
The new schemes of calculation of double integrals and triple integrals are proposed in this paper. The formulas in which the double integral is converted into a line integral with respect to the arc length, and the triple integral is converted into a surface integral with respect to the area or a line integral with respect to the arc length are given separately. The effectiveness of the proposed methods is verified by several examples. Under certain conditions, these methods become the normal iterated integrals in Cartesian coordinate system or polar coordinate system, and the commonly used triple iterated integrals in Cartesian coordinate system, Cylindrical coordinate system or Spherical coordinate system. The transformation calculation method promoted in this paper points out the intrinsic relationship among double integral, triple integral, line integral and surface integral, which further enriches the theories of multivariate integrals.
Citation: Rong-jian Ning, Xiao-yan Liu, Zhi Liu. Conversion calculation method of multivariate integrals[J]. AIMS Mathematics, 2021, 6(3): 3009-3024. doi: 10.3934/math.2021183
[1] | Sunyoung Bu . A collocation methods based on the quadratic quadrature technique for fractional differential equations. AIMS Mathematics, 2022, 7(1): 804-820. doi: 10.3934/math.2022048 |
[2] | SAIRA, Wenxiu Ma, Suliman Khan . An efficient numerical method for highly oscillatory logarithmic-algebraic singular integrals. AIMS Mathematics, 2025, 10(3): 4899-4914. doi: 10.3934/math.2025224 |
[3] | Junkang Tian, Zerong Ren, Shouming Zhong . Stability analysis for time delay systems via a generalized double integral inequality. AIMS Mathematics, 2020, 5(6): 6448-6456. doi: 10.3934/math.2020415 |
[4] | Deren Gong, Xiaoliang Wang, Peng Dong, Shufan Wu, Xiaodan Zhu . Shifted Legendre polynomials-based single and double integral inequalities with arbitrary approximation order: Application to stability of linear systems with time-varying delays. AIMS Mathematics, 2020, 5(5): 4371-4398. doi: 10.3934/math.2020279 |
[5] | Muhammad Tariq, Sotiris K. Ntouyas, Hijaz Ahmad, Asif Ali Shaikh, Bandar Almohsen, Evren Hincal . A comprehensive review of Grüss-type fractional integral inequality. AIMS Mathematics, 2024, 9(1): 2244-2281. doi: 10.3934/math.2024112 |
[6] | Stevo Stević . Norms of some operators between weighted-type spaces and weighted Lebesgue spaces. AIMS Mathematics, 2023, 8(2): 4022-4041. doi: 10.3934/math.2023201 |
[7] | Hanan Alohali, Sharief Deshmukh . Some generic hypersurfaces in a Euclidean space. AIMS Mathematics, 2024, 9(6): 15008-15023. doi: 10.3934/math.2024727 |
[8] | Hasanen A. Hammad, Hassen Aydi, Choonkil Park . Fixed point approach for solving a system of Volterra integral equations and Lebesgue integral concept in F$ _{\text{CM}} $-spaces. AIMS Mathematics, 2022, 7(5): 9003-9022. doi: 10.3934/math.2022501 |
[9] | Zengtai Gong, Xuyang Kou, Ting Xie . Interval-valued Choquet integral for set-valued mappings: definitions, integral representations and primitive characteristics. AIMS Mathematics, 2020, 5(6): 6277-6297. doi: 10.3934/math.2020404 |
[10] | Ling Peng, Qiong Liu . The construction conditions of a Hilbert-type local fractional integral operator and the norm of the operator. AIMS Mathematics, 2025, 10(1): 1779-1791. doi: 10.3934/math.2025081 |
The new schemes of calculation of double integrals and triple integrals are proposed in this paper. The formulas in which the double integral is converted into a line integral with respect to the arc length, and the triple integral is converted into a surface integral with respect to the area or a line integral with respect to the arc length are given separately. The effectiveness of the proposed methods is verified by several examples. Under certain conditions, these methods become the normal iterated integrals in Cartesian coordinate system or polar coordinate system, and the commonly used triple iterated integrals in Cartesian coordinate system, Cylindrical coordinate system or Spherical coordinate system. The transformation calculation method promoted in this paper points out the intrinsic relationship among double integral, triple integral, line integral and surface integral, which further enriches the theories of multivariate integrals.
Multivariable calculus [1] is a rich and fascinating subject. Multivariate integrals are widely used in various disciplines. For example, in optimization and (multidimensional) optimal control theory, authors [2,3] studied the optimization of some multiple or curvilinear integral functionals subject to ODEs, PDEs or isoperimetric constraints. In 2018, Mititelu and Treanţă [4] proposed optimality conditions in multiobjective control problems which involve multiple integrals. In the computation method of multivariate integrals, we often convert the double integral into iterated integrals in a Cartesian coordinate system by Fubini's Theorem or a polar coordinate system [5]. The triple integral is converted into iterated integrals in a Cartesian coordinate system, or a cylindrical coordinate system, or a spherical coordinate system. In general, a line integral is converted into a definite integral, and a surface integral is converted into a double integral. In addition, Green's formula, Gauss's formula and Stokes's formula could be applied to perform the calculation [6].
In the last twenty years, more researchers worked on the conversion relationships for several types of multivariate integrals. For example, the surface integral with respect to area is converted into the line integral with respect to the arc length in [7]. In [8], Li and Shi proposed a method of computing the surface integral by curvilinear integral in the special case when the projection of the surface on the coordinate plane was a curve. In the current paper, we propose new transformation algorithms for multivariate integrals, which further enhances the computation methods of multivariate integrals.
The main idea is to use the change of variables formula to transform double and triple integrals to iterated integrals which can be done more easily.
Theorem 1. (1) If the bounded closed region $ D $ is generated by a family of planar smooth curves
$ L_{t}: y = y(x, t), a \leq t \leq b, x_{1}(t) \leq x \leq x_{2}(t), y_{t}^{\prime} \neq 0 $ |
with no double point (i.e., a point traced out twice as a closed curve is traversed), and the function $ f(x, y) $ is continuous on the closed region $ D $, then
$ \begin{equation} \iint\limits_D f(x, y) \mathrm{d} x \mathrm{\; d} y = \int_{a}^{b} \mathrm{\; d} t \int_{x_{1}(t)}^{x_{2}(t)} f(x, y(x, t))\left|y_{t}^{\prime}\right| \mathrm{d} x \end{equation} $ | (2.1) |
$ \begin{equation} \iint\limits_D f(x, y) \mathrm{d} x \mathrm{\; d} y = \int_{a}^{b} \mathrm{\; d} t \int\limits_{L_{t}} f(x, y) \frac{\left|y_{t}^{\prime}\right|}{\sqrt{1+y_{x}^{\prime 2}}} \mathrm{\; d} s \end{equation} $ | (2.2) |
where the symbol 'ds' represents the arclength differential.
