개념 해석 · Concepts

NOTE 일반 영역으로의 확장 Integrating over General Regions절 도입 · General Regions · p.1089–1090
한국어

1변수 적분에서는 적분 구간이 항상 구간 하나로 정해지지만, 이중적분은 직사각형보다 훨씬 다양한 모양의 평면 영역 위에서도 다룰 수 있어야 한다.

$D$를 유계인 영역, 즉 어떤 직사각형 $R$ 안에 완전히 담기는 영역이라 하고, $f$가 $D$에서 정의되어 있다고 하자. $R$ 전체에서 정의된 새 함수 $F$를 $D$ 안에서는 $f$의 값을, $D$ 밖에서는 $0$을 취하도록 정의한다.

$$F(x,y)=\begin{cases} f(x,y) & (x,y)\text{가 }D\text{ 안에 있을 때},\\[2pt] 0 & (x,y)\text{가 }R\text{ 안에 있지만 }D\text{ 밖에 있을 때}.\end{cases}\tag{1}$$

확장함수 $F$는 직사각형 위에서 정의되므로, §15.1에서 이미 직사각형 위의 적분이 무엇인지 알고 있다. 이 관찰이 이 절 전체의 핵심 아이디어이다.

English

For a single integral the region of integration is always an interval. A double integral, however, should be available over plane regions of far more general shape than a rectangle.

Let $D$ be a bounded region — that is, a region that can be enclosed in some rectangle $R$ — and let $f$ be defined on $D$. Define a new function $F$ on all of $R$ by giving it the values of $f$ inside $D$ and the value $0$ everywhere else:

$$F(x,y)=\begin{cases} f(x,y) & \text{if }(x,y)\text{ is in }D,\\[2pt] 0 & \text{if }(x,y)\text{ is in }R\text{ but not in }D.\end{cases}\tag{1}$$

The extension $F$ lives on a rectangle, and §15.1 already says what it means to integrate over a rectangle. That observation is the whole idea of this section.

DEF 2 일반 영역 위의 이중적분 Double Integral over a General RegionDefinition (2) · p.1090
한국어

정의 2 — 일반 영역 위의 이중적분. $R$을 유계 영역 $D$를 포함하는 직사각형이라 하고, $F$를 (1)에서 정의한 $f$의 확장함수라 하자. $F$가 $R$ 위에서 적분가능하면, $f$의 $D$ 위 이중적분을 다음과 같이 정의한다.

$$\iint_D f(x,y)\,dA=\iint_R F(x,y)\,dA.\tag{2}$$

이 정의는 우변이 §15.1에서 이미 정의된, 직사각형 $R$ 위의 이중적분이므로 문제없이 성립한다. 또한 어떤 직사각형을 택하는지에 의존하지 않는다. $F$는 $D$ 밖 $R$의 모든 점에서 $0$이므로 그 점들은 적분값에 아무런 기여를 하지 않고, 따라서 $D$를 포함하는 어떤 직사각형을 잡더라도 같은 값을 얻는다.

English

Definition (2) — Double Integral over a General Region. Let $R$ be a rectangle containing the bounded region $D$, and let $F$ be the extension of $f$ given by (1). If $F$ is integrable over $R$, the double integral of $f$ over $D$ is defined by

$$\iint_D f(x,y)\,dA=\iint_R F(x,y)\,dA.\tag{2}$$

The definition makes sense because the right-hand side was already defined in §15.1, $R$ being a rectangle. It also does not depend on which rectangle is used: $F$ takes the value $0$ at every point of $R$ outside $D$, so those points contribute nothing, and any rectangle containing $D$ produces the same number.

REM 3 부피 해석과 적분가능성 Volume Interpretation and IntegrabilityDefinition 2 뒤 단락 · p.1090–1091
한국어

$D$에서 $f(x,y)\ge 0$이면, $\iint_D f(x,y)\,dA$는 여전히 $D$ 위·곡면 $z=f(x,y)$ 아래에 놓인 입체의 부피이다. $f$의 그래프와 확장함수 $F$의 그래프를 비교해 보면 그 이유가 분명해진다. $D$ 밖에서 $F$의 그래프는 $xy$평면에 납작하게 붙어 부피를 만들지 않으므로, $R$ 위 $F$의 그래프 아래 부피는 정확히 $D$ 위 $f$의 그래프 아래 부피와 같다.

확장의 대가는, $F$가 대개 $D$의 경계점에서 불연속이 된다는 점이다. 경계에서 값이 갑자기 $0$으로 떨어지기 때문이다. 그렇다고 치명적인 것은 아니다. $f$가 $D$에서 연속이고 $D$의 경계곡선이 (이 책에서 엄밀히 규정하지는 않지만) "충분히 좋으면", $\iint_R F(x,y)\,dA$가 존재하고, 따라서 $\iint_D f(x,y)\,dA$도 존재한다. 다음에 소개할 두 종류의 영역이 바로 이러한 경우에 해당한다.

English

If $f(x,y)\ge 0$ on $D$, then $\iint_D f(x,y)\,dA$ is still the volume of the solid lying above $D$ and below the graph $z=f(x,y)$. Comparing the graph of $f$ with the graph of its extension $F$ makes this plain: outside $D$ the graph of $F$ sits flat on the $xy$-plane and encloses no volume, so the volume under the graph of $F$ over $R$ is exactly the volume under the graph of $f$ over $D$.

The price of the extension is that $F$ will usually be discontinuous at the boundary points of $D$, where its values fall abruptly to $0$. That is not fatal: if $f$ is continuous on $D$ and the boundary curve of $D$ is "well behaved" (in a sense not made precise in this book), then $\iint_R F(x,y)\,dA$ exists, and therefore $\iint_D f(x,y)\,dA$ exists as well. The two families of regions described next are of exactly this kind.

DEF 4 제1형 영역 Type I Regiontype I region · Figure 5 · p.1091
한국어

제1형 영역(Type I 영역). 평면 영역 $D$가 $x$의 두 연속함수의 그래프 사이에 놓여 있을 때, 즉

$$D=\{(x,y)\mid a\le x\le b,\ g_1(x)\le y\le g_2(x)\},$$

$g_1,g_2$가 $[a,b]$에서 연속일 때 $D$를 제1형이라 한다.

그러므로 제1형 영역은 좌우로는 수직선 $x=a$, $x=b$로, 아래로는 곡선 $y=g_1(x)$로, 위로는 곡선 $y=g_2(x)$로 잘려 있다.

두 경계함수는 연속이기만 하면 되고, 반드시 하나의 식으로 주어질 필요는 없다. $g_1$이나 $g_2$의 자리에 조각적으로 정의된 연속함수가 와도 무방하다.

x y 0 a b y=g₂(x) y=g₁(x) D
English

type I region. A plane region $D$ is of type I when it is caught between the graphs of two continuous functions of $x$, that is,

$$D=\{(x,y)\mid a\le x\le b,\ g_1(x)\le y\le g_2(x)\},$$

where $g_1$ and $g_2$ are continuous on $[a,b]$.

