Finding limit using integration

songoku
Messages
2,467
Reaction score
382
Homework Statement
Please see below
Relevant Equations
Riemann Sum
1723095708837.png


I want to ask why the answer is not zero. If n approaches infinity, it means each term will approach zero so why the answer is not zero?

Thanks
 
Physics news on Phys.org
songoku said:
I want to ask why the answer is not zero. If n approaches infinity, it means each term will approach zero so why the answer is not zero?
True, each term does approach zero, but the number of terms in the sum approaches infinity. Can't the two effects compensate each other to yield a finite result?
 
  • Like
Likes FactChecker and songoku
renormalize said:
True, each term does approach zero, but the number of terms in the sum approaches infinity. Can't the two effects compensate each other to yield a finite result?
I am thinking something like 0 + 0 + 0 + 0 + ..... = 0. Not sure how the compensation works

Thanks
 
songoku said:
I am thinking something like 0 + 0 + 0 + 0 + ..... = 0. Not sure how the compensation works
Consider this simpler example where each term goes to zero, but the sum of the n terms is always 1.
(n terms): 1/n+1/n+1/n+...+1/n = n/n = 1
 
  • Like
Likes songoku and fresh_42
songoku said:
I want to ask why the answer is not zero.
Because that's not one of the listed potential answers. :oldbiggrin:
songoku said:
I am thinking something like 0 + 0 + 0 + 0 + ..... = 0.
That's not reasonable -- you can't take the limit of each individual term.
songoku said:
Not sure how the compensation works
What is meant by compensation here is that although each term is approaching zero, there are more and more terms. The decreasing size of each term is compensated by the contrasting effect of adding more and more of them. Kind of like inflating a balloon that has a hole in it. If the amount of air escaping the balloon is more than the amount coming in, the balloon won't be inflated, but if more air is coming in than going out, the balloon inflates. The air coming in can compensate for the air going out.
 
  • Like
Likes DeBangis21 and songoku
Mark44 said:
That's not reasonable -- you can't take the limit of each individual term.
Is the reason that the number of terms is infinite?

Thanks
 
songoku said:
Is the reason that the number of terms is infinite?
Yes.
 
Thank you very much for all the help and explanation renormalize, FactChecker, Mark44
 
@songoku if you already figured it out, can you post the solution for reference?
 
  • Like
Likes songoku and SammyS
  • #10
docnet said:
@songoku if you already figured it out, can you post the solution for reference?
$$\lim_{n\to \infty} \left(\frac{1}{2n+1}+\frac{1}{2n+2} + ... + \frac{1}{2n+n} \right)$$
$$=\lim_{n\to \infty} \sum_{r=1}^{n} \left(\frac{1}{2n+r}\right)$$
$$=\lim_{n\to \infty} \sum_{r=1}^{n} \left(\frac{1}{2+\frac{r}{n}} . \frac{1}{n} \right)$$
$$=\int_{0}^{1} \frac{1}{2+x} dx$$
$$=\ln(2+x)|^{1}_{0}$$
$$=\ln(3) - \ln(2)$$
$$=\ln \left(\frac{3}{2} \right)$$
 
  • #11
songoku said:
$$=\lim_{n\to \infty} \sum_{r=1}^{n} \left(\frac{1}{2+\frac{r}{n}} . \frac{1}{n} \right)$$
$$=\int_{0}^{1} \frac{1}{2+x} dx$$
This step is not immediately clear to me. Could someone explain why the limit is the integral?
 
  • #12
docnet said:
This step is not immediately clear to me. Could someone explain why the limit is the integral?
For me, I imagine the rectangles of Riemann sum
 
  • #13
It must be as ##x=\frac{r}{n}## and ##\Delta x = \frac{1}{n}##. The boundaries are ##0## and ##1## because
$$0< r \leq n\Leftrightarrow 0<\frac{r}{n}\leq 1\Leftrightarrow 0<x\leq 1,$$
which goes along with $$\sum_{r=1}^n\frac{1}{n}=1$$
for the traditional right Riemann sum $$\sum_{r=1}^n \frac{1}{2+x_r}\Delta x\,$$
and ##(x_0,...,x_n)## an equally spaced partition of ##[0,1].##
 
Last edited:
  • #14
If f : [0, \infty) \to \mathbb{R} is strictly decreasing, we have <br /> \int_1^{n+1} f(x)\,dx \leq \sum_{r=1}^n f(r) \leq \int_0^n f(x)\,dx which for f(x) = (2n + x)^{-1} is <br /> \ln \left( \frac{3n + 1}{2n + 1} \right) \leq \sum_{r=1}^n \frac{1}{2n + r} \leq \ln \tfrac 32. Taking the limit n \to \infty establishes the result.
 
  • #15
songoku said:
$$=\lim_{n\to \infty} \sum_{r=1}^{n} \left(\frac{1}{2+\frac{r}{n}} . \frac{1}{n} \right)$$
$$=\int_{0}^{1} \frac{1}{2+x} dx$$

$$ \int_{0}^{1}\frac{1}{2+x}dx \approx \sum_{r=1}^{n}\left(\frac{1}{2+\frac{r}{n}}\cdot\frac{1}{n}\right) = \sum_{r=1}^{n}\left(\frac{1}{2+(0+r\frac{1-0}{n})}\cdot\frac{1-0}{n}\right) $$ is a right rectangular approximation or a right Riemann sum.
 
Back
Top