Divergence and Curl
볡μμ 곡νκ³ μλ μνκ³Όμ μ‘Έμ μνμ μν΄ νλΆ μν κ³Όλͺ©λ€μ λ€μ 곡λΆνκ³ μμ΅λλ€. λ―Έμ λΆν ν¬μ€νΈ μ 체 보기
μ΄μ©λ©΄ μ§κΈ μ΄ μνμΌμ§λβ¦
λλμ΄!! λ―Έμ 2μ κ½μ΄λΌκ³ ν μ μλ λ°μ°(Divergence)κ³Ό νμ (Curl)μ΄λ€!! λ κ°λ λͺ¨λ 벑ν°μ₯μ μ 체μ νλ¦μ΄λΌκ³ λ³΄κ³ , κ·Έ μ μ²΄κ° ν μ μ λν΄ ν½μ°½/μμΆ νκ³ μλμ§, μλλ©΄ ν μ μμ μ μ²΄κ° μΌλ§λ λΉ λ₯΄κ² νμ νκ³ μλμ§λ₯Ό μΈ‘μ νλ μ§νμ΄λ€. π
Divergence
If $\mathbf{F} = P \, \mathbf{i} + Q \, \mathbf{j} + R \, \mathbf{k}$ is a vector field on $\mathbb{R}^3$,
and the below partial derivatives exist
\[\frac{\partial P}{\partial x} \text{ , } \frac{\partial Q}{\partial y} \text{ , } \frac{\partial R}{\partial z}\]then the βdivergence of $\mathbf{F}$β is a scalar value defined by
\[\text{div } \mathbf{F} = \frac{\partial P}{\partial x} + \frac{\partial Q}{\partial y} + \frac{\partial R}{\partial z}\]λ°μ°(divergence)λ 벑ν°μ₯μ λν΄ μ μνλ μ€μΉΌλΌ ν¨μμ΄λ€. Gradient Vector $\nabla$μμ λ΄μ μΌλ‘ μμ λ μ½κ² ννν μ μλ€.
\[\text{div } \mathbf{F} = \nabla \cdot \mathbf{F}\]Geometric meaning
벑ν°μ₯μ μ 체μ νλ¦ μμ΄λΌκ³ νννλ€λ©΄, λ°μ°(div)λ ν μ $(x, y, z)$μ λͺ¨μ΄λ(μμΆ) λ°©ν₯μΌλ‘ μ μ²΄κ° νλ₯΄λμ§, μλλ©΄ μ μμ λκ°λ(ν½μ°½)νλ λ°©ν₯μΌλ‘ μ μ²΄κ° νλ₯΄λμ§λ₯Ό μμΉν νλ κ°μ΄λ€.
- $\text{div }\mathbf{F} > 0$
- ν μ μμ μ μ²΄κ° ν½μ°½νλ€: outflowing-ness.
- $\text{div }\mathbf{F} < 0$
- ν μ μμ μ μ²΄κ° μμΆλλ€: inflowing-ness.
- $\text{div }\mathbf{F} = 0$
- ν μ μ΄ μ 체μ μ ν μλ ₯μ λ°κ³ μμ§ μλ€: incompressible.
λ°μ°μ λν μμ 3μ°¨μ 곡κ°μ λν κ²μΌλ‘ μ μμ§λ§, 2μ°¨μ νλ©΄μμ μ μν 벑ν°μ₯μ λν΄μλ λ°μ°(divergence) κ°μ κ³μ°ν μ μλ€.
Gilbert Strang - Calculus Vol 3.: Negative Divergence
μμ 벑ν°μ₯μ $\mathbf{F}(x, y) = \left<-y, x\right>$λ‘ divλ₯Ό κ³μ°ν΄λ³΄λ©΄, $-2$κ° λλ€. λ°λΌμ λͺ¨λ μ μ΄ μμΆνλ λ°©ν₯μΌλ‘ μλ ₯μ λ°λλ€.
Gilbert Strang - Calculus Vol 3.: Zero Divergence
λ°λ©΄μ μμΆμ μ ν λ°μ§ μλ 벑ν°μ₯λ μλ€. $\mathbf{F}(x, y) = \left<-x, y\right>$μΈ λ²‘ν°μ₯μ divλ₯Ό κ³μ°ν΄λ³΄λ©΄ $0$μ΄ λλ€. λ°λΌμ λͺ¨λ μ μμ μμΆμ λ°μ§ μλλ€.
