Integral of 1/(something)

  • Thread starter Jules18
  • Start date
  • #1
102
0
[tex]\int[/tex][tex]\stackrel{du}{\sqrt{u^{2}-4}}[/tex] = ln(u + sqrt.(u2-4))

My answer key makes the jump directly from that integral to that natural log fxn, which would make sense to me if only the "u +" weren't inside the brackets of the ln fxn.
Does anyone know where it comes from?
 
Last edited:

Answers and Replies

  • #2
102
0
I'm sorry about the latex. I edited it but the changes aren't showing up. Hope you can still decipher it.
The latex program is ridiculously glitchy, but I'm sure I'm not the first person to say that.
 
Last edited:
  • #3
1,101
3
Um, that's the correct antiderivative? Have you differentiated it to check? Perhaps you forgot the chain rule.
 
  • #4
324
0
Try trig substitution
 
  • #5
907
2
If [itex]u = \cosh{t}[/itex], then [itex]u^2-1 = \sinh^2{t}[/itex], and [itex]du = \sinh{t} dt[/itex].
 
  • #6
HallsofIvy
Science Advisor
Homework Helper
41,833
964
[tex]\int[/tex][tex]\stackrel{du}{\sqrt{u^{2}-4}}[/tex] = ln(u + sqrt.(u2-4))

My answer key makes the jump directly from that integral to that natural log fxn, which would make sense to me if only the "u +" weren't inside the brackets of the ln fxn.
Does anyone know where it comes from?
Actually, that shouldn't make sense.
[tex]\int \frac{1}{f(x)}dx[/tex]
is NOT, in general, equal to ln(f(x))!

hamster143 has suggested you use u= cosh(t) because [itex]cosh^2(t)- 1= sinh^2(t)[/itex] and you can get rid of the square root.

It is also true that [itex]sec^2(t)- 1= csc^2(t)[/itex] so the substitution u= sec(t) would also work. It might be a little more complicated because then du= sec(t)tan(t)dt.

(This board has a slight problem with its LaTex. When you edit LaTex, always click the "refresh" button on your screen (looks like two little 'circulating' green arrows). That should fix it.)
 
  • #7
102
0
Actually, that shouldn't make sense.
[tex]\int \frac{1}{f(x)}dx[/tex]
is NOT, in general, equal to ln(f(x))!

hamster143 has suggested you use u= cosh(t) because [itex]cosh^2(t)- 1= sinh^2(t)[/itex] and you can get rid of the square root.

It is also true that [itex]sec^2(t)- 1= csc^2(t)[/itex] so the substitution u= sec(t) would also work. It might be a little more complicated because then du= sec(t)tan(t)dt.

(This board has a slight problem with its LaTex. When you edit LaTex, always click the "refresh" button on your screen (looks like two little 'circulating' green arrows). That should fix it.)

Really? the equality I posted is straight from my prof's answer key. ....

Also, I'm not sure I understand what you mean by cosh^2(t). Is that another way of expressing cos^2(ht)?

In that case I thought cos^2(x) + sin^2(x) = 1,
so how could cos^2(x) - 1 = sin^2(x) ? I was thinking it would equal -sin^2(x) , and so I was getting confused because it's under a sqrt. sign so it can't be -ve, so I resorted to the answer key only to find my prof's weird equality ... -_-

(thanks for the LaTex tip)
 
  • #8
907
2
Cosh and sinh are hyperbolic sine and cosine, defined as

[tex]\sinh{t} = (e^t-e^{-t})/2[/tex]
[tex]\cosh{t} = (e^t+e^{-t})/2[/tex]
 
  • #9
343
0
Ya your answer isn't right. Remember that when integrating, derivatives are your best friend to check your work. If I were to take a derivative of that I would get a u + sqrt.(u2-4) on the bottom * whatever the derivative of that is on the top.

You need to use a trig substitution here of some sort.
 
  • #10
144
0
woahhhh no youre answer is definetly right.

[tex]\frac{d}{dx}(ln(x+\sqrt{x^{2}-4})=\frac{1}{x+\sqrt{x^{2}-4}}*(1+\frac{x}{\sqrt{x^{2}-4}})=\frac{1}{x+\sqrt{x^{2}-4}}*\frac{x+\sqrt{x^{2}-4}}{\sqrt{x^{2}-4}}=\frac{1}{\sqrt{x^{2}-4}}[/tex]
 
  • #11
144
0
use x=2*sec(u) and dx=2*sec(u)*tan(u)

so the problem would become something like [tex]\int \frac{2*\sec u *\tan u}{2*\sqrt{(\sec x)^{2}-1}}du[/tex]

you'll get [tex]\int \sec u[/tex]
 

Related Threads on Integral of 1/(something)

Replies
2
Views
1K
  • Last Post
Replies
2
Views
16K
Replies
11
Views
2K
  • Last Post
Replies
3
Views
2K
  • Last Post
Replies
3
Views
9K
  • Last Post
Replies
2
Views
5K
  • Last Post
Replies
5
Views
2K
  • Last Post
Replies
3
Views
1K
Replies
5
Views
991
Replies
9
Views
2K
Top