2

Coordinate Transformations

2.1 Introduction
Partial dierential equations in the physical domain X n can be solved on a
structured numerical grid obtained by mapping a reference grid in the logical
region n into X n with a coordinate transformation x() : n X n . The
structured grid concept also gives an alternative way to obtain a numerical
solution to a partial dierential equation, by solving the transformed equation
with respect to the new independent variables i on the reference grid in the
logical domain n . Some notions and relations concerning the coordinate
transformations yielding structured grids are discussed in this chapter. These
notions and relations are used to represent some conservation-law equations
in the new logical coordinates in a convenient form. The relations presented
will be used in Chap. 3 to formulate various grid properties.
Conservation-law equations in curvilinear coordinates are typically de-
duced from the equations in Cartesian coordinates through the classical for-
mulas of tensor calculus, by procedures which include the substitution of ten-
sor derivatives for ordinary derivatives. The formulation and evaluation of
the tensor derivatives is rather dicult, and they retain some elements of
mystery. However, these derivatives are based on specic transformations of
tensors, modeling in the equations some dependent variables, e.g. the compo-
nents of a uid velocity vector, which after the transformation have a clear
interpretation in terms of the contravariant components of the vector. With
this concept, the conservation-law equations are readily written out in this
chapter without application of the tensor derivatives, but utilizing instend
only some specic transformations of the dependent variables, ordinary deriv-
atives, and one basic identity of coordinate transformations derived from the
formula for dierentiation of the Jacobian.
For generality, the transformations of the coordinates are mainly consid-
ered for arbitrary n-dimensional domains, though in practical applications the
dimension n equals 1, 2, 3, or 4 for time-dependent transformations of three-
dimensional domains. We also apply chiey a standard vector notation for

V.D. Liseikin, Grid Generation Methods, Scientic Computation,

DOI 10.1007/978-90-481-2912-6 2, 
c Springer Science+Business Media B.V. 2010
32 2 Coordinate Transformations

the coordinates, as variables with indices. Sometimes, however, particularly

in gures, the ordinary designation for three-dimensional coordinates, namely
x, y, z for the physical coordinates and , , for the logical ones, is used to
simplify the presentation.

2.2 General Notions and Relations

This section presents some basic relations between Cartesian and curvilinear
coordinates.

2.2.1 Jacobi Matrix

Let
x() : n X n , = ( 1 , . . . , n ), x = (x1 , . . . , xn ),
be a smooth invertible coordinate transformation of the physical region X n
Rn from the parametric domain n Rn . If n is a standard logical domain,
then, in accordance with Chap. 1, this coordinate transformation can be used
to generate a structured grid in X n . Here and later Rn presents the Euclidean
space with the Cartesian basis e1 , . . . , en , which represents an orthogonal
system of vectors, i.e. 
1 if i = j,
ei ej =
0 if i = j.
Thus we have

x = x1 e1 + + xn en ,
= 1 e1 + + n en .

The values xi , i = 1, . . . , n, are called the Cartesian coordinates of the vec-

tor x. The coordinate transformation x() denes, in the domain X n , new
coordinates 1 , . . . , n , which are called the curvilinear coordinates. The ma-
trix  i
x
j= , i, j = 1, . . . , n,
j
is referred to as the Jacobi matrix, and its Jacobian is designated by J:
 i
x
J = det , i, j = 1, . . . , n. (2.1)
j

The inverse transformation to the coordinate mapping x() is denoted by

(x) : X n n .
 2.2 General Notions and Relations 33

This transformation can be considered analogously as a mapping introducing

a curvilinear coordinate system x1 , . . . , xn in the domain n Rn . It is
obvious that the inverse to the matrix j is
 i
1
j = , i, j = 1, . . . , n,
xj
and consequently
 
i 1
det = , i, j = 1, . . . , n. (2.2)
xj J

In the case of two-dimensional space the elements of the matrices (xi / j )

and ( i /xj ) are connected by

i x3j 
j
= (1)i+j 3i J,
x
(2.3)
xi 3j
j
= (1)i+j J 3i , i, j = 1, 2.
x
Similar relations between the elements of the corresponding three-dimensional
matrices have the form
 
i 1 xj+1 xj+2 xj+1 xj+2
= ,
xj J i+1 i+2 i+2 i+1
  (2.4)
xi j+1 j+2 j+1 j+2
=J , i, j = 1, 2, 3,
j xi+1 xi+2 xi+2 xi+1
where for each superscript or subscript index, say l, l + 3 is equivalent to l.
With this condition the sequence of indices (l, l + 1, l + 2) is the result of a
cyclic permutation of (1, 2, 3) and vice versa; the indices of a cyclic sequence
(i, j, k) satisfy the relation j = i + 1, k = i + 2.