Gilbert Strang - Calculus Vol 3.: Zero Divergence
λ, μμ κ°μ΄ μ 체μ νλ¦μ΄ μμμΈ κ²½μ°λ μ μ²΄κ° λ°λ μλ ₯μ΄ $0$μΌλ‘ κ³μ°λλ€.
λκ° $\text{div } \mathbf{F} = 0$μΈ μν©μ βμ μ²΄κ° λ°λ μλ ₯μ΄ $0$βμ΄λ€ λΌκ³ νννλ©΄ μ’ λͺ¨νΈν κ°μ΄ μλ€. κ·Έλμ μ΄λ κ² νννλ κ²λ κ°λ₯νλ€.
ν μ§μ μμ μ 체μ μ μΆμ΄λ μ μ λλ μ λμ΄ μλ€.
μ¦, μ μ²΄κ° λκ°λ λ§νΌ, μ΄λκ°μμ μ μ²΄κ° λ€μ΄μ λκ°λ λμ μμμν¨λ€λ κ²μ΄λ€.
Source-free Field
μμμ μ€λͺ ν κ²κ³Ό κ°μ΄ μ΄λ€ μ§μ λλ μμμμ μ μΆλμ΄ $0$μΈ λ²‘ν°μ₯μ βsource-free Vector FieldβλΌκ³ νλ€. μ΄λ° 벑ν°μ₯μ μ무 λ«ν 곑μ μ μ‘λλΌλ, κ·Έ 곑μ μ λν΄ μ μ /μ μΆνλ μ 체λμ΄ λμΌνλ€.
βsourceβλΌλ ννμ ν μ μμ λκ°λ μ λμ λ§νλ€. λ°λλ‘ βsinkβλ ν μ μΌλ‘ λ€μ΄μ€λ μ λμ λ§νλ€. wikipedia
λ, μ΄λ° $\text{div }\mathbf{F} = 0$μΈ λ²‘ν°μ₯μ βSolenoidal Vector FieldβλΌκ³ λ λΆλ₯Έλ€. μλ λ Έμ΄λλ κ³ λ±νκ΅ λ¬Όλ¦¬μκ°μ λ΄€λ κ·Έ βμλ λ Έμ΄λβλ₯Ό λ§νλ€ γ γ
Curl
If $\mathbf{F} = P \, \mathbf{i} + Q \, \mathbf{j} + R \, \mathbf{k}$ is a vector field on $\mathbb{R}^3$, and the partial derivatives of $P$, $Q$ and $R$ all exists,
then the βcurl of $\mathbf{F}$β is the vector field on $\mathbb{R}^3$ defined by
\[\text{curl } \mathbf{F} = \left( \frac{\partial R}{\partial y} - \frac{\partial Q}{\partial z} \right) \mathbf{i} + \left( \frac{\partial P}{\partial z} - \frac{\partial R}{\partial x} \right) \mathbf{j} + \left( \frac{\partial Q}{\partial x} - \frac{\partial P}{\partial y} \right) \mathbf{k}\]μμ΄ μ‘°κΈ λ³΅μ‘νλ°, μμ μμ μΈμ°κΈ° 보λ€λ μΈμ μΌλ‘ νκΈ°ν μλμ μμΌλ‘ μ΄ν΄νλκ² λ μ μ©νλ€.
\[\text{curl } \mathbf{F} = \nabla \times \mathbf{F}\]μ¦, Gradient 벑ν°μ 벑ν°μ₯μ μΈμ ν κ²μ΄ νμ (curl)μ΄λ€.
\[\nabla \times \mathbf{F} = \left| \begin{matrix} \mathbf{i} & \mathbf{j} & \mathbf{k} \\ \frac{\partial}{\partial x} & \frac{\partial}{\partial y} & \frac{\partial}{\partial z} \\ P & Q & R \end{matrix} \right|\]Geometric meaning
벑ν°μ₯μ μ 체μ νλ¦ μμ΄λΌκ³ νννλ€λ©΄, νμ (curl)μ ν μ $(x, y, z)$ μ£Όλ³μ νμ νλ μ 체μ νμ νλ¦μ νννλ 벑ν°λ€.
