DC power application from a capacitor bank, so there will be no power up until the next mag needs charged. The plan is to use a big motor contactor to handle the current, and they are salvage, so we don't care if the service life is short.
More questions later, I'm sure.

JR

 

[EdStainless]

Back up a bit, capacitors? Then you don't want iron cores.

Just use an air core coil and be done with it.
You would need a very wide pulse to generate the maximum field in the iron, and that would lower your peak current and field.
Our iron core magnetizers used a variac and full 3 phase rectifier bridge as a current source. It took a few seconds to ramp up to full field.

If you do use a pulse use a magnetron tube for the firing switch. They will handle 500,000A without a problem.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[JerryRoy]

Thanks for the addendum, Ed.
Originally, I had thought to use a DC welder as the power source, But my accomplice had 2 Farad cap bank at a couple of hundred volts, and he wanted to use that.
My thought was that the foot pedal on the welder could be used to ramp up the current until it stopped rising, (saturation?), but then the question arose, does a slow down from max magnetizer current reduce the magnetism in the work piece, or does it remain at it's highest attained level as the magnetizer field is reduced? And would the reactance of the magnetizer coil do bad things to the welder? (300 amp Hobart, old school with transformer)
Any advice?

Jerry Roy