2.2.2 Tangential Vectors

The value of the function x() = [x1 (), . . . , xn ()] in the Cartesian basis
(e1 , . . . , en ), i.e.
x() = x1 ()e1 + + xn ()en ,
is a position vector for every n . This vector-valued function x() gen-
erates the nodes, edges, faces, etc. of the cells of the coordinate grid in the
domain X n . Each edge of the cell corresponds to a coordinate line i for some
i and is dened by the vector

i x = x( + hei ) x(),

where h is the step size of the uniform grid in the i direction in the logical
domain n . We have
34 2 Coordinate Transformations

Fig. 2.1. Grid cell and contracted parallelogram

i x = hxi + t,
where  
x1 xn
x i = , ,
i i
is the vector tangential to the coordinate curve i , and t is a residual vector
whose length does not exceed the following quantity:
1
max|xi i |h2 .
2
Thus the cells in the domain X n whose edges are formed by the vectors
hxi , i = 1, . . . , n, are approximately the same as those obtained by map-
ping the uniform coordinate cells in the computational domain n with the
transformation x(). Consequently, the uniformly contracted parallelepiped
spanned by the tangential vectors xi , i = 1, . . . , n, represents to a high order
of accuracy with respect to h the cell of the coordinate grid at the correspond-
ing point in X n (see Fig. 2.1). In particular, for the length li of the ith grid
edge we have
li = h|xi | + O(h2 ).
The volume Vh (area in two dimensions) of the cell is expressed as follows:
Vh = hn V + O(hn+1 ),
where V is the volume of the n-dimensional parallelepiped determined by the
tangential vectors xi , i = 1, . . . , n.
The tangential vectors xi , i = 1, . . . , n, are called the base covariant
vectors since they comprise a vector basis. The sequence x1 , . . . , xn of the
tangential vectors has a right-handed orientation if the Jacobian of the trans-
formation x() is positive. Otherwise, the base vectors xi have a left-handed
orientation.
The operation of the dot product on these vectors produces elements of
the covariant metric tensor. These elements generate the coecients that ap-
pear in the transformed governing equations that model the conservation-law
equations of mechanics. Besides this, the metric elements play a primary role
in studying and formulating various geometric characteristics of the grid cells.
 2.2 General Notions and Relations 35

Fig. 2.2. Disposition of the base tangential and normal vectors in two dimensions

2.2.3 Normal Vectors

For a xed i the vector  

i i
, , ,
x1 xn
which is the gradient of i (x) with respect to the Cartesian coordinates
x1 , . . . , xn , is denoted by i . The set of the vectors i , i = 1, . . . , n, is
called the set of base contravariant vectors.
Similarly, as the tangential vectors relate to the coordinate curves, the
contravariant vectors i , i = 1, . . . , n, are connected with their respective
coordinate surfaces (curves in two dimensions). A coordinate surface is de-
ned by the equation i = 0i ; i.e. along the surface all of the coordinates
1 , . . . , n except i are allowed to vary. For all of the tangent vectors xj to
the coordinate lines on the surface i = 0i we have the obvious identity
xj i = 0, i = j,
and thus the vector i is a normal to the coordinate surface i = 0i . There-
fore the vectors i , i = 1, . . . , n, are also called the normal base vectors.
Since
xi i = 1
for each xed i = 1, . . . , n, the vectors xi and i intersect each other at
an angle which is less than /2. Now, taking into account the orthogonality
of the vector i to the surface i = 0i , we nd that these two vectors xi
and i are directed to the same side of the coordinate surface (curve in two
dimensions). An illustration of this fact in two dimensions is given in Fig. 2.2.
The length of any normal base vector i is linked to the distance di
between the corresponding opposite boundary segments (joined by the vector
xi ) of the n-dimensional parallelepiped formed by the base tangential vectors,
namely, 
di = 1/| i |, | i | = i i .
To prove this relation we recall that the vector i is a normal to all of the
vectors xj , j = i, and therefore to the boundary segments formed by these
36 2 Coordinate Transformations

