Infinite product

Calculus Level 4

r = lim n 1 + n 2 ( n + 1 ) 2 n 1 + n 2 ( n + 2 ) 2 n 1 + n 2 ( n + n 4 ) 2 n \large r = \lim_{n\to\infty} \sqrt[n]{1 + \dfrac{n^2}{(n+1)^2}} \cdot \sqrt[n]{1 + \dfrac{n^2}{(n+2)^2}} \cdots \sqrt[n]{1 + \dfrac{n^2}{(n+n^4)^2}}

Find the value of the closed form of r r to 3 decimal places.


The answer is 2.405.

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

Oleg Turcan
Feb 17, 2017

L e t : f ( x ) : = 1 + 1 x 2 t i = 1 + i n r = lim n 1 + n 2 ( n + 1 ) 2 n 1 + n 2 ( n + 2 ) 2 n . . . 1 + n 2 ( n + n 4 ) 2 n = = lim n i = 1 n 4 ( 1 + n 2 ( n + i ) 2 ) 1 n ln r = lim n i = 1 n 4 1 n ln ( 1 + n 2 ( n ( 1 + i n ) ) 2 ) = = lim n i = 1 n 4 1 n ln ( 1 + 1 ( 1 + i n ) 2 ) = lim n i = 1 n 4 1 n ln ( 1 + 1 ( t i ) 2 ) = lim n i = 1 n 4 1 n ln f ( t i ) = = lim n 1 n 3 + 1 ln ( 1 + x 2 ) d x n o w l e t : { u = ln ( 1 + x 2 ) d u = 2 x 3 1 + x 2 d x d v = d x v = x i n t e g r a t i n g b y p a r t s , i t s : u d v = u v v d u = x ln ( 1 + x 2 ) + 2 x x 3 1 + x 2 d x = = x ln ( 1 + x 2 ) + 2 1 x 2 + 1 d x = x ln ( 1 + x 2 ) + 2 arctan x + C S o : ln r = lim n 1 n 3 + 1 ln ( 1 + x 2 ) d x = x ln ( 1 + x 2 ) 0 + 2 arctan π 2 ( ln 2 + 2 arctan 1 π 4 ) lim x + x ln ( 1 + x 2 ) = lim x + ln ( 1 + x 2 ) x 1 = lim x + 2 x 3 1 + x 2 x 2 b y L H o p i t a l R u l e = lim x + 2 x 3 x 2 + x 4 = lim x + 2 x x 2 + 1 = 0 ln r = π ln 2 π 2 r = e π 2 2.405 Let:\quad f\left( x \right) :=1+\frac { 1 }{ { x }^{ 2 } } \qquad { t }_{ i }=1+\frac { i }{ n } \\ r=\lim _{ n\rightarrow \infty }{ \sqrt [ n ]{ 1+\frac { { n }^{ 2 } }{ \left( { n }+1 \right) ^{ 2 } } } \cdot \sqrt [ n ]{ 1+\frac { { n }^{ 2 } }{ \left( { n }+2 \right) ^{ 2 } } } \cdot ...\cdot \sqrt [ n ]{ 1+\frac { { n }^{ 2 } }{ \left( { n }+{ n }^{ 4 } \right) ^{ 2 } } } } =\\ =\lim _{ n\rightarrow \infty }{ \prod _{ i=1 }^{ { n }^{ 4 } }{ \left( 1+\frac { { n }^{ 2 } }{ \left( { n }+i \right) ^{ 2 } } \right) ^{ \frac { 1 }{ n } } } } \Leftrightarrow \ln { r } =\lim _{ n\rightarrow \infty }{ \sum _{ i=1 }^{ { n }^{ 4 } }{ \frac { 1 }{ n } \cdot \ln { \left( 1+\frac { { n }^{ 2 } }{ \left( n\left( 1+\frac { i }{ n } \right) \right) ^{ 2 } } \right) } } } =\\ =\lim _{ n\rightarrow \infty }{ \sum _{ i=1 }^{ { n }^{ 4 } }{ \frac { 1 }{ n } \cdot \ln { \left( 1+\frac { 1 }{ \left( 1+\frac { i }{ n } \right) ^{ 2 } } \right) } } } =\lim _{ n\rightarrow \infty }{ \sum _{ i=1 }^{ { n }^{ 4 } }{ \frac { 1 }{ n } \cdot \ln { \left( 1+\frac { 1 }{ \left( { t }_{ i } \right) ^{ 2 } } \right) } } } =\lim _{ n\rightarrow \infty }{ \sum _{ i=1 }^{ { n }^{ 4 } }{ \frac { 1 }{ n } \cdot \ln { f\left( { t }_{ i } \right) } } } =\\ =\lim _{ n\rightarrow \infty }{ \int _{ 1 }^{ { n }^{ 3 }+1 }{ \ln { \left( 1+{ x }^{ -2 } \right) } } dx } \qquad now\quad let:\begin{cases} u=\ln { \left( 1+{ x }^{ -2 } \right) } \Leftrightarrow du=\frac { -2{ x }^{ -3 } }{ 1+{ x }^{ -2 } } dx \\ dv=dx\Leftrightarrow v=x \end{cases}\\ integrating\quad by\quad parts,\quad it's:\quad \int { udv } =uv-\int { vdu } =x\ln { \left( 1+{ x }^{ -2 } \right) } +2\int { \frac { x\cdot { x }^{ -3 } }{ 1+{ x }^{ -2 } } dx } =\\ =x\ln { \left( 1+{ x }^{ -2 } \right) } +2\int { \frac { 1 }{ { x }^{ 2 }+1 } dx } =x\ln { \left( 1+{ x }^{ -2 } \right) } +2\arctan { x } +C\\ So:\quad \ln { r } =\lim _{ n\rightarrow \infty }{ \int _{ 1 }^{ { n }^{ 3 }+1 }{ \ln { \left( 1+{ x }^{ -2 } \right) } } dx } =\overbrace { x\ln { \left( 1+{ x }^{ -2 } \right) } }^{ \infty \cdot 0 } +2\overbrace { \arctan { \infty } }^{ \frac { \pi }{ 2 } } -\left( \ln { 2 } +2\overbrace { \arctan { 1 } }^{ \frac { \pi }{ 4 } } \right) \\ \lim _{ x\rightarrow +\infty }{ x\ln { \left( 1+{ x }^{ -2 } \right) } } =\lim _{ x\rightarrow +\infty }{ \frac { \ln { \left( 1+{ x }^{ -2 } \right) } }{ { x }^{ -1 } } } =\overbrace { \lim _{ x\rightarrow +\infty }{ \frac { \frac { -2x^{ -3 } }{ 1+{ x }^{ -2 } } }{ -{ x }^{ -2 } } } }^{ by\quad L'Hopital\quad Rule } =\lim _{ x\rightarrow +\infty }{ \frac { 2x^{ -3 } }{ { x }^{ -2 }+{ x }^{ -4 } } = } \lim _{ x\rightarrow +\infty }{ \frac { 2x }{ { x }^{ 2 }+1 } = } 0\\ \\ \therefore \ln { r } =\pi -\ln { 2 } -\frac { \pi }{ 2 } \Leftrightarrow \boxed { r=\frac { \sqrt { { e }^{ \pi } } }{ 2 } \cong 2.405 }

