LECTURE 7.
VECTOR-
TENSOR ANALYSIS PART 5
Transport Phenomena
CHEB417-01
2024 Spring Semester / MW 11:00 ~ 12:15
Outline
• Index notation
• Index notation concept
• Summation convention (Einstein convention)
• Preliminary notions
• Vector operations with index notation
• Proofs for the vector identities (example problem)
• Application to fluid mechanics
• Tensor (2nd order tensor)
• Transformation matrix for vector components (using index notation)
Tensor (2nd order tensor)
• We are familiar with the concept of a vector, which has a direction and a
magnitude
• But sometimes, we need to consider two independent directions
simultaneously to describe the situation.
• For example, let us consider the situation where we are interested in the
gradient of a given velocity field as show below. In Cartesian coordinate
system, the velocity field can be described as,
𝐮𝐮 = 𝑢𝑢1 (𝐱𝐱)𝐞𝐞1 + 𝑢𝑢2 (𝐱𝐱)𝐞𝐞2 + 𝑢𝑢3 (𝐱𝐱)𝐞𝐞3 𝐱𝐱 = 𝑥𝑥1 𝐞𝐞1 + 𝑥𝑥2 𝐞𝐞2 + 𝑥𝑥3 𝐞𝐞3
𝐮𝐮 = 𝑢𝑢1 (𝐱𝐱)𝐞𝐞1 + 𝑢𝑢2 (𝐱𝐱)𝐞𝐞2 + 𝑢𝑢3 (𝐱𝐱)𝐞𝐞3
𝐱𝐱 = 𝑥𝑥1 𝐞𝐞1 + 𝑥𝑥2 𝐞𝐞2 + 𝑥𝑥3 𝐞𝐞3
• Now we may be interested in the gradient in the 𝐞𝐞1 –direction. Then we have
𝜕𝜕𝐮𝐮 𝜕𝜕𝑢𝑢1 𝜕𝜕𝑢𝑢2 𝜕𝜕𝑢𝑢3
= 𝐞𝐞1 + 𝐞𝐞2 + 𝐞𝐞
𝜕𝜕𝑥𝑥1 𝜕𝜕𝑥𝑥1 𝜕𝜕𝑥𝑥1 𝜕𝜕𝑥𝑥1 3
• If we are interested in the gradient in the 𝐞𝐞2 –direction, then we have
𝜕𝜕𝐮𝐮 𝜕𝜕𝑢𝑢1 𝜕𝜕𝑢𝑢2 𝜕𝜕𝑢𝑢3
= 𝐞𝐞1 + 𝐞𝐞2 + 𝐞𝐞
𝜕𝜕𝑥𝑥2 𝜕𝜕𝑥𝑥2 𝜕𝜕𝑥𝑥2 𝜕𝜕𝑥𝑥2 3
• In this way, we may find 3 × 3 = 9 independent components to describe the
𝜕𝜕𝑢𝑢𝑗𝑗
situation. If we define as the 𝐞𝐞𝑖𝑖 -directional gradient of 𝐞𝐞𝑗𝑗 –directional
𝜕𝜕𝑥𝑥𝑖𝑖
component of the velocity vector,
𝜕𝜕𝑢𝑢2 the 𝐞𝐞3 -directional gradient of 𝐞𝐞2 –directional
𝜕𝜕𝑥𝑥3 component of the velocity vector
𝜕𝜕𝑢𝑢3 the 𝐞𝐞2 -directional gradient of 𝐞𝐞3 –directional
𝜕𝜕𝑥𝑥2 component of the velocity vector
𝜕𝜕𝑢𝑢𝑗𝑗
• We can also see that the component is for the ordered pair of unit vectors
𝜕𝜕𝑥𝑥𝑖𝑖
𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 if we make a rule that the first unit vector denotes the direction in which
the gradient is taken and the second unit vector the 𝑗𝑗-th component of the
velocity vector.
