Mathematics Formulary

By ir. J.C.A. Wevers


c 1999, 2007 J.C.A. Wevers Version: December 30, 2007

Dear reader,
This document contains 66 pages with mathematical equations intended for physicists and engineers. It
is intended to be a short reference for anyone who often needs to look up mathematical equations.
This document can also be obtained from the author, Johan Wevers ([email protected]).
It can also be found on the WWW on http://www.xs4all.nl/~johanw/index.html.
This document is Copyright by J.C.A. Wevers. All rights reserved. Permission to use, copy and distribute
this unmodified document by any means and for any purpose except profit purposes is hereby granted.
Reproducing this document by any means, included, but not limited to, printing, copying existing prints,
publishing by electronic or other means, implies full agreement to the above non-profit-use clause, unless
upon explicit prior written permission of the author.
The C code for the rootfinding via Newtons method and the FFT in chapter 8 are from “Numerical Recipes
in C ”, 2nd Edition, ISBN 0-521-43108-5.
The Mathematics Formulary is made with teTEX and LATEX version 2.09.
If you prefer the notation in which vectors are typefaced in boldface, uncomment the redefinition of the
\vec command and recompile the file.
If you find any errors or have any comments, please let me know. I am always open for suggestions and
possible corrections to the mathematics formulary.
Johan Wevers
Chapter 1

Basics

1.1 Goniometric functions

For the goniometric ratios for a point p on the unit circle holds:
yp
cos(φ) = xp , sin(φ) = yp , tan(φ) =
xp

sin2 (x) + cos2 (x) = 1 and cos−2 (x) = 1 + tan2 (x).

cos(a ± b) = cos(a) cos(b) ∓ sin(a) sin(b) , sin(a ± b) = sin(a) cos(b) ± cos(a) sin(b)
tan(a) ± tan(b)
tan(a ± b) =
1 ∓ tan(a) tan(b)
The sum formulas are:

sin(p) + sin(q) = 2 sin( 12 (p + q)) cos( 12 (p − q))

sin(p) − sin(q) = 2 cos( 12 (p + q)) sin( 12 (p − q))
cos(p) + cos(q) = 2 cos( 21 (p + q)) cos( 12 (p − q))
cos(p) − cos(q) = −2 sin( 12 (p + q)) sin( 12 (p − q))

From these equations can be derived that

2 cos2 (x) = 1 + cos(2x) , 2 sin2 (x) = 1 − cos(2x)

sin(π − x) = sin(x) , cos(π − x) = − cos(x)
sin( 21 π − x) = cos(x) , cos( 12 π − x) = sin(x)

Conclusions from equalities:

sin(x) = sin(a) ⇒ x = a ± 2kπ or x = (π − a) ± 2kπ, k ∈ IN

cos(x) = cos(a) ⇒ x = a ± 2kπ or x = −a ± 2kπ
π
tan(x) = tan(a) ⇒ x = a ± kπ and x 6= ± kπ
2
The following relations exist between the inverse goniometric functions:
   
x 1 p
arctan(x) = arcsin √ = arccos √ , sin(arccos(x)) = 1 − x2
x2 + 1 x2 + 1

1
2 Mathematics Formulary by ir. J.C.A. Wevers

1.2 Hyperbolic functions

The hyperbolic functions are defined by:

ex − e−x ex + e−x sinh(x)

sinh(x) = , cosh(x) = , tanh(x) =
2 2 cosh(x)

From this follows that cosh2 (x) − sinh2 (x) = 1. Further holds:

p p
arsinh(x) = ln |x + x2 + 1| , arcosh(x) = arsinh( x2 − 1)

1.3 Calculus
The derivative of a function is defined as:

df f (x + h) − f (x)
= lim
dx h→0 h

Derivatives obey the following algebraic rules:

 
x ydx − xdy
d(x ± y) = dx ± dy , d(xy) = xdy + ydx , d =
y y2

For the derivative of the inverse function f inv (y), defined by f inv (f (x)) = x, holds at point P = (x, f (x)):

df inv (y)
   
df (x)
· =1
dy P dx P

Chain rule: if f = f (g(x)), then holds

df df dg
=
dx dg dx

Further, for the derivatives of products of functions holds:

n  
X n
(f · g)(n) = f (n−k) · g (k)
k
k=0

For the primitive function F (x) holds: F 0 (x) = f (x). An overview of derivatives and primitives is:
Chapter 1: Basics 3

dy/dx = f 0 (x)
R
y = f (x) f (x)dx

axn anxn−1 a(n + 1)−1 xn+1

1/x −x−2 ln |x|
a 0 ax
ax ax ln(a) ax / ln(a)
ex ex ex
a
log(x) (x ln(a))−1 (x ln(x) − x)/ ln(a)
ln(x) 1/x x ln(x) − x
sin(x) cos(x) − cos(x)
cos(x) − sin(x) sin(x)
tan(x) cos−2 (x) − ln | cos(x)|
sin−1 (x) − sin−2 (x) cos(x) ln | tan( 12 x)|
sinh(x) cosh(x) cosh(x)
cosh(x) sinh(x)
√ sinh(x)√
arcsin(x) 1/ √1 − x2 x arcsin(x) + √1 − x2
arccos(x) −1/ 1 − x2 x arccos(x) − 1 − x2
arctan(x) (1 + x2 )−1 x arctan(x) − 12 ln(1 + x2 )
√
(a + x2 )−1/2 −x(a + x2 )−3/2 ln |x + a + x2 |
1
(a2 − x2 )−1 2x(a2 + x2 )−2 ln |(a + x)/(a − x)|
2a

(1 + (y 0 )2 )3/2
The curvature ρ of a curve is given by: ρ =
|y 00 |
f (x) f 0 (x)
The theorem of De ’l Hôpital: if f (a) = 0 and g(a) = 0, then is lim = lim 0
x→a g(x) x→a g (x)

