0% found this document useful (0 votes)
173 views6 pages

Magnetfeld und Vektorpotential in Kabeln

Uploaded by

muelerjoerg
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
173 views6 pages

Magnetfeld und Vektorpotential in Kabeln

Uploaded by

muelerjoerg
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Musterlösung zu Aufgabenblatt 10 – Klassische Elektrodynamik WS 21/22

Abgabe: Dienstag, 18.01.2022

Prof. Dr. W. Hofstetter - Institut für Theoretische Physik, Universität Frankfurt

Übungsleiter: Dr. Morad Ebrahimkhas

[Link] Elektrodynamik WS 21 22

Aufgabe 30: Magnetische Suszeptibilität und Permeabilität (9 Pkte)

Ein koaxiales Kabel besteht aus zwei unendlich ausgedehnten zylindrischen Rohrleitungen (Radien a und b, a < b),
zwischen denen sich ein Isolator mit magnetischer Suszeptibilität χm befindet. Im inneren Zylinder fließt ein Strom
I, im äußeren ein entgegengerichteter Strom gleichen Betrages. Der jeweilige Strom verteilt sich gleichmäßig auf
der Oberfläche des jeweiligen Leiters (siehe Abbildung).
z
a) (3 Pkte) Berechnen Sie das Magnetfeld im gesamten Raum über das H-Feld.
SOLUTION: b a
We shall use the Ampère’s law to find the auxiliary H-field, that is:
I
H · dl = ILf ree ,
L I I

where ILf ree is the total free current passing through the closed Ampèrian loop.
Before applying Ampère’s law, it is useful to find the direction of the H-field from symmetry
arguments, as was done in the lecture for the infinite plane with a constant homogeneous cur-
rent (Skript S. 144). We will use cylindrical coordinates (r, ϕ, z). Firstly, the field cannot have any z-component,
as the Biot-Savart law tells us that the magnetic field is perpendicular to the direction of the current. Second-
ly, the field cannot have any r-component, as if the field was pointing away from the cable, by reversing the
direction of the current one could make it point toward the cable, because in the Biot-Savart law changing the
sign of the current switches the sign of the field. However, the r-component of H cannot depend on the ±z
direction of the current in the cable, as one can see from rotation symmetry. Thus the H field can only have a
ϕ component.
For the Ampèrian loop, we thus take a simple circular contour of radius r, and obtain

2πrHϕ = Irf ree .

For region 1, r < a, and region 3, r > b, we have Irf ree = 0, thus

H1,ϕ = H3,ϕ = 0.

For region 2, a < r < b, one gets


I
2πrH2,ϕ = I ⇒ H2,ϕ = .
2πr
In vacuum, the H-field and magnetic field B are related by B = µ0 H, while in a linear magnetic material we have
B = µ0 (1 + χm )H where χm is the magnetic susceptibility. Therefore, the magnetic field is here given by

0,
 r < a,
I
B(r) = µ0 (1 + χm ) 2πr ϕ̂, a < r < b,

0, r > b.

b) (3 Pkte) Berechnen Sie nun die Magnetisierung und die gebundenen Ströme im gesamten Raum. Berechnen Sie
nun das Magnetfeld explizit aus den freien und gebundenen Strömen und vergleichen Sie das Ergebnis mit dem
Resultat aus (a).
SOLUTION:
In linear media, the magnetization depends on the H-field as M = χm H. Using the results from the Ampère’s law

1
for the auxiliary H-field that we have obtained in question a), we directly obtain the magnetization in regions 1, 2
and 3:
I
M2 = χm ϕ̂, M1 = M3 = 0.
2πr
Corresponding to this magnetization M is a volume bound current,

jb = ∇ × M,

throughout the material, and a surface bound current,

Kb = M × n̂,

on the boundaries, where n̂ is the unit vector normal to the surface boundary.
Note that in cylindrical coordinates (r, ϕ, z), we have:
     
