Double integrals in polar coordinates
A disk, a ring or a wedge is awkward to cut into vertical strips, but in polar coordinates it is a rectangle. When you integrate over it in r and θ, what replaces dA = dx dy, and why?
Where you are 6 in Multiple integrals
Why this exists
Yuna and Kofi plan cell sites for a wireless carrier. A panel antenna on a tower serves a wedge of ground with its point at the tower, and the number of people it serves is the integral of a population density over that wedge. People thin out with distance from a town center, so the density is a function of \displaystyle \sqrt{x^2 + y^2}, and the region is bounded by circles and rays.
In rectangular coordinates both are awkward. A ring is not a type I region (the ring 1 \le x^2 + y^2 \le 4 meets the line x = 0 in two segments, as Iterated integrals and Fubini's theorem noted), circles bring square roots into the limits, and \displaystyle \sqrt{x^2 + y^2} in the integrand rarely has an antiderivative in y alone. In Polar coordinates the same wedge is a rectangle, a \le r \le b and \alpha \le \theta \le \beta, and the density depends on r alone. The price is a factor r: equal steps in r and \theta do not cut out equal areas.
Definition
The polar map is P(r, \theta) = (r\cos\theta, r\sin\theta). For r \ge 0 it sends (r, \theta) to the point at distance r from the origin (the pole) in the direction \theta.
A polar rectangle is the image under P of a rectangle Q = [a, b] \times [\alpha, \beta] of the r\theta-plane, with 0 \le a < b and 0 < \beta - \alpha \le 2\pi:
It is a piece of a ring cut off by two rays; with a = 0 it is a sector, and with \beta - \alpha = 2\pi a whole ring or disk.
A region D is radially simple if
where 0 < \beta - \alpha \le 2\pi and h_1, h_2 are continuous on [\alpha, \beta] with 0 \le h_1 \le h_2. Each ray from the pole at an angle \theta in [\alpha, \beta] meets D in one segment, from r = h_1(\theta) out to r = h_2(\theta). We assume, as holds for every region in this lesson, that D is admissible in the sense of Double integrals as volume (its boundary is made of finitely many graphs of continuous functions), so \iint_D f\,dA exists for continuous f. The matching set of the r\theta-plane,
is a type I region with \theta as the outer variable, and P maps E onto D.
Intuition
Cut [a, b] into steps \Delta r and [\alpha, \beta] into steps \Delta\theta. The circles and rays cut the polar rectangle into cells (Figure 1). A cell at distance about r from the pole has two straight sides of length \Delta r and two arcs of length about r\,\Delta\theta, so it is nearly a rectangle of area r\,\Delta r\,\Delta\theta. Cells far from the pole are larger than cells near it, though they come from rectangles of the same size in the r\theta-plane. The factor r records how much P stretches area.
A Riemann sum over the cells is then \sum f(P(r_i^*, \theta_j^*))\,r_i^*\,\Delta r\,\Delta\theta, a Riemann sum for f(r\cos\theta, r\sin\theta)\,r on the rectangle Q, and its limit is an iterated integral in r and \theta. The units agree: with r in kilometers and \theta in radians, which carry no unit, r\,dr\,d\theta is in square kilometers.
Results
Lemma 1 (area of a polar rectangle). The polar rectangle a \le r \le b, \alpha \le \theta \le \beta has area
where \Delta r = b - a, \Delta\theta = \beta - \alpha and \displaystyle \bar{r} = \frac{a + b}{2} is the middle radius.
Theorem 2 (double integrals in polar coordinates). Let D be radially simple and let f be continuous on D. Then
On a polar rectangle the limits are constants, and the order may be reversed to \int_a^b \int_{\alpha}^{\beta} \cdots\,r\,d\theta\,dr by Fubini’s theorem (Theorem 3 of Iterated integrals and Fubini's theorem) applied to the continuous function f(r\cos\theta, r\sin\theta)\,r on Q. A region c \le r \le d, g_1(r) \le \theta \le g_2(r) gives \int_c^d \int_{g_1(r)}^{g_2(r)} \cdots\,r\,d\theta\,dr; the proof below, with rows of cells in place of columns, proves it.
