The Student Room Group

The hard integral thread.

Scroll to see replies

Reply 200
Original post by atsruser
Yes, that's essentially the method that I used. (and PRSOM)

However, I'm beginning to get a little suspicious. You seem to have memorised a whole load of useful results; you can exchange the order of integrals in the blink of an eye; you handle Laplace and Fourier methods with ease ... so ... can the the rumour really be true? That the so-called KUMMER* project is not a myth? That in 1957, the Americans cryogenically preserved the brain of John von Neumann, and have recently managed to "light it up" again? Perhaps we'll never know, but when Kummer posts, what I'm hearing is:

"Hee-ee-e-re's Johnny!!"

(*Kold 'Ungarian Man Must Evaluate Residues - yeah, I know - acronym technology wasn't too impressive in the 1950s)


you will also see my neat detailed version after work this evening
Original post by lovemaths123
can you show your workings?


Not going to write it all out.

Substitution u2=xu^2=x, then by parts.
Reply 202
Original post by atsruser
Yes, that's essentially the method that I used. (and PRSOM)

However, I'm beginning to get a little suspicious. You seem to have memorised a whole load of useful results; you can exchange the order of integrals in the blink of an eye; you handle Laplace and Fourier methods with ease ... so ... can the the rumour really be true? That the so-called KUMMER* project is not a myth? That in 1957, the Americans cryogenically preserved the brain of John von Neumann, and have recently managed to "light it up" again? Perhaps we'll never know, but when Kummer posts, what I'm hearing is:

"Hee-ee-e-re's Johnny!!"

(*Kold 'Ungarian Man Must Evaluate Residues - yeah, I know - acronym technology wasn't too impressive in the 1950s)
:lol: Thanks.
Evaluate 0π2etanx3e3tanxsin2x dx\displaystyle \int_0^{\frac{\pi}{2}} \frac { e^{-\frac{\tan x}{\sqrt{3}}}-e^{-\sqrt{3}\tan x} } {\sin 2x} \ dx
Original post by atsruser
Evaluate 0exn+1e(2n+1xn+1)x dx\displaystyle \int_0^\infty \frac{e^{-x^{n+1}}-e^{-(2^{n+1}x^{n+1})}}{x} \ dx for nNn \in \mathbb{N}


A couple of extra points:

1. For a sensible definition of N={1,2,3,}\mathbb{N} = \{ 1, 2, 3, \cdots \}, we can write this more nicely as:

0exne2nxnx dx\displaystyle \int_0^\infty \frac{e^{-x^n}-e^{-2^nx^n}}{x} \ dx

2. I also note that 0e2nxnx dx\displaystyle \int_0^\infty \frac{e^{-2^nx^n}}{x} \ dx does not converge according to Wolfram. So an extra question is: prove that the first integral above does indeed converge, and that the second one doesn't. (I don't have answers for these, BTW)
Reply 205
and here is my solution (with full details)
Original post by atsruser
Any hints? I don't have a clue. Presumably some series expansion to start?


Spoiler



Original post by Zacken
Not you! I've seen you on TPIT a few years ago and your skills were beastly enough back then. I shudder to think what they are now...


Pft! Nonsense :smile:

Original post by Star-girl
:hi:


Hey! Hope all is well!
(edited 8 years ago)
Reply 207
Original post by atsruser
Evaluate 0π2etanx3e3tanxsin2x dx\displaystyle \int_0^{\frac{\pi}{2}} \frac { e^{-\frac{\tan x}{\sqrt{3}}}-e^{-\sqrt{3}\tan x} } {\sin 2x} \ dx
It's easy to show that 12absec2xeytanxdy=eatanxebtanxsin2x\displaystyle \frac{1}{2}\int_a^b \sec^2{x} e^{-y \tan{x}} \, dy =\frac { e^{-a \tan x}-e^{-b\tan x} } {\sin 2x} . Now,

I=120π/2absec2xeytanxdy  dx=12ab0π/2sec2xeytanxdx  dy=12ab1y  dy=12log(b/a).[br]\begin{aligned}\displaystyle I & = \frac{1}{2} \int_0^{\pi/2} \int_a^b \sec^2{x} e^{-y \tan{x}} \, dy \;{dx} \\& = \frac{1}{2} \int_a^{b} \int_0^{\pi/2} \sec^2{x} e^{-y \tan{x}} \, dx \;{dy} \\& = \frac{1}{2} \int_a^b \frac{1}{y} \; dy \\& = \frac{1}{2} \log(b/a).[br] \end{aligned}
So the answer to your integral is 12log(3).\frac{1}{2} \log(3).
Reply 208
One of mine (undergrad level)
Reply 209
Original post by Kummer
It's easy to show that 12absec2xeytanxdy=eatanxebtanxsin2x\displaystyle \frac{1}{2}\int_a^b \sec^2{x} e^{-y \tan{x}} \, dy =\frac { e^{-a \tan x}-e^{-b\tan x} } {\sin 2x} . Now,

I=120π/2absec2xeytanxdy  dx=12ab0π/2sec2xeytanxdx  dy=12ab1y  dy=12log(b/a).[br]\begin{aligned}\displaystyle I & = \frac{1}{2} \int_0^{\pi/2} \int_a^b \sec^2{x} e^{-y \tan{x}} \, dy \;{dx} \\& = \frac{1}{2} \int_a^{b} \int_0^{\pi/2} \sec^2{x} e^{-y \tan{x}} \, dx \;{dy} \\& = \frac{1}{2} \int_a^b \frac{1}{y} \; dy \\& = \frac{1}{2} \log(b/a).[br] \end{aligned}
So the answer to your integral is 12log(3).\frac{1}{2} \log(3).


obviously ... (what the hell is going on ....?)

