ここでは、ベクトル解析について解説する。物理数学I ベクトル解析 及び解析学基礎/多変数関数の微積分 も参照。
ここでは、実数のみで考える。
高等学校数学C/ベクトル 、線型代数学/ベクトル 、解析学基礎/ベクトル などを参照。本項では、矢印を付した表記(
a
→
{\displaystyle {\vec {a}}}
など)と太字立体による表記(
p
{\displaystyle \mathbf {p} }
など)及び太字斜体による表記(
x
{\displaystyle {\boldsymbol {x}}}
など)を混用する。
高等学校数学C/数学的な表現の工夫 及び線形代数学 の各ページを参照。本項では大文字の太字立体(
A
{\displaystyle \mathbf {A} }
など)で表記する。
関数
f
(
x
)
=
∑
i
f
i
(
x
)
e
i
{\displaystyle {\boldsymbol {f}}(x)=\sum _{i}f_{i}(x){\boldsymbol {e}}_{i}}
をベクトル(値)関数 (vector-valued function, vector function)という。
ただし、各
e
i
{\displaystyle {\boldsymbol {e}}_{i}}
は各座標軸方向の単位ベクトルであり、
f
(
x
)
{\displaystyle {\boldsymbol {f}}(x)}
の定義域は各
f
i
(
x
)
{\displaystyle f_{i}(x)}
の定義域の積集合である。
ベクトルの演算が成分ごとの演算結果を並べたベクトルで定義されたように、ベクトル関数の演算は(結果となる値が存在するならば)各成分の演算結果を並べたベクトルに等しい。これはベクトルの線型性による。
例えば、
f
(
x
)
=
(
sin
x
cos
x
)
{\displaystyle {\boldsymbol {f}}(x)={\begin{pmatrix}\sin x\\\cos x\end{pmatrix}}}
の
x
→
0
{\displaystyle x\to 0}
の極限は、
lim
x
→
0
f
(
x
)
=
(
lim
x
→
0
cos
x
lim
x
→
0
sin
x
)
{\displaystyle \lim _{x\to 0}{\boldsymbol {f}}(x)={\begin{pmatrix}\lim _{x\to 0}\cos x\\\lim _{x\to 0}\sin x\end{pmatrix}}}
(
∵
{\displaystyle \because }
線型性)
=
(
1
0
)
{\displaystyle ={\begin{pmatrix}1\\0\end{pmatrix}}}
=
e
x
{\displaystyle ={\boldsymbol {e}}_{x}}
と求まる。
このとき、
e
x
{\displaystyle {\boldsymbol {e}}_{x}}
を
f
(
x
)
{\displaystyle {\boldsymbol {f}}(x)}
の
x
→
0
{\displaystyle x\to 0}
に於ける極限ベクトル という。
また、ベクトル関数の導関数
f
′
(
x
)
{\displaystyle {\boldsymbol {f}}'(x)}
は
d
d
x
f
(
x
)
=
d
d
x
(
∑
i
f
i
(
x
)
e
i
)
{\displaystyle {\frac {d}{dx}}{\boldsymbol {f}}(x)={\frac {d}{dx}}\left(\sum _{i}f_{i}(x)\mathbf {e} _{i}\right)}
=
∑
i
(
d
d
x
f
i
(
x
)
e
i
)
{\displaystyle =\sum _{i}\left({\frac {d}{dx}}f_{i}(x)\mathbf {e} _{i}\right)}
(
∵
{\displaystyle \because }
線型性)
=
(
f
1
′
(
x
)
f
2
′
(
x
)
⋮
f
i
′
(
x
)
⋮
)
{\displaystyle ={\begin{pmatrix}f'_{1}(x)\\f'_{2}(x)\\\vdots \\f'_{i}(x)\\\vdots \end{pmatrix}}}
である。
この結果は、ベクトル関数に於ける導関数の正式な定義
lim
h
→
0
f
(
x
+
h
)
−
f
(
x
)
h
{\displaystyle \lim _{h\to 0}{\frac {{\boldsymbol {f}}(x+h)-{\boldsymbol {f}}(x)}{h}}}
からも導かれる。
高校数学や高校物理で扱った平面(
R
2
{\displaystyle \mathbb {R} ^{2}}
)上の運動は、実は2項ベクトル関数の理論に他ならない。
例えば、平面上の運動に於ける加速度・速度・変位の関係は
x
(
t
)
=
(
f
(
t
)
g
(
t
)
)
{\displaystyle {\boldsymbol {x}}(t)={\begin{pmatrix}f(t)\\g(t)\end{pmatrix}}}
として
a
(
t
)
=
d
d
t
v
(
t
)
=
d
2
d
t
2
x
(
t
)
{\displaystyle {\boldsymbol {a}}(t)={\frac {d}{dt}}{\boldsymbol {v}}(t)={\frac {d^{2}}{dt^{2}}}{\boldsymbol {x}}(t)}
と表すことができる。
曲線の媒介変数表示は、パラメータを変数とするベクトル関数とみることができる。
例えば、円の媒介変数表示
{
x
=
r
cos
θ
y
=
r
sin
θ
{\displaystyle {\begin{cases}x=r\cos \theta \\y=r\sin \theta \end{cases}}}
は、
r
(
θ
)
=
r
(
cos
θ
sin
θ
)
{\displaystyle {\boldsymbol {r}}(\theta )=r{\begin{pmatrix}\cos \theta \\\sin \theta \end{pmatrix}}}
という2次元ベクトル関数とみることができる。更に
θ
=
ω
t
{\displaystyle \theta =\omega t}
とすると、これは等速円運動の式である。
1変数ベクトル関数の微分に関して、以下の公式が成り立つ。
定数倍の導関数:
{
k
a
(
x
)
}
′
=
k
a
′
(
x
)
{\displaystyle \{k{\boldsymbol {a}}(x)\}'=k{\boldsymbol {a}}'(x)}
和・差の導関数:
{
a
(
x
)
±
b
(
x
)
}
′
=
a
′
(
x
)
±
b
′
(
x
)
{\displaystyle \{{\boldsymbol {a}}(x)\pm {\boldsymbol {b}}(x)\}'={\boldsymbol {a}}'(x)\pm {\boldsymbol {b}}'(x)}
積の導関数:
{
f
(
x
)
a
(
x
)
}
′
=
f
′
(
x
)
a
(
x
)
+
f
(
x
)
a
′
(
x
)
{\displaystyle \{f(x){\boldsymbol {a}}(x)\}'=f'(x){\boldsymbol {a}}(x)+f(x){\boldsymbol {a}}'(x)}
商の導関数:
{
a
(
x
)
f
(
x
)
}
′
=
f
(
x
)
a
′
(
x
)
−
f
′
(
x
)
a
(
x
)
{
f
(
x
)
}
2
(
f
(
x
)
≠
0
)
{\displaystyle \left\{{\frac {{\boldsymbol {a}}(x)}{f(x)}}\right\}'={\frac {f(x){\boldsymbol {a}}'(x)-f'(x){\boldsymbol {a}}(x)}{\{f(x)\}^{2}}}\quad (f(x)\neq 0)}
内積の導関数:
{
a
(
x
)
⋅
b
(
x
)
}
′
=
a
′
(
x
)
⋅
b
(
x
)
+
a
(
x
)
⋅
b
′
(
x
)
{\displaystyle \{{\boldsymbol {a}}(x)\cdot {\boldsymbol {b}}(x)\}'={\boldsymbol {a}}'(x)\cdot {\boldsymbol {b}}(x)+{\boldsymbol {a}}(x)\cdot {\boldsymbol {b}}'(x)}
外積の導関数:
{
a
(
x
)
×
b
(
x
)
}
′
=
a
′
(
x
)
×
b
(
x
)
+
a
(
x
)
×
b
′
(
x
)
{\displaystyle \{{\boldsymbol {a}}(x)\times {\boldsymbol {b}}(x)\}'={\boldsymbol {a}}'(x)\times {\boldsymbol {b}}(x)+{\boldsymbol {a}}(x)\times {\boldsymbol {b}}'(x)}
スカラー三重積の導関数:
[
a
(
x
)
⋅
{
b
(
x
)
×
c
(
x
)
}
]
′
=
a
′
(
x
)
⋅
{
b
(
x
)
×
c
(
x
)
}
+
a
(
x
)
⋅
{
b
′
(
x
)
×
c
(
x
)
}
+
a
(
x
)
⋅
{
b
(
x
)
×
c
′
(
x
)
}
{\displaystyle [{\boldsymbol {a}}(x)\cdot \{{\boldsymbol {b}}(x)\times {\boldsymbol {c}}(x)\}]'={\boldsymbol {a}}'(x)\cdot \{{\boldsymbol {b}}(x)\times {\boldsymbol {c}}(x)\}+{\boldsymbol {a}}(x)\cdot \{{\boldsymbol {b}}'(x)\times {\boldsymbol {c}}(x)\}+{\boldsymbol {a}}(x)\cdot \{{\boldsymbol {b}}(x)\times {\boldsymbol {c}}'(x)\}}
ベクトル三重積の導関数:
[
a
(
x
)
×
{
b
(
x
)
×
c
(
x
)
}
]
′
=
a
′
(
x
)
×
{
b
(
x
)
×
c
(
x
)
}
+
a
(
x
)
×
{
b
′
(
x
)
×
c
(
x
)
}
+
a
(
x
)
×
{
b
(
x
)
×
c
′
(
x
)
}
{\displaystyle [{\boldsymbol {a}}(x)\times \{{\boldsymbol {b}}(x)\times {\boldsymbol {c}}(x)\}]'={\boldsymbol {a}}'(x)\times \{{\boldsymbol {b}}(x)\times {\boldsymbol {c}}(x)\}+{\boldsymbol {a}}(x)\times \{{\boldsymbol {b}}'(x)\times {\boldsymbol {c}}(x)\}+{\boldsymbol {a}}(x)\times \{{\boldsymbol {b}}(x)\times {\boldsymbol {c}}'(x)\}}
ベクトル関数のノルム (大きさ、長さ)は、通常のベクトルと同様に
‖
f
(
x
)
‖
=
f
(
x
)
⋅
f
(
x
)
=
∑
i
{
f
i
(
x
)
}
2
{\displaystyle \|{\boldsymbol {f}}(x)\|={\sqrt {{\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}(x)}}={\sqrt {\sum _{i}\{f_{i}(x)\}^{2}}}}
で与えられる。
つまり、ノルムは元のベクトル関数と同一変数のスカラー関数である。
ここで、このスカラー関数が定数関数である条件に就て見てみよう。
‖
f
(
x
)
‖
=
c
(
c
∈
R
>
0
)
{\displaystyle \|{\boldsymbol {f}}(x)\|=c\quad (c\in \mathbb {R} _{>0})}
の両辺を2乗して
‖
f
(
x
)
‖
2
=
f
(
x
)
⋅
f
(
x
)
=
c
2
{\displaystyle \|{\boldsymbol {f}}(x)\|^{2}={\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}(x)=c^{2}}
両辺微分すると内積の微分より
{
f
(
x
)
⋅
f
(
x
)
}
′
=
f
′
(
x
)
⋅
f
(
x
)
+
f
(
x
)
⋅
f
′
(
x
)
=
2
f
(
x
)
⋅
f
′
(
x
)
=
0
{\displaystyle \{{\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}(x)\}'={\boldsymbol {f}}'(x)\cdot {\boldsymbol {f}}(x)+{\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}'(x)=2{\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}'(x)=0}
∴
f
(
x
)
⋅
f
′
(
x
)
=
0
{\displaystyle \therefore {\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}'(x)=0}
逆に、
f
(
x
)
⋅
f
′
(
x
)
=
0
⟺
2
f
(
x
)
⋅
f
′
(
x
)
=
f
′
(
x
)
⋅
f
(
x
)
+
f
(
x
)
⋅
f
′
(
x
)
=
{
f
(
x
)
⋅
f
(
x
)
}
′
=
0
{\displaystyle {\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}'(x)=0\iff 2{\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}'(x)={\boldsymbol {f}}'(x)\cdot {\boldsymbol {f}}(x)+{\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}'(x)=\{{\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}(x)\}'=0}
∫
{
f
(
x
)
⋅
f
(
x
)
}
′
d
x
=
∫
0
d
x
{\displaystyle \int \{{\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}(x)\}'dx=\int 0dx}
f
(
x
)
⋅
f
(
x
)
=
C
1
{\displaystyle {\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}(x)=C_{1}}
(
C
1
≥
0
{\displaystyle C_{1}\geq 0}
は積分定数)
‖
f
(
x
)
‖
2
=
C
1
{\displaystyle \|{\boldsymbol {f}}(x)\|^{2}=C_{1}}
より
‖
f
(
x
)
‖
=
C
1
=
c
o
n
s
t
.
{\displaystyle \|{\boldsymbol {f}}(x)\|={\sqrt {C_{1}}}=\mathrm {const} .}
よって、以下が成り立つ。
直交条件の定理
‖
f
(
x
)
‖
=
c
o
n
s
t
.