1.4 Limits

sin(x) ex − 1 tan(x)  n x
lim =1 , lim =1 , lim =1 , lim (1 + k)1/k = e , lim 1+ = en
x→0 x x→0 x x→0 x k→0 x→∞ x

lnp (x) ln(x + a) xp

lim xa ln(x) = 0 , lim = 0 , lim =a , lim = 0 als |a| > 1.
x↓0 x→∞ xa x→0 x x→∞ ax

  arcsin(x) √
lim a1/x − 1 = ln(a) , lim =1 , lim x
x=1
x→0 x→0 x x→∞

1.5 Complex numbers and quaternions

1.5.1 Complex numbers
The complex
√ number z = a + bi with a and b ∈ IR. a is the real part, b the imaginary part of z.
|z| = a2 + b2 . By definition holds: i2 = −1. Every complex number can be written as z = |z| exp(iϕ),
4 Mathematics Formulary by ir. J.C.A. Wevers

with tan(ϕ) = b/a. The complex conjugate of z is defined as z = z ∗ := a − bi. Further holds:
(a + bi)(c + di) = (ac − bd) + i(ad + bc)
(a + bi) + (c + di) = a + c + i(b + d)
a + bi (ac + bd) + i(bc − ad)
=
c + di c2 + d2
Goniometric functions can be written as complex exponents:
1 ix
sin(x) = (e − e−ix )
2i
1 ix
cos(x) = (e + e−ix )
2
From this follows that cos(ix) = cosh(x) and sin(ix) = i sinh(x). Further follows from this that
e±ix = cos(x) ± i sin(x), so eiz 6= 0∀z. Also the theorem of De Moivre follows from this:
(cos(ϕ) + i sin(ϕ))n = cos(nϕ) + i sin(nϕ).
Products and quotients of complex numbers can be written as:
z 1 · z2 = |z1 | · |z2 |(cos(ϕ1 + ϕ2 ) + i sin(ϕ1 + ϕ2 ))
z1 |z1 |
= (cos(ϕ1 − ϕ2 ) + i sin(ϕ1 − ϕ2 ))
z2 |z2 |
The following can be derived:
|z1 + z2 | ≤ |z1 | + |z2 | , |z1 − z2 | ≥ | |z1 | − |z2 | |
And from z = r exp(iθ) follows: ln(z) = ln(r) + iθ, ln(z) = ln(z) ± 2nπi.

1.5.2 Quaternions
Quaternions are defined as: z = a + bi + cj + dk, with a, b, c, d ∈ IR and i2 = j 2 = k 2 = −1. The products
of i, j, k with each other are given by ij = −ji = k, jk = −kj = i and ki = −ik = j.

1.6 Geometry
1.6.1 Triangles
The sine rule is:
a b c
= =
sin(α) sin(β) sin(γ)
Here, α is the angle opposite to a, β is opposite to b and γ opposite to c. The cosine rule is: a2 =
b2 + c2 − 2bc cos(α). For each triangle holds: α + β + γ = 180◦ .
Further holds:
tan( 12 (α + β)) a+b
=
tan( 12 (α − β)) a−b
p
The surface of a triangle is given by 21 ab sin(γ) = 12 aha = s(s − a)(s − b)(s − c) with ha the perpendicular
on a and s = 21 (a + b + c).
Chapter 1: Basics 5

1.6.2 Curves
Cycloid: if a circle with radius a rolls along a straight line, the trajectory of a point on this circle has the
following parameter equation:
x = a(t + sin(t)) , y = a(1 + cos(t))
Epicycloid: if a small circle with radius a rolls along a big circle with radius R, the trajectory of a point
on the small circle has the following parameter equation:
   
R+a R+a
x = a sin t + (R + a) sin(t) , y = a cos t + (R + a) cos(t)
a a
Hypocycloid: if a small circle with radius a rolls inside a big circle with radius R, the trajectory of a
point on the small circle has the following parameter equation:
   
R−a R−a
x = a sin t + (R − a) sin(t) , y = −a cos t + (R − a) cos(t)
a a
A hypocycloid with a = R is called a cardioid. It has the following parameterequation in polar coordinates:
r = 2a[1 − cos(ϕ)].

1.7 Vectors
X
The inner product is defined by: ~a · ~b = ai bi = |~a | · |~b | cos(ϕ)
i

where ϕ is the angle between ~a and ~b. The external product is in IR3 defined by:
   
ay bz − az by  ~ex ~ey ~ez 
~

~a × b =  az bx − ax bz  =  ax ay az 
ax by − ay bx  bx by bz 

Further holds: |~a × ~b | = |~a | · |~b | sin(ϕ), and ~a × (~b × ~c ) = (~a · ~c )~b − (~a · ~b )~c.

1.8 Series
1.8.1 Expansion
The Binomium of Newton is:
n  
n
X n
(a + b) = an−k bk
k
k=0
 
n n!
where := .
k k!(n − k)!
n n
rk and r rk one finds:
P P
By subtracting the series
k=0 k=0
n
X 1 − rn+1
rk =
1−r
k=0
6 Mathematics Formulary by ir. J.C.A. Wevers

∞
X 1
and for |r| < 1 this gives the geometric series: rk = .
1−r
k=0
N
X
The arithmetic series is given by: (a + nV ) = a(N + 1) + 21 N (N + 1)V .
n=0

The expansion of a function around the point a is given by the Taylor series:

(x − a)2 00 (x − a)n (n)

f (x) = f (a) + (x − a)f 0 (a) + f (a) + · · · + f (a) + R
2 n!
where the remainder is given by:
hn (n+1)
Rn (h) = (1 − θ)n f (θh)
n!
and is subject to:
mhn+1 M hn+1
≤ Rn (h) ≤
(n + 1)! (n + 1)!
From this one can deduce that
∞  
X α
(1 − x)α = xn
n=0
n
One can derive that:
∞ ∞ ∞
X 1 π2 X 1 π4 X 1 π6
2
= , 4
= , 6
=
n=1
n 6 n=1
n 90 n=1
n 945
n ∞ ∞
X X (−1)n+1 π2 X (−1)n+1
k 2 = 16 n(n + 1)(2n + 1) , 2
= , = ln(2)
n=1
n 12 n=1
n
k=1
∞ ∞ ∞ ∞
X 1 1
X 1 π2 X 1 π4 X (−1)n+1 π3
2
= 2 , 2
= , 4
= , 3
=
n=1
4n − 1 n=1
(2n − 1) 8 n=1
(2n − 1) 96 n=1
(2n − 1) 32

1.8.2 Convergence and divergence of series

P P
If |un | converges, un also converges.
n n
P
If lim un 6= 0 then un is divergent.
n→∞ n

An alternating series of which the absolute values of the terms drop monotonously to 0 is convergent
(Leibniz).
R∞ P
If p f (x)dx < ∞, then fn is convergent.
n
P P
If un > 0 ∀n then is un convergent if ln(un + 1) is convergent.
n n
 
n
P 1 p
n
 cn+1 
If un = cn x the radius of convergence ρ of un is given by: = lim |cn | = lim 
 .
n ρ n→∞ n→∞ cn 
Chapter 1: Basics 7

∞
X 1
The series is convergent if p > 1 and divergent if p ≤ 1.
n=1
np
un P P
If: lim = p, than the following is true: if p > 0 than un and vn are both divergent or both
n→∞ vn n n
P P
convergent, if p = 0 holds: if vn is convergent, than un is also convergent.
n n
 
p
n
 un+1  P
If L is defined by: L = lim |nn |, or by: L = lim 
 , then is un divergent if L > 1 and
n→∞ n→∞ un 
n
convergent if L < 1.