n 1 vectors. Hence, the unit normal vector ni to these segments is expressed

by
ni = i /| i |.
Now, taking into account that

di = x i n i ,

we readily obtain
di = xi i /| i | = 1/| i |.
Let li denote the distance between a grid point on the coordinate surface
i = c and the nearest point on the neighboring coordinate surface i = c + h;
then
li = hdi + O(h2 ) = h/| i | + O(h2 ).
This equation shows that the inverse length of the normal vector i mul-
tiplied by h represents with high accuracy the distance between the corre-
sponding faces of the coordinate cells in the domain X n .
Note that the volume of the parallelepiped spanned by the tangential vec-
tors equals J, so we nd from (2.2) that the volume of the n-dimensional par-
allelepiped dened by the normal vectors i , i = 1, . . . , n, is equal to 1/J.
Thus both the base normal vectors i and the base tangential vectors xi
have the same right-handed or left-handed orientation.
If the coordinate system 1 , . . . , n is orthogonal, i.e.

xi xj = P (x)ji , P (x) > 0, i, j = 1, . . . , n,

then for each xed i = 1, . . . , n the vector i is parallel to xi . Here and

later, ji is the Kronecker symbol, i.e.

ji = 0 if i = j, ji = 1 if i = j.

2.2.4 Representation of Vectors Through the Base Vectors

If there are n independent base vectors a1 , . . . , an of the Euclidean space Rn

then any vector b with components b1 , . . . , bn in the Cartesian basis e1 , . . . , en
is represented through the vectors ai , i = 1, . . . , n, by

b = aij (b aj )ai , i, j = 1, . . . , n, (2.5)

where aij are the elements of the matrix (aij ) which is the inverse of the
tensor (aij ), aij = ai aj , i, j = 1, . . . , n. It is assumed in (2.5) and later that
a summation is carried out over repeated indices unless otherwise noted.
The components of the vector b in the natural basis of the tangential
vectors xi , i = 1, . . . , n, are called contravariant. Let them be denoted by
i
b , i = 1, . . . , n. Thus
 2.2 General Notions and Relations 37
1 n
b = b x 1 + + b x n .
Assuming in (2.5) ai = xi , i = 1, . . . , n, we obtain
 k
 i
i mj k x x
b =a b , i, j, k, m = 1, . . . , n, (2.6)
j m

where b1 , . . . , bn are the components of the vector b in the Cartesian basis

e1 , . . . , en . Since
xk xk
aij = , i, j, k = 1, . . . , n,
i j
we have
i j
aij = , k = 1, . . . , n.
xk xk
Therefore, from (2.6),

i i
b = bj , i, j = 1, . . . , n, (2.7)
xj
or, using the dot product notation
i
b = b i , i = 1, . . . , n. (2.8)

Thus, in this case (2.5) has the form

b = (b i )xi , i = 1, . . . , n. (2.9)

For example, the normal base vector i is expanded through the base
tangential vectors xj , j = 1, . . . , n, by the following formula:

i k
i = x k, i, j, k, = 1, . . . , n. (2.10)
xj xj
Analogously, a component bi of the vector b in the basis i , i = 1, . . . , n, is
expressed by the formula
xj
bi = bj = b x i , i = 1, . . . , n, (2.11)
i
and consequently

b = bi i = (b xi ) i , i = 1, . . . , n. (2.12)

These components bi , i = 1, . . . , n, of the vector b are called covariant.

In particular, the base tangential vector xi is expressed through the base
normal vectors j , j = 1, . . . , n, as follows:
xj xj
x i = k , i, j, k = 1, . . . , n. (2.13)
i k
38 2 Coordinate Transformations

2.2.5 Metric Tensors

Many grid generation algorithms, in particular those based on the calculus

of variations, are typically formulated in terms of fundamental features of
coordinate transformations and the corresponding mesh cells. These features
are compactly described with the use of the metric notation, which is discussed
in this subsection.

