Calculating magnetic flux shielding thickness

[Salvador12]

Imagine there is a wire that passes by an area that has magnetic flux in it. My goal is to shield that part of the wire that cuts flux.
Now I know a magnetic field can't be cancelled but it can be redirected which acts as shielding a certain area from flux.

The flux is 50hz AC magnetic flux,(essentially a transformer core) and I take the flux to be around 1 Tesla for simplicity although it might be somewhat less than that.

Now I have bought a permalloy metal foil of 0.1mm thickness and my idea is to create a tube of that foil where multiple layers are wrapped around it + add some copper foil tape on the outside. But ideally I would like to calculate the thickness of foil I actually need to drop the flux within the shielded area to a minimal value.

The foil is ASTM A753 Alloy 3 type of permalloy from what I could gather.

I read that for high frequency flux this is easy as even just micrometers of thickness for high permeability metals suffice to shield a high frequency AC flux, but for lower frequency this thickness increases and for DC it becomes large.

I asked this to chatGPT but it gave me somewhat differing answers from time to time so I just want to double check.
Can anybody help me out ?

 

[D.A.(Tony)Stewart]

1. Normally choosing the right coax is a good start with controlled impedances.
2. Another option is STP cable with balanced signals and PE gnd to shield.

1 T is a very large flux found inside a transformer and not outside.

1. Can you define more precisely define the environment and tolerances?
2. How much crosstalk do you get on this wire and what do you need it to be?
3. Show a physical block diagram with wire gap and distance to this interference.

 

[Salvador12]

So to make matters clearer I'm still coming to terms with the setup myself.
In terms of loop induction that is not a problem, because I have equal flux entering and exiting the loop area on either side of it, therefore induction will not happen. What will happen is the flux that crosses the part of the loop (wire segment) that traverses my said core will experience a perpendicular flux and that might cause local induction the that wire segment contributing to some current in the loop.
Now you said correctly that if this was a high frequency (RF for example) case then simple coax outer shielding would shield the inner conductor completely from any flux, but this is low frequency AC flux. 50 Hz to be exact so I need some thicker and more layers of high permeability metal to create a tube of sorts where on the inside would be my wire that passes through.

STP doesn't help here because I only got 1 wire.


Well there is no distance to the interference, as I said the wire goes through a core, i can only make a hole and wrap some shielding metal around the wire so that the flux warps around it without directly traversing it, that is the idea.

 

[D.A.(Tony)Stewart]

I have no idea what your wire purpose is and why you cannot change the wire to guard it.
You could notch out the interference with a tuned notch filter. by > 100 dB as long as f does not shift in the short term.

I would need more specs. on "must haves" and limitations.

 

[FvM]

STP doesn't help here because I only got 1 wire.


Well there is no distance to the interference, as I said the wire goes through a core, i can only make a hole and wrap some shielding metal around the wire so that the flux warps around it without directly traversing it, that is the idea.

I fear your model is missing the nature of induction. A varying field induces voltage in a conductor loop, shielding a single wire has no effect at all. The shield must enclose the loop.

 

[Easy peasy]

@ FvM - he is talking about magnetic material to shield the wire - this will of course work.

 

[FvM]

he is talking about magnetic material to shield the wire - this will of course work.

I strongly doubt. Ultimately shielding a single wire increases its impedance but not induced voltage.

 

[Easy peasy]

you appear to be confusing electrostatic shielding with magnetic shielding - look up Mu metal

 

[Salvador12]

I fear your model is missing the nature of induction. A varying field induces voltage in a conductor loop, shielding a single wire has no effect at all. The shield must enclose the loop.

The flux is parallel to the loop not perpendicular to it, and it's equal on both sides of it. So there can be no induction in the loop as total because field lines are not "looping" through it like they do in a coil wrapped around a transformer core.
I made the loop geometry specifically so that the flux wouldn't cause induction because as we know you cannot shield a loop from induction if the field lines pass through it. But as @Easy peasy already said you can shield a wire segment from flux by redirecting flux around it and that is exactly what I want to do.

So AC or DC flux doesn't matter because we are not dealing with effects of induction here , we are simply dealing with an area in need of flux shielding by flux redirection around the area.
I think this should be easily achieved by a high permeability multi layer metal with some additional high conductivity foil wrapped around it.

I just wanted to recheck what chatGPT already gave me.

Basically I have a 0.1mm thick high nickel content foil (ASTM A753 Alloy 3) which i have rolled in a small tube of some 6 layers around the wire. To keep the rolled permalloy foil from unrolling + to cause additional shielding I wrapped the foil with a copper tape , also with some few layers.
The copper tape will get eddy currents in it which will deflect some of the field , hopefully the rest of the field will be diverted by the high permeability ASTM A753 Alloy 3 foil that I just described.

 

[Akanimo]

Hi,
Could you make a little sketch to make things clearer? It's difficult to picture the situation.

 

[FvM]

The flux is parallel to the loop not perpendicular to it, and it's equal on both sides of it. So there can be no induction in the loop as total because field lines are not "looping" through it like they do in a coil wrapped around a transformer core.

What are you shielding at all? Either there's no flux through the loop, then you don't experience induction. Or there is some, at least residual, flux, then shielding the wire won't help.

 

[D.A.(Tony)Stewart]

I'd like to see the model or experiment of orthogonal flux with stray capacitance into high impedance mu shielded wire.

It seems there is insufficient data to define the I/O experiment.

This site has many EM 2D 3D models . How many variables are you considering?

I like this one that creates quantum gravity waves from a black hole on your TV. j/k

2-D Vector Field Simulation

 

[Easy peasy]

Now I have bought a permalloy metal foil of 0.1mm thickness and my idea is to create a tube of that foil where multiple layers are wrapped around it + add some copper foil tape on the outside.

and what were your results ?

 

[D.A.(Tony)Stewart]

I had a similar problem to monitor current in a 10kA Zirc-steel cylinder diffusion plasma welder 60 Hz, with my current sense leads in the middle of the loop sensing mV voltage drop across 80cm diameter solid current path. The welder used solid cylinder Cu wheels inside and out of 2 pipes with a Zirconium sleeve shim to diffuse the bond to support 10kA atm. of heavy water. The process used water jets and it looked like a Star Wars battle increasing current as the resistance dropped around the circumference. used for Candu reactor heat exchangers in the late 70's.

The solution was orthogonal RG-58 to an HP strip chart recorder to measure fusion power.

too bad you can't use coax.

 

[Salvador12]

Ok, I guess I will first do the test the way I had planned to see the results and then come back to figure out improvements.

too bad you can't use coax.

Who said I can't use coax, I can use pretty much anything just that it has to be well shielded.
For high frequency coax would suffice as the skin depth is very small. But for something like 50Hz, I will try the method with wrapping a multilayer permalloy sheet into a tube, then insert a plastic spacer inside the tube and put the wire inside the spacer. So it will almost be like a coax. The idea being that the outer layers should redirect the flux through their high permeability path, then airgap to stop as much leftover flux as possible and then in the middle the conductor.

basically like a coax but with better outer shielding because standard coax uses copper for screen and that doesn't cut it for low frequency flux because the copper has a permeability of just 1 so is essentially just air.

 

[D.A.(Tony)Stewart]

I expect grounded coax with a load R to perform better. Good luck.