赞
踩
定义 e ^ i ⨂ e ^ j \hat e_i \bigotimes \hat e_j e^i⨂e^j的迹:
T r ( e ^ i ⨂ e ^ j ) = e ^ i ⋅ e ^ j = δ i j Tr(\hat e_i \bigotimes \hat e_j) = \hat e_i \cdot \hat e_j = \delta_{ij} Tr(e^i⨂e^j)=e^i⋅e^j=δij
所以,可以定义二阶张量的迹为:对角线元素加和
T r ( A ) = T r ( A i j e ^ i ⨂ e ^ j ) = A i j T r ( e ^ i ⨂ e ^ j ) = A i j δ i j = A i i Tr(A) = Tr(A_{ij}\hat e_i \bigotimes \hat e_j) = A_{ij}Tr(\hat e_i \bigotimes \hat e_j) = A_{ij}\delta_{ij} = A_{ii} Tr(A)=Tr(Aije^i⨂e^j)=AijTr(e^i⨂e^j)=Aijδij=Aii
并矢的迹为:
T
r
(
u
⃗
⨂
v
⃗
)
=
u
⃗
i
v
⃗
j
T
r
(
e
^
i
⨂
e
^
j
)
=
u
⃗
i
v
⃗
j
δ
i
j
=
u
⃗
i
v
⃗
i
=
u
⃗
⋅
v
⃗
Tr(\vec u \bigotimes \vec v) = \vec u_i \vec v_jTr(\hat e_i \bigotimes \hat e_j)= \vec u_i \vec v_j\delta_{ij} = \vec u_i \vec v_i = \vec u \cdot \vec v
Tr(u
⨂v
)=u
iv
jTr(e^i⨂e^j)=u
iv
jδij=u
iv
i=u
⋅v
NOTE: 张量的迹是一个不变量,是独立于坐标系的
张量的迹的性质:
同样地,可以验证:
单位张量:
1
=
δ
i
j
e
^
i
⨂
e
^
j
=
e
^
i
⨂
e
^
i
=
1
e
^
i
⨂
e
^
j
1 = \delta_{ij}\hat e_i \bigotimes \hat e_j = \hat e_i \bigotimes \hat e_i = 1 \hat e_i \bigotimes \hat e_j
1=δije^i⨂e^j=e^i⨂e^i=1e^i⨂e^j
四阶单位张量的定义:
可以计算:
I
:
A
=
(
δ
i
k
δ
j
l
e
^
i
⨂
e
^
j
⨂
e
^
k
⨂
e
^
l
)
:
(
A
p
q
e
^
p
⨂
e
^
q
)
=
δ
i
k
δ
j
l
A
p
q
δ
k
p
δ
l
q
e
^
i
⨂
e
^
j
=
δ
i
k
δ
j
l
A
k
l
e
^
i
⨂
e
^
j
=
A
i
j
e
^
i
⨂
e
^
j
=
A
I:A = (\delta_{ik}\delta_{jl}\hat e_i \bigotimes \hat e_j \bigotimes \hat e_k\bigotimes\hat e_l ): (A_{pq}\hat e_p \bigotimes \hat e_q) \\ =\delta_{ik}\delta_{jl}A_{pq} \delta_{kp} \delta_{lq} \hat e_i \bigotimes \hat e_j \\ =\delta_{ik}\delta_{jl}A_{kl} \hat e_i \bigotimes \hat e_j \\ =A_{ij} \hat e_i \bigotimes \hat e_j \\= A
I:A=(δikδjle^i⨂e^j⨂e^k⨂e^l):(Apqe^p⨂e^q)=δikδjlApqδkpδlqe^i⨂e^j=δikδjlAkle^i⨂e^j=Aije^i⨂e^j=A
I ˉ : A = ( δ i l δ j k e ^ i ⨂ e ^ j ⨂ e ^ k ⨂ e ^ l ) : ( A p q e ^ p ⨂ e ^ q ) = δ i l δ j k A p q δ k p δ l q e ^ i ⨂ e ^ j = δ i l δ j k A k l e ^ i ⨂ e ^ j = A j i e ^ i ⨂ e ^ j = A T \bar I:A = (\delta_{il}\delta_{jk}\hat e_i \bigotimes \hat e_j \bigotimes \hat e_k\bigotimes\hat e_l ): (A_{pq}\hat e_p \bigotimes \hat e_q) \\ =\delta_{il}\delta_{jk}A_{pq} \delta_{kp} \delta_{lq} \hat e_i \bigotimes \hat e_j \\ =\delta_{il}\delta_{jk}A_{kl} \hat e_i \bigotimes \hat e_j \\ =A_{ji} \hat e_i \bigotimes \hat e_j \\= A^T Iˉ:A=(δilδjke^i⨂e^j⨂e^k⨂e^l):(Apqe^p⨂e^q)=δilδjkApqδkpδlqe^i⨂e^j=δilδjkAkle^i⨂e^j=Ajie^i⨂e^j=AT
