Download A Primer on Index Notation and more Study notes Calculus in PDF only on Docsity!
A Primer on Index Notation
John Crimaldi
August 28, 2006
- Index versus Vector Notation
Index notation (a.k.a. Cartesian notation) is a powerful tool for manip- ulating multidimensional equations. However, there are times when the more conventional vector notation is more useful. It is therefore impor- tant to be able to easily convert back and forth between the two. This primer will use both index and vector formulations, and will adhere to the notation conventions summarized below:
Vector Index Notation Notation scalar a a vector ~a ai tensor A Aij
In either notation, we tend to group quantities into one of three categories:
scalar A magnitude that does not change with a rotation of axes. vector Associates a scalar with a direction. tensor Associates a vector (or tensor) with a direction.
- Free Indices
(a) A free index appears once and only once within each additive term in an expression. In the equation below, i is a free index:
ai = ijkbj ck + Dij ej
(b) A free index implies three distinct equations. That is, the free index sequentially assumes the values 1, 2, and 3. Thus,
aj = bj + cj implies
a 1 = b 1 + c 1 a 2 = b 2 + c 2 a 3 = b 3 + c 3
(c) The same letter must be used for the free index in every additive term. The free index may be renamed if and only if it is renamed in every term. (d) Terms in an expression may have more than one free index so long as the indices are distinct. For example the vector-notation expres- sion A = BT^ is written Aij = (Bij )T^ = Bji in index notation. This expression implies nine distinct equations, since i and j are both free indices.
(e) The number of free indices in a term equals the rank of the term:
Notation Rank scalar a 0 vector ai 1 tensor Aij 2 tensor Aijk 3
Technically, a scalar is a tensor with rank 0, and a vector is a tensor of rank 1. Tensors may assume a rank of any integer greater than or equal to zero. You may only sum together terms with equal rank. (f) The first free index in a term corresponds to the row, and the second corresponds to the column. Thus, a vector (which has only one free index) is written as a column of three rows,
~a = ai =
a 1 a 2 a 3
and a rank-2 tensor is written as
A = Aij =
A 11 A 12 A 13
A 21 A 22 A 23
A 31 A 32 A 33
- Dummy Indices
(a) A dummy index appears twice within an additive term of an expres- sion. In the equation below, j and k are both dummy indices:
ai = ijkbj ck + Dij ej
(b) A dummy index implies a summation over the range of the index:
aii ≡ a 11 + a 22 + a 33
- Commutation and Association in Vector and Index Notation
(a) In general, operations in vector notation do not have commutative or associative properties. For example,
~a × ~b 6 = ~b × ~a
(b) All of the terms in index notation are scalars (although the term may represent multiple scalars in multiple equations), and only mul- tiplication/division and addition/subtraction operations are defined. Therefore, commutative and associative properties hold. Thus,
aibj = bj ai
and,
(aibj )ck = ai(bj ck)
A caveat to the commutative property is that calculus operators (discussed later) are not, in general, commutative.
- Vector Operations using Index Notation
(a) Multiplication of a vector by a scalar:
Vector Notation Index Notation a~b = ~c abi = ci
The index i is a free index in this case.
(b) Scalar product of two vectors (a.k.a. dot or inner product):
Vector Notation Index Notation ~a · ~b = c aibi = c The index i is a dummy index in this case. The term “scalar prod- uct” refers to the fact that the result is a scalar.
(c) Scalar product of two tensors (a.k.a. inner or dot product):
Vector Notation Index Notation A : B = c Aij Bji = c
The two dots in the vector notation indicate that both indices are to be summed. Again, the result is a scalar.
(d) Tensor product of two vectors (a.k.a. dyadic product):
Vector Notation Index Notation ~a~b = C aibj = Cij
The term “tensor product” refers to the fact that the result is a ten- sor.
(e) Tensor product of two tensors:
Vector Notation Index Notation A · B = C Aij Bjk = Cik
The single dot refers to the fact that only the inner index is to be summed. Note that this is not an inner product.
(f) Vector product of a tensor and a vector:
Vector Notation Index Notation ~a · B = ~c aiBij = cj
Given a unit vector ˆn, we can form the vector product ˆn · B = ~c. In the language of the definition of a tensor, we say here that then ten- sor B associates the vector ~c with the direction given by the vector ˆn. Also, note that ~a · B 6 = B · ~a.