So a type I region is cut off on the left and right by the vertical lines $x=a$ and $x=b$, below by the curve $y=g_1(x)$, and above by the curve $y=g_2(x)$.

The two boundary functions must be continuous, but neither of them has to be given by a single formula: a continuous piecewise defined function is perfectly allowed for $g_1$ or $g_2$.

x y 0 a b y=g₂(x) y=g₁(x) D
THM 5 제1형 영역의 반복적분 공식 Iterated Integral over a Type I RegionEquation (3) · p.1091
한국어

식 (3) — 제1형 영역의 반복적분 공식. $f$가 다음과 같은 제1형 영역

$$D=\{(x,y)\mid a\le x\le b,\ g_1(x)\le y\le g_2(x)\},$$

$D$ 위에서 연속이면, 다음이 성립한다.

$$\iint_D f(x,y)\,dA=\int_a^b\!\!\int_{g_1(x)}^{g_2(x)} f(x,y)\,dy\,dx.\tag{3}$$

우변은 §15.1에서 다룬 것과 같은 종류의 반복적분이지만, 결정적으로 다른 점이 하나 있다. 안쪽 적분을 계산할 때 $x$는 $f(x,y)$ 안에서뿐 아니라 두 적분 한계 $g_1(x)$, $g_2(x)$ 안에서도 상수로 고정된다는 점이다. 그래서 안쪽 적분은 $x$만의 함수를 만들어 내고, 바깥 적분이 이를 $a$에서 $b$까지 적분한다.

증명. $g_1,g_2$는 $[a,b]$에서 연속이므로 유계이고, 따라서 모든 $x\in[a,b]$에서 $c\le g_1(x)\le g_2(x)\le d$가 되는 수 $c,d$를 잡을 수 있다. $R=[a,b]\times[c,d]$는 $D$를 포함하는 직사각형이므로, 정의 2에 의해

$$\iint_D f(x,y)\,dA=\iint_R F(x,y)\,dA$$

이고, 여기서 $F$는 (1)의 확장함수이다. $F$는 $R$에서 유계이고, $f$가 $D$에서 연속이므로 $F$가 불연속일 수 있는 곳은 $D$의 경계, 곧 두 그래프 $y=g_1(x)$, $y=g_2(x)$와 두 선분 $x=a$, $x=b$뿐이다. 그러므로 연속함수에 대한 판이 아니라, 유계이고 유한 개의 매끄러운 곡선 위에서만 불연속인 함수를 다루는 §15.1 정리 10(푸비니 정리)의 일반형을 적용하면

$$\iint_R F(x,y)\,dA=\int_a^b\int_c^d F(x,y)\,dy\,dx.$$

고정된 $x\in[a,b]$에서 안쪽 적분을 계산하자. (1)에 의해 $c\le y\lt g_1(x)$ 또는 $g_2(x)\lt y\le d$인 점 $(x,y)$는 $D$ 밖에 있으므로 $F(x,y)=0$이고, $g_1(x)\le y\le g_2(x)$이면 $F(x,y)=f(x,y)$이다. 그러므로 안쪽 적분을 세 조각으로 쪼개면 바깥 두 조각이 사라져

$$\int_c^d F(x,y)\,dy=\int_{g_1(x)}^{g_2(x)}f(x,y)\,dy$$

만 남는다. 이를 바깥 적분에 대입하면 식 (3)

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

이 얻어진다.

$\square$
English

Equation (3) — Iterated Integral over a Type I Region. If $f$ is continuous on a type I region $D$ described by

$$D=\{(x,y)\mid a\le x\le b,\ g_1(x)\le y\le g_2(x)\},$$

then

$$\iint_D f(x,y)\,dA=\int_a^b\!\!\int_{g_1(x)}^{g_2(x)} f(x,y)\,dy\,dx.\tag{3}$$

The right-hand side is an iterated integral of the kind met in §15.1, with one crucial difference: while the inner integration is carried out, $x$ is held constant not only inside $f(x,y)$ but also in the two limits of integration $g_1(x)$ and $g_2(x)$. The inner integral therefore produces a function of $x$ alone, which the outer integral then integrates from $a$ to $b$.

Proof. Since $g_1,g_2$ are continuous on $[a,b]$, they are bounded, so there exist numbers $c,d$ with $c\le g_1(x)\le g_2(x)\le d$ for every $x\in[a,b]$. Then $R=[a,b]\times[c,d]$ is a rectangle containing $D$, and Definition 2 gives

$$\iint_D f(x,y)\,dA=\iint_R F(x,y)\,dA,$$

where $F$ is the extension (1). Note that $F$ is bounded on $R$ and, $f$ being continuous on $D$, the only points at which $F$ can fail to be continuous are those of the boundary of $D$: the two graphs $y=g_1(x)$, $y=g_2(x)$ and the two segments $x=a$, $x=b$. What applies is therefore not the version of §15.1 Theorem 10 (Fubini's Theorem) for continuous functions but its general form, for a bounded function discontinuous only on finitely many smooth curves, which turns this into

$$\iint_R F(x,y)\,dA=\int_a^b\int_c^d F(x,y)\,dy\,dx.$$

Fix $x\in[a,b]$ and evaluate the inner integral. By (1), a point $(x,y)$ with $c\le y\lt g_1(x)$ or $g_2(x)\lt y\le d$ lies outside $D$, so $F(x,y)=0$ there, while $F(x,y)=f(x,y)$ for $g_1(x)\le y\le g_2(x)$. Splitting the inner integral into three pieces therefore makes the outer two vanish, leaving

$$\int_c^d F(x,y)\,dy=\int_{g_1(x)}^{g_2(x)}f(x,y)\,dy.$$

Substituting this into the outer integral yields equation (3),

$$\iint_D f(x,y)\,dA=\int_a^b\int_{g_1(x)}^{g_2(x)}f(x,y)\,dy\,dx.$$
$\square$
DEF 6 제2형 영역 Type II Regiontype II region · Figure 7 · p.1092
한국어

제2형 영역(Type II 영역). 평면 영역 $D$를 다음과 같이 나타낼 수 있을 때

$$D=\{(x,y)\mid c\le y\le d,\ h_1(y)\le x\le h_2(y)\},$$

$h_1,h_2$가 연속함수이면 $D$를 제2형이라 한다.

이번에는 두 변수의 역할이 서로 바뀐다. 수평선 $y=c$, $y=d$가 영역의 아래와 위를 잘라내고, 왼쪽과 오른쪽 경계는 $y$의 연속함수의 그래프인 $x=h_1(y)$와 $x=h_2(y)$이다.

x y 0 d c x=h₁(y) x=h₂(y) D
English

type II region. A plane region $D$ is said to be of type II if it can be expressed as

$$D=\{(x,y)\mid c\le y\le d,\ h_1(y)\le x\le h_2(y)\},$$

where $h_1$ and $h_2$ are continuous.

