Homework 4: Benjamin Stahl

U NIVERSITY OF C ALIFORNIA - S ANTA C RUZ
D EPARTMENT OF P HYSICS
PHYS 116A
Homework 4
Benjamin Stahl
February 7, 2014
1 B OAS : C H . 11, S EC . 7, Q. 6
Z
0
yd y
(1 + y 3 )2
(1.1)
The above integral (1.1) will be expressed as a Beta function by making the substitution of
x = y 3:
x = y3
y =x
x 3 dx
2
3x 3 (1 + x)2
Z 1
x 3 dx
1
=
3 0 (1 + x)2
0
d x = 3y d y
dy =
1
3
dx
2
3x 3
(1.2)
The result found in 1.2 contains a Beta function where p + q = 2 and p 1 = 13 . Thus, p =
and q = 34
2 4

23 13 13
1 2 4
1 3 3
1 2
1
,
=
=
=
3 3 3
3
(2)
31
9 3
3
2
3
(1.3)
The result in 1.3 is of the form (p)(1 p) =
si n(p) .
Thus:
p

1 2
2
2 3
1
1
=
= p =
9 3
3
9 si n 2
27
9 3
3
(1.4)
2 B OAS : C H . 11, S EC . 7, Q. 9
n, 12
.
22n1
Prove that (n, n) =

Applying the definition of a Beta function, making the appropriate
trigonometric substitution, and simplifying yields:
n, 12
(n, n) =
2
0
22n1
1
R
2 02 [si n()]2n1 [cos()]2 2 1 d
[si n()]2n1 [cos()]2n1 d =
22n1
R
[si n()cos()]
2n1
0
0
Z
0
si n(2)
2
2n1
22n1
d =
d =
(2.1)
[si n()]2n1 d
22n1
[si n(2)]2n1 d =
[si n()]2n1 [cos()]0 d

22n1
[si n(2)]2n1 d
Z
0
[si n()]2n1 d
22n1
[si n()]2n1 d
Making the substitution = 2 to the left integral:

Z
Z
2
1
[si n()]2n1 d =
[si n()]2n1 d
2 0
0
"Z
# Z
Z
2
2
1
2n1
2n1
[si n()]
d +
[si n()]
d =
[si n()]2n1 d
2 0
0
2
(2.2)
By symmetry, the equivalence is proven:

Z
Z
2
2
1
2
[si n()]2n1 d =
[si n()]2n1 d
2
0
0
Z
Z
2
2
[si n()]2n1 d =
[si n()]2n1 d
0
(2.3)
Using this result, the duplication formula for Gamma functions will be derived:
(n, n) =
n, 21
22n1
(n) 12
(n)(n)
=
(2n)
22n1 n + 12
p
(n)
=
2n1
(2n) 2
n + 12
22n1 (n) n + 12
(2n) =
p
(2.4)
Checking this result with n = 14 :

1

22 4 1 ( 41 ) 14 + 12
( 41 ) 34
1
2
=
=
= p
p
4
2
This result is of the form (p)(1 p) =
si n(p) .
(2.5)
Thus:

( 14 ) 34
p
2
=p
p
= p p =
2
2si n 4
2 2
(2.6)
3 B OAS : C H . 11, S EC . 11, Q. 4

p
(2n)! n
lim
n 22n (n!)2
(3.1)
p
Stirlings formula (n n n e n 2n) will be utilized to evaluate the limit in 3.1:
p
p
p
p
(2n)2n e 2n 4n n
(2n)! n
22n n 2n e 2n 2n
lim
= lim
= lim 2n 2n 2n
p
n 22n [n n e n 2n]2
n 22n (n!)2
n 2 n e
2n
p
1
=
= p
(3.2)
(3.3)
4 B OAS : C H . 11, S EC . 12, Q. 1

The integrands of K and E will be expanded using power series and then integrated in order
to determine the power series approximations for the complete elliptic integrals:
K=
d
p
E=
1 k 2 si n 2 ()
p
1 k 2 si n 2 ()d
(4.1)
The integrands in 4.1 are compared with the binomial expansion:
(1 + x)p = 1 + px +
p(p 1) 2 p(p 1)(p 2) 3

x +
x +...
2!
3!
(4.2)
Utilizing the binomial expansion (4.2) with the substitution x = k 2 si n 2 () and p = 12 yields:
1
1
3
(1 + x) 2 = 1 x + x 2 . . .
2
8
1
3
1
(1 k 2 si n 2 ()) 2 = 1 + k 2 si n 2 () + k 4 si n 4 () + . . .
2
8
Z
2
3
1 2
K=
[1 + k si n 2 () + k 4 si n 4 () + . . . ]d
2
8
0
Using Wolfram Alpha: K =
(64 + 16k 2 + 9k 4 )
128
(4.3)
Utilizing the binomial expansion (4.2) with the substitution x = k 2 si n 2 () and p =
1
2
yields:
1
1
1
(1 + x) 2 = 1 + x x 2 . . .
2
8
1
1
1
(1 k 2 si n 2 ()) 2 = 1 k 2 si n 2 () k 4 si n 4 () + . . .
2
8
Z
2
1 2
1
E=
[1 k si n 2 () k 4 si n 4 () + . . . ]d
2
8
0
Using Wolfram Alpha: E =
(64 + 16k 2 + 3k 4 )
128
(4.4)
5 B OAS : C H . 11, S EC . 12, Q. 12

