Magnetic shield for power losses reductions

Hi Marco
Think you're in the wrong path. Only thing one could (?) do is optimising geometry of your machine cores so to minimize escaping magnetic flux. And this is not really easy since their function sets shape limitations. Once magnetic flux is out the only thing that can be done to minimize losses is keeping magnetic-contactive materials as far as possible so they do not pose as."loads". .Not much gain to expect though, electromagnetic losses at mains frequencies are relatively small overall factor.
And any "shielding" in the same sense you shield your speaker magnets has opposite results, (loss increase)

This is a problem that needs to fixed at the souce. Use low leakage magnetic components in the power supply. Use good wiring discipline to avaoid stray fields.

SimpleMind and TVP45 have the correct idea. Once your magnetic field generators (inductors) deliver deliver energy into free space, any external shield will just direct this energy into something and rob some power. This is not a simple problem that will be easily fixed by a few patches.

Think of it this way. If you make an inductor from a straight iron rod, like the letter "I". The magnetic field from the top of the "I" to the bottom migrate through any and all paths outside of the bar. It will have the highest concentrations at the top and bottom of the "I" but all of the energy must travel from one end to the other. Now if your inductor was fabricated from an iron rod in the shape of the letter "C". Now a clear easy path across the gap of the "C" will be established for the magnetic field. Most of your magnetic energy will be directed across this gap with only fringe field effects now permeating into free space. But this is just the beginning of the "black art" procedure to minimize stray fields.

I don't know whether this is on track with your needs, but...

I've messed around with those insanely powerful neodymium magnets. I have a couple of 700 lb pullers that are about three inches in diameter. If you drop them down a copper pipe, the eddy currents are so intense that the magnets float slowly to the bottom... So, like would copper shielding help?

Vermin, Redfres, TVP45

Thank you for the pointers.

Unfortunately, I need the leakage from the transformer and the air core inductor to "tune" the capacitive load. I know that the field has to flow through space and its basic distribution.

Unfortunately, space constraints prevent using a much larger cabinet and most of them are made of mild steel.

What I am looking for is a way to conduct the fields just before they reach the cabinet walls. This would prevent power losses if this material is a better magnetic conductor than the mild steel and has very little eddy current losses in the KHz region.

Here is a simple sketch:

Nice drawings that makes my points easier to follow. (Wikipedia presents a nice article on magnetic fields.) Unfortunately you imply a whole batch of unknowns (e.g. frequencies, energy level, cabinet geometry, cabinet material) that will make it difficult if not impossible to give you a definitive answer. If this was physically in front of me, I'd be probing around this chassis's free space with a fixed small test loop attached to an integrator circuit and an oscilloscope to first determine what flux strengths you were dealing with and from where. (Remember to wear your PPE. Yes, work last week coerced me to attend a safety class.)

Since the core material geometry appears to be more of a bar magnet than a horseshoe magnet, the magnetic flux resides predominantly in free space in all directions outside of the bar magnet. Now remember most of the magnetic flux resides in free space, not in the magnetic core itself. This will be important later. So your soft iron cabinet walls are already acting as a shield. Adding another shield inside this shield will at best mean that your power losses will now be contained inside your chassis. You may not have reduced you losses at all. Now instead of shielding, I believe that you will get lower energy losses by pursuing two procedures. First, steer the magnetic fields outside of the bar magnets you presently have to an air gap, internal to your chassis shield. You do this with magnetic material of a higher permeability than free space (iron). Remember all magnetic fields are dipoles. The field runs from one pole to the other. This alone will dramatically lower the stray field your chassis will be attenuating. Second, reduce the power losses in doing this steering by using laminar magnetic material to break up the eddy currents. Just like the lacquered laminar sheets of iron that make the core of a transformer. This will mean that the magnetic field steered by your addition should not be getting overly hot (wasted power). As a final improvement, examine the remnant field that your chassis (shield) will be stopping. (You did build that test loop, didn't you. ) Identify the eddy current paths your shield contains and find as many places as possible to put non-conductive breaks (paint, insulating stand offs, plastic shims) so that these currents cannot form.

Best practical materials for magnetics at high frequencies are the ferrites. They have a much higher permeability that air (but not as good as iron) and the lowest eddy current losses. A layer of ferrite material might shunt enough of the field away from your cabinet wall to reduce the heating.

Wooden cabinets?

