For n ∈ N ≥ 1 , m ∈ N ≥ 2 , define I ( n , m ) = ∫ 0 ∞ x n m + 1 x n − 1 lo g ( x ) d x . The definite integral given by I ( 3 , 3 ) converges to the value − c a π b , where a , b and c denote positive integers such that a and c are coprime. Find a + b + c .
Bonus : Show that I ( n , 2 ) = 0 for all n ∈ N ≥ 1 .
Bonus : Find a general formula for I ( n , m ) for all n ∈ N ≥ 1 , m ∈ N ≥ 2 , which only contains elementary functions.
Remark : You may want to use that for n ∈ N ≥ 2 , ∫ 0 ∞ x n + 1 d x = sin ( π / n ) π / n .
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.
In the lemma below we will prove that for all n ∈ N ≥ 1 , m ∈ N ≥ 2 , I ( n , m ) = − n 2 m 2 sin 2 ( π / m ) π 2 cos ( π / m ) . From this, it automatically follows that I ( 3 , 3 ) = − 2 4 3 2 π 2 and I ( n , 2 ) = 0 for all n ∈ N ≥ 1 . Hence, a + b + c = 2 + 2 + 2 4 3 = 2 4 7 .
Lemma : For all n ∈ N ≥ 1 , m ∈ N ≥ 2 , I ( n , m ) = − n 2 m 2 sin 2 ( π / m ) π 2 cos ( π / m ) . First of all, by substituting u = x n , we get rid of the first variable of the integral I ( n , m ) = ∫ 0 ∞ x n m + 1 x n − 1 lo g ( x ) d x = n 2 1 ∫ 0 ∞ u m + 1 lo g ( u ) d u . We will now focus on the last given integral, which we will denote by J .
We define f : C → C , f ( z ) = z m + 1 lo g ( z ) and lo g : C → C , lo g ( z ) = lo g ∣ z ∣ + i ⋅ ar g ( z ) , where − 2 π < ar g ( z ) < 2 3 π denotes the argument of z . Consider the integral ∮ C f ( z ) d z , over the contour C going anti-clockwise as seen in the following picture, where r < 1 , R > 1 and the "diagonal" line is given by z = λ e 2 π i / m , with r ≤ λ ≤ R .
The two parts of the contour which are not represented by lines are part of the circles with center O and radius R , r , respectively. In the picture the point z 0 = e π i / m can be seen, which is the only singularity within the contour.
From the definition of the contour, it follows that ∮ C f ( z ) d z = J + ∫ C R f ( z ) d z + ∫ C 2 π i / m f ( z ) d z + ∫ C r f ( z ) d z , where C R , C 2 π i / m , C r define the different parts in an obvious way. We will now show that both ∫ C R f ( z ) d z and ∫ C r f ( z ) d z tend to 0 as R → ∞ , r → 0 . For the first integral, observe that ∣ z m + 1 ∣ ≥ ∣ z m ∣ − 1 so that ∣ ∣ ∣ ∣ ∫ C R f ( z ) d z ∣ ∣ ∣ ∣ ≤ m 2 π R z ∈ C R max ∣ ∣ ∣ ∣ z m + 1 lo g ( z ) ∣ ∣ ∣ ∣ ≤ m 2 π R z ∈ C R max R m − 1 ∣ lo g ( z ) ∣ ≤ m 2 π R z ∈ C R max R m − 1 lo g ( R ) + ∣ i ⋅ ar g ( z ) ∣ R → ∞ 0 . The other integral can be dealt with in a similar fashion. Now, note that by the remark in the problem, ( r , R ) → ( 0 , ∞ ) lim ∫ C 2 π i / m f ( z ) d z = − e 2 π i / m ∫ 0 ∞ R m + 1 lo g ( R ) + 2 π i / m d r = − e 2 π i / m ( J + sin ( π / m ) 2 π 2 i / m 2 ) . Now, by the Residue theorem, it follows that ∮ C f ( z ) d z = 2 π i ⋅ Res z = e π i / m z m + 1 lo g ( z ) = 2 π i m e π i ( m − 1 ) / m π i / m = m 2 2 π 2 e π i / m . Gathering all information we got so far, we notice that J = m 2 2 π 2 ⋅ 1 − e 2 π i / m e π i / m + i e 2 π i / m csc ( π / m ) . Using that csc ( z ) = e i z − e − i z 2 i , after some algebraic manipulations it can be shown that 1 − e 2 i z e i z + i e 2 i z csc ( z ) = ( e i z − e − i z ) 2 e i z + e − i z = − 2 sin 2 ( z ) cos ( z ) . So, J = − m 2 sin 2 ( π / m ) π 2 cos ( π / m ) , and we conclude that for all n ∈ N ≥ 1 , m ∈ N ≥ 2 , ∫ 0 ∞ x n m + 1 x n − 1 lo g ( x ) d x = − n 2 m 2 sin 2 ( π / m ) π 2 cos ( π / m ) .
Problem Loading...
Note Loading...
Set Loading...
The substitution y = x m n gives I ( n , m ) = m 2 n 2 1 ∫ 0 ∞ y + 1 y m 1 − 1 ln y d y = m 2 n 2 1 d α d B ( α , 1 − α ) ∣ ∣ ∣ α = m 1 = m 2 n 2 1 d α d Γ ( α ) Γ ( 1 − α ) ∣ ∣ ∣ α = m 1 = m 2 n 2 1 d α d π c o s e c π α ∣ ∣ ∣ α = m 1 = − m 2 n 2 π 2 c o s e c m π cot m π which makes I ( n , 2 ) = 0 for n ≥ 1 and I ( 3 , 3 ) = − 2 4 3 2 π 2 , giving the answer 2 + 2 + 2 4 3 = 2 4 7 .