• Therefore, the ordered pair 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 is a basis to describe the situation. On the
other hand, if we consider the velocity gradient tensor 𝐆𝐆, it is quite natural to
represent as:
3 3 3 3
𝜕𝜕 𝜕𝜕𝑢𝑢𝑗𝑗
𝐆𝐆 = 𝛁𝛁𝐮𝐮 = � 𝐞𝐞 � 𝑢𝑢𝑗𝑗 𝐞𝐞𝑗𝑗 = �� 𝐞𝐞 𝐞𝐞
𝑖𝑖
𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑖𝑖 𝑖𝑖 𝑗𝑗
𝑖𝑖=1 𝑗𝑗=1 𝑖𝑖=1 𝑗𝑗=1
3 3 3 3
𝜕𝜕 𝜕𝜕𝑢𝑢𝑗𝑗
𝐆𝐆 = 𝛁𝛁𝐮𝐮 = � 𝐞𝐞𝑖𝑖 𝜕𝜕𝑥𝑥 � 𝑢𝑢𝑗𝑗 𝐞𝐞𝑗𝑗 = �� 𝐞𝐞 𝐞𝐞
𝑖𝑖 𝜕𝜕𝑥𝑥𝑖𝑖 𝑖𝑖 𝑗𝑗
𝑖𝑖=1 𝑗𝑗=1 𝑖𝑖=1 𝑗𝑗=1
𝜕𝜕𝑢𝑢𝑗𝑗
• Therefore, once again we can regard the component as associated with
𝜕𝜕𝑥𝑥𝑖𝑖
the ordered pair 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 .
• Now let’s define the unit dyads with which second order tensors can be
conveniently represented
(1) The unit dyads 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 (ordered pair)
• The second order tensor can be represented with 3 × 3 unit dyads as
shown below
𝜕𝜕𝑢𝑢𝑗𝑗 the 𝐞𝐞𝑖𝑖 -directional gradient of 𝐞𝐞𝑗𝑗 –directional
𝐞𝐞 𝐞𝐞
𝜕𝜕𝑥𝑥𝑖𝑖 𝑖𝑖 𝑗𝑗 component of the velocity vector
𝐞𝐞𝑖𝑖
𝐞𝐞𝑗𝑗
(Note) The dyad 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 is an ordered pair and it is totally different from 𝐞𝐞𝑖𝑖 � 𝐞𝐞𝑗𝑗
or 𝐞𝐞𝑖𝑖 × 𝐞𝐞𝑗𝑗 .
(2) Representation of a second order tensor (index notation)
3 3
𝐓𝐓 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 → � � 𝑇𝑇𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗
𝑖𝑖=1 𝑗𝑗=1
(i) 𝑇𝑇𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 = 𝑇𝑇𝑗𝑗𝑗𝑗 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 symmetric tensor 𝐓𝐓 = 𝐓𝐓 𝑇𝑇
