Question:
Does the series n / sqrt((n^3)+2) converge or diverge?
anonymous
1970-01-01 00:00:00 UTC
Does the series n / sqrt((n^3)+2) converge or diverge?
Nine answers:
Anon E. Moose アナンイムース
2011-05-18 20:57:24 UTC
This series is roughly equivalent to [n / n^(3/2)] → [1 / n^(1/2)] which diverges because of the p-test (the power is 0 < p <= 1).
nyc_kid
2011-05-18 20:44:50 UTC
n / sqrt((n^3)+2) > n/sqrt(3(n^3)) = (1/3)(1/sqrt(n)) > (1/3)(1/n)

You know that the series (1/n) diverges and thus, your original series must diverge.
Graduate
2011-05-18 20:42:11 UTC
The Series of n / sqrt((n^3)+2) >= the series 1/n.



Since we know that the series 1/n diverges, the series which is greater than 1/n will also diverge.



RegardZ

EDEM
?
2016-12-14 13:54:04 UTC
Leibniz Test For Alternating Series
anonymous
2011-05-18 22:03:11 UTC
diverge
Fred
2011-05-18 20:47:40 UTC
Terminology.



The sequence converges:

lim[n→∞] n/√(n^3 + 2) = 0

TychaBrahe covers this, although she calls this a series.



The series diverges:

∑[n=1,∞] n/√(n^3 + 2) = ∞

Barney Good demonstrates this quite thoroughly.



Because you didn't include a summation sign in your question, it could be taken either way, if you erred in calling a sequence a series.



BTW, Barney, the nth term test *can* show convergence, but only for an alternating series, which, of course, this one is not. So you're right that in this case, that test fails to decide the question.

* * * CAUTION: The claim that this test can show convergence for an alternating series, turns out to be false, as developed below. * * *



EDIT:

@ B.G.: OK, I get that you're not referring to that, but correct me if I'm wrong, say, with a counterexample. If a sequence converges to 0, and if beyond some finite point, its signs alternate, then the series converges. This is called conditional convergence if the absolute series diverges, but it is convergence, nonetheless. That is, the limit of the sequence of partial sums converges, and no further tests are needed in that case.



BTW, BG, I want to give a shout out to the clowns who've given you 3 TD's, because your answer is absolutely solid, and instructive. Shame on those clowns!



EDIT2:

BG: An infinite sequence can be any infinite list of numbers, and can do wild things before settling down to a pattern, or it can just never settle into a pattern. You could take a well-behaved I.S. and append n terms of any kind to the beginning, and you'd have a new I.S. You could also take that well-behaved I.S. and stick funny terms into it every so often, forever, and that would be another legitimate I.S. So when arriving at rules for finding the behavior of a series, your rules have to be able to handle anything that might come along. Your condition 2 is meant to do that -- I believe you meant "the absolute sequence (that is, the sequence of absolute values) has to be monotonic decreasing." So if there's a way to violate condition 2 while still satisfying condition 1, then that should form the basis for a counterexample to what I said. It's been too long since my course in Infinite Series, so I'll have to do some reviewing.



But as for settling down only after some finite point -- the tests for overall behavior of an infinite sequence or series still apply, as long as you can 'quarantine' a finite number of unruly terms in such a way that the remaining ones satisfy the test. That is sufficient to draw the conclusion from the test.



Also, I see that you could strengthen your main answer by revising the first comparison in your chain:

n/√(n^3 + 2) ≥ n/(3n^3) = 1/(3√n) > 1/3n --> diverges.



[Actually, I see that nyc_kid has done this.]



EDIT3:

BG: I see what can 'cook' my statement for convergence for an alternating series, by way of a counterexample. If you let all the odd terms be

a[n] = 1/n; n = 1, 3, 5, ...

and all the even terms be

a[n] = -1/2^n; n = 2, 4, 6, ...

so overall it goes:

1 - 1/4 + 1/3 - 1/16 + 1/5 - 1/64 + - ...

then

∑[j=1,∞] a[n] = ∑[j=1,∞] 1/(2j - 1) - ∑[j=1,∞] 1/4^j

and the second sum converges to 1/3, while the first sum diverges (by comparison with (1/2)*(1/n), for n=1,2,3...; 1 > 1/2, 1/3 > 1/4, 1/5 > 1/6, ...), making the overall sum diverge (∞ - 1/3 = ∞). The Leibniz test's 2nd requirement ("L2"), namely, that the absolute sequence be monotonic decreasing (forever starting from some finite point), smokes this out and rejects it. That is, L2 fails to decide [con/di]vergence. (Since L2 is a sufficient, but not necessary condition for convergence, because we could have made the odd terms = 1/n^2, and it would still fail L2, but would converge to:

(π^2)/8 - 1/3 )
Barney Good
2011-05-18 20:33:12 UTC
It is divergent. Here is how you know:



edit:

n/sqrt(n^3 + 2) >= n/sqrt(3n^3)

Do this because the +2 prevents simplification (Note: You could multiply the denominator by any (positive) integer here and it would work just as well. There's no reason you couldn't do n/sqrt(5n^3), for example.)

You can simplify n/sqrt(3n^3) = n/(3n)^3/2 = 1/sqrt(3n)



Compare this to 1/3n, which is known to be divergent because it is a nonzero multiple of the harmonic series.



1/sqrt(3n) > 1/3n, therefore the series n/sqrt(3n^3) diverges. Because the original series is greater than this one, it is also divergent.



Hope this helped.





note: While it is true that the limit as n > infinity of 1/sqrt(n) = 0, that does not allow you to conclude convergence (as in the answer above). The nth term test can ONLY conclude divergence, never convergence. (If you aren't familiar with it: if the limit as n > infinity of the series is NOT 0, the series is divergent. If it IS zero you have to use other tests to determine if it converges or diverges: ie the test fails)

edit: Yes, it can be used for alternating series, but in my courses we called that the alternating series test and I am not referring to that here. There are other conditions that must be met in that case.

@Fred: There are two conditions for Leibniz's alternating series test: 1) the limit as n>infinity must be zero, 2) the series must be monotonic decreasing [Edit: Yes, I meant the series of the absolute values here]. I'm not entirely sure about your example; how would the signs only begin alternating at that point? I'm not a mathematician, I'm only a second year who got an A+ in this stuff last term.
TychaBrahe
2011-05-18 20:24:30 UTC
When something confuses me, I try to think of a similar version.



Does the series x^2/x converge or diverge? It diverges, because it reduces to x.



Does the series x/x^2 converge or diverge? It converges, because it reduces to 1/x.



sqrt(n^3) = n * sqrt(n), which is bigger than n, so the series converges, because the bottom is going to get bigger faster than the top.
ted s
2011-05-18 20:23:10 UTC
diverges....greater than Σ 1 / n...or it ' looks like ' Σ 1 / √ n for large n


This content was originally posted on Y! Answers, a Q&A website that shut down in 2021.
Loading...