2.2.5.1 Covariant Metric Tensor

The matrix
(gij ), i, j = 1, . . . , n,
whose elements gij are the dot products of the pairs of the basic tangential
vectors xi ,

xk xk
gij = xi xj = , i, j, k = 1, . . . , n, (2.14)
i j
is called a covariant metric tensor of the domain X n in the coordinates
1 , . . . , n . Geometrically, each diagonal element gii of the matrix (gij ) is the
length of the tangent vector xi squared:

gii = |xi |2 , i = 1, . . . , n.

Also,

gij = |xi ||xj |cos = gii gjj cos , (2.15)
where is the angle between the tangent vectors xi and xj . In these expres-
sions for gii and gij the subscripts ii and jj are xed, i.e. here the summation
over the repeated indices is not carried out.
The matrix (gij ) is called the metric tensor because it denes distance
measurements with respect to the coordinates 1 , . . . , n :

ds = gij d i d j , i, j = 1, . . . , n.

Thus the length s of the curve in X n prescribed by the parametrization

x[(t)] : [a, b] X n

is computed by the formula


b
d i d j
s= gij dt.
a dt dt

We designate by g the Jacobian of the covariant matrix (gij ). It is evident

that
(gij ) = jjT ,
 2.2 General Notions and Relations 39

and hence
J 2 = g.
The covariant metric tensor is a symmetric matrix, i.e. gij = gji . If a
coordinate system at a point is orthogonal then the tensor (gij ) has a simple
diagonal form at this point. Note that these advantageous properties are in
general not possessed by the Jacobi matrix (xi / i ) from which the covariant
metric tensor (gij ) is dened.

2.2.5.2 Contravariant Metric Tensor

The contravariant metric tensor of the domain X n in the coordinates 1 , . . . , n

is the matrix
(g ij ), i, j = 1, . . . , n,
inverse to (gij ), i.e.

gij g jk = ik , i, j, k = 1, . . . , n. (2.16)

Therefore
1
det(g ij ) = .
g
It is easily shown that (2.16) is satised if and only if

i j
g ij = i j = , i, j, k = 1, . . . , n. (2.17)
xk xk
Thus, each diagonal element g ii (where i is xed) of the matrix (g ij ) is the
square of the length of the vector i :

g ii = | i |2 . (2.18)

Geometric Interpretation

Now we discuss the geometric meaning of a xed diagonal element g ii , say

g 11 , of the matrix (g ij ). Let us consider a three-dimensional coordinate trans-
formation x() : 3 X 3 . Its tangential vectors x1 , x2 , x3 represent
geometrically the edges of the parallelepiped formed by these vectors. For the
distance d1 between the opposite faces of the parallelepiped which are dened
by the vectors x2 and x3 , we have

d1 = x1 n1 ,

where n1 is the unit normal to the plane spanned by the vectors x2 and x3 .
It is clear, that
1 xj = 0, j = 2, 3,
and hence the unit normal n1 is parallel to the normal base vector 1 . Thus
we obtain
40 2 Coordinate Transformations

n1 = 1 /|1 | = 1 / g 11 .
Therefore  
d1 = 1 1 / g 11 = 1/ g 11 ,
and consequently
g 11 = 1/(d1 )2 .
Analogous relations are valid for g 22 and g 33 , i.e. in three dimensions the
diagonal element g ii for a xed i means the inverse square of the distance di
between those faces of the parallelepiped which are connected by the vector
xi . In two-dimensional space the element g ii (where i is xed) is the inverse
square of the distance between the edges of the parallelogram dened by the
tangential vectors x1 and x2 .
The same interpretation of g ii is valid for general multidimensional coor-
dinate transformations:

g ii = 1/(di )2 , i = 1, . . . , n, (2.19)

where the index i is xed, and di is the distance between those faces of the
n-dimensional parallelepiped which are linked by the tangential vector xi .