⟺
f
(
x
)
⋅
f
′
(
x
)
=
0
{\displaystyle \|{\boldsymbol {f}}(x)\|=\mathrm {const} .\iff {\boldsymbol {f}}(x)\cdot {\boldsymbol {f}}'(x)=0}
これは、物理的には「位置ベクトルと速度ベクトルが直交する」ことを表す。
また、ベクトル関数が変数に依らず常に同一方向を向く条件も見てみよう。
所与の条件は
f
(
x
)
=
f
(
x
)
c
{\displaystyle {\boldsymbol {f}}(x)=f(x){\boldsymbol {c}}}
と表される。
両辺微分して
f
′
(
x
)
=
{
f
(
x
)
c
}
=
f
′
(
x
)
c
{\displaystyle {\boldsymbol {f}}'(x)=\{f(x){\boldsymbol {c}}\}=f'(x){\boldsymbol {c}}}
ここで元のベクトル関数との外積をとると
f
(
x
)
×
f
′
(
x
)
=
{
f
(
x
)
c
}
×
{
f
′
(
x
)
c
}
=
f
(
x
)
f
′
(
x
)
(
c
×
c
)
=
f
(
x
)
f
′
(
x
)
0
=
0
{\displaystyle {\boldsymbol {f}}(x)\times {\boldsymbol {f}}'(x)=\{f(x){\boldsymbol {c}}\}\times \{f'(x){\boldsymbol {c}}\}=f(x)f'(x)({\boldsymbol {c}}\times {\boldsymbol {c}})=f(x)f'(x){\boldsymbol {0}}={\boldsymbol {0}}}
逆に、
f
(
x
)
×
f
′
(
x
)
=
0
{\displaystyle {\boldsymbol {f}}(x)\times {\boldsymbol {f}}'(x)={\boldsymbol {0}}}
とすると一次従属なので
f
′
(
x
)
=
g
(
x
)
f
(
x
)
{\displaystyle {\boldsymbol {f}}'(x)=g(x){\boldsymbol {f}}(x)}
という微分方程式が成り立つ。
各成分ごとに微分方程式を解くと、
これは変数分離型微分方程式なので第i成分は
d
f
i
d
x
=
g
(
x
)
f
i
(
x
)
{\displaystyle {\frac {df_{i}}{dx}}=g(x)f_{i}(x)}
∫
d
f
i
f
i
(
x
)
=
∫
g
(
x
)
d
x
{\displaystyle \int {\frac {df_{i}}{f_{i}(x)}}=\int g(x)dx}
G
′
(
x
)
=
g
(
x
)
,
C
i
{\displaystyle G'(x)=g(x),C_{i}}
を積分定数として
log
|
f
i
(
x
)
|
=
G
(
x
)
+
C
i
{\displaystyle \log |f_{i}(x)|=G(x)+C_{i}}
C
i
=
±
e
C
i
{\displaystyle C_{i}=\pm e^{C_{i}}}
と改めておくと
f
i
(
x
)
=
C
i
e
G
(
x
)
{\displaystyle f_{i}(x)=C_{i}e^{G(x)}}
よって
f
(
x
)
=
∑
i
f
i
(
x
)
e
i
=
∑
i
C
i
e
G
(
x
)
e
i
=
e
G
(
x
)
∑
i
C
i
e
i
{\displaystyle {\boldsymbol {f}}(x)=\sum _{i}f_{i}(x){\boldsymbol {e}}_{i}=\sum _{i}C_{i}e^{G(x)}{\boldsymbol {e}}_{i}=e^{G(x)}\sum _{i}C_{i}{\boldsymbol {e}}_{i}}
ここで
f
(
x
)
=
e
G
(
x
)
,
c
=
∑
i
C
i
e
i
{\displaystyle f(x)=e^{G(x)},{\boldsymbol {c}}=\sum _{i}C_{i}{\boldsymbol {e}}_{i}}
と置くと
f
(
x
)
=
f
(
x
)
c
{\displaystyle {\boldsymbol {f}}(x)=f(x){\boldsymbol {c}}}
となり、方向は変数に依らず一定であることが判る//
よって、以下が成り立つ。
平行条件の定理
f
(
x
)
=
f
(
x
)
c
⟺
f
(
x
)
×
f
′
(
x
)
=
0
{\displaystyle {\boldsymbol {f}}(x)=f(x){\boldsymbol {c}}\iff {\boldsymbol {f}}(x)\times {\boldsymbol {f}}'(x)=0}
f
(
x
)
{\displaystyle {\boldsymbol {f}}(x)}
の不定積分は、各
f
i
(
x
)
{\displaystyle f_{i}(x)}
の原始関数を並べたベクトルを
F
(
x
)
{\displaystyle {\boldsymbol {F}}(x)}
として
∫
f
(
x
)
d
x
=
F
(
x
)
+
C
{\displaystyle \int {\boldsymbol {f}}(x)dx={\boldsymbol {F}}(x)+{\boldsymbol {C}}}
(
C
{\displaystyle {\boldsymbol {C}}}
は各積分定数を並べたベクトル)
と表される。
また、定積分は
∫
b
a
f
(
x
)
d
x
=
F
(
b
)
−
F
(
a
)
{\displaystyle \int _{b}^{a}{\boldsymbol {f}}(x)dx={\boldsymbol {F}}(b)-{\boldsymbol {F}}(a)}
と表される。
1変数ベクトル関数の積分について、以下が成り立つ(不定積分表記だが定積分に就ても同様)。
定数倍の原始関数:
∫
{
k
a
(
x
)
}
d
x
=
k
∫
a
(
x
)
d
x
{\displaystyle \int \{k{\boldsymbol {a}}(x)\}dx=k\int \mathbf {a} (x)dx}
和・差の原始関数:
∫
{
a
(
x
)
±
b
(
x
)
}
d
x
=
∫
a
(
x
)
d
x
±
∫
b
(
x
)
d
x
{\displaystyle \int \{{\boldsymbol {a}}(x)\pm {\boldsymbol {b}}(x)\}dx=\int {\boldsymbol {a}}(x)dx\pm \int {\boldsymbol {b}}(x)dx}
内積の原始関数
∫
{
k
⋅
a
(
x
)
}
d
x
=
k
⋅
(
∫
a
(
x
)
d
x
)
{\displaystyle \int \{{\boldsymbol {k}}\cdot {\boldsymbol {a}}(x)\}dx={\boldsymbol {k}}\cdot \left(\int {\boldsymbol {a}}(x)dx\right)}
外積の原始関数
∫
{
k
×
a
(
x
)
}
d
x
=
k
×
(
∫
a
(
x
)
d
x
)
{\displaystyle \int \{{\boldsymbol {k}}\times {\boldsymbol {a}}(x)\}dx={\boldsymbol {k}}\times \left(\int {\boldsymbol {a}}(x)dx\right)}
スカラー倍の部分積分:
∫
f
(
x
)
a
′
(
x
)
d
x
=
f
(
x
)
a
(
x
)
−
∫
f
′
(
x
)
a
(
x
)
d
x
⟺
∫
f
′
(
x
)
a
(
x
)
d
x
=
f
(
x
)
a
(
x
)
−
∫
f
(
x
)
a
′
(
x
)
d
x
{\displaystyle \int f(x){\boldsymbol {a}}'(x)dx=f(x){\boldsymbol {a}}(x)-\int f'(x){\boldsymbol {a}}(x)dx\iff \int f'(x){\boldsymbol {a}}(x)dx=f(x){\boldsymbol {a}}(x)-\int f(x){\boldsymbol {a}}'(x)dx}
内積の部分積分:
∫
{
a
(
x
)
⋅
b
′
(
x
)
}
d
x
=
a
(
x
)
⋅
b
(
x
)
−
∫
{
a
′
(
x
)
⋅
b
(
x
)
}
d
x
{\displaystyle \int \{{\boldsymbol {a}}(x)\cdot {\boldsymbol {b}}'(x)\}dx={\boldsymbol {a}}(x)\cdot {\boldsymbol {b}}(x)-\int \{{\boldsymbol {a}}'(x)\cdot {\boldsymbol {b}}(x)\}dx}
外積の部分積分:
∫
{
a
(
x
)
×
b
′
(
x
)
}
d
x
=
a
(
x
)
×
b
(
x
)
−
∫
{
a
′
(
x
)
×
b
(
x
)
}
d
x
{\displaystyle \int \{{\boldsymbol {a}}(x)\times {\boldsymbol {b}}'(x)\}dx={\boldsymbol {a}}(x)\times {\boldsymbol {b}}(x)-\int \{{\boldsymbol {a}}'(x)\times {\boldsymbol {b}}(x)\}dx}
平行条件の定理の証明でベクトル関数の微分方程式を解いたように、成分毎に考えることで解空間ベクトルを求められる。
1階線型常微分方程式
f
′
(
x
)
+
a
(
x
)
f
(
x
)
=
b
(
x
)
{\displaystyle {\boldsymbol {f}}'(x)+a(x){\boldsymbol {f}}(x)={\boldsymbol {b}}(x)}
の解に就て考える。
第
i
{\displaystyle i}
成分を考えると
f
i
′
(
x
)
+
a
(
x
)
f
i
(
x
)
=
b
i
(
x
)
{\displaystyle f'_{i}(x)+a(x)f_{i}(x)=b_{i}(x)}
という1階線型常微分方程式を得るので、
両辺に積分因子
μ
(
x
)
=
exp
(
∫
a
(
x
)
d
x
)
{\displaystyle \mu (x)=\exp \left(\int a(x)dx\right)}
を掛けることで
f
i
(
x
)
=
1
μ
(
x
)
{
∫
μ
(
x
)
b
i
(
x
)
d
x
+
C
i
}
{\displaystyle f_{i}(x)={\frac {1}{\mu (x)}}\left\{\int \mu (x)b_{i}(x)dx+C_{i}\right\}}
を得る。
よって一般解は
f
(
x
)
=
∑
i
1
μ
(
x
)
{
∫
μ
(
x
)
b
i
(
x
)
d
x
+
C
i
}
e
i
=
1
μ
(
x
)
{
∫
μ
(
x
)
b
(
x
)
d
x
+
C
}
{\displaystyle {\boldsymbol {f}}(x)=\sum _{i}{\frac {1}{\mu (x)}}\left\{\int \mu (x)b_{i}(x)dx+C_{i}\right\}{\boldsymbol {e}}_{i}={\frac {1}{\mu (x)}}\left\{\int \mu (x){\boldsymbol {b}}(x)dx+{\boldsymbol {C}}\right\}}
ベクトル関数は必ずしもここで扱った
f
:
R
→
R
n
{\displaystyle {\boldsymbol {f}}:\mathbb {R} \to \mathbb {R} ^{n}}
のみを指さない。
例えば、ベクトル変数
x
=
∑
i
g
i
(
x
)
e
i
{\displaystyle {\boldsymbol {x}}=\sum _{i}g_{i}(x){\boldsymbol {e}}_{i}}
に対して
h
(
x
)
=
∑
i
(
h
∘
g
)
i
(
x
)
e
i
{\displaystyle {\boldsymbol {h}}({\boldsymbol {x}})=\sum _{i}(h\circ g)_{i}(x){\boldsymbol {e}}_{i}}
と定義されるような関数
h
:
R
m
→
R
m
{\displaystyle {\boldsymbol {h}}:\mathbb {R} ^{m}\to \mathbb {R} ^{m}}
もベクトル関数である。
また、始域と終域の次元が異なる
i
:
R
m
→
R
n
(
m
,
n
≠
1
)
{\displaystyle {\boldsymbol {i}}:\mathbb {R} ^{m}\to \mathbb {R} ^{n}\,(m,n\neq 1)}
のようなベクトル関数も考えられる。
一般に、始域が
R
n
(
n
≠
1
)
{\displaystyle \mathbb {R} ^{n}(n\neq 1)}
であっても終域が
R
{\displaystyle \mathbb {R} }
である関数はベクトル関数に対してスカラー関数 (scalar function)と呼ばれる。
この節で述べた理論は、多くは一般のベクトル関数に対しても成り立つ。但し、後述のヤコビ行列のような拡張を必要とするものもある。
写像
X
:
R
m
→
R
n
{\displaystyle \mathbf {X} :\mathbb {R} ^{m}\to \mathbb {R} ^{n}}
をベクトル場 (vector field)という。
実際には
m
=
n
{\displaystyle m=n}
の場合を考えることが多い。
写像
F
:
R
n
→
R
{\displaystyle F:\mathbb {R} ^{n}\to \mathbb {R} }
をスカラー場 (scalar field)という。
それぞれ上で述べたベクトル関数・スカラー関数に対応する。
ベクトル場の例としては、電場・磁場・重力場などが挙げられる。
スカラー場の例としては、電位分布・温度分布・圧力分布などが挙げられる。
例えば、ベクトル場
f
=
(
x
,
0
,
0
)
{\displaystyle {\boldsymbol {f}}=(x,0,0)}
は、
R
3
{\displaystyle \mathbb {R} ^{3}}
内の平面
x
=
k
{\displaystyle x=k}
上の任意点に、その点を始点とするベクトル
(
k
0
0
)
{\displaystyle {\begin{pmatrix}k\\0\\0\end{pmatrix}}}
が存在する状態を表す。
ベクトル場の第i成分に第k成分の変数が入っていたとしても、それはiと異なる独立変数なのでスカラーと見做してよい。
ベクトル場の和は、ベクトル場の成分の和をとったものに等しい。
例えば、
f
=
(
1
,
2
,
3
)
,
g
=
(
3
,
4
,
5
)
{\displaystyle {\boldsymbol {f}}=(1,2,3),{\boldsymbol {g}}=(3,4,5)}
について
f
+
g
=
(
1
,
2
,
3
)
+
(
3
,
4
,
5
)
=
(
1
+
3
,
2
+
4
,
3
+
5
)
=
(
4
,
6
,
8
)
{\displaystyle {\boldsymbol {f}}+{\boldsymbol {g}}=(1,2,3)+(3,4,5)=(1+3,2+4,3+5)=(4,6,8)}
である。
領域
D
{\displaystyle {\boldsymbol {D}}}
上のベクトル場
X
{\displaystyle \mathbf {X} }
について、微分方程式
d
r
d
t
=
X
(
r
)
(
r
∈
D
)
{\displaystyle {\frac {d\mathbf {r} }{dt}}=\mathbf {X} (\mathbf {r} )(\mathbf {r} \in {\boldsymbol {D}})}
で与えられる曲線
z
=
r
(
t
)
{\displaystyle z=\mathbf {r} (t)}
を
X
{\displaystyle \mathbf {X} }
の流線 (streamline)という。流線の接線ベクトルは元のベクトル場そのものになるので、方程式
d
x
1
X
x
1
=
d
x
2
X
x
2
=
⋯
=
d
x
n
X
x
n
{\displaystyle {\frac {dx_{1}}{X_{x_{1}}}}={\frac {dx_{2}}{X_{x_{2}}}}=\cdots ={\frac {dx_{n}}{X_{x_{n}}}}}
を流線方程式 という。
例えば、
f
(
x
,
y
)
=
(
y
,
x
)
{\displaystyle {\boldsymbol {f}}(x,y)=(y,x)}
の流線は、
定義から
d
x
d
t
=
y
,
d
y
d
t
=
x
{\displaystyle {\frac {dx}{dt}}=y,{\frac {dy}{dt}}=x}
合成関数の微分公式から
d
y
d
x
=
d
y
d
t
d
x
d
t
=
x
y
{\displaystyle {\frac {dy}{dx}}={\frac {\frac {dy}{dt}}{\frac {dx}{dt}}}={\frac {x}{y}}}
∫
y
d
y
=
∫
x
d
x
{\displaystyle \int ydy=\int xdx}
1
2
y
2
=
1
2
x
2
+
C
{\displaystyle {\frac {1}{2}}y^{2}={\frac {1}{2}}x^{2}+C}
よって双曲線
y
2
−
x
2
=
2
C
{\displaystyle y^{2}-x^{2}=2C}
//
この場合、
t
{\displaystyle t}
に関する2つの微分方程式を組み合わせることで
d
2
x