RedFred is on the right track as well. This type of scenario occurs in transformer covers when the HV bushings are close together & heavy current is passed through. The best way to break it up is with Non-magnetic inserts. If the cabinets are mild steel then you need non-magnetic stainless steel inserts. I have found that some stainless cabinets are magnetic & so they are useless in this situation.

Where the cable entry & exits are is where you should find the most heating, so inserts between the cables is a good start. Then look at the areas outside of that & another insert to the edge of the flat surface would be a good idea. Each suface should have the inserts at different directions so as to break-up or elongate the flux path. That will reduce the heating. Your inserts should also be about 1 to 2mm (or 1/16th inch) thicker so that welding it in is easier & helps with heat dissapation.

If you leave an air gap or have plastic inserts etc you may build up a charge that can short across as the Flux is quite powerful & can "leap" small gaps. The inserts should be at least 2/3rd the width/ Dia of the cable passing through. That should do the trick.

Dear Marco

Die Antworten sind meiner Meinung nach richtig.Verluste Elektromagnetische Art sind eine Folge der Konstruktion des Generators und unvermeidlich.Abschirmungen erzeugen Gegenfelder und saugen das Feld aus.

But the generator must be contained.Right answer:regain Thermal

Energie for heating.Canada is a cold area!

Dear Gunter,

I will try to get some meaning from your post by sending it to translation. My understanding of German language is very limited. You can try French or Italian and of course English. I never had a German girlfriend to motivate mu studies of your language...

Bye.

Thank you to all who are trying to help me solve this. I appreciate the suggestions.

Welcome to the world of cost engineering. Perhaps the following approach may help.

Rather than combat the effects of the electromagnetic radiation by shorting it with mu metal or other wasteful methods, capture and use it. Think of surrounding the volume with a set of copper wires and capture and use it, like another set of transformer windings. By embedding the entire assembly as essentially the primary side of a transformer, think of the copper wire shielding as the secondary. This energy could then be captured and used in traditional fashion. Just a thought.

Honestly, I was reading this thread hoping to get to the last post without finding such a solution like the one that you offer here, because it is absolutely correct (though, not so easy to accomplish) and exactly what the OP requested. I would only add that the wound flat coil would ned to be as close as posible to the surface of the 250 KW Power supply and that he would need to rectify the output and use it as DC power supply for a section of his control or lighting or indicator panel Etc.

Great minds think alike

If I were registered you'd get a GA

Yahlasit

Well what's keeping you from registering. Last time I heard it was free. Think all of us can use a good laugh with "great minds" like your's that comment on issues they don't have a clue about.

Don't take ofense, I didn't mean "great minds..." as opposed to your user name; you surely know it is just a saying, but now that you mention,I had to go thru the learning of eddy currents, heat treatment of core laminations, magnetic flux and a couple more things to work on a core-loss test machine for the motors that we manufacture.

But you might still be right and I don't have a clue, specially on "magnetic-contactive materials"; could you please tell me what they are?

Yahlasit

None taken. By "conductive" I mean of course electrically conductive. A material can be conductive,magnetic,both or neither . Now for the subject if you put conductive material at any orientation combinations as you suggest to collect "lost" power a) where do you think this power will come from? b)do you think this will keep the load that the "transmitting" machine sees constant and you 'll just be there to collect the lost power without loading it more on top of the former loss? Wow!

Oh, ok, is just that I thought I read "contactive" and then imagined about whole new technology, but never mind.

Don't assume I meant "any orientation combinations; you know how a magnetic field has to cross a conductor in order to efectively induce.

About your two points:

A)from the magnetic fields produced by the magnetic components of the OP's equipment, which lose a few KW as stated by the OP.

B) No, he probably will have to work below those variations and build a power supply with feedback to take advantage of them or use capacitors to store surplus and release it on the next v.drop.

Yahlasit

Gentlemen,

Don't get too excited with the brain storming, I am not going to build a power supply within a power supply. There is nothing useful to do with this power that would probably comes at the expense of the overall efficiency.

I am looking for a simple way of redirecting the magnetic lines to the sources with a minimum of losses. i.e. minimum dampening.

Thank you for your effort. It is appreciated.

Marco

In reference to the cabinet material, it was common in the Automotive industry for composite panels to utilize a conductive in mold coating, this requirement allowed for electrostatic painting to occur on the composite, the actual conductive properties I am unsure of but Omnova Solutions, was the vendor that we utilized that may have some insight. Composites can be a cost effective replacement for mild steel in an electrical cabinet application, and the formulations for in mold coating and composites can be manipulated to suit many different scenarios.