Corollary 3 (area). With f = 1,
the formula of Area in polar coordinates, which there came from thin sectors.
Why it is true
Lemma 1. S is the sector of radius b and angle \beta - \alpha with the sector of radius a and the same angle removed. A sector of radius \rho and angle \varphi has area \displaystyle \frac12 \rho^2 \varphi (Sectors, segments and their areas), so \displaystyle A(S) = \frac12(b^2 - a^2)(\beta - \alpha), and b^2 - a^2 = (b + a)(b - a) = 2\bar{r}\,\Delta r. \blacksquare
Theorem 2. Besides Results 1 and 3 of Double integrals as volume, the proof uses uniform continuity, as in Iterated integrals and Fubini's theorem: a function continuous on a closed bounded set varies by less than any \varepsilon > 0 between points closer than some \delta > 0. It also uses a distance bound for P. For 0 \le r, r' \le b,
since going from P(r, \theta) to P(r', \theta) along the ray covers \lvert r - r' \rvert, and P(r', \theta) to P(r', \theta') is a chord of the circle of radius r', no longer than its arc.
Let b = \max h_2 and Q = [0, b] \times [\alpha, \beta], whose image S = P(Q) is a sector containing D. Let G(r, \theta) = f(r\cos\theta, r\sin\theta)\,r on E and G = 0 on the rest of Q, and let f^D be f on D and 0 on the rest of S. With M = \max_D \lvert f \rvert, both are bounded (by Mb and M) and continuous except on finitely many graphs, so both are integrable (Result 1), \iint_S f^D\,dA = \iint_D f\,dA, and \iint_Q G\,dA = \iint_E G\,dA, which is the iterated integral of the theorem by Theorem 4 of Iterated integrals and Fubini's theorem. It is enough to show \iint_S f^D\,dA = \iint_Q G\,dA.
Cut [0, b] and [\alpha, \beta] into n equal pieces, \displaystyle \Delta r = \frac{b}{n} and \displaystyle \Delta\theta = \frac{\beta - \alpha}{n}, giving cells Q_{ij} of Q and polar cells S_{ij} = P(Q_{ij}). Let r_i^* be the middle radius of Q_{ij}, \theta_j^* any angle in it, and p_{ij} = P(r_i^*, \theta_j^*). Since \beta - \alpha \le 2\pi, the polar cells cover S and two of them share only boundary points, so additivity (Result 3) and Lemma 1 give
Compare this sum term by term with the Riemann sum \Sigma_n = \sum_{i,j} G(r_i^*, \theta_j^*)\,\Delta r\,\Delta\theta, which tends to \iint_Q G\,dA as n \to \infty. There are three kinds of cell.
- Outside (Q_{ij} misses E). Both terms are 0: a point of S_{ij} can lie in D only if it has a second preimage in Q, which happens only at the pole or on the ray \theta = \alpha when \beta - \alpha = 2\pi, a set of zero area.
- Inside (Q_{ij} \subseteq E). Here f^D = f on S_{ij}. Given \varepsilon > 0, take \delta for f on D. Once \Delta r + b\,\Delta\theta < \delta, every point of S_{ij} is within \delta of p_{ij}, so \lvert f - f(p_{ij}) \rvert < \varepsilon on S_{ij}, and the bounds of Result 3 give \left\lvert \iint_{S_{ij}} f\,dA - f(p_{ij})\,A(S_{ij}) \right\rvert \le \varepsilon\,A(S_{ij}), where f(p_{ij})\,A(S_{ij}) = G(r_i^*, \theta_j^*)\,\Delta r\,\Delta\theta.