1
2
Z
0
p
100 t 2
dt
p
1 t2
(5.1)
The above integral (5.1) will be made to look like an elliptic integral and then solved using a
computer:
1
2
Z
0
q
p
Z 1 1 t2
Z si n() p
2
100 t 2
1 k2t 2
100
d t = 10
d t = 10
d t = 10E (, k)
p
p
p
0
0
1 t2
1 t2
1 t2
(5.2)
From this result it is clear that:

1
si n() =
2
1
= ar c si n
=
2
6
1
100
1
k=
10
k2 =
(5.3)
Calculating this integral using Wolfram Alpha yields:

q
Z 1 1 t2
2
1
100
,
= 10
d t = 5.2337
10E
p
6 10
0
1 t2
(5.4)
6 B OAS : C H . 11, S EC . 13, Q. 3

lim n x (x, n) = (x)
(6.1)
Using the definition of the beta function in terms of gamma functions as well as Stirlings
p
formula ((p + 1) p p e p 2p) the above equality (6.1) will be verified:
lim n x (x, n) = (x)
n
x
n (x)(n)
= (x)
(x + n)
n (n + 1)(x + n)
(x) lim
= (x)
n
n(x + n + 1)
p
n x1 n n e n 2n(x + n)
(x) lim
= (x)
p
n (x + n)x+n e xn 2(x + n)
p
n x+n1 n
(x) lim
= (x)
p
n (x + n)x+n1 e x (x + n)
r
n x+n1
n
(x)e x lim
= (x)
n
x +n
x +n
x + n 1xn
= (x)
(x)e x lim
n
n
x n
x 1x
(x)e x lim 1 +
1+
= (x)
n
n
n
x 1x
= (x)
(x)e x lim e x 1 +
n
n
x 1x
(x) lim 1 +
= (x)
n
n
(x) = (x)
lim
n
x
(6.2)
7
(x)
d
ln (x)
dx
(7.1)
7.1 A
Equation 7.1 will be used to find 0(x + 1):
(x + 1) =
d
d
d
1
d
ln (x + 1) =
ln[x(x)] =
ln x +
ln (x) = + (x)
dx
dx
dx
dx
x
1
0(x + 1) = 2 + 0(x)
x
(7.2)
7.2 B
1
+ 0(1)
12
1
1
1
0(2 + 1) = 2 + 0(2) = 2 2 + 0(1)
2
1
2
1
1
1
1
0(3 + 1) = 2 + 0(3) = 2 + 2 + 2 + 0(1)
3
3
2
1
n 1
X
0(n + 1) = 0(1)
2
k=1 k
0(1 + 1) =
(7.3)
7.3 C
d
ln (n)
dn
p
d
d
d
1
n n
(n + 1) =
ln (n + 1) =
ln[n e
2n] =
n ln n n + ln 2n
dn
dn
dn
2
1 2
1
= ln n + 1 1 +
= ln n +
22n
2n
1
1
1
d
ln n +
= 2
0(n + 1) =
dn
2n
n 2n
1
1
lim 0(n + 1) = lim
2 =0
n
n n
2n
(n) =
(7.4)
8
(x + 1) =
x ln t t
dt =
e f (t ) d t
(8.1)
8.1 A
The argument f (t ) of the exponential will be expanded using a Taylor series about t = x:
f (x) = x ln x x
x
f 0(x) = 1 = 0
x
x
1
f 00(x) = 2 =
x
x
2
2x
f 000(x) = 3 = 2
x
x
6x
6
(i v)
f
(x) = 4 = 3
x
x
f (t ) = x ln t t
x
f 0(t ) = 1
t
x
f 00(t ) = 2
t
2x
f 000(t ) = 3
t
6x
f (i v) (t ) = 4
t
f (t ) ' x ln x x
(8.2)
(t x)2 (t x)3 (t x)4