I will write my own version which I estimate to be around 3 sides.
Reply 210
Original post by TeeEm
...



I found another version of the Fresnel integrals in my resources

So 3 versions here

by Laplace transforms,
by contour integration
By DUTIS
Reply 211
Original post by atsruser
Evaluate 0π2etanx3e3tanxsin2x dx\displaystyle \int_0^{\frac{\pi}{2}} \frac { e^{-\frac{\tan x}{\sqrt{3}}}-e^{-\sqrt{3}\tan x} } {\sin 2x} \ dx


Practically the same as the previous one/here are 2 ways (one by LAPLACE)

I think I have now provided lengthy detailed solutions (which I hope they are useful to others trying to get a grasp of these) to the last 7 or 8 integrals except to that of Lord of the Flies.
Although I need more practice I have to quit as I am falling behind with my own work schedule.
Original post by Lord of the Flies

Hey! Hope all is well!


Hope all is well with you too. I'll message you at some point soon to catch up (have family round at the moment). :hugs:
Reply 213
Original post by TeeEm
One of mine (undergrad level)
This time we'll use 1x2=0text  dt\displaystyle \frac{1}{x^2} = \int_0^{\infty} te^{-xt}\;{dt} - we have:

I=0exx2(3x+e3x1)  dx=00textx(3x+e3x1)  dt  dx=00textx(3x+e3x1)  dx  dt=09t(1+t)2(4+t)  dt=log(256)3.\begin{aligned} \displaystyle I & = \int_0^{\infty} \frac{e^{-x}}{x^2}\left(3x+e^{-3x}-1\right)\;{dx} = \int_0^{\infty} \int_0^{\infty} te^{-xt-x}\left(3x+e^{-3x}-1\right)\;{dt} \;{dx} \\& =\int_0^{\infty} \int_0^{\infty} te^{-xt-x}\left(3x+e^{-3x}-1\right)\;{dx} \;{dt} = \int_0^{\infty} \frac{9t}{(1+t)^2(4+t)}\;{dt} \\& = \log(256)-3.\end{aligned}
(edited 8 years ago)
Original post by Kummer
This time we'll use 1x2=0text  dt\displaystyle \frac{1}{x^2} = \int_0^{\infty} te^{-xt}\;{dt} - we have:

I=0exx2(3x+e3x1)  dt=00textx(3x+e3x1)  dt  dx=00textx(3x+e3x1)  dx  dt=09t(1+t)2(4+t)=log(256)3.\begin{aligned} \displaystyle I & = \int_0^{\infty} \frac{e^{-x}}{x^2}\left(3x+e^{-3x}-1\right)\;{dt} = \int_0^{\infty} \int_0^{\infty} te^{-xt-x}\left(3x+e^{-3x}-1\right)\;{dt} \;{dx} \\& =\int_0^{\infty} \int_0^{\infty} te^{-xt-x}\left(3x+e^{-3x}-1\right)\;{dx} \;{dt} = \int_0^{\infty} \frac{9t}{(1+t)^2(4+t)} \\& = \log(256)-3.\end{aligned}


I have a notificiation everytime this thread is used, since integration is nicez
But integration should be banned 12-6am.


Posted from TSR Mobile
Reply 215
Original post by physicsmaths
I have a notificiation everytime this thread is used, since integration is nicez
But integration should be banned 12-6am.


Posted from TSR Mobile
Haha, there's nothing like a juicy integral to ponder upon late at night!
Original post by Kummer
This time we'll use 1x2=0text  dt\displaystyle \frac{1}{x^2} = \int_0^{\infty} te^{-xt}\;{dt} - we have:

I=0exx2(3x+e3x1)  dt=00textx(3x+e3x1)  dt  dx=00textx(3x+e3x1)  dx  dt=09t(1+t)2(4+t)=log(256)3.\begin{aligned} \displaystyle I & = \int_0^{\infty} \frac{e^{-x}}{x^2}\left(3x+e^{-3x}-1\right)\;{dt} = \int_0^{\infty} \int_0^{\infty} te^{-xt-x}\left(3x+e^{-3x}-1\right)\;{dt} \;{dx} \\& =\int_0^{\infty} \int_0^{\infty} te^{-xt-x}\left(3x+e^{-3x}-1\right)\;{dx} \;{dt} = \int_0^{\infty} \frac{9t}{(1+t)^2(4+t)} \\& = \log(256)-3.\end{aligned}


That first result, did you just know it? Or did you have to find it/figure it out before employing it. Unless it is a known result in the integral world?
Also your successive lines are a page of work to me lol. I hope you working is not how it is presented on here, it is too good and hard to be that short of a human mind.


Posted from TSR Mobile
Reply 217
Original post by physicsmaths
That first result, did you just know it? Or did you have to find it/figure it out before employing it. Unless it is a known result in the integral world?
Also your successive lines are a page of work to me lol. I hope you working is not how it is presented on here, it is too good and hard to be that short of a human mind.


Posted from TSR Mobile
I constructed the first one; it's not something that I recalled from memory. Its proof is simple. But obviously don't take it the way it's presented :lol: For example, I could have added more details when integrating with respect to x.
Original post by Kummer
I constructed the first one; it's not something that I recalled from memory. Its proof is simple. But obviously don't take it the way it's presented :lol: For example, I could have added more details when integrating with respect to x.


Yh integrate by parts right?


Posted from TSR Mobile
Reply 219
Attachment not found



another undergrad question from my secret personal stash

Quick Reply

Latest