- curl 벑ν°μ λ°©ν₯ = νμ μΆ
- curl 벑ν°μ λΆνΈ = νμ λ°©ν₯: CCW(+), CW(-)
- curl 벑ν°μ ν¬κΈ° = νμ νλ μλ
Gilbert Strang - Calculus Vol 3.: Zero Divergence, Positive Curl
$\mathbf{F} = \left< -y, x, 0\right>$λΌλ 벑ν°μ₯μ μμ $O$μμμ curlμ κ³μ°ν΄λ³΄λ©΄,
\[\text{curl } \mathbf{F} = 0 \, \mathbf{i} + 0 \, \mathbf{j} + 2 \, \mathbf{k}\]λ‘, $+z$ λ°©ν₯μ curl 벑ν°λ₯Ό μ»μΌλ©°, μ μ²΄κ° μμ μ κΈ°μ€μΌλ‘ μκ³ λ°©ν₯μΌλ‘ νμ νκ³ μμμ ν μ μλ€.
curl 벑ν°κ° μ벑ν°μΈ κ²½μ°λ₯Ό μ μ²΄κ° ν΄λΉ μ μ£Όλ³μ νμ νμ§ μλ(irrotational)νλ€κ³ λ§νλ€. μ΄κ²μ μ μ²΄κ° νμ μ΄λμμ μμ λ‘λ€λ κ²μΌλ‘, μ μ²΄κ° μ§μ μ΄λμ νκ³ μλ€λ©΄, κ·Έ λ°©ν₯μ΄ κ±°μ λ°λμ§ μμ κ²μμ λ§νλ€.
Gilbert Strang - Calculus Vol 3.: Zero Divergence, Zero Curl
μλ° λ²‘ν°μ₯λ curl 벑ν°κ° μ벑ν°μΈλ°, μ μ²΄κ° μμ§μ΄λ λ°©ν₯κ³Ό λ€λ₯Έ λ°©ν₯μΌλ‘ μμ§μ΄κ² νλ €λ μΈλ ₯μ΄ μ‘΄μ¬νμ§ μκΈ° λλ¬Έμ΄λ€.
If Conservative field, curl is zero vector
If $f$ a function of three variables that has continuous second-order derivatives, then
\[\text{curl}(\nabla f) = \mathbf{0}\]μ 리μμλ potential function $f(x, y, z)$λ₯Ό κΈ°μ€μΌλ‘ λμ΄ μμ§λ§, Conservative field $\mathbf{F} = \nabla f$μ κ²½μ°, curlμ΄ μ벑ν°μμ λ§νκ³ μλ€.
μ¦λͺ μ Conservative Fieldμ μ μμ μΆ©μ€νκ² μμ μ κ°νκΈ°λ§ νλ©΄ λλ€.
\[\mathbf{F} = \nabla f = \frac{\partial f}{\partial x} \, \mathbf{i} + \frac{\partial f}{\partial y} \, \mathbf{j} + \frac{\partial f}{\partial z} \, \mathbf{k}\] \[\text{curl } \mathbf{F} = \left(\frac{\partial^2 f}{\partial y\partial z} - \frac{\partial^2 f}{\partial z\partial y}\right) \, \mathbf{i} + \left(\frac{\partial^2 f}{\partial z\partial x} - \frac{\partial^2 f}{\partial x\partial z}\right) \, \mathbf{j} + \left(\frac{\partial^2 f}{\partial x\partial y} - \frac{\partial^2 f}{\partial y\partial x}\right) \, \mathbf{k} = \mathbf{0}\]EzEz $\blacksquare$ (μκ² κ°λ₯ν μ΄μ λ νΈλ―ΈλΆ μμλ₯Ό λ°κΏλ μκ΄ μλ€λ ν΄λ λ‘μ μ 리(Clairautβs Theorem) λλ¬Έ.)
μμ μ 리λ Conservative 벑ν°μ₯μ curl 벑ν°κ° νμ μ벑ν°λΌ κ³μ°ν νμ μλ€λ κ²λ λ§ν΄μ£Όμ§λ§, μ€νλ € 벑ν°μ₯μ curlμ΄ μ벑ν°κ° μλ μν©μμ μ μ©νλ€. μλνλ©΄, λμ° λͺ μ μ λ°λΌ $\text{curl }\mathbf{F} \ne 0$λΌλ©΄, 벑ν°μ₯μ΄ non-conservativeμ΄κΈ° λλ¬Έμ΄λ€.
μμ μ 리μ μλͺ μ λ μ±λ¦½νμ§ μλλ€. $\text{curl } \mathbf{F} = \mathbf{0}$μ΄λλΌλ κ·Έ 벑ν°μ₯μ΄ conservativeκ° μλ λ°λ‘κ° μκΈ° λλ¬Έ.
λ¨, 벑ν°μ₯μ΄ μλμ 쑰건μ λ§μ‘±νλ€λ©΄ μλͺ μ λ μ±λ¦½νλ€κ³ νλ€.