1 ∂Mz ∂Mϕ ∂Mr ∂Mz 1 ∂ ∂Mr
∇×M≡ − r̂ + − ϕ̂ + (rMϕ ) − ẑ.
r ∂ϕ ∂z ∂z ∂r r ∂r ∂ϕ

From the magnetization M calculated above, one sees that the volume bound current is zero in all three regions,
as
∂Mϕ 1 ∂
jb = ∇ × M = − r̂ + (rMϕ )ẑ = 0.
∂z r ∂r
(a)
On the other hand, the surface bound currents (with n̂2→1 being vector perpendicular to the inner surface at r = a
(b)
and n̂2→3 to the outer surface at r = b, both pointing away from the magnetic medium) are obtained as

(a) (a) I
Kb = M × n̂2→1 = M × (−r̂) = χm ẑ,
2πa
(b) (b) I
Kb = M × n̂2→3 = M × (+r̂) = −χm ẑ.
2πb

Now from these bound currents we can find the magnetic field directly from Ampère’s law,
I
B · dl = µ0 IL ,
L

and confirm that the bound current together with the free currents generate the correct magnetic field as calculated
in a). As Ampèrian loop, we use again a circular contour of radius r.
There is no bound or free current in region 1 (r < a), thus B1,ϕ = 0.
In the third region (r > b) all enclosed currents (i.e. the upward and downward-going free currents ±I, as well as
(a) (b)
the surface bound currents Kb and Kb at the two boundaries) cancel each other, like so:

tot I I
I3,r = I − I + χm 2πa − χm 2πb = 0,
2πa 2πb
so we get B3,ϕ = 0.
Hence, the only non-vanishing enclosed currents are given by a loop in the second region (a < r < b):

tot I
I2,r = I + χm 2πa = (1 + χm )I.
2πa
tot
Finally, by applying Ampère’s law with the enclosed current I2,r above, we obtain the same result for the magnetic
field as in a): 
 0, r < a;
I
B(r) = µ0 (1 + χm ) 2πr ϕ̂, a < r < b;
0, r > b.

c) (3 Pkte) Berechnen Sie das Vektorpotential A(r) im gesamten Raum.


SOLUTION:
We here assume that there exists a choice of gauge such that the vector potential A(r) can be chosen of the form
(in cylindrical coordinates): A(r) = (0, 0, Az (r)), with ∇ · A = 0. We can then calculate the vector potential
A(r) in two ways.

2
First alternative:
We will start with
∇ × A(r) = B(r),
with A(r) = (0, 0, Az (r)). Using the result for the magnetic field obtained in a) and the expression for the curl
in cylindrical coordinates (see formula in a)), we get:


− ∂r Az (r) = 0 ⇒ Az (r) = C1 ,
 for r < a,
∂ I I
− ∂r Az (r) = µ0 (1 + χm ) 2πr ⇒ Az (r) = −µ0 (1 + χm ) 2π ln(r) + C2 for a < r < b,
 ∂

− ∂r Az (r) = 0 ⇒ Az (r) = C3 for r > b.

In a gauge where A is divergenceless (∇ · A = 0, like in the Coulomb gauge of magnetostatics), the vector
potential is continuous across any boundary and we obtain the matching conditions:
(
I
C1 = −µ0 (1 + χm ) 2π ln(a) + C2 at r = a,
I
C3 = −µ0 (1 + χm ) 2π ln(b) + C2 at r = b.
This finally gives the expression for the vector potential in all three regions:

I
−µ0 (1 + χm ) 2π ln(a) + C2 for r < a,

I
Az (r) = −µ0 (1 + χm ) 2π ln(r) + C2 for a < r < b,
I

−µ0 (1 + χm ) 2π ln(b) + C2 for r > b.

As the observable magnetic field B depends on the vector potential A through B = ∇ × A, one can set for
example C2 = 0 without physical consequence.

Second alternative:
As in all three regions (except boundaries) the free current density is zero, jf = 0, Ampère’s law in the Coulomb
gauge (Skript S. 146) gives that each component of the vector potential satisfies the Laplace equation,
∆A = 0.
Using the expression for the Laplacian in cylindrical coordinates,
1 ∂2f ∂2f
 
2 1 ∂ ∂f
∇ f≡ r + 2 + ,
r ∂r ∂r r ∂ϕ2 ∂z 2
and the fact that A(r) = (0, 0, Az (r)), one finds that

1 ∂

∂Az (r)

∂Az (r) A1,z = C1 ln r + C2 ,

r =0 ⇒ r = C1 ⇒ A2,z = C3 ln r + C4 ,
r ∂r ∂r ∂r 
A3,z = C5 ln r + C6 .

As the potential must be finite for r → 0 and r → +∞, we have to set C1 = C5 = 0.


To obtain constant C3 , we can use the boundary conditions for the H-field. These are, at r = a,
I I 1 I 1 ∂A2,z I
H2,ϕ −H1,ϕ = ⇒ H2,ϕ = ⇒ B2,ϕ = ⇒ − = .
2πa 2πa µ0 (1 + χm ) 2πa µ0 (1 + χm ) ∂r r=a 2πa

From which we get C3 = − µ0 (1+χ



m )I
.
Alternatively, to fix C3 one can use the results for the magnetic field
obtained in a) and apply ∇ × A = B.
We thus have three remaining unknown constants:

A1,z = C2 ,

A2,z = − µ0 (1+χ

m )I
ln r + C4 ,

A3,z = C6 .

Using the matching conditions for the vector potential on the two boundaries as we did in the first variant, one
finds: 
I
−µ0 (1 + χm ) 2π ln(a) + C4 for r < a,

I
Az (r) = −µ0 (1 + χm ) 2π ln(r) + C4 for a < r < b,
I

−µ0 (1 + χm ) 2π ln(b) + C4 for r > b.

Once again, one can set for example C4 = 0.

3
Aufgabe 31: Magnetischer Dipol in magnetischem Medium (6 Pkte)

Ein magnetischer Dipol m ist im Zentrum einer Kugel mit Radius R eingebettet, die aus einem homogenen li-
nearen magnetischen Medium mit Permeabilität µ besteht. Es befindet sich ein gebundener Dipol mb = χm m am
Mittelpunkt der Kugel und ein gebundener Oberflächenstrom Kb = M×r̂ bei r = R (d.h. auf der Kugeloberfläche).
a) (3 Pkte) Zeigen Sie, dass das Magnetfeld im Inneren der Kugel für 0 < r ≤ R gegeben ist durch
 
µ 1 2(µ0 − µ)m
B(r) = (3(m · r̂) − m) + .
4π r3 (2µ0 + µ)R3
SOLUTION:
In view of the following expression
Jb = ∇ × M = ∇ × (χm H) = χm Jf , (1)

There is a bound dipole at the center: mb = χm m. So the net dipole moment at the center is
mcenter = m + mb = (1 + χm )m = µµ0 m. This produces a field given by a magnetic dipole
µ 1
Bcenter dipole = (3(m · r̂)r̂ − m) .
4π r3
This accounts for the first term in the field. The remainder must be due to the bound surface current Kb at
r = R (since there can be no volume bound current, according to Eq. 1). Let us make a guess (based either on
the answer provided or on an analogous electrical problem) that the field due to the surface bound current is
(for interior points) of the form Bsurf ace current = Am (i.e. a constant, proportional to m). In that case the
magnetization will be
χm χm 1 χm
M = χm H = B= (3(m · r̂)r̂ − m) + Am.
µ 4π r3 µ
This will produce bound currents Jb = ∇ × M = 0, as it should, for 0 < r < R (no need to calculate this
curl–the second term is constant, and the first is essentially the field of a dipole, which we know is curl-less,
except at r = 0), and
 
χm χm A 1 A
Kb = M(R) × r̂ = (−m × r̂) + (m × r̂) = χm m − + sin θϕ̂
4πR3 µ 4πR3 µ
But this is exactly the surface current produced by a spinning sphere
 
A 1
K = σv = σωR sin θϕ̂, with σRω ↔ χm m − .
µ 4πR3
So the field it produces for points inside is (see exercise 25, problem sets 8),
 
2 2 A 1
Bsurface current = µ0 σRω = µ0 χm m −
3 3 µ 4πR3
   
2µ0 2 µ0 χm
Everything is consistent, therefore, provided A = 23 µ0 χm A − 1
4πR3 , or A 1 − 3µ m = − 3 4πR3 . But
χ
     µ 
µ 2(µ0 −µ)
χm = µµ0 − 1, so A 1 − 23 + 23 µµ0 = − 23 (µ−µ 0)
4πR3 , or A 1 + µ
2µ0
= 2 (µ4πR
0 )−µ
3 ; A = 4π R3 (2µ +µ) , and hence
0
 
µ 1 2(µ0 − µ)m
B= 3
(3(m · r̂) − m) + 3 .
4π r R (2µ0 + µ)

b) (3 Pkte) Welches Magnetfeld liegt außerhalb der Kugel vor?


SOLUTION:
The exterior field is that of the central dipole plus that of the surface current, which, according to exercise 25
of problem set 8, is also a perfect dipole field, of dipole moment

2πR3 µ 2(µ0 − µ)m


 
4 4 3 µ(µ0 − µ)m
msurface current = πR3 (σRω) = πR3 Bsurface current = 3
= .
3 3 2µ0 µ0 4π R (2µ0 + µ) µ0 (2µ0 + µ)
So the total dipole moment is
µ µ (µ0 − µ) 3µm
mtot = m+ m = ,
µ0 µ0 (2µ0 + µ) (2µ0 + µ)
and hence the field for r > R is
 
µ0 3µ 1
B= (3(m · r̂)r̂ − m) .
4π 2µ0 + µ r3

4
Aufgabe 32: Magnetisches Skalarpotential einer homogen magnetisierten Kugel (5 Pkte)

Falls überall Jf = 0 gilt, verschwindet die Rotation des H-Feldes, und wir können das H-Feld als z
Gradient eines skalaren Potentials φM ausdrücken gemäß

H = −∇φM .
M
Es folgt ∇2 φM = ∇ · M, also genügt φM der Poisson-Gleichung. Finden Sie das Feld im Inneren
einer homogen magnetisierten Kugel durch einen Separationsansatz.

Hinweis: Es gilt ∇ · M = 0 überall außer an der Oberfläche (r = R), also erfüllt φP


M die
∞ l Bl

Laplace-Gleichung in den Bereichen r < R und  r > R. Verwenden Sie φ M (r, θ) = l=0 Al r + r l+1
Pl (cos θ) und
out in out in
leiten Sie aus H⊥ − H⊥ = − M⊥ − M⊥ die richtige Randbedingung für φM her.
SOLUTION:
In terms of the H, Ampère’s law reads,
∇ × H = Jf .
Since Jf = 0 everywhere, we have ∇ × H = −∇ · M = 0, where ∇ · M = 0 everywhere except at the surface r = R.
According to the hint φM satisfies Laplace’s equation in inside and outside of the sphere and one finds
(P
Al rl Pl (cos θ) r < R
φM (r, θ) = Pl Bl .
l r l+1 Pl (cos θ) r > R

From the boundary conditions one can find,


( (
(i) φout in
M (r, θ)|r=R = φM (r, θ)|r=R (i) φout in
M (R, θ) = φM (R, θ)
⇒ ∂φout ∂φin .
M (r,θ) M (r,θ)
out in out in

(ii) H⊥ − H⊥ = − M⊥ − M⊥ in
(ii) − ∂r |r=R + ∂r |r=R = M⊥ = Mẑ · r̂ = M cos θ
R r2 R r2
The continuity of φM follows from the gradient theorem: φM (r2 ) − φM (r1 ) = r1
∇φM · dl = − r1
H · dl; if the
Rr
two points are infinitesimally separated, the last integral r12 H · dl → 0.
(
(i) Al Rl = RBl+1l
⇒ Bl = R2l+1 Al
P Bl P l−1
(ii) l (l + 1) Rl+2 Pl (cos θ) + l lAl R Pl (cos θ) = M cos θ

