Calculus › Section O · Multiple integrals › Topic O.2

Double integrals and volume

Rectangles were a warm-up. Real regions sit between curves, and their double integrals measure volumes of solids that no single-variable method can reach.

Adding columns over a region

The previous topic evaluated integrals over rectangles, but most interesting regions are not rectangles: triangles, disks, the shapes trapped between two curves. The idea is the same at heart. Chop the region \(R\) into tiny patches of area \(\Delta A\). Above each patch, the surface \(z = f(x, y)\) has a height, so the patch carries a thin column of volume \(f(x, y)\Delta A\). Adding all the columns and letting the patches shrink gives the double integral

\iint_R f(x, y)\,dA

When \(f \ge 0\), this number is the volume of the solid standing on \(R\) with roof the surface. Integrating the constant 1 still returns the region's area.

Regions between curves

A region between two curves has variable inner limits. Suppose a vertical strip through the region enters at \(y = g_1(x)\) and leaves at \(y = g_2(x)\), while the strips themselves run from \(x = a\) to \(x = b\). Then

\iint_R f\,dA = \int_a^b \int_{g_1(x)}^{g_2(x)} f(x, y)\,dy\,dx

The rule of thumb is strict: the inner limits may involve the outer variable, and the outer limits must be constants, because they span the whole region. Reading the limits is a drawing exercise: sketch the region, draw one typical vertical strip, and record where it enters and where it leaves.

Worked example: volume between two parabolas

Find the volume under \(z = xy\) above the region between \(y = x^2\) and \(y = x\). The curves meet where \(x^2 = x\), at \(x = 0\) and \(x = 1\), and between them \(x^2\) is the lower curve. A vertical strip enters at \(y = x^2\) and leaves at \(y = x\), so

V = \int_0^1 \int_{x^2}^{x} xy\,dy\,dx = \int_0^1 x\Big[\tfrac{y^2}{2}\Big]_{x^2}^{x} dx = \int_0^1 \frac{x^3 - x^5}{2}\,dx = \frac{1}{2}\Big(\frac{1}{4} - \frac{1}{6}\Big) = \frac{1}{24}

Re-solving with horizontal strips, \(\int_0^1 \int_y^{\sqrt{y}} xy\,dx\,dy\), gives \(\int_0^1 \frac{y^2 - y^3}{2}\,dy = \frac{1}{24}\) as well: the two orders agree, as they must.

Horizontally simple regions

Some regions read more naturally with horizontal strips. If a horizontal strip enters at \(x = h_1(y)\) and leaves at \(x = h_2(y)\) while \(y\) runs from \(c\) to \(d\), then

\iint_R f\,dA = \int_c^d \int_{h_1(y)}^{h_2(y)} f(x, y)\,dx\,dy

The same region can often be described both ways, and each description is called simple in that direction. Which to use? Whichever gives the friendlier inner antiderivative, and, as the next card shows, whichever is not actually impossible. When both work, evaluating both is a strong check.

Worked example: a triangle, both ways

Integrate \(f(x, y) = 2x\) over the triangle with vertices \((0, 0)\), \((2, 0)\) and \((2, 2)\). Vertically: a strip at \(x\) enters at \(y = 0\) and leaves at the hypotenuse \(y = x\), with \(0 \le x \le 2\):

\int_0^2 \int_0^x 2x\,dy\,dx = \int_0^2 2x^2\,dx = \frac{16}{3}

Horizontally: a strip at \(y\) runs from the hypotenuse \(x = y\) to the edge \(x = 2\), with \(0 \le y \le 2\): \(\int_0^2 \int_y^2 2x\,dx\,dy = \int_0^2 (4 - y^2)\,dy = 8 - \frac{8}{3} = \frac{16}{3}\). Both orders agree. Integrating 1 instead of \(2x\) gives the area \(\int_0^2 x\,dx = 2\), exactly half the \(2 \times 2\) bounding square, as a triangle's area should be.