I ‾ ‾ : A = ( δ i j δ k l e ^ i ⨂ e ^ j ⨂ e ^ k ⨂ e ^ l ) : ( A p q e ^ p ⨂ e ^ q ) = δ i j δ k l A p q δ k p δ l q e ^ i ⨂ e ^ j = δ i j δ k l A k l e ^ i ⨂ e ^ j = A k k δ i j e ^ i ⨂ e ^ j = T r ( A ) 1 \overline{\overline{I}} :A = (\delta_{ij}\delta_{kl}\hat e_i \bigotimes \hat e_j \bigotimes \hat e_k\bigotimes\hat e_l ): (A_{pq}\hat e_p \bigotimes \hat e_q) \\ =\delta_{ij}\delta_{kl}A_{pq} \delta_{kp} \delta_{lq} \hat e_i \bigotimes \hat e_j \\ =\delta_{ij}\delta_{kl}A_{kl} \hat e_i \bigotimes \hat e_j \\ =A_{kk}\delta_{ij}\hat e_i \bigotimes \hat e_j \\= Tr(A) 1 I:A=(δijδkle^i⨂e^j⨂e^k⨂e^l):(Apqe^p⨂e^q)=δijδklApqδkpδlqe^i⨂e^j=δijδklAkle^i⨂e^j=Akkδije^i⨂e^j=Tr(A)1
四阶单位张量的对称部分:
张量乘积符号
⨂
‾
\overline{\bigotimes}
⨂ 的定义如下:
与以下是一样的:
张量乘积符号
⨂
‾
\underline{\bigotimes}
⨂ 的定义如下:
四阶单位张量的反对称部分:
以下等式成立:
b
⃗
⋅
1
=
b
⃗
\vec b \cdot 1 = \vec b
b
⋅1=b
I
:
A
=
A
I:A = A
I:A=A
I
s
y
m
:
A
=
A
s
y
m
I^{sym}:A = A^{sym}
Isym:A=Asym
A
:
1
=
T
r
(
A
)
=
A
i
i
A:1 = Tr(A) =A_{ii}
A:1=Tr(A)=Aii
A
2
:
1
=
T
r
(
A
2
)
=
T
r
(
A
⋅
A
)
=
A
i
l
A
l
i
A^2:1 = Tr(A^2)=Tr(A\cdot A) = A_{il}A_{li}
A2:1=Tr(A2)=Tr(A⋅A)=AilAli
A
3
:
1
=
T
r
(
A
3
)
=
T
r
(
A
⋅
A
⋅
A
)
=
A
i
j
A
j
k
A
k
l
A^3:1 = Tr(A^3)=Tr(A\cdot A \cdot A)=A_{ij}A_{jk}A_{kl}
A3:1=Tr(A3)=Tr(A⋅A⋅A)=AijAjkAkl
Levi-Civita 伪张量(置换张量):
是一个三阶伪张量:
ϵ = ϵ i j k e ^ i ⨂ e ^ j ⨂ e ^ k \epsilon = \epsilon_{ijk} \hat e_i \bigotimes \hat e_j \bigotimes \hat e_k ϵ=ϵijke^i⨂e^j⨂e^k
张量的行列式:
det ( A ) ≡ ∣ A ∣ = ϵ i j k A 1 i A 2 j A 3 k = ϵ i j k A i 1 A j 2 A k 3 \det (A) \equiv|A| = \epsilon_{ijk}A_{1i}A_{2j}A_{3k} = \epsilon_{ijk}A_{i1}A_{j2}A_{k3} det(A)≡∣A∣=ϵijkA1iA2jA3k=ϵijkAi1Aj2Ak3
张量的行列式也是一个不变量(独立于坐标系)
行列式的性质:
det
1
=
1
\det 1 = 1
det1=1
det