The roles of the two variables are now exchanged: the horizontal lines $y=c$ and $y=d$ cut the region off below and above, while the left and right boundaries are the graphs $x=h_1(y)$ and $x=h_2(y)$ of continuous functions of $y$.

x y 0 d c x=h₁(y) x=h₂(y) D
THM 7 제2형 영역의 반복적분 공식 Iterated Integral over a Type II RegionEquation (4) · p.1092
한국어

식 (4) — 제2형 영역의 반복적분 공식. $f$가 다음과 같은 제2형 영역

$$D=\{(x,y)\mid c\le y\le d,\ h_1(y)\le x\le h_2(y)\},$$

$D$ 위에서 연속이면, 다음이 성립한다.

$$\iint_D f(x,y)\,dA=\int_c^d\!\!\int_{h_1(y)}^{h_2(y)} f(x,y)\,dx\,dy.\tag{4}$$

이번에는 안쪽 적분이 $x$ 방향으로 진행되며 $y$가 상수로 고정된다. 역시 $f(x,y)$ 안에서뿐 아니라 한계 $h_1(y)$, $h_2(y)$ 안에서도 $y$는 고정되고, 바깥 적분이 그 결과를 $y=c$에서 $y=d$까지 모은다.

증명. 식 (3)의 증명에서 $x$와 $y$의 역할만 맞바꾸면 같은 방식으로 얻어진다. $h_1,h_2$가 $[c,d]$에서 연속이므로 유계이고, 모든 $y\in[c,d]$에서 $a\le h_1(y)\le h_2(y)\le b$가 되는 수 $a,b$를 잡아 $R=[a,b]\times[c,d]$라 하면 $D\subset R$이다. 정의 2에 의해

$$\iint_D f(x,y)\,dA=\iint_R F(x,y)\,dA$$

이고, 이번에도 같은 이유로 $F$는 유계이며 유한 개의 곡선 위에서만 불연속이므로, §15.1 정리 10의 일반형을 $x$가 안쪽 변수가 되도록 적용하면

$$\iint_R F(x,y)\,dA=\int_c^d\int_a^b F(x,y)\,dx\,dy$$

이다. 고정된 $y\in[c,d]$에서 $F(x,y)$는 $h_1(y)\le x\le h_2(y)$ 밖에서 $0$이므로 안쪽 적분은

$$\int_a^b F(x,y)\,dx=\int_{h_1(y)}^{h_2(y)}f(x,y)\,dx$$

로 줄어들고, 바깥 적분을 취하면 식 (4)

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

가 얻어진다.

$\square$
English

Equation (4) — Iterated Integral over a Type II Region. If $f$ is continuous on a type II region $D$ described by

$$D=\{(x,y)\mid c\le y\le d,\ h_1(y)\le x\le h_2(y)\},$$

then

$$\iint_D f(x,y)\,dA=\int_c^d\!\!\int_{h_1(y)}^{h_2(y)} f(x,y)\,dx\,dy.\tag{4}$$

Here the inner integration runs in the $x$-direction with $y$ held constant — again in the limits $h_1(y)$ and $h_2(y)$ as well as in $f(x,y)$ — and the outer integration collects the result from $y=c$ to $y=d$.

Proof. This is the proof of equation (3) with the roles of $x$ and $y$ exchanged. Since $h_1,h_2$ are continuous on $[c,d]$, they are bounded, so there exist numbers $a,b$ with $a\le h_1(y)\le h_2(y)\le b$ for every $y\in[c,d]$; put $R=[a,b]\times[c,d]$, so $D\subset R$. Definition 2 gives

$$\iint_D f(x,y)\,dA=\iint_R F(x,y)\,dA,$$

and, $F$ being bounded and discontinuous only on finitely many curves for the same reason as before, the general form of §15.1 Theorem 10 applied with $x$ as the inner variable yields

$$\iint_R F(x,y)\,dA=\int_c^d\int_a^b F(x,y)\,dx\,dy.$$

For fixed $y\in[c,d]$, $F(x,y)=0$ outside $h_1(y)\le x\le h_2(y)$, so the inner integral collapses to

$$\int_a^b F(x,y)\,dx=\int_{h_1(y)}^{h_2(y)}f(x,y)\,dx,$$

and taking the outer integral gives equation (4),

$$\iint_D f(x,y)\,dA=\int_c^d\int_{h_1(y)}^{h_2(y)}f(x,y)\,dx\,dy.$$
$\square$
REM 8 적분 한계를 그림에서 읽는 법 Reading the Limits from a SketchExample 1 뒤 NOTE · p.1092, 1094–1095
한국어

일반 영역 위의 이중적분을 세울 때는 $D$의 그림을 그리는 일이 필수적이다. 적분 한계는 짐작으로 정하는 것이 아니라 그림에서 그대로 읽어내는 것이다.

제1형 영역이면 $D$를 가로지르는 세로 화살표를 긋는다. 화살표는 아래 경계 $y=g_1(x)$에서 출발하는데 이것이 안쪽 적분의 하한을 주고, 위 경계 $y=g_2(x)$에서 끝나는데 이것이 상한을 준다. 바깥 한계 $a$와 $b$는 $D$에 나타나는 $x$값 가운데 가장 작은 값과 가장 큰 값이다.

제2형 영역이면 화살표를 가로로 긋는다. 왼쪽 경계 $x=h_1(y)$에서 오른쪽 경계 $x=h_2(y)$까지이며, 바깥 한계는 $y$의 극단값인 $c$와 $d$이다.

한 영역이 두 형을 동시에 만족할 수도 있다. 이때는 경계를 나타내는 데 필요한 식의 개수가 더 적은 쪽을 고르는 편이 낫다. 예제 3에서는 제1형으로 기술하면 아래 경계가 두 조각으로 쪼개지는 반면, 제2형으로 기술하면 양쪽 경계가 각각 하나의 식으로 끝난다.

English

When a double integral over a general region is set up, drawing a diagram of $D$ is essential; the limits are then read off the picture rather than guessed.

For a type I region, draw a vertical arrow across $D$. The arrow starts on the lower boundary $y=g_1(x)$, which supplies the lower limit of the inner integral, and ends on the upper boundary $y=g_2(x)$, which supplies the upper limit; the outer limits $a$ and $b$ are the smallest and largest values of $x$ occurring in $D$.

For a type II region the arrow is drawn horizontally, from the left boundary $x=h_1(y)$ to the right boundary $x=h_2(y)$, and the outer limits are the extreme values $c$ and $d$ of $y$.

A region can be of both types at once. Then the better description is the one whose boundaries need fewer formulas: in Example 3 the type I description forces the lower boundary to be split into two pieces, while the type II description uses one formula on each side.

NOTE 적분 순서 바꾸기 Changing the Order of IntegrationChanging the Order of Integration · p.1095
한국어

푸비니 정리에 따르면 이중적분은 어느 순서로도 반복적분으로 쓸 수 있지만, 두 순서의 계산 품이 늘 같은 것은 아니다. 한쪽 순서가 다른 쪽보다 훨씬 어렵거나, 초등함수로는 아예 계산할 수 없는 경우도 있다.