Reversing the order to escape a trap

Some double integrals are impossible as written. Consider

\int_0^1 \int_{\sqrt{y}}^1 e^{x^3}\,dx\,dy

The inner integral needs an antiderivative of \(e^{x^3}\), and none exists in elementary functions. But the region, sketched, is just the area under the parabola \(y = x^2\) for \(0 \le x \le 1\). Reversing the order there:

\int_0^1 \int_0^{x^2} e^{x^3}\,dy\,dx = \int_0^1 x^2 e^{x^3}\,dx = \Big[\tfrac{e^{x^3}}{3}\Big]_0^1 = \frac{e - 1}{3}

The inner integration in \(y\) leaves the integrand untouched and produces exactly the \(x^2\) factor that substitution needs. One stuck problem has become a short one-variable integral.

Properties and a shortcut

Double integrals inherit the familiar rules: constants pull out, sums split, and an integral over a union of non-overlapping regions is the sum of the pieces. These make hard regions manageable by cutting them into friendly pieces. There is also a shortcut worth knowing: if the integrand separates as \(f(x, y) = g(x)h(y)\) and the region is a rectangle \([a, b] \times [c, d]\), the double integral splits into a product

\iint_R g(x)h(y)\,dA = \left(\int_a^b g(x)\,dx\right)\left(\int_c^d h(y)\,dy\right)

For \(\iint xy\,dA\) over \([0, 2] \times [0, 3]\), that is \(\big[\tfrac{x^2}{2}\big]_0^2 \cdot \big[\tfrac{y^2}{2}\big]_0^3 = 2 \cdot \tfrac{9}{2} = 9\), with no iterated integration at all.

Signed volume and applications ahead

The double integral adds signed height over area. Where the surface dips below the \(xy\)-plane, the columns count negatively, so a naive reading of \(\iint f\,dA\) as "volume" can understate the true enclosed volume; split the region where \(f\) changes sign, as with signed area in one variable. The same machinery measures far more than geometry: a density \(\rho(x, y)\) on a flat plate integrates to the plate's mass, and any quantity spread over an area, from rainfall to pollutant concentration, adds up the same way. Those applications get their own topic after polar coordinates.

Common mistakes to avoid

The classic error is putting a variable in the outer limits: the outer integral must span the whole region, so its limits are constants. Second, when reversing the order, the limits must be rebuilt from a sketch; mechanically swapping the two limit pairs without redrawing describes a different region. Third, check which curve is on top: entering at \(y = x^2\) and leaving at \(y = x\) is only correct on \(0 \le x \le 1\), where \(x^2 \le x\). Fourth, do not treat \(x\) as a constant in the outer integral: after the inner integration it is the live variable. Finally, remember that as-written order can be impossible while the reversed order is easy; a stuck antiderivative means redraw, not give up.

Key takeaways

Over a general region, write the double integral by drawing one strip: entry and exit points give the inner limits, which may depend on the outer variable; the outer limits are constants. Volume under \(z = f(x, y)\) above a region between curves is \(\int_a^b \int_{g_1(x)}^{g_2(x)} f\,dy\,dx\), and horizontal strips give the alternative form. Reversing the order, with limits rebuilt from a sketch, can turn an impossible integral into an easy one, and agreement between orders is a check. Constants pull out, unions split, and separable integrands on rectangles factor into products.

Flashcards

Flashcard

Double integral ∬_R f dA

Adds f(x, y) ΔA over the region R; volume when f ≥ 0

Limit of sums of thin columns.

Test your understanding

Question 1 of 5

What is the value of the integral of 2y, with y running from 0 to x and x running from 0 to 1?

Question 2 of 5

What is the area of the region bounded by y = x² and y = x?

Question 3 of 5

What is the volume under z = xy above the region between y = x² and y = x?

Question 4 of 5

The integral of e^(x³), with x running from √y to 1 and y from 0 to 1, cannot be evaluated as written. What is the value of this double integral?

Question 5 of 5

Which integral correctly reverses the order of ∫₀¹ ∫₀ˣ f dy dx?

Standards: Beyond AP

← Previous topicNext topic →