Does #a_n=lnn/(n^2+1) # converge?

1 Answer
Dec 13, 2015

Yes. #lim_(n->oo)(ln(n))/(n^2+1) = 0#

Explanation:

Intuitively, we should know that the given sequence converges to #0# as any form of polynomial growth (quadratic in this case) is significantly faster than logarithmic growth. This should make sense as being the inverse of an exponential function, a logarithmic function must increase very slowly.

However, while this gives us an idea of what the result should be, it is not a proof. Let's show the result more formally. To do so, we will look at a function with the same end behavior as #a_n#.

Let #f(x)= ln(x)/(x^2 + 1)#.

By the construction of #f(x)#, it should be clear that if #f(x)# converges as #x->oo#, then #a_n# converges as #n->oo#, and #lim_(n->oo)a_n = lim_(x->oo)f(x)#.
(Note that this can also be proven formally with an #epsilon-delta# proof, but such detail is likely unnecessary for a student of calculus).

Now, all that remains is to show that #lim_(x->oo)f(x) = 0#.

As #lim_(x->oo)ln(x) = lim_(x->oo)(x^2+1) = oo#, we may apply L'hospital's rule to #f(x)# as an #oo/oo# case.

#lim_(x->oo)ln(x)/(x^2+1) = lim_(x->oo)(d/dxln(x))/(d/dx(x^2+1))#

#=lim_(x->oo)(1/x)/(2x)#

#=lim_(x->oo)1/(2x^2)#

#= 1/oo#

#=0#

And so

#lim_(n->oo)a_n = lim_(x->oo)f(x) = 0#