- Boundary (the rest). Each term is at most Mb\,\Delta r\,\Delta\theta in size, since A(S_{ij}) \le b\,\Delta r\,\Delta\theta. Such a cell holds a point of E and a point outside it; along the segment joining them, r - h_1(\theta) or h_2(\theta) - r changes sign, so by the intermediate value theorem the cell meets the graph r = h_1(\theta) or r = h_2(\theta). Let \omega_n be the largest change of h_1 or h_2 over an interval of length \Delta\theta; \omega_n \to 0 by uniform continuity. Over one column of cells each graph stays in an interval of r of length at most \omega_n, so it meets at most \displaystyle \frac{\omega_n}{\Delta r} + 2 cells of that column. There are at most \displaystyle 2n\left(\frac{\omega_n}{\Delta r} + 2\right) boundary cells.
Adding up, with n\,\Delta\theta = \beta - \alpha,
As n \to \infty the last term tends to 0 and \Sigma_n \to \iint_Q G\,dA, so \left\lvert \iint_S f^D\,dA - \iint_Q G\,dA \right\rvert \le \varepsilon\,A(S) for every \varepsilon > 0, and the two are equal. \blacksquare
Corollary 3. The inner integral of r from h_1(\theta) to h_2(\theta) is \displaystyle \frac12\left(h_2(\theta)^2 - h_1(\theta)^2\right), and \iint_D 1\,dA = A(D). \blacksquare
Worked examples
A ring
Evaluate \iint_D x^2\,dA over the ring D = \{4 \le x^2 + y^2 \le 16\}.
SolutionThe ring is the polar rectangle 2 \le r \le 4, 0 \le \theta \le 2\pi. In rectangular coordinates it is not type I, so it would have to be cut into four pieces. In polar coordinates x^2 = r^2\cos^2\theta, and with the factor r the integrand is r^3\cos^2\theta, a function of r times a function of \theta. By Theorem 2, and Result 4 of Double integrals as volume used in the r\theta-plane,
Replacing \cos^2\theta by \sin^2\theta, whose integral over a full turn is also \pi, gives \iint_D y^2\,dA = 60\pi. So \iint_D (x^2 + y^2)\,dA = 120\pi, which agrees with \int_0^{2\pi} \int_2^4 r^3\,dr\,d\theta = 2\pi \cdot 60.
Yuna’s sector antenna
Yuna’s tower stands at the center of a town, at the origin, with distances in kilometers. One panel antenna serves the wedge \displaystyle -\frac{\pi}{3} \le \theta \le \frac{\pi}{3} out to 6 kilometers, and the population density is
people per square kilometer. How many people does the antenna serve, and within what distance of the tower do half of them live?
SolutionThe wedge is the polar rectangle 0 \le r \le 6, \displaystyle -\frac{\pi}{3} \le \theta \le \frac{\pi}{3}, and on it \displaystyle \delta = \frac{3000}{1 + r^2}. By Theorem 2 the number served is
since \displaystyle \int_0^6 \frac{r}{1 + r^2}\,dr = \frac12 \ln 37 by the substitution u = 1 + r^2. The factor r is what made the antiderivative elementary.
Out to a distance \rho the same computation gives 1000\pi \ln(1 + \rho^2) people. This is half of N when \displaystyle \ln(1 + \rho^2) = \frac12 \ln 37, that is \displaystyle 1 + \rho^2 = \sqrt{37}, so \displaystyle \rho = \sqrt{\sqrt{37} - 1} \approx 2.25 kilometers. Half the people live in a wedge with \displaystyle \left(\frac{\rho}{6}\right)^2 \approx 0.14 of the area.
The tower on the edge of town
Kofi’s second site is a town that is a disk of radius 2 kilometers, with the tower on its edge. If the residents are spread evenly, what is their average distance from the tower?
SolutionPut the tower at the origin and the town’s center at (0, 2), so the town is x^2 + (y - 2)^2 \le 4, that is x^2 + y^2 \le 4y. In polar coordinates this reads r^2 \le 4r\sin\theta, so r \le 4\sin\theta (or r = 0, the pole). Since 4\sin\theta \ge 0 exactly for 0 \le \theta \le \pi, the town is radially simple with h_1 = 0 and h_2 = 4\sin\theta (Figure 2).