(ii) 𝑇𝑇𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 = −𝑇𝑇𝑗𝑗𝑗𝑗 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 anti-symmetric (skew-symmetric) tensor 𝐓𝐓 = −𝐓𝐓 𝑇𝑇
(iii) 𝐓𝐓 𝑇𝑇 = 𝑇𝑇𝑗𝑗𝑗𝑗 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 transpose
(iv) 𝐈𝐈 = 𝛿𝛿𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 identity tensor
(3) Dyadic product of two vectors is a 2nd order tensor
3 3
𝐚𝐚𝐚𝐚 = 𝑎𝑎𝑖𝑖 𝐞𝐞𝑖𝑖 𝑏𝑏𝑗𝑗 𝐞𝐞𝑗𝑗 → � 𝑎𝑎𝑖𝑖 𝐞𝐞𝑖𝑖 � 𝑏𝑏𝑗𝑗 𝐞𝐞𝑗𝑗
𝑖𝑖=1 𝑗𝑗=1
3 3
= 𝑎𝑎𝑖𝑖 𝑏𝑏𝑗𝑗 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 → � � 𝑎𝑎𝑖𝑖 𝑏𝑏𝑗𝑗 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗
𝑖𝑖=1 𝑗𝑗=1
e.g.) Gradient of a vector (2nd order tensor)
3 3
𝜕𝜕 𝜕𝜕𝑢𝑢𝑗𝑗 𝜕𝜕𝑢𝑢𝑗𝑗
𝛁𝛁𝐮𝐮 = 𝐞𝐞𝑖𝑖 (𝑢𝑢𝑗𝑗 𝐞𝐞𝑗𝑗 ) = 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 → � � 𝐞𝐞 𝐞𝐞
𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑖𝑖 𝑖𝑖 𝑗𝑗
𝑖𝑖=1 𝑗𝑗=1
(4) Operations of dyads
(i) 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 � 𝐞𝐞𝑘𝑘 𝐞𝐞𝑙𝑙 = 𝐞𝐞𝑖𝑖 𝛿𝛿𝑗𝑗𝑗𝑗 𝐞𝐞𝑙𝑙 = 𝛿𝛿𝑗𝑗𝑗𝑗 𝐞𝐞𝑖𝑖 𝐞𝐞𝑙𝑙
3 3 3
(ii) 𝐓𝐓 � 𝐒𝐒 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 � 𝑆𝑆𝑘𝑘𝑘𝑘 𝐞𝐞𝑘𝑘 𝐞𝐞𝑙𝑙 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝑆𝑆𝑘𝑘𝑘𝑘 𝛿𝛿𝑗𝑗𝑗𝑗 𝐞𝐞𝑖𝑖 𝐞𝐞𝑙𝑙 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝑆𝑆𝑗𝑗𝑙𝑙 𝐞𝐞𝑖𝑖 𝐞𝐞𝑙𝑙 → � � � 𝑇𝑇𝑖𝑖𝑖𝑖 𝑆𝑆𝑗𝑗𝑗𝑗 𝐞𝐞𝑖𝑖 𝐞𝐞𝑙𝑙
𝑖𝑖=1 𝑙𝑙=1 𝑗𝑗=1
3 3 3 3 3
� � � 𝑇𝑇𝑖𝑖𝑖𝑖 𝑆𝑆𝑗𝑗𝑗𝑗 𝐞𝐞𝑖𝑖 𝐞𝐞𝑙𝑙 = � � 𝑇𝑇𝑖𝑖1 𝑆𝑆1𝑙𝑙 + 𝑇𝑇𝑖𝑖2 𝑆𝑆2𝑙𝑙 + 𝑇𝑇𝑖𝑖3 𝑆𝑆3𝑙𝑙 𝐞𝐞𝑖𝑖 𝐞𝐞𝑙𝑙
𝑖𝑖=1 𝑙𝑙=1 𝑗𝑗=1 𝑖𝑖=1 𝑙𝑙=1
𝑇𝑇11 𝑇𝑇12 𝑇𝑇13 𝑆𝑆11 𝑆𝑆12 𝑆𝑆13
𝑇𝑇21 𝑇𝑇22 𝑇𝑇23 𝑆𝑆21 𝑆𝑆22 𝑆𝑆23
𝑇𝑇31 𝑇𝑇32 𝑇𝑇33 𝑆𝑆31 𝑆𝑆32 𝑆𝑆33
𝑖𝑖 = 1, 𝑙𝑙 = 1 𝑇𝑇11 𝑆𝑆11 + 𝑇𝑇12 𝑆𝑆21 + 𝑇𝑇13 𝑆𝑆31 ⋯
= ⋮ ⋱
𝑇𝑇𝑖𝑖1 𝑆𝑆1𝑙𝑙 + 𝑇𝑇𝑖𝑖2 𝑆𝑆2𝑙𝑙 + 𝑇𝑇𝑖𝑖3 𝑆𝑆3𝑙𝑙