1.8.3 Convergence and divergence of functions

f (x) is continuous in x = a only if the upper - and lower limit are equal: lim f (x) = lim f (x). This is
x↑a x↓a
written as: f (a− ) = f (a+ ).

If f (x) is continuous in a and: lim f 0 (x) = lim f 0 (x), than f (x) is differentiable in x = a.
x↑a x↓a

We define: kf kW := sup(|f (x)| |x ∈ W ), and lim fn (x) = f (x). Than holds: {fn } is uniform convergent
x→∞
if lim kfn − f k = 0, or: ∀(ε > 0)∃(N )∀(n ≥ N )kfn − f k < ε.
n→∞
P P
Weierstrass’ test: if kun kW is convergent, than un is uniform convergent.

∞
X Zb
We define S(x) = un (x) and F (y) = f (x, y)dx := F . Than it can be proved that:
n=N a

Theorem For Demands on W Than holds on W

rows fn continuous, f is continuous
{fn } uniform convergent
C series S(x) uniform convergent, S is continuous
un continuous
integral f is continuous F is continuous
rows fn can be integrated, fRn can be integrated,
R
{fn } uniform convergent f (x)dx = lim fn dx
n→∞
R PR
I series S(x) is uniform convergent, S can be integrated, Sdx = un dx
un can be integrated
R RR
integral f is continuous F dy = f (x, y)dxdy

rows {fn } ∈C−1 ; {fn0 } unif.conv → φ f 0 = φ(x)

un ∈C−1 ;
P 0
S 0 (x) = u0n (x)
P P
D series un conv; un u.c.
R
integral ∂f /∂y continuous Fy = fy (x, y)dx
8 Mathematics Formulary by ir. J.C.A. Wevers

1.9 Products and quotients

For a, b, c, d ∈ IR holds:
The distributive property: (a + b)(c + d) = ac + ad + bc + bd
The associative property: a(bc) = b(ac) = c(ab) and a(b + c) = ab + ac
The commutative property: a + b = b + a, ab = ba.
Further holds:
n
a2n − b2n a2n+1 − b2n+1 X
= a2n−1 ± a2n−2 b + a2n−3 b2 ± · · · ± b2n−1 , = a2n−k b2k
a±b a+b
k=0

a3 ± b3
(a ± b)(a2 ± ab + b2 ) = a3 ± b3 , (a + b)(a − b) = a2 + b2 , = a2 ∓ ba + b2
a+b

1.10 Logarithms
a
Definition: log(x) = b ⇔ ab = x. For logarithms with base e one writes ln(x).
Rules: log(xn ) = n log(x), log(a) + log(b) = log(ab), log(a) − log(b) = log(a/b).

1.11 Polynomials
Equations of the type
n
X
ak xk = 0
k=0

have n roots which may be equal to each other. Each polynomial p(z) of order n ≥ 1 has at least one root
in C . If all ak ∈ IR holds: when x = p with p ∈ C a root, than p∗ is also a root. Polynomials up to and
including order 4 have a general analytical solution, for polynomials with order ≥ 5 there does not exist a
general analytical solution.
For a, b, c ∈ IR and a 6= 0 holds: the 2nd order equation ax2 + bx + c = 0 has the general solution:
√
−b ± b2 − 4ac
x=
2a
For a, b, c, d ∈ IR and a 6= 0 holds: the 3rd order equation ax3 + bx2 + cx + d = 0 has the general analytical
solution:
3ac − b2 b
x1 = K− −
9a2 K 3a √ 
3ac − b2 3ac − b2

K b 3
x2 = x∗3 = − + − + i K +
2 18a2 K 3a 2 9a2 K

√ √ !1/3
9abc − 27da2 − 2b3 3 4ac3 − c2 b2 − 18abcd + 27a2 d2 + 4db3
with K = +
54a3 18a2
Chapter 1: Basics 9

1.12 Primes
A prime is a number ∈ IN that can only be divided by itself and 1. There are an infinite number of primes.
Q
Proof: suppose that the collection of primes P would be finite, than construct the number q = 1 + p,
p∈P
than holds q = 1(p) and so Q cannot be written as a product of primes from P . This is a contradiction.
If π(x) is the number of primes ≤ x, than holds:

π(x) π(x)
lim = 1 and lim =1
x→∞ x/ ln(x) x→∞ Rx dt
ln(t)
2

For each N ≥ 2 there is a prime between N and 2N .

The numbers Fk := 2k + 1 with k ∈ IN are called Fermat numbers. Many Fermat numbers are prime.
The numbers Mk := 2k −1 are called Mersenne numbers. They occur when one searches for perfect numbers,
which are numbers n ∈ IN which are the sum of their different dividers, for example 6 = 1+2+3. There are
23 Mersenne numbers for k < 12000 which are prime: for k ∈ {2, 3, 5, 7, 13, 17, 19, 31, 61, 89, 107, 127, 521,
607, 1279, 2203, 2281, 3217, 4253, 4423, 9689, 9941, 11213}.
To check if a given number n is prime one can use a sieve method. The first known sieve method was
developed by Eratosthenes. A faster method for large numbers are the 4 Fermat tests, who don’t prove
that a number is prime but give a large probability.
1. Take the first 4 primes: b = {2, 3, 5, 7},
2. Take w(b) = bn−1 mod n, for each b,
3. If w = 1 for each b, then n is probably prime. For each other value of w, n is certainly not prime.
Chapter 3

Calculus

3.1 Integrals
3.1.1 Arithmetic rules
The primitive function F (x) of f (x) obeys the rule F 0 (x) = f (x). With F (x) the primitive of f (x) holds
for the definite integral
Zb
f (x)dx = F (b) − F (a)
a

If u = f (x) holds:
Zb f
Z(b)
g(f (x))df (x) = g(u)du
a f (a)

Partial integration: with F and G the primitives of f and g holds:

Z Z
df (x)
f (x) · g(x)dx = f (x)G(x) − G(x) dx
dx

A derivative can be brought under the intergral sign (see section 1.8.3 for the required conditions):
 
x=h(y)
Z x=h(y)
Z
d  ∂f (x, y) dg(y) dh(y)
f (x, y)dx = dx − f (g(y), y) + f (h(y), y)

dy ∂y dy dy

x=g(y) x=g(y)

3.1.2 Arc lengts, surfaces and volumes

The arc length ` of a curve y(x) is given by:
s  2
Z
dy(x)
`= 1+ dx
dx

The arc length ` of a parameter curve F (~x(t)) is:

Z Z
` = F ds = F (~x(t))|~x˙ (t)|dt

14
Chapter 3: Calculus 15

with
˙
~t = d~x = ~x(t) , |~t | = 1
ds |~x˙ (t)|
Z Z Z
~ ˙
~
(~v , t)ds = (~v , t(t))dt = (v1 dx + v2 dy + v3 dz)