2.2.5.3 Relations Between Covariant and Contravariant Elements

Now, in analogy with (2.3) and (2.4), we write out very convenient formulas
for natural relations between the contravariant elements g ij and the covariant
ones gij in two and three dimensions.
For n = 2,
g3i 3j
g ij = (1)i+j ,
g (2.20)
gij = (1)i+j gg 3i 3j , i, j = 1, 2,

where the indices i, j on the right-hand side of the relations (2.20) are xed,
i.e. summation over the repeated indices is not carried out here. For n = 3 we
have
1
g ij = (gi+1 j+1 gi+2 j+2 gi+1 j+2 gi+2 j+1 ),
g (2.21)
gij = g(g i+1 j+1 g i+2 j+2 g i+1 j+2 g i+2 j+1 ), i, j = 1, 2, 3,
with the convention that any index, say l, is identied with l 3, so, for
instance, g45 = g12 .
We also note that, in accordance with the expressions (2.14), (2.17) for gij
and g ij , respectively, the relations (2.10) and (2.13) between the basic vectors
xi and j can be written in the form

xi = gik k ,
(2.22)
i = g ik xk , i, k = 1, . . . , n.
 2.2 General Notions and Relations 41

So the rst derivatives xi / j and k /xm of the transformations x()

and (x), respectively, are connected through the metric elements:

xi m
= g mj ,
j xi
(2.23)
i mi x
j
= g , i, j, m = 1, . . . , n.
xj m

2.2.6 Cross Product

In addition to the dot product there is another important operation on three-

dimensional vectors. This is the cross product, , which for any two vectors
a = (a1 , a2 , a3 ), b = (b1 , b2 , b3 ) is expressed as the determinant of a matrix:

e1 e2 e3
a b = det a1 a2 a3 , (2.24)
b1 b2 b3

where (e1 , e2 , e3 ) is the Cartesian vector basis of the Euclidean space R3 .

Thus
a b = (a2 b3 a3 b2 , a3 b1 a1 b3 , a1 b2 a2 b1 ),
or, with the previously mentioned convention in three dimensions of the iden-
tication of any index j with j 3,

a b = (ai+1 bi+2 ai+2 bi+1 )ei , i = 1, 2, 3. (2.25)

We will now state some facts connected with the cross product operation.

2.2.6.1 Geometric Meaning

We can readily see that a b = 0 if the vectors a and b are parallel. Also,
from (2.25) we nd that a (a b) = 0 and b (a b) = 0, i.e. the vector
a b is orthogonal to each of the vectors a and b. Thus, if these vectors are
not parallel then
a b = |a b|n, (2.26)
where = 1 or = 1 and n is a unit normal vector to the plane determined
by the vectors a and b.
Now we show that the length of the vector a b equals the area of the
parallelogram formed by the vectors a and b, i.e.

|a b| = |a||b| sin , (2.27)

where is the angle between the two vectors a and b. To prove (2.27) we rst
note that
42 2 Coordinate Transformations

|a|2 |b|2 sin2 = |a|2 |b|2 (1 cos2 ) = |a|2 |b|2 (a b)2 .

We have, further,
    2

3 
3 
3
|a|2 |b|2 |a b|2 = ai ai bj bj a k bk
i=1 j=1 k=1


3
= [(al )2 (bm )2 + (am )2 (bl )2 2al bl am bm ]
k=1

3
= (al bm am bl )2 ,
k=1

where (k, l, m) are cyclic, i.e. l = k + 1, m = k + 2 with the convention

that j + 3 is equivalent to j for any index j. According to (2.25) the quantity
al bm am bl for the cyclic sequence (k, l, m) is the kth component of the vector
a b, so we nd that

|a||b| sin2 = |a|2 |b|2 |a b|2 = |a b|2 , (2.28)

what proves (2.27). Thus we obtain the result that if the vectors a and b are
not parallel then the vector a b is orthogonal to the parallelogram formed
by these vectors and its length equals the area of the parallelogram. Therefore
the three vectors a, b and a b are independent in this case and represent a
base vector system in the three-dimensional space R3 . Moreover, the vectors
a, b and a b form a right-handed triad since a b = 0, and consequently
the Jacobian of the matrix determined by a, b, and a b is positive; it equals

a b a b = (a b)2 .

2.2.6.2 Relation to Volumes

Let c = (c1 , c2 , c3 ) be one more vector. The volume V of the parallelepiped

whose edges are the vectors a, b and c equals the area of the parallelogram
formed by the vectors a and b multiplied by the modulas of the dot product
of the vector c and the unit normal n to the parallelogram. Thus

V = |a b||n c|

and from (2.26) we obtain

V = |(a b) c|. (2.29)

Taking into account (2.25), we obtain

(a b) c = c1 (a2 b3 a3 b2 ) + c2 (a3 b1 a1 b3 ) + c3 (a1 b2 a2 b1 ).