(2) If the bounded closed region $ D $ is generated by a family of planar smooth curves
$ L_{t}: x = x(y, t), a \leq t \leq b, y_{1}(t) \leq y \leq y_{2}(t), x_{t}^{\prime} \neq 0 $ |
with no double points, and the function $ f(x, y) $ is continuous on the closed region $ D $, then
$ \begin{equation} \iint\limits_D f(x, y) \mathrm{d} x \mathrm{\; d} y = \int_{a}^{b} \mathrm{\; d} t \int_{y_{1}(t)}^{y_{2}(t)} f(x(y, t), y)\left|x_{t}^{\prime}\right| \mathrm{d} y \end{equation} $ | (2.3) |
$ \begin{equation} \iint\limits_D f(x, y) \mathrm{d} x \mathrm{\; d} y = \int_{a}^{b} \mathrm{\; d} t \int\limits_{L_{t}} f(x, y) \frac{\left|x_{t}^{\prime}\right|}{\sqrt{1+x_{y}^{\prime 2}}} \mathrm{\; d} s \end{equation} $ | (2.4) |
Proof: Let transformation $ y = y(x, t) $, which transforms the $ D $ on the $ x O y $ plane to
$ D: a \leq t \leq b, x_{1}(t) \leq x \leq x_{2}(t) $ |
on the $ t O x $ plane. Notice
$ \frac{\partial(x, y)}{\partial(x, t)} = \left|\begin{array}{cc}1 & 0 \\y_{x}^{\prime} & y_{t}^{\prime}\end{array}\right| = y_{t}^{\prime} \neq 0. $ |
According to the integration by substitution, the formula (2.1) holds. On the other hand, since $ \mathrm{d} s = \sqrt{1+y_{x}^{\prime 2}} \mathrm{\; d} x $ on the curve $ L_{t} $,
$ \int\limits_{L_{t}} f(x, y) \frac{\left|y_{t}^{\prime}\right|}{\sqrt{1+y_{x}^{\prime 2}}} \mathrm{\; d} s = \int_{x_{1}(t)}^{x_{2}(t)} f(x, y(x, t))\left|y_{t}^{\prime}\right| \mathrm{d} x. $ |
The formula (2.2) can be obtained from (2.1). In the same way we can get (2.3) and (2.4).
Example 1. Calculate the double integral $ I = \iint\limits_D \sqrt{x^{2}+y^{2}} \mathrm{\; d} x \mathrm{\; d} y $, where the closed region $ D $ is bounded by three curves $ y = \sqrt{1-x^{2}}, x = \sqrt{y} $ and $ y = \sqrt{-x} $.
Method 1. Take a family of concentric circles arcs
$ L_{t}: y = \sqrt{t^{2}-x^{2}}, \frac{1}{2}-\sqrt{t^{2}+\frac{1}{4}} \leq x \leq \sqrt{\sqrt{t^{2}+\frac{1}{4}}-\frac{1}{2}}, $ |
where $ 0 \leq t \leq 1 $ as shown in Figure 1. Obviously, the central angle of each arc segment is $ \frac{\pi}{2} $, which results in the length of arc $ L_{t} $ is $ \frac{\pi t}{2} $. It follows from (2.2) that
$ I = \int_{0}^{1} \mathrm{\; d} t \int\limits_{L_{t}} \sqrt{x^{2}+y^{2}} \mathrm{\; d} s = \int_{0}^{1} \mathrm{\; d} t \int\limits_{L_{t}} t \mathrm{\; d} s = \int_{0}^{1} t \cdot \frac{1}{2} \pi t \mathrm{\; d} t = \frac{\pi}{6}. $ |
Method 2. Label regions $ D_{1}, D_{2} $ and $ D_{3} $ as shown in Figure 2. Rotate $ D_{1} $ clockwise by $ \frac{\pi}{2} $ to get $ D_{3} $. It is not difficult to find that $ D = D_{1} \cup D_{2} $ and $ D_{2} \cup D_{3}: \quad x^{2}+y^{2} \leq 1, \quad x \geq 0 \quad y \geq 0 $. Make a rotation transformation $ \left\{\begin{array}{l}x = -v, \\ y = u, \end{array}\right. $, then
$ \iint\limits_{D_{1}} \sqrt{x^{2}+y^{2}} \mathrm{\; d} x \mathrm{\; d} y = \iint\limits_{D_{3}} \sqrt{u^{2}+v^{2}} \mathrm{\; d} u \mathrm{\; d} v = \iint\limits_{D_{3}} \sqrt{x^{2}+y^{2}} \mathrm{\; d} x \mathrm{\; d} y. $ |
So,
$ \begin{aligned} I & = \iint\limits_{D_{1}} \sqrt{x^{2}+y^{2}} \mathrm{\; d} x \mathrm{\; d} y+\iint\limits_{D_{2}} \sqrt{x^{2}+y^{2}} \mathrm{\; d} x \mathrm{\; d} y = \iint\limits_{D_{3}} \sqrt{x^{2}+y^{2}} \mathrm{\; d} x \mathrm{\; d} y+\iint\limits_{D_{2}} \sqrt{x^{2}+y^{2}} \mathrm{\; d} x \mathrm{\; d} y \\ & = \iint\limits_{D_{2} \cup D_{3}} \sqrt{x^{2}+y^{2}} \mathrm{\; d} x \mathrm{\; d} y = \int_{0}^{\frac{\pi}{2}} \mathrm{\; d} \theta \int_{0}^{1} r \cdot r \mathrm{\; d} r = \frac{\pi}{6}. \end{aligned} $ |
Example 2. Calculate the double integral $ I = \iint\limits_D \frac{1}{x} \mathrm{\; d} x \mathrm{\; d} y $, where the closed region $ D $ is in the first quadrant and is bounded by two curves $ x y = 2, x y = 12 $, and two lines $ y = x+1, y = \frac{1}{4}(x+2) $, as shown in Figure 3.
Since the shape of the integral region D is very complicated (at least it needs to be represented as two sub-regions), as shown in Figure 4, the method of directly converting to the iterated integral in the Cartesian coordinate system or the polar coordinate system is quite cumbersome. Moreover, it is difficult to calculate with the integration by substitution.
The methods developed in Theorem 1 work well for this kind of regions.
Solution: Let's take a family of hyperbolics
$ L_{t}: y = \frac{t}{x}, \frac{\sqrt{4 t+1}-1}{2} \leq x \leq \sqrt{4 t+1}-1 (2 \leq t \leq 12) $ |
as shown in Figure 5.
It follows from (2.1) that
$ I = \int_{2}^{12} \mathrm{\; d} t \int_{\frac{\sqrt{4 t+1}-1}{2}}^{\sqrt{4 t+1}-1} \frac{1}{x} \cdot \frac{1}{x} \mathrm{\; d} x = \int_{2}^{12} \frac{1}{\sqrt{4 t+1}-1} \mathrm{\; d} t = 2+\frac{1}{2} \ln 3. $ |
Example 3. Calculate the double integral $ I = \iint\limits_D \frac{\sqrt{x^{2}+y^{2}}}{x} \mathrm{\; d} \sigma $, where the closed region $ D $ is bounded by four lines $ x = 1, x = 2, y = 0 $ and $ y = x $, as shown in Figure 6.