The average distance is \displaystyle \frac{1}{A(D)} \iint_D \sqrt{x^2 + y^2}\,dA, the average value over the town, and A(D) = 4\pi. In polar coordinates the integrand is r, so with the factor r,
using \sin^3\theta = (1 - \cos^2\theta)\sin\theta and u = \cos\theta. The average distance is \displaystyle \frac{256/9}{4\pi} = \frac{64}{9\pi} \approx 2.26 kilometers, a little more than the 2 kilometers from the tower to the town’s center.
A region cut off by a line
Find the centroid of the circular segment D inside the circle x^2 + y^2 = 4 and to the right of the line \displaystyle x = \sqrt{2}.
SolutionIn polar form the line is \displaystyle r\cos\theta = \sqrt{2}, that is \displaystyle r = \sqrt{2}\sec\theta, and it meets the circle r = 2 where \displaystyle \cos\theta = \frac{\sqrt{2}}{2}, at \displaystyle \theta = \pm\frac{\pi}{4} (Figure 3). A ray at an angle in that range enters D at the line and leaves at the circle, so D is radially simple with \displaystyle h_1 = \sqrt{2}\sec\theta and h_2 = 2. By Corollary 3,
For the moment about the y-axis, x = r\cos\theta, and \displaystyle \left(\sqrt{2}\sec\theta\right)^3 = 2\sqrt{2}\sec^3\theta:
The integrand r^2\sin\theta of M_x is odd in \theta on an interval symmetric about 0, so M_x = 0. The centroid is \displaystyle \left(\frac{4\sqrt{2}}{3(\pi - 2)}, 0\right) \approx (1.652, 0), between \displaystyle \sqrt{2} \approx 1.414 and 2 and nearer the chord, where the segment is tallest.
Converting an iterated integral
Evaluate \displaystyle \int_0^2 \int_x^{\sqrt{8 - x^2}} \frac{1}{1 + x^2 + y^2}\,dy\,dx.
SolutionThe limits describe the region with 0 \le x \le 2 between the line y = x and the circle x^2 + y^2 = 8 of radius \displaystyle 2\sqrt{2}, which meet at (2, 2). It is the sector \displaystyle 0 \le r \le 2\sqrt{2}, \displaystyle \frac{\pi}{4} \le \theta \le \frac{\pi}{2}: a point of the sector has x \ge 0, y \ge x because \displaystyle \theta \ge \frac{\pi}{4}, and \displaystyle x \le 2\sqrt{2}\cos\frac{\pi}{4} = 2; conversely a point of the region has y \ge x \ge 0 and \displaystyle r \le 2\sqrt{2}. The integrand is \displaystyle \frac{1}{1 + r^2}, so
In the original order the inner integral is
and the outer integral of that is not one to attempt.
The Gaussian integral
Prove that \displaystyle \int_{-\infty}^{\infty} e^{-x^2}\,dx = \sqrt{\pi}.
SolutionThe function e^{-x^2} has no elementary antiderivative, so the value comes from squaring. Let I(a) = \int_{-a}^{a} e^{-x^2}\,dx for a > 0. By Result 4 of Double integrals as volume, with e^{-x^2 - y^2} = e^{-x^2}e^{-y^2},
Let D_\rho be the disk of radius \rho about the origin. By Theorem 2, with u = r^2,
The disk D_a lies inside the square, which lies inside \displaystyle D_{a\sqrt{2}} (Figure 4), and the integrand is positive, so additivity and monotonicity (Result 3) give
Both bounds tend to \pi as a \to \infty, so \displaystyle I(a) \to \sqrt{\pi}. The integrand is even, so \displaystyle J(t) = \int_0^t e^{-x^2}\,dx = \frac12 I(t); it is increasing and bounded by \displaystyle \frac{\sqrt{\pi}}{2}, so \int_0^{\infty} e^{-x^2}\,dx converges (Improper integrals) to \displaystyle \frac{\sqrt{\pi}}{2}, and so does \int_{-\infty}^{0}. Their sum is \displaystyle \sqrt{\pi}. \blacksquare
The substitution \displaystyle x = \frac{t}{\sqrt{2}} turns this into \displaystyle \int_{-\infty}^{\infty} e^{-t^2/2}\,dt = \sqrt{2\pi}, the constant in the density of the The normal distribution.
