Does the following improper integral converge?
∫ 0 ∞ sin ( x ) sin ( x 2 ) d x
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.
Whoa, thanks!!
I was expecting some kind of bounding / alternating series argument, but wasn't able to push through.
Log in to reply
Well technically this was a bounding argument. :) No problem!
Log in to reply
Right. I was looking for a more statement like
If f ( x ) is a continuous periodic function (hence bounded) of period 2 π , then ∫ 0 ∞ f ( x ) sin ( x 2 ) d x exists.
I believe that this claim is true, because the "(not so) random testing of value of sin x 2 " makes everything work out.
In particular,
1. The value should be something like
∫
0
2
π
f
(
x
)
g
(
x
)
d
x
where
g
(
x
)
=
"Cecaro summation" of
sin
y
2
when
y
2
≡
x
(
m
o
d
2
π
)
.
2. There are no "limit does not exist" issues like with
∫
0
∞
sin
x
d
x
.
Problem Loading...
Note Loading...
Set Loading...
We can use the following formula: sin ( a ) sin ( b ) = 2 cos ( a − b ) − cos ( a + b ) :
∫ 0 ∞ sin ( x ) sin ( x 2 ) d x = 2 1 ∫ 0 ∞ ( cos ( x 2 − x ) − cos ( x 2 + x ) ) d x = r → ∞ lim 2 1 ( ∫ 0 r cos ( x 2 − x ) d x − ∫ 0 r cos ( x 2 + x ) d x ) = 2 1 r → ∞ lim ( ∫ 0 r cos ( x 2 − x ) d x − ∫ 1 r + 1 cos ( x 2 − x ) d x ) = 2 1 r → ∞ lim ( ∫ 0 1 cos ( x 2 − x ) d x − ∫ r r + 1 cos ( x 2 − x ) d x ) = 2 1 ∫ 0 1 cos ( x 2 − x ) d x − 2 1 r → ∞ lim ∫ r r + 1 cos ( x 2 − x ) d x
For 0 ≤ x ≤ 1 we have − 4 1 ≤ x 2 − x ≤ 0 , so cos ( x 2 − x ) ≥ 0 and thus ∫ 0 1 cos ( x 2 − x ) d x > 0 .
Fix r > 3 π and let I ( r ) = ∫ r r + 1 cos ( x 2 − x ) d x . Now we just need to prove that r → ∞ lim I ( r ) = 0 :
First let y = x 2 − x . Then x = 2 1 + y + 4 1 . Then, using integration by substitution, we get: I ( r ) = ∫ r 2 − r r 2 + r 4 y + 1 cos ( y ) d y
Define the sequence ( a n ) as follows. Let a 1 = p be the smallest number in the interval [ r 2 − r , r 2 + r ] of the form 2 π k − 2 π for k ∈ Z , and let a 2 m − 1 = q be the largest number in that interval of that form (where m = 2 π q − p + 1 ). And for 2 ≤ n ≤ 2 m − 2 let a n = a n − 1 + π . Note that a 2 m − 1 = a 2 m − 2 + π as well. Also let a 0 = r 2 − r and a 2 m = r 2 + r .
Now for 0 ≤ n ≤ 2 m − 1 , let f ( n ) = ∫ a n a n + 1 4 y + 1 cos ( y ) d y .
When n = 0 we have a 1 − a 0 < π . Therefore: ∣ f ( 0 ) ∣ ∣ f ( 0 ) ∣ = ∣ ∣ ∣ ∣ ∫ a 0 a 1 4 y + 1 cos ( y ) d y ∣ ∣ ∣ ∣ ≤ ∫ a 0 a 1 4 y + 1 ∣ cos ( y ) ∣ d y ≤ ∫ a 0 a 1 4 a 0 + 1 1 d y = 4 a 0 + 1 a 1 − a 0 < 4 a 0 + 1 π < 2 r − 1 π . . . → ( A )
Similarly, ∣ f ( 2 m − 1 ) ∣ < 4 a 2 m − 1 + 1 π < 4 ( r 2 + r − π ) + 1 π < 2 r π . . . → ( B ) .
Now let k be an integer 1 ≤ k ≤ m − 1 . For a 2 k − 1 ≤ y ≤ a 2 k we have cos ( y ) ≥ 0 . Then, ∫ a 2 k − 1 a 2 k 4 a 2 k + 1 cos ( y ) d y 4 a 2 k + 1 sin ( a 2 k ) − sin ( a 2 k − 1 ) 4 a 2 k + 1 2 4 ( r 2 + r ) + 1 2 2 r + 1 2 ≤ ≤ ≤ < < ∫ a 2 k − 1 a 2 k 4 y + 1 cos ( y ) d y f ( 2 k − 1 ) ≤ ≤ ≤ < < ∫ a 2 k − 1 a 2 k 4 a 2 k − 1 + 1 cos ( y ) d y 4 a 2 k − 1 + 1 sin ( a 2 k ) − sin ( a 2 k − 1 ) 4 a 2 k − 1 + 1 2 4 ( r 2 − r ) + 1 2 2 r − 1 2
Similarly we can show that: 2 r − 1 − 2 < f ( 2 k ) < 2 r + 1 − 2
Therefore, combining all these equations gives: ∣ ∣ ∣ ∣ ∣ j = 1 ∑ 2 m − 2 f ( j ) ∣ ∣ ∣ ∣ ∣ ≤ ( m − 1 ) ( 2 r − 1 2 − 2 r + 1 2 ) . . . → ( C )
Now since a 2 m − 1 = a 1 + ( 2 m − 2 ) π , then we have: 2 r = a 2 m − a 0 = ( a 2 m − a 2 m − 1 ) + ( a 2 m − 1 − a 1 ) + ( a 1 − a 0 ) ≥ a 2 m − 1 − a 1 = 2 ( m − 1 ) π Therefore m ≤ 1 + π r . Substituting this into ( C ) gives:
∣ ∣ ∣ ∣ ∣ j = 1 ∑ 2 m − 2 f ( j ) ∣ ∣ ∣ ∣ ∣ ≤ ( π r ) ( 2 r − 1 2 − 2 r + 1 2 ) = π ( 4 r 2 − 1 ) 4 r . . . → ( D )
Therefore, combining ( A ) , ( B ) and ( D ) gives:
∣ I ( r ) ∣ ∣ I ( r ) ∣ = ∣ ∣ ∣ ∣ ∣ ∫ r 2 − r r 2 + r 4 y + 1 cos ( y ) d y ∣ ∣ ∣ ∣ ∣ = ∣ ∣ ∣ ∣ ∣ j = 0 ∑ 2 m − 1 f ( j ) ∣ ∣ ∣ ∣ ∣ ≤ ∣ f ( 0 ) ∣ + ∣ ∣ ∣ ∣ ∣ j = 1 ∑ 2 m − 2 f ( j ) ∣ ∣ ∣ ∣ ∣ + ∣ f ( 2 m − 1 ) ∣ < 2 r − 1 π + π ( 4 r 2 − 1 ) 4 r + 2 r π
As r approaches infinity, this expression approaches 0 . Therefore by the Squeeze Theorem, r → ∞ lim I ( r ) = 0 .
Therefore, ∫ 0 ∞ sin ( x ) sin ( x 2 ) d x = 2 1 ∫ 0 1 cos ( x 2 − x ) d x , which as we proved earlier, converges to a positive value.