If we adopt the method of converting to the iterated integral in the polar coordinate system, the integral will eventually be transformed into the definite integral $ \int_{0}^{\pi / 4} \sec ^{3} x \mathrm{\; d} x $. And the calculation process is complicated.
Solution: Let's take a family of lines
$ L_{t}: y = t x, 1 \leq x \leq 2 (0 \leq t \leq 1) $ |
as shown in Figure 7. It follows from (2.1) that
$ I = \int_{0}^{1} \mathrm{\; d} t \int_{1}^{2} \frac{\sqrt{x^{2}+t^{2} x^{2}}}{x} \cdot|x| \mathrm{d} x = \int_{0}^{1} \sqrt{1+t^{2}} \mathrm{\; d} t \cdot \int_{1}^{2} x \mathrm{\; d} x = \frac{3}{4}(\sqrt{2}+\ln (1+\sqrt{2}). $ |
It can be seen from the above three examples that the calculation method of double integral varies. The calculation method proposed by Theorem 1 is an important component. This method can simplify the calculation process, which has certain theoretical significance for studying the multivariate integrals.
According to the differential method of the implicit function, the following inference is easily obtained by Theorem 1.
Corollary 1. The bounded closed region $ D $ is generated by a family of planar smooth curves
$ L_{t}: F(x, y, t) = 0, a \leq t \leq b\left(F_{t}^{\prime} \neq 0, F_{x}^{\prime 2}+F_{y}^{\prime 2} \neq 0\right) $ |
with no double points. If function $ f(x, y) $ is continuous on $ D $, then
$ \iint_{D} f(x, y) \mathrm{d} x \mathrm{\; d} y = \int_{a}^{b} \mathrm{\; d} t \int_{L_{t}} f(x, y) \frac{\left|F_{t}^{\prime}\right|}{\sqrt{F_{x}^{2}+F_{y}^{2}}} \mathrm{\; d} s. $ |
In addition, in Theorem 1, if the family of curves take the family of vertical lines, the family of horizontal lines and the family of lines passing through the origin respectively, three common calculation methods of double integral can be obtained.
Corollary 2. (1) The function $ f(x, y) $ is continuous on the bounded closed region
$ D_{x} = \left\{(x, y) \mid a \leq x \leq b, y_{1}(x) \leq y \leq y_{2}(x)\right\}, $ |
then
$ \begin{equation} \iint_{D_{x}} f(x, y) \mathrm{d} x \mathrm{\; d} y = \int_{a}^{b} \mathrm{\; d} x \int_{y_{1}(x)}^{y_{2}(x)} f(x, y) \mathrm{d} y. \end{equation} $ | (2.5) |
(2) The function $ f(x, y) $ is continuous on the bounded closed region
$ D_{y} = \left\{(x, y) \mid c \leq y \leq d, x_{1}(y) \leq x \leq x_{2}(y)\right\}, $ |
then
$ \begin{equation} \iint\limits_{D_{y}} f(x, y) \mathrm{d} x \mathrm{\; d} y = \int_{c}^{d} \mathrm{\; d} y \int_{x_{1}(y)}^{x_{2}(y)} f(x, y) \mathrm{d} x. \end{equation} $ | (2.6) |
(3) The function $ f(x, y) $ is continuous on the bounded closed region $ D $. If $ D $ can be expressed as
$ D = \left\{(r, \theta) \mid \alpha \leq \theta \leq \beta, r_{1}(\theta) \leq r \leq r_{2}(\theta)\right\} $ |
with the polar transformation $ x = r \cos \theta, y = r \sin \theta $, then
$ \iint\limits_D f(x, y) \mathrm{d} x \mathrm{\; d} y = \int_{\alpha}^{\beta} \mathrm{d} \theta \int_{r_{1}(\theta)}^{r_{2}(\theta)} f(r \cos \theta, r \sin \theta) \cdot r \mathrm{\; d} r. $ |
Proof: Taking the family of vertical lines
$ L_{t}: x = t, a \leq t \leq b, y_{1}(t) \leq y \leq y_{2}(t), $ |
we can know from (2.3)
$ \iint\limits_D f(x, y) \mathrm{d} x \mathrm{\; d} y = \int_{a}^{b} \mathrm{\; d} t \int_{y_{1}(t)}^{y_{2}(t)} f(t, y) \cdot 1 \mathrm{\; d} y = \int_{a}^{b} \mathrm{\; d} x \int_{y_{1}(x)}^{y_{2}(x)} f(x, y) \mathrm{d} y. $ |
That is, the formula (2.5) is established. Similarly, taking the family of horizontal lines
$ L_{t}: y = t, c \leq t \leq d, x_{1}(t) \leq x \leq x_{2}(t), $ |
we can get (2.6) from (2.1).
If we use
$ L_{\theta}: x = r \cos \theta, y = r \sin \theta, $ |
where
$ \alpha \leq \theta \leq \beta, r_{1}(\theta) \leq r \leq r_{2}(\theta), $ |
we can get $ y = x \tan \theta $ and
$ y_{\theta}^{\prime} = x \sec ^{2} \theta = \frac{r}{\cos \theta}. $ |
It follows that from (2.1)
$ \begin{aligned} \iint\limits_D f(x, y) \mathrm{d} x \mathrm{\; d} y & = \int_{\alpha}^{\beta} \mathrm{d} \theta \int_{r_{1}(\theta)}^{r_{2}(\theta)} f(r \cos \theta, r \sin \theta) \cdot \frac{r}{|\cos \theta|} \cdot|\cos \theta| \mathrm{d} r \\ & = \int_{\alpha}^{\beta} \mathrm{d} \theta \int_{r_{1}(\theta)}^{r_{2}(\theta)} f(r \cos \theta, r \sin \theta) \cdot r \mathrm{\; d} r. \end{aligned} $ |
Theorem 2. If the bounded closed region $ \Omega $ is generated by a family of smooth surfaces
$ \Sigma_{t}: z = z(x, y, t), a \leq t \leq b,(x, y) \in D_{x y}(t) $ |
with no double points, where $ D_{x y}(t) $ is the projection area of the surface $ \Sigma_{t} $ on $ x O y $ coordinate plane and $ z_{t}^{\prime} \neq 0 $, and the function $ f(x, y, z) $ is continuous on the closed region $ \Omega $, then
$ \begin{equation} \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z = \int_{a}^{b} \mathrm{\; d} t \iint\limits_{D_{x y}(t)} f(x, y, z(x, y, t))\left|z_{t}^{\prime}\right| \mathrm{d} x \mathrm{\; d} y, \end{equation} $ | (3.1) |
$ \begin{equation} \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z = \int_{a}^{b} \mathrm{\; d} t \iint\limits_{\Sigma_{t}} f(x, y, z) \frac{\left|z_{t}^{\prime}\right|}{\sqrt{1+z_{x}^{\prime 2}+z_{y}^{\prime 2}}} \mathrm{\; d} S. \end{equation} $ | (3.2) |
Proof: Let $ \left\{\begin{array}{l}x = x, \\ y = y, \\ z = z(x, y, t), \end{array}\right. $then the Jacobian determinant
$ \frac{\partial(x, y, z)}{\partial(x, y, t)} = \left|\begin{array}{ccc} 1 & 0 & 0 \\ 0 & 1 & 0 \\ z_{x}^{\prime} & z_{y}^{\prime} & z_{t}^{\prime} \end{array}\right| = z_{t}^{\prime} \neq 0. $ |
And it transforms $ \Omega $ in $ O-x y z $ to
$ \Omega: a \leq t \leq b,(x, y) \in D_{x y}(t) $ |
in $ O-x y t $. By substitution method of triple integral, we can obtain formula (3.1). Since $ \mathrm{d} S = \sqrt{1+z_{x}^{\prime 2}+z_{y}^{\prime 2}} \mathrm{\; d} x \mathrm{\; d} y $ on $ \Sigma_{t} $, we know
$ \iint\limits_{\Sigma_{t}} f(x, y, z) \frac{\left|z_{t}^{\prime}\right|}{\sqrt{1+z_{x}^{\prime 2}+z_{y}^{\prime 2}}} \mathrm{\; d} S = \iint\limits_{D_{x y}(t)} f(x, y, z(x, y, t))\left|z_{t}^{\prime}\right| \mathrm{d} x \mathrm{\; d} y. $ |
Substituting the above formula into (3.1) yields (3.2).