The surface A of a solid of revolution is:

s  2
Z
dy(x)
A = 2π y 1+ dx
dx

The volume V of a solid of revolution is:

Z
V =π f 2 (x)dx

3.1.3 Separation of quotients

Every rational function P (x)/Q(x) where P and Q are polynomials can be written as a linear combination
of functions of the type (x − a)k with k ∈ ZZ, and of functions of the type
px + q
((x − a)2 + b2 )n

with b > 0 and n ∈ IN . So:

n n
p(x) X Ak p(x) X Ak x + B
n
= , =
(x − a) (x − a)k 2 2
((x − b) + c )n ((x − b)2 + c2 )k
k=1 k=1

Recurrent relation: for n 6= 0 holds:

2n − 1
Z Z
dx 1 x dx
2 n+1
= 2 n
+
(x + 1) 2n (x + 1) 2n (x + 1)n
2

3.1.4 Special functions

Elliptic functions
Elliptic functions can be written as a power series as follows:
∞
(2n − 1)!!
q X
1 − k 2 sin2 (x) = 1 − k 2n sin2n (x)
n=1
(2n)!!(2n − 1)

∞
1 X (2n − 1)!! 2n 2n
q =1+ k sin (x)
1 − k 2 sin2 (x) n=1
(2n)!!

with n!! = n(n − 2)!!.

16 Mathematics Formulary by ir. J.C.A. Wevers

The Gamma function

The gamma function Γ(y) is defined by:

Z∞
Γ(y) = e−x xy−1 dx
0

One can derive that Γ(y + 1) = yΓ(y) = y!. This is a way to define faculties for non-integers. Further one
can derive that
√ Z∞
π
Γ(n + 2 ) = n (2n − 1)!! and Γ (y) = e−x xy−1 lnn (x)dx
1 (n)
2
0

The Beta function

The betafunction β(p, q) is defined by:

Z1
β(p, q) = xp−1 (1 − x)q−1 dx
0

with p and q > 0. The beta and gamma functions are related by the following equation:

Γ(p)Γ(q)
β(p, q) =
Γ(p + q)

The Delta function

The delta function δ(x) is an infinitely thin peak function with surface 1. It can be defined by:

0 for |x| > ε

(
δ(x) = lim P (ε, x) with P (ε, x) = 1
ε→0 when |x| < ε
2ε
Some properties are:
Z∞ Z∞
δ(x)dx = 1 , F (x)δ(x)dx = F (0)
−∞ −∞

3.1.5 Goniometric integrals

When solving goniometric integrals it can be useful to change variables. The following holds if one defines
tan( 21 x) := t:
2dt 1 − t2 2t
dx = 2
, cos(x) = 2
, sin(x) =
1+t 1+t 1 + t2
Chapter 3: Calculus 17

R √
Each integral of the type R(x, ax2 + bx + c)dx can be converted into one of the types that were treated
in section 3.1.3. After this conversion one can substitute in the integrals of the type:
Z p dϕ p
R(x, x2 + 1)dx : x = tan(ϕ) , dx = of x2 + 1 = t + x
cos(ϕ)
Z p p
R(x, 1 − x2 )dx : x = sin(ϕ) , dx = cos(ϕ)dϕ of 1 − x2 = 1 − tx
Z p 1 sin(ϕ) p
R(x, x2 − 1)dx : x= , dx = dϕ of x2 − 1 = x − t
cos(ϕ) cos2 (ϕ)
These definite integrals are easily solved:

Zπ/2 
n m (n − 1)!!(m − 1)!! π/2 when m and n are both even
cos (x) sin (x)dx = ·
(m + n)!! 1 in all other cases
0

Some important integrals are:

Z∞ Z∞ Z∞
xdx π2 x2 dx π2 x3 dx π4
= , = , =
eax + 1 12a2 (ex + 1)2 3 ex + 1 15
0 −∞ 0

3.2 Functions with more variables

3.2.1 Derivatives
The partial derivative with respect to x of a function f (x, y) is defined by:
 
∂f f (x0 + h, y0 ) − f (x0 , y0 )
= lim
∂x x0 h→0 h

The directional derivative in the direction of α is defined by:

∂f f (x0 + r cos(α), y0 + r sin(α)) − f (x0 , y0 ) ~ (sin α, cos α)) = ∇f · ~v
= lim = (∇f,
∂α r↓0 r |~v |
When one changes to coordinates f (x(u, v), y(u, v)) holds:
∂f ∂f ∂x ∂f ∂y
= +
∂u ∂x ∂u ∂y ∂u
If x(t) and y(t) depend only on one parameter t holds:
∂f ∂f dx ∂f dy
= +
∂t ∂x dt ∂y dt
The total differential df of a function of 3 variables is given by:
∂f ∂f ∂f
df = dx + dy + dz
∂x ∂y ∂z
18 Mathematics Formulary by ir. J.C.A. Wevers

So
df ∂f ∂f dy ∂f dz
= + +
dx ∂x ∂y dx ∂z dx
The tangent in point ~x0 at the surface f (x, y) = 0 is given by the equation fx (~x0 )(x−x0 )+fy (~x0 )(y−y0 ) = 0.
The tangent plane in ~x0 is given by: fx (~x0 )(x − x0 ) + fy (~x0 )(y − y0 ) = z − f (~x0 ).

3.2.2 Taylor series

A function of two variables can be expanded as follows in a Taylor series:
n
∂p ∂p
 
X 1
f (x0 + h, y0 + k) = h p + k p f (x0 , y0 ) + R(n)
p=0
p! ∂x ∂y

with R(n) the residual error and

p  
∂p ∂p p m p−m ∂ p f (a, b)
  X
h p +k p f (a, b) = h k
∂x ∂y m=0
m ∂xm ∂y p−m

3.2.3 Extrema
When f is continuous on a compact boundary V there exists a global maximum and a global minumum
for f on this boundary. A boundary is called compact if it is limited and closed.
Possible extrema of f (x, y) on a boundary V ∈ IR2 are:

1. Points on V where f (x, y) is not differentiable,

~ = ~0,
2. Points where ∇f

3. If the boundary V is given by ϕ(x, y) = 0, than all points where ∇f ~ (x, y) + λ∇ϕ(x,
~ y) = 0 are
possible for extrema. This is the multiplicator method of Lagrange, λ is called a multiplicator.

The same as in IR2 holds in IR3 when the area to be searched is constrained by a compact V , and V is
defined by ϕ1 (x, y, z) = 0 and ϕ2 (x, y, z) = 0 for extrema of f (x, y, z) for points (1) and (2). Point (3) is
rewritten as follows: possible extrema are points where ∇f ~ (x, y, z) + λ1 ∇ϕ
~ 1 (x, y, z) + λ2 ∇ϕ
~ 2 (x, y, z) = 0.