If given vector field $\mathbf{F}$ is
- defined on all of $\mathbb{R}^3$
- and whose component functions have continuous partial derivatives
- and $\text{curl } \mathbf{F} = \mathbf{0}$
then $\mathbf{F}$ is a conservative vector field.
κ·Έλμ λ²‘ν° ν¨μκ° λΆμ°μμ±μ κ°μ§κ±°λ, μ±λΆ ν¨μμ νΈλ―ΈλΆμ΄ λΆμ°μμ±μ κ°λ κ·Έλ° νΉμν μν©λ§ μλλΌλ©΄, $\text{curl } \mathbf{F} = \mathbf{0}$μΈμ§ νμΈνμ¬ conservative μ¬λΆλ₯Ό νλ¨ν μ μμ κ²μ΄λ€ γ γ
μ¬μ€ μ μμμ΄ $\mathbb{R}^3$μΌ νμ μκ³ , simply-connected regionμ΄κΈ°λ§ νλ©΄ λλ€κ³ νλ€. (μ¦λͺ μ μ΄νμ μ€ν μ€ν¬ μ 리 ν λ λ°°μ΄λ€κ³ ν¨.)
Curl and Div
μ΄λ€ 벑ν°μ₯μ Curlμ μ μ©ν ν, Divergenceλ₯Ό μ μ©νλ©΄ μ¬λ°λ κ²°κ³Όκ° λμ¨λ€.
If $\mathbf{F} = P \, \mathbf{i} + Q \, \mathbf{j} + R \, \mathbf{k}$ is a vector field on $\mathbb{R}^3$ and $p$, $Q$, and $R$ have continuous second-order partial derivatives, then
\[\text{div } \text{curl } \mathbf{F} = 0\]μ¦λͺ μ curlμ divμ μ μμ λ§μ μλ μμ μ κ°νλ©΄ λλ€. Ez
\[\text{div } \text{curl } \mathbf{F} = \nabla \cdot (\nabla \times \mathbf{F}) = 0\]μ μ 리λ 벑ν°μ₯μ΄ curl μ°μ°μΌλ‘ μΈν΄ λ§λ€μ΄μ§ κ²μ΄λΌλ©΄, κ·Έκ²μ divκ° 0μμ λ§νλ€. μ΄κ²μ λμ° λͺ μ λ₯Ό νμ©νλ©΄, μ΄λ€ 벑ν°μ₯ $\mathbf{G}$κ° $\text{div } \mathbf{G} \ne 0$λΌλ©΄, κ·Έ 벑ν°μ₯μ curl 벑ν°λ‘ μ λλ κ²μ΄ μλλΌκ³ νλ¨ν μ μλ€.
Vector Potential
벑ν°μ₯μ curlμ μ μ©νλ©΄ νμ 벑ν°μ₯μ΄λΌλ λλ€λ₯Έ 벑ν°μ₯μ΄ λ°μ³λμ¨λ€. κ·Έλμ μ£Όμ΄μ§ 벑ν°μ₯μ΄ νμ 벑ν°μ₯μ΄λΌλ©΄, curlμ μ μ©ν μλ³Έ 벑ν°μ₯μ βVector PotentialβλΌκ³ λΆλ₯Έλ€.
The vector field $\mathbf{A}$ is said to be a βvector potentialβ for the vector field $\mathbf{B}$ if
\[\mathbf{B} = \nabla \times \mathbf{A}\]μμ Curl-Div νλ±μμ Vector Potentialλ‘ λ€μ κΈ°μ νλ©΄ μλμ κ°λ€.
If there exists a vector potential for the vector field $\mathbf{B}$, then
\[\nabla \cdot \mathbf{B} = 0\]μμ λμλ $\nabla \times \mathbf{F} = 0$ μΌμ΄μ€μ μ’ ν·κ°λ¦΄ μλ μμ΄μ μ 리λ₯Ό μ’ ν΄λ³΄μ.
For a vector field $\mathbf{F}$
- If scalar Potential(= potential function) exists, then
- the field is a conservative field
- $\nabla \times \mathbf{F} = \mathbf{0}$
- If vector Potential exists, then
- the field is a curl vector field
- $\nabla \cdot \mathbf{F} = 0$
Vector Potentialμ΄λ Scalar Potentialμ΄λ μ΄λ€ 벑ν°μ₯μ μλ³Έμ΄ λλ vector/scalar functionλΌλκ² κ³΅ν΅μ !