e.g.) 𝐈𝐈 � 𝐒𝐒= 𝛿𝛿𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 � 𝑆𝑆𝑘𝑘𝑘𝑘 𝐞𝐞𝑘𝑘 𝐞𝐞𝑙𝑙 = 𝛿𝛿𝑖𝑖𝑖𝑖 𝛿𝛿𝑗𝑗𝑗𝑗 𝑆𝑆𝑘𝑘𝑙𝑙 𝐞𝐞𝑖𝑖 𝐞𝐞𝑙𝑙 = 𝑆𝑆𝑖𝑖𝑙𝑙 𝐞𝐞𝑖𝑖 𝐞𝐞𝑙𝑙 = 𝐒𝐒
(iii) Double dot product 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 ∶ 𝐞𝐞𝑘𝑘 𝐞𝐞𝑙𝑙 = 𝛿𝛿𝑗𝑗𝑗𝑗 𝛿𝛿𝑖𝑖𝑖𝑖
3 3
(iv) 𝐓𝐓 ∶ 𝐒𝐒 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 ∶ 𝑆𝑆𝑘𝑘𝑘𝑘 𝐞𝐞𝑘𝑘 𝐞𝐞𝑙𝑙 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝑆𝑆𝑘𝑘𝑘𝑘 𝛿𝛿𝑗𝑗𝑗𝑗 𝛿𝛿𝑖𝑖𝑖𝑖 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝑆𝑆𝑗𝑗𝑗𝑗 → � � 𝑇𝑇𝑖𝑖𝑖𝑖 𝑆𝑆𝑗𝑗𝑗𝑗
𝑖𝑖=1 𝑗𝑗=1
3
e.g.) 𝐈𝐈 ∶ 𝐒𝐒 = 𝛿𝛿𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 ∶ 𝑆𝑆𝑘𝑘𝑘𝑘 𝐞𝐞𝑘𝑘 𝐞𝐞𝑙𝑙 = 𝛿𝛿𝑖𝑖𝑖𝑖 𝑆𝑆𝑘𝑘𝑘𝑘 𝛿𝛿𝑗𝑗𝑗𝑗 𝛿𝛿𝑖𝑖𝑖𝑖 = 𝑆𝑆𝑖𝑖𝑖𝑖 → � 𝑆𝑆𝑖𝑖𝑖𝑖 = tr (𝐒𝐒)
𝑖𝑖=1
The trace (often abbreviated to “tr”) of
a square matrix 𝐀𝐀 is defined to be the sum
of elements on the main diagonal (from the
upper left to the lower right) of 𝐀𝐀 tr 𝐀𝐀 = 𝑎𝑎11 + 𝑎𝑎22 + 𝑎𝑎33 = −1 + 5 + −5 = −1
https://en.wikipedia.org/
(v) 𝐓𝐓 � 𝐚𝐚 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 � 𝑎𝑎𝑘𝑘 𝐞𝐞𝑘𝑘 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝑎𝑎𝑘𝑘 𝐞𝐞𝑖𝑖 𝛿𝛿𝑗𝑗𝑗𝑗 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝑎𝑎𝑗𝑗 𝐞𝐞𝑖𝑖
e.g.) 𝐈𝐈 � 𝐚𝐚 = 𝛿𝛿𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 � 𝑎𝑎𝑘𝑘 𝐞𝐞𝑘𝑘 = 𝛿𝛿𝑖𝑖𝑖𝑖 𝑎𝑎𝑘𝑘 𝐞𝐞𝑖𝑖 𝛿𝛿𝑗𝑗𝑗𝑗 = 𝑎𝑎𝑖𝑖 𝐞𝐞𝑖𝑖 = 𝐚𝐚
(vi) 𝐓𝐓 × 𝐚𝐚 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 × 𝑎𝑎𝑘𝑘 𝐞𝐞𝑘𝑘 = 𝑇𝑇𝑖𝑖𝑖𝑖 𝑎𝑎𝑘𝑘 𝐞𝐞𝑖𝑖 𝐞𝐞𝑗𝑗 × 𝐞𝐞𝑘𝑘 = 𝜖𝜖𝑗𝑗𝑘𝑘𝑘𝑘 𝑇𝑇𝑖𝑖𝑖𝑖 𝑎𝑎𝑘𝑘 𝐞𝐞𝑖𝑖 𝐞𝐞𝑙𝑙
(5) Divergence of velocity gradient tensor
From previous slides,
𝛁𝛁 � 𝐆𝐆 = 𝛁𝛁 � 𝛁𝛁𝐮𝐮
𝜕𝜕 𝜕𝜕𝑢𝑢𝑗𝑗
𝜕𝜕 𝜕𝜕𝑢𝑢𝑘𝑘 𝐆𝐆 = 𝛁𝛁𝐮𝐮 = 𝐞𝐞𝑖𝑖 (𝑢𝑢𝑗𝑗 𝐞𝐞𝑗𝑗 ) = 𝐞𝐞 𝐞𝐞
