0
$\begingroup$

I'm trying to test the following series for convergence: $$\sum_{n=2}^{\infty}\frac{n}{\sqrt{n^5-n^3}}$$

I've progressed through several tests but am having trouble developing an intuition of how to approach a problem like this. I've checked the following cases so far:

  • Divergence Test ($\lim a_n = 0$, so not helpful)
  • Geometric Series Test (I can't find a straightforward way to find a common ratio)
  • p-Series Test (it does not appear to be a p-Series)
  • Limit Comparison and Comparison Tests (I can't find another series with which to prove convergence or divergence)
  • Integral Test (I'm unable to integrate the expression)

Obviously I'm missing something here, but I'm just not sure which it is.

$\endgroup$
8
  • 2
    $\begingroup$ Terms are roughly on the order of $\frac{n}{n^{5/2}} = \frac{1}{n^{3/2}}$ which should converge. How you get there is combination of comparison and $p$-series tests. $\endgroup$
    – Randall
    Jan 23, 2018 at 14:28
  • $\begingroup$ OK. It appears that I can just use the comparison test with something like $\frac{1}{x^\left(1.1\right)}$. Thank you! $\endgroup$
    – Alex Johnson
    Jan 23, 2018 at 14:31
  • 1
    $\begingroup$ Oftentimes, limit comparison is easier to use than comparison because you don't have to care if the "right" series dominates. $\endgroup$
    – Randall
    Jan 23, 2018 at 14:33
  • 1
    $\begingroup$ You should try to work it all out and answer your own question here. Would be good for you. $\endgroup$
    – Randall
    Jan 23, 2018 at 14:33
  • $\begingroup$ @Randall The trouble I have with limit comparison is that I have a hard time ending up with a constant. Seems I'm always getting infinity or infinity over infinity. Adding in the radical in the denominator makes it even easier to get lost down dead-ends. $\endgroup$
    – Alex Johnson
    Jan 23, 2018 at 14:46

4 Answers 4

Reset to default
3
$\begingroup$

First of all, you seem to have misused some of the tests. For example, the divergence test only says that the series $\sum a_n$ diverges if $\lim a_n\neq 0$, which is not true here since $$\lim_{n\to\infty}\frac{n}{\sqrt{n^5-n^3}} = 0.$$

Second of all, you can make a comparison test:

$$\frac{n}{\sqrt{n^5-n^3}} = \frac{n}{\sqrt{n^5}\sqrt{1 - \frac{1}{n^2}}}=\frac{1}{\sqrt{n^3}}\cdot \frac{1}{\sqrt{1-\frac{1}{n^2}}}\leq 2\cdot \frac{1}{\sqrt{n^3}}$$

and of course, $$\sum_{n=1}^\infty \frac{2}{\sqrt{n^3}}$$ converges as it is two times a $p$-series.

$\endgroup$
2
  • $\begingroup$ Thanks for the explanation and for pointing out my error with the divergence test. I'd just been checking if the sequence increases, but now I I see what I was doing wrong there. As for implementing the comparison test, I think I just need more practice so that I can "see" what you're seeing without having to try a dozen other strategies first. I appreciate your help. $\endgroup$
    – Alex Johnson
    Jan 23, 2018 at 14:52
  • $\begingroup$ @AlexJohnson In general, you just need to find the biggest factor of the numerator and denominator. In this case, the numerator is $n$, and the denominator, for large $n$, is dominated by $\sqrt{n^5}$. Since $\frac{5}{2} - 1> 1$, you have convergence. $\endgroup$
    – 5xum
    Jan 23, 2018 at 14:54
0
$\begingroup$

You forgot the equivalence test: $$\sqrt{n^5-n^3\strut}\sim_\infty n^{5/2},\enspace\text{hence }\;\frac n{\sqrt{n^5-n^3\strut}}\sim_\infty \frac n{n^{5/2}}=\frac 1{n^{3/2}}, $$ which is a convergent Riemann series.

$\endgroup$
0
$\begingroup$

Let $a_n$ be the $n$th term of your series and let $b_n= \frac{1}{n^{3/2}}$. Why is this my choice of $b_n$? See my first comment under your question.

Now, limit compare by $\lim_{n \to \infty} \frac{b_n}{a_n}$. Do the algebra and eventually get $$ \frac{b_n}{a_n}= \frac{\sqrt{n^5-n^3}}{n^{5/2}} = \frac{\sqrt{n^5-n^3}}{\sqrt{n^5}} = \sqrt{\frac{n^5-n^3}{n^5}} = \sqrt{1-\frac{1}{n^2}}. $$ Now take the limit, and notice that it's easily $1$. Limit comparison now tells you that your series converges because $\sum_n \frac{1}{n^{3/2}}$ is a convergent $p$-series.

$\endgroup$
2
  • 1
    $\begingroup$ I fudged a step or two but was able to get to the same place as you. This was all a tremendous help. Thank you again. $\endgroup$
    – Alex Johnson
    Jan 23, 2018 at 15:00
  • $\begingroup$ Excellent. You're well on your way. $\endgroup$
    – Randall
    Jan 23, 2018 at 15:00
0
$\begingroup$

Simply note that

$$\frac{n}{\sqrt{n^5-n^3}}\sim \frac{1}{n^\frac32}$$

thus $\sum_{n=2}^{\infty}\frac{n}{\sqrt{n^5-n^3}}$ converges by comparison with

$$\sum_{n=2}^{\infty}\frac{1}{n^\frac32}$$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.