Moderator note:

This solution takes the standard approximation techniques and carefully deals with the integrals to find the limits. Using logarithms to convert a product into a summation, allows us to apply the Riemann sums to determine the limit.

How did you get n^3 +1 as upper limit

Ishan Chopra - 4 years, 3 months ago

Log in to reply

Ah, you're right, this is not immediately clear. @Oleg Turcan , can you elaborate on this point? I think it would make your solution more easily understood...

Pi Han Goh - 4 years, 3 months ago

Sorry for being nooby but could you give a bit more details as to how you got to the integral?

Romain Farthoat - 4 years, 3 months ago

Log in to reply

It's called Riemann sums .

Pi Han Goh - 4 years, 3 months ago

Log in to reply

Riemann sums uses an interval with fixed limits. Let's consider lim n 1 n k = 1 n 4 s i n ( 2 k π n ) = 0 \lim_{n \rightarrow \infty} \dfrac{1}{n} \sum_{k=1}^{n^4} sin(\dfrac {2k\pi}{n}) = 0

but 0 s i n ( 2 π t ) d t 0 \int_0^{\infty} sin(2\pi t) dt \neq 0

Abdelhamid Saadi - 4 years, 3 months ago

Log in to reply

@Abdelhamid Saadi I noticed this issue as well... That particular problem doesn't occur in this solution because the integrand is non-negative, but this is a nontrivial theorem.

To avoid having to think about this, my solution involved rewriting the sum as a Lebesgue integral of a step function and then using DC to evaluate the limit, but I don't know if this would be an appropriate solution for brilliant.org...

Brian Moehring - 4 years, 3 months ago

Log in to reply

@Brian Moehring I believe it would. It is always nice to have different sorts of solutions

Agnishom Chattopadhyay - 4 years, 3 months ago

@Brian Moehring What is DC? Dominated Convergence theorem ? If yes, can you please post it? I don't know how you can invoke this theorem in this problem.

Pi Han Goh - 4 years, 3 months ago

Log in to reply

@Pi Han Goh Yes, I was talking about the Dominated Convergence theorem.

I don't know if I'll have time today or tomorrow, but I'll post it as soon as I find some time to LaTeX it up...

Brian Moehring - 4 years, 3 months ago

Allright!!!!

Gustav Liljedahl - 4 years, 3 months ago

Why is n^3 + 1 the upper limit?!