3.2.4 The ∇-operator

In cartesian coordinates (x, y, z) holds:

~ ∂ ∂ ∂
∇ = ~ex + ~ey + ~ez
∂x ∂y ∂z
∂f ∂f ∂f
gradf = ~ex + ~ey + ~ez
∂x ∂y ∂z
∂ax ∂ay ∂az
div ~a = + +
∂x ∂y ∂z
Chapter 3: Calculus 19

     
∂az ∂ay ∂ax ∂az ∂ay ∂ax
curl ~a = − ~ex + − ~ey + − ~ez
∂y ∂z ∂z ∂x ∂x ∂y
∂2f ∂2f ∂2f
∇2 f = 2
+ 2 + 2
∂x ∂y ∂z

In cylindrical coordinates (r, ϕ, z) holds:

~ ∂ 1 ∂ ∂
∇ = ~er + ~eϕ + ~ez
∂r r ∂ϕ ∂z
∂f 1 ∂f ∂f
gradf = ~er + ~eϕ + ~ez
∂r r ∂ϕ ∂z
∂ar ar 1 ∂aϕ ∂az
div ~a = + + +
∂r r r ∂ϕ ∂z
     
1 ∂az ∂aϕ ∂ar ∂az ∂aϕ aϕ 1 ∂ar
curl ~a = − ~er + − ~eϕ + + − ~ez
r ∂ϕ ∂z ∂z ∂r ∂r r r ∂ϕ
∂2f 1 ∂f 1 ∂2f ∂2f
∇2 f = 2
+ + 2 2
+ 2
∂r r ∂r r ∂ϕ ∂z

In spherical coordinates (r, θ, ϕ) holds:

~ ∂ 1 ∂ 1 ∂
∇ = ~er + ~eθ + ~eϕ
∂r r ∂θ r sin θ ∂ϕ
∂f 1 ∂f 1 ∂f
gradf = ~er + ~eθ + ~eϕ
∂r r ∂θ r sin θ ∂ϕ
∂ar 2ar 1 ∂aθ aθ 1 ∂aϕ
div ~a = + + + +
∂r r r ∂θ r tan θ r sin θ ∂ϕ
   
1 ∂aϕ aθ 1 ∂aθ 1 ∂ar ∂aϕ aϕ
curl ~a = + − ~er + − − ~eθ +
r ∂θ r tan θ r sin θ ∂ϕ r sin θ ∂ϕ ∂r r
 
∂aθ aθ 1 ∂ar
+ − ~eϕ
∂r r r ∂θ
∂2f 2 ∂f 1 ∂2f 1 ∂f 1 ∂2f
∇2 f = + + + +
∂r2 r ∂r r2 ∂θ2 r2 tan θ ∂θ r2 sin2 θ ∂ϕ2
General orthonormal curvilinear coordinates (u, v, w) can be derived from cartesian coordinates by the
transformation ~x = ~x(u, v, w). The unit vectors are given by:

1 ∂~x 1 ∂~x 1 ∂~x

~eu = , ~ev = , ~ew =
h1 ∂u h2 ∂v h3 ∂w
where the terms hi give normalization to length 1. The differential operators are than given by:

1 ∂f 1 ∂f 1 ∂f
gradf = ~eu + ~ev + ~ew
h1 ∂u h ∂v h3 ∂w
 2 
1 ∂ ∂ ∂
div ~a = (h2 h3 au ) + (h3 h1 av ) + (h1 h2 aw )
h1 h2 h3 ∂u ∂v ∂w
20 Mathematics Formulary by ir. J.C.A. Wevers

   
1 ∂(h3 aw ) ∂(h2 av ) 1 ∂(h1 au ) ∂(h3 aw )
curl ~a = − ~eu + − ~ev +
h2 h3 ∂v ∂w h3 h1 ∂w ∂u
 
1 ∂(h2 av ) ∂(h1 au )
− ~ew
h1 h2 ∂u ∂v
      
1 ∂ h2 h3 ∂f ∂ h3 h1 ∂f ∂ h1 h2 ∂f
∇2 f = + +
h1 h2 h3 ∂u h1 ∂u ∂v h2 ∂v ∂w h3 ∂w
Some properties of the ∇-operator are:

div(φ~v ) = φdiv~v + gradφ · ~v curl(φ~v ) = φcurl~v + (gradφ) × ~v curl gradφ = ~0

div(~u × ~v ) = ~v · (curl~u) − ~u · (curl~v ) curl curl~v = grad div~v − ∇2~v div curl~v = 0
div gradφ = ∇2 φ ∇2~v ≡ (∇2 v1 , ∇2 v2 , ∇2 v3 )

Here, ~v is an arbitrary vectorfield and φ an arbitrary scalar field.

3.2.5 Integral theorems

Some important integral theorems are:
ZZ ZZZ
Gauss: (~v · ~n)d2 A = (div~v )d3 V
I ZZ
Stokes for a scalar field: (φ · ~et )ds = (~n × gradφ)d2 A
I ZZ
Stokes for a vector field: (~v · ~et )ds = (curl~v · ~n)d2 A
ZZ
this gives: (curl~v · ~n)d2 A = 0
ZZ ZZZ
2
Ostrogradsky: (~n × ~v )d A = (curl~v )d3 A
ZZ ZZZ
(φ~n )d2 A = (gradφ)d3 V

d2 A is bounded by the Jordan curve s(t).

RR
Here the orientable surface

3.2.6 Multiple integrals

Let A be a closed curve given by f (x, y) = 0, than the surface A inside the curve in IR2 is given by
ZZ ZZ
A= d2 A = dxdy

Let the surface A be defined by the function z = f (x, y). The volume V bounded by A and the xy plane
is than given by: ZZ
V = f (x, y)dxdy
Chapter 3: Calculus 21

The volume inside a closed surface defined by z = f (x, y) is given by:

ZZZ ZZ ZZZ
3
V = d V = f (x, y)dxdy = dxdydz

3.2.7 Coordinate transformations

The expressions d2 A and d3 V transform as follows when one changes coordinates to ~u = (u, v, w) through
the transformation x(u, v, w):
ZZZ ZZZ  
 ∂~x 
V = f (x, y, z)dxdydz = f (~x(~u))   dudvdw
∂~u

In IR2 holds:  
∂~x  xu xv 
=
∂~u yu yv 
Let the surface A be defined by z = F (x, y) = X(u, v). Than the volume bounded by the xy plane and F
is given by:
ZZ ZZ   ZZ
2
 ∂X ∂X 
q
f (~x)d A = f (~x(~u)) 
 × dudv = f (x, y, F (x, y)) 1 + ∂x F 2 + ∂y F 2 dxdy
∂u ∂v 
S G G

3.3 Orthogonality of functions

The inner product of two functions f (x) and g(x) on the interval [a, b] is given by:

Zb
(f, g) = f (x)g(x)dx
a

or, when using a weight function p(x), by:

Zb
(f, g) = p(x)f (x)g(x)dx
a

The norm kf k follows from: kf k2 = (f, f ). A set functions fi is orthonormal if (fi , fj ) = δij .
Each function f (x) can be written as a sum of orthogonal functions:
∞
X
f (x) = ci gi (x)
i=0

c2i ≤ kf k2 . Let the set gi be orthogonal, than it follows:

P
and

f, gi
ci =
(gi , gi )
22 Mathematics Formulary by ir. J.C.A. Wevers

3.4 Fourier series

Each function can be written as a sum of independent base functions. When one chooses the orthogonal
basis (cos(nx), sin(nx)) we have a Fourier series.
A periodical function f (x) with period 2L can be written as:
∞ h
X  nπx   nπx i
f (x) = a0 + an cos + bn sin
n=1
L L

Due to the orthogonality follows for the coefficients:

ZL ZL   ZL  
1 1 nπt 1 nπt
a0 = f (t)dt , an = f (t) cos dt , bn = f (t) sin dt
2L L L L L
−L −L −L

A Fourier series can also be written as a sum of complex exponents:

∞
X
f (x) = cn einx
n=−∞

with
Zπ
1
cn = f (x)e−inx dx
2π
−π

The Fourier transform of a function f (x) gives the transformed function fˆ(ω):
Z∞
1
fˆ(ω) = √ f (x)e−iωx dx
2π
−∞

The inverse transformation is given by:

Z∞
1 1
f (x+ ) + f (x− ) = √ fˆ(ω)eiωx dω

2 2π
−∞

where f (x+ ) and f (x− ) are defined by the lower - and upper limit:

f (a− ) = lim f (x) , f (a+ ) = lim f (x)

x↑a x↓a

For continuous functions is 1

2 [f (x+ ) + f (x− )] = f (x).
Chapter 4

Differential equations

4.1 Linear differential equations

4.1.1 First order linear DE
The general solution of a linear differential equation is given by yA = yH + yP , where yH is the solution of
the homogeneous equation and yP is a particular solution.
A first order differential equation is given by: y 0 (x) + a(x)y(x) = b(x). Its homogeneous equation is
y 0 (x) + a(x)y(x) = 0.
The solution of the homogeneous equation is given by
Z 
yH = k exp a(x)dx

Suppose that a(x) = a =constant.

Substitution of exp(λx) in the homogeneous equation leads to the characteristic equation λ + a = 0
⇒ λ = −a.
Suppose b(x) = α exp(µx). Than one can distinguish two cases:

1. λ 6= µ: a particular solution is: yP = exp(µx)

2. λ = µ: a particular solution is: yP = x exp(µx)

When a DE is solved by variation of parameters one writes: yP (x) = yH (x)f (x), and than one solves f (x)
from this.

4.1.2 Second order linear DE

A differential equation of the second order with constant coefficients is given by: y 00 (x) + ay 0 (x) + by(x) =
c(x). If c(x) = c =constant there exists a particular solution yP = c/b.
Substitution of y = exp(λx) leads to the characteristic equation λ2 + aλ + b = 0.
There are now 2 possibilities:

1. λ1 6= λ2 : than yH = α exp(λ1 x) + β exp(λ2 x).

2. λ1 = λ2 = λ: than yH = (α + βx) exp(λx).

23
24 Mathematics Formulary by ir. J.C.A. Wevers

If c(x) = p(x) exp(µx) where p(x) is a polynomial there are 3 possibilities:

1. λ1 , λ2 6= µ: yP = q(x) exp(µx).

2. λ1 = µ, λ2 6= µ: yP = xq(x) exp(µx).

3. λ1 = λ2 = µ: yP = x2 q(x) exp(µx).

where q(x) is a polynomial of the same order as p(x).

Rx
When: y 00 (x) + ω 2 y(x) = ωf (x) and y(0) = y 0 (0) = 0 follows: y(x) = f (x) sin(ω(x − t))dt.
0

4.1.3 The Wronskian

We start with the LDE y 00 (x) + p(x)y 0 (x) + q(x)y(x) = 0 and the two initial conditions y(x0 ) = K0 and
y 0 (x0 ) = K1 . When p(x) and q(x) are continuous on the open interval I there exists a unique solution
y(x) on this interval.
The general solution can than be written as y(x) = c1 y1 (x) + c2 y2 (x) and y1 and y2 are linear independent.
These are also all solutions of the LDE.
The Wronskian is defined by:
 
 y y2 
W (y1 , y2 ) =  10 = y1 y20 − y2 y10
y1 y20 

y1 and y2 are linear independent if and only if on the interval I when ∃x0 ∈ I so that holds:
W (y1 (x0 ), y2 (x0 )) = 0.

4.1.4 Power series substitution

an xn is substituted in the LDE with constant coefficients y 00 (x) + py 0 (x) + qy(x) = 0
P
When a series y =
this leads to: X
n(n − 1)an xn−2 + pnan xn−1 + qan xn = 0

n

Setting coefficients for equal powers of x equal gives:

(n + 2)(n + 1)an+2 + p(n + 1)an+1 + qan = 0

This gives a general relation between the coefficients. Special cases are n = 0, 1, 2.

4.2 Some special cases

4.2.1 Frobenius’ method
Given the LDE
d2 y(x) b(x) dy(x) c(x)
+ + 2 y(x) = 0
dx2 x dx x
Chapter 4: Differential equations 25

with b(x) and c(x) analytical at x = 0. This LDE has at least one solution of the form
∞
X
ri
yi (x) = x an xn with i = 1, 2
n=0

with r real or complex and chosen so that a0 6= 0. When one expands b(x) and c(x) as b(x) = b0 + b1 x +
b2 x2 + ... and c(x) = c0 + c1 x + c2 x2 + ..., it follows for r:

r2 + (b0 − 1)r + c0 = 0

There are now 3 possibilities:

1. r1 = r2 : than y(x) = y1 (x) ln |x| + y2 (x).
2. r1 − r2 ∈ IN : than y(x) = ky1 (x) ln |x| + y2 (x).
3. r1 − r2 6= ZZ: than y(x) = y1 (x) + y2 (x).

4.2.2 Euler
Given the LDE
d2 y(x) dy(x)
x2 + ax + by(x) = 0
dx2 dx
Substitution of y(x) = xr gives an equation for r: r2 + (a − 1)r + b = 0. From this one gets two solutions
r1 and r2 . There are now 2 possibilities:
1. r1 6= r2 : than y(x) = C1 xr1 + C2 xr2 .
2. r1 = r2 = r: than y(x) = (C1 ln(x) + C2 )xr .