A
T
=
det
A
\det A^T = \det A
detAT=detA
det
(
A
⋅
B
)
=
det
A
⋅
det
B
\det (A \cdot B) =\det A \cdot \det B
det(A⋅B)=detA⋅detB
det
(
α
A
)
=
α
3
det
A
\det (\alpha A) = \alpha^3 \det A
det(αA)=α3detA
如果
det
A
=
0
\det A = 0
detA=0,则A是奇异的
如果交换两行或两列,则行列式的符号改变
如果某一行或某一列的所有元素为0, 则行列式为0
如果对某一行或某一列所有元素乘以c,则行列式变为
c
∣
A
∣
c|A|
c∣A∣
如果两行(列)或以上是线性相关的,则行列式为0
以上用到:
其中用到:
由于以下等式也成立:
det
[
1
+
α
(
a
⃗
⨂
b
⃗
)
+
β
(
b
⃗
⨂
a
⃗
)
]
=
1
+
α
(
a
⃗
⋅
b
⃗
)
+
β
(
a
⃗
⋅
b
⃗
)
+
α
β
[
(
a
⃗
⋅
b
⃗
)
2
−
(
a
⃗
⋅
a
⃗
)
(
b
⃗
⋅
b
⃗
)
]
\det [1 + \alpha (\vec a \bigotimes \vec b) + \beta (\vec b \bigotimes \vec a)] = 1 + \alpha (\vec a \cdot \vec b) +\beta (\vec a \cdot \vec b) + \alpha \beta [(\vec a \cdot \vec b)^2 - (\vec a \cdot \vec a)(\vec b \cdot \vec b)]
det[1+α(a
⨂b
)+β(b
⨂a
)]=1+α(a
⋅b
)+β(a
⋅b
)+αβ[(a
⋅b
)2−(a
⋅a
)(b
⋅b
)]
令
β
=
0
\beta = 0
β=0, 也可以得到
det
(
μ
1
+
α
a
⃗
⨂
b
⃗
)
=
μ
3
+
μ
2
α
a
⃗
⋅
b
⃗
\det (\mu 1 + \alpha \vec a \bigotimes \vec b) = \mu^3 + \mu^2 \alpha \vec a \cdot \vec b
det(μ1+αa
⨂b
)=μ3+μ2αa
⋅b
如果
α
=
β
\alpha = \beta
α=β, 有:
det
[
1
+
α
(
a
⃗
⨂
b
⃗
)
+
α
(
b
⃗
⨂
a
⃗
)
]
=
1
+
α
(
a
⃗
⋅
b
⃗
)
+
α
(
a
⃗
⋅
b
⃗
)
+
α
2
[
(
a
⃗
⋅
b
⃗
)
2
−
(
a
⃗
⋅
a
⃗
)
(
b
⃗
⋅
b
⃗
)
]
=
1
+
α
[
2
(
a
⃗
⋅
b
⃗
)
−
α
(
a
⃗
∧
b
⃗
)
2
]
\det [1 + \alpha (\vec a \bigotimes \vec b) + \alpha (\vec b \bigotimes \vec a)] \\= 1 + \alpha (\vec a \cdot \vec b) +\alpha (\vec a \cdot \vec b) + \alpha^2 [(\vec a \cdot \vec b)^2 - (\vec a \cdot \vec a)(\vec b \cdot \vec b)] \\ = 1 + \alpha [2(\vec a \cdot \vec b) - \alpha (\vec a \wedge \vec b)^2]
det[1+α(a
⨂b
)+α(b
⨂a
)]=1+α(a
⋅b
)+α(a
⋅b
)+α2[(a
⋅b
)2−(a
⋅a
)(b
⋅b
)]=1+α[2(a
⋅b
)−α(a
∧b
)2]
其中,用到:
以下等式成立:
det
(
α
A
+
β
B
)
=
α
3
det
A
+
α
2
β
T
r
(
B
⋅
a
d
j
(
A
)
)
+
α
β
2
[
A
⋅
a
d
j
(
B
)
]
+
β
3
det
B
\det (\alpha A + \beta B) \\ =\alpha^3\det A + \alpha^2 \beta Tr(B \cdot adj(A))+ \alpha \beta^2 [A \cdot adj(B)] + \beta^3 \det B
det(αA+βB)=α3detA+α2βTr(B⋅adj(A))+αβ2[A⋅adj(B)]+β3detB
当
α
=
1
,
A
=
1
,
B
=
a
⃗
⨂
b
⃗