I am way out of my field here. I am not an electrical engineer. Magnetic fields have always fascinated me especially when I was working below the water line on ships.

All our training told us that mines and torpedoes were just myths invented to make the black gang behave themselves and you shouldn't think about them to much.

Meanwhile, over in the corner of the engine room was a great big grey cabinet that was hooked up to a coil that ran around the ship. With this coil we would induce a field around the ship that was equal and opposite to the magnetic signature of the ship. This would take a few hours of cruising up and down and across the degaussing range.

If a coil on the inside of a ship, which is just a funny shaped conductive metal box, can do that would not a similar coil on the inside of a cabinet be able to neutralise the unwanted field?

If the problem is the cost of producing this neutralising field then an alternative would be to absorb it and use use it for something useful rather than heating up the walls of the cabinet.

BAB

You've got a point there, Very theoretically it could be done this way, symmetric fields cancelling each other, we can call it active shielding. Not sure if possible to accomplish in 3D space and at what power cost. Power on the other hand is what we want to save. I'll give a GA for creative thinking.

This will work to cancel the magnetic field being generated outside of the supply box. Kudos for being creative. However, this will be less efficient than the original supply.

Use toroidal cores for the transformers and inductors. This will confine the field and keep it localized to the component. The losses from the connecting wiring or circuit traces could be moderated by designing the conductors as transmission lines.

Have FUN!
TT3

Thank you for the suggestion but do you know of any thoroidal transformer core capable of handeling 450KVA at 1500Hz?

That would be a "special" I believe! Just because no one is currently making it, doesn't mean that it is not a good idea. You are obviously working off the map with your project, what ever it is.

You asked for ways to shield the stray fields to prevent losses, I gave you the only way I know of that contains the field and is not terribly inefficient. I believe that one could construct a tape wound core that would handle the power level and frequency you specified. It would be heavy and big but it would largely contain the field.

At any rate it was only a Turbotrollish suggestion. Perhaps useful, perhaps not, but you did not pay a lot for it either!

Audios!
TT3

Please don't be offended by my sincere question. I wasn't cynical, I was inquiring if you had a source.

By the way, this is a real product. We sell almost a hundred power stages a year. It ends up being many hundred KW of continuous losses since they operate almost continually in water treatment plants and oil refineries.

I am not offended, well after thinking about it a while, I am not. I have no idea who could or would make toroids of that size. My suggestion of using a tape wound core should be possible. I personally have never used, seen or heard of a core in the size required for your application, but I'll bet that someone somewhere has made one or knows why it won't work. The trick will be finding them.

I would start by contacting the manufacturers of toroidal transformers.

Edited to add: The way common pole pig power transformers are currently made may give some clues. They are designed for efficency and one of the ways they do that is by containing the field so that the eddy currents in the case are minimized. Wrong frequency I know, but different number of turns and different core material may give you some ideas.

Good luck,

TT3

OK. One more uneducated answer... What about a couple of large pieces of circuit board material with that square, spiral pattern etched into it?

That will transfer some of the heating of the cabinet, to the copper. Any gain???

Using a material with better electrical conduction is opposite to what the transformer manufacturers do. They use thin silicon steel with good magnetic property but with high electrical resistivity. This reduces the eddy current losses.

This is the type of material I am looking for in a friendlier format than 3' x 8' sheets of thin metal.

Thank you.

Marco

Not quite up to the initial 250kW or the later 450kVA rating, these people claim to have at least 150kW capabilities...

http://www.cefem.fr/pages/en/produit/magnetique/directindustrie/tfohf.htm

Just curious, which power level is it? Is that CW, quasi-resonant, or pulsed?

"I am looking for a magnetic material in sheet format that have a high electrical resistivity to minimize the eddy current. I would line my cabinets with it and conduct most of the magnetic lines back to the source away from the cabinet's steel walls."

If you cannot afford a layer of ferrite plates {material spec} on two or more interior walls of the cabinet, then I suggest you turn the thermostat (heat) down a few degrees and circulate the warmth from the cabinets around the room. Just add a small fan and your winter heating bill might go down a little.

Thank you for the link.

Our transformer steps up the voltage from the inverter from 800V to almost 4KV, The load is an ozone generator which has a poor power factor (leading). We need to circulate 450KVA to transfer 250KW to this highly capacitive load.

The leakage inductance from the transformer is used to improve the power factor seen by the inverter.

We often group a few of these power stages together to supply over 1MW to large generators making many thousand of pounds of ozone a day.