4.2.3 Legendre’s DE
Given the LDE
d2 y(x) dy(x)
(1 − x2 ) − 2x + n(n − 1)y(x) = 0
dx2 dx
The solutions of this equation are given by y(x) = aPn (x) + by2 (x) where the Legendre polynomials P (x)
are defined by:
dn (1 − x2 )n
 
Pn (x) = n
dx 2n n!
For these holds: kPn k2 = 2/(2n + 1).

4.2.4 The associated Legendre equation

This equation follows from the θ-dependent part of the wave equation ∇2 Ψ = 0 by substitution of
ξ = cos(θ). Than follows:
 
2 d 2 dP (ξ)
(1 − ξ ) (1 − ξ ) + [C(1 − ξ 2 ) − m2 ]P (ξ) = 0
dξ dξ
26 Mathematics Formulary by ir. J.C.A. Wevers

Regular solutions exists only if C = l(l + 1). They are of the form:

|m| d|m| P 0 (ξ) (1 − ξ 2 )|m|/2 d|m|+l 2

Pl (ξ) = (1 − ξ 2 )m/2 = (ξ − 1)l
dξ |m| 2l l! dξ |m|+l

|m|
For |m| > l is Pl (ξ) = 0. Some properties of Pl0 (ξ) zijn:

Z1 ∞
2 X 1
Pl0 (ξ)Pl00 (ξ)dξ = δll0 , Pl0 (ξ)tl = p
2l + 1 1 − 2ξt + t2
−1 l=0

This polynomial can be written as:

Zπ
1 p
Pl0 (ξ) = (ξ + ξ 2 − 1 cos(θ))l dθ
π
0

4.2.5 Solutions for Bessel’s equation

Given the LDE
d2 y(x) dy(x)
x2 2
+x + (x2 − ν 2 )y(x) = 0
dx dx
also called Bessel’s equation, and the Bessel functions of the first kind
∞
X (−1)m x2m
Jν (x) = xν
m=0
22m+ν m!Γ(ν + m + 1)

for ν := n ∈ IN this becomes:

∞
X (−1)m x2m
Jn (x) = xn
m=0
22m+n m!(n + m)!

When ν 6= ZZ the solution is given by y(x) = aJν (x) + bJ−ν (x). But because for n ∈ ZZ holds:
J−n (x) = (−1)n Jn (x), this does not apply to integers. The general solution of Bessel’s equation is given
by y(x) = aJν (x) + bYν (x), where Yν are the Bessel functions of the second kind:

Jν (x) cos(νπ) − J−ν (x)

Yν (x) = and Yn (x) = lim Yν (x)
sin(νπ) ν→n

The equation x2 y 00 (x) + xy 0 (x) − (x2 + ν 2 )y(x) = 0 has the modified Bessel functions of the first kind
Iν (x) = i−ν Jν (ix) as solution, and also solutions Kν = π[I−ν (x) − Iν (x)]/[2 sin(νπ)].

Sometimes it can be convenient to write the solutions of Bessel’s equation in terms of the Hankel functions

Hn(1) (x) = Jn (x) + iYn (x) , Hn(2) (x) = Jn (x) − iYn (x)
Chapter 4: Differential equations 27

4.2.6 Properties of Bessel functions

Bessel functions are orthogonal with respect to the weight function p(x) = x.

J−n (x) = (−1)n Jn (x). The Neumann functions Nm (x) are definied as:
∞
1 1 X
Nm (x) = Jm (x) ln(x) + m αn x2n
2π x n=0

The following holds: lim Jm (x) = xm , lim Nm (x) = x−m for m 6= 0, lim N0 (x) = ln(x).
x→0 x→0 x→0

r r
e±ikr eiωt 2 2
lim H(r) = √ , lim Jn (x) = cos(x − xn ) , lim J−n (x) = sin(x − xn )
r→∞ r x→∞ πx x→∞ πx

with xn = 12 π(n + 12 ).

2n dJn (x)
Jn+1 (x) + Jn−1 (x) = Jn (x) , Jn+1 (x) − Jn−1 (x) = −2
x dx
The following integral relations hold:

Z2π Zπ
1 1
Jm (x) = exp[i(x sin(θ) − mθ)]dθ = cos(x sin(θ) − mθ)dθ
2π π
0 0

4.2.7 Laguerre’s equation

Given the LDE
d2 y(x) dy(x)
x + (1 − x) + ny(x) = 0
dx2 dx
Solutions of this equation are the Laguerre polynomials Ln (x):
∞
ex dn (−1)m n m
 
n −x

X
Ln (x) = x e = x
n! dxn m=0
m! m

4.2.8 The associated Laguerre equation

Given the LDE
d2 y(x) n + 12 (m + 1)
   
m+1 dy(x)
+ −1 + y(x) = 0
dx2 x dx x
Solutions of this equation are the associated Laguerre polynomials Lm
n (x):

(−1)m n! −x −m dn−m
Lm e−x xn


n (x) = e x n−m
(n − m)! dx
28 Mathematics Formulary by ir. J.C.A. Wevers

4.2.9 Hermite
The differential equations of Hermite are:

d2 Hn (x) dHn (x) d2 Hen (x) dHen (x)

2
− 2x + 2nHn (x) = 0 and −x + nHen (x) = 0
dx dx dx2 dx
Solutions of these equations are the Hermite polynomials, given by:

dn (exp(− 21 x2 )) √
 
1 2
Hn (x) = (−1)n exp x n
= 2n/2 Hen (x 2)
2 dx

dn (exp(−x2 )) −n/2
√
Hen (x) = (−1)n (exp x2 = 2 Hn (x/ 2)
dxn

4.2.10 Chebyshev
The LDE
d2 Un (x) dUn (x)
(1 − x2 ) 2
− 3x + n(n + 2)Un (x) = 0
dx dx
has solutions of the form
sin[(n + 1) arccos(x)]
Un (x) = √
1 − x2
The LDE
d2 Tn (x) dTn (x)
(1 − x2 ) −x + n2 Tn (x) = 0
dx2 dx
has solutions Tn (x) = cos(n arccos(x)).