Corollary 3. (1) Let the bounded closed region $ \Omega $ be generated by a family of smooth surfaces
$ \Sigma_{t}: x = x(y, z, t), a \leq t \leq b,(y, z) \in D_{y z}(t), $ |
with no double points, where $ D_{y z}(t) $ is the projection area of the surface $ \Sigma_{t} $ on $ yOz $ coordinate plane and $ x_{t}^{\prime} \neq 0 $. If the function $ f(x, y, z) $ is continuous on the closed region $ \Omega $, then
$ \begin{equation} \begin{array}{l} \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z = \int_{a}^{b} \mathrm{\; d} t \iint\limits_{D_{yz}(t)} f(x(y, z, t), y, z)\left|x_{t}^{\prime}\right| \mathrm{d} y \mathrm{\; d} z, \\ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z = \int_{a}^{b} \mathrm{\; d} t \iint\limits_{\Sigma_{t}} f(x, y, z) \frac{\left|x_{t}^{\prime}\right|}{\sqrt{1+x_{y}^{\prime 2}+x_{z}^{\prime 2}}} \mathrm{\; d} S. \end{array} \end{equation} $ | (3.3) |
(2) Let the bounded closed region $ \Omega $ be generated by a family of smooth surfaces
$ \Sigma_{t}: y = y(x, z, t), a \leq t \leq b, (x, z) \in D_{x z}(t), $ |
with no double points, where $ D_{x z}(t) $ is the projection area of the surface $ \Sigma_{t} $ on $ xOz $ coordinate plane and $ y_{t}^{\prime} \neq 0 $. If the function $ f(x, y, z) $ is continuous on the closed region $ \Omega $, then
$ \begin{equation} \begin{array}{l} \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z = \int_{a}^{b} \mathrm{\; d} t \iint\limits_{D_{xz}(t)} f(x, y(x, z, t), z)\left|y_{t}^{\prime}\right| \mathrm{d} x \mathrm{\; d} z, \\ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z = \int_{a}^{b} \mathrm{\; d} t \iint\limits_{\Sigma_{t}} f(x, y, z) \frac{\left|y_{t}^{\prime}\right|}{\sqrt{1+y_{x}^{\prime 2}+y_{z}^{\prime 2}}} \mathrm{\; d} S. \end{array} \end{equation} $ | (3.4) |
Example 4. Calculate the triple integral $ I = \iiint\limits_{\Omega}(x+y+z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z $, where
$ \Omega: \frac{1}{2} \leq x+y+z \leq 1, x \geq 0, y \geq 0, z \geq 0, $ |
as shown in Figure 8.
Method 1. As shown in Figure 9, Taking a family of triangle planes
$ \Sigma_{t}: z = t-x-y, \frac{1}{2} \leq t \leq 1,0 \leq x \leq 1,0 \leq y \leq t-x, $ |
then $ \left|z_{t}^{\prime}\right| = 1, \sqrt{1+z_{x}^{\prime 2}+z_{y}^{\prime 2}} = \sqrt{3} $ hold, and the area of $ \Sigma_{t} $ is $ \frac{\sqrt{3}}{2} t^{2} $. Substituting the above formula into (3.2) gives
$ I = \int_{\frac{1}{2}}^{1} \mathrm{\; d} t \iint\limits_{\Sigma_{t}}(x+y+z) \frac{\mathrm{d} S}{\sqrt{3}} = \frac{1}{\sqrt{3}} \int_{\frac{1}{2}}^{1} t \mathrm{\; d} t \iint\limits_{\Sigma_{t}} \mathrm{\; d} S = \frac{1}{\sqrt{3}} \int_{\frac{1}{2}}^{1} t \cdot \frac{\sqrt{3}}{2} t^{2} \cdot \mathrm{d} t = \frac{15}{128}. $ |
According to the additivity of the integral region, this integral can also be calculated as follows.