주어진 반복적분이 그대로는 계산되지 않을 때 해법은 식 (3)과 (4)를 거꾸로 적용하는 것이다. 주어진 한계로부터 적분이 이루어지는 영역 $D$를 읽어 반복적분을 $\iint_D f(x,y)\,dA$로 되돌린 뒤, $D$를 그려 보고, 같은 $D$를 다른 형으로 다시 기술하여, 그 형에 맞는 공식을 적용하면 반대 순서의 반복적분을 얻는다.

예제 5가 바로 이 과정을 $\int_0^1\!\int_x^1\sin(y^2)\,dy\,dx$에 적용한다. 안쪽 적분은 $\sin(y^2)$의 초등함수 형태의 원시함수가 없어 계산되지 않지만, 순서를 바꾸면 곧바로 적분된다.

English

Fubini's Theorem allows a double integral to be written as an iterated integral in either order, but the two orders need not cost the same: one of them can be much harder than the other, or impossible to carry out in closed form.

When an iterated integral resists evaluation as it stands, the remedy is to run formulas (3) and (4) backward. Read off from the given limits the region $D$ over which the integration takes place, so that the iterated integral becomes $\iint_D f(x,y)\,dA$; sketch $D$; describe the same $D$ as a region of the other type; and apply the formula for that type to obtain an iterated integral in the reverse order.

Example 5 does this for $\int_0^1\!\int_x^1\sin(y^2)\,dy\,dx$, whose inner integral cannot be evaluated because $\sin(y^2)$ has no elementary antiderivative, while the reversed order integrates at once.

THM 10 이중적분의 선형성과 비교 Linearity and ComparisonProperties (5)(6)(7) · p.1096
한국어

성질 (5)–(7) — 선형성과 비교. 아래 적분이 모두 존재한다고 하자. 영역 $D$ 위의 함수 $f,g$와 상수 $c$에 대해 다음이 성립한다.

$$\iint_D\big[f(x,y)+g(x,y)\big]\,dA=\iint_D f(x,y)\,dA+\iint_D g(x,y)\,dA,\tag{5}$$
$$\iint_D c\,f(x,y)\,dA=c\iint_D f(x,y)\,dA.\tag{6}$$

또한 $D$의 모든 점에서 $f(x,y)\ge g(x,y)$이면, 다음이 성립한다.

$$\iint_D f(x,y)\,dA\ge\iint_D g(x,y)\,dA.\tag{7}$$

성질 (5)와 (6)은 이중적분이 선형이라는 뜻이고, 성질 (7)은 이중적분이 부등호를 보존한다는 뜻이다. 바로 이 성질 덕분에 성질 (10)과 같은 어림이 가능해진다.

증명. $R$을 $D$를 포함하는 직사각형이라 하고, $F,G$를 각각 (1)에 따른 $f,g$의 확장함수라 하자. $D$ 안에서는 $F+G=f+g$이고 $D$ 밖에서는 $F+G=0+0=0$이므로 $F+G$는 정확히 $f+g$의 확장함수이며, 같은 이유로 $cF$는 $cf$의 확장함수이다.

직사각형 위에서는 (5)와 (6)이 리만 합의 선형성에서 곧바로 따라온다(§4.2의 단일적분 성질을 얻는 방법과 같다). 따라서 정의 2를 적용한 뒤 이 선형성을 이용하면

$$\iint_D[f+g]\,dA=\iint_R[F+G]\,dA=\iint_R F\,dA+\iint_R G\,dA=\iint_D f\,dA+\iint_D g\,dA,$$
$$\iint_D cf\,dA=\iint_R cF\,dA=c\iint_R F\,dA=c\iint_D f\,dA$$

가 되어 (5)와 (6)이 성립한다.

(7)을 보이기 위해 $D$의 모든 점에서 $f\ge g$라 하자. $D$ 안에서는 $F=f\ge g=G$이고 $D$ 밖에서는 $F=G=0$이므로 $R$ 전체에서 $F\ge G$이다. 직사각형 위의 비교 성질(같은 방법으로 얻는다)을 $F,G$에 적용하면 $\displaystyle\iint_R F\,dA\ge\iint_R G\,dA$, 즉

$$\iint_D f\,dA\ge\iint_D g\,dA$$

를 얻는다.

$\square$
English

Properties (5)–(7) — Linearity and Comparison. Assume all the integrals below exist. For functions $f$ and $g$ on a region $D$ and a constant $c$,

$$\iint_D\big[f(x,y)+g(x,y)\big]\,dA=\iint_D f(x,y)\,dA+\iint_D g(x,y)\,dA,\tag{5}$$
$$\iint_D c\,f(x,y)\,dA=c\iint_D f(x,y)\,dA.\tag{6}$$

Moreover, if $f(x,y)\ge g(x,y)$ at every point of $D$, then

$$\iint_D f(x,y)\,dA\ge\iint_D g(x,y)\,dA.\tag{7}$$

Properties (5) and (6) say that the double integral is linear; property (7) says that it preserves inequalities, which is what makes estimates such as (10) possible.

Proof. Let $R$ be a rectangle containing $D$, and let $F,G$ be the extensions of $f,g$ given by (1). Inside $D$, $F+G=f+g$, and outside $D$, $F+G=0+0=0$, so $F+G$ is exactly the extension of $f+g$; for the same reason $cF$ is the extension of $cf$.

Over a rectangle, (5) and (6) drop straight out of the linearity of the Riemann sums, just as the corresponding properties of single integrals do in §4.2. Applying Definition 2 and then that linearity,

$$\iint_D[f+g]\,dA=\iint_R[F+G]\,dA=\iint_R F\,dA+\iint_R G\,dA=\iint_D f\,dA+\iint_D g\,dA,$$
$$\iint_D cf\,dA=\iint_R cF\,dA=c\iint_R F\,dA=c\iint_D f\,dA,$$

which are (5) and (6).

For (7), suppose $f\ge g$ at every point of $D$. Inside $D$, $F=f\ge g=G$, and outside $D$, $F=G=0$, so $F\ge G$ throughout $R$. The comparison property over a rectangle, obtained in the same way, applied to $F$ and $G$, gives $\displaystyle\iint_R F\,dA\ge\iint_R G\,dA$, that is,

$$\iint_D f\,dA\ge\iint_D g\,dA.$$
$\square$
THM 11 영역 분할 Additivity over SubregionsProperty (8) · Figures 17–18 · p.1096
한국어

성질 (8) — 영역 분할. $D=D_1\cup D_2$이고, $D_1$과 $D_2$가 겹치더라도 그 경계에서만 겹친다면, 다음이 성립한다.

$$\iint_D f(x,y)\,dA=\iint_{D_1} f(x,y)\,dA+\iint_{D_2} f(x,y)\,dA.\tag{8}$$

이는 1변수 적분에서 $\int_a^b f(x)\,dx=\int_a^c f(x)\,dx+\int_c^b f(x)\,dx$가 성립하는 것의 2차원 대응이다.