Soumil Sahu - 3 years, 8 months ago
Brian Moehring
Mar 3, 2017

Note : I've added this in response to some questions in the comments in Oleg's post. I'm going to show one way to justify using an improper Riemann integral to find the answer by way of the Lebesgue Dominated Convergence theorem. At this point, the solution would follow Oleg's almost exactly, so I will just reference his solution for that part.


By applying the natural logarithm and using continuity, ln r = lim n i = 1 n 4 1 n ln ( 1 + n 2 ( n + i ) 2 ) = lim n i = 1 n 4 1 n ln ( 1 + 1 ( 1 + i n ) 2 ) . \ln r = \lim_{n\rightarrow\infty} \sum_{i=1}^{n^4} \frac{1}{n} \ln\left(1 + \frac{n^2}{(n+i)^2}\right) = \lim_{n\rightarrow\infty} \sum_{i=1}^{n^4} \frac{1}{n} \ln\left(1 + \frac{1}{\left(1+\tfrac{i}{n}\right)^2}\right).

To investigate this limit, we define f ( x ) = ln ( 1 + 1 ( 1 + x ) 2 ) , f n ( x ) = { f ( 1 n n x ) 0 < x n 3 0 otherwise f(x) = \ln\left(1 + \frac{1}{(1+x)^2}\right), \qquad f_n(x) = \begin{cases} f\left(\frac{1}{n}\lceil{nx}\rceil\right) & 0 < x \leq n^3 \\ 0 & \mbox{otherwise}\end{cases} and note that ( 0 , ) f n ( x ) d x = ( 0 , n 3 ] f ( 1 n n x ) d x = i = 1 n 4 ( i 1 n , i n ] f ( 1 n n x ) d x = i = 1 n 4 ( i 1 n , i n ] f ( i n ) d x = i = 1 n 4 f ( i n ) ( i n i 1 n ) = i = 1 n 4 1 n ln ( 1 + 1 ( 1 + i n ) 2 ) , \begin{aligned}\int\limits_{(0,\infty)} f_n(x)\,dx &= \int\limits_{(0, n^3]} f\left(\tfrac{1}{n}\lceil{nx}\rceil\right)\,dx \\ &= \sum_{i=1}^{n^4} \int\limits_{\left(\tfrac{i-1}{n}, \tfrac{i}{n}\right]} f\left(\tfrac{1}{n}\lceil{nx}\rceil\right)\,dx \\ &= \sum_{i=1}^{n^4} \int\limits_{\left(\tfrac{i-1}{n}, \tfrac{i}{n}\right]} f\left(\tfrac{i}{n}\right)\,dx \\ &= \sum_{i=1}^{n^4} f\left(\tfrac{i}{n}\right)\left(\frac{i}{n} - \frac{i-1}{n}\right) \\ &= \sum_{i=1}^{n^4} \frac{1}{n} \ln\left(1 + \frac{1}{\left(1+\tfrac{i}{n}\right)^2}\right),\end{aligned} so ln r = lim n ( 0 , ) f n ( x ) d x . \ln r = \lim_{n\rightarrow\infty} \int\limits_{(0,\infty)} f_n(x)\,dx.

Then, since f f is non-negative, non-increasing, and right-continuous on x > 0 x>0 , we can see that

  • lim n f n ( x ) = f ( x ) \displaystyle\lim_{n\rightarrow\infty} f_n(x) = f(x) for all x > 0 x > 0 . [by right-continuity]
  • 0 f n ( x ) f ( x ) 0 \leq f_n(x) \leq f(x) for all n n and x > 0 x > 0 . [the inequalities follow by non-negativity and non-increasing, respectively]

Therefore, by the Dominated Convergence theorem, we can conclude that IF ( 0 , ) f ( x ) d x < \int_{(0,\infty)} f(x)\, dx < \infty , then ln r = ( 0 , ) f ( x ) d x , \ln r = \int_{(0,\infty)} f(x)\,dx, and since f ( x ) f(x) is positive, we may evaluate this Lebesgue integral as the improper Riemann integral 0 ln ( 1 + 1 ( 1 + x ) 2 ) d x = 1 ln ( 1 + 1 x 2 ) d x = lim N 1 N ln ( 1 + 1 x 2 ) d x \int_0^\infty \ln\left(1 + \frac{1}{(1+x)^2}\right) \,dx = \int_1^\infty \ln\left(1 + \frac{1}{x^2}\right)\,dx = \lim_{N\rightarrow\infty} \int_1^N \ln\left(1 + \frac{1}{x^2}\right)\,dx

Since this improper integral equals π 2 ln 2 \frac{\pi}{2} - \ln 2 [see Oleg's solution], which is finite, we may conclude by the above remarks that ln r = π 2 ln 2 \ln r = \frac{\pi}{2} - \ln 2

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...