Method 2. Let
$ \begin{array}{c} \Omega_{1} = \{(x, y, z) \mid 0 \leq z \leq 1-x-y, 0 \leq y \leq 1-x, 0 \leq x \leq 1\}, \\ \Omega_{2} = \left\{(x, y, z) \mid 0 \leq z \leq \frac{1}{2}-x-y, 0 \leq y \leq \frac{1}{2}-x, 0 \leq x \leq \frac{1}{2}\right\}, \end{array} $ |
then $ \Omega = \Omega_{1}-\Omega_{2} $. According to the symmetry of the triple integral, we have
$ \begin{aligned} I & = \iiint\limits_{\Omega} 3 x \mathrm{\; d} x \mathrm{\; d} y \mathrm{\; d} z = \iiint\limits_{\Omega_{1}} 3 x \mathrm{\; d} x \mathrm{\; d} y \mathrm{\; d} z-\iiint\limits_{\Omega_{2}} 3 x \mathrm{\; d} x \mathrm{\; d} y \mathrm{\; d} z \\ & = \int_{0}^{1} \mathrm{\; d} x \int_{0}^{1-x} \mathrm{\; d} y \int_{0}^{1-x-y} 3 x \mathrm{\; d} z-\iint_{0}^{\frac{1}{2}} \mathrm{\; d} x \int_{0}^{\frac{1}{2}-x} \mathrm{\; d} y \int_{0}^{\frac{1}{2}-x-y} 3 x \mathrm{\; d} z \\ & = \int_{0}^{1} 3 x \mathrm{\; d} x \int_{0}^{1-x}(1-x-y) \mathrm{d} y-\int_{0}^{\frac{1}{2}} 3 x \mathrm{\; d} x \int_{0}^{\frac{1}{2}-x}\left(\frac{1}{2}-x-y\right) \mathrm{d} y \\ & = \int_{0}^{1} \frac{3}{2}\left(x-2 x^{2}+x^{3}\right) \mathrm{d} x-\int_{0}^{\frac{1}{2}} \frac{3}{2}\left(\frac{1}{4} x-x^{2}+x^{3}\right) \mathrm{d} x \\ & = \frac{1}{8}-\frac{1}{128} = \frac{15}{128}. \end{aligned} $ |
Example 5. Calculate the triple integral $ I = \iiint\limits_{\Omega} \sqrt{x^{2}+y^{2}+z^{2}} \mathrm{\; d} x \mathrm{\; d} y \mathrm{\; d} z $, where
$ \Omega: x^{2}+y^{2}+z^{2} \leq 1,-z \leq x+y \leq 2 z. $ |
Method 1. Taking a family of spheres
$ \Sigma_{t}: z = \sqrt{t^{2}-x^{2}-y^{2}}, 0 \leq t \leq 1,-z \leq x+y \leq 2 z, $ |
then $ \left|z_{t}^{\prime}\right| = \frac{t}{\sqrt{t^{2}-x^{2}-y^{2}}}, \sqrt{1+z_{x}^{\prime 2}+z_{y}^{\prime 2}} = \frac{t}{\sqrt{t^{2}-x^{2}-y^{2}}} $ hold. In addition, the plane $ x+y+z = 0 $ and the plane $ 2z-x-y = 0 $ are perpendicular to each other, so the area of $ \Sigma_{t} $ is $ \pi t^{2} $. Substituting the above results into (3.2) gives
$ I = \int_{0}^{1} \mathrm{\; d} t \iint\limits_{\Sigma_{t}} \sqrt{x^{2}+y^{2}+z^{2}} \mathrm{\; d} S = \int_{0}^{1} \mathrm{\; d} t \iint\limits_{\Sigma_{t}} t \mathrm{\; d} S = \int_{0}^{1} t \cdot \pi t^{2} \mathrm{\; d} t = \pi \int_{0}^{1} t^{3} \mathrm{\; d} t = \frac{\pi}{4}. $ |
Method 2. Let a orthogonal transformation $ u = \frac{1}{\sqrt{2}}(x-y), v = \frac{1}{\sqrt{3}}(x+y+z), w = \frac{1}{\sqrt{6}}(2 z-x-y). $ And it transforms $ \Omega $ in $ O-x y z $ to
$ \Omega: u^{2}+v^{2}+w^{2} \leq 1, v \geq 0, w \geq 0 $ |
in $ O-uvw $, thus
$ I = \iiint\limits_{\Omega} \sqrt{u^{2}+v^{2}+w^{2}} \mathrm{\; d} u \mathrm{\; d} v \mathrm{\; d} w. $ |
Using spherical coordinates, we can obtain
$ I = \int_{0}^{\pi} d \theta \int_{0}^{\frac{\pi}{2}} d \varphi \int_{0}^{1} \rho \cdot \rho^{2} \sin \varphi d \rho = \int_{0}^{\pi} d \theta \cdot \int_{0}^{\frac{\pi}{2}} \sin \varphi d \varphi \cdot \int_{0}^{1} \rho^{3} d \rho = \frac{\pi}{4}. $ |
Example 6. Calculate the triple integral $ I = \iiint\limits_{\Omega} \sqrt{x^{2}+y^{2}} \mathrm{\; d} x \mathrm{\; d} y \mathrm{\; d} z $, where
$ \Omega: 0 \leq z \leq \sqrt{x^{2}+y^{2}}, x^{2}+y^{2} \leq 1, x \geq 0, y \geq 0. $ |
Method 1. Taking a family of cylindrical surfaces
$ \Sigma_{t}: x^{2}+y^{2} = t^{2}, x \geq 0, y \geq 0,0 \leq z \leq t, 0 \leq t \leq 1 $ |
then the area of $ \Sigma_{t} $ is $ \frac{1}{2} \pi t^{2} $. Because of $ x = \sqrt{t^{2}-y^{2}} $, $ \frac{\left|x_{t}^{\prime}\right|}{\sqrt{1+x_{y}^{\prime 2}+x_{z}^{\prime 2}}} = 1 $ holds. Substituting the above results into (3.3) gives
$ I = \int_{0}^{1} \mathrm{\; d} t \iint\limits_{\Sigma_{t}} \sqrt{x^{2}+y^{2}} \mathrm{\; d} S = \int_{0}^{1} \mathrm{\; d} t \iint\limits_{\Sigma_{t}} t \mathrm{\; d} S = \int_{0}^{1} t \cdot \frac{1}{2} \pi t^{2} \mathrm{\; d} t = \frac{\pi}{8}. $ |
Method 2. Taking a family of conic surfaces
$ \Sigma_{t}: z = t \sqrt{x^{2}+y^{2}}, x^{2}+y^{2} \leq 1, x \geq 0, y \geq 0,0 \leq t \leq 1. $ |
then $ z_{t}^{\prime} = \sqrt{x^{2}+y^{2}} $ holds. The projection area on the xoy coordinate plane:
$ D_{x y}(t): x^{2}+y^{2} \leq 1, x \geq 0, y \geq 0. $ |
Substituting the above results into (3.1) gives
$ I = \int_{0}^{1} \mathrm{\; d} t \iint\limits_{D_{xy}(t)} \sqrt{x^{2}+y^{2}} \cdot\left|\sqrt{x^{2}+y^{2}}\right| \mathrm{d} x \mathrm{\; d} y = \int_{0}^{1} \mathrm{\; d} t \int_{0}^{\frac{\pi}{2}} \mathrm{\; d} \theta \int_{0}^{1} r^{3} \mathrm{\; d} r = \frac{\pi}{8}. $ |
Similarly, in Theorem 2, if we choose three kinds of special plane families, three common calculation methods of triple integral will be obtained.