4.2.11 Weber
The LDE Wn00 (x) + (n + 1
2 − 14 x2 )Wn (x) = 0 has solutions: Wn (x) = Hen (x) exp(− 41 x2 ).

4.3 Non-linear differential equations

Some non-linear differential equations and a solution are:
p
y 0 = apy 2 + b2 y= b sinh(a(x − x0 ))
y 0 = apy 2 − b2 y= b cosh(a(x − x0 ))
y 0 = a b2 − y 2 y= b cos(a(x − x0 ))
y 0 = a(y 2 + b2 ) y= b tan(a(x − x0 ))
y 0 = a(y 2 − b2 ) y= b coth(a(x − x0 ))
y 0 = a(b
2
− y 2 ) y= b tanh(a(x − x0 ))
b−y b
y 0 = ay y=
b 1 + Cb exp(−ax)
Chapter 4: Differential equations 29

4.4 Sturm-Liouville equations

Sturm-Liouville equations are second order LDE’s of the form:
 
d dy(x)
− p(x) + q(x)y(x) = λm(x)y(x)
dx dx
The boundary conditions are chosen so that the operator
 
d d
L=− p(x) + q(x)
dx dx
is Hermitean. The normalization function m(x) must satisfy
Zb
m(x)yi (x)yj (x)dx = δij
a

When y1 (x) and y2 (x) are two linear independent solutions one can write the Wronskian in this form:
 
= C
 y 1 y2 
W (y1 , y2 ) =  0
y1 y20  p(x)


p
where C is constant. By changing to another dependent variable u(x), given by: u(x) = y(x) p(x), the
LDE transforms into the normal form:
2
d2 u(x) 1 p0 (x) 1 p00 (x) q(x) − λm(x)

+ I(x)u(x) = 0 with I(x) = − −
dx2 4 p(x) 2 p(x) p(x)
If I(x) > 0, than y 00 /y < 0 and the solution has an oscillatory behaviour, if I(x) < 0, than y 00 /y > 0 and
the solution has an exponential behaviour.

4.5 Linear partial differential equations

4.5.1 General
The normal derivative is defined by:
∂u ~ ~n)
= (∇u,
∂n
A frequently used solution method for PDE’s is separation of variables: one assumes that the solution
can be written as u(x, t) = X(x)T (t). When this is substituted two ordinary DE’s for X(x) and T (t) are
obtained.

4.5.2 Special cases

The wave equation
The wave equation in 1 dimension is given by
∂2u ∂2u
2
= c2 2
∂t ∂x
30 Mathematics Formulary by ir. J.C.A. Wevers

When the initial conditions u(x, 0) = ϕ(x) and ∂u(x, 0)/∂t = Ψ(x) apply, the general solution is given by:
x+ct
Z
1 1
u(x, t) = [ϕ(x + ct) + ϕ(x − ct)] + Ψ(ξ)dξ
2 2c
x−ct

The diffusion equation

The diffusion equation is:
∂u
= D∇2 u
∂t
Its solutions can be written in terms of the propagators P (x, x0 , t). These have the property that
P (x, x0 , 0) = δ(x − x0 ). In 1 dimension it reads:

−(x − x0 )2
 
0 1
P (x, x , t) = √ exp
2 πDt 4Dt

In 3 dimensions it reads:
−(~x − ~x 0 )2
 
0 1
P (x, x , t) = exp
8(πDt)3/2 4Dt
With initial condition u(x, 0) = f (x) the solution is:
Z
u(x, t) = f (x0 )P (x, x0 , t)dx0
G

The solution of the equation

∂u ∂2u
− D 2 = g(x, t)
∂t ∂x
is given by Z Z
u(x, t) = dt0 dx0 g(x0 , t0 )P (x, x0 , t − t0 )

The equation of Helmholtz

The equation of Helmholtz is obtained by substitution of u(~x, t) = v(~x) exp(iωt) in the wave equation.
This gives for v:
∇2 v(~x, ω) + k 2 v(~x, ω) = 0
This gives as solutions for v:

1. In cartesian coordinates: substitution of v = A exp(i~k · ~x ) gives:

Z Z
~
v(~x ) = · · · A(k)eik·~x dk

with the integrals over ~k 2 = k 2 .

Chapter 4: Differential equations 31

2. In polar coordinates:
∞
X
v(r, ϕ) = (Am Jm (kr) + Bm Nm (kr))eimϕ
m=0

3. In spherical coordinates:
∞ X
l
X Y (θ, ϕ)
v(r, θ, ϕ) = [Alm Jl+ 21 (kr) + Blm J−l− 12 (kr)] √
r
l=0 m=−l

4.5.3 Potential theory and Green’s theorem

Subject of the potential theory are the Poisson equation ∇2 u = −f (~x ) where f is a given function, and the
Laplace equation ∇2 u = 0. The solutions of these can often be interpreted as a potential. The solutions of
Laplace’s equation are called harmonic functions.
When a vector field ~v is given by ~v = gradϕ holds:
Zb
(~v , ~t )ds = ϕ(~b ) − ϕ(~a )
a

In this case there exist functions ϕ and w

~ so that ~v = gradϕ + curlw.
~
The field lines of the field ~v (~x ) follow from:

~x˙ (t) = λ~v (~x )

The first theorem of Green is:

ZZZ ZZ
∂v
[u∇2 v + (∇u, ∇v)]d3 V = u d2 A
∂n
G S

The second theorem of Green is:

ZZZ ZZ  
∂v ∂u
[u∇2 v − v∇2 u]d3 V = u −v d2 A
∂n ∂n
G S

A harmonic function which is 0 on the boundary of an area is also 0 within that area. A harmonic function
with a normal derivative of 0 on the boundary of an area is constant within that area.
The Dirichlet problem is:

∇2 u(~x ) = −f (~x ) , ~x ∈ R , u(~x ) = g(~x ) for all ~x ∈ S.

It has a unique solution.

The Neumann problem is:
∂u(~x )
∇2 u(~x ) = −f (~x ) , ~x ∈ R , = h(~x ) for all ~x ∈ S.
∂n
32 Mathematics Formulary by ir. J.C.A. Wevers

The solution is unique except for a constant. The solution exists if:
ZZZ ZZ
− f (~x )d V = h(~x )d2 A
3

R S

A fundamental solution of the Laplace equation satisfies:

∇2 u(~x ) = −δ(~x )

This has in 2 dimensions in polar coordinates the following solution:

ln(r)
u(r) =
2π
This has in 3 dimensions in spherical coordinates the following solution:
1
u(r) =
4πr

The equation ∇2 v = −δ(~x − ξ~ ) has the solution

1
v(~x ) =
4π|~x − ξ~ |

After substituting this in Green’s 2nd theorem and applying the sieve property of the δ function one can
derive Green’s 3rd theorem:
∇2 u 3
ZZZ ZZ   
~ 1 1 1 ∂u ∂ 1
u(ξ ) = − d V + −u d2 A
4π r 4π r ∂n ∂n r
R S

The Green function G(~x, ξ~ ) is defined by: ∇2 G = −δ(~x − ξ~ ), and on boundary S holds G(~x, ξ~ ) = 0. Than
G can be written as:
1
G(~x, ξ~ ) = + g(~x, ξ~ )
4π|~x − ξ~ |
Than g(~x, ξ~ ) is a solution of Dirichlet’s problem. The solution of Poisson’s equation ∇2 u = −f (~x ) when
on the boundary S holds: u(~x ) = g(~x ), is:

∂G(~x, ξ~ ) 2
ZZZ ZZ
u(ξ~ ) = G(~x, ξ~ )f (~x )d3 V − g(~x ) d A
∂n
R S