= 𝐞𝐞𝑖𝑖 � 𝐞𝐞 𝐞𝐞 𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑖𝑖 𝑖𝑖 𝑗𝑗
𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑗𝑗 𝑗𝑗 𝑘𝑘
𝜕𝜕 2 𝑢𝑢𝑘𝑘 𝜕𝜕 2 𝑢𝑢𝑘𝑘 𝜕𝜕 2 (𝑢𝑢𝑘𝑘 𝐞𝐞𝑘𝑘 ) 2
= 𝛿𝛿𝑖𝑖𝑖𝑖 𝐞𝐞 = 𝐞𝐞𝑘𝑘 = = 𝛁𝛁 𝐮𝐮
𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑗𝑗 𝑘𝑘 𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑖𝑖 2
𝑇𝑇 𝜕𝜕 𝜕𝜕𝑢𝑢𝑗𝑗
𝛁𝛁 � 𝛁𝛁𝐮𝐮 = 𝐞𝐞𝑖𝑖 � 𝐞𝐞 𝐞𝐞
𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑘𝑘 𝑗𝑗 𝑘𝑘
𝜕𝜕 2 𝑢𝑢𝑗𝑗 𝜕𝜕 2 𝑢𝑢𝑖𝑖 𝜕𝜕 𝜕𝜕𝑢𝑢𝑖𝑖
= 𝛿𝛿𝑖𝑖𝑖𝑖 𝐞𝐞𝑘𝑘 = 𝐞𝐞𝑘𝑘 = 𝐞𝐞𝑘𝑘 = 𝛁𝛁 𝛁𝛁 � 𝐮𝐮
𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑘𝑘 𝜕𝜕𝑥𝑥𝑖𝑖 𝜕𝜕𝑥𝑥𝑘𝑘 𝜕𝜕𝑥𝑥𝑘𝑘 𝜕𝜕𝑥𝑥𝑖𝑖
Transformation matrix for vector components
• Let’s revisit the transformation rule for the components of a vector since
we now are familiar with the index notation:
• Consider a vector 𝐯𝐯. To express the vector 𝐯𝐯, we introduce
the 𝐞𝐞-, 𝐞𝐞� -coordinate systems as shown in the figure below
In this case, 𝐯𝐯 = 𝑣𝑣1 𝐞𝐞1 + 𝑣𝑣2 𝐞𝐞2 + 𝑣𝑣3 𝐞𝐞3 = 𝑣𝑣𝑖𝑖 𝐞𝐞𝑖𝑖 → � 𝑣𝑣𝑖𝑖 𝐞𝐞𝑖𝑖
𝑖𝑖=1
3
= 𝑣𝑣�1 𝐞𝐞� 1 + 𝑣𝑣�2 𝐞𝐞� 2 + 𝑣𝑣�3 𝐞𝐞� 3 = 𝑣𝑣�𝛼𝛼 𝐞𝐞� 𝛼𝛼 → � 𝑣𝑣�𝛼𝛼 𝐞𝐞� 𝛼𝛼
𝛼𝛼=1
• Here we should note that the vector itself 𝐯𝐯 is not changed but the
components of the vector are changed depending on the coordinate system.
• In linear algebra, it is customary to call the column vectors that consists of the
components, as the component vectors.
𝑣𝑣�1 𝑣𝑣1
𝑣𝑣� = 𝑣𝑣�2 𝑣𝑣 = 𝑣𝑣2
𝑣𝑣�3 𝑣𝑣3
• The vectors for the components (in some sense, component vectors) in linear
algebra are denoted by italic fonts (e.g. 𝑣𝑣,𝑣𝑣,
� etc)
• Let’s determine the relationships between the component vectors, 𝑣𝑣 and 𝑣𝑣,
� for
the same vector 𝐯𝐯.
