Christmas integral

Calculus Level 4

Find the value of the following expression

8 ( 0 e x sin x x d x ) 2 0 x e x 1 d x \frac{\displaystyle 8 \left(\int_{0}^\infty \frac{e^{-x}\sin x}{x} dx \right)^2}{\displaystyle \int_{0}^\infty \frac{x}{e^x-1} dx}


The answer is 3.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Karan Chatrath
Dec 25, 2019

Let the integral in the numerator be

I 1 = 0 e t sin t t d t I_1 = \int_{0}^{\infty} \frac{e^{-t}\sin{t}}{t}dt

This integral will be solved using the concept of Laplace transforms as such:

Consider the Laplace transform of sin t \sin{t} :

L [ sin t ] = 0 e s t sin ( t ) d t L\left[\sin{t}\right]= \int_{0}^{\infty} e^{-st}\sin(t)dt

By referring to a table of Laplace transforms, this standard integral evaluates to:

0 e s t sin ( t ) d t = 1 1 + s 2 \int_{0}^{\infty} e^{-st}\sin(t)dt = \frac{1}{1+s^2}

Generally for Laplace transforms:

L [ f ( t ) ] = 0 e s t f ( t ) d t = F ( s ) L\left[f(t)\right] = \int_{0}^{\infty} e^{-st}f(t)dt = F(s)

Now, using the table of Laplace transforms, one can use the identity:

L [ f ( t ) t ] = 0 e s t f ( t ) t d t = s F ( p ) d p L\left[\frac{f(t)}{t}\right] = \int_{0}^{\infty} \frac{e^{-st}f(t)}{t}dt = \int_{s}^{\infty}F(p)dp

Using this identity for the Laplace transform of a sine function, one gets:

L [ sin t t ] = 0 e s t sin t t d t = s 1 1 + p 2 d p L\left[\frac{\sin{t}}{t}\right] = \int_{0}^{\infty} \frac{e^{-st}\sin{t}}{t}dt = \int_{s}^{\infty}\frac{1}{1+p^2}dp

Taking s = 1 s=1 gives:

I 1 = 0 e t sin t t d t = 1 1 1 + p 2 d p = π 4 \boxed{I_1=\int_{0}^{\infty} \frac{e^{-t}\sin{t}}{t}dt = \int_{1}^{\infty}\frac{1}{1+p^2}dp=\frac{\pi}{4}}


Let the integral in the denominator be:

I 2 = 0 x e x 1 d x = 0 e x x 1 e x d x I_2 = \int_{0}^{\infty} \frac{x}{e^x-1}dx=\int_{0}^{\infty} \frac{e^{-x}x}{1-e^{-x}}dx

The function e x e^{-x} is less than unity in the interval ( 0 , ) (0,\infty) . Therefore, the following term can be written as an infinite geometric series as such:

e x 1 e x = e x + e 2 x + e 3 x + \frac{e^{-x}}{1-e^{-x}} = e^{-x}+e^{-2x}+e^{-3x}+\dots

Replacing this in the integral gives:

I 2 = 0 ( e x + e 2 x + e 3 x + ) x d x = k = 1 ( 0 x e k x ) d x I_2 = \int_{0}^{\infty} \left(e^{-x}+e^{-2x}+e^{-3x}+\dots\right)xdx = \sum_{k=1}^{\infty}\left(\int_{0}^{\infty}xe^{-kx}\right)dx

Now, by introducing a change of variables k x = z kx=z , and after some simplification, the sum of integrals transforms to the sum:

I 2 = ( k = 1 1 k 2 ) 0 z e z d z I_2 = \left(\sum_{k=1}^{\infty}\frac{1}{k^2}\right)\int_{0}^{\infty}ze^{-z}dz

The integral:

0 z e z d z = 1 \int_{0}^{\infty}ze^{-z}dz=1

and the following sum refers to the Basel problem: k = 1 1 k 2 = π 2 6 \sum_{k=1}^{\infty}\frac{1}{k^2}=\frac{\pi^2}{6}

Therefore:

I 2 = 0 x e x 1 d x = π 2 6 \boxed{I_2 = \int_{0}^{\infty} \frac{x}{e^x-1}dx = \frac{\pi^2}{6}}


The required expression to be evaluated is:

8 I 1 2 I 2 = 3 \boxed{\frac{8I_1^2}{I_2}=3}

Relevant references just a google search away:

  • The Reimann-Zeta function
  • Laplace Transforms
  • The Basel problem

The integral is the denominator is a particular case of a very nice integral. For n N n \in \N we have 0 x n 1 e x 1 d x = Γ ( n ) ζ ( n ) \int_0^{\infty} \dfrac{x^{n-1}}{e^x -1} \,dx = \Gamma (n) \zeta (n) However, the expression is true for any n C n \in \mathbb{C} where both functions are defined and analytic.

Tapas Mazumdar - 1 year, 5 months ago
Alapan Das
Dec 25, 2019

See, 0 s i n ( x ) e x x d x \int_{0}^\infty \frac{sin(x)e^{-x}}{x} dx

= 0 ( 1 x 2 3 ! + x 4 5 ! x 6 7 ! + . . . . ) e x d x =\int_{0}^\infty (1-\frac{x^2}{3!}+\frac{x^4}{5!}-\frac{x^6}{7!}+....)e^{-x} dx

= 0 ( 1 ) k 0 x ( 2 k + 1 ) 1 e x d x ( 2 k + 1 ) ! =\sum_{0}^{\infty} \frac{(-1)^k \int_{0}^\infty x^{(2k+1)-1} e^{-x} dx}{(2k+1)!}

= 0 ( 1 ) k Γ ( 2 k + 1 ) ( 2 k + 1 ) ! =\sum_{0}^\infty (-1)^{k}\frac{\Gamma(2k+1)}{(2k+1)!}

= Γ ( 1 ) Γ ( 3 ) 3 ! + Γ ( 5 ) 5 ! Γ ( 7 ) 7 ! . . . . =\Gamma(1)-\frac{\Gamma(3)}{3!}+\frac{\Gamma(5)}{5!}-\frac{\Gamma(7)}{7!}-....

= 1 1 3 + 1 5 1 7 + . . . =1-\frac{1}{3}+\frac{1}{5}-\frac{1}{7}+... = 0 1 1 1 + x 2 d x = π 4 \int_{0}^1 \frac{1}{1+x^2} dx=\frac{π}{4}

Again we know 0 x e x 1 d x = ζ ( 2 ) = π 2 6 \int_{0}^\infty \frac{x}{e^x-1} dx=\zeta(2)=\frac{π^2}{6} .

Then the desired result is 8 π 2 16 π 2 6 = 3 \frac{8\frac{π^2}{16}}{\frac{π^2}{6}}=3 .

That solution is extremely unthinkable.

William Ly - 6 months, 2 weeks ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...