another interesting limit

Darya

Junior Member
Joined
Jan 17, 2020
Messages
154
I'm having trouble with the following:
[MATH] \lim_{x \to 1} (x-2x^{1/100}+1)/(x^{1/2}-2x^{1/100}+1) [/MATH]Please help with a hint. I tried substitution, rewriting it as (x-x^1/100)+(1--x^1/100) and so on in order to use A^2-B^2 formula but nothing works... :(

So sorry! Edit: it is not allowed to use LHopitals rule
 
Last edited:
I'm having trouble with the following:
[MATH] \lim_{x \to 1} (x-2x^{1/100}+1)/(x^{1/2}-2x^{1/100}+1) [/MATH]Please help with a hint. I tried substitution, rewriting it as (x-x^1/100)+(1--x^1/100) and so on in order to use A^2-B^2 formula but nothing works... :(
Have you tried L'Hopital's method?
 
If Jomo's method in post #6 works, use it. I couldn't find a multiplier that accomplishes what he suggests.

Two things that make things a bit easier I think.

i) [MATH]u = x^{1/100} \Rightarrow \dfrac{u^{100}-2u+1}{u^{50}-2u+1}[/MATH]
[MATH]x \to 1 \Rightarrow u \to 1[/MATH]
ii) [MATH]\dfrac{u^{100}-2u+1}{u^{50}-2u+1} = (u^{50}+2 u-1) + \dfrac{4 u^2-6 u+2}{u^{50} - 2 u + 1}[/MATH]
[MATH]\lim \limits_{u\to 1} \left(u^{50}+2 u-1 + \dfrac{4 u^2-6 u+2}{u^{50} - 2 u + 1}\right) = 2 + \lim \limits_{u\to1} \dfrac{4 u^2-6 u+2}{u^{50} - 2 u + 1}[/MATH]
You still have a limit to find by some means but it's a bit less unwieldy.
 
If Jomo's method in post #6 works, use it. I couldn't find a multiplier that accomplishes what he suggests.

Two things that make things a bit easier I think.

i) [MATH]u = x^{1/100} \Rightarrow \dfrac{u^{100}-2u+1}{u^{50}-2u+1}[/MATH]
[MATH]x \to 1 \Rightarrow u \to 1[/MATH]
ii) [MATH]\dfrac{u^{100}-2u+1}{u^{50}-2u+1} = (u^{50}+2 u-1) + \dfrac{4 u^2-6 u+2}{u^{50} - 2 u + 1}[/MATH]
[MATH]\lim \limits_{u\to 1} \left(u^{50}+2 u-1 + \dfrac{4 u^2-6 u+2}{u^{50} - 2 u + 1}\right) = 2 + \lim \limits_{u\to1} \dfrac{4 u^2-6 u+2}{u^{50} - 2 u + 1}[/MATH]
You still have a limit to find by some means but it's a bit less unwieldy.

Ok the next thing I'd do is a nasty little trick.

You want to basically expand the numerator and denominator as Taylor series about [MATH]u-1[/MATH]We only need the linear terms so we don't really need to use derivatives... we can pretend we're just dividing alot.. :D

[MATH]4u^2-6u+2 = 2(u-1) + 4(u-1)^2 \\ u^{50}-2u + 1 = 48(u-1) + 1225(u-1)^2 + \dots[/MATH]
[MATH]\dfrac{4 u^2-6 u+2}{u^{50} - 2 u + 1} = \dfrac{2(u-1) + 4(u-1)^2}{48(u-1) + 1225(u-1)^2 + \dots}[/MATH]
Now we pull our usual "divide by the highest power of [MATH]x[/MATH]" trick, but here [MATH]x=u-1[/MATH]
we end up with [MATH]\lim \limits_{u\to1} \dfrac{4 u^2-6 u+2}{u^{50} - 2 u + 1} = \dfrac{2}{48}=\dfrac{1}{24}[/MATH]
then adding the 2 from the previous section we have

[MATH]\lim \limits_{u\to 1} \dfrac{u^{100}-2u+1}{u^{50}-2u+1} = 2 + \dfrac{1}{24} = \dfrac{49}{24}[/MATH]
Now of course I have waved a wand to produce the 48 that appears above. You have to figure out how to obtain that.
 
You want to multiply the denominator (and numerator!) by something so that the limit as x->1 of the denominator is not zero. Think conjugate.
Although I liked your idea, but I (like Romsek) could not find a useful conjugate.
 
Although I liked your idea, but I (like Romsek) could not find a useful conjugate.
i also sought a suitable conjugate at first. Since Taylor series (although I'm familiar with them) have not been in the course yet, it's not allowed to use them either :0 I'll text if I come up with something :) Have a nice day, thanks for help!
 
Using post # 7, I'm going to show some steps in factoring the algebraic fraction,
that the limit is being taken of, as an alternate method.

\(\displaystyle \dfrac{2(2u^2 - 3u + 1)}{u^{50} - u^{49} + u^{49} - u^{48} + u^{48} \ - \ ... \ - \ u^3 + u^3 - u^2 + u^2 - 2u + 1} \ =\)

\(\displaystyle \dfrac{2(u - 1)(2u - 1)}{u^{49}(u -1) \ + \ u^{48}(u - 1) \ + \ ... \ + \ u^2(u - 1) \ + \ (u - 1)(u - 1)} \ =\)

\(\displaystyle \dfrac{2(u - 1)(2u - 1)}{(u - 1)(u^{49} + u^{48} \ + \ ... \ + \ u^2 + u^1 - 1)} \ = \ ?\)

In front of these fractions is the limit as u approaches 1. And this partial result is
added to 2 for the full answer.

Darya, can you complete the end of the limit steps of my alternative to arrive at Romsek's answer?
 
Last edited:
Using post # 7, I'm going to show some steps in factoring the algebraic fraction,
that the limit is being taken of, as an alternate method.

\(\displaystyle \dfrac{2(2u^2 - 3u + 1)}{u^{50} - u^{49} + u^{49} - u^{48} + u^{48} \ - \ ... \ - \ u^3 + u^3 - u^2 + u^2 - 2u + 1} \ =\)

\(\displaystyle \dfrac{2(u - 1)(2u - 1)}{u^{49}(u -1) \ + \ u^{48}(u - 1) \ + \ ... \ + \ u^2(u - 1) \ + \ (u - 1)(u - 1)} \ =\)

\(\displaystyle \dfrac{2(u - 1)(2u - 1)}{(u - 1)(u^{49} + u^{48} \ + \ ... \ + \ u^2 + u^1 - 1)} \ = \ ?\)

In front of these fractions is the limit as u approaches 1. And this partial result is
added to 2 for the full answer.

Darya, can you complete the end of the limit steps of my alternative to arrive at Romsek's answer?
Yes! Alternatively, with substitution from post 7 and then long division by (u-1) we get (u^99+u^98+u^97+...+u-1)/((u^49+u^48+u^47+...+u-1), which is equal to 98/48 or 49/24. Thaaaankssss!!!!
 
The key point to all this is that we converted the problem from one where the limit ended up with a bunch of terms that did not go to zero in the limit to a problem where most of the terms did go to zero in the limit. Thus we were sort of able to pluck off the important terms.

Taking a step back this is all basically why L'Hopital's rule works.
 
Top