0
$\begingroup$

[EDITED]

I am interested in studying the physics behind wireless charging for autonomous vehicles. Therefore, I want to compute the mutual inductance between two coaxial circular coils.

I have read that the mutual inductance between two coaxial circular cables (i.e. individual cables, not coils) is

$M = 4\pi \sqrt{Aa}((\frac{2}{k}-k)F-\frac{2}{k}E)$

Where we have $a$ is the radius of the first circle, $A$ is the radius of the second circle, $h$ the distance between them, and $k^2 = \frac{4Aa}{h^2+(a+A)^2}$.

F and E are elliptic integrals of the first and second type (which must be evaluated numerically), or

$F(k) = \int_0^{\pi/2} \frac{d\phi}{\sqrt{1-k^2\sin^2\phi}}$

$E(k) = \int_0^{\pi/2} d\phi\sqrt{1-k^2\sin^2\phi}$

I can do these integrals numerically in Matlab no problem, which gives me an M value.

Now, I am interested in computing the voltage/power input from the first (driving) coil into the second (receiving) coil. Once I have M, how do i get the emf in the secondary coil? Where does the number of turns in the coil come into play? Or the frequency of the AC signal? The constant for the permeability of free space?

Finally, should I account for back-emf (self-inductance), capacity (resonant transfer), or anything else I haven't considered?

$\endgroup$
4
  • $\begingroup$ If I have understood correctly, you are triying to compute the mutual inductance of two circular current loop. The analytic computation is not easy. For the case of two parallel coaxial circular loops, I think you could easily find the result in a textbook (for example, a very old reference is Becker, electromagnetism and interaction). The result use elliptic integrals. For an efficient wireless transfert, you also have to take into account the capacity of the coils (resonant transfer). $\endgroup$ Jan 31, 2019 at 6:33
  • $\begingroup$ Thank you Vincent. I've been looking online for more information and have also stumbled across some elliptic integrals. ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20010038494.pdf For example, in the above paper, do you think I could just numerically integrate equation (14) over a circle's area to get the magnetic flux through it? Alternatively, I will also look into the textbook you mentioned, although my old reference is in storage right now. Maybe I can pay for it online. $\endgroup$
    – Tucker
    Jan 31, 2019 at 6:56
  • $\begingroup$ Forget my most recent comment. After your suggestion, I changed my search queries to 'mutual inductance of coaxial circular loops', and have been getting much more useful results. I'm in the process of reviewing them, and will reply soon. $\endgroup$
    – Tucker
    Jan 31, 2019 at 7:07
  • $\begingroup$ BTW, elliptic integrals of the 1st kind can be computed using the arithmetic-geometric mean, which converges very quickly. I assume Matlab uses this algorithm, since it is superior to a power series approach, or a generic numerical integrator. $\endgroup$
    – PM 2Ring
    Jan 31, 2019 at 11:34

1 Answer 1

0
$\begingroup$

I'm not sure which one to use

That depends on the units you are using. For SI units you will use the equation without the $c^2$ on the bottom, and if you are using Gaussian or electrostatic units then you will use the one with it.

Once I have M, how do i get the emf in the secondary coil? Where does the number of turns in the coil come into play? Or the frequency of the AC signal? The constant for the permeability of free space?

Once you have M everything else is simply circuit analysis. The emf comes simply as the voltage across the secondary side of the mutual inductance. The number of turns, and the permeability are part of M. The frequency maps inductance to impedance in the usual manner.

$\endgroup$
2
  • $\begingroup$ Thanks for your response. I'll look into the 'frequency mapping inductance to impedance' as I am not familiar with that process. However, I am not convinced that the number of turns is 'encapsulated' within M. The equation listed in the original question is from Maxwell and describes the mutual inductance of two single-loop coils, not coils of N loops. I reason that if the number of loops in the first coil increases by N1 times, then M also increases by N1 times. Since M12 = M21 = M, this should also be true for N2. So the total M is M*N1*N2. Does that sound correct to you? $\endgroup$
    – Tucker
    Feb 1, 2019 at 10:06
  • $\begingroup$ The equation above is the equation to calculate M for a single loop only. You would need a different equation to calculate M for a multi loop coil. I don’t know what that would be, but multiplication by the number of turns would probably be a good approximation. $\endgroup$
    – Dale
    Feb 1, 2019 at 12:09

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.