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

Skin Effect calculation

  1. Mar 11, 2010 #1
    References are available to calculate skin effect, RAC/RDC given wire gauge and frequency. But my problem is a complication of this simple calculation.

    Say I have a 1 Amp RMS AC current component and a selected wire size that gives me a skin effect of 7X the DC resistance. But I also have a hefty 10 Amp DC current component.

    I'm interested in calculating the total I2R losses due to both the AC and DC currents.

    How are these two combined?? For the sake or argument, the DC resistance of the wire is 0.01 Ohms.

    1) I can treat them as separate and independent currents and get a small power loss:

    P = 0.01 Ohms * 102 Amps2 + 0.07 Ohms * 12 Amps2 = 1.07 Watts.

    Or 2) I can imagine that the 7X skin effect also impedes the DC current flow.

    P = 0.07 Ohms * 102 Amps2 + 0.07 Ohms * 12 Amps2 = 7.07 Watts.

    The results are wildly different between the two.

    Which method--or a third method, is correct to first order? I mean, never mind the finer details. I just want to know what my power loss is to +/-20%.
    Last edited: Mar 11, 2010
  2. jcsd
  3. Mar 11, 2010 #2


    User Avatar

    Staff: Mentor

    I believe they are independent effects. Except for the fact that each effect heats up the wire, which increases its resistance some. That is the only interaction that I see.
  4. Mar 11, 2010 #3
    The 1.07 watts is correct. Only the ac current flows in the "skin depth" on the outer surface of the conductor. If you go to stranded wire of the same gauge, the total skin depth resistance is less.

    Bob S
  5. Mar 13, 2010 #4
    This http://www.dartmouth.edu/~sullivan/litzwire/skin.html" is the only semi-useful link I've found on skin effect.

    Its intent is to examine skin effect in transformers where wires are subjected to the magnetic influence of their neighbors, it’s pretty thick reading.

    Given, without proof or motivation, is the resistive power loss due to non-sinusoidal waveforms as the Fourier sum of the power losses due to each sine wave independently.

    P = RDC Σj I2j Frj)

    Supposedly, this includes DC.

    I could over-analyze this problem to death and include the influences of other neighboring wires 'n everything, or get back to finishing the project. But I do like overanalyzing. :)

    Is it too difficult to derive the equation for skin effect of an arbitrary waveform, where at any given moment, the radial force due to a radial distribution of charge cancels the radial force due to the time-rate-change of magnetic field due to longitudinal current?
    Last edited by a moderator: Apr 24, 2017
  6. Mar 15, 2010 #5
    Hey Phrak,

    I had to do this one in college. For the simple wire, a breakdown of the currents by frequency is sufficient.
    When designing transformers / inductors, flux at gaps can increase the loss by cutting across the copper, but this is usually neglected unless you have a very large gap, and when this is a problem, I've never seen a definitive way to compute the loss.

    - Mike
  7. Mar 17, 2010 #6
    Thanks Mike. Do you have a reference source for the frequency independence that specifically includes f=0?

    Secondly, after taking a short look at the simpler problem for a single frequency and single wire, I have to solve some sort of Bernoulli equation using about 3 or 4 out of 4 of Maxwell's equations that are both time and radially variant: Jz(r,t), Er(r,t), Ez(r,t) and Bphi(r,t). Does this sound familiar?
    Last edited: Mar 17, 2010
  8. Mar 17, 2010 #7
    In Smythe "Static and Dynamic Electricity" third edition on pages 372-4 the equations for the ac current density vs. radius in a solid cylindrical conductor are written out using modified Bessel functions of order zero with complex arguments. He gives the effective ac skin effect resistance vs. frequency on page 373.

    Bob S.
  9. Mar 17, 2010 #8
    Electrical Engineer rule of thumb Bessel function for copper (wire)

    skindepth in mm = 260 / [tex]\sqrt{f}[/tex] (in cycles per second)
  10. Mar 18, 2010 #9
    Doh! Of course I meant Bessel rather than Bernolli.

    .....I finally found some cheats here http://home.swipnet.se/swi/bessel/bessel.html#Basic_theory" but he does kind of jump into the middle of it, unexpectedly beginning with the electric field wave equation in a conductive medium, that I'm not familiar with.
    Last edited by a moderator: Apr 24, 2017
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook