1
$\begingroup$

I'm working on a field configuration that needs to die off rapidly and I got to a $1/r^5$ dependence with canceling the dipole moment of the system cleverly, but to go get better arrangements I need to figure out the higher order moments of a simple loop as well.

Griffiths (page 243, 3rd ed) refers to this integral $\oint (r')^{n}P_n(cos \theta ') d L'$, with $P_n$ as the Legendre polynomials but I am unclear on how to go about evaluating it for higher moments because for the dipole term Griffiths does a clever trick that I'm not super sure extrapolates. I feel like the current loop is such a simple, tidy system it's higher moments would be tabulated somewhere just as an example of how higher orders are calculated, but I'm unable to find them after a few a days of googling and librarying.

$\endgroup$

1 Answer 1

1
$\begingroup$

Whenever $|\mathbf{r}|$ lies far outside the loop, we can expand the vector potential $\mathbf{A}(\mathbf{r})$ in an irregular solid harmonic series $$\mathbf{A}(\mathbf{r})=\frac{\mu_0}{4\pi}\int\frac{\mathbf{J}(\mathbf{r}')}{|\mathbf{r}-\mathbf{r} '|}dV=\frac{\mu_0}{4\pi}\sum_{L=0}^\infty\sum_{m=-L}^L I_{Lm}(\mathbf{r})\langle R_{Lm},\mathbf{J}\rangle$$ where $R_{Lm},I_{Lm}$ are the regular and irregular real solid harmonic functions and $\langle R_{Lm},\mathbf{J}\rangle$ is the integral inner product between the two (aka, the $Lm$ multipole moment).

Letting the loop with radius $\rho$ lie in the $xy$-plane with $z$-axis height $h$, the interior moments $\langle R_{Lm},\mathbf{J}\rangle$ can be computed as $$\langle R_{Lm},\mathbf{J}\rangle=\int_0^{2\pi}R_{Lm}\left(\sqrt{\rho^2+h^2},\arctan(h,\rho),\phi\right)\mathbf{J}(\phi)\,\rho d\phi.$$ Letting the loop current flow counterclockwise, in Cartesian coordinates we have $$\mathbf{J}(\phi)= \left( \begin{array}{cc} -J_0\sin(\phi)\\ J_0\cos(\phi)\\0 \end{array} \right) $$ and we obtain the following moments up to octupole order (copy paste into Mathematica and hit Cell > Convert To > StandardForm to properly format it):

SphericalHarmonicYr[L_, m_, \[Theta]_, \[Phi]_] := (-1)^m*
   Sqrt[((2*L + 1)*(L - Abs[m])!)/((2*Pi)*(L + Abs[m])!)]*
   LegendreP[L, Abs[m], Cos[\[Theta]]]*
       Piecewise[{{Sin[Abs[m]*\[Phi]], m < 0}, {Sqrt[1/2], 
      m == 0}, {Cos[Abs[m]*\[Phi]], m > 0}}]; 
SolidHarmonicIr[l_, m_, 
   r_, \[Theta]_, \[Phi]_] := (Sqrt[(4*Pi)/(2*l + 1)]*
     SphericalHarmonicYr[l, m, \[Theta], \[Phi]])/r^(l + 1); 
SolidHarmonicRr[l_, m_, r_, \[Theta]_, \[Phi]_] := 
  Sqrt[(4*Pi)/(2*l + 1)]*r^l*
   SphericalHarmonicYr[l, m, \[Theta], \[Phi]]; 
J[\[Phi]_] := J0*{-Sin[\[Phi]], Cos[\[Phi]], 0}; 
Grid[Simplify[
  Table[MatrixForm[
    Integrate[\[Rho]*J[\[Phi]]*
      SolidHarmonicRr[L, m, Sqrt[\[Rho]^2 + h^2], 
       ArcTan[h, \[Rho]], \[Phi]], {\[Phi], 0, 2*Pi}]], 
       {L, 0, 3}, {m, -L, L}], 
  Assumptions -> {\[Rho] > 0, Element[h, Reals]}], Frame -> All]

enter image description here

The code above computes the moments out to $L=3$, but you can change 3 to whatever multipole order is necessary. Here is a numerical verification that the irregular solid harmonic series $\frac{\mu_0}{4\pi}\sum_{L=0}^\infty\sum_{m=-L}^L I_{Lm}(\mathbf{r})\langle R_{Lm},\mathbf{J}\rangle$ and the integral $\frac{\mu_0}{4\pi}\int\frac{\mathbf{J}(\mathbf{r}')}{|\mathbf{r}-\mathbf{r} '|}dV$ are actually the same thing (again, hit Cell > Convert To > StandardForm to properly format the appearance):

Subscript[Q, L_, m_] := 
  Integrate[\[Rho]*J[\[Phi]]*
    SolidHarmonicRr[L, m, Sqrt[\[Rho]^2 + h^2], 
     ArcTan[h, \[Rho]], \[Phi]], {\[Phi], 0, 2*Pi}]; 
r = N[{5, 6, 7}]; 
rSph = CoordinateTransform["Cartesian" -> "Spherical", r]; 
h = 2; 
\[Rho] = 1; 
J0 = 2; 
NIntegrate[(J[\[Phi]]/
    Norm[r - {\[Rho]*Cos[\[Phi]], \[Rho]*Sin[\[Phi]], 
       h}])*\[Rho], {\[Phi], 0, 2*Pi}]
Sum[(Subscript[Q, L, m]*SolidHarmonicIr[L, m, ##1] & ) @@ rSph, {L, 
  0, 5}, {m, -L, L}]
Clear[h, \[Rho], J0]

{-0.0471743, 0.039312, 0.}

{-0.0471859, 0.0393216, 0.}

Alternately, as an additional simplification, note that $\mathbf{J}$ can be expressed in spherical vector coordinates as $$\mathbf{J}(\phi)=\left( \begin{array}{cc} 0\\ 0\\ J_0 \end{array} \right)$$ giving (this time up to $L=4$):

J[\[Phi]_] := {0, 0, J0}; 
Grid[Simplify[
  Table[MatrixForm[
    Integrate[\[Rho]*J[\[Phi]]*
      SolidHarmonicRr[L, m, Sqrt[\[Rho]^2 + h^2], 
       ArcTan[h, \[Rho]], \[Phi]], {\[Phi], 0, 2*Pi}]], 
       {L, 0, 4}, {m, -L, L}], 
  Assumptions -> {\[Rho] > 0, Element[h, Reals]}], Frame -> All]

enter image description here

although for this coordinate choice I'm a bit confused as to why the $L=0$ coefficient is not zero (will have to check that later).

The magnetic field can then be obtained from the above coefficients $\langle R_{Lm},\mathbf{J}\rangle$ by $$\mathbf{B}(\mathbf{r})=\nabla\times\mathbf{A}(\mathbf{r})=\frac{\mu_0}{4\pi}\sum_{L=0}^\infty\sum_{m=-L}^L \nabla\times I_{Lm}(\mathbf{r})\langle R_{Lm},\mathbf{J}\rangle.$$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.