Corollary 4. (1) If the function $ f(x, y, z) $ is continuous on the bounded closed region
$ \Omega = \left\{(x, y, z) \mid a \leq z \leq b,(x, y) \in D_{z}\right\}, $ |
then
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z = \int_{c}^{d} d z \iint\limits_{D_{z}} f(x, y, z) \mathrm{d} x \mathrm{\; d} y. $ |
(2) The function $ f(x, y, z) $ is continuous on the bounded closed region $ \Omega $. If $ \Omega $ can be expressed as
$ \Omega = \left\{(r, \theta, z) \mid \alpha \leq \theta \leq \beta, r_{1}(\theta) \leq r \leq r_{2}(\theta), z_{1}(r, \theta) \leq z \leq z_{2}(r, \theta)\right\} $ |
with the cylindrical coordinate transformation $ x = r \cos \theta, y = r \sin \theta, z = z $, then
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \int_{\alpha}^{\beta} \mathrm{d} \theta \int_{r_{1}(\theta)}^{r_{2}(\theta)} \mathrm{d} r \int_{z_{1}(r, \theta)}^{z_{2}(r, \theta)} f(r \cos \theta, r \sin \theta, z) r \mathrm{\; d} z. $ |
(3) The function $ f(x, y, z) $ is continuous on the bounded closed region $ \Omega $. If $ \Omega $ can be expressed as
$ \Omega = \left\{(\theta, \varphi, \rho) \mid \alpha \leq \theta \leq \beta, \varphi_{1}(\theta) \leq \varphi \leq \varphi_{2}(\theta), \rho_{1}(\varphi, \theta) \leq \rho \leq \rho_{2}(\varphi, \theta)\right\} $ |
with the spherical coordinate transformation $ x = \rho \sin \varphi \cos \theta, y = \rho \sin \varphi \sin \theta, z = \rho \cos \varphi $, then
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \int_{\alpha}^{\beta} \mathrm{d} \theta \int_{\varphi_{1}(\theta)}^{\varphi_{2}(\theta)} \mathrm{d} \varphi \int_{\rho_{1}(\varphi, \theta)}^{\rho_{2}(\varphi, \theta)} f(\rho \sin \varphi \cos \theta, \rho \sin \varphi \sin \theta, \rho \cos \varphi) \rho^{2} \sin \varphi \mathrm{d} \rho. $ |
Proof: (1) In Theorem 2, we take plane family
$ \Sigma_{t}:\left\{\begin{array}{ll} x = x, \\ y = y, & c \leq t \leq d,(x, y) \in D_{x y}(t) \\ z = z(x, y, t) = t \end{array}\right. $ |
where $ D_{x y}(t) $ is the projection area of the cross section of $ \Sigma_{z} $ and $ \Omega $ on the $ xOy $ coordinate plane. It is easy to calculate $ z_{t}^{\prime} = 1 $. Substituting this result into (3.1) gives
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z = \int_{c}^{d} d z \iint_{D_{z}} f(x, y, z) \mathrm{d} x \mathrm{\; d} y. $ |
(2) In Corollary 3, we take plane family
$ \Sigma_{\theta}:\left\{\begin{array}{l} x = r \cos \theta, \\ y = y(r, z, \theta) = r \sin \theta, \alpha \leq \theta \leq \beta,(r, z) \in D_{r z}(\theta) \\ z = z, \end{array}\right. $ |
where
$ D_{r z}(\theta) = \left\{(r, z) \mid r_{1}(\theta) \leq r \leq r_{2}(\theta), z_{1}(r, \theta) \leq z \leq z_{2}(r, \theta)\right\} $ |
is the projection area of the cross section of $ \Sigma_{\theta} $ and $ \Omega $ on the $ rOz $ plane. Obviously, $ y = x \tan \theta $ is true. Thus $ y_{\theta}^{\prime} = \frac{x}{\cos ^{2} \theta} = \frac{r}{\cos \theta} $. Substituting this result into (3.4) gives
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \int_{\alpha}^{\beta} \mathrm{d} \theta \iint\limits_{D_{xz}(\theta)} f(x, y(x, z, \theta), z) \cdot \frac{r}{|\cos \theta|} \mathrm{d} x \mathrm{\; d} z $ |
Since $ \frac{\partial(x, z)}{\partial(r, z)} = \left|\begin{array}{cc}\cos \theta & 0 \\ 0 & 1\end{array}\right| = \cos \theta $, $ \mathrm{d} x \mathrm{\; d} z = |\cos \theta| \mathrm{d} r \mathrm{\; d} z $ holds. Thus
$ \begin{array}{l} \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \int_{\alpha}^{\beta} \mathrm{d} \theta \iint\limits_{D_{rz}(\theta)} f(r \cos \theta, r \sin \theta, z) \cdot r \mathrm{\; d} r \mathrm{\; d} z. \\ \quad = \int_{\alpha}^{\beta} \mathrm{d} \theta \int_{r_{1}(\theta)}^{r_{2}(\theta)} \mathrm{d} r \int_{z_{1}(r, \theta)}^{z_{2}(r, \theta)} f(r \cos \theta, r \sin \theta, z) r \mathrm{\; d} z. \end{array} $ |
(3) In Corollary 3, we take plane family
$ \Sigma_{\theta}:\left\{\begin{array}{l} x = \rho \sin \varphi \cos \theta, \\ y = y(\rho, \varphi, \theta) = \rho \sin \varphi \sin \theta, \alpha \leq \theta \leq \beta,(\varphi, \rho) \in D_{\varphi \rho}(\theta) \\ z = \rho \cos \varphi, \end{array}\right. $ |
where
$ D_{\varphi \rho}(\theta) = \left\{(\varphi, \rho) \mid \varphi_{1}(\theta) \leq \varphi \leq \varphi_{2}(\theta), \rho_{1}(\varphi, \theta) \leq \rho \leq \rho_{2}(\varphi, \theta)\right\} $ |
is the projection area of the cross section of $ \Sigma_{\theta} $ and $ \Omega $ on the $ \varphi O \rho $ plane. Obviously, $ y = x \tan \theta $ is true. Thus $ y_{\theta}^{\prime} = \frac{x}{\cos ^{2} \theta} = \rho \frac{\sin \varphi}{\cos \theta} $. Substituting this result into (3.4) gives
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \int_{\alpha}^{\beta} \mathrm{d} \theta \iint\limits_{D_{xz}(\theta)} f(x, y(x, z, \theta), z) \cdot \frac{\rho \sin \varphi}{|\cos \theta|} \mathrm{d} x \mathrm{\; d} z $ |
Since $ \frac{\partial(x, z)}{\partial(\rho, \varphi)} = \left|\begin{array}{cc}\sin \varphi \cos \theta & \cos \varphi \\ \rho \cos \varphi \cos \theta & -\rho \sin \varphi\end{array}\right| = \rho \cos \theta $, $ \mathrm{d} x \mathrm{\; d} z = \rho|\cos \theta| \mathrm{d} \rho \mathrm{d} \varphi $ holds. Thus
$ \begin{array}{l} \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \int_{\alpha}^{\beta} \mathrm{d} \theta \iint\limits_{D_{\varphi \rho}(\theta)} f(\rho \sin \varphi \cos \theta, \rho \sin \varphi \sin \theta, \rho \cos \varphi) \cdot \rho \sin \varphi \mathrm{d} \varphi \mathrm{d} \rho. \\ \quad = \int_{\alpha}^{\beta} \mathrm{d} \theta \int_{\varphi_{1}(\theta)}^{\varphi_{2}(\theta)} \mathrm{d} \varphi \int_{\rho_{1}(\varphi, \theta)}^{\rho_{2}(\varphi, \theta)} f(\rho \sin \varphi \cos \theta, \rho \sin \varphi \sin \theta, \rho \cos \varphi) \rho^{2} \sin \varphi \mathrm{d} \rho. \end{array} $ |
Theorem 3. Let the bounded closed region $ \Omega $ be generated by a family of smooth surfaces
$ \Sigma_{t}: z = z(x, y, t), a \leq t \leq b, x_{1}(t) \leq x \leq x_{2}(t) $ |
with no double points and $ z_{t}^{\prime} \neq 0 $. Let $ \Gamma_{x t} $ be the curves where the planes, which pass through the point $ (x, 0, 0) $ and are perpendicular to the X-axis, and surface $ \Sigma_{t} $ intersect. If the function $ f(x, y, z) $ is continuous on the closed region $ \Omega $, then
$ \begin{equation} \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \iint\limits_{D} \mathrm{\; d} x \mathrm{\; d} t \int_{\Gamma_{x t}} f(x, y, z) \frac{\left|z_{t}^{\prime}\right|}{\sqrt{1+z_{y}^{\prime 2}}} \mathrm{\; d} s, \end{equation} $ | (4.1) |
where $ D = \left\{(t, x) \mid a \leq t \leq b, x_{1}(t) \leq x \leq x_{2}(t)\right\} $.