 2.2 General Notions and Relations 43

The right-hand side of this equation is the Jacobian of the matrix whose rows
are formed by the vectors a, b, and c, i.e.
1
a a2 a3
(a b) c = det b1 b2 b3 . (2.30)
c1 c2 c3

From this equation we readily obtain

(a b) c = a (b c) = (c a) b.

Thus the volume of the parallelepiped determined by the vectors a, b, and c

equals the Jacobian of the matrix formed by the components of these vectors.
In particular, we obtain from (2.1) that the Jacobian of a three-dimensional
coordinate transformation x() is expressed as follows:

J = x1 (x2 x3 ). (2.31)

2.2.6.3 Relation to Base Vectors

Applying the operation of the cross product to two base tangential vectors xl
and xm , we nd that the vector xl xm is a normal to the coordinate surface
i = 0i with (i, l, m) cyclic. The base normal vector i is also orthogonal to
the surface and therefore it is a scalar multiple of xl xm , i.e.

i = c(xl xm ).

Multiplying this equation for a xed i by xi , using the operation of the dot
product, we obtain, using (2.31),

1 = cJ,

and therefore
1
(x l xm ).
i = (2.32)
J
Thus the elements of the three-dimensional contravariant metric tensor (g ij )
are computed through the tangential vectors xi by the formula

1
g ij = (x i+1 xi+2 ) (xj+1 xj+2 ), i, j = 1, 2, 3.
g
Analogously, every base vector xi , i = 1, 2, 3, is expressed by the tensor
product of the vectors j , j = 1, 2, 3:

xi = J( l k ), i = 1, 2, 3, (2.33)

where l = i + 1, k = i + 2, and m is equivalent to m + 3 for any index m.

Accordingly, we have
44 2 Coordinate Transformations

gij = g( i+1 i+2 ) ( j+1 j+2 ), i, j = 1, 2, 3.

Using the relations (2.32) and (2.33) in (2.31), we also obtain

1
= 1 2 3 . (2.34)
J
Thus the volume of the parallelepiped formed by the base normal vectors
1 , 2 , and 3 is the modulus of the inverse of the Jacobian J of the
transformation x().

2.3 Relations Concerning Second Derivatives

The elements of the covariant and contravariant metric tensors are dened
by the dot products of the base tangential and normal vectors, respectively.
These elements are suitable for describing the internal features of the cells
such as the lengths of the edges, the areas of the faces, their volumes, and
the angles between the edges and the faces. However, as they are derived
from the rst derivatives of the coordinate transformation x(), the direct
use of the metric elements is not sucient for the description of the dynamic
features of the grid (e.g. curvature), which reect changes between adjacent
cells. This is because the formulation of these grid features relies not only
on the rst derivatives but also on the second derivatives of x(). Therefore
there is a need to study relations connected with the second derivatives of the
coordinate parametrizations.
This section presents some notations and formulas which are concerned
with the second derivatives of the components of the coordinate transforma-
tions. These notations and relations will be used to describe the curvature
and eccentricity of the coordinate lines and to formulate some equations of
mechanics in new independent variables.

2.3.1 Christoel Symbols

The edge of a grid cell in the i direction can be represented with high accuracy
by the base vector xi contracted by the factor h, which represents the step
size of a uniform grid in n . Therefore the local change of the edge in the j
direction is characterized by the derivative of xi with respect to j , i.e. by
x i j .
Since the second derivatives may be used to formulate quantitative mea-
sures of the grid, we describe these vectors xi j through the base tangential
and normal vectors using certain three-index quantities known as Christoel
symbols. The Christoel symbols are commonly used in formulating measures
of the mutual interaction of the cells and in formulas for dierential equations.
Let us denote by ijk the kth contravariant component of the vector xi j
in the base tangential vectors xk , k = 1, . . . , n. The superscript k in this
 2.3 Relations Concerning Second Derivatives 45

designation relates to the base vector xk and the subscript ij corresponds to

the mixed derivative with respect to i and j . Thus

xi j = ijk xk , i, j, k = 1, . . . , n, (2.35)

and consequently

2 xp m x
p

j k
= kj m
, j, k, m, p = 1, . . . , n. (2.36)