d
t
2
=
x
,
d
2
y
d
t
2
=
y
{\displaystyle {\frac {d^{2}x}{dt^{2}}}=x,{\frac {d^{2}y}{dt^{2}}}=y}
という二階線型常微分方程式となり、一般解が双曲線関数を用いて
{
x
=
C
1
cosh
t
+
C
2
sinh
t
y
=
C
1
sinh
t
+
C
2
cosh
t
{\displaystyle {\begin{cases}x=C_{1}\cosh t+C_{2}\sinh t\\y=C_{1}\sinh t+C_{2}\cosh t\end{cases}}}
であることが容易に判る。 なお、流線方程式を用いても
d
x
y
=
d
y
x
{\displaystyle {\frac {dx}{y}}={\frac {dy}{x}}}
から
d
y
d
x
=
x
y
{\displaystyle {\frac {dy}{dx}}={\frac {x}{y}}}
が出てくる。
xyz空間上のスカラー場を
F
:
R
3
→
R
{\displaystyle F:\mathbb {R} ^{3}\to \mathbb {R} }
、ベクトル場を
X
:
R
3
→
R
3
{\displaystyle \mathbf {X} :\mathbb {R} ^{3}\to \mathbb {R} ^{3}}
とする。また、
X
{\displaystyle \mathbf {X} }
のx, y, z成分をそれぞれ
X
x
,
X
y
,
X
z
{\displaystyle X_{x},X_{y},X_{z}}
と書くことにする。また、註がない限り凡ての議論は直交座標系のxyz空間で考えるものとする。
三変数関数
f
(
x
,
y
,
z
)
{\displaystyle f(x,y,z)}
の偏導関数 (partial derivative)はそれぞれ以下のように定義される。
∂
f
∂
x
=
lim
h
→
0
f
(
x
+
h
,
y
,
z
)
−
f
(
x
,
y
,
z
)
h
{\displaystyle {\frac {\partial {f}}{\partial {x}}}=\lim _{h\to 0}{\frac {f(x+h,y,z)-f(x,y,z)}{h}}}
∂
f
∂
y
=
lim
h
→
0
f
(
x
,
y
+
h
,
z
)
−
f
(
x
,
y
,
z
)
h
{\displaystyle {\frac {\partial {f}}{\partial {y}}}=\lim _{h\to 0}{\frac {f(x,y+h,z)-f(x,y,z)}{h}}}
∂
f
∂
z
=
lim
h
→
0
f
(
x
,
y
,
z
+
h
)
−
f
(
x
,
y
,
z
)
h
{\displaystyle {\frac {\partial {f}}{\partial {z}}}=\lim _{h\to 0}{\frac {f(x,y,z+h)-f(x,y,z)}{h}}}
偏導関数は、着目する変数軸方向の傾きを表す。
偏導関数を求める操作を偏微分 (partial differentiation)という。
つまり、偏微分とは着目する変数以外を定数と見做し、一変数関数とみて微分する 操作である。
記号「
∂
{\displaystyle \partial }
」は「デル(del)」「ラウン(ド)ディー(rounded d)」「パーシャルディー(partial d)」などと読まれる。
着目する変数以外を明示したい場合は、
∂
x
f
(
x
,
y
,
z
)
,
u
x
|
x
,
y
,
z
{\displaystyle \partial _{x}f(x,y,z),\;u_{x}|_{x,y,z}}
などの記法が用いられる。
偏微分でも、一変数関数の微分と同様の性質が成り立つ(特に線型性と積の微分)。証明は別項に譲る。
例えば、
∂
∂
x
(
x
2
+
2
y
x
+
y
2
)
=
∂
∂
x
x
2
+
y
∂
∂
x
x
+
y
2
∂
∂
x
1
=
2
x
+
y
{\displaystyle {\frac {\partial }{\partial {x}}}(x^{2}+2yx+y^{2})={\frac {\partial }{\partial {x}}}x^{2}+y{\frac {\partial }{\partial {x}}}x+y^{2}{\frac {\partial }{\partial {x}}}1=2x+y}
である。
偏導関数に偏微分変数の微小量を掛けて全ての変数について足し合わせた
d
f
=
∂
f
∂
x
d
x
+
∂
f
∂
y
d
y
+
∂
f
∂
z
d
z
{\displaystyle df={\frac {\partial {f}}{\partial {x}}}dx+{\frac {\partial {f}}{\partial {y}}}dy+{\frac {\partial {f}}{\partial {z}}}dz}
を
f
(
x
,
y
,
z
)
{\displaystyle f(x,y,z)}
の(完)全微分 (total derivative)という。
全微分は、関数
f
(
x
,
y
,
z
)
{\displaystyle f(x,y,z)}
を
∆
x
,
∆
y
,
∆
z
{\displaystyle {\text{∆}}x,{\text{∆}}y,{\text{∆}}z}
だけ変化させたときの関数全体の変化量
∆
f
=
f
(
x
+
∆
x
,
y
+
∆
y
,
z
+
∆
z
)
−
f
(
x
,
y
,
z
)
{\displaystyle {\text{∆}}f=f(x+{\text{∆}}x,y+{\text{∆}}y,z+{\text{∆}}z)-f(x,y,z)}
の
∆
x
,
∆
y
,
∆
z
→
0
{\displaystyle {\text{∆}}x,{\text{∆}}y,{\text{∆}}z\to 0}
とした三重極限に等しい。
つまり、
f
(
x
+
d
x
,
y
+
d
y
,
z
+
d
z
)
−
f
(
x
,
y
,
z
)
=
∂
f
∂
x
d
x
+
∂
f
∂
y
d
y
+
∂
f
∂
z
d
z
{\displaystyle f(x+dx,y+dy,z+dz)-f(x,y,z)={\frac {\partial {f}}{\partial {x}}}dx+{\frac {\partial {f}}{\partial {y}}}dy+{\frac {\partial {f}}{\partial {z}}}dz}
が成り立つ。
これを形式的に変形すると、
f
(
x
+
d
x
,
y
+
d
y
,
z
+
d
z
)
−
f
(
x
,
y
,
z
)
d
x
d
y
d
z
=
∂
f
∂
x
d
y
d
z
+
∂
f
∂
y
d
z
d
x
+
∂
f
∂
z
d
x
d
y
{\displaystyle {\frac {f(x+dx,y+dy,z+dz)-f(x,y,z)}{dxdydz}}={\frac {\partial {f}}{\partial {x}dydz}}+{\frac {\partial {f}}{\partial {y}dzdx}}+{\frac {\partial {f}}{\partial {z}dxdy}}}
となるが、右辺を見ると変数名が循環していることが、左辺を見ると一変数関数の導関数の定義式に酷似していることが判る。
f
=
(
f
1
(
x
,
y
,
z
)
f
2
(
x
,
y
,
z
)
f
3
(
x
,
y
,
z
)
)
{\displaystyle {\boldsymbol {f}}={\begin{pmatrix}f_{1}(x,y,z)\\f_{2}(x,y,z)\\f_{3}(x,y,z)\end{pmatrix}}}
としたとき、
J
f
(
x
,
y
,
z
)
=
∂
(
f
1
,
f
2
,
f
3
)
∂
(
x
,
y
,
z
)
=
(
∂
f
1
∂
x
∂
f
1
∂
y
∂
f
1
∂
z
∂
f
2
∂
x
∂
f
2
∂
y
∂
f
2
∂
z
∂
f
3
∂
x
∂
f
3
∂
y
∂
f
3
∂
z
)
{\displaystyle \mathbf {J} _{\boldsymbol {f}}(x,y,z)={\frac {\partial {(f_{1},f_{2},f_{3})}}{\partial {(x,y,z)}}}={\begin{pmatrix}{\frac {\partial {f_{1}}}{\partial {x}}}&{\frac {\partial {f_{1}}}{\partial {y}}}&{\frac {\partial {f_{1}}}{\partial {z}}}\\{\frac {\partial {f_{2}}}{\partial {x}}}&{\frac {\partial {f_{2}}}{\partial {y}}}&{\frac {\partial {f_{2}}}{\partial {z}}}\\{\frac {\partial {f_{3}}}{\partial {x}}}&{\frac {\partial {f_{3}}}{\partial {y}}}&{\frac {\partial {f_{3}}}{\partial {z}}}\end{pmatrix}}}
を
f
{\displaystyle {\boldsymbol {f}}}
の関数行列 (functional matrix)またはヤコビ行列 (ヤコビアン 、Jacobian matrix)という。
これを用いると、3次元ベクトル関数
f
(
x
,
y
,
z
)
{\displaystyle {\boldsymbol {f}}(x,y,z)}
の全微分は以下のように求められる。
d
f
=
J
f
(
x
0
,
y
0
,
z
0
)
d
v
{\displaystyle d{\boldsymbol {f}}=\mathbf {J} _{\boldsymbol {f}}(x_{0},y_{0},z_{0})d\mathbf {v} }
ただし、
d
v
=
(
d
x
d
y
d
z
)
{\displaystyle d\mathbf {v} ={\begin{pmatrix}dx\\dy\\dz\end{pmatrix}}}
である。このベクトルを方向ベクトル (direction vector)という。
関数
f
(
x
,
y
,
z
)
{\displaystyle f(x,y,z)}
が三次元ユークリッド空間
R
3
{\displaystyle \mathbb {R} ^{3}}
上に曲線
C
{\displaystyle C}
を描いているとする。
C
{\displaystyle C}
上の点
A
{\displaystyle A}
から点
B
{\displaystyle B}
まで、
C
{\displaystyle C}
に沿って積分することを考える。
経路
A
→
B
{\displaystyle A\to B}
を微小線分
∆
l
{\displaystyle {\text{∆}}l}
に分けたとき、区分求積法 により以下のように線積分 (line integral)が定義される。
∫
A
B
f
(
x
,
y
,
z
)
d
l
=
lim
∆
l
→
0
∑
A
B
f
(
x
,
y
,
z
)
∆
l
{\displaystyle \int _{A}^{B}f(x,y,z)dl=\lim _{{\text{∆}}l\to 0}\sum _{A}^{B}f(x,y,z){\text{∆}}l}
このとき、経路
A
→
B
{\displaystyle A\to B}
を積分(経)路 (path)や積分領域 (domain of integral)という。
なお、線積分は量子力学 の経路積分(path integral)とは異なる概念なので、混同に注意。
右辺の極限が具体的にどのような値に収束し、どのように計算されるかを見ていく。
曲線
C
{\displaystyle C}
の媒介変数表示を
p
(
t
)
=
(
x
(
t
)
y
(
t
)
z
(
t
)
)
{\displaystyle {\boldsymbol {p}}(t)={\begin{pmatrix}x(t)\\y(t)\\z(t)\end{pmatrix}}}
とし、初期条件を
p
(
a
)
=
O
A
→
,
p
(
b
)
=
O
B
→
{\displaystyle {\boldsymbol {p}}(a)={\vec {\mathrm {OA} }},{\boldsymbol {p}}(b)={\vec {\mathrm {OB} }}}
とする。
このとき、閉区間
[
a
,
b
]
{\displaystyle [a,b]}
を長さ
∆
t
=
b
−
a
n
{\displaystyle {\text{∆}}t={\frac {b-a}{n}}}
の
n
{\displaystyle n}
個の小区間
[
t
n
−
1
,
t
n
]
{\displaystyle [t_{n-1},t_{n}]}
に分割し、
C
{\displaystyle C}
上に標本点
p
(
t
i
)
{\displaystyle {\boldsymbol {p}}(t_{i})}
をとる。各標本点を結ぶ線分の長さを
∆
s
i
{\displaystyle {\text{∆}}s_{i}}
とおけば、
F
(
p
(
t
i
)
)
∆
s
i
{\displaystyle F({\boldsymbol {p}}(t_{i})){\text{∆}}s_{i}}
の総和はリーマン和である。
∆
t
→
0
{\displaystyle {\text{∆}}t\to 0}
の極限を考えると、
∆
s
i
=
‖
p
(
t
i
+
∆
t
)
−
p
(
t
i
)
‖
=
‖
p
(
t
i
+
∆
t
)
−
p
(
t
i
+
0
)
∆
t
−
0
‖
(
∆
t
−
0
)
=
‖
d
p
(
t
i
)
d
t
‖
d
t
{\displaystyle {\text{∆}}s_{i}=\|{\boldsymbol {p}}(t_{i}+{\text{∆}}t)-{\boldsymbol {p}}(t_{i})\|=\left\|{\frac {{\boldsymbol {p}}(t_{i}+{\text{∆}}t)-{\boldsymbol {p}}(t_{i}+0)}{{\text{∆}}t-0}}\right\|({\text{∆}}t-0)=\left\|{\frac {d{\boldsymbol {p}}(t_{i})}{dt}}\right\|dt}
よりこのリーマン和の極限は収束し、
lim
∆
t
→
0
∑
i
=
1
n
F
(
p
(
t
i
)
)
∆
s
i
=
∫
a
b
F
(
p
(
t
)
)
‖
d
p
(
t
)
d
t
‖
d
t
{\displaystyle \lim _{{\text{∆}}t\to 0}\sum _{i=1}^{n}F({\boldsymbol {p}}(t_{i})){\text{∆}}s_{i}=\int _{a}^{b}F({\boldsymbol {p}}(t))\left\|{\frac {d{\boldsymbol {p}}(t)}{dt}}\right\|dt}
//
ここで、この線積分の値は(同じ向き付けを与える限り)媒介変数のとり方に依らない 。
p
(
u
)
=
(
x
(
t
(
u
)
)
y
(
t
(
u
)
)
z
(
t
(
u
)
)
)
,
u
∈
[
c
,
d
]
{\displaystyle {\boldsymbol {p}}(u)={\begin{pmatrix}x(t(u))\\y(t(u))\\z(t(u))\end{pmatrix}},u\in [c,d]}
という変数変換を考えると、 恒等的に
d
t
d
u
>
0
{\displaystyle {\frac {dt}{du}}>0}
のとき
∫
c
d
F
(
p
(
t
(
u
)
)
)
‖
d
p
(
u
)
d
u
‖
d
u
=
∫
a
b
F
(
p
(
t
(
u
)
)
)
‖
d
p
(
t
)
d
t
‖
d
t
d
u
d
u
=
∫
a
b
F
(
p
(
t
)
)
‖
d
p
(
t
)
d
t
‖
d
t
{\displaystyle \int _{c}^{d}F({\boldsymbol {p}}(t(u)))\left\|{\frac {d{\boldsymbol {p}}(u)}{du}}\right\|du=\int _{a}^{b}F({\boldsymbol {p}}(t(u)))\left\|{\frac {d{\boldsymbol {p}}(t)}{dt}}\right\|{\frac {dt}{du}}du=\int _{a}^{b}F({\boldsymbol {p}}(t))\left\|{\frac {d{\boldsymbol {p}}(t)}{dt}}\right\|dt}
となり、一致する。
「恒等的に
d
t
d
u
>
0
{\displaystyle {\frac {dt}{du}}>0}
」は「
t
,
u
{\displaystyle t,u}
それぞれを用いた媒介変数表示が同じ向き付けを与える」ことと同値なので、先述の内容が確かめられた。
そこで、
‖
d
p
(
t
)
d
t
‖
d
t
{\displaystyle \left\|{\frac {d{\boldsymbol {p}}(t)}{dt}}\right\|dt}
を更に(形式的に)変形する。
‖
d
p
(
t
)
d
t
‖
d
t
=
‖
(
d
d
t
x
(
t
)
d
d
t
y
(
t
)
d
d
t
z
(
t
)
)
‖
d
t
{\displaystyle \left\|{\frac {d{\boldsymbol {p}}(t)}{dt}}\right\|dt=\left\|{\begin{pmatrix}{\frac {d}{dt}}x(t)\\{\frac {d}{dt}}y(t)\\{\frac {d}{dt}}z(t)\end{pmatrix}}\right\|dt}
=
(
d
x
(
t
)
d
t
)
2
+
(
d
y
(
t
)
d
t
)
2
+
(
d
z
(
t
)
d
t
)
2
d
t
{\displaystyle ={\sqrt {\left({\frac {dx(t)}{dt}}\right)^{2}+\left({\frac {dy(t)}{dt}}\right)^{2}+\left({\frac {dz(t)}{dt}}\right)^{2}}}\;dt}
=
(
d
x
(
t
)
)
2
+
(
d
y
(
t
)
)
2
+
(
d
z
(
t
)
)
2
{\displaystyle ={\sqrt {(dx(t))^{2}+(dy(t))^{2}+(dz(t))^{2}}}}
=
d
x
2
+
d
y
2
+
d
z
2
{\displaystyle ={\sqrt {dx^{2}+dy^{2}+dz^{2}}}}
最後の式を
d
l
{\displaystyle dl}
で置くと、
d
l
=
‖
d
p
‖
{\displaystyle dl=\|d{\boldsymbol {p}}\|}
が形式的に成り立つ。
この
d
l
{\displaystyle dl}
を線素 (line element)と呼ぶ。
線素を用いることで、曲線
C
{\displaystyle C}
(の一部)を積分経路とするスカラー場
F
{\displaystyle F}
の線積分は以下のように略記される。
∫
C
F
d
l
{\displaystyle \int _{C}Fdl}
特に積分経路が閉曲線である(閉経路 )とき、この線積分を周回積分 (閉路積分 、closed line integral)といい
∮
C
F
d
l
{\displaystyle \oint _{C}Fdl}
と表す。
線積分
∫
C
d
l
{\displaystyle \int _{C}dl}
は
C
{\displaystyle C}
の長さに等しい。
∫
C
d
l
=
∫
a
b
‖
d
p
(
t
)
‖
=
∫
a
b
‖
d
p
(
t
)
d
t
‖
d
t
{\displaystyle \int _{C}dl=\int _{a}^{b}\|d{\boldsymbol {p}}(t)\|=\int _{a}^{b}\left\|{\frac {d{\boldsymbol {p}}(t)}{dt}}\right\|dt}
=
∫
a
b
(
d
x
(
t
)
d
t
)
2
+
(
d
y
(
t
)
d
t
)
2
+
(
d
z
(
t
)
d
t
)
2
d
t
{\displaystyle =\int _{a}^{b}{\sqrt {\left({\frac {dx(t)}{dt}}\right)^{2}+\left({\frac {dy(t)}{dt}}\right)^{2}+\left({\frac {dz(t)}{dt}}\right)^{2}}}dt}
これは平面上の曲線の長さの式 を三次元に拡張した式であることがわかる。
C
{\displaystyle C}
を
C
{\displaystyle C}
上任意点からの符号付き距離
s
{\displaystyle s}
(弧長パラメータ )で媒介変数表示することを考える。ただし、積分経路の端点をそれぞれ
s
=
α
,
β
{\displaystyle s=\alpha ,\beta }
の場合とする。
すなわち、
q
(
s
)
=
(
x
(
s
)
y
(
s
)
z
(
s
)
)
{\displaystyle {\boldsymbol {q}}(s)={\begin{pmatrix}x(s)\\y(s)\\z(s)\end{pmatrix}}}
と表示する。
このとき定義から
s
=
∫
0
s
‖
d
q
d
s
‖
d
s
{\displaystyle s=\int _{0}^{s}\left\|{\frac {d{\boldsymbol {q}}}{ds}}\right\|ds}
なので、両辺を
s
{\displaystyle s}
で微分すると微分積分学の基本定理より
1
=
‖
d
q
d
s
‖
{\displaystyle 1=\left\|{\frac {d{\boldsymbol {q}}}{ds}}\right\|}
が恒等的に成り立つ。
これは、接線方向のベクトル(接ベクトル )が単位ベクトルであることを示す。
よって、
∫
C
F
d
l
=
∫
α
β
F
(
q
(
s
)
)
d
s
{\displaystyle \int _{C}Fdl=\int _{\alpha }^{\beta }F({\boldsymbol {q}}(s))ds}
である。
スカラー場の場合と同様の議論により、ベクトル場の線積分は以下のように計算されるとわかる。
∫
C
X
⋅
d
l
=
∫
a
b
X
(
p
(
t
)
)
⋅
p
′
(
t
)
d
t
{\displaystyle \int _{C}{\boldsymbol {X}}\cdot d{\boldsymbol {l}}=\int _{a}^{b}{\boldsymbol {X}}({\boldsymbol {p}}(t))\cdot {\boldsymbol {p}}'(t)dt}
ここで
d
l
{\displaystyle d{\boldsymbol {l}}}
は線素ベクトル であり、
d
l
=
(
d
x
d
y
d
z
)
=
d
x
e
x
+
d
y
e
y
+
d
z
e
z
{\displaystyle d{\boldsymbol {l}}={\begin{pmatrix}dx\\dy\\dz\end{pmatrix}}=dx{\boldsymbol {e}}_{x}+dy{\boldsymbol {e}}_{y}+dz{\boldsymbol {e}}_{z}}
と定義される。
そのため、この線積分は
∫
C
X
x
d
x
+
X
y
d
y
+
X
z
d
z
{\displaystyle \int _{C}X_{x}dx+X_{y}dy+X_{z}dz}
とも書かれる。
線素ベクトルは積分経路の接ベクトルとなるので、ベクトル場の線積分を接線積分 (integral of tangential component)という場合もある。
C
{\displaystyle C}
を弧長パラメータ
s
{\displaystyle s}
で媒介変数表示すると、
∫
C
X
⋅
d
l
=
∫
α
β
X
(
q
(
s
)
)
⋅
d
q
d
s
d
s
{\displaystyle \int _{C}{\boldsymbol {X}}\cdot d{\boldsymbol {l}}=\int _{\alpha }^{\beta }{\boldsymbol {X}}({\boldsymbol {q}}(s))\cdot {\frac {d{\boldsymbol {q}}}{ds}}ds}
ここで
‖
d
q
d
s
‖
=
1
{\displaystyle \left\|{\frac {d{\boldsymbol {q}}}{ds}}\right\|=1}
より、内積
X
(
q
(
s
)
)
⋅
d
q
d
s
{\displaystyle {\boldsymbol {X}}({\boldsymbol {q}}(s))\cdot {\frac {d{\boldsymbol {q}}}{ds}}}
は
X
(
q
(
s
)
)
{\displaystyle {\boldsymbol {X}}({\boldsymbol {q}}(s))}
を
q
(
s
)
{\displaystyle {\boldsymbol {q}}(s)}
方向に射影したものである。
すなわち、線積分はベクトル場
X
{\displaystyle {\boldsymbol {X}}}
の
C
{\displaystyle C}
の接線方向成分を積分したものである。これは「仕事」や「流れ」といった物理的な概念に通じ、線積分の本質を形作る。
なお、スカラー場の線積分に登場した線素
d
l
{\displaystyle dl}
は、線素ベクトルの大きさ
‖
d
l
‖
{\displaystyle \|d{\boldsymbol {l}}\|}
に等しい。これは、スカラー場の線積分が経路に沿った「長さの重み付け」積分であることを示している。
位置ベクトルを
r
(
x
,
y
,
z
)
{\displaystyle {\boldsymbol {r}}(x,y,z)}
とする。積分経路を流線上にとると、流線の定義から
d
r
/
/
X
(
r
)
{\displaystyle d{\boldsymbol {r}}//{\boldsymbol {X}}({\boldsymbol {r}})}
則ち一次従属なので、
d
r
=
k
X
(
r
)
{\displaystyle d{\boldsymbol {r}}=k{\boldsymbol {X}}({\boldsymbol {r}})}
と書ける。
よって線積分の被積分関数は
X
⋅
d
r
=
k
‖
X
‖
2
{\displaystyle {\boldsymbol {X}}\cdot d{\boldsymbol {r}}=k\|{\boldsymbol {X}}\|^{2}}
となり、積分値が最大化される他、ベクトル場の沿流方向成分を抽出できる。実際に計算する際は、
∫
s
t
r
e
a
m
l
i
n
e
k
‖
X
‖
2
{\displaystyle \int _{streamline}k\|{\boldsymbol {X}}\|^{2}}
とは書かず、
∫
s
t
r
e
a
m
l
i
n
e
X
⋅
d
r
d
t
d
t
{\displaystyle \int _{streamline}{\boldsymbol {X}}\cdot {\frac {d{\boldsymbol {r}}}{dt}}dt}
と書くのが望ましい(インテグラルと積分微小要素はセットのため)。
C
{\displaystyle C}
の単位接線ベクトルを
t
{\displaystyle {\boldsymbol {t}}}
とおくと、先ほどの議論より
t
=
d
p
(
s
)
d
s
{\displaystyle {\boldsymbol {t}}={\frac {d{\boldsymbol {p}}(s)}{ds}}}
であり、直交条件の定理から
t
⋅
d
t
d
t
=
0
{\displaystyle {\boldsymbol {t}}\cdot {\frac {d{\boldsymbol {t}}}{dt}}=0}
、則ち
d
p
(
t
)
d
t
=
d
2
p
(
s
)
d
s
2
{\displaystyle {\frac {d{\boldsymbol {p}}(t)}{dt}}={\frac {d^{2}{\boldsymbol {p}}(s)}{ds^{2}}}}
である。このベクトルを主法線ベクトル という。主法線ベクトルを
n
{\displaystyle {\boldsymbol {n}}}
とおくとき、
b
=
t
×
n
{\displaystyle {\boldsymbol {b}}={\boldsymbol {t}}\times {\boldsymbol {n}}}
を従法線ベクトル という。正規化すると
e
b
=
e
t
×
e
n
{\displaystyle {\boldsymbol {e}}_{b}={\boldsymbol {e}}_{t}\times {\boldsymbol {e}}_{n}}
、則ち
{
e
t
,
e
n
,
e
b
}
{\displaystyle \{{\boldsymbol {e}}_{t},{\boldsymbol {e}}_{n},{\boldsymbol {e}}_{b}\}}
は
R
3
{\displaystyle \mathbb {R} ^{3}}
の正規直交基底である。
主法線ベクトルの大きさ
κ
=
‖
n
‖
{\displaystyle \kappa =\|{\boldsymbol {n}}\|}
を
C
{\displaystyle C}
の曲率 、
τ
=
e
b
⋅
d
e
n
d
t
{\displaystyle \tau ={\boldsymbol {e}}_{b}\cdot {\frac {d{\boldsymbol {e}}_{n}}{dt}}}
を
C
{\displaystyle C}
の捩 ( れい ) 率 という。
例えば、円
x
2
+
y
2
=
r
2
{\displaystyle x^{2}+y^{2}=r^{2}}
の曲率と捩率は以下のように求まる。
媒介変数表示して
p
(
t
)
=
r
(
cos
t
sin
t
0
)
{\displaystyle {\boldsymbol {p}}(t)=r{\begin{pmatrix}\cos t\\\sin t\\0\end{pmatrix}}}
接線ベクトルは
d
p
(
t
)
d
t
=
r
(
−
sin
t
cos
t
0
)
{\displaystyle {\frac {d{\boldsymbol {p}}(t)}{dt}}=r{\begin{pmatrix}-\sin t\\\cos t\\0\end{pmatrix}}}
弧長パラメータ
s
{\displaystyle s}
は
s
=
∫
0
t
‖
d
p
(
u
)
d
t
‖
d
u
=
∫
0
t
r
d
t
=
r
t
{\displaystyle s=\int _{0}^{t}\left\|{\frac {d{\boldsymbol {p}}(u)}{dt}}\right\|du=\int _{0}^{t}rdt=rt}
∴
t
=
s
r
{\displaystyle \therefore t={\frac {s}{r}}}
p
(
s
)
=
r
(
cos
s
r
sin
s
r
0
)
{\displaystyle {\boldsymbol {p}}(s)=r{\begin{pmatrix}\cos {\frac {s}{r}}\\\sin {\frac {s}{r}}\\0\end{pmatrix}}}
p
(
s
)
d
s
=
(
−
sin
s
r
cos
s
r
0
)
{\displaystyle {\frac {{\boldsymbol {p}}(s)}{ds}}={\begin{pmatrix}-\sin {\frac {s}{r}}\\\cos {\frac {s}{r}}\\0\end{pmatrix}}}
d
2
p
(
s
)
d
s
2
=
−
1
r
(
cos
s
r
sin
s
r
0
)
{\displaystyle {\frac {d^{2}{\boldsymbol {p}}(s)}{ds^{2}}}=-{\frac {1}{r}}{\begin{pmatrix}\cos {\frac {s}{r}}\\\sin {\frac {s}{r}}\\0\end{pmatrix}}}
κ
=
‖
d
2
p
(
s
)
d
s
2
‖
=
1
r
{\displaystyle \kappa =\left\|{\frac {d^{2}{\boldsymbol {p}}(s)}{ds^{2}}}\right\|={\frac {1}{r}}}
t
=
r
(
−
sin
t
cos
t
0
)
,
n
=
d
t
d
t
=
−
r
(
cos
t
sin
t
0
)
{\displaystyle {\boldsymbol {t}}=r{\begin{pmatrix}-\sin t\\\cos t\\0\end{pmatrix}},{\boldsymbol {n}}={\frac {d{\boldsymbol {t}}}{dt}}=-r{\begin{pmatrix}\cos t\\\sin t\\0\end{pmatrix}}}
b
=
t
×
n
=
(
0
0
r
2
)
{\displaystyle {\boldsymbol {b}}={\boldsymbol {t}}\times {\boldsymbol {n}}={\begin{pmatrix}0\\0\\r^{2}\end{pmatrix}}}
τ
=
(
0
0
1
)
⋅
−
r
(
−
sin
t
cos
t
0
)
r
2
⋅
1
r
=
0
{\displaystyle \tau ={\frac {{\begin{pmatrix}0\\0\\1\end{pmatrix}}\cdot -r{\begin{pmatrix}-\sin t\\\cos t\\0\end{pmatrix}}}{r^{2}\cdot {\frac {1}{r}}}}=0}
//
曲率・捩率を用いることで、
t
,
n
,
b
{\displaystyle {\boldsymbol {t}},{\boldsymbol {n}},{\boldsymbol {b}}}
のs-方向微分を求めることができる。
フレネ=セレーの公式
d
d
s
(
e
t
e
n
e
b
)
=
(
0
κ
0
−
κ
0
τ
0
−
τ
0
)
(
e
t
e
n
e
b
)
{\displaystyle {\frac {d}{ds}}{\begin{pmatrix}{\boldsymbol {e}}_{t}\\{\boldsymbol {e}}_{n}\\{\boldsymbol {e}}_{b}\end{pmatrix}}={\begin{pmatrix}0&\kappa &0\\-\kappa &0&\tau \\0&-\tau &0\end{pmatrix}}{\begin{pmatrix}{\boldsymbol {e}}_{t}\\{\boldsymbol {e}}_{n}\\{\boldsymbol {e}}_{b}\end{pmatrix}}}
全微分がヤコビ行列による一次変換で表されたように、こちらも一次変換の形式で表すことができた。一般に、微分操作は
∞
{\displaystyle \infty }
次行列を表現行列とする線形写像である。詳しくは関数解析学 で扱う。
また、曲線は
(
κ
,
τ
)
{\displaystyle (\kappa ,\tau )}
の値と初期条件のみで定まるので、
(
κ
,
τ
)
{\displaystyle (\kappa ,\tau )}
の定める曲線族を
C
[
κ
,
τ
]
{\displaystyle {\mathcal {C}}[\kappa ,\tau ]}
と書く。
例えば、円を含む螺旋は
c
1
,
c
2
=
c
o
n
s
t
.
{\displaystyle c_{1},c_{2}=\mathrm {const} .}
として一般に
C
[
c
1
,
c
2
]
{\displaystyle {\mathcal {C}}[c_{1},c_{2}]}
と書ける。
R
=
1
κ
{\displaystyle R={\frac {1}{\kappa }}}
を曲率半径 という。曲率半径は曲線の特定のカーブにおける曲がり具合の急さを表す指標であり、カーブを近似する真円(密着円 )の半径でもある。高速道路の標識で
R
=
∘
∘
m
{\displaystyle R=\circ \circ \mathrm {m} }
は、そのカーブの曲率半径を示している。道路設計では、快適なドライブの為曲率半径がなるべく大きくなるように工夫を凝らしている。
関数
f
(
x
,
y
,
z
)
{\displaystyle f(x,y,z)}
が三次元ユークリッド空間
R
3
{\displaystyle \mathbb {R} ^{3}}
上に曲面
S
{\displaystyle S}
を描いているとする。
S
{\displaystyle S}
に沿って積分することを考える。
領域
S
{\displaystyle S}
を微小領域に分割した表面積を
∆
S
i
{\displaystyle {\text{∆}}S_{i}}
としたとき、区分求積法により以下のように(曲)面積分 (surface integral)が定義される。
∫
S
f
(
x
,
y
,
z
)
d
S
=
lim
∆
S
→
0
∑
i
f
(
x
i
,
y
i
,
z
i
)
∆
S
i
{\displaystyle \int _{S}f(x,y,z)dS=\lim _{{\text{∆}}S\to 0}\sum _{i}f(x_{i},y_{i},z_{i}){\text{∆}}S_{i}}
右辺の極限が具体的にどのような値に収束し、どのように計算されるかを見ていく。
曲面
S
{\displaystyle S}
の滑らかな媒介変数表示を
p
(
s
,
t
)
=
(
x
(
s
,
t
)
y
(
s
,
t
)
z
(
s
,
t
)
)
{\displaystyle \mathbf {p} (s,t)={\begin{pmatrix}x(s,t)\\y(s,t)\\z(s,t)\end{pmatrix}}}
とする。
ここで、パラメータベクトル
(
s
t
)
{\displaystyle {\begin{pmatrix}s\\t\end{pmatrix}}}
が属する領域
T
(
⊂
R
2
)
{\displaystyle T(\subset \mathbb {R} ^{2})}
を微小矩形領域
∆
T
i
{\displaystyle {\text{∆}}T_{i}}
に分割する。このとき、各iについて
∆
T
i
=
[
s
i
,
s
i
+
∆
s
i
]
×
[
t
i
,
t
i
+
∆
t
i
]
{\displaystyle {\text{∆}}T_{i}=[s_{i},s_{i}+{\text{∆}}s_{i}]\times [t_{i},t_{i}+{\text{∆}}t_{i}]}
であり、各
∆
T
i
{\displaystyle {\text{∆}}T_{i}}
に対応する曲面上の領域を
∆
S
i
=
p
(
∆
T
i
)
{\displaystyle {\text{∆}}S_{i}=\mathbf {p} ({\text{∆}}T_{i})}
とする。
このとき、パラメータ領域の面積変化率は、各変数に関する接ベクトル同士の張る接平面上の面積ベクトルで表される。
すなわち、
∆
s
i
,
∆
t
i
→
0
{\displaystyle {\text{∆}}s_{i},{\text{∆}}t_{i}\to 0}
のとき
∆
S
i
=
‖
∂
p
(
s
i
,
t
i
)
∂
s
×
∂
p
(
s
i
,
t
i
)
∂
t
‖
∆
s
i
∆
t
i
{\displaystyle {\text{∆}}S_{i}=\left\|{\frac {\partial {\mathbf {p} (s_{i},t_{i})}}{\partial {s}}}\times {\frac {\partial {\mathbf {p} (s_{i},t_{i})}}{\partial {t}}}\right\|{\text{∆}}s_{i}{\text{∆}}t_{i}}
である。
このリーマン和の二重極限は収束し、
lim
∆
s
i
→
0
∆
t
i
→
0
∑
i
f
(
x
i
,
y
i
,
z
i
)
∆
S
i
=
lim
∆
s
i
→
0
∆
t
i
→
0
∑
i
f
(
p
(
s
i
,
t
i
)
)
‖
∂
X
(
s
i
,
t
i
)
∂
s
×
∂
X
(
s
i
,
t
i
)
∂
t
‖
∆
s
i
∆
t
i
{\displaystyle \lim _{\scriptstyle {\text{∆}}s_{i}\to 0 \atop \scriptstyle {\text{∆}}t_{i}\to 0}\sum _{i}f(x_{i},y_{i},z_{i}){\text{∆}}{S_{i}}=\lim _{\scriptstyle {\text{∆}}s_{i}\to 0 \atop \scriptstyle {\text{∆}}t_{i}\to 0}\sum _{i}f(\mathbf {p} (s_{i},t_{i}))\left\|{\frac {\partial {\mathbf {X} (s_{i},t_{i})}}{\partial {s}}}\times {\frac {\partial {\mathbf {X} (s_{i},t_{i})}}{\partial {t}}}\right\|{\text{∆}}{s_{i}}{\text{∆}}{t_{i}}}
=
∬
T
f
(
p
(
s
,
t
)
)
‖
∂
p
∂
s
×
∂
p
∂
t
‖
d
s
d
t
{\displaystyle =\iint _{T}f(\mathbf {p} (s,t))\left\|{\frac {\partial {\mathbf {p} }}{\partial {s}}}\times {\frac {\partial {\mathbf {p} }}{\partial {t}}}\right\|dsdt}
//
すなわち、面積分は二重積分に帰着される。
面積分の値は(Sの向き付けが同じならば)パラメータのとり方に依らない。
証明は恒等条件
det
∂
(
s
1
,
t
1
)
∂
(
s
2
,
t
2
)
>
0
{\displaystyle \det {\frac {\partial (s_{1},t_{1})}{\partial (s_{2},t_{2})}}>0}
を用いて線積分と同様に行われる。
そこで、
F
{\displaystyle F}
の
S
{\displaystyle S}
上での面積分を以下のように略記する。
∫
S
F
d
S
{\displaystyle \int _{S}FdS}
1の面積分を考えると、
∫
S
d
S
{\displaystyle \int _{S}dS}
は曲面
S
{\displaystyle S}
の表面積に等しいことが知られている。
そのため、
d
S
{\displaystyle dS}
を面(積要)素 (surface element)と呼ぶ。
S
{\displaystyle S}
が閉曲面のときは
∮
S
F
d
S
{\displaystyle \oint _{S}FdS}
とも書く。
例えば、関数
z
=
f
(
x
,
y
)
{\displaystyle z=f(x,y)}
が
R
3
{\displaystyle \mathbb {R} ^{3}}
上に描く曲面の表面積は、
p
=
(
x
y
z
)
{\displaystyle \mathbf {p} ={\begin{pmatrix}x\\y\\z\end{pmatrix}}}
として
∬
T
‖
∂
p
∂
x
×
∂
p
∂
y
‖
d
x
d
y
{\displaystyle \iint _{T}\left\|{\frac {\partial {\mathbf {p} }}{\partial {x}}}\times {\frac {\partial {\mathbf {p} }}{\partial {y}}}\right\|dxdy}
=
∬
T
‖
(
−
∂
f
∂
x
−
∂
f
∂
y
1
)
‖
d
x
d
y
{\displaystyle =\iint _{T}\left\|{\begin{pmatrix}-{\frac {\partial {f}}{\partial {x}}}\\-{\frac {\partial {f}}{\partial {y}}}\\1\end{pmatrix}}\right\|dxdy}
=
∬
T
(
∂
f
∂
x
)
2
+
(
∂
f
∂
y
)
2
+
1
d
x
d
y
{\displaystyle =\iint _{T}{\sqrt {\left({\frac {\partial {f}}{\partial {x}}}\right)^{2}+\left({\frac {\partial {f}}{\partial {y}}}\right)^{2}+1}}\;dxdy}
と求められる。
スカラー場の場合と同様の議論により、ベクトル場の面積分は以下のように計算されるとわかる。
∫
S
X
⋅
d
S
→
=
∬
X
(
p
(
s
,
t
)
)
⋅
(
∂
p
∂
s
×
∂
p
∂
t
)
d
s
d
t
{\displaystyle \int _{S}\mathbf {X} \cdot d{\vec {S}}=\iint \mathbf {X} (\mathbf {p} (s,t))\cdot \left({\frac {\partial {\mathbf {p} }}{\partial {s}}}\times {\frac {\partial {\mathbf {p} }}{\partial {t}}}\right)dsdt}
=
∬
det
(
X
,
∂
p
∂
s
,
∂
p
∂
t
)
d
s
d
t
{\displaystyle =\iint \det {\begin{pmatrix}\mathbf {X} ,{\frac {\partial {\mathbf {p} }}{\partial {s}}},{\frac {\partial {\mathbf {p} }}{\partial {t}}}\end{pmatrix}}dsdt}
(
∵
{\displaystyle \because }
ベクトル三重積の性質)
ここで
d
S
→
{\displaystyle d{\vec {S}}}
は面素ベクトルであり、
d
S
→
=
n
d
S
{\displaystyle d{\vec {S}}=\mathbf {n} dS}
と定義される。
n
{\displaystyle \mathbf {n} }
は有向曲面
S
{\displaystyle S}
と右手系(後述)をなすように定められた単位法線ベクトルである。
n
{\displaystyle \mathbf {n} }
は、x, y, z軸それぞれとなす角度を
α
,
β
,
γ
{\displaystyle \alpha ,\beta ,\gamma }
とすると通常の余弦関数で表されるベクトル
(
cos
α
cos
β
cos
γ
)
{\displaystyle {\begin{pmatrix}\cos \alpha \\\cos \beta \\\cos \gamma \end{pmatrix}}}
に等しい。これを方向余弦 という。
方向余弦に関して、以下のような関係が成り立つ(※証明の加筆お願いします)。
{
cos
α
d
S
=
d
y
d
z
cos
β
d
S
=
d
z
d
x
cos
γ
d
S
=
d
x
d
y
{\displaystyle {\begin{cases}\cos \alpha dS=dydz\\\cos \beta dS=dzdx\\\cos \gamma dS=dxdy\end{cases}}}
これを用いることで、
d
S
→
=
e
x
d
y
d
z
+
e
y
d
z
d
x
+
e
z
d
x
d
y
=
(
d
y
d
z
d
z
d
x
d
x
d
y
)
{\displaystyle d{\vec {S}}=\mathbf {e} _{x}dydz+\mathbf {e} _{y}dzdx+\mathbf {e} _{z}dxdy={\begin{pmatrix}dydz\\dzdx\\dxdy\end{pmatrix}}}
と表せる。
敢えてx⇒y⇒zの順で記したように、変数名が循環していることがわかる。
関数
f
(
x
,
y
,
z
)
{\displaystyle f(x,y,z)}
が三次元ユークリッド空間
R
3
{\displaystyle \mathbb {R} ^{3}}
上に立体領域
V
{\displaystyle V}
を描いているとする。
V
{\displaystyle V}
に沿って積分することを考える。
立体
V
{\displaystyle V}
を微小立体に分割した体積を
∆
V
i
{\displaystyle {\text{∆}}V_{i}}
としたとき、区分求積法により以下のように体(積)積分 (volume integral)が定義される。
∫
V
f
(
x
,
y
,
z
)
d
V
=
lim
∆
V
→
0
∑
i
f
(
x
i
,
y
i
,
z
i
)
∆
V
i
{\displaystyle \int _{V}f(x,y,z)dV=\lim _{{\text{∆}}V\to 0}\sum _{i}f(x_{i},y_{i},z_{i}){\text{∆}}V_{i}}
体積分では、リーマン和の三重極限を用いて右辺の極限値を求めることが困難である(3つの極限を任意に交換できるとは限らないため)。
そのため、以下のように考える。
微小立体の体積
d
V
{\displaystyle dV}
は、各辺の長さが
d
x
,
d
y
,
d
z
{\displaystyle dx,dy,dz}
である直方体の体積で近似できる(どのような分割の仕方をしても微小領域では直方体に近似できるため)。
よって、
d
V
=
d
x
d
y
d
z
{\displaystyle dV=dxdydz}
。この
d
V
{\displaystyle dV}
を体(積)素 (volume element)という。
つまり、体積分は
V
{\displaystyle V}
についての三重積分
∭
V
f
(
x
,
y
,
z
)
d
x
d
y
d
z
{\displaystyle \iiint _{V}f(x,y,z)dxdydz}
に等しい。
V
{\displaystyle V}
の媒介変数表示を
r
(
s
,
t
,
u
)
=
(
x
(
s
,
t
,
u
)
y
(
s
,
t
,
u
)
z
(
s
,
t
,
u
)
)
{\displaystyle \mathbf {r} (s,t,u)={\begin{pmatrix}x(s,t,u)\\y(s,t,u)\\z(s,t,u)\end{pmatrix}}}
とすると、スカラー場の
V
{\displaystyle V}
に関する体積分は
∫
V
F
(
r
)
d
V
=
∭
V
F
(
r
(
s
,
t
,
u
)
)
‖
det
∂
(
x
,
y
,
z
)
∂
(
s
,
t
,
u
)
‖
d
s
d
t
d
u
{\displaystyle \int _{V}F(\mathbf {r} )dV=\iiint _{V}F(\mathbf {r} (s,t,u))\left\|\det {\frac {\partial {(x,y,z)}}{\partial {(s,t,u)}}}\right\|dsdtdu}
で表される。
積分領域
V
{\displaystyle V}
が閉立体であるときは
∮
V
F
d
V
{\displaystyle \oint _{V}FdV}
とも書く。
また、ベクトル場
の
V
{\displaystyle V}
に関する体積分は、ベクトル場の成分となるスカラー場の体積分を並べたベクトル
∫
V
X
(
r
)
d
V
=
(
∫
V
X
x
(
r
)
d
V
∫
V
X
y
(
r
)
d
V
∫
V
X
z
(
r
)
d
V
)
{\displaystyle \int _{V}\mathbf {X} (\mathbf {r} )dV={\begin{pmatrix}\int _{V}X_{x}(\mathbf {r} )dV\\\int _{V}X_{y}(\mathbf {r} )dV\\\int _{V}X_{z}(\mathbf {r} )dV\end{pmatrix}}}
で表される。
平面の座標系として直交座標の他に斜交座標、極座標(円座標)、重心座標などがあったように、空間の座標系も複数通り考えられる。
空間座標を3つの実数の組
(
r
,
θ
,
z
)
{\displaystyle (r,\theta ,z)}
で表した座標系を円筒座標系 (円柱座標系 、cylindrical coordinate system)という。ただし、
r
{\displaystyle r}
は原点からの(符号付き)距離、
θ
{\displaystyle \theta }
は基準面に於ける始線からの正の回転角、
z
{\displaystyle z}
は基準面からの高さである。
円筒座標系は、円座標系のrθ平面にz軸方向の次元を加えた3次元空間の座標と考えられる。
そのため、円筒座標系と直交座標系の対応は以下のようになる。
{
x
=
r
cos
θ
y
=
r
sin
θ
z
=
z
⟺
{
r
=
x
2
+
y
2
θ
=
arctan
y
x
z
=
z
{\displaystyle {\begin{cases}x=r\cos \theta \\y=r\sin \theta \\z=z\end{cases}}\iff {\begin{cases}r={\sqrt {x^{2}+y^{2}}}\\\theta =\arctan {\frac {y}{x}}\\z=z\end{cases}}}
但し、
r
{\displaystyle r}
は基準面の裏側では負とする。
直交座標のz軸を基に円筒座標系の座標軸を決定する。一般に座標軸は直線とは限らないので、「軸」ではなく「曲線」と呼ぶことにする(座標線 )。なお、「曲線」という語は一般に「直線」を包含する。
r曲線:z軸上の点を通りz軸に垂直な半直線。
θ曲線:z軸を法線とする平面内にあり、中心がz軸上にある円周。
z曲線:z軸に平行な直線。
ここで、各曲線に沿った微小変化がどのように表されるか見ていく。
f
:
(
r
,
θ
,
z
)
→
(
x
,
y
,
z
)
{\displaystyle {\boldsymbol {f}}:(r,\theta ,z)\to (x,y,z)}
とする。
f
{\displaystyle {\boldsymbol {f}}}
全体の微小変化は全微分
d
f
=
∂
(
x
,
y
,
z
)
∂
(
r
,
θ
,
z
)
(
d
r
d
θ
d
z
)
{\displaystyle d{\boldsymbol {f}}={\frac {\partial {(x,y,z)}}{\partial {(r,\theta ,z)}}}{\begin{pmatrix}dr\\d\theta \\dz\end{pmatrix}}}
で表される。
上で求めた関係式より
f
(
r
θ
z
)
=
(
x
y
z
)
=
(
r
cos
θ
r
sin
θ
z
)
{\displaystyle {\boldsymbol {f}}{\begin{pmatrix}r\\\theta \\z\end{pmatrix}}={\begin{pmatrix}x\\y\\z\end{pmatrix}}={\begin{pmatrix}r\cos \theta \\r\sin \theta \\z\end{pmatrix}}}
なので、
∂
(
x
,
y
,
z
)
∂
(
r
,
θ
,
z
)
=
(
∂
∂
r
r
cos
θ
∂
∂
θ
r
cos
θ
∂
∂
z
r
cos
θ
∂
∂
r
r
sin
θ
∂
∂
θ
r
sin
θ
∂
∂
z
r
sin
θ
∂
∂
r
z
∂
∂
θ
z
∂
∂
z
z
)
=
(
cos
θ
−
r
sin
θ
0
sin
θ
r
cos
θ
0
0
0
1
)
{\displaystyle {\frac {\partial {(x,y,z)}}{\partial {(r,\theta ,z)}}}={\begin{pmatrix}{\frac {\partial }{\partial {r}}}r\cos \theta &{\frac {\partial }{\partial {\theta }}}r\cos \theta &{\frac {\partial }{\partial {z}}}r\cos \theta \\{\frac {\partial }{\partial {r}}}r\sin \theta &{\frac {\partial }{\partial {\theta }}}r\sin \theta &{\frac {\partial }{\partial {z}}}r\sin \theta \\{\frac {\partial }{\partial {r}}}z&{\frac {\partial }{\partial {\theta }}}z&{\frac {\partial }{\partial {z}}}z\end{pmatrix}}={\begin{pmatrix}\cos \theta &-r\sin \theta &0\\\sin \theta &r\cos \theta &0\\0&0&1\end{pmatrix}}}
よって
d
f
=
(
cos
θ
−
r
sin
θ
0
sin
θ
r
cos
θ
0
0
0
1
)
(
d
r
d
θ
d
z
)
=
(
cos
θ
sin
θ
0
)
d
r
+
(
−
sin
θ
cos
θ
0
)
r
d
θ
+
(
0
0
1
)
d
z
{\displaystyle d{\boldsymbol {f}}={\begin{pmatrix}\cos \theta &-r\sin \theta &0\\\sin \theta &r\cos \theta &0\\0&0&1\end{pmatrix}}{\begin{pmatrix}dr\\d\theta \\dz\end{pmatrix}}={\begin{pmatrix}\cos \theta \\\sin \theta \\0\end{pmatrix}}dr+{\begin{pmatrix}-\sin \theta \\\cos \theta \\0\end{pmatrix}}rd\theta +{\begin{pmatrix}0\\0\\1\end{pmatrix}}dz}
再右辺の各項を
d
r
→
,
d
θ
→
,
d
z
→
{\displaystyle d{\vec {r}},d{\vec {\theta }},d{\vec {z}}}
と書くことにすると、
|
d
r
→
|
,
|
d
θ
→
|
,
|
d
z
→
|
{\displaystyle |d{\vec {r}}|,|d{\vec {\theta }}|,|d{\vec {z}}|}
はそれぞれ
d
r
,
r
d
θ
,
d
z
{\displaystyle dr,rd\theta ,dz}
とわかる。
r
,
θ
,
z
{\displaystyle r,\theta ,z}
を物理量として捉えたとき
d
θ
{\displaystyle d\theta }
の単位はラジアン(無次元量)であり、
r
d
θ
{\displaystyle rd\theta }
の単位はメートル(長さの次元1)なので、次元の整合性が取れる。
纏めると以下のようになる。
e
x
{\displaystyle \mathbf {e} _{x}}
成分
e
y
{\displaystyle \mathbf {e} _{y}}
成分
e
z
{\displaystyle \mathbf {e} _{z}}
成分
ノルム
d
r
→
{\displaystyle d{\vec {r}}}
cos
θ
d
r
{\displaystyle \cos \theta dr}
sin
θ
d
r
{\displaystyle \sin \theta dr}
0
{\displaystyle 0}
d
r
{\displaystyle dr}
d
θ
→
{\displaystyle d{\vec {\theta }}}
−
r
sin
θ
d
θ
{\displaystyle -r\sin \theta d\theta }
r
cos
θ
d
θ
{\displaystyle r\cos \theta d\theta }
0
{\displaystyle 0}
r
d
θ
{\displaystyle rd\theta }
d
z
→
{\displaystyle d{\vec {z}}}
0
{\displaystyle 0}
0
{\displaystyle 0}
1
{\displaystyle 1}
d
z
{\displaystyle dz}
ここで
d
r
→
,
d
θ
→
,
d
z
→
{\displaystyle d{\vec {r}},d{\vec {\theta }},d{\vec {z}}}
は内積を取ればどの組も互いに直交することが容易にわかるので、この3つのベクトルは3次元ユークリッド空間の直交基底である。
そこで、それぞれのベクトルを正規化することで各軸曲線方向の基本ベクトルを得る。
e
r
=
d
r
→
d
r
=
cos
θ
e
x
+
sin
θ
e
y
{\displaystyle \mathbf {e} _{r}={\frac {d{\vec {r}}}{dr}}=\cos \theta \mathbf {e} _{x}+\sin \theta \mathbf {e} _{y}}
e
θ
=
d
θ
→
r
d
θ
=
−
sin
θ
e
x
+
cos
θ
e
y
{\displaystyle \mathbf {e} _{\theta }={\frac {d{\vec {\theta }}}{rd\theta }}=-\sin \theta \mathbf {e} _{x}+\cos \theta \mathbf {e} _{y}}
e
z
=
d
z
→
d
z
=
e
z
{\displaystyle \mathbf {e} _{z}={\frac {d{\vec {z}}}{dz}}=\mathbf {e} _{z}}
これを一次変換の形式で表すと
(
e
r
e
θ
e
z
)
=
(
cos
θ
sin
θ
0
−
sin
θ
cos
θ
0
0
0
1
)
(
e
x
e
y
e
z
)
{\displaystyle {\begin{pmatrix}\mathbf {e} _{r}\\\mathbf {e} _{\theta }\\\mathbf {e} _{z}\end{pmatrix}}={\begin{pmatrix}\cos \theta &\sin \theta &0\\-\sin \theta &\cos \theta &0\\0&0&1\end{pmatrix}}{\begin{pmatrix}\mathbf {e} _{x}\\\mathbf {e} _{y}\\\mathbf {e} _{z}\end{pmatrix}}}
となる。
このときの変換行列を、座標系または基底ベクトルを変換することから座標変換行列 又は基底変換行列 という。場合によっては変換テンソル とも呼ぶ。
なお、この直交座標⇒円筒座標の座標変換行列は、xyz空間に於けるz軸まわりの回転行列に等しい。
逆基底変換はこの座標変換行列の逆行列を求めれば定式化できる。
回転行列は直交行列なので
(
cos
θ
sin
θ
0
−
sin
θ
cos
θ
0
0
0
1
)
−
1
=
t
(
cos
θ
sin
θ
0
−
sin
θ
cos
θ
0
0
0
1
)
=
(
cos
θ
−
sin
θ
0
sin
θ
cos
θ
0
0
0
1
)
{\displaystyle {\begin{pmatrix}\cos \theta &\sin \theta &0\\-\sin \theta &\cos \theta &0\\0&0&1\end{pmatrix}}^{-1}={}^{t}\!{\begin{pmatrix}\cos \theta &\sin \theta &0\\-\sin \theta &\cos \theta &0\\0&0&1\end{pmatrix}}={\begin{pmatrix}\cos \theta &-\sin \theta &0\\\sin \theta &\cos \theta &0\\0&0&1\end{pmatrix}}}
であり、
(
e
x
e
y
e
z
)
=
(
cos
θ
−
sin
θ
0
sin
θ
cos
θ
0
0
0
1
)
(
e
r
e
θ
e
z
)
{\displaystyle {\begin{pmatrix}\mathbf {e} _{x}\\\mathbf {e} _{y}\\\mathbf {e} _{z}\end{pmatrix}}={\begin{pmatrix}\cos \theta &-\sin \theta &0\\\sin \theta &\cos \theta &0\\0&0&1\end{pmatrix}}{\begin{pmatrix}\mathbf {e} _{r}\\\mathbf {e} _{\theta }\\\mathbf {e} _{z}\end{pmatrix}}}
円筒座標系は、電磁気学 の解析でよく使用される。
例えば、高校物理の公式
B
(
r
)
=
μ
0
I
τ
r
{\displaystyle B(r)={\frac {\mu _{0}I}{\tau r}}}
は、円筒座標系を知っていれば「電流がz軸に平行であることから磁場はz方向に一葉且つθ方向に回転対称なので、r方向の項のみが寄与する」と理解される。
高等学校までの地理で習ったように、球面上の位置は緯度
θ
{\displaystyle \theta }
と経度
ϕ
{\displaystyle \phi }
で定めることができる。
一般に三次元空間で球面上のある点の座標を定めるには、更に球の中心からの距離
r
{\displaystyle r}
が必要である。
このようにして定められる座標系を三次元極座標系 (polar coordinate system in 3D)または球(面)座標系 (spherical coordinate system)という。
xyz座標空間の点
P
(
x
,
y
,
z
)
{\displaystyle P(x,y,z)}
の極座標表示を考える。
r
=
|
O
P
→
|
=
x
2
+
y
2
+
z
2
{\displaystyle r=|{\vec {OP}}|={\sqrt {x^{2}+y^{2}+z^{2}}}}
は容易にわかる。
θ
{\displaystyle \theta }
はz軸からの天頂角 (zenith angle)であり、
O
P
→
{\displaystyle {\vec {OP}}}
のz軸への正射影ベクトルを
O
H
→
=
h
→
{\displaystyle {\vec {OH}}={\vec {h}}}
とすると
tan
θ
=
|
h
→
|
z
{\displaystyle \tan \theta ={\frac {|{\vec {h}}|}{z}}}
が成り立つ。
|
h
→
|
=
x
2
+
y
2
{\displaystyle |{\vec {h}}|={\sqrt {x^{2}+y^{2}}}}
より、
θ
=
arctan
x
2
+
y
2
z
{\displaystyle \theta =\arctan {\frac {\sqrt {x^{2}+y^{2}}}{z}}}
である。
ϕ
{\displaystyle \phi }
はx軸からの方位角 (azimuth angle)であり、
P
{\displaystyle P}
からxy平面に下した垂線の足
Q
{\displaystyle Q}
を考えると
tan
ϕ
=
y
x
{\displaystyle \tan \phi ={\frac {y}{x}}}
とわかる。よって、
ϕ
=
arctan
y
x
{\displaystyle \phi =\arctan {\frac {y}{x}}}
である。
逆に、これを
P
(
x
,
y
,
z
)
{\displaystyle P(x,y,z)}
に復元することを考える。
△
O
H
P
{\displaystyle \triangle OHP}
について
∠
H
O
P
=
θ
{\displaystyle \angle HOP=\theta }
なので、
z
=
O
H
¯
=
r
cos
θ
{\displaystyle z={\overline {OH}}=r\cos \theta }
。
また、
△
O
P
Q
{\displaystyle \triangle OPQ}
について
∠
O
P
Q
=
θ
{\displaystyle \angle OPQ=\theta }
なので、
O
Q
¯
=
r
sin
θ
{\displaystyle {\overline {OQ}}=r\sin \theta }
。
よって
x
=
O
Q
¯
cos
ϕ
=
r
sin
θ
cos
ϕ
,
y
=
O
Q
¯
sin
ϕ
=
r
sin
θ
sin
ϕ
{\displaystyle x={\overline {OQ}}\cos \phi =r\sin \theta \cos \phi ,y={\overline {OQ}}\sin \phi =r\sin \theta \sin \phi }
。
以上を纏めると、直交座標と球面座標の対応関係は
{
x
=
r
sin
θ
cos
ϕ
y
=
r
sin
θ
sin
ϕ
z
=
r
cos
θ
⟺
{
r
=
x
2
+
y
2
+
z
2
θ
=
arctan
x
2
+
y
2
z
ϕ
=
arctan
y
x
{\displaystyle {\begin{cases}x=r\sin \theta \cos \phi \\y=r\sin \theta \sin \phi \\z=r\cos \theta \end{cases}}\iff {\begin{cases}r={\sqrt {x^{2}+y^{2}+z^{2}}}\\\theta =\arctan {\frac {\sqrt {x^{2}+y^{2}}}{z}}\\\phi =\arctan {\frac {y}{x}}\end{cases}}}
一般に緯度は
−
90
∘
≦
θ
≦
90
∘
{\displaystyle -90^{\circ }\leqq \theta \leqq 90^{\circ }}
、経度は
−
180
∘
≦
ϕ
≦
180
∘
{\displaystyle -180^{\circ }\leqq \phi \leqq 180^{\circ }}
の範囲で表される(但し、符号の違いは東⇔西、南⇔北の違いである)。同様に、球面座標の天頂角
θ
{\displaystyle \theta }
、方位角
ϕ
{\displaystyle \phi }
は
θ
∈
[
−
π
4
,
τ
4
]
,
ϕ
∈
[
−
τ
2
,
τ
2
]
{\displaystyle \theta \in [-{\frac {\pi }{4}},{\frac {\tau }{4}}],\phi \in [-{\frac {\tau }{2}},{\frac {\tau }{2}}]}
で考える。赤道面より下側(すなわち南半球)では
θ
<
0
{\displaystyle \theta <0}
とする。
直交座標のz軸を基に極座標系の座標線を決定する。
r曲線:原点を通る半直線。
θ曲線:原点中心でz軸上の一点から出る半円弧。
φ曲線:z軸を法線とする平面内でz軸上に中心を持つ円周。
各曲線に沿った微小変化は円柱座標系の場合と同様、ベクトル場の全微分を用いて
{
d
r
→
=
(
sin
θ
cos
ϕ
sin
θ
sin
ϕ
cos
θ
)
d
r
d
θ
→
=
(
cos
θ
cos
ϕ
cos
θ
sin
ϕ
−
sin
θ
)
r
d
θ
d
ϕ
→
=
(
−
sin
ϕ
cos
ϕ
0
)
r
sin
θ
d
ϕ
{\displaystyle {\begin{cases}d{\vec {r}}={\begin{pmatrix}\sin \theta \cos \phi \\\sin \theta \sin \phi \\\cos \theta \end{pmatrix}}dr\\d{\vec {\theta }}={\begin{pmatrix}\cos \theta \cos \phi \\\cos \theta \sin \phi \\-\sin \theta \end{pmatrix}}rd\theta \\d{\vec {\phi }}={\begin{pmatrix}-\sin \phi \\\cos \phi \\0\end{pmatrix}}r\sin \theta d\phi \end{cases}}}
と求まる。
これらは(内積の結果から)互いに直交するので、円柱座標の場合と同様の議論により
(
sin
θ
cos
ϕ
sin
θ
sin
ϕ
cos
θ
cos
θ
cos
ϕ
cos
θ
sin
ϕ
−
sin
θ
−
sin
ϕ
cos
ϕ
0
)
{\displaystyle {\begin{pmatrix}\sin \theta \cos \phi &\sin \theta \sin \phi &\cos \theta \\\cos \theta \cos \phi &\cos \theta \sin \phi &-\sin \theta \\-\sin \phi &\cos \phi &0\end{pmatrix}}}
と変換テンソルが求まる。
これも直交行列であるので、逆変換テンソルは
(
sin
θ
cos
ϕ
cos
θ
cos
ϕ
cos
ϕ
sin
θ
sin
ϕ
cos
θ
sin
ϕ
−
sin
ϕ
−
sin
θ
cos
θ
0
)
{\displaystyle {\begin{pmatrix}\sin \theta \cos \phi &\cos \theta \cos \phi &\cos \phi \\\sin \theta \sin \phi &\cos \theta \sin \phi &-\sin \phi \\-\sin \theta &\cos \theta &0\end{pmatrix}}}
と求まる。
球面座標系は、力学 の解析でよく使用される。
例えば、赤道平面における角運動量保存則は
m
r
2
d
ϕ
d
t
=
c
o
n
s
t
.
{\displaystyle mr^{2}{\frac {d\phi }{dt}}=\mathrm {const} .}
と表示されるが、これはz軸方向の角運動量を球面座標系の演算により取り出したものである。
球面座標系は特に機械制御に応用されている。
円柱座標系・球面座標系ともに各基底ベクトルが互いに直交しているので、そのことを明示して(且つ通常の直交座標系と区別して)直交曲線座標系 (orthogonal curvilinear coordinate system)と呼ぶ場合がある。直交曲線座標系には、他にも放物柱座標系や楕円座標系などがある。
スカラー場の勾配 (gradient)を以下のように定義する。
g
r
a
d
F
=
(
∂
F
∂
x
∂
F
∂
y
∂
F
∂
z
)
=
∂
F
∂
x
e
x
+
∂
F
∂
y
e
y
+
∂
F
∂
z
e
z
{\displaystyle \mathrm {grad} \,F={\begin{pmatrix}{\frac {\partial {F}}{\partial {x}}}\\{\frac {\partial {F}}{\partial {y}}}\\{\frac {\partial {F}}{\partial {z}}}\end{pmatrix}}={\frac {\partial {F}}{\partial {x}}}\mathbf {e} _{x}+{\frac {\partial {F}}{\partial {y}}}\mathbf {e} _{y}+{\frac {\partial {F}}{\partial {z}}}\mathbf {e} _{z}}
勾配は、幾何的にはスカラー場の各点に於ける傾きを表す。これは、最右辺の式が「各軸方向の傾きを表すベクトルの和」になっていることからわかる。「スカラー場の各点の傾き」とは、則ち「スカラー場上の変化が最も急峻な方向及び変化の速さ」のことである。このことから、「gradient」の和訳として「勾配」が充当されたのは妥当である。
F
{\displaystyle F}
に一変数
f
(
x
)
,
g
(
y
)
,
h
(
z
)
{\displaystyle f(x),g(y),h(z)}
をそれぞれ代入すると、各座標軸方向の傾き(すなわち通常の導関数)を得る。このことからも、勾配がスカラー場の各点に於ける傾きを表すことが判る。
勾配はベクトル量であるため、そのことを明示するため新たな演算子を導入する。
空間偏微分演算子
∇
{\displaystyle \nabla }
を以下のように定義する。
∇
=
(
∂
∂
x
∂
∂
y
∂
∂
z
)
=
∂
∂
x
e
x
+
∂
∂
y
e
y
+
∂
∂
z
e
z
{\displaystyle \nabla ={\begin{pmatrix}{\frac {\partial }{\partial {x}}}\\{\frac {\partial }{\partial {y}}}\\{\frac {\partial }{\partial {z}}}\end{pmatrix}}={\frac {\partial }{\partial {x}}}\mathbf {e} _{x}+{\frac {\partial }{\partial {y}}}\mathbf {e} _{y}+{\frac {\partial }{\partial {z}}}\mathbf {e} _{z}}
この演算子自体は偏微分作用素の対象を後接しないと意味を成さない。しかし、便宜的にベクトル量と考えてベクトルと同様の演算則が成り立つと見做す。そのため、この演算子に関する演算はw:記号の濫用 (symbol abuse)であることを註しておく。
∇
{\displaystyle \nabla }
の正式な読み方はナブラ (nabla)であるが、
∆
{\displaystyle {\text{∆}}}
(delta)を反転した形であることからアルテッド (alted)とも読まれる。なお、ナブラの由来は「竪琴」を意味する古代ギリシャ語「νάβλα」である。
空間偏微分演算子を用いることで、勾配は以下のように略記される。
g
r
a
d
F
=
∇
F
{\displaystyle \mathrm {grad} \,F=\nabla F}
F
{\displaystyle F}
はスカラー場なのでスカラー量であり、
∇
{\displaystyle \nabla }
は便宜的にベクトル量と考えることから、その積
∇
F
{\displaystyle \nabla F}
はベクトル量のスカラー倍、すなわちベクトル量であることが見て取れる。
勾配を用いて定義される概念として、電場が有名である。
電位
ϕ
{\displaystyle \phi }
に対する電場
E
{\displaystyle E}
は
E
:=
−
g
r
a
d
ϕ
{\displaystyle E:=-\mathrm {grad} \,\phi }
スカラー場の全微分は
d
F
=
∂
F
∂
x
d
x
+
∂
F
∂
y
d
y
+
∂
F
∂
z
d
z
{\displaystyle dF={\frac {\partial {F}}{\partial {x}}}dx+{\frac {\partial {F}}{\partial {y}}}dy+{\frac {\partial {F}}{\partial {z}}}dz}
で定義された。
これを内積の成分表示とみると、ベクトル表示は
d
F
=
(
∂
F
∂
x
∂
F
∂
y
∂
F
∂
z
)
⋅
(
d
x
d
y
d
z
)
{\displaystyle dF={\begin{pmatrix}{\frac {\partial {F}}{\partial {x}}}\\{\frac {\partial {F}}{\partial {y}}}\\{\frac {\partial {F}}{\partial {z}}}\end{pmatrix}}\cdot {\begin{pmatrix}dx\\dy\\dz\end{pmatrix}}}
である。
すなわち、スカラー場の全微分は
d
F
=
∇
F
⋅
d
v
{\displaystyle dF=\nabla F\cdot d\mathbf {v} }
と表される。
一般に、任意方向Sの単位ベクトルが方向余弦ベクトル
n
{\displaystyle \mathbf {n} }
で表されるとすると、
n
{\displaystyle \mathbf {n} }
方向の方向微分係数は勾配と方向余弦ベクトルの内積で表される。更に、スカラー場の最大傾斜が勾配ベクトル方向であると判る。
ベクトル場の全微分は
d
X
=
∂
(
X
x
,
X
y
,
X
z
)
∂
(
x
,
y
,
z
)
⋅
d
v
{\displaystyle d\mathbf {X} ={\frac {\partial {(X_{x},X_{y},X_{z})}}{\partial {(x,y,z)}}}\cdot d\mathbf {v} }
で定義された。
ここでn次行列
A
{\displaystyle \mathbf {A} }
とn項列ベクトル
a
{\displaystyle \mathbf {a} }
に対し、
(
A
⊛
a
)
i
j
=
(
A
i
j
)
a
i
{\displaystyle (\mathbf {A} \circledast \mathbf {a} )_{ij}=(\mathbf {A} _{ij})a_{i}}
となるような可換二項演算子
⊛
{\displaystyle \circledast }
を定義する。これは本項独自の記法である。但し、
(
P
)
i
j
{\displaystyle (\mathbf {P} )_{ij}}
は行列
P
{\displaystyle \mathbf {P} }
の第ij成分を表す。
この記法を用いると、ヤコビ行列
∂
(
X
x
,
X
y
,
X
z
)
∂
(
x
,
y
,
z
)
{\displaystyle {\frac {\partial {(X_{x},X_{y},X_{z})}}{\partial {(x,y,z)}}}}
は
∇
⊛
X
{\displaystyle \nabla \circledast \mathbf {X} }
と表すことができる。
則ち、ベクトル場の全微分は
d
X
=
(
∇
⊛
X
)
⋅
d
v
{\displaystyle d\mathbf {X} =(\nabla \circledast \mathbf {X} )\cdot d\mathbf {v} }
と表される。
ベクトル変数による微分を考える。ベクトル変数による微分は、変数ベクトルの各成分による偏微分を並べたベクトルで表される。
例えば、
d
f
d
r
=
(
∂
f
∂
r
x
∂
f
∂
r
y
∂
f
∂
r
z
)
{\displaystyle {\frac {df}{d\mathbf {r} }}={\begin{pmatrix}{\frac {\partial {f}}{\partial {r_{x}}}}\\{\frac {\partial {f}}{\partial {r_{y}}}}\\{\frac {\partial {f}}{\partial {r_{z}}}}\end{pmatrix}}}
である。
これは
r
{\displaystyle \mathbf {r} }
の各成分が直交基底をなすとき、勾配であると考えてよい。
ベクトル場の発散 (divergence)を以下のように定義する。
d
i
v
X
=
∂
X
x
∂
x
+
∂
X
y
∂
y
+
∂
X
z
∂
z
{\displaystyle \mathrm {div} \,\mathbf {X} ={\frac {\partial {X_{x}}}{\partial {x}}}+{\frac {\partial {X_{y}}}{\partial {y}}}+{\frac {\partial {X_{z}}}{\partial {z}}}}
発散は、ベクトル場の各点における指力線ベクトルの流出入量の収支を表す。これは、右辺の式が「各軸方向のベクトルの大きさの変化率の和」になっていることからわかる。「発散」の辞書的定義は「自身から何かを湧出すること」であり、ベクトル場の発散の正負は夫々其の点に於ける指力線ベクトル全体としての流出・流入に対応している。故に、「divergence」を「発散」と訳出したことは妥当と考えられる。
発散はスカラー量であるが、それを明示するため空間偏微分演算子による表記を考える。
X
=
(
X
x
X
y
X
z
)
{\displaystyle \mathbf {X} ={\begin{pmatrix}X_{x}\\X_{y}\\X_{z}\end{pmatrix}}}
より、発散は空間偏微分演算子を用いると以下のように表せる。
d
i
v
X
=
∇
⋅
X
{\displaystyle \mathrm {div} \,\mathbf {X} =\nabla \cdot \mathbf {X} }
発散を用いた概念として、ガウスの法則が有名である。
電荷密度
ρ
{\displaystyle \rho }
、電場
E
{\displaystyle \mathbf {E} }
、空間の誘電率
ε
{\displaystyle \varepsilon }
とすると
d
i
v
E
=
ρ
ε
{\displaystyle \mathrm {div} \,\mathbf {E} ={\frac {\rho }{\varepsilon }}}
極座標で考えることにより、この式の重要な性質「荷電粒子の座標を原点にとると原点以外では
d
i
v
E
=
0
{\displaystyle \mathrm {div} \,\mathbf {E} =0}
」を証明することができる。
ベクトル場の存在する3次元領域内に体積
∆
v
{\displaystyle {\text{∆}}v}
の閉曲面をとる。
閉曲面の有向微小表面積を
d
S
→
{\displaystyle d{\vec {S}}}
とするとき、この微小面を通過する指力線の量は
X
n
d
S
=
X
⋅
d
S
→
{\displaystyle \mathbf {X} _{n}dS=\mathbf {X} \cdot d{\vec {S}}}
(
X
n
{\displaystyle \mathbf {X} _{n}}
は法線成分)
である。
この閉曲面全体を出入りする指力線の総量は面積分
∮
S
X
⋅
d
S
→
{\displaystyle \oint _{S}\mathbf {X} \cdot d{\vec {S}}}
で与えられる。
発散は単位体積当たりの指力線ベクトルの湧出量とも考えられるので、上式を規格化(体積で除)し体積を0に近づけた極限は発散に等しい。
d
i
v
X
=
lim
∆
V
→
0
1
∆
V
∮
S
X
⋅
d
S
→
{\displaystyle \mathrm {div} \,\mathbf {X} =\lim _{{\text{∆}}V\to 0}{\frac {1}{{\text{∆}}V}}\oint _{S}\mathbf {X} \cdot d{\vec {S}}}
これを積分形に変形すると、以下のガウスの発散定理 (Gauss' divergence theorem)を得る。
∫
∆
V
d
i
v
X
d
V
=
∮
S
X
⋅
d
S
→
{\displaystyle \int _{{\text{∆}}V}\mathrm {div} \,\mathbf {X} dV=\oint _{S}\mathbf {X} \cdot d{\vec {S}}}
発散はまた、ヤコビ行列の跡に等しい。
d
i
v
X
=
t
r
(
∇
⊛
X
)
{\displaystyle \mathrm {div} \,\mathbf {X} =\mathrm {tr} (\nabla \circledast \mathbf {X} )}
回転運動は、以下の3つにより一意に定まる。
回転軸の方向
回転の速さ
回転方向(軸に対して右回転か左回転か)
「回転方向」を(暗黙の諒解として)回転軸に対して右螺子の向きにとることにする(右手系 )。すると、回転運動を表すのに必要な情報は「方向」と「速さ(=スカラー量)」のみとなる。
従って、任意の回転運動はベクトルで表示される (ベクトルの大きさは「回転の速さ」で定める)。このようなベクトルを角速度ベクトル という。
ベクトル場の回転 (curl、rotation、rotor)を以下のように定義する。
c
u
r
l
X
=
r
o
t
X
=
(
∂
X
z
∂
y
−
∂
X
y
∂
z
∂
X
x
∂
z
−
∂
X
z
∂
x
∂
X
y
∂
x
−
∂
X
x
∂
y
)
{\displaystyle \mathrm {curl} \,\mathbf {X} =\mathrm {rot} \,\mathbf {X} ={\begin{pmatrix}{\frac {\partial {X_{z}}}{\partial {y}}}-{\frac {\partial {X_{y}}}{\partial {z}}}\\{\frac {\partial {X_{x}}}{\partial {z}}}-{\frac {\partial {X_{z}}}{\partial {x}}}\\{\frac {\partial {X_{y}}}{\partial {x}}}-{\frac {\partial {X_{x}}}{\partial {y}}}\end{pmatrix}}}
c
u
r
l
{\displaystyle \mathrm {curl} }
は理学系、
r
o
t
{\displaystyle \mathrm {rot} }
は工学系が好んで使用する。
z軸方向の成分について考える。
z軸方向の回転はフレミングの左手の法則からxy平面内の成分のみが寄与すると考えられる。z軸の正方向の回転を生むには、x軸方向の変化率を大きくしてy軸方向の変化率を小さくすればよい。
ある軸方向の変化率とはその軸の変数に関する偏導関数のことであったので、回転のz成分を
∂
X
y
∂
x
−
∂
X
x
∂
y
{\displaystyle {\frac {\partial {X_{y}}}{\partial {x}}}-{\frac {\partial {X_{x}}}{\partial {y}}}}
とすれば満足する。
x軸方向・y軸方向も同様であるので、回転は上の式で表されるとわかる。
ここまで定性的に見てきたが、今度は定量的に考える。
回転の正式な定義は「微小領域境界を線積分した値の微小面積密度を並べたベクトル」である。これは以下のように陰に表示される。
(
c
u
r
l
X
)
⋅
n
:=
lim
S
→
0
1
|
S
|
∮
∂
S
X
⋅
d
l
→
{\displaystyle (\mathrm {curl} \,\mathbf {X} )\cdot \mathbf {n} :=\lim _{S\to 0}{\frac {1}{|S|}}\oint _{\partial {S}}\mathbf {X} \cdot d{\vec {l}}}
但し、
S
{\displaystyle S}
は
n
{\displaystyle \mathbf {n} }
を単位法線ベクトルとし点
l
(
l
→
)
{\displaystyle l({\vec {l}})}
を含む微小有向面、
∂
S
{\displaystyle \partial {S}}
はその境界となる微小閉曲線、
|
S
|
{\displaystyle |S|}
は
S
{\displaystyle S}
の表面積。
定義に従ってz軸方向の回転の式を導く。
C
{\displaystyle C}
を積分経路とする。
(
c
u
r
l
X
)
z
=
lim
∆
S
→
0
1
|
∆
S
|
∮
C
X
⋅
d
l
→
{\displaystyle (\mathrm {curl} \,\mathbf {X} )_{z}=\lim _{{\text{∆}}S\to 0}{\frac {1}{|{\text{∆}}S|}}\oint _{C}\mathbf {X} \cdot d{\vec {l}}}
=
lim
∆
x
→
0
∆
y
→
0
1
∆
x
∆
y
∮
C
X
⋅
d
l
→
{\displaystyle =\lim _{\scriptstyle {\text{∆}}x\to 0 \atop \scriptstyle {\text{∆}}y\to 0}{\frac {1}{{\text{∆}}x{\text{∆}}y}}\oint _{C}\mathbf {X} \cdot d{\vec {l}}}
=
lim
∆
x
→
0
∆
y
→
0
1
∆
x
∆
y
(
∫
∆
x
X
x
(
x
,
y
,
z
)
d
l
+
∫
∆
y
X
y
(
x
+
∆
x
,
y
,
z
)
d
l
−
∫
∆
x
X
x
(
x
,
y
+
∆
y
,
z
)
d
l
−
∫
∆
y
X
y
(
x
,
y
,
z
)
d
l
)
{\displaystyle =\lim _{\scriptstyle {\text{∆}}x\to 0 \atop \scriptstyle {\text{∆}}y\to 0}{\frac {1}{{\text{∆}}x{\text{∆}}y}}\left(\int _{{\text{∆}}x}X_{x}(x,y,z)dl+\int _{{\text{∆}}y}X_{y}(x+{\text{∆}}x,y,z)dl-\int _{{\text{∆}}x}X_{x}(x,y+{\text{∆}}y,z)dl-\int _{{\text{∆}}y}X_{y}(x,y,z)dl\right)}
=
lim
∆
x
→
0
∆
y
→
0
1
∆
x
∆
y
(
X
x
(
x
,
y
,
z
)
∆
x
+
X
y
(
x
+
∆
x
,
y
,
z
)
∆
y
−
X
x
(
x
,
y
+
∆
y
,
z
)
∆
x
−
X
y
(
x
,
y
,
z
)
∆
y
)