Proof: Let $ x = x, y = y, z = z(x, y, t), $ then the Jacobian determinant
$ \frac{\partial(x, y, z)}{\partial(x, y, t)} = \left|\begin{array}{ccc} 1 & 0 & 0 \\ 0 & 1 & 0 \\ z_{x}^{\prime} & z_{y}^{\prime} & z_{t}^{\prime} \end{array}\right| = z_{t}^{\prime} \neq 0. $ |
And it transforms $ \Omega $ to
$ \Omega^{\prime} = \left\{(x, y, z) \mid a \leq t \leq b, x_{1}(t) \leq x \leq x_{2}(t), y_{1}(x, t) \leq y \leq y_{2}(x, t)\right\}. $ |
By substitution method of triple integral we can obtain
$ \begin{aligned} \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z & = \iiint\limits_{\Omega^{\prime}} f(x, y, z(x, y, t))\left|z_{t}^{\prime}\right| \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} t \\ & = \int_{a}^{b} \mathrm{\; d} t \int_{x_{1}(t)}^{x_{2}(t)} \mathrm{d} x \int_{y_{1}(x, t)}^{y_{2}(x, t)} f(x, y, z(x, y, t))\left|z_{t}^{\prime}\right| \mathrm{d} y. \end{aligned} $ |
On the other hand,
$ \int\limits_{\Gamma_{x t}} f(x, y, z) \frac{\left|z_{t}^{\prime}\right|}{\sqrt{1+z_{y}^{\prime 2}}} \mathrm{\; d} s = \int_{y_{1}(x, t)}^{y_{2}(x, t)} f(x, y, z(x, y, t))\left|z_{t}^{\prime}\right| \mathrm{d} y, $ |
then
$ \iint\limits_{D} \mathrm{\; d} x \mathrm{\; d} t \int\limits_{\Gamma_{xt}} f(x, y, z) \frac{\left|z_{t}^{\prime}\right|}{\sqrt{1+z_{y}^{\prime 2}}} \mathrm{\; d} s = \int_{a}^{b} \mathrm{\; d} t \int_{x_{1}(t)}^{x_{2}(t)} \mathrm{d} x \int_{y_{1}(x, t)}^{y_{2}(x, t)} f(x, y, z(x, y, t))\left|z_{t}^{\prime}\right| \mathrm{d} y. $ |
thus
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \iint\limits_{D} \mathrm{\; d} x \mathrm{\; d} t \int_{\Gamma_{x t}} f(x, y, z) \frac{\left|z_{t}^{\prime}\right|}{\sqrt{1+z_{y}^{\prime 2}}} \mathrm{\; d} s. $ |
Example 7. Calculate the triple integral $ I = \iiint\limits_{\Omega} z \mathrm{\; d} v $, where $ \Omega: x^{2}+y^{2}+z^{2} \leq 1, z \geq|y| $.
Method 1. Taking the surface family
$ \Sigma_{t}: z = \sqrt{t^{2}-x^{2}-y^{2}} (0 \leq t \leq 1). $ |
For any given $ t \in[0, 1] $, $ x \in[-t, t] $ and
$ -\frac{1}{\sqrt{2}} \sqrt{t^{2}-x^{2}} \leq y \leq \frac{1}{\sqrt{2}} \sqrt{t^{2}-x^{2}}. $ |
Let $ \Gamma_{x t} $ be the curves where the planes, which pass through the point $ (x, 0, 0) $ and are perpendicular to the X-axis, and surface $ \Sigma_{t} $ intersect. That means
$ \Gamma_{x t}:\left\{\begin{array}{l} x = x, \\ z = \sqrt{t^{2}-x^{2}-y^{2}}. \end{array}\right. $ |
Then $ z_{t}^{\prime} = \frac{t}{\sqrt{t^{2}-x^{2}-y^{2}}} $, thus
$ \int\limits_{\Gamma_{x t}} z \cdot \frac{\left|z_{t}^{\prime}\right|}{\sqrt{1+z_{y}^{\prime 2}}} \mathrm{\; d} s = \int_{-\frac{1}{\sqrt{2}} \sqrt{t^{2}-x^{2}}}^{-\frac{1}{\sqrt{2}} \sqrt{t^{2}-x^{2}}} t \mathrm{\; d} y = \sqrt{2 t} \sqrt{t^{2}-x^{2}}. $ |
Substituting the above formula into (4.1) gives
$ I = \iint\limits_{D} \sqrt{2} t \sqrt{t^{2}-x^{2}} \mathrm{\; d} x \mathrm{\; d} t, $ |
where $ D = \{(t, x) \mid 0 \leq t \leq 1, -t \leq x \leq t\} $. So,
$ I = \int_{0}^{1} \mathrm{\; d} t \int_{-t}^{t} \sqrt{2} t \sqrt{t^{2}-x^{2}} \mathrm{\; d} x = \frac{1}{\sqrt{2}} \pi \int_{0}^{1} t^{3} \mathrm{\; d} t = \frac{\sqrt{2}}{8} \pi. $ |
Method 2. According to the symmetry of the integral, it is easy to see that
$ I = \iiint\limits_{\Omega} z \mathrm{\; d} v = 2 \iiint\limits_{\Omega^{\prime}} z \mathrm{\; d} v, $ |
where
$ \Omega^{\prime}: x^{2}+y^{2}+z^{2} \leq 1, z \geq y \geq 0. $ |
And the projection area of $ \Omega^{\prime} $ on $ xOy $ plane is
$ D_{x y} = \left\{(x, y) \mid x^{2}+2 y^{2} \leq 1, y \geq 0\right\}, $ |
thus
$ I = 2 \iint\limits_{D_{x y}} d x d y \int_{y}^{\sqrt{1-x^{2}-y^{2}}} z d z = \iint\limits_{D_{x y}}\left(1-x^{2}-2 y^{2}\right) d x d y. $ |
Using the generalized polar transformation, the above integral can be calculated as follows:
$ I = \int_{0}^{\pi} d \theta \int_{0}^{1}\left(1-r^{2}\right) \frac{1}{\sqrt{2}} r d r = \frac{\sqrt{2}}{8} \pi. $ |
Corollary 5. (1) The bounded closed region $ \Omega $ is generated by a family of smooth surfaces
$ \Sigma_{t}: z = z(x, y, t), a \leq t \leq b, y_{1}(t) \leq y \leq y_{2}(t) $ |
with no double points and $ z_{t}^{\prime} \neq 0 $. Let $ \Gamma_{y t} $ be the curves where the planes, which pass through the point $ (0, y, 0) $ and are perpendicular to the Y-axis, and surface $ \Sigma_{t} $ intersect. If the function $ f(x, y, z) $ is continuous on the closed region $ \Omega $, then
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \iint\limits_{D} \mathrm{\; d} y \mathrm{\; d} t \int\limits_{\Gamma_{yt}} f(x, y, z) \frac{\left|z_{t}^{\prime}\right|}{\sqrt{1+z_{x}^{\prime 2}}} \mathrm{\; d} s, $ |
where $ D = \left\{(t, y) \mid a \leq t \leq b, y_{1}(t) \leq y \leq y_{2}(t)\right\} $.
