Confusing Trig/rational Integral

AI Thread Summary
The discussion centers on the integral of the expression involving sine and cosine functions, specifically \int (sin(t)-cos(t)) \sqrt{cos^2(t)-sin^2(t)} dt. Participants explore the use of trigonometric identities, particularly cos^2(t) - sin^2(t) = cos(2t), to simplify the integral. There is a focus on integration by parts and substitutions, with one user suggesting that using u = cos(t) could simplify the process. The conversation reveals some confusion about the correct identities and substitutions, but ultimately leads to a more manageable form of the integral. The exchange emphasizes the importance of choosing the right substitution and understanding the derivatives involved in the integration process.
Chaz706
Messages
13
Reaction score
0
\int (sin(t)-cos(t)) \sqrt{cos^2(t)-sin^2(t)} dt

Is there a trig idendity I can use? I've distributed that root to both terms to get:
\int sin(t) \sqrt{cos^2(t)-sin^2(t)} dt - \int cos(t) \sqrt{cos^2(t)-sin^2(t)}

If I take one of the terms and integrate by parts, I'm trying to put u=\sqrt{cos^2(t)-sin^2(t)} and dv= sin(t) or dv= cos(t) but that ugly root's derivative appears inside the \int vdu part.

Is there a trig identity I'm missing, or some other tactic I could use? or could this just be really really ugly math?
 
Physics news on Phys.org
Its either

cos^2(t) - sin^2(t) = cos(2t)**

or

sin^2(t) - cos^2(t) = cos(2t)
 
Whozum:
cos^2(t) - sin^2(t) = cos(2t) is the correct identity (and thanks! It may just help!)

Your other identity: sin^2(t) - cos^2(t) = cos(2t) is erroneous, but not by much. sin^2(t) - cos^2(t) yields -cos(2t)

This could actually help as well, however. Thank you for posting both (even if one was inaccurate).
 
You needn't that identity.

\int \left(\sin t-\cos t\right)\sqrt{\cos^{2}t-\sin^{2}t} \ dt = \int \sin t\sqrt{2\cos^{2}t-1} \ dt -\int \cos t\sqrt{1-2\sin^{2}t} \ dt

Can u take it from here (HINT:obvious substitutions necessary) ?

Daniel.
 
That would have been much easier dex, I think...
 
Well it was one of the two, I justw anst sure which one. dex's way is easier though its basically the same thing.
 
The Method is substitution, but then what's u? I'm guessing the basic trig function outside the root is du.

But when I derive the function within that root, I get something different. The derivative of either one ends up to be -4cos(t)sin(t).

Am I choosing the wrong U? I'm integrating with respect to t, and simply inserting the extra parts onto by basic trig function isn't exactly kosher.
 
Okay,let's take the first,i'll leave with the second,which is basically similar (maybe other substitution).

I=\int \sin t\sqrt{2\cos^{2}t-1} \ dt (1)

becomes after the substitution \cos t=u;\sin t \ dt= -du (2)

I=-\int \sqrt{2u^{2}-1} \ du(3)

Make the substitution

\sqrt{2} \ u=\cosh v;du=\frac{1}{\sqrt{2}} \sinh v \ dv(4)

Then

I=-\frac{\sqrt{2}}{2}\int \sinh^{2}v \ dv=-\frac{\sqrt{2}}{4}\int (\cosh 2v-1) \ dv=-\frac{\sqrt{2}}{8}\sinh 2v+\frac{\sqrt{2}}{4}v +C (5)

And now invert the 2 substitutions

I=-\frac{\sqrt{2}}{8}\sinh \left(2 \ \mbox{arg}\cosh\left(\sqrt{2}\cos t\right)\right)+\frac{\sqrt{2}}{4}\mbox{arg}\cosh\left(\sqrt{2}\cos t\right) + C (6)

I'm sure the other one will be simpler.

Daniel.
 
Last edited:
Didn't know that if you took cos(t) = u that you could allow u^2 to simply be cos^2(t).

Of course, I didn't think about that either.
 
  • #10
The other way around is incorrect,of course,but i simply squared.

Daniel.
 
Back
Top