이 성질의 실용적 가치는 제1형·제2형이라는 두 특수한 모양의 제약에서 벗어나게 해 준다는 데 있다. 어느 형에도 속하지 않는 영역 $D$라도, 각각이 제1형이거나 제2형인 조각들로 나눌 수만 있다면, 각 조각 위에서 따로 적분한 뒤 그 결과를 더하면 된다.

English

Property (8) — Additivity over Subregions. If $D=D_1\cup D_2$, where $D_1$ and $D_2$ do not overlap except perhaps along their boundaries, then

$$\iint_D f(x,y)\,dA=\iint_{D_1} f(x,y)\,dA+\iint_{D_2} f(x,y)\,dA.\tag{8}$$

This is the two-dimensional counterpart of $\int_a^b f(x)\,dx=\int_a^c f(x)\,dx+\int_c^b f(x)\,dx$ for single integrals.

Its practical value is that it removes the restriction to the two special shapes: a region $D$ that is of neither type, but that can be cut into pieces each of which is of type I or of type II, is handled by integrating over the pieces separately and adding the results.

THM 12 넓이 공식 Area as a Double IntegralProperty (9) · Figure 19 · p.1096
한국어

성질 (9) — 넓이 공식. 상수함수 $1$을 영역 $D$ 위에서 적분하면 $D$의 넓이가 나온다.

$$\iint_D 1\,dA=A(D).\tag{9}$$

이 시점부터는 넓이를 이중적분으로 구할 수 있다. $D$의 경계가 여러 곡선으로 이루어져 있을 때는 이 방법이 단일적분보다 세우기 쉬운 경우가 많다.

증명. 상수함수 $1$은 $D$ 위에서 $1\ge0$이므로, 이 절 도입에서 설명한 부피 해석에 의해 $\iint_D1\,dA$는 밑면이 $D$이고 윗면이 평면 $z=1$인 입체의 부피와 같다. 이 입체는 밑면 $D$, 높이 $1$인 (일반적인 뜻의) 기둥이므로 그 부피는 (밑넓이)$\times$(높이)$=A(D)\times1=A(D)$이며, 이것이 곧 (9)이다.

같은 결론을 정의 2로부터 직접 얻을 수도 있다. $R$을 $D$를 포함하는 직사각형이라 하면 상수함수 $1$의 확장함수 $F$는 $D$ 안에서 $1$, $D$ 밖에서 $0$인 함수, 즉 $D$의 지시함수이다. $R$의 분할에 대한 리만 합 $\sum F(x_{ij}^*,y_{ij}^*)\,\Delta A$에서는 표본점 $(x_{ij}^*,y_{ij}^*)$가 $D$에 속하는 항만 $\Delta A$를 내놓고 나머지는 $0$이므로, 이 합은 $D$를 덮는 부분직사각형들의 넓이의 합에 다가간다. 분할을 한없이 잘게 하면 그 극한 $\iint_R F\,dA$는 곧 $D$의 넓이 $A(D)$이다.

$\square$
English

Property (9) — Area as a Double Integral. Integrating the constant function $1$ over a region $D$ returns the area of $D$:

$$\iint_D 1\,dA=A(D).\tag{9}$$

From this point on, an area can be computed as a double integral, which is often easier to set up than a single integral when the boundary of $D$ is described by several curves.

Proof. The constant function $1$ satisfies $1\ge0$ on $D$, so by the volume interpretation described at the opening of this section, $\iint_D1\,dA$ equals the volume of the solid with base $D$ and top the plane $z=1$. That solid is a cylinder (in the general sense) with base $D$ and height $1$, whose volume is (base area)$\times$(height)$=A(D)\times1=A(D)$, which is (9).

The same conclusion also follows directly from Definition 2. Let $R$ be a rectangle containing $D$; the extension $F$ of the constant function $1$ equals $1$ on $D$ and $0$ elsewhere on $R$, that is, $F$ is the indicator function of $D$. In a Riemann sum $\sum F(x_{ij}^*,y_{ij}^*)\,\Delta A$ over a partition of $R$, a term contributes $\Delta A$ when its sample point $(x_{ij}^*,y_{ij}^*)$ lies in $D$ and $0$ otherwise, so the sum approaches the total area of the sub-rectangles covering $D$; refining the partition, its limit $\iint_R F\,dA$ is the area $A(D)$.

$\square$
THM 13 이중적분의 상·하한 추정 Bounds for a Double IntegralProperty (10) · Figure 20 · p.1097
한국어

성질 (10) — 이중적분의 상·하한 추정. $D$의 모든 점 $(x,y)$에서 $m\le f(x,y)\le M$이면, 다음이 성립한다.

$$m\,A(D)\le\iint_D f(x,y)\,dA\le M\,A(D).\tag{10}$$

$m\gt 0$일 때는 기하적 의미가 곧바로 보인다. $D$ 위·곡면 $z=f(x,y)$ 아래에 놓인 입체는 밑면이 $D$이고 높이가 각각 $m$, $M$인 두 기둥(일반적인 뜻의 기둥) 사이에 끼여 있으며, 그 부피가 각각 $m\,A(D)$, $M\,A(D)$이다.

이 어림에 드는 대가는 $f$의 상·하계와 $D$의 넓이뿐이므로, 예제 6처럼 적분 자체를 계산할 수 없는 경우에도 이 어림은 여전히 쓸 수 있다.

증명. 연습문제 73 — 성질 (6), (7), (9)를 결합한다.

English

Property (10) — Bounds for a Double Integral. If $m\le f(x,y)\le M$ for all $(x,y)$ in $D$, then

$$m\,A(D)\le\iint_D f(x,y)\,dA\le M\,A(D).\tag{10}$$

For $m\gt 0$ the geometric content is visible at once: the solid lying above $D$ and below the graph $z=f(x,y)$ is trapped between two cylinders with base $D$, one of height $m$ and one of height $M$, whose volumes are $m\,A(D)$ and $M\,A(D)$.

The estimate costs nothing but a bound on $f$ and the area of $D$, so it is available even when the integral itself cannot be evaluated, as in Example 6.

Proof: Exercise 73 — combine Properties (6), (7), and (9).

예제 답안 · Worked Examples

EXAM 1 두 포물선 사이 영역에서의 적분 Integrating between Two ParabolasExample 1 · p.1092
한국어

두 포물선 $y=2x^2$과 $y=1+x^2$으로 둘러싸인 영역을 $D$라 하자. 이중적분 $\displaystyle\iint_D (x+2y)\,dA$ 의 값을 구하라.