μ΄λ, νμ 벑ν°μ₯ $\mathbf{B}$λ₯Ό λ§λλ Vector Potentialμ μ μΌνκ² κ²°μ λμ§ μλλ€.
μμμ μ€μΉΌλΌ ν¨μ $\psi$μ λν΄ μλμ λ±μμ΄ μ±λ¦½νλ€.
\[\mathbf{B} = \nabla \times \mathbf{A} = \nabla \times (\mathbf{A} + \nabla \psi)\]μ΄κ²μ΄ κ°λ₯ν μ΄μ λ $\nabla \times \nabla \psi = \mathbf{0}$μ΄κΈ° λλ¬Έμ΄λ€. $\mathbf{A}$μ $\mathbf{A} + \nabla \psi$ λλ€ νμ 벑ν°μ₯ $\mathbf{B}$λ₯Ό μ λνλ―λ‘, λλ€ Vector Potential μ΄κ³ , Vector Potentialμ μ μΌνκ² κ²°μ λμ§ μκ³ λ¬΄νν λ§λ€.
λ°λΌμ, μλμ λ°λ¦ μ λ¦¬κ° μ±λ¦½νλλ°,
If the vector field $\mathbf{B}$ has a vector potential,
then, there is a vector potential $\mathbf{A}$ for $\mathbf{B}$ with $\mathbf{A}_3 = 0$.
μ¦, νμ 벑ν°μ₯μ Vector Potentialλ‘ $\mathbf{k}$ μ»΄ν¬λνΈκ° 0μΈ λ²‘ν°μ₯μ΄ νμ μ‘΄μ¬νλ€λ κ²μ΄λ€. μ΄ κ²°κ³Όλ $\mathbf{k}$ μ»΄ν¬λνΈκ° μλλΌ $\mathbf{i}$λ $\mathbf{j}$κ° 0μ΄μ΄λ μ±λ¦½νλ€.
Laplace Operator
벑ν°μ₯ $\mathbf{F}$κ° Conservative FieldλΌλ©΄,
\[\mathbf{F} = \nabla f\]μΈλ°, μ¬κΈ°μ div μ°μ°μ νλ²λ μνν΄λ³΄μ.[* curl μ°μ°μ νλ©΄, μ벑ν°κ° λμλ€: $\nabla \times \nabla f = \mathbf{0}$] κ·Έλ¬λ©΄ μμ μλμ κ°μ΄ potential function $f$μ μ±λΆλ³λ‘ νΈλ―ΈλΆμ λ λ² μ μ©ν κ²μ ν©μΌλ‘ ννλλ€.
\[\text{div}(\nabla f) = \nabla \cdot (\nabla f) = \frac{\partial^2 f}{\partial x^2} + \frac{\partial^2 f}{\partial y^2} + \frac{\partial^2 f}{\partial z^2}\]μλ κ² potential functionμ Gradient μ°μ°κ³Ό Div μ°μ°μ μ°μν΄ μννλ κ²μ κ°λ¨νκ² μλμ νκΈ°λ‘ νννλ€.
\[\nabla^2 f = \nabla \cdot (\nabla f) = \text{div}(\nabla f)\]μ μ°μ°μ μννλ μ°μ°μλ₯Ό βLaplace OperatorβλΌκ³ λΆλ₯Έλ€. μλνλ©΄, Laplace Equationμμ μ λν κ²μ΄κΈ° λλ¬Έ.
\[\nabla^2 f = \frac{\partial^2 f}{\partial x^2} + \frac{\partial^2 f}{\partial y^2} + \frac{\partial^2 f}{\partial z^2} = 0\]μ°Έκ³ λ‘ λΌνλΌμ€ λ°©μ μμ λ―Έλ°© λ λ§μ΄ λ³Ό μμ β¦ ^^ 1νλ λ―Έμ λΆνμμλ μ ν μ¬μ©νμ§ μμ μ°μ°μμ΄λ κ·Έλ₯ μ¬κΈ°μμ κ°λ λ§ λ³΄κ³ λμ΄κ°μλ€!
Vector form of Greenβs Theorem
κ·Έλ¦° μ 리λ₯Ό λ°μ°κ³Ό νμ μ νκΈ°λ₯Ό μ¬μ©ν΄ ννν μλ μμ΅λλ€!
ν¬μ€νΈμ λ΄μ©μ΄ κΈΈμ΄μ Έμ ν΄λΉ λ΄μ©μ λ³λμ ν¬μ€νΈλ‘ λΆλ¦¬νμμ΅λλ€ ^^