This goal may be achieved by the relation,
3
𝑣𝑣�𝛼𝛼 = 𝐞𝐞� 𝛼𝛼 � 𝐯𝐯 = 𝐞𝐞� 𝛼𝛼 � (𝑣𝑣𝑖𝑖 𝐞𝐞𝑖𝑖 ) = (�𝐞𝐞𝛼𝛼 � 𝐞𝐞𝑖𝑖 )𝑣𝑣𝑖𝑖 = 𝑙𝑙𝛼𝛼𝑖𝑖 𝑣𝑣𝑖𝑖 → � 𝑙𝑙𝛼𝛼𝑖𝑖 𝑣𝑣𝑖𝑖
𝑖𝑖=1
Thus,
𝑣𝑣� = 𝐿𝐿𝐿𝐿 ,where 𝑙𝑙𝛼𝛼𝑖𝑖 = 𝐿𝐿 𝛼𝛼𝑖𝑖
• On the other hand, we can also express the components in 𝐞𝐞-coordinate
system in terms of the 𝐞𝐞� –coordinate system.
3
𝑣𝑣𝑖𝑖 = 𝐞𝐞𝑖𝑖 � 𝐯𝐯 = 𝐞𝐞𝑖𝑖 � (𝑣𝑣�𝛼𝛼 𝐞𝐞� 𝛼𝛼 ) = 𝑣𝑣�𝛼𝛼 𝐞𝐞� 𝛼𝛼 � 𝐞𝐞𝑖𝑖 = 𝑣𝑣�𝛼𝛼 𝑙𝑙𝛼𝛼𝑖𝑖 = 𝑙𝑙𝛼𝛼𝑖𝑖 𝑣𝑣�𝛼𝛼 = 𝐿𝐿𝑇𝑇 𝑖𝑖𝛼𝛼 𝑣𝑣
�𝛼𝛼 → � 𝐿𝐿𝑇𝑇 𝑖𝑖𝛼𝛼 𝑣𝑣
�𝛼𝛼
𝛼𝛼=1
In the previous slide, we defined 𝑙𝑙𝛼𝛼𝑖𝑖 as, 𝐞𝐞� 𝛼𝛼 � 𝐞𝐞𝑖𝑖 = 𝑙𝑙𝛼𝛼𝑖𝑖
𝑣𝑣 = 𝐿𝐿𝑇𝑇 𝑣𝑣�
In the previous slide, since 𝑣𝑣� = 𝐿𝐿𝐿𝐿, we can show that
𝑣𝑣� = 𝐿𝐿𝐿𝐿 = 𝐿𝐿𝐿𝐿𝑇𝑇 𝑣𝑣� 𝐿𝐿𝐿𝐿𝑇𝑇 = 𝐼𝐼
,where 𝐼𝐼 is the identity matrix
• Two kinds of coordinate change for the Cartesian coordinate system:
(1) The rotation operation about a certain rotation axis
det 𝐿𝐿 = 1
(2) The reflection operation about a certain plane
det 𝐿𝐿 = −1
For your interest, the determinant of a 3 × 3 matrix 𝐴𝐴 is,
e.g.)
www.chilimath.com
Example problem
• Let us consider two coordinate systems, where 𝐞𝐞
� –coordinate system is
obtained by rotating 𝐞𝐞–coordinate system 120º about the rotation axis
𝐞𝐞3 as shown below:
Then the transformation matrix 𝐿𝐿 is obtained as
𝑣𝑣� = 𝐿𝐿𝐿𝐿 ,where 𝑙𝑙𝛼𝛼𝑖𝑖 = 𝐿𝐿 𝛼𝛼𝑖𝑖
𝑙𝑙11 = 𝐞𝐞� 1 � 𝐞𝐞1 = cos 120° = −1/2
𝑙𝑙12 = 𝐞𝐞� 1 � 𝐞𝐞2 = cos 30° = 3/2
𝑙𝑙21 = 𝐞𝐞� 2 � 𝐞𝐞1 = cos 210° = − 3/2
𝑙𝑙22 = 𝐞𝐞� 2 � 𝐞𝐞2 = cos 120° = −1/2
In this way, we prepare,
−1/2 3/2 0
𝐿𝐿 = − 3/2 −1/2 0 1 1 3 3
det 𝐿𝐿 = − − − − =1
0 0 1 2 2 2 2
Homework assignment 2
• Due Mar 25th before class