(2) The bounded closed region $ \Omega $ is generated by a family of smooth surfaces
$ \Sigma_{t}: y = y(x, z, t), a \leq t \leq b, x_{1}(t) \leq x \leq x_{2}(t) $ |
with no double points and $ y_{t}^{\prime} \neq 0 $. Let $ \Gamma_{x t} $ be the curves where the planes, which pass through the point $ (x, 0, 0) $ and are perpendicular to the X-axis, and surface $ \Sigma_{t} $ intersect. If the function $ f(x, y, z) $ is continuous on the closed region $ \Omega $, then
$ \begin{equation} \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \iint\limits_{D} \mathrm{\; d} x \mathrm{\; d} t \int\limits_{\Gamma_{x t}} f(x, y, z) \frac{\left|y_{t}^{\prime}\right|}{\sqrt{1+y_{z}^{\prime 2}}} \mathrm{\; d} s, \end{equation} $ | (4.2) |
where $ D = \left\{(t, x) \mid a \leq t \leq b, x_{1}(t) \leq x \leq x_{2}(t)\right\} $.
In fact, there are several other cases of Theorem 3, which are not repeated here.
Method 3 of Example 6. Taking a family of cylindrical surfaces
$ \Sigma_{t}: y = \sqrt{t^{2}-x^{2}}, \quad D = \{(t, x) \mid 0 \leq t \leq 1,0 \leq x \leq t\}. $ |
then $ \frac{\left|y_{t}^{\prime}\right|}{\sqrt{1+y_{z}^{\prime 2}}} = \frac{t}{\sqrt{t^{2}-x^{2}}} $ holds. Let $ \Gamma_{x t} $ be $ \left\{\begin{array}{l}x = x, \\ y = \sqrt{t^{2}-x^{2}}, 0 \leq x \leq t \\ z = t, \end{array}\right. $. And the length of the straight-line segment $ \Gamma_{x t} $ is $ t $. Substituting the above results into (4.2) gives
$ I = \iint\limits_{D} \mathrm{\; d} x \mathrm{\; d} t \int\limits_{\Gamma_{x t}} \sqrt{x^{2}+y^{2}} \cdot \frac{t}{\sqrt{t^{2}-x^{2}}} \mathrm{\; d} s = \iint\limits_{D} \mathrm{\; d} x \mathrm{\; d} t \int\limits_{\Gamma_{x t}} \frac{t^{2}}{\sqrt{t^{2}-x^{2}}} \mathrm{\; d} s = \iint\limits_{D} \frac{t^{2}}{\sqrt{t^{2}-x^{2}}} \cdot t \mathrm{\; d} x \mathrm{\; d} t = \frac{\pi}{8}. $ |
Corollary 6. Let the function $ f(x, y, z) $ be continuous on the bounded closed region
$ \Omega = \left\{(x, y, z) \mid z_{1}(x, y) \leq z \leq z_{2}(x, y),(x, y) \in D\right\}, $ |
then
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} x \mathrm{\; d} y \mathrm{\; d} z = \iint\limits_{D} \mathrm{\; d} x \mathrm{\; d} y \int_{z_{1}(x, y)}^{z_{2}(x, y)} f(x, y, z) \mathrm{d} z, $ |
where $ D $ is the projection area of $ \Omega $ on the $ xOy $ coordinate plane.
Proof: In the Corollary 5(2), if we take
$ \Sigma_{t}: y = y(x, z, t) = t, $ |
$ \Gamma_{x t} $ is straight lines $ \Gamma_{x y} $ parallel to the Z-axis. Notice
$ y_{t}^{\prime} = 1, y_{z}^{\prime} = 0, \mathrm{\; d} s = \mathrm{d} z, $ |
It follows from (4.2) that
$ \iiint\limits_{\Omega} f(x, y, z) \mathrm{d} v = \iint\limits_{D} \mathrm{\; d} x \mathrm{\; d} y \int\limits_{\Gamma_{x y}} f(x, y, z) \frac{1}{\sqrt{1+0^{2}}} \mathrm{\; d} s = \iint\limits_{D} \mathrm{\; d} x \mathrm{\; d} y \int_{z_{1}(x, y)}^{z_{2}(x, y)} f(x, y, z) \mathrm{d} z. $ |
In this paper we establish several formulas for converting the double integral to a line integral with respect to arc length, and the triple integral to a surface integral with respect to area or a line integral with respect to arc length. Some commonly used calculation methods are special cases of our methods. Examples show that the methods presented here are simple and effective in computing certain complex double integrals and triple integrals.
This work was supported in part by the National Natural Science Foundation of China under Grant Nos. 12001151. The authors would like to thanks to the experts and scholars who took time out of their busy schedules to review this paper which greatly help improve the clarity and quality of the paper.
The authors declare that they have no conflict of interest.
[1] | P. Dyke, Two and three dimensional calculus with applications in science and engineering, Wiley, 2018. |
[2] |
S. Treanţă, Constrained variational problems governed by second-order Lagrangians, Appl. Anal., 99 (2020), 1467–1484. doi: 10.1080/00036811.2018.1538501
![]() |
[3] |
S. Treanţă, On a modified optimal control problem with first-order PDE constraints and the associated saddle-point optimality criterion, Eur. J. Control, 51 (2020), 1–9. doi: 10.1016/j.ejcon.2019.07.003
![]() |
[4] |
Ş. Mititelu, S. Treanţă, Efficiency conditions in vector control problems governed by multiple integrals, J. Appl. Math. Comput., 57 (2018), 647–665. doi: 10.1007/s12190-017-1126-z
![]() |
[5] | R. Larson, B. Edwards, Calculus, 11 Eds., Cengage Learning Inc., 2014. |
[6] | S. Zhu, S. Tang, R. Ning, P. Ren, Z. Yin, Advanced mathematics, Beijing: Higher Education Press, (2015), 127–252. |
[7] | Y. Peng, X. Ma, R. Ning, Calculation method of surface integral with respect to area to line integral with respect to arc length, Stud. Coll. mathe., 13 (2010), 61–63. |
[8] | Y. Li, R. Shi, The application of curvilinear integral to surface integral, Coll. math., 19 (2003), 106–108. |
1. | Gauhar Rahman, Saud Fahad Aldosary, Muhammad Samraiz, Kottakkaran Sooppy Nisar, Some Double Generalized Weighted Fractional Integral Inequalities Associated with Monotone Chebyshev Functionals, 2021, 5, 2504-3110, 275, 10.3390/fractalfract5040275 |