\alpha = 1, A = 1, B = \vec a \bigotimes \vec b
α=1,A=1,B=a
⨂b
, 且由于
det
(
a
⃗
⨂
b
⃗
)
=
0
\det (\vec a \bigotimes \vec b) = 0
det(a
⨂b
)=0 以及
c
o
f
(
a
⃗
⨂
b
⃗
)
=
0
cof(\vec a \bigotimes \vec b) = 0
cof(a
⨂b
)=0, 有
det ( 1 + β a ⃗ ⨂ b ⃗ ) = det 1 + β T r [ a ⃗ ⨂ b ⃗ ⋅ 1 ] = 1 + β T r [ a i b j ] = 1 + β a ⃗ ⋅ b ⃗ \det (1 + \beta \vec a \bigotimes \vec b) = \det 1 + \beta Tr[\vec a \bigotimes \vec b \cdot 1] = 1+\beta Tr[a_ib_j] = 1+ \beta \vec a \cdot \vec b det(1+βa ⨂b )=det1+βTr[a ⨂b ⋅1]=1+βTr[aibj]=1+βa ⋅b
性质:
(
A
⋅
a
⃗
)
⋅
[
(
A
⋅
b
⃗
)
∧
(
A
⋅
c
⃗
)
]
=
det
(
A
)
[
a
⃗
⋅
b
⃗
∧
c
⃗
]
\boxed{(A \cdot \vec a) \cdot [(A \cdot \vec b) \wedge (A \cdot \vec c)] = \det (A)[\vec a \cdot \vec b \wedge \vec c]}
(A⋅a
)⋅[(A⋅b
)∧(A⋅c
)]=det(A)[a
⋅b
∧c
]
证明:
由于
a
⃗
⋅
(
b
⃗
∧
c
⃗
)
=
ϵ
i
j
k
a
i
b
j
c
k
\vec a \cdot (\vec b \wedge \vec c) = \epsilon_{ijk} a_i b_j c_k
a
⋅(b
∧c
)=ϵijkaibjck
两边同乘
∣
A
∣
|A|
∣A∣:
a
⃗
⋅
(
b
⃗
∧
c
⃗
)
∣
A
∣
=
ϵ
i
j
k
a
i
b
j
c
k
∣
A
∣
\vec a \cdot (\vec b \wedge \vec c) |A| = \epsilon_{ijk} a_i b_j c_k |A|
a
⋅(b
∧c
)∣A∣=ϵijkaibjck∣A∣
又由于
∣
A
∣
ϵ
i
j
k
=
ϵ
p
q
r
A
p
i
A
q
j
A
r
k
|A|\epsilon_{ijk} = \epsilon_{pqr} A_{pi}A_{qj}A_{rk}
∣A∣ϵijk=ϵpqrApiAqjArk, 因此:
a
⃗
⋅
(
b
⃗
∧
c
⃗
)
∣
A
∣
=
ϵ
i
j
k
a
i
b
j
c
k
∣
A
∣
=
ϵ
p
q
r
A
p
i
A
q
j
A
r
k
a
i
b
j
c
k
=
ϵ
p
q
r
A
p
i
a
i
A
q
j
b
j
A
r
k
c
k
=
(
A
⋅
a
⃗
)
⋅
[
(
A
⋅
b
⃗
)
∧
(
A
⋅
c
⃗
)
]
=
[
(
A
⋅
b
⃗
)
∧
(
A
⋅
c
⃗
)
]
⋅
(
A
⋅
a
⃗
)
\vec a \cdot (\vec b \wedge \vec c) |A| = \epsilon_{ijk} a_i b_j c_k |A| \\=\epsilon_{pqr} A_{pi}A_{qj}A_{rk} a_i b_j c_k \\=\epsilon_{pqr} A_{pi}a_iA_{qj}b_jA_{rk}c_k \\ =(A \cdot \vec a) \cdot [(A \cdot \vec b) \wedge (A \cdot \vec c)] \\= [(A \cdot \vec b) \wedge (A \cdot \vec c)]\cdot (A \cdot \vec a)
a
⋅(b
∧c
)∣A∣=ϵijkaibjck∣A∣=ϵpqrApiAqjArkaibjck=ϵpqrApiaiAqjbjArkck=(A⋅a
)⋅[(A⋅b
)∧(A⋅c
)]=[(A⋅b
)∧(A⋅c
)]⋅(A⋅a
)
张量的逆:
A
−
1
A^{-1}
A−1
指标形式:
为了计算张量的逆,我们从伴随张量(Adjugate Tensor)开始:
a
d
j
(
A
T
)
⋅
(
a
⃗
∧
b
⃗
)
=
(
A
⋅
a
⃗
)
∧
(
A
⋅
b
⃗
)
adj(A^T)\cdot (\vec a \wedge \vec b)=(A \cdot \vec a)\wedge (A \cdot \vec b)