Boundaries
The factor r is a Jacobian. For the polar map, \det \begin{bmatrix} x_r & x_\theta \\ y_r & y_\theta \end{bmatrix} = \cos\theta \cdot r\cos\theta - (-r\sin\theta)\sin\theta = r. Change of variables and the Jacobian states the general rule, which multiplies the integrand by the absolute value of this determinant; Theorem 2 is the polar case, proved directly.
Each point once. Theorem 2 needs \beta - \alpha \le 2\pi and 0 \le h_1 \le h_2. Without them a region can be counted twice. The circle r = 2\cos\theta bounds a disk of area \pi, drawn once for \displaystyle -\frac{\pi}{2} \le \theta \le \frac{\pi}{2}, but \int_0^{2\pi} \int_0^{2\cos\theta} r\,dr\,d\theta = \int_0^{2\pi} 2\cos^2\theta\,d\theta = 2\pi. For \displaystyle \frac{\pi}{2} < \theta < \frac{3\pi}{2} the upper limit is negative, the hypothesis fails, and \displaystyle \int_0^{h} r\,dr = \frac{h^2}{2} is still positive, so the disk is swept a second time.
Integrands unbounded at the pole. Theorem 2 is for f continuous on D, hence bounded. For an f unbounded near the pole, such as (x^2 + y^2)^{-p/2}, the integral over a disk about the pole is improper: integrate over \varepsilon \le r and let \varepsilon \to 0^+ (Exercise 18). The factor r helps: (x^2 + y^2)^{-1/2} times r is 1.
Other centers and other shapes. A ray from the pole may meet a region in two pieces, as for the ring 1 \le (x - 3)^2 + y^2 \le 4. Then split the region, or put the pole elsewhere: with x = x_0 + r\cos\theta, y = y_0 + r\sin\theta Theorem 2 holds as before, since a translation leaves every Riemann sum unchanged.
Where this goes next
Triple integrals adds a third variable, and Cylindrical and spherical coordinates pairs polar coordinates in the plane with z, giving dV = r\,dz\,dr\,d\theta. Change of variables and the Jacobian puts the factor r in its general setting as a Jacobian. In probability, Joint distributions and The bivariate normal use r\,dr\,d\theta for the distance and direction of a random point.
Exercises
Polar rectangles
- Evaluate \iint_D y^2\,dA over the disk x^2 + y^2 \le 9.
- Evaluate \displaystyle \iint_D \frac{x^2}{x^2 + y^2}\,dA over the ring 1 \le x^2 + y^2 \le 9.
- Evaluate \iint_D xy\,dA over the quarter disk x^2 + y^2 \le 4, x \ge 0, y \ge 0, in polar coordinates, and check the answer in rectangular coordinates.
- Evaluate \iint_D \ln(1 + x^2 + y^2)\,dA over the disk x^2 + y^2 \le 1.
- Find the volume of the solid between the paraboloid z = 9 - x^2 - y^2 and the plane z = 5.
- Find the volume of the solid inside both the sphere x^2 + y^2 + z^2 = 4 and the cylinder x^2 + y^2 = 1.
Regions bounded by polar curves
- Evaluate \iint_D x\,dA over the disk (x - 1)^2 + y^2 \le 1.
- A lamina fills the region inside the Limacons and cardioids r = 1 + \cos\theta, with density \displaystyle \delta(x, y) = \sqrt{x^2 + y^2}. Find its mass.
- Evaluate \iint_D xy\,dA over the petal of the rose r = \sin 2\theta in the first quadrant.