풀이. 두 포물선의 교점을 먼저 구한다. $2x^2=1+x^2$에서 $x^2=1$, 즉 $x=\pm1$이다. $-1\le x\le1$에서 $(1+x^2)-2x^2=1-x^2\ge0$이므로 이 구간에서는 $y=1+x^2$이 위 경계, $y=2x^2$이 아래 경계이다. 즉

$$D=\{(x,y)\mid -1\le x\le1,\ 2x^2\le y\le1+x^2\}$$

는 제1형 영역이고, 식 (3)에 의해

$$\iint_D(x+2y)\,dA=\int_{-1}^{1}\int_{2x^2}^{1+x^2}(x+2y)\,dy\,dx.$$

안쪽 적분은 원시함수 $xy+y^2$을 두 한계에서 계산하면 되며, 정리하면

$$\int_{2x^2}^{1+x^2}(x+2y)\,dy=\Big[xy+y^2\Big]_{y=2x^2}^{y=1+x^2}=-3x^4-x^3+2x^2+x+1$$

을 얻는다. 이 다항식을 $-1$부터 $1$까지 적분하는데, 홀수차항 $-x^3+x$는 대칭 구간에서 적분값이 $0$이므로 짝수차항만 남는다.

$$\int_{-1}^{1}(-3x^4+2x^2+1)\,dx=2\int_0^1(-3x^4+2x^2+1)\,dx=2\left[-\frac35x^5+\frac23x^3+x\right]_0^1=2\left(-\frac35+\frac23+1\right)=\frac{32}{15}.$$

따라서 $\displaystyle\iint_D(x+2y)\,dA=\frac{32}{15}$이다.

English

Let $D$ be the region enclosed by the two parabolas $y=2x^2$ and $y=1+x^2$. Evaluate $\displaystyle\iint_D (x+2y)\,dA$.

Solution. Find where the two parabolas meet: $2x^2=1+x^2$ gives $x^2=1$, so $x=\pm1$. On $-1\le x\le1$, $(1+x^2)-2x^2=1-x^2\ge0$, so on this interval $y=1+x^2$ is the upper boundary and $y=2x^2$ the lower one. Thus

$$D=\{(x,y)\mid -1\le x\le1,\ 2x^2\le y\le1+x^2\}$$

is a type I region, and equation (3) gives

$$\iint_D(x+2y)\,dA=\int_{-1}^{1}\int_{2x^2}^{1+x^2}(x+2y)\,dy\,dx.$$

The inner integral is evaluated from the antiderivative $xy+y^2$ at the two limits; simplifying,

$$\int_{2x^2}^{1+x^2}(x+2y)\,dy=\Big[xy+y^2\Big]_{y=2x^2}^{y=1+x^2}=-3x^4-x^3+2x^2+x+1.$$

Integrating this polynomial from $-1$ to $1$, the odd terms $-x^3+x$ contribute $0$ over the symmetric interval, leaving only the even-degree terms.

$$\int_{-1}^{1}(-3x^4+2x^2+1)\,dx=2\int_0^1(-3x^4+2x^2+1)\,dx=2\left[-\frac35x^5+\frac23x^3+x\right]_0^1=2\left(-\frac35+\frac23+1\right)=\frac{32}{15}.$$

Hence $\displaystyle\iint_D(x+2y)\,dA=\frac{32}{15}$.

EXAM 2 두 가지 형으로 구하는 부피 A Volume Computed Two WaysExample 2 · p.1093
한국어

$xy$평면에서 직선 $y=2x$와 포물선 $y=x^2$으로 둘러싸인 영역을 $D$라 하자. 포물면 $z=x^2+y^2$ 아래, 그리고 $D$ 위에 놓인 입체의 부피를 구하라. 단, $D$를 제1형 영역으로 볼 때와 제2형 영역으로 볼 때의 두 가지 방법으로 각각 계산하라.

풀이. 두 곡선의 교점은 $2x=x^2$에서 $x=0,2$이고, $0\le x\le2$에서 $2x-x^2=x(2-x)\ge0$이므로 이 구간에서 $y=2x$가 위, $y=x^2$이 아래이다.

(a) 제1형으로. $D=\{(x,y)\mid0\le x\le2,\ x^2\le y\le2x\}$이므로 식 (3)에 의해

$$V=\int_0^2\int_{x^2}^{2x}(x^2+y^2)\,dy\,dx.$$

안쪽 적분은 $x^2y+\dfrac{y^3}{3}$을 두 한계에서 계산하여

$$\int_{x^2}^{2x}(x^2+y^2)\,dy=\frac{14}{3}x^3-x^4-\frac13x^6$$

이 되고, 이를 $0$부터 $2$까지 적분하면 $V=\dfrac{216}{35}$을 얻는다.

(b) 제2형으로. 같은 $D$를 $x$에 대해 풀면, 주어진 $y$에서 $x$는 포물선 $x=\sqrt y$와 직선 $x=y/2$ 사이에 있으므로

$$D=\Big\{(x,y)\Bigm|0\le y\le4,\ \frac{y}{2}\le x\le\sqrt y\Big\}$$

이고, 식 (4)에 의해

$$V=\int_0^4\int_{y/2}^{\sqrt y}(x^2+y^2)\,dx\,dy.$$

안쪽 적분은 $\dfrac{x^3}{3}+y^2x$를 두 한계에서 계산하여 $y^{5/2}+\dfrac13y^{3/2}-\dfrac{13}{24}y^3$이 되고, 이를 $0$부터 $4$까지 적분하면 다시 $V=\dfrac{216}{35}$을 얻는다.

두 방법이 일치하므로, 입체의 부피는 $\dfrac{216}{35}$이다.

English

Let $D$ be the region of the $xy$-plane bounded by the line $y=2x$ and the parabola $y=x^2$. Find the volume of the solid that lies under the paraboloid $z=x^2+y^2$ and above $D$, computing it twice: once with $D$ regarded as a type I region and once with $D$ regarded as a type II region.

Solution. The two curves meet where $2x=x^2$, that is, $x=0,2$, and $2x-x^2=x(2-x)\ge0$ on $0\le x\le2$, so on this interval $y=2x$ is the upper curve and $y=x^2$ the lower one.

(a) As a type I region. Here $D=\{(x,y)\mid0\le x\le2,\ x^2\le y\le2x\}$, so equation (3) gives

$$V=\int_0^2\int_{x^2}^{2x}(x^2+y^2)\,dy\,dx.$$

Evaluating the antiderivative $x^2y+\dfrac{y^3}{3}$ at the two limits,

$$\int_{x^2}^{2x}(x^2+y^2)\,dy=\frac{14}{3}x^3-x^4-\frac13x^6,$$

and integrating this from $0$ to $2$ gives $V=\dfrac{216}{35}$.

(b) As a type II region. Solving the same $D$ for $x$, a given $y$ has $x$ ranging between the parabola $x=\sqrt y$ and the line $x=y/2$, so

$$D=\Big\{(x,y)\Bigm|0\le y\le4,\ \frac{y}{2}\le x\le\sqrt y\Big\},$$

and equation (4) gives

$$V=\int_0^4\int_{y/2}^{\sqrt y}(x^2+y^2)\,dx\,dy.$$

Evaluating the antiderivative $\dfrac{x^3}{3}+y^2x$ at the two limits gives $y^{5/2}+\dfrac13y^{3/2}-\dfrac{13}{24}y^3$, and integrating this from $0$ to $4$ again gives $V=\dfrac{216}{35}$.

The two computations agree, so the solid has volume $\dfrac{216}{35}$.

