3

When integration limits were from $0$ to $2\pi$ and the function was $\sin x \cdot \cos^2 x$ and also when the function was $\cos x \cdot \sin^2 x$,answer is $0$, so is there a standard formula when limits are $0$ to $2\pi$ and function is product of higher powers of $\sin x$ and $\cos x$? I have searched about it but got no such information..

David K
  • 98,388
ogirkar
  • 2,681
  • 14
  • 27
  • To write math formulas, follow the advice at https://math.stackexchange.com/help/notation and the other pages it links to. – David K Sep 02 '17 at 12:03
  • Also, for goodness' sake, please put better titles on your questions. Look at the "Related" questions to the right. Can you see how they are related to your question? Two ways: (1) they're about integration (like 36,000 other questions on this site); (2) they all have titles in which the word "question" occurs, with very little useful information about the actual question. – David K Sep 02 '17 at 12:18
  • i will take care next time...but can u help me about my question – ogirkar Sep 02 '17 at 13:35

2 Answers2

4

Here's an answer not just for higher powers of $\sin x$ and $\cos x,$ but indeed for any non-negative integer powers of $\sin x$ and $\cos x.$

In this answer, assume at all times that $m$ and $n$ are non-negative integers.

If $m$ and $n$ both are even, then $$ \int_0^{2\pi} \sin^m t \cos^n t\,dt $$ evaluates to a positive real number according to procedures shown later in this answer. In any other case ($m$ odd, $n$ odd, or $m$ and $n$ both odd), $$ \int_0^{2\pi} \sin^m t \cos^n t\,dt = 0. $$


Consider three cases that cover all possibilities:

Case 1: $m$ is odd.

Since $\sin(-t) = -\sin t$ and $m$ is odd, $\sin^m(-t) = -\sin^m t,$ that is, $\sin$ is an odd function. But $\cos(-t) = \cos(t)$ and therefore $\cos^n(-t) = \cos^n t.$ It follows that \begin{align} \sin^m(-t)\cos^n(-t) &= -\sin^m t \cos^n t; \\ \int_{-\pi}^0 \sin^m t \cos^n t\,dt &= - \int_0^\pi \sin^m t \cos^n t\,dt; \\ \int_{-\pi}^\pi \sin^m t \cos^n t\,dt &= 0. \\ \end{align}

Now observe that $\sin(t - 2\pi) = \sin t$ and $\cos(t - 2\pi) = \cos t,$ so $\sin^m (t - 2\pi) \cos^n (t - 2\pi) = \sin^m t \cos^n t$ and \begin{align} \int_0^{2\pi} \sin^m t \cos^n t\,dt & = \int_0^\pi \sin^m t \cos^n t\,dt + \int_\pi^{2\pi} \sin^m t \cos^n t\,dt \\ & = \int_0^\pi \sin^m t \cos^n t\,dt + \int_{-\pi}^0 \sin^m t \cos^n t\,dt \\ &= \int_{-\pi}^{\pi} \sin^m t \cos^n t\,dt. \end{align}

Therefore $$ \int_0^{2\pi} \sin^m t \cos^n t\,dt = 0. $$

Case 2: $n$ is odd.

Since (as we already know) $\sin^m (t - 2\pi) \cos^n (t - 2\pi) = \sin^m t \cos^n t,$ it follows that \begin{align} \int_0^{2\pi} \sin^m t \cos^n t\,dt & = \int_0^{3\pi/2} \sin^m t \cos^n t\,dt + \int_{3\pi/2}^{2\pi} \sin^m t \cos^n t\,dt \\ & = \int_0^{3\pi/2} \sin^m t \cos^n t\,dt + \int_{-\pi/2}^0 \sin^m t \cos^n t\,dt \\ &= \int_{-\pi/2}^{3\pi/2} \sin^m t \cos^n t\,dt \\ &= \int_0^{2\pi} \sin^m \left(t-\frac\pi2\right) \cos^n \left(t-\frac\pi2\right)\,dt \\ &= \int_0^{2\pi} \left(-\cos t\right)^m \sin^n t\,dt \\ &= \pm \int_0^{2\pi} \cos^m t \sin^n t\,dt \\ \end{align} (where the $\pm$ sign depends on whether $m$ is even or odd). But we now have $\sin t$ raised to an odd power, so Case 1 shows that $$ \int_0^{2\pi} \cos^m t \sin^n t\,dt = 0.$$ Therefore $$ \int_0^{2\pi} \sin^m t \cos^n t\,dt = 0.$$

Case 3: $m$ and $n$ are both even.