- Let D be the region inside the circle r = 4\cos\theta and outside the circle r = 2\cos\theta. Evaluate \iint_D y^2\,dA.
Converting iterated integrals
- Evaluate \displaystyle \int_{-3}^{3} \int_0^{\sqrt{9 - x^2}} \sqrt{x^2 + y^2}\,dy\,dx.
- Evaluate \displaystyle \int_0^1 \int_y^{\sqrt{2 - y^2}} (x + y)\,dx\,dy.
- Evaluate \displaystyle \int_0^2 \int_0^{\sqrt{2x - x^2}} \sqrt{x^2 + y^2}\,dy\,dx.
- Write \int_0^1 \int_x^1 f(x, y)\,dy\,dx as an iterated integral in polar coordinates.
Applications
- Find the average distance from the center of a disk of radius a to its points.
- Find the centroid of the half ring 1 \le x^2 + y^2 \le 4, y \ge 0.
- Kofi models a town whose population density is \displaystyle 500\left(6 - \sqrt{x^2 + y^2}\right) people per square kilometer within 6 kilometers of its center, and 0 beyond. Find its population.
- For which p > 0 does \displaystyle \lim_{\varepsilon \to 0^+} \iint_{\varepsilon^2 \le x^2 + y^2 \le 1} \frac{dA}{(x^2 + y^2)^{p/2}} exist? Find the limit when it does.
Proofs
- Let g be continuous on [0, a]. Prove that \displaystyle \iint_{x^2 + y^2 \le a^2} g\left(\sqrt{x^2 + y^2}\right) dA = 2\pi\int_0^a r\,g(r)\,dr, and explain how this gives the volume formula of the The shell method when g \ge 0.
- Let f be continuous on the disk x^2 + y^2 \le a^2 with f(-x, -y) = -f(x, y). Prove that \iint f\,dA = 0 over the disk.
- Prove that \displaystyle \int_0^{\infty} e^{-cx^2}\,dx = \frac12\sqrt{\frac{\pi}{c}} for c > 0, and deduce that \displaystyle \int_{-\infty}^{\infty} e^{-(x - \mu)^2/(2\sigma^2)}\,dx = \sigma\sqrt{2\pi} for \sigma > 0.
Answers
- \displaystyle \frac{81\pi}{4}. The integral is \displaystyle \int_0^{2\pi} \sin^2\theta\,d\theta \int_0^3 r^3\,dr = \pi \cdot \frac{81}{4}.
- 4\pi. The integrand is \cos^2\theta, so the integral is \int_0^{2\pi} \cos^2\theta\,d\theta \int_1^3 r\,dr = \pi \cdot 4.
- 2. In polar coordinates, \displaystyle \int_0^{\pi/2} \cos\theta\sin\theta\,d\theta \int_0^2 r^3\,dr = \frac12 \cdot 4. In rectangular coordinates the inner integral \displaystyle \int_0^{\sqrt{4 - x^2}} xy\,dy is \displaystyle \frac{x(4 - x^2)}{2}, and \displaystyle \int_0^2 \frac{x(4 - x^2)}{2}\,dx = \frac12(8 - 4).
- \pi(2\ln 2 - 1). The integral is 2\pi\int_0^1 r\ln(1 + r^2)\,dr, and with u = 1 + r^2 this is \pi\int_1^2 \ln u\,du = \pi\Big[u\ln u - u\Big]_1^2.
- 8\pi. The surfaces meet where 9 - r^2 = 5, on r = 2, and the height is 4 - r^2, so V = 2\pi\int_0^2 (4 - r^2)\,r\,dr = 2\pi(8 - 4).
- \displaystyle \frac{4\pi}{3}\left(8 - 3\sqrt{3}\right). Over the disk r \le 1 the solid runs from \displaystyle z = -\sqrt{4 - r^2} to \displaystyle z = \sqrt{4 - r^2}, so \displaystyle V = 2\pi\int_0^1 2r\sqrt{4 - r^2}\,dr = 4\pi \cdot \frac{8 - 3\sqrt{3}}{3}.