EXAM 3 어느 형으로 볼 것인가 Choosing between Type I and Type IIExample 3 · p.1094–1095
한국어

직선 $y=x-1$과 포물선 $y^2=2x+6$으로 둘러싸인 영역을 $D$라 하자. 이중적분 $\displaystyle\iint_D xy\,dA$ 의 값을 구하고, $D$를 어느 형으로 기술하는 편이 계산이 짧은지 밝혀라.

풀이. 교점을 구하기 위해 직선에서 $x=y+1$을 얻어 포물선의 방정식에 대입하면 $y^2=2(y+1)+6$, 즉 $y^2-2y-8=0$이므로 $(y-4)(y+2)=0$에서 $y=-2,4$이다. 교점은 $(-1,-2)$와 $(5,4)$이다.

$D$를 제1형으로 기술하면 $-3\le x\le-1$에서는 위·아래 경계가 모두 포물선의 두 갈래 $y=\pm\sqrt{2x+6}$이고, $-1\lt x\le5$에서는 아래 경계가 직선 $y=x-1$로 바뀌어 적분을 두 조각으로 나누어야 한다. 반면 제2형으로 보면 $-2\le y\le4$인 모든 $y$에서 왼쪽 경계는 포물선 $x=\dfrac{y^2}{2}-3$, 오른쪽 경계는 직선 $x=y+1$ 하나로 끝나므로

$$D=\Big\{(x,y)\Bigm| -2\le y\le4,\ \frac{y^2}{2}-3\le x\le y+1\Big\}$$

가 훨씬 짧은 기술이다. 식 (4)에 의해

$$\iint_D xy\,dA=\int_{-2}^{4}\int_{y^2/2-3}^{y+1}xy\,dx\,dy.$$

안쪽 적분은 $\dfrac{x^2}{2}y$를 두 한계에서 계산하여

$$\int_{y^2/2-3}^{y+1}xy\,dx=-\frac18y^5+2y^3+y^2-4y$$

가 되고, 이를 $-2$부터 $4$까지 적분하면 $36$을 얻는다. 따라서 $\displaystyle\iint_D xy\,dA=36$이며, 경계가 한 식씩으로 끝나는 제2형 기술이 두 조각으로 쪼개지는 제1형 기술보다 계산이 짧다.

English

Let $D$ be the region bounded by the line $y=x-1$ and the parabola $y^2=2x+6$. Evaluate $\displaystyle\iint_D xy\,dA$, and say which description of $D$ — type I or type II — makes the computation shorter.

Solution. To find the intersection points, solve the line for $x=y+1$ and substitute into the parabola: $y^2=2(y+1)+6$, that is, $y^2-2y-8=0$, so $(y-4)(y+2)=0$ gives $y=-2,4$. The intersection points are $(-1,-2)$ and $(5,4)$.

Describing $D$ as a type I region, on $-3\le x\le-1$ both boundaries come from the two branches of the parabola, $y=\pm\sqrt{2x+6}$, while on $-1\lt x\le5$ the lower boundary switches to the line $y=x-1$, forcing the integral to be split into two pieces. As a type II region, however, every $y$ with $-2\le y\le4$ has left boundary the parabola $x=\dfrac{y^2}{2}-3$ and right boundary the line $x=y+1$, a single formula on each side, so

$$D=\Big\{(x,y)\Bigm| -2\le y\le4,\ \frac{y^2}{2}-3\le x\le y+1\Big\}$$

is the much shorter description. Equation (4) gives

$$\iint_D xy\,dA=\int_{-2}^{4}\int_{y^2/2-3}^{y+1}xy\,dx\,dy.$$

Evaluating the antiderivative $\dfrac{x^2}{2}y$ at the two limits,

$$\int_{y^2/2-3}^{y+1}xy\,dx=-\frac18y^5+2y^3+y^2-4y,$$

and integrating this from $-2$ to $4$ gives $36$. Hence $\displaystyle\iint_D xy\,dA=36$, and the type II description — with one formula on each boundary — is shorter than the type I description, which splits into two pieces.

EXAM 4 사면체의 부피 Volume of a TetrahedronExample 4 · p.1094–1095
한국어

네 평면 $x+2y+z=2$, $x=2y$, $x=0$, $z=0$ 으로 둘러싸인 사면체의 부피를 구하라.

풀이. 사면체는 위에서 평면 $x+2y+z=2$로, 아래에서 $z=0$으로, 옆에서 $x=0$과 $x=2y$로 잘려 있다. 밑면 $D$는 $xy$평면($z=0$)에서 이 옆면들과 위 평면이 만드는 삼각형이다. $x=2y$는 $y=x/2$이고 $x+2y=2$(즉 $z=0$일 때 위 평면)는 $y=1-x/2$이며, 두 직선은 $x/2=1-x/2$, 즉 $x=1$에서 만난다. 따라서

$$D=\Big\{(x,y)\Bigm| 0\le x\le1,\ \frac{x}{2}\le y\le1-\frac{x}{2}\Big\}$$

이고, 위 평면에서 잰 높이는 $z=2-x-2y$이다. 식 (3)에 의해 부피는

$$V=\int_0^1\int_{x/2}^{1-x/2}(2-x-2y)\,dy\,dx.$$

안쪽 적분은 원시함수 $2y-xy-y^2$을 두 한계에서 계산하면

$$\int_{x/2}^{1-x/2}(2-x-2y)\,dy=x^2-2x+1=(x-1)^2$$

이 되고, 이를 $0$부터 $1$까지 적분하면

$$V=\int_0^1(x-1)^2\,dx=\left[\frac{(x-1)^3}{3}\right]_0^1=0-\left(-\frac13\right)=\frac13$$

이다. 따라서 사면체의 부피는 $\dfrac13$이다.

English

Find the volume of the tetrahedron bounded by the four planes $x+2y+z=2$, $x=2y$, $x=0$, and $z=0$.

Solution. The tetrahedron is cut off above by the plane $x+2y+z=2$, below by $z=0$, and on the sides by $x=0$ and $x=2y$. Its base $D$ is the triangle in the $xy$-plane ($z=0$) cut out by these side planes and the top plane. Now $x=2y$ gives $y=x/2$, and $x+2y=2$ (the top plane with $z=0$) gives $y=1-x/2$; the two lines meet where $x/2=1-x/2$, that is, at $x=1$. Hence

$$D=\Big\{(x,y)\Bigm| 0\le x\le1,\ \frac{x}{2}\le y\le1-\frac{x}{2}\Big\},$$

and the height under the top plane is $z=2-x-2y$. By equation (3) the volume is

$$V=\int_0^1\int_{x/2}^{1-x/2}(2-x-2y)\,dy\,dx.$$

Evaluating the antiderivative $2y-xy-y^2$ at the two limits,

$$\int_{x/2}^{1-x/2}(2-x-2y)\,dy=x^2-2x+1=(x-1)^2,$$

and integrating this from $0$ to $1$,

$$V=\int_0^1(x-1)^2\,dx=\left[\frac{(x-1)^3}{3}\right]_0^1=0-\left(-\frac13\right)=\frac13.$$

Hence the tetrahedron has volume $\dfrac13$.

