PDA

View Full Version : deciding if the series converges or diverges


Lord Dark
Jan2-10, 09:27 AM
1. The problem statement, all variables and given/known data
Hi everyone ,, got the following question

Sum(k=1,infinity) (2009+5^(-k^3))/(k^2(1+e^(-k^2)))


2. Relevant equations



3. The attempt at a solution

first I think it diverges so I did a comparison which is :
ak = (2009+5^(-k^3))/(k^2(1+e^(-k^2)))
e^(k^2))/(5^(k^3)*k^2) which is less than ak ,,

but now what should I do ?? which test should I use to make it diverges (that is if it diverges)

note : I tried already ratio test and it didn't work and k-th term but I get (infinity over infinity) so it's no use ,, can someone help please

HallsofIvy
Jan2-10, 09:34 AM
It should be easy to see that
\frac{2009+ 5^{-k^3}}{1+ e^{-k^2}}< 2010
compare this series to \sum 2010/k^2.

Thanks, jgens, for catching the typo.

Lord Dark
Jan2-10, 09:39 AM
lol ,, first of all ,, can you explain it a little further ,, how can you choose a number where the function is less than it ( i mean 20010) how did you know that number ??

and why did you remove 5^(-k^3) and e^(-k^2) ?? do they cancel each other or what ?

jgens
Jan2-10, 01:10 PM
Halls mistyped and probably meant that

\frac{2009 + 5^{-k^3}}{1 + e^{-k^2}} < 2010

This inequality should be obvious since 5^{-k^3} < 1 which implies that 2009 + 5^{-k^3} < 2010. Now, because 1 + e^{-k^2} > 1 we know that

\frac{2009 + 5^{-k^3}}{1 + e^{-k^2}} < 2010

Now, use the comparrison test as Halls suggested.

Lord Dark
Jan3-10, 09:02 AM
aha ,, now I got you ... thanks :)