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

Maxwell 3D eddy current simulation for a cylindrical wire

  1. Nov 28, 2011 #1
    Hi, I am trying to do a simulation for the current distribution in a cylindrical conductor wire using MAXWELL 3D. I used eddy current solver and assigned excitation of current 1A at both ends of the round wire. (I set length L=2mm, radius R=1mm) The field plot seems ok, having highest current density J near the surface of the wire, and lowest in the center. However, the total current calculated by integrate(J,cross-section) is NOT 1A for higher frequencies. Could someone tell me what was wrong in this simulation? Thank you!
  2. jcsd
  3. Nov 28, 2011 #2


    User Avatar

    I've never used maxwell 3d so I doubt I'll be much help but, how far off is it?

    This is just a guess, but is there a setting to specify how many points are used to represent the field? As the frequency goes higher you could have much of the period between the points which would distort the integration. This is assuming the low frequency simulations give the correct answer.
  4. Nov 28, 2011 #3
    Thanks for the reply. It about 30% of the assigned current. I used very fine meshes and the maximum size of the mesh is kept smaller than the skin depth..I am wondering whether the length of the wire will effect the simulation result?
  5. Nov 28, 2011 #4
    I don't know the program but I won't use eddy current model, it is not. You use normal EM model and solve it. I don't understand why you set current on both ends. You set the current at one end and let it propagate down. Also you have to be careful on the reflection along the line if you don't have a termination. Even if you use a single line, you still have a transmission line going and you'll have reflection, current is not going to be the same at every point along the line. You'll have a standing wave pattern of the current. You really have to be careful to set the parameters.

    You can easily do calculation by hand for this to verify. Look up in the Engineering EM book and there all formulas about EM wave and skin depth for good conductors. If you can't find them, come back and let us know, I will pull out some of my notes and list the formulas. I just don't have it from my memory right at this moment. This involve forward and reflected phasor.
    Last edited: Nov 28, 2011
  6. Nov 28, 2011 #5
    Thank you. The purpose of this work is to check/verify whether the Maxwell 3D code is able to simulate AC current correctly so that more complicated simulation can be done later. It seems that only the eddy current solver in the code handles skin effect.. I had the analytical
    expression for the simple cylinder also..Maybe the inconsistency is caused by reflection as you mentioned..Thank you for reminding me this..
  7. Nov 28, 2011 #6
    I am not positive, but isn't eddy current only happened with ferromagnetic material like iron etc. I am not sure, just a question. I assume your wire is normal copper wire, or you assume ferromagnetic material?

    But either way, even a single wire hanging in air, you still have to have a return path even though it's not obvious. So the impedance might be close to free space or something 377Ω characteristic impedance or something like that. So unless you have perfect match, you are going to have a standing wave. Try varying the length and if you see the current change, it's likely to be reflection.
  8. Nov 29, 2011 #7


    User Avatar
    Gold Member


    Eddy Current Solution type in Maxwell.. Frequency domain. It assumes materials are linear and excitations are sinusoidal.

    well.. that's how maxwell excitations are set - on both ends. No return path is needed then

    Ok, my guess (I haven't tried to solve high frequency problems) - check displacement current options (Maxwell3D->Excitations->Set Eddy Effects). Another one - be careful when using Maxwell Calculator - many problems have sources in formulas used there.

    You might also note, that not only EC solver supports AC excitations (check transient solver).
    Last edited: Nov 29, 2011
  9. Nov 29, 2011 #8
    Hi. Thanks. It seems little difference with and without displacement current options, as I can manually check the displacement is indeed negligible in the setup. I am not sure whether it is due to the problem of Maxwell Calculator.. I guess I could try the transient solver..
  10. Nov 29, 2011 #9


    User Avatar
    Gold Member

    What is the frequency of Your current ? For higher freqs there will be difference (with or without displacement), it's obvious - the question is - how big.

    One more lead.. try to use ComplexMag of J or some other formulas in Calculator. I think, that here is the source of Your problem.

    About transient solver: it's far more complicated than EC, but provides more possibilities as well. Anyway, for this kind of problem I would rather stick to EC - it will save You a lot of time, believe me.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook