European Journal of Physics

LETTERS AND COMMENTS Related content

- Various paths to Faraday's law
A little help for a better understanding and Dragan V Redži

application of Faraday's law - Rethinking Faraday's law for teaching

motional electromotive force
Kristina Zuza, Jenaro Guisasola, Marisa
Michelini et al.
To cite this article: E Benedetto et al 2012 Eur. J. Phys. 33 L15
- Note
Dragan V Redži

View the article online for updates and enhancements.

This content was downloaded from IP address 168.176.5.118 on 23/09/2019 at 08:22

IOP PUBLISHING EUROPEAN JOURNAL OF PHYSICS
Eur. J. Phys. 33 (2012) L15–L20 doi:10.1088/0143-0807/33/3/L15

LETTERS AND COMMENTS

A little help for a better understanding

and application of Faraday’s law
E Benedetto 1 , M Capriolo 1 , A Feoli 2 and D Tucci 2
1 Department of Mathematics and Computer Science, University of Salerno,
via Ponte Don Melillo - 84084 Fisciano (SA), Italy
2 Department of Engineering, University of Sannio, Piazza Roma 21, 82100 Benevento, Italy

E-mail: [email protected]

Received 3 January 2012, in final form 6 March 2012

Published 29 March 2012
Online at stacks.iop.org/EJP/33/L15

Abstract
In this letter, we examine Faraday’s law of induction, analysing the
electromotive force generated by a Lorentz force and the one generated by
an electric field due to a changing magnetic field. We obtain the result in a
didactically simple and appealing way. The final formula is derived considering
explicitly the dependence of the magnetic field on the space coordinates, which
is often neglected in standard textbooks.
(Some figures may appear in colour only in the online journal)

1. Introduction

Let us remember the following relation that describes well-known electromagnetic induction:

d(t ) d −
→ −
EMF = − =− B (t, →
r (t )) · 
n dS, (1)
dt dt S(t )
where EMF is the electromotive force, S(t ) is a surface that has a circuit, which we will
denote with l, as its boundary, and the integral is the magnetic flux. First of all, we stress
that Faraday’s law describes two different phenomena: the EMF generated by a Lorentz
force on a moving wire and the one generated by an electric field due to a time-varying
magnetic field. This is well explained in many standard textbooks (see, e.g. [1, 2]). While
there are no doubts about the way to calculate the derivative of the flux (1) in the two
extremal cases, many interesting new books (see, e.g. [3, 4]) and papers [5–16] focussed
on the study of the right approach to the more general case, the right way to define the
EMF as well as the relativistic point of view. The debate is still alive after so many years.
In this letter, we want to suggest an appealing way to demonstrate Faraday’s law in the
general case of surface and field both simultaneously changing, considering explicitly the
dependence of the magnetic field on the space coordinates. This dependence is often neglected

0143-0807/12/030015+06$33.00 
c 2012 IOP Publishing Ltd Printed in the UK & the USA L15
L16 Letters and Comments

in most standard textbooks that usually put in evidence only the time-dependent part of the
field.
So we have to consider a closed circuit embedded in a non-constant magnetic field, with
a wire moving with velocity v . This motion of the circuit and the changing field induce along
the wire a displacement of charges with velocity u = u l (where dl is an infinitesimal part of
charge = v + u. A point of the circuit,
the circuit) so that the total velocity of the charge is V
which has a position r (t ) at the initial time, will be found at a position r (t + t ) after a
time interval t. Now our aim is to perform the time differentiation of the flux integral (1)
in order to obtain explicitly a didactically useful formula to understand better the physics of
electromagnetic induction. We also want to advise that other different forms of Faraday’s law,
often considered equivalent, may create confusion because they give the right results only if
some very special hypotheses are satisfied.
This letter is organized as follows. First, we write the demonstration of Faraday’s law in
a general form in section 2, and then, in section 3, conclusions are drawn.

2. Faraday’s law

We consider the following integral function:


−
→ −
(t ) = B (t, →
r (t )) · 
n dS (2)
S(t )

and its derivative

 −
→ −
→  − → →
d(t ) n dS − S(t ) B (t, −
S(t+t ) B (t + t, r (t + t )) ·  r (t )) · 
ndS
= lim . (3)
dt t→0 t
−
→− →  −
→
We start from the second Maxwell equation ∇ · B = 0 and its integral form σ = σ B · ndS =
0, where σ is a closed surface. From figure 1, we have
−
→
n3 d = −d−
 → r ∧d l , (4)
 −
→
r
−
→
r (t + t ) − −
→
r (t ) = −
→
r = r =−
d−
→ →
v t, (5)
0

 −
→
α  −
→
α
−
→
r (t + τ ) − −
→
r (t ) = −
→
α = d−
→
r = α =−
d−
→ →
v τ, (6)
0 0

where − →v = d− →r /dt = d−
→α /dτ is constant during the small interval of time t
and 0 < τ < t. So we can compute the area of the surface  that, for small
t, is
   
1 −
→ −
→ −
→
A = |−
→
r ∧d l |+ |−
→
r ∧d l |  |−
→
v t ∧ d l |, (7)
2 l(t ) l(t+t ) l(t )

and considering the closed surface σ formed by S(t ), S(t + t ) and the lateral surface  at
the fixed instant t + t, we can write
 
−
→ −
→ −
→ −
→
σ = n1 dS +
B (t + t, r +  r ) ·  B (t + t, −
→
r ) ·
n2 dS
S(t+t ) S(t )

−
→
+ B (t + t, −
→
r + −→ n3 d = 0.
α ) · (8)

Letters and Comments L17

Figure 1. The evolution of the circuit, embedded in a magnetic field, from the time t to the time
t + t. (The magnetic field is not explicitly drawn because the demonstration of Faraday’s law in

the text is general and does not depend on the particular form of the vector field B.)

Then, using equation (4) and the convention for the normal n of a surface enclosed in a circuit,
which must have the current flow counterclockwise in the loop  n = n1 = −n2 , we obtain
 
−
→ −
→
B (t + t, −
→
r + − →
r ) ·
ndS − B (t + t, −
→
r ) ·
ndS
S(t+t ) S(t )

−
→ −
→
− B (t + t, −
→r + −→α ) · (d−→α ∧ d l ) = 0. (9)


We have, using Taylor’s theorem,

→
−
−
→ −
→ −
→ − → ∂ B 
B (t + t, r ) = B (t, r ) +  t, (10)
∂t 
t=0


−
→ −
→ −
→ −
→
B (t + t, −
→
r + −
→ r ) + (−
α ) = B (t + t, −
→ →
v · ∇)B τ. (11)
τ =0
L18 Letters and Comments

Substituting these equations into (9), we can write

 
−
→ −
→ −
→ −
→ −
B (t + t, r +  r ) ·  ndS − B (t, →r ) ·ndS
S(t+t ) S(t )
 −
→  
∂ B  −
→ −
→
=  t · ndS + B (t + t, − →r ) · dτ (− →
v ∧d l )
S(t ) ∂t  
 t=0

−
→ −
→  −
→
+ (−→v · ∇)B τ dτ · (− → v ∧d l )
 τ =0
 →
−   t
∂ B  −
→ −
→
=  t · ndS + dτ B (t + t, − →r ) · (−
→
v ∧d l )
S(t ) ∂t  l 0
t=0
  t 
−
→ − → −
→
+ τ d(τ ) (− →v · ∇)B · (−
→
v ∧d l )
τ =0
→
l 0
 − 
∂ B  −
→ → −
→
= t  ·ndS + t B · − v ∧d l
S(t ) ∂t  l
t=0
 
t 2 −
→ − → −
→
+ (−→v · ∇)B ·− →v ∧d l . (12)
2 l τ =0
Then, we can compute the derivative of the integral function using the definition of
equation (3):
  
d 1 −
→ −
→ −
→ −
→ − →
= lim B (t + t, r +  r ) ·  ndS − B (t, r ) · 
ndS
dt t→0 t S(t+t ) S(t )
 
∂B −
→ −
→
= ·
ndS − (−
→
v ∧ B)·d l . (13)
S(t ) ∂t l
Finally, we obtain
 −
→

d(t ) ∂B −
→ −
→
EMF = − =− ndS + (−
· →
v ∧ B)·d l . (14)
dt S(t ) ∂t l

3. Conclusions

In our opinion, equation (14) is the only correct way to write down Faraday’s law in the general
case, and comparing it with the correct definition of the EMF in terms of the Lorentz force
divided by the charge q
 