(g) Cross product of two vectors:
Vector Notation Index Notation ~a × ~b = ~c ijkaj bk = ci Recall that
(a 1 , a 2 , a 3 ) × (b 1 , b 2 , b 3 ) = (a 2 b 3 − a 3 b 2 , a 3 b 1 − a 1 b 3 , a 1 b 2 − a 2 b 1 )
Now, note that the notation ijkaj bk represents three terms, the first
These three equations can be written collectively as ∂φ ∂xi
≡ ∂iφ
In vector notation, ∂iφ is written ∇φ or grad φ. Note that ∂iφ is a vector (rank=1). Some equivalent notations for ∂iφ are
∂iφ ≡ ∂xiφ
and, occasionally, ∂iφ ≡ ∂φ,i
(c) Gradient (spatial derivatives) of a vector field ~a(x 1 , x 2 , x 3 , t): ∂~a ∂x 1
≡ ∂ 1 ai
∂~a ∂x 2
≡ ∂ 2 ai
∂~a ∂x 3
≡ ∂ 3 ai
These three equations can be written collectively as ∂ai ∂xj
≡ ∂j ai
In vector notation, ∂j ai is written ∇~a or grad ~a Note that ∂j ai is a tensor (rank=2):
grad ~a =
∂ai ∂xj = ∂j ai =
∂ 1 a 1 ∂ 1 a 2 ∂ 1 a 3 ∂ 2 a 1 ∂ 2 a 2 ∂ 2 a 3 ∂ 3 a 1 ∂ 3 a 2 ∂ 3 a 3
The index on the denominator of the derivative is the row index. Note that the gradient increases by one the rank of the expression on which it operates.
(d) Divergence of a vector field ~a(x 1 , x 2 , x 3 , t):
div ~a = ∇ · ~a = ∂iai = b Notice that ∂iai is a scalar (rank=0). Important note: The divergence decreases by one the rank of the expression on which it operates by one. It is not possible to take the divergence of a scalar.
(e) Curl of a vector field ~a(x 1 , x 2 , x 3 , t):
curl ~a = ∇ × ~a = ijk∂j ak = bi
Notice that ijk∂j ak is a vector (rank=1 ). Important note: The curl does not change the rank of the expres- sion on which it operates. It is not possible to take the curl of a scalar.
(f) Laplacian of a vector field ~a(x 1 , x 2 , x 3 , t):
∇^2 ~a ≡ ∇ · (∇~a) = div (grad ~a) = ∂i∂iaj = bj
- The ordering of terms in expression involving calculus operators
Index notation is used to represent vector (and tensor) quantities in terms of their constitutive scalar components. For example, ai is the ith^ com- ponent of the vector ~a. Thus, ai is actually a collection of three scalar quantities that collectively represent a vector. Since index notation represents quantities of all ranks in terms of their scalar components, the order in which these terms are written within an expression is usually unimportant. This differs from vector notation, where the order of terms in an expression is often very important. An extremely important caveat to the above discussion on independence of order is to pay special attention to operators (e.g. div, grad, curl). In particular, remember that the rules of calculus (e.g. product rule, chain rule) still apply.
Example 1:
∂ ∂xk
(aibj ) ≡ ∂k (aibj ) = ai∂kbj + bj ∂kai (product rule)
Example 2: Show that ∇ · ~a 6 = ~a · ∇
∇ · ~a = ∂iai =
∂a 1 ∂x 1
∂a 2 ∂x 2
∂a 3 ∂x 3
= a scalar
whereas
~a · ∇ = ai∂i = a 1
∂x 1
() + a 2
∂x 2
() + a 3
∂x 3
() = an operator
Thus, ∂iai 6 = ai∂i
(f) A more general form of the previous relationship can be stated as follows. The expression Aijkl...Bjklm... = 0 is equal to zero if A and B are symmetric and antisymmetric (respectively) with respect to the same indices. For example,
AijklBjklm = 0
if Aijkl = Aikjl and Bjklm = −Bkjlm. (The proof for this is assigned as Problem 4).
Problems
- Write the following vector expressions in Cartesian Index Notation:
(a) ~a ·∇~a (b) (~a ·∇)~a (c) ~a ∇~a (d) A : A (e) A : AT (f) curl grad div(~a × ~b)
- Is the expression “curl [~a · (grad φ)]” valid? Why or why not?
- Show that Q : R = R : Q = 0 if Q is symmetric and R is antisymmetric.
- Show that AijklBjklm = 0 if A is symmetric with respect to indices j and k, and B is antisymemtric with respect to j and k. Note that this is a more general version of problem 3.
- Show that ∇ · ρ~u~u = ρ~u · ∇~u + ρ~u(∇ · ~u )
- Show that ai∂j ai = ∇
2 ~a^ ·^ ~a^
- Show that curl (grad φ)=
- Show that ~a × (~b × ~c ) = (~a · ~c )~b − (~a · ~b )~c
- Show that (~a × ~b ) · (~c × d~ ) = (~a · ~c )(~b · d~ ) − (~a · d~ )(~b · ~c )
- Show that ∇ · (∇ × ~a ) = 0
- Show that ∇ · (~a × ~b ) = ~b · (∇ × ~a ) − ~a · (∇ × ~b )
- Show that ∇ × (~a × ~b ) = ~a(∇ · ~b ) + ~b · ∇~a − ~a · ∇~b − ~b(∇ · ~a )
- Show that if div ~u = 0, div ~v = 0, and curl w~ = 0 then
∇ · [(~u × ~v ) × w~] = w~ · [(~v · ∇ )~u − (~u · ∇ )~v]