By combining these boundary conditions we find,


X M
(2l + 1)Rl−1 Al Pl (cos θ) = M cos θ, so Al = 0(l 6= 1), and 3A1 = M ⇒ A1 = .
3
l
M M M 1
Therefore φin
M (r, θ) = 3 r cos θ = 3 z,and hence Hin = −∇φin
M (r, θ) = − 3 ẑ = − 3 M, so
 
1 2
B = µ0 (H + M) = µ0 − M + M = µ0 M.
3 3

Zusatzaufgabe: Superpositionsprinzip in der Magnetostatik (+4 Pkte)

Die Abbildung zeigt den Querschnitt eines unendlich langen Zylinders mit dem Radius 3a mit einem unendlich
langen zylindrischen Loch mit dem Radius a, das so verschoben ist, dass sich sein Mittelpunkt in einem Abstand a
vom Mittelpunkt des großen Zylinders befindet. Der massive Teil des Zylinders wird von einem Strom I durchflossen,
der gleichmäßig über den Querschnitt verteilt und aus der Papierebene heraus gerichtet ist.

5
y
a) (+2 Pkte) Berechnen Sie das H-Feld an allen Punkten in der xz-Ebene, die auf einer
der Achsen der beiden Zylinder liegen.
SOLUTION:
according to the principle of superposition the field can be regarded as the difference
of two fields H2 and H1 , where H2 is the field produced by a solid (without the hole) x
cylinder of radius 3a and H1 is that produced by a cylinder of radius a at the position
of the hole. The current in each of these two cylinders is uniformly distributed over
the cross section. The currents I1 and I2 in the small and large cylinders have current
densities −j and +j respectively. Then as I = I2 − I1 = 9πa2 j − πa2 j = 8πa2 j, we a 2a
I
have j = 8πa2 and
I 9
I1 = πa2 j = , I2 = 9πa2 j = I.
8 8
The plane P p is the xz plane, p
i.e., the plane y = 0. Ampère’s law gives H1 and H2 as follows
(noting r = x2 + y 2 , r1 = (x − a)2 + y 2 , being the distances of the field point from the cylinder and hole
respectively:
(
Iy Ix
H2x = − 16πa 2, H2y = 16πa 2, r ≤ 3a
9Iy 9Ix
H2x = − 16π(x2 +y2 ) , H2y = 16π(x2 +y2 ) , r > 3a

( I(x−a)
Iy
H1x = − 16πa 2, H1y = 16πa2 , r1 ≤ a
Iy I(x−a)
H1x = − 16π((x−a) 2 +y 2 ) , H1y = 16π((x−a)2 +y 2 ) , r1 > a

On the plane P , H2x = H1x = 0. Hence Hx = 0, Hy = H2y − H1y . We therefore have the following:
(1) Inside the hole (0 < x < 2a),

Ix I(x − a) Ia I
Hy = − = = .
16πa2 16πa2 16πa2 16πa
(2) Inside the solid part (2a ≤ x ≤ 3a or −3a ≤ x ≤ 0),

Ix I(x − a) I(x2 − ax − a2 )
Hy = − = .
16πa2 16π ((x − a)2 + y 2 ) 16πa2 (x − a)

(3) Outside the cylinder (|x| > 3a),

9Ix I(x − a) (8x − 9a)I


Hy = − = .
16π(x2 + y 2 ) 16π ((x − a)2 + y 2 ) 16πx(x − a)

b) (+2 Pkte) Bestimmen Sie das H-Feld an allen Punkten des Lochs.
SOLUTION:
The magnetic field at all points inside the hole (r1 ≤ a) is

Iy Iy Ix I(x − a) I
Hx = − + = 0, Hy = − = .
16πa2 16πa2 16πa2 16πa2 16πa
This field is uniform inside the hole and is along the positive y-direction.

You might also like