In this case we can use the identity $\sin^2 t = 1 - \cos^2 t$ to show that $$ \int_0^{2\pi} \sin^m t \cos^n t\,dt = \int_0^{2\pi} (1 - \cos^2 t)^{m/2} \cos^n t\,dt, $$ and since $\frac m2$ is an integer we can expand $(1 - \cos^2 t)^{m/2}$ to get a polynomial in $\cos^2 t.$ Since $n$ is even, $\cos^n t$ is a power of $\cos^2 t$ and therefore the entire integrand $$ (1 - \cos^2 t)^{m/2} \cos^n t $$ is a polynomial in $\cos^2 t.$ We can integrate each term of this polynomial separately; the integral for any non-constant term can be evaluated using the formula $$ \int_0^{2\pi} \cos^{2p} t\,dt = 2\pi \left(\frac{1 \cdot 3 \cdot 5 \cdots(2p-1)}{2 \cdot 4 \cdot 6 \cdots 2p}\right) $$ where $p$ is a positive integer (see Definite integral of even powers of Cosine., integration of $\int_0^{2\pi} cos^{2n}(t)dt$, and the answers to those questions for details). In the case $n=0$ there is also a constant term whose integral (of course) is $\int_0^{2\pi} 1\,dt = 2\pi.$

Example: Integrate $\sin^4 t \cos^2 t.$

We have \begin{align} \int_0^{2\pi} \sin^4 t \cos^2 t \,dt &= \int_0^{2\pi} (1 - \cos^2 t)^2 \cos^2 t \,dt \\ &= \int_0^{2\pi} (\cos^2 t - 2\cos^4 t + \cos^6 t)\,dt \\ &= \int_0^{2\pi} \cos^2 t\,dt - \int_0^{2\pi} 2\cos^4 t\,dt + \int_0^{2\pi} \cos^6 t\,dt \\ &= \int_0^{2\pi} \cos^2 t\,dt - \int_0^{2\pi} 2\cos^4 t\,dt + \int_0^{2\pi} \cos^6 t\,dt \\ &= 2\pi\left(\frac12\right) - 2\left(2\pi\left(\frac{1\cdot3}{2\cdot4}\right)\right) + 2\pi\left(\frac{1\cdot3\cdot5}{2\cdot4\cdot6}\right) \\ &= 2\pi\left(\frac{24 - 2(18) + 15}{48}\right) \\ &= \frac18 \pi. \end{align}

David K
  • 98,388
  • Thanks but I have one doubt..you said that integral will come zero for 3 cases..first two I agree but for third case where you said it will be zero even if m and n both are odd..if m and n both are odd m+n will be even so as per the ans. given by cyclone this will be non zero number... – ogirkar Sep 03 '17 at 18:53
  • @omkarGirkar It was not clear to me why the other answer said the integral would be positive whenever $m+n$ is even. Aside from the argument in this answer, you can try yourself the integrals of $\sin x \cos x$ or $\sin^3 x \cos^3 x$ (or other examples where both powers are odd) and see. – David K Sep 03 '17 at 20:33
  • @ omkar Girkar: I have edited my post, the result by David is correct and more general than what I showed in my post. Maybe you should accept his answer for this question. – Cyclone Sep 04 '17 at 18:33
3

If I understand correctly, you are looking for a statement along the lines of:

$$ \int_0^{2\pi} \sin^n(t)\cos^m(t) \mathrm{d}t =0, $$ if $m,n\in \mathbb{N}$, $m+n$ odd. This is easily seen as follows. Let $m,n\in \mathbb{N}$, $m+n$ odd and $$I=\int_0^{2\pi} \sin^n(t)\cos^m(t) \mathrm{d}t.$$ Then, letting $t=2\pi-u$ and using addition formulae, $$ \int_0^{2\pi} \sin^n(t)\cos^m(t) \mathrm{d}t = -\int_{2\pi}^0\sin^n(2\pi-u) \cos^m(2\pi-u) \mathrm{du} = \int_0^{2\pi} \underbrace{(-1)^{n+m}}_{=-1} \sin^n(u)\cos^m(u)\mathrm{d}u $$ Hence, $I=-I$, which implies $I=0$. The case where $n+m$ even means that both $m,n$ are even or odd. In this case one may efficiently use the "power-reduction" formulae. A general formula could be derived using the identities for $\sin^n(\theta)$ and $\cos^n(\theta)$ given on the linked page, but I must admit I have never seen it and I'm not sure it is in any way useful. One can certainly always calculate the integral in a closed form for given values of $m$ and $n$, e.g. $I(n=2,m=0)=I(n=0,m=2)=\pi$. I hope this answers your question.

EDIT: I incorrectly stated that in the case $m+n$ even, the integral would always be non-zero. I agree that $m=n=1$ is an obvious counter-example. In fact, the integral is always zero if at least one of $m,n$ is odd. Only if they are both even, the integral is non-zero as is shown nicely in David K's post.

Cyclone
  • 1,853
  • That's exactly what I was searching.. thanks.. – ogirkar Sep 02 '17 at 16:47
  • For $m$ and $n$ both even, it is clear the integral is positive, since even powers are never negative and the integrand is not zero a.e. But try some cases in which $m$ and $n$ are both odd, for example $m=n=1,$ $m=n=3,$ or $m=1$ and $n=3$; the results are not as predicted here. – David K Sep 03 '17 at 20:23
  • Note that the power-reduction formulas for even powers have constant terms, but the formulas for odd powers do not. Moreover, for odd powers the power reduction for sine gives sines and for cosines gives cosines. And in general, $\int_0^{2\pi}\sin(ax)\cos(bx),dx=0$ for positive integers $a,b.$ – David K Sep 03 '17 at 20:43
  • I have edited my post, I agree with you that my claim was not entirely correct. Thanks for pointing that out! – Cyclone Sep 04 '17 at 18:32
  • Nevertheless was always a useful answer (though more useful now that it is completely correct), because it is a reminder that the power reduction formulas are also a tool that we can use when evaluating integrals like these. – David K Sep 04 '17 at 20:43