- \pi. The disk is \displaystyle -\frac{\pi}{2} \le \theta \le \frac{\pi}{2}, 0 \le r \le 2\cos\theta, and \displaystyle \int_{-\pi/2}^{\pi/2} \frac{8\cos^3\theta}{3}\cos\theta\,d\theta = \frac83 \cdot \frac{3\pi}{8}. As a check, the centroid of the disk is its center (1, 0) and its area is \pi.
- \displaystyle \frac{5\pi}{3}. M = \int_0^{2\pi} \int_0^{1 + \cos\theta} r^2\,dr\,d\theta, which is \displaystyle \frac13\int_0^{2\pi} (1 + \cos\theta)^3\,d\theta. Expanded, the terms 1, 3\cos\theta, 3\cos^2\theta, \cos^3\theta integrate to 2\pi, 0, 3\pi, 0.
- \displaystyle \frac{1}{15}. The petal is \displaystyle 0 \le \theta \le \frac{\pi}{2}, 0 \le r \le \sin 2\theta. The inner integral of r^3\cos\theta\sin\theta is \displaystyle \frac{\sin^4 2\theta}{4} \cdot \frac{\sin 2\theta}{2}, and \displaystyle \frac18\int_0^{\pi/2} \sin^5 2\theta\,d\theta = \frac{1}{16}\int_0^{\pi} \sin^5 u\,du = \frac{1}{16} \cdot \frac{16}{15}.
- \displaystyle \frac{15\pi}{4}. For \displaystyle -\frac{\pi}{2} \le \theta \le \frac{\pi}{2}, 2\cos\theta \le r \le 4\cos\theta. The inner integral of r^3\sin^2\theta is 60\cos^4\theta\sin^2\theta, and \displaystyle \int_{-\pi/2}^{\pi/2} \cos^4\theta\sin^2\theta\,d\theta = \frac{3\pi}{8} - \frac{5\pi}{16} = \frac{\pi}{16}.
- 9\pi. The region is the upper half disk of radius 3, so the integral is \int_0^{\pi} \int_0^3 r^2\,dr\,d\theta = \pi \cdot 9.
- \displaystyle \frac{2\sqrt{2}}{3}. The line x = y and the circle x^2 + y^2 = 2 meet at (1, 1), and the region is the sector \displaystyle 0 \le \theta \le \frac{\pi}{4}, \displaystyle 0 \le r \le \sqrt{2}. The integral is \displaystyle \int_0^{\pi/4} (\cos\theta + \sin\theta)\,d\theta \int_0^{\sqrt{2}} r^2\,dr = 1 \cdot \frac{2\sqrt{2}}{3}.
- \displaystyle \frac{16}{9}. The curve \displaystyle y = \sqrt{2x - x^2} is the upper half of the circle (x - 1)^2 + y^2 = 1, that is r = 2\cos\theta, so the region is \displaystyle 0 \le \theta \le \frac{\pi}{2}, 0 \le r \le 2\cos\theta, and the integral is \displaystyle \int_0^{\pi/2} \frac{8\cos^3\theta}{3}\,d\theta = \frac83 \cdot \frac23.
- \int_{\pi/4}^{\pi/2} \int_0^{\csc\theta} f(r\cos\theta, r\sin\theta)\,r\,dr\,d\theta. The region is the triangle with vertices (0, 0), (0, 1) and (1, 1). Its rays run from \displaystyle \theta = \frac{\pi}{4} (the side y = x) to \displaystyle \theta = \frac{\pi}{2}, and each leaves through the top side y = 1, which is r\sin\theta = 1.
- \displaystyle \frac{2a}{3}. The average is \displaystyle \frac{1}{\pi a^2}\int_0^{2\pi} \int_0^a r^2\,dr\,d\theta = \frac{1}{\pi a^2} \cdot \frac{2\pi a^3}{3}.