+
2x
3x 2
4x 3
(8.3)
8.2 B
Substituting the Taylor series for f (t ) back into the exponential and then making the substit x
tution u = p
yields:
2x
(x + 1) =
x ln xx (t x)
+ (t x)2 (t x)3
2x
3x
4x
Z
p
x ln xx
= 2xe
dt =
p
x
p
2
x ln xxu 2 + 32px2 u 3 x1 u 4 p
2xd u
(8.4)
p
x
p
2
u 2 + 32px2 u 3 x1 u 4
du
8.3 C
Z
p
x ln xx
(x + 1) = 2xe
u 2
p
2p 2 3 1 4
u xu
3 x
Z
p
x ln xx
d u ' 2xe
u 2
A(u) B (u)
1+ p +
du
x
x
(8.5)
The second exponential of the integrand is expanded according to: e z ' 1 + z + 12 z 2 + . . .
Z
p
x ln xx
2xe
A(u) B (u)
e
1+ p +
du
x
x
p
#
"
!2
p
Z
p
2 2 3 1 4 1 2 2 3 1 4
x ln xx
u 2
= 2xe
e
+ ... du
1+ p u u +
p u u
x
2 3 x
x
3 x
#
"
p
Z
p
2 2 3 1 4 1 8 6
x ln xx
u 2
e
u + ... du
= 2xe
1+ p u u +
x
2 9x
3 x
u 2
(8.6)
Therefore:
p
2 2 3
A(u) =
u
3
4
B (u) = u 4 + u 6
9
(8.7)
8.4 D
#
p
2 2 3 1 4 4 6
e
1+ p u u +
u du
x
9x
3 x
#
"Z
p Z
Z
Z
p
2
2
2
2
4
2
2
1
e u u 4 d u +
e u u 6 d u
= 2xe x ln xx
e u d u + p
e u u 3 d u
x
9x
3 x
(8.8)
Z
p
2xe x ln xx
u 2
"
Applying symmetry to double the integrals of the even functions and half their intervals of
integration, and send the integrals of the odd functions to 0 yields:
Z
p
x ln xx
2 2xe
e
0
u 2
1
du
x
e
0
u 2
4
u du +
9x
4
u 2
u du
(8.9)
Next, the following substitution is made:
z = u2
d z = 2ud u
z =u
du =
dz
dz
= p
2u 2 z
Z
Z
Z
p
1 z 3
2 z 5
x ln xx 1
z 21
2
2
2 2xe
e z dz
e z dz +
e z dz
2
2x 0
9x 0
0

p
1 1
1
5
2
7
= 2 2xx x e x
2 2
2x 2
9x 2

p
1
1
1
3
1
1
2 531 1
x x
= 2 2xe x
+
2 2
2x 2 2 2
9x 2 2 2 2
p
p
1 3p
2 5 3p
= 2xe x x x
2x 2
9x 2 2
p
p
3
5
1
x x
x x
= 2xe x 1
+
= 2xe x 1 +
4x 6x
12x
(8.10)
(8.11)
E 1 (x)
e xt
dt
t
(9.1)
The asymptotic series for 9.1 will be found via integrating by parts:
u=
1
t
du =
E 1 (x) =
d v = e xt d t
1
v = e xt
x
1
dt
t2
Z
Z
e xt
1 xt 1 e xt
1
1 e xt
dt = e
= x
t
xt
x 1
t2
xe
x 1
t2
1
u=
1
t2
d v = e xt d t
1
v = e xt
x
2
du = 3 dt
t
E 1 (x) =
(9.2)
u=
(9.4)
Z
1
1 xt 2 e xt
e xt
1
2e
dt = x
t
xe
x
xt
x 1
t3
1
Z xt
1
1 1
2
e
= x
xe
x xe x x 1
t3
Z xt
1
1
2
e
= x 2 x+ 2
xe
x e
x 1
t3
1
t3
du =
(9.5)
d v = e xt d t
1
v = e xt
x
3
dt
t4
(9.3)
(9.6)
Z
1
2
1 xt 2 e xt
1
+
e
xe x x 2 e x x 2
xt 3
x 1
t3
1
Z xt
1
1
2
2
e
= x 2 x+ 3 x
xe
x e
x e
x 1
t3
E 1 (x) =
(9.7)
This is enough to establish a pattern. Thus:

N (1)n n!
1 X
E 1 (x) = x
+ R N (x)
e n=0 x n1
R N (x) =
(1)N +1 (N + 1)!
ex xN
Z
1
e xt
t N +2
dt
(9.8)

Homework 4: Benjamin Stahl

Uploaded by

Copyright:

Available Formats

Homework 4: Benjamin Stahl

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Homework 4: Benjamin Stahl

Uploaded by

Copyright:

Available Formats

U NIVERSITY OF C ALIFORNIA - S ANTA C RUZ

The result in 1.3 is of the form (p)(1 p) =

Prove that (n, n) =

[si n()]2n1 [cos()]2n1 d =

[si n()]2n1 [cos()]0 d

Making the substitution = 2 to the left integral:

By symmetry, the equivalence is proven:

Checking this result with n = 14 :

This result is of the form (p)(1 p) =

3 B OAS : C H . 11, S EC . 11, Q. 4

4 B OAS : C H . 11, S EC . 12, Q. 1

The integrands in 4.1 are compared with the binomial expansion:

p(p 1) 2 p(p 1)(p 2) 3

Using Wolfram Alpha: K =

Utilizing the binomial expansion (4.2) with the substitution x = k 2 si n 2 () and p =

Using Wolfram Alpha: E =

5 B OAS : C H . 11, S EC . 12, Q. 12

From this result it is clear that:

Calculating this integral using Wolfram Alpha yields:

6 B OAS : C H . 11, S EC . 13, Q. 3

(t x)2 (t x)3 (t x)4

The second exponential of the integrand is expanded according to: e z ' 1 + z + 12 z 2 + . . .

Next, the following substitution is made:

This is enough to establish a pattern. Thus:

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.