adj(AT)⋅(a
∧b
)=(A⋅a
)∧(A⋅b
)
两边同乘以一个任意的向量
d
⃗
\vec d
d
上一节,证明了:
a
⃗
⋅
(
b
⃗
∧
c
⃗
)
∣
A
∣
=
[
(
A
⋅
b
⃗
)
∧
(
A
⋅
c
⃗
)
]
⋅
(
A
⋅
a
⃗
)
\vec a \cdot (\vec b \wedge \vec c) |A| = [(A \cdot \vec b) \wedge (A \cdot \vec c)]\cdot (A \cdot \vec a)
a
⋅(b
∧c
)∣A∣=[(A⋅b
)∧(A⋅c
)]⋅(A⋅a
)
所以有:
c
⃗
⋅
(
a
⃗
∧
b
⃗
)
∣
A
∣
=
[
(
A
⋅
a
⃗
)
∧
(
A
⋅
b
⃗
)
]
⋅
(
A
⋅
c
⃗
)
\vec c \cdot (\vec a \wedge \vec b) |A| = [(A \cdot \vec a) \wedge (A \cdot \vec b)]\cdot (A \cdot \vec c)
c
⋅(a
∧b
)∣A∣=[(A⋅a
)∧(A⋅b
)]⋅(A⋅c
)
因此:
[
(
a
d
j
(
A
)
)
T
⋅
(
a
⃗
∧
b
⃗
)
]
⋅
d
⃗
=
[
(
A
⋅
a
⃗
)
∧
(
A
⋅
b
⃗
)
]
⋅
(
A
⋅
A
−
1
⋅
d
⃗
)
=
A
−
1
⋅
d
⃗
⋅
(
a
⃗
∧
b
⃗
)
∣
A
∣
=
∣
A
∣
(
a
⃗
∧
b
⃗
)
⋅
A
−
1
⋅
d
⃗
[ (adj(A))^T\cdot (\vec a \wedge \vec b) ] \cdot \vec d =[(A \cdot \vec a) \wedge (A \cdot \vec b)]\cdot (A \cdot A^{-1} \cdot \vec d) \\=A^{-1} \cdot \vec d \cdot (\vec a \wedge \vec b) |A| \\ =|A| (\vec a \wedge \vec b)\cdot A^{-1} \cdot \vec d
[(adj(A))T⋅(a
∧b
)]⋅d
=[(A⋅a
)∧(A⋅b
)]⋅(A⋅A−1⋅d
)=A−1⋅d
⋅(a
∧b
)∣A∣=∣A∣(a
∧b
)⋅A−1⋅d
令
p
⃗
=
a
⃗
∧
b
⃗
\vec p = \vec a \wedge \vec b
p
=a
∧b
, 上式可以重写为:
因此:
[
a
d
j
(
A
)
]
=
∣
A
∣
A
−
1
[adj(A)] = |A| A^{-1}
[adj(A)]=∣A∣A−1
所以, 张量的逆,
A
−
1
=
1
∣
A
∣
[
a
d
j
(
A
)
]
=
1
∣
A
∣
[
c
o
f
(
A
)
]
T
\boxed{A^{-1} = \frac{1}{|A|}[adj(A)] = \frac{1}{|A|}[cof(A)]^T}
A−1=∣A∣1[adj(A)]=∣A∣1[cof(A)]T
可逆张量,有以下性质:
(
A
⋅
B
)
−
1
=
B
−
1
⋅
A
−
1
(A \cdot B)^{-1} = B^{-1} \cdot A^{-1}
(A⋅B)−1=B−1⋅A−1
(
A
−
1
)
−
1
=
A
(A^{-1})^{-1} = A
(A−1)−1=A
(
β
A
)
−
1
=
1
β
A
−
1
(\beta A)^{-1} = \frac{1}{\beta}A^{-1}
(βA)−1=β1A−1
det
(
A
−
1
=
[
det
A
]
−
1
)
\det (A^{-1} = [\det A]^{-1})
det(A−1=[detA]−1)
可逆转置:
A
−
T
≡
(
A
−
1
)
T
≡
(
A
T
)
−
1
A^{-T} \equiv (A^{-1})^T \equiv (A^T)^{-1}
A−T≡(A−1)T≡(AT)−1
证明
a
d
j
(
A
⋅
B
)
=
a
d
j
(
A
)
⋅
a
d
j
(
B
)
adj(A \cdot B) = adj(A) \cdot adj(B)
adj(A⋅B)=adj(A)⋅adj(B):
其中,用到了性质:
∣
A
⋅
B
∣
=
∣
A
∣
∣
B
∣
|A \cdot B| = |A| |B|
∣A⋅B∣=∣A∣∣B∣
同样的, 可以证明:
c
o
f
(
A
⋅
B
)
=
[
c
o
f
(
A
)
]
⋅
[
c
o
f
(
B
)
]