- \displaystyle \left(0, \frac{28}{9\pi}\right). The area is \displaystyle \frac{3\pi}{2}, and \displaystyle \iint_D y\,dA = \int_0^{\pi} \sin\theta\,d\theta \int_1^2 r^2\,dr = 2 \cdot \frac73. \bar{x} = 0 by symmetry about the y-axis.
- 36{,}000\pi \approx 113{,}097 people. The population is 2\pi\int_0^6 500(6 - r)\,r\,dr, which is \displaystyle 1000\pi\left[3r^2 - \frac{r^3}{3}\right]_0^6 = 1000\pi(108 - 72).
- The limit exists exactly when p < 2, and equals \displaystyle \frac{2\pi}{2 - p}. The integral is 2\pi\int_{\varepsilon}^1 r^{1 - p}\,dr. For p < 2 it is \displaystyle \frac{2\pi\left(1 - \varepsilon^{2 - p}\right)}{2 - p}, which tends to \displaystyle \frac{2\pi}{2 - p}; for p = 2 it is \displaystyle 2\pi\ln\frac{1}{\varepsilon}, and for p > 2 it is \displaystyle \frac{2\pi\left(\varepsilon^{2 - p} - 1\right)}{p - 2}, and both tend to \infty.
- The disk is the polar rectangle 0 \le r \le a, 0 \le \theta \le 2\pi, on which \displaystyle g\left(\sqrt{x^2 + y^2}\right) = g(r), a continuous function. By Theorem 2 the integral is \int_0^{2\pi} \left(\int_0^a g(r)\,r\,dr\right) d\theta. The inner integral does not depend on \theta, so this is 2\pi\int_0^a r\,g(r)\,dr. If g \ge 0, Result 2 of Double integrals as volume says the integral is the volume of the solid \displaystyle 0 \le z \le g\left(\sqrt{x^2 + y^2}\right) over the disk, which is the solid swept out when the region under z = g(r), 0 \le r \le a, turns about the z-axis. So that solid has volume 2\pi\int_0^a r\,g(r)\,dr, the shell formula.
- By Theorem 2 on the polar rectangle 0 \le r \le a, 0 \le \theta \le 2\pi, the integral is \int_0^{2\pi} H(\theta)\,d\theta with H(\theta) = \int_0^a f(r\cos\theta, r\sin\theta)\,r\,dr, continuous by Lemma 2 of Iterated integrals and Fubini's theorem. Since \cos(\theta + \pi) = -\cos\theta, \sin(\theta + \pi) = -\sin\theta and f(-x, -y) = -f(x, y), the integrand at \theta + \pi is the negative of the integrand at \theta, so H(\theta + \pi) = -H(\theta). Split [0, 2\pi] at \pi and put \theta = \varphi + \pi in the second piece: \int_{\pi}^{2\pi} H(\theta)\,d\theta = \int_0^{\pi} H(\varphi + \pi)\,d\varphi = -\int_0^{\pi} H(\varphi)\,d\varphi. The two pieces cancel.
- For t > 0 the substitution \displaystyle u = \sqrt{c}\,x gives \displaystyle \int_0^t e^{-cx^2}\,dx = \frac{1}{\sqrt{c}}\int_0^{\sqrt{c}\,t} e^{-u^2}\,du. As t \to \infty the right side tends to \displaystyle \frac{1}{\sqrt{c}} \cdot \frac{\sqrt{\pi}}{2} by Example 6, so \displaystyle \int_0^{\infty} e^{-cx^2}\,dx = \frac12\sqrt{\frac{\pi}{c}}. For the second integral, the substitution x = \mu + s on each finite interval shows it equals \int_{-\infty}^{\infty} e^{-s^2/(2\sigma^2)}\,ds. That integrand is even, so with \displaystyle c = \frac{1}{2\sigma^2} the integral is \displaystyle 2 \cdot \frac12\sqrt{2\pi\sigma^2} = \sigma\sqrt{2\pi}.