Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Looking for what I thought would be a simple Permeability Q

  1. Mar 26, 2015 #1
    Hi, I've been thinking about how to calculate a permeability curve, I thought this would be an easy to find online but unfortunately I haven't had any luck.
    From what I can graphically see below:
    perm.PNG
    It appears to me that mu for a ferromagnetic material is proportional to the derivative of B vs H, but if this is true, and if so what the actual relation is I haven't been able to find.

    Anyone know?

    Thanks
     
  2. jcsd
  3. Mar 26, 2015 #2

    Svein

    User Avatar
    Science Advisor

  4. Mar 26, 2015 #3

    NascentOxygen

    User Avatar

    Staff: Mentor

    That's an astute observation! The graph does indeed bear that resemblance.

    Let's see .....

    B and H are related in the usual way, viz., B=μH

    Since μ changes with H, we can say μ is a function of H, writing this as μ(H). Accordingly, we can
    write the equation more generally as B = μ(H).H

    So, dB/dH = H.dμ/dH + μ

    Now, on the right side if the first term were small in relation to the second (and you would need to look at typical values and graphs to see where this approximate could apply) you'd have your approximation

    μ ≈ dB/dH http://imageshack.com/a/img29/6853/xn4n.gif [Broken]
     
    Last edited by a moderator: May 7, 2017
  5. Mar 26, 2015 #4
    Hi, I remember what a hysteresis curve is, but I don't see what it has to do with finding the mu of a steel.

    I'm sorry it's been so long since I've done multivariable calculus, I'm sure that was simple but could you tell me the working, did you use implicit or partial differentiation? Or just the old product or chain rule?

    So since μ = dB/dH - H.dμ/dH
    to calculate current μ you could just use a previous value of μ in H.dμ/dH, for a really close approximation. What units would dB/dH be in? Because B is in Teslas, I'm assuming it would be in too, as the graph indicates, but I notice it is H/m, does this mean the graph isn't scaled correctly?

    Thanks
     
    Last edited by a moderator: May 7, 2017
  6. Mar 27, 2015 #5

    NascentOxygen

    User Avatar

    Staff: Mentor

    In that differentiation I used the derivative of a product rule.

    dB/dH will have the same units as B/H.
     
  7. Mar 27, 2015 #6
    Ok so the actual working is:
    dB/dH = H.dμ(H)/dH + μ(H)*dH/dB

    but we don't worry that μ is a function of H, but what happened to the dH/dB, why was that zero?

    Ok, so if μ's units are [H/m] but on the graph μmax is like probably over a Tesla, they've taken liberties with its scale I assume.
     
  8. Mar 27, 2015 #7

    NascentOxygen

    User Avatar

    Staff: Mentor

    Nothing happened to dH/dB, because it isn't there. That term is dH/dH
     
  9. Mar 27, 2015 #8

    NascentOxygen

    User Avatar

    Staff: Mentor

    Going right back to the start ....

    You have a graph of B vs H, so why can't you take values of B and divide by the corresponding H to obtain μ?
     
  10. Mar 27, 2015 #9
    Yeah I suppose you're right, I was just thinking it was strange that μ is an inherent property of the material that causes be, yet we use B to find it. So I was thinking (not that I achieved it) that there was some expression for μ without B.
    Though, how is that term dH/dH ?
    I was thinking let u = μ, v = H
    d(u.v)/dB = μ*dH/dB + H*dμ/dB

    what am I doing wrong?
    Thanks
     
  11. Mar 27, 2015 #10

    NascentOxygen

    User Avatar

    Staff: Mentor

    d(u.v)/dH
     
  12. Mar 29, 2015 #11
    OMG, what was I thinking. Thanks!
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Looking for what I thought would be a simple Permeability Q
Loading...