−
→ − → −
→ −
→ −
→ → − → −
→
EMF = ( E + V charge ∧ B ) · d l = ( E + − v ∧ B)·d l , (15)
l l
the Maxwell equation
−
→
−
→ − → ∂B
∇ × E =− (16)
∂t
is recovered.
But some other versions of Faraday’s law often proposed in several contexts do not always
give the right result. For example, we examine
  − → 
d(t ) ∂B −
→ −
→ −
→
EMF = − =− − ∇ ∧ ( v ∧ B ) · ndS, (17)
dt S(t ) ∂t
Letters and Comments L19

obtained from (14) using Stokes’ theorem and

  − → 
d(t ) ∂B −
→ −
→
EMF = − =− + (−
→
v · ∇ ) B ·
ndS (18)
dt S(t ) ∂t

derived from (17) using the well-known vectorial identity

−
→ −
→ → −
− → → −
→ −
→ → − → − → → −
− → →
∇ ∧ (−→v ∧ B ) = ( B · ∇ )− v − (−→v · ∇)B +− v (∇ · B )− B (∇ ·−
v) (19)

and supposing a constant velocity field. When the field B  is constant everywhere and the
velocity has a constant value just on a moving line (a part of the circuit) and is zero elsewhere
(i.e. −
→v is not a continuous field with continuous space derivatives), equations (17) and (18)
erroneously predict a vanishing EMF.
We remember that Stokes’ theorem can be applied to equation (14) only if the vector
−
→
field (−→
v ∧ B ) is a continuously differentiable vector function. Even if we generalize our
framework taking into account also discontinuous functions, the result does not change. If
we consider the famous example of a rectangular circuit on the xy-plane with only one side
moving along the x-axis with constant velocity and a magnetic field directed along the z-axis,
we can approximate the velocity field with a constant in an infinitesimal part (the thickness
of the moving wire) of the domain. In this way, the velocity field can be described by a
rectangular function v = vorect[(x − xo (t ))/ ] = u(x − xo (t ) + /2) − u(x − xo (t ) − /2),
where xo (t ) = xo + vot and u(x) is the step function. Computing the divergence of velocity in
the last term of equation (19), we obtain a difference of two delta functions that, after being
integrated, give the difference between two values of the magnetic field at a distance from
each other. Hence, if B is constant, the surface integral in (17) is vanishing.
So, from the didactic point of view, it is better to show Faraday’s law in a way that obtains
as its final result formula (14), without introducing other almost equivalent equations that
can generate confusion when applied by the students in solving exercises. In conclusion, we
have derived the right formula (14) in the general case of a field dependent on time and space
coordinates and we have also clarified that some other ways to express Faraday’s law cannot
be considered fully equivalent and could lead to incorrect results.

Acknowledgments

The authors thank Dr Amorosia of Messina University for his help in achieving, with computer
graphics, a readable and good quality figure.

References

[1] Feynman R P, Leighton R B and Sand M 1964 Feynman Lectures on Physics vol II (Reading, MA: Addison-
Wesley)
[2] Jackson J D 1975 Classical Electrodynamics 2nd edn (New York: Wiley)
[3] Pramanik A 2004 Electro-Magnetism: Theory and Applications (New Delhi: Prentice-Hall)
[4] Chow T L 2006 Introduction to Electromagnetic Theory: A Modern Perspective (Mississauga: Jones and Bartlett)
[5] Scanlon P J, Henriksen R N and Allen J R 1969 Approaches to electromagnetic induction Am. J.
Phys. 37 698–708
[6] Flanders H 1973 Differentiation under the integral sign Am. Math. Mon. 80 615–27
[7] Gamo H 1979 A general formulation of Faraday’s law of induction Proc. IEEE 67 676–7
[8] Rosen N and Schieber D 1982 Some remarks on Faraday’s law Am. J. Phys. 50 974–5
[9] Gelman H 1991 Faraday’s law for relativistic and deformed motion of a circuit Eur. J. Phys. 12 230–3
[10] Giuliani G 2000 On electromagnetic induction arXiv:physics/0008006v1
[11] Munley F 2004 Challenges to Faraday’s flux rule Am. J. Phys. 72 1478–83
L20 Letters and Comments

[12] Redzic D V 2007 Faraday’s law via the magnetic vector potential Eur. J. Phys. 28 N7–10
[13] Redzic D V 2008 Reply to comment on ‘Faraday’s law via the magnetic vector potential’ Eur. J. Phys. 29 L5
[14] Lopez-Ramos A, Menendez J R and Piquè C 2008 Conditions for the validity of Faraday’s law of induction and
their experimental confirmation Eur. J. Phys. 29 1069–76
[15] Cross D J 2009 On the flux rule www.brynmawr.edu/physics/DJCross/docs/misc/flux.pdf
[16] Rodrigues F G 2010 On equivalent expressions for Faraday’s law of induction arXiv:1002.2792v2