EXAM 5 적분 순서를 바꾸어 계산하기 Reversing the Order of IntegrationExample 5 · p.1095
한국어

반복적분 $\displaystyle\int_0^1\!\!\int_x^1 \sin(y^2)\,dy\,dx$ 의 값을 구하라. 안쪽 적분 $\int\sin(y^2)\,dy$ 는 초등함수로 나타낼 수 없으므로, 적분 영역을 다시 기술하여 적분 순서를 바꾸어라.

x y 0 1 1 y=x y=1 D

풀이. 주어진 반복적분의 적분 영역은 위 그림의 삼각형 $D=\{(x,y)\mid0\le x\le1,\ x\le y\le1\}$이며, 이것이 (바깥이 $x$, 안쪽이 $y$인) 제1형 기술이다. 안쪽 적분 $\displaystyle\int_x^1\sin(y^2)\,dy$는 $\sin(y^2)$의 원시함수가 초등함수로 존재하지 않아 그대로는 계산되지 않으므로, 같은 $D$를 제2형으로 다시 기술해 순서를 바꾼다.

위 그림의 가로 화살표가 보여 주듯, $D$ 안의 점은 $0\le y\le1$인 각 $y$에 대해 $x$가 $0$부터 $y$까지 움직인 점들이므로

$$D=\{(x,y)\mid 0\le y\le1,\ 0\le x\le y\}$$

이다. 이 기술로 식 (4)를 적용하면

$$\int_0^1\int_x^1\sin(y^2)\,dy\,dx=\iint_D\sin(y^2)\,dA=\int_0^1\int_0^y\sin(y^2)\,dx\,dy.$$

안쪽 적분에서 $\sin(y^2)$은 $x$에 대해 상수이므로 $\displaystyle\int_0^y\sin(y^2)\,dx=y\sin(y^2)$이 되고, 남은 적분은 곧바로 원시함수를 찾을 수 있다.

$$\int_0^1 y\sin(y^2)\,dy=\left[-\frac12\cos(y^2)\right]_0^1=-\frac12\cos1+\frac12=\frac12(1-\cos1).$$

따라서 $\displaystyle\int_0^1\int_x^1\sin(y^2)\,dy\,dx=\frac12(1-\cos1)\approx0.2298$이다.

English

Evaluate the iterated integral $\displaystyle\int_0^1\!\!\int_x^1 \sin(y^2)\,dy\,dx$. Since $\int\sin(y^2)\,dy$ cannot be expressed in terms of elementary functions, redescribe the region of integration and reverse the order.

x y 0 1 1 y=x y=1 D

Solution. The region of integration for the given iterated integral is the triangle $D=\{(x,y)\mid0\le x\le1,\ x\le y\le1\}$ shown in the figure above, which is the type I description (outer variable $x$, inner variable $y$). The inner integral $\displaystyle\int_x^1\sin(y^2)\,dy$ cannot be evaluated as it stands, since $\sin(y^2)$ has no elementary antiderivative, so the same $D$ is redescribed as a type II region to reverse the order.

As the horizontal arrow in the figure above shows, for each $y$ with $0\le y\le1$ the points of $D$ are exactly those whose $x$ runs from $0$ to $y$, so

$$D=\{(x,y)\mid 0\le y\le1,\ 0\le x\le y\}.$$

With this description, equation (4) gives

$$\int_0^1\int_x^1\sin(y^2)\,dy\,dx=\iint_D\sin(y^2)\,dA=\int_0^1\int_0^y\sin(y^2)\,dx\,dy.$$

In the inner integral $\sin(y^2)$ is constant with respect to $x$, so $\displaystyle\int_0^y\sin(y^2)\,dx=y\sin(y^2)$, and the remaining integral has an antiderivative in sight at once.

$$\int_0^1 y\sin(y^2)\,dy=\left[-\frac12\cos(y^2)\right]_0^1=-\frac12\cos1+\frac12=\frac12(1-\cos1).$$

Hence $\displaystyle\int_0^1\int_x^1\sin(y^2)\,dy\,dx=\frac12(1-\cos1)\approx0.2298$.

EXAM 6 성질 (10)으로 적분값 어림하기 Estimating an Integral by Property (10)Example 6 · p.1097
한국어

원점을 중심으로 하고 반지름이 $2$인 원판을 $D$라 하자. 성질 (10)을 이용하여 적분 $\displaystyle\iint_D e^{\sin x\cos y}\,dA$ 의 값을 위아래에서 어림하라.

풀이. 모든 실수 $x,y$에서 $|\sin x|\le1$, $|\cos y|\le1$이므로 $-1\le\sin x\cos y\le1$이고, 이는 특히 $D$ 위의 모든 점에서도 성립한다. $t\mapsto e^t$가 증가함수이므로 이 부등식 전체에 지수를 취하면

$$m=e^{-1}\le e^{\sin x\cos y}\le e=M\qquad\big((x,y)\in D\big)$$

를 얻는다. $D$는 원점을 중심으로 하는 반지름 $2$인 원판이므로 $A(D)=\pi(2)^2=4\pi$이다. 성질 (10)은

$$m\,A(D)\le\iint_D e^{\sin x\cos y}\,dA\le M\,A(D)$$

를 주므로, 여기에 위의 $m,M,A(D)$를 대입하면

$$\frac1e\cdot4\pi\le\iint_D e^{\sin x\cos y}\,dA\le e\cdot4\pi$$

이다. 따라서 $\displaystyle\frac{4\pi}{e}\le\iint_D e^{\sin x\cos y}\,dA\le4\pi e$이며, 수치로는 대략 $4.62\le\iint_D e^{\sin x\cos y}\,dA\le34.16$이다.

English

Let $D$ be the disk with center the origin and radius $2$. Use Property (10) to estimate the integral $\displaystyle\iint_D e^{\sin x\cos y}\,dA$ from below and from above.

Solution. For every real $x,y$, $|\sin x|\le1$ and $|\cos y|\le1$, so $-1\le\sin x\cos y\le1$; in particular this holds at every point of $D$. Since $t\mapsto e^t$ is increasing, exponentiating throughout gives

$$m=e^{-1}\le e^{\sin x\cos y}\le e=M\qquad\big((x,y)\in D\big).$$

$D$ is the disk of radius $2$ centered at the origin, so $A(D)=\pi(2)^2=4\pi$. Property (10) reads

$$m\,A(D)\le\iint_D e^{\sin x\cos y}\,dA\le M\,A(D),$$

and substituting these values of $m,M,A(D)$ gives

$$\frac1e\cdot4\pi\le\iint_D e^{\sin x\cos y}\,dA\le e\cdot4\pi.$$

Hence $\displaystyle\frac{4\pi}{e}\le\iint_D e^{\sin x\cos y}\,dA\le4\pi e$, or numerically $4.62\le\iint_D e^{\sin x\cos y}\,dA\le34.16$ approximately.

연결 · Links