Mumetal grain size and magnetic properties correlation

[marcus4]

I have a sample of transformer laminate (mumetal) with unknown magnetic properties. My goal is to achieve the same magnetic properties with raw material that has the same chemical composition and thickness, but I ca't measure the permeability results. Is it possible to do that by comparing the grain size/shape and repeat the heat treatment until they become the same?

 

[MagBen]

Probably not.
Grain size is indeed one of factors which determine magnetics, at a certian range, the bigger the grain size, the less the grain boundaries, the less the pinning sites of magnetic domain displacement, and so the higher the permeability. However, as for Mumetal, the ratio of short-ranged disordered/ordered takes a big role to determine soft magnetics. A same grain size with a different disorder/order structure can create a big difference in permeability.

 

[marcus4]

Thanks. So, if the target's sample permeability is unknown, do you have approach to achieve same magnetics qualities with raw material? I'm aware that repeating the heat treatment and trimming it's parameters will be the solution, but I would like to do it under the lab environment (lab tube furnace with protective atmosphere). Diameter of the tube is large enough for few parts but to prove final results (measuring inductivity of the choke) I need 80 parts. That leads me to large furnace I don't have :/

 

[IRstuff]

Is this a science project? Why are you not just buying the material from one of the many mumetal suppliers?

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[marcus4]

quote/Is this a science project? Why are you not just buying the material from one of the many mumetal suppliers?/quote
Good one. well of corse, material will be from manufacturer or supplier but final parts with target properties have to be done "in the house". If it's possible, of course

 

[MagBen]

optimizing heat treatment depends on chemistry and thickness. if oyu would offer both info. i could recommend heat treatment parameters to fully recover the magnetics.

 

[marcus4]

optimizing heat treatment depends on chemistry and thickness. if oyu would offer both info. i could recommend heat treatment parameters to fully recover the magnetics.

Great, thickmess is 0,35mm but properties of material are posted really rough on manufacturer site. I can achieve wavelength-Dispersive X-Ray Flurescence Spectrometry if it's needed but here is chemistry from their site:
Ni76.6 Cu4.5 Mo3.3 Fe14.7 other% Mn, Si
Target permeability has to be approx. u5=11000 but as I mentioned in first post I can't measure it. Yes, I know it's low permeability for this kind of material but that's the subject of my post here.

 

[MagBen]

try 1100-1200C/4h + 150C/h cool to 700C, then 180C/h cool. better use dry H2 atomsphere (dew point < -40C).
I assume u5 is the permeabilty at H =.005 Oe at 60 Hz. 60-70,000 is expected. initial permeability at B=40Gs, 60 Hz is about 50-60,000. these are impedance permeabilties, if using inductance permeability, the vaules are even higher.

 

[marcus4]

Thanks MagBen. Since I'm not sure about desired properties (permeabilty at H =.005 Oe at 60 Hz. 60-70,000) is i clever to start with lower temperature, say, 1000, than if the permeability is lower than expected to repeat treatment with higher temperature? Of course with H2 protection and clean furnace.

 

[MagBen]

Sounds a feasible plan. Note the cooling rate below 700C is critical to yield partially ordered structure.
High temperature is for grain growth as well as effective removal of C, S. If the temperature is too low, you will get very small grain size and high amount of remaining C, S, both of which could bring you properties far away from desired.

b.t.w., even low Ni grades, say Carpenter High Perm 49, can give you 11,000 at .35mm heavy.

 

[marcus4]

Note the cooling rate below 700C is critical to yield partially ordered structure.

I marked that with red

b.t.w., even low Ni grades, say Carpenter High Perm 49, can give you 11,000 at .35mm heavy.

Yes, I know, but since my goal is to reconstruct the old design I decided to go with the same material as original. Thanks, I will post my results here.

 

[marcus4]

Hi. Here I am gain. I sucseed with 5h at 1000c in pure hydrogen. One thing I would also like to know is the surface color of the mumetal. Mine samples are silver color but I saw some parts with yellowish surface. How to achieve that surface color? Is that some kind of oxide or the effect of the different protective atmosphere? Thanks and Cheers, Marcus

 

[MagBen]

This was probably due to furnace leakage and/or wet/impure H2. heat treated material is supposed to be silver color and shining.

 

[marcus4]

Thanks Magben, so since I need some oxide as isolation layer I can leave the lams at approx 500c to cool down on air and they will probably became yellowish. Am I right?

 

[MagBen]

500C could damage disordered/ordered structure which is critical to magnetics

 

[marcus4]

Thanks Magben. Maybe I sounded confusing. I Mmant to treat the laminations at 1000c for 5 hours and in the cooling process to replace hydrogen with air at 500c. Will that obtain thin oxid isolation layer? Or you have some other advice?

 

[marcus4]

Polite bump I know that my oxide coating question is off topic but please can you give me the advice regarding that? Of course, I can try to experiment but if you are familiar with process it will be great for me to save some material from possible damage. I need this thin oxide layer because laminates need to be isolated from each other because of reducing eddy current effect. Thanks

 

[MagBen]

I doubt your approach will work, the process seems hard to control.
There are lots of other methods to coat your strip.
I donot have a good idea how to create a thin oxide layer for high Ni lams, but you may try to bake at lower temperature, say 350C or lower(to prevent from damaging the formed structure during your previous cooling) for 30-60 minutes at oxygen bearing atmosphere. The exact tempreature and time may be determined by the thickness of oxide layer desired.

 

[marcus4]

Thanks Magben. I will try as you recommend.