In accordance with (2.7), we have

i 2 xl i
kj = , i, j, k, l = 1, . . . , n, (2.37)
k j xl
or in vector form,
i
kj = xk j i . (2.38)
i p
Equation (2.37) is also obtained by multiplying (2.36) by /x and sum-
ming over p.
i
The quantities kj are called the space Christoel symbols of the second
kind and the expression (2.35) is a form of the Gauss relation representing
the second derivatives of the position vector x() through the tangential vec-
tors xi .
Analogously, the components of the second derivatives of the position vec-
tor x() expanded in the base normal vectors i , i = 1, . . . , n, are referred
to as the space Christoel symbols of the rst kind. The mth component of
the vector xk j in the base vectors i , i = 1, . . . , n, is denoted by [kj, m].
Thus, according to (2.11),

2 xl xl
[kj, m] = xk j xm = , j, k, l, m = 1, . . . , n, (2.39)
k j m
and consequently
xk j = [kj, m] m . (2.40)
So, in analogy with (2.36), we obtain

2 xl m
j k
= [kj, m] i , i, j, k, m = 1, . . . , n. (2.41)
x

Multiplying (2.39) by g im and summing over m we nd that the space Christof-

fel symbols of the rst and second kind are connected by the following relation:
i
kj = g im [kj, m], i, j, k, m = 1, . . . , n. (2.42)

Conversely, from (2.37),

l
[kj, m] = gml kj , j, k, l, m = 1, . . . , n. (2.43)
46 2 Coordinate Transformations

The space Christoel symbols of the rst kind [kj, m] can be expressed through
the rst derivatives of the covariant elements gij of the metric tensor (gij ) by
the following readily veried formula:
 
1 gjm gkm gkj
[kj, m] = + m , i, j, k, m = 1, . . . , n. (2.44)
2 k j
Thus, taking into account (2.42), we see that the space Christoel symbols of
i
the second kind kj can be written in terms of metric elements and their rst
derivatives. In particular, in the case of an orthogonal coordinate system i ,
we obtain from (2.42), (2.44)
 
1 ii gii gii gkj
i
kj = g + .
g k j i
Here the index i is xed, i.e. the summation over i is not carried out.

2.3.2 Dierentiation of the Jacobian

Of critical importance in obtaining compact conservation-law equations with
coecients derived from the metric elements in new curvilinear coordinates
1 , . . . , n is the formula for dierentiation of the Jacobian
 i
J 2 xi m x x
J J Jdivx k ,
k k m xi xi k
i, k, m = 1, . . . , n. (2.45)
In accordance with (2.37), this identity can also be expressed through the
i
space Christoel symbols of the second kind kj by
J i
= Jik , i, k = 1, . . . , n,
k
with the summation convention over the repeated index i.
In order to prove the identity (2.45) we note that in the case of an arbitrary
matrix (aij ) the rst derivative of its Jacobian with respect to k is obtained
by the process of dierentiating the rst row (the others are left unchanged),
then performing the same operation on the second row, and so on with all
of the rows of the matrix. The summation of the Jacobians of the matrices
derived in such a manner gives the rst derivative of the Jacobian of the
original matrix (aij ). Thus
aim im
det(aij ) = G , i, j, k, m = 1, . . . , n, (2.46)
k k
where Gim is the cofactor of the element aim . For the Jacobi matrix (xi / j )
of the coordinate transformation x() we have
m
Gim = J , i, j = 1, . . . , n.
xi
Therefore, applying (2.46) to the Jacobi matrix, we obtain (2.45).
 2.3 Relations Concerning Second Derivatives 47

2.3.3 Basic Identity

The identity (2.45) implies the extremely important relation

 
j
J i 0, i, j = 1, . . . , n, (2.47)
j x
which leads to specic forms of new dependent variables for conservation-law
equations. To prove (2.47) we rst note that

2 p xl 2 xl m p
k j p
= p m .
x x xk xj

Multiplying this equation by i /xl and summing over l, we obtain a for-

mula representing the second derivative 2 i /xk xm of the functions i (x)
through the second derivatives 2 xm / l p of the functions xm (), m =
1, . . . , n:

2i 2 xp j l i
= , i, j, k, l, m, p = 1, . . . , n. (2.48)
xk xm l j xk xm xp
Now, using this relation and the formula (2.45) for dierentiation of the Ja-
cobian in the identity
 
j