cof(A \cdot B) = [cof(A)] \cdot [cof(B)]
cof(A⋅B)=[cof(A)]⋅[cof(B)]
计算矩阵A的逆的步骤:
计算矩阵的余子式: cof(A)
定义一个矩阵M, 其元素
M
i
j
M_{ij}
Mij为矩阵A消去第i行第j列之后的行列式:
所以,定义矩阵A的余子式为:
c
o
f
(
A
)
=
(
−
1
)
i
+
j
M
i
j
cof (A) = (-1)^{i+j}M_{ij}
cof(A)=(−1)i+jMij
计算伴随矩阵, adj(A),
a
d
j
(
A
)
=
[
c
o
f
(
A
)
]
T
adj(A) = [cof(A)]^T
adj(A)=[cof(A)]T
计算矩阵的逆:
A
−
1
=
a
d
j
(
A
)
∣
A
∣
A^{-1} = \frac{adj(A)}{|A|}
A−1=∣A∣adj(A)
由于:
可以将矩阵的余子式的每一行写成:
M
1
i
=
ϵ
i
j
k
A
2
j
A
3
k
M_{1i} = \epsilon_{ijk}A_{2j}A_{3k}
M1i=ϵijkA2jA3k
M
2
i
=
ϵ
i
j
k
A
1
j
A
3
k
M_{2i} = \epsilon_{ijk}A_{1j}A_{3k}
M2i=ϵijkA1jA3k
M
3
i
=
ϵ
i
j
k
A
1
j
A
2
k
M_{3i} = \epsilon_{ijk}A_{1j}A_{2k}
M3i=ϵijkA1jA2k
在连续介质力学,正交张量发挥了很重要的角色
若二阶张量
Q
Q
Q满足其转置与其逆相等:
Q
T
=
Q
−
1
Q^T = Q^{-1}
QT=Q−1,则称
Q
Q
Q 是正交张量
Q
⋅
Q
T
=
Q
T
⋅
Q
=
1
Q \cdot Q^T = Q^T \cdot Q = 1
Q⋅QT=QT⋅Q=1
Q
i
k
Q
j
k
=
Q
k
i
Q
k
j
=
δ
i
j
Q_{ik}Q_{jk} = Q_{ki}Q_{kj}= \delta_{ij}
QikQjk=QkiQkj=δij
性质:
两个正交张量点积得到的张量也是正交的:
C
−
1
=
(
A
⋅
B
)
−
1
=
B
−
1
⋅
A
−
1
=
B
T
⋅
A
T
=
(
A
⋅
B
)
T
=
C
T
C^{-1} = (A \cdot B)^{-1}= B^{-1} \cdot A^{-1} = B^T \cdot A^T= (A\cdot B)^T=C^T
C−1=(A⋅B)−1=B−1⋅A−1=BT⋅AT=(A⋅B)T=CT
正交变换:
a
~
⃗
=
Q
⋅
a
⃗
\vec{\tilde a}=Q \cdot \vec a
a~
=Q⋅a
b
~
⃗
=
Q
⋅
b
⃗
\vec{\tilde b}=Q \cdot \vec b
b~
=Q⋅b
以上向量点积:
a
~
⃗
⋅
b
~
⃗
=
(
Q
⋅
a
⃗
)
⋅
(
Q
⋅
b
⃗
)
=
a
⃗
⋅
Q
T
⋅
Q
⋅
b
⃗
=
a
⃗
⋅
1
⋅
b
⃗
=
a
⃗
⋅
b
⃗
\vec{\tilde a} \cdot \vec{\tilde b}=(Q \cdot \vec a) \cdot (Q \cdot \vec b)=\vec a\cdot Q^T\cdot Q\cdot \vec b=\vec a\cdot1\cdot \vec b = \vec a \cdot \vec b
a~
⋅b~
=(Q⋅a
)⋅(Q⋅b
)=a
⋅QT⋅Q⋅b
=a
⋅1⋅b
=a
⋅b
a
~
i
b
~
i
=
(
Q
i
k
a
k
)
(
Q
i
j
b
j
)
=
a
k
(
Q
i
k
Q
i
j
)
b
j
=
a
k
δ
k
j
b
j
=
a
k
b
k
\tilde a_i \tilde b_i=(Q_{ik}a_k) (Q_{ij}b_j)=a_k(Q_{ik}Q_{ij})b_j=a_k\delta_{kj}b_j=a_kb_k
a~ib~i=(Qikak)(Qijbj)=ak(QikQij)bj=akδkjbj=akbk
同样的,对于
a
~
⃗
=
b
~
⃗
\vec{\tilde a} = \vec{\tilde b}
a~
=b~
,,有:
a
~
⃗
⋅
a
~
⃗
=
∣
∣
a
~
⃗
∣
∣
2
=
a
⃗
⋅
a
⃗
=
∣
∣
a
⃗
∣
∣
2
\vec{\tilde a} \cdot \vec{\tilde a}=||\vec{\tilde a}||^2=\vec a \cdot \vec a=||\vec a||^2
a~
⋅a~
=∣∣a~