{\displaystyle =\lim _{\scriptstyle {\text{∆}}x\to 0 \atop \scriptstyle {\text{∆}}y\to 0}{\frac {1}{{\text{∆}}x{\text{∆}}y}}\left(X_{x}(x,y,z){\text{∆}}x+X_{y}(x+{\text{∆}}x,y,z){\text{∆}}y-X_{x}(x,y+{\text{∆}}y,z){\text{∆}}x-X_{y}(x,y,z){\text{∆}}y\right)}
=
lim
∆
x
→
0
∆
y
→
0
(
X
x
(
x
,
y
,
z
)
∆
y
+
X
y
(
x
+
∆
x
,
y
,
z
)
∆
x
−
X
x
(
x
,
y
+
∆
y
,
z
)
∆
y
−
X
y
(
x
,
y
,
z
)
∆
x
)
{\displaystyle =\lim _{\scriptstyle {\text{∆}}x\to 0 \atop \scriptstyle {\text{∆}}y\to 0}\left({\frac {X_{x}(x,y,z)}{{\text{∆}}y}}+{\frac {X_{y}(x+{\text{∆}}x,y,z)}{{\text{∆}}x}}-{\frac {X_{x}(x,y+{\text{∆}}y,z)}{{\text{∆}}y}}-{\frac {X_{y}(x,y,z)}{{\text{∆}}x}}\right)}
=
lim
∆
x
→
0
∆
y
→
0
(
X
y
(
x
,
y
+
∆
y
,
z
)
−
X
y
(
x
,
y
,
z
)
∆
x
−
X
x
(
x
+
∆
x
,
y
,
z
)
−
X
x
(
x
,
y
,
z
)
∆
y
)
{\displaystyle =\lim _{\scriptstyle {\text{∆}}x\to 0 \atop \scriptstyle {\text{∆}}y\to 0}\left({\frac {X_{y}(x,y+{\text{∆}}y,z)-X_{y}(x,y,z)}{{\text{∆}}x}}-{\frac {X_{x}(x+{\text{∆}}x,y,z)-X_{x}(x,y,z)}{{\text{∆}}y}}\right)}
=
∂
X
y
∂
x
−
∂
X
x
∂
y
{\displaystyle ={\frac {\partial {X_{y}}}{\partial {x}}}-{\frac {\partial {X_{x}}}{\partial {y}}}}
//
回転はベクトル量であるが、それを明示するため空間偏微分演算子による表記を考える。
成分に着目すると外積の形式なので、
c
u
r
l
X
=
∇
×
X
{\displaystyle \mathrm {curl} \,\mathbf {X} =\nabla \times \mathbf {X} }
と表されることが判る。
外積の性質から、回転は行列式
det
(
e
x
e
y
e
z
∂
∂
x
∂
∂
y
∂
∂
z
X
x
X
y
X
z
)
{\displaystyle \det {\begin{pmatrix}\mathbf {e} _{x}&\mathbf {e} _{y}&\mathbf {e} _{z}\\{\frac {\partial }{\partial {x}}}&{\frac {\partial }{\partial {y}}}&{\frac {\partial }{\partial {z}}}\\X_{x}&X_{y}&X_{z}\end{pmatrix}}}
でも表示される。
回転を用いた概念として、ファラデーの電磁誘導の法則が有名である。
電場
E
{\displaystyle \mathbf {E} }
、磁束密度
B
{\displaystyle \mathbf {B} }
として
r
o
t
E
=
−
∂
B
∂
t
{\displaystyle \mathrm {rot} \,\mathbf {E} =-{\frac {\partial {\mathbf {B} }}{\partial {t}}}}
流体の渦の記述にも応用されている。
回転は常に角速度ベクトルと一致するとは限らない。そのため、必ずしも回転運動そのものを記述しない。
回転の正式な定義から、以下のストークスの定理 (Stokes' theorem)が成り立つと判る。
∫
S
c
u
r
l
X
⋅
d
S
→
=
∮
C
X
⋅
d
l
→
{\displaystyle \int _{S}\mathrm {curl} \,\mathbf {X} \cdot d{\vec {S}}=\oint _{C}\mathbf {X} \cdot d{\vec {l}}}
但し、
C
{\displaystyle C}
は閉曲面
S
{\displaystyle S}
の境界曲線。
回転は、ヤコビ行列を用いると以下のように表される。
r
o
t
X
=
2
a
x
l
(
∇
⊛
X
−
t
(
∇
⊛
X
)
2
)
{\displaystyle \mathrm {rot} \,\mathbf {X} =2\mathrm {axl} \left({\frac {\nabla \circledast \mathbf {X} -{}^{t}\!(\nabla \circledast \mathbf {X} )}{2}}\right)}
ここで
∇
⊛
X
−
t
(
∇
⊛
X
)
{\displaystyle \nabla \circledast \mathbf {X} -{}^{t}\!(\nabla \circledast \mathbf {X} )}
は反対称行列であり、
a
x
l
{\displaystyle \mathrm {axl} }
は反対称行列を軸ベクトルに移す写像である。
以下のようにラプラス演算子 (ラプラス作用素 、Laplacian)を定義する。但し、
∆
{\displaystyle {\text{∆}}}
は通常のデルタ(Unicode文字名: GREEK CAPITAL LETTER DELTA, U+394)ではなく増分記号 (Unicode文字名:INCREMENT, U+2206)である(が通常のデルタで書かれることが多い)。
∆
=
∇
⋅
∇
=
∂
2
∂
x
2
+
∂
2
∂
y
2
+
∂
2
∂
z
2
{\displaystyle {\text{∆}}=\nabla \cdot \nabla ={\frac {\partial ^{2}}{\partial {x^{2}}}}+{\frac {\partial ^{2}}{\partial {y^{2}}}}+{\frac {\partial ^{2}}{\partial {z^{2}}}}}
第二式は通常、w:記号の濫用 により単に
∇
2
{\displaystyle \nabla ^{2}}
と書かれる。
スカラー場のラプラシアン を以下のように定義する。
∆
F
=
∂
2
F
∂
x
2
+
∂
2
F
∂
y
2
+
∂
2
F
∂
z
2
{\displaystyle {\text{∆}}F={\frac {\partial ^{2}F}{\partial {x^{2}}}}+{\frac {\partial ^{2}F}{\partial {y^{2}}}}+{\frac {\partial ^{2}F}{\partial {z^{2}}}}}
ラプラス演算子の定義により、左辺は以下のように変形される。
∆
F
=
∇
2
F
=
∇
⋅
(
∇
F
)
=
d
i
v
(
g
r
a
d
F
)
{\displaystyle {\text{∆}}F=\nabla ^{2}F=\nabla \cdot (\nabla F)=\mathrm {div} \,(\mathrm {grad} \,F)}
スカラー場のラプラシアンは、各成分方向の曲率の和、すなわちスカラー場の各点に於ける突出・沈み込みの度合いを表す。
ベクトル場のラプラシアンは、「各成分のラプラシアンを並べたベクトル」で表される。
∆
X
=
(
∆
X
x
∆
X
y
∆
X
z
)
{\displaystyle {\text{∆}}\mathbf {X} ={\begin{pmatrix}{\text{∆}}X_{x}\\{\text{∆}}X_{y}\\{\text{∆}}X_{z}\end{pmatrix}}}
ベクトル場では以下の等式が成り立つ。
∆
X
=
g
r
a
d
(
d
i
v
X
)
−
r
o
t
(
r
o
t
X
)
{\displaystyle {\text{∆}}\mathbf {X} =\mathrm {grad} \,(\mathrm {div} \,\mathbf {X} )-\mathrm {rot} \,(\mathrm {rot} \,\mathbf {X} )}
回転が
0
{\displaystyle {\boldsymbol {0}}}
ならば
∆
X
=
∇
(
∇
⋅
X
)
{\displaystyle {\text{∆}}\mathbf {X} =\nabla (\nabla \cdot \mathbf {X} )}
であり、形式的に
(
∇
⋅
)
∇
=
∇
(
∇
⋅
)
=
∇
2
{\displaystyle (\nabla \cdot )\nabla =\nabla (\nabla \cdot )=\nabla ^{2}}
が成り立つ。
ラプラシアンは、ポアソン方程式やシュレディンガー方程式、拡散方程式・波動方程式など、物理学の様々な局面で登場する。
ベクトル場
X
{\displaystyle \mathbf {X} }
にスカラー関数
ϕ
{\displaystyle \phi }
を掛けた関数について、その発散は以下であると容易に証明される。
d
i
v
(
ϕ
X
)
=
g
r
a
d
ϕ
⋅
X
+
ϕ
d
i
v
X
{\displaystyle \mathrm {div} \,(\phi \mathbf {X} )=\mathrm {grad} \,\phi \cdot \mathbf {X} +\phi \mathrm {div} \,\mathbf {X} }
ここで、
X
{\displaystyle \mathbf {X} }
をスカラー関数
ψ
{\displaystyle \psi }
の勾配
g
r
a
d
ψ
{\displaystyle \mathrm {grad} \,\psi }
と見做すと、
d
i
v
(
ϕ
g
r
a
d
ψ
)
=
g
r
a
d
ϕ
⋅
g
r
a
d
ψ
+
ϕ
d
i
v
(
g
r
a
d
ψ
)
{\displaystyle \mathrm {div} \,(\phi \mathrm {grad} \,\psi )=\mathrm {grad} \,\phi \cdot \mathrm {grad} \,\psi +\phi \mathrm {div} \,(\mathrm {grad} \,\psi )}
空間偏微分演算子及びラプラス演算子を用いて表記すると、
∇
⋅
(
ϕ
∇
ψ
)
=
∇
ϕ
⋅
∇
ψ
+
ϕ
∆
ψ
{\displaystyle \nabla \cdot (\phi \nabla \psi )=\nabla \phi \cdot \nabla \psi +\phi {\text{∆}}\psi }
両辺を三次元領域Vに関して体積分すると、
∫
V
∇
(
ϕ
∇
ψ
)
d
V
=
∫
V
(
∇
ϕ
⋅
∇
ψ
+
ϕ
∆
ψ
)
d
V
{\displaystyle \int _{V}\nabla (\phi \nabla \psi )dV=\int _{V}(\nabla \phi \cdot \nabla \psi +\phi {\text{∆}}\psi )dV}
Vの境界曲面をSとして左辺にガウスの発散定理を適用すると、
∮
S
(
ϕ
∇
ψ
)
⋅
d
S
→
=
∫
V
(
ϕ
∆
ψ
+
∇
ϕ
⋅
∇
ψ
)
d
V
{\displaystyle \oint _{S}(\phi \nabla \psi )\cdot d{\vec {S}}=\int _{V}(\phi {\text{∆}}\psi +\nabla \phi \cdot \nabla \psi )dV}
これをグリーンの第一定理 (Green's first theorem)という。
グリーンの第一定理に於いて、
ϕ
{\displaystyle \phi }
と
ψ
{\displaystyle \psi }
を入れ替えた式は
∮
S
(
ψ
∇
ϕ
)
⋅
d
S
→
=
∫
V
(
ψ
∆
ϕ
+
∇
ψ
⋅
∇
ϕ
)
d
V
{\displaystyle \oint _{S}(\psi \nabla \phi )\cdot d{\vec {S}}=\int _{V}(\psi {\text{∆}}\phi +\nabla \psi \cdot \nabla \phi )dV}
元の式との差をとると
∫
S
(
ϕ
∇
ψ
−
ψ
∇
ϕ
)
⋅
d
S
→
=
∫
V
(
ϕ
∆
ψ
−
ψ
∆
ϕ
)
d
V
{\displaystyle \int _{S}(\phi \nabla \psi -\psi \nabla \phi )\cdot d{\vec {S}}=\int _{V}(\phi {\text{∆}}\psi -\psi {\text{∆}}\phi )dV}
これをグリーンの第二定理 (Green's second theorem)という。
S
{\displaystyle S}
の単位法線ベクトルを
n
{\displaystyle \mathbf {n} }
、その方向の軸を
n
{\displaystyle n}
とすると、
(
ϕ
∇
ψ
)
⋅
d
S
→
=
(
ϕ
∇
ψ
)
⋅
n
d
S
=
ϕ
∂
ψ
∂
n
d
S
{\displaystyle (\phi \nabla \psi )\cdot d{\vec {S}}=(\phi \nabla \psi )\cdot \mathbf {n} dS=\phi {\frac {\partial {\psi }}{\partial {n}}}dS}
グリーンの第二定理の左辺を書き換えて
∫
S
(
ϕ
∂
ψ
∂
n
−
ψ
∂
ϕ
∂
n
)
d
S
=
∫
V
(
ϕ
∆
ψ
−
ψ
∆
ϕ
)
d
V
{\displaystyle \int _{S}\left(\phi {\frac {\partial {\psi }}{\partial {n}}}-\psi {\frac {\partial {\phi }}{\partial {n}}}\right)dS=\int _{V}(\phi {\text{∆}}\psi -\psi {\text{∆}}\phi )dV}
これをグリーンの定理 (Green's theorem)という。
この節は書きかけです。この節を編集してくれる方 を心からお待ちしています。
この節では、上述したガウスの発散定理、ストークスの定理、グリーンの定理を厳密な議論により証明する。
補足事項
一般に
r
o
t
(
g
r
a
d
F
)
=
d
i
v
(
r
o
t
X
)
=
0
{\displaystyle \mathrm {rot} \,(\mathrm {grad} \,F)=\mathrm {div} \,(\mathrm {rot} \,\mathbf {X} )=0}
が成り立つ。証明は下記の演習問題で行う。
r
o
t
x
=
0
{\displaystyle \mathrm {rot} \,\mathbf {x} ={\boldsymbol {0}}}
を満たすベクトル場
x
{\displaystyle \mathbf {x} }
を層状ベクトル場 (lamellar vector field)という。回転を持たないことから非回転的ベクトル場 (irrotational vector field)ともいう。
上記の性質から、
x
=
g
r
a
d
f
{\displaystyle \mathbf {x} =\mathrm {grad} \,f}
となるスカラー場
f
{\displaystyle f}
が存在する。この
f
{\displaystyle f}
を、
x
{\displaystyle \mathbf {x} }
のスカラーポテンシャル (scalar potential)という。
d
i
v
y
=
0
{\displaystyle \mathrm {div} \,\mathbf {y} =0}
を満たすスカラー場
y
{\displaystyle \mathbf {y} }
を管状ベクトル場 (solenoidal vector field)という。発散を持たないことから回転的ベクトル場 (rotational vector field)ともいう。
上記の性質から、
y
=
r
o
t
g
{\displaystyle \mathbf {y} =\mathrm {rot} \,\mathbf {g} }
となるベクトル場
g
{\displaystyle \mathbf {g} }
が存在する。この
g
{\displaystyle \mathbf {g} }
を、
y
{\displaystyle \mathbf {y} }
のベクトルポテンシャル (vector potential)という。
スカラーポテンシャル及びベクトルポテンシャルの存在は、w:ポアンカレの補題 の特別な場合であり、ユークリッド空間
R
n
{\displaystyle \mathbb {R} ^{n}}
では一般に成り立つ。
演習問題
f
(
x
,
y
,
z
)
=
(
y
z
z
x
x
y
)
{\displaystyle {\boldsymbol {f}}(x,y,z)={\begin{pmatrix}yz\\zx\\xy\end{pmatrix}}}
の曲率・捩率・流線方程式を求めよ。但し、
∫
−
1
1
∫
−
1
−
z
2
1
−
z
2
∫
−
1
−
z
2
−
y
2
1
−
z
2
−
y
2
d
x
d
y
d
z
{\displaystyle \int _{-1}^{1}\int _{-{\sqrt {1-z^{2}}}}^{\sqrt {1-z^{2}}}\int _{-{\sqrt {1-z^{2}-y^{2}}}}^{\sqrt {1-z^{2}-y^{2}}}dxdydz}
を円筒座標または極座標に変数変換し、計算せよ。
r
o
t
(
X
+
Y
)
=
r
o
t
X
+
r
o
t
Y
{\displaystyle \mathrm {rot} \,(\mathbf {X} +\mathbf {Y} )=\mathrm {rot} \,\mathbf {X} +\mathrm {rot} \,\mathbf {Y} }
を示せ。
c
u
r
l
(
g
r
a
d
F
)
=
d
i
v
(
c
u
r
l
X
)
=
0
{\displaystyle \mathrm {curl} \,(\mathrm {grad} \,F)=\mathrm {div} \,(\mathrm {curl} \,\mathbf {X} )=0}
を証明せよ。
∆
X
=
g
r
a
d
(
d
i
v
X
)
−
r
o
t
(
r
o
t
X
)
{\displaystyle {\text{∆}}\mathbf {X} =\mathrm {grad} \,(\mathrm {div} \,\mathbf {X} )-\mathrm {rot} \,(\mathrm {rot} \,\mathbf {X} )}
を確かめよ。
線素ベクトル・面素ベクトル・体素を円筒座標系の基本ベクトルを用いて表せ。
極座標系に於ける勾配・発散・回転・ラプラシアンを導出せよ。
層状ベクトル場上の線積分は積分経路に依存せず、始点と終点のみに依存することを証明せよ。
f
=
(
3
x
5
y
7
z
)
{\displaystyle {\boldsymbol {f}}={\begin{pmatrix}3x\\5y\\7z\end{pmatrix}}}
とする。円筒側面
x
2
+
y
2
=
1
,
−
2
≤
z
≤
2
{\displaystyle x^{2}+y^{2}=1,-2\leq z\leq 2}
上での面積分を求めよ。但し、面積ベクトルは円筒面に対して外向きとする。
複素解析学 も参照。
この節は書きかけです。この節を編集してくれる方 を心からお待ちしています。
この節は書きかけです。この節を編集してくれる方 を心からお待ちしています。
Wikipedia
多様体論 に発展した内容の記事があります。