∣∣2=a
⋅a
=∣∣a
∣∣2
因此,在一个正交变换中,向量的大小和它们之间的角度不会改变。
正定的 positive definite: 对于所有
x
^
≠
0
⃗
\hat x \neq \vec 0
x^=0
张量表示:
x
^
⋅
T
⋅
x
^
>
0
\hat x \cdot T \cdot \hat x >0
x^⋅T⋅x^>0
指标表示:
x
i
T
i
j
x
j
>
0
x_i T_{ij}x_j > 0
xiTijxj>0
负定的 negative definite: 对于所有
x
^
≠
0
⃗
\hat x \neq \vec 0
x^=0
张量表示:
x
^
⋅
T
⋅
x
^
<
0
\hat x \cdot T \cdot \hat x <0
x^⋅T⋅x^<0
指标表示:
x
i
T
i
j
x
j
<
0
x_i T_{ij}x_j < 0
xiTijxj<0
半正定的 semi-positive definite: 对于所有
x
^
≠
0
⃗
\hat x \neq \vec 0
x^=0
张量表示:
x
^
⋅
T
⋅
x
^
≥
0
\hat x \cdot T \cdot \hat x \geq 0
x^⋅T⋅x^≥0
指标表示:
x
i
T
i
j
x
j
≥
0
x_i T_{ij}x_j \geq 0
xiTijxj≥0
负定的 negative definite: 对于所有
x
^
≠
0
⃗
\hat x \neq \vec 0
x^=0
张量表示:
x
^
⋅
T
⋅
x
^
≤
0
\hat x \cdot T \cdot \hat x \leq 0
x^⋅T⋅x^≤0
指标表示:
x
i
T
i
j
x
j
≤
0
x_i T_{ij}x_j \leq 0
xiTijxj≤0
若
α
=
x
^
⋅
T
⋅
x
^
=
T
:
(
x
^
⨂
x
^
)
=
T
i
j
x
i
x
j
\alpha =\hat x \cdot T \cdot \hat x = T:(\hat x \bigotimes \hat x) = T_{ij}x_ix_j
α=x^⋅T⋅x^=T:(x^⨂x^)=Tijxixj,则
α
\alpha
α关于
x
^
\hat x
x^的导数为:
∂
α
∂
x
k
=
T
i
j
∂
x
i
∂
x
k
x
j
+
T
i
j
x
i
∂
x
j
∂
x
k
=
T
i
j
δ
i
k
x
j
+
T
i
j
x
i
δ
j
k
=
T
k
j
x
j
+
T
i
k
x
j
=
(
T
k
i
+
T
i
k
)
x
i
\frac{\partial \alpha}{\partial x_k} = T_{ij}\frac{\partial x_i}{\partial x_k}x_j+T_{ij}x_i\frac{\partial x_j}{\partial x_k}\\= T_{ij}\delta_{ik}x_j+T_{ij}x_i\delta_{jk} \\ =T_{kj}x_j+T_{ik}x_j \\ =(T_{ki}+T_{ik})x_i
∂xk∂α=Tij∂xk∂xixj+Tijxi∂xk∂xj=Tijδikxj+Tijxiδjk=Tkjxj+Tikxj=(Tki+Tik)xi
所以:
∂
α
∂
x
^
=
2
T
s
y
m
⋅
x
^
⟹
∂
2
α
∂
x
^
⨂
∂
x
^
=
2
T
s
y
m
\frac{\partial \alpha}{\partial \hat x}=2T^{sym}\cdot \hat x \implies \frac{\partial^2\alpha}{\partial \hat x \bigotimes \partial \hat x} = 2T^{sym}
∂x^∂α=2Tsym⋅x^⟹∂x^⨂∂x^∂2α=2Tsym
同样的,以下等式成立:
x
^
⋅
T
⋅
x
^
=
x
^
⋅
T
s
y
m
⋅
x
^
\hat x \cdot T \cdot \hat x = \hat x \cdot T^{sym} \cdot \hat x
x^⋅T⋅x^=x^⋅Tsym⋅x^
因为如果张量的对称部分是正定的,那么该张量也是正定的
NOTE: T的特征值必须为正,T才能为正定的
任意两个张量
S
,
T
≠
0
S, T\neq 0
S,T=0,将
S
S
S表示成张量的加性分解:
S
=
α
T
+
U
,其中
U
=
S
−
α
T
S = \alpha T + U, 其中 U = S - \alpha T
S=αT+U,其中U=S−αT
所以,
S
S
S由
α
\alpha
α决定,有无穷多种分解方式
然而,当
T
r
(
T
⋅
U
T
)
=
T
r
(
U
⋅
T
T
)
=
0
Tr(T \cdot U^T) = Tr(U \cdot T^T)=0
Tr(T⋅UT)=Tr(U⋅TT)=0, 则加性分解是唯一的
S
⋅
T
T
=
α
T
⋅
T
T
+
U
⋅
T
T
⟹
T
r
(
S
⋅
T
T
)
=
α
T
r
(
T
⋅
T
T
)
+
T
r
(
U
⋅
T
T
)
=
α
T
r
(
T
⋅
T
T
)
S\cdot T^T =\alpha T \cdot T^T+U\cdot T^T \\ \implies Tr(S\cdot T^T ) = \alpha Tr(T \cdot T^T) +Tr(U \cdot T^T)=\alpha Tr(T\cdot T^T)
S⋅TT=αT⋅TT+U⋅TT⟹Tr(S⋅TT)=αTr(T⋅TT)+Tr(U⋅TT)=αTr(T⋅TT)
⟹
α
=
T
r
(
S
⋅
T
T
)
T
r
(
T
⋅
T
T
)
\implies\alpha = \frac{Tr(S\cdot T^T ) }{Tr(T\cdot T^T)}
⟹α=Tr(T⋅TT)Tr(S⋅TT)
假设
T
=
1
T = 1
T=1
α
=
T
r
(
S
⋅
1
)
T
r
(
1
⋅
1
)
=
T
r
(
S
)
T
r
(
1
)
=
T
r
(
S
)
3
\alpha = \frac{Tr(S\cdot 1 ) }{Tr(1\cdot 1)} =\frac{Tr(S ) }{Tr(1)} = \frac{Tr(S ) }{3}
α=Tr(1⋅1)Tr(S⋅1)=Tr(1)Tr(S)=3Tr(S)
U
=
S
−
α
T
=
S
−
T
r
(
S
)
3
1
≡
S
d
e
v
U = S - \alpha T=S - \frac{Tr(S ) }{3} 1 \equiv S^{dev}
U=S−αT=S−3Tr(S)1≡Sdev
所以:
S
=
T
r
(
S
)
3
1
+
S
d
e
v
=
S
s
p
h
+
S
d
e
v
S = \frac{Tr(S ) }{3} 1 + S^{dev} =S^{sph} + S^{dev}
S=3Tr(S)1+Sdev=Ssph+Sdev
NOTE:
S
s
p
h
=
T
r
(
S
)
3
1
S^{sph} = \frac{Tr(S ) }{3} 1
Ssph=3Tr(S)1是球形张量;
S
d
e
v
=
S
−
T
r
(
S
)
3
1
S^{dev} = S - \frac{Tr(S ) }{3} 1
Sdev=S−3Tr(S)1是偏张量
如果
T
=
1
2
(
S
+
S
T
)
T = \frac{1}{2}(S + S^T)
T=21(S+ST),则
α
=
=
T
r
(
S
⋅
T
T
)
T
r
(
T
⋅
T
T
)
=
1
2
T
r
(
S
⋅
(
S
+
S
T
)
T
)
1
4
T
r
(
(
S
+
S
T
)
⋅
(
S
+
S
T
)
T
)
=
1
\alpha = \frac{}{} = \frac{Tr(S\cdot T^T ) }{Tr(T\cdot T^T)} = \frac{\frac{1}{2}Tr(S\cdot (S + S^T)^T ) }{\frac{1}{4}Tr((S + S^T)\cdot (S + S^T)^T)}=1
α==Tr(T⋅TT)Tr(S⋅TT)=41Tr((S+ST)⋅(S+ST)T)21Tr(S⋅(S+ST)T)=1
可以定义
U
=
S
−
α
T
=
S
−
T
=
S
−
1
2
(
S
+
S
T
)
=
1
2
(
S
−
S
T
)
U = S - \alpha T=S-T=S-\frac{1}{2}(S + S^T)=\frac{1}{2}(S-S^T)
U=S−αT=S−T=S−21(S+ST)=21(S−ST)
所以可以得到
S
S
S的加性分解为:
S
=
1
2
(
S
+
S
T
)
+
1
2
(
S
−
S
T
)
=
S
s
y
m
+
S
s
k
e
w
S = \frac{1}{2}(S + S^T)+\frac{1}{2}(S - S^T) = S^{sym} + S^{skew}
S=21(S+ST)+21(S−ST)=Ssym+Sskew
其中,用到:
参考教材:
Eduardo W.V. Chaves, Notes On Continuum Mechanics
Copyright © 2003-2013 www.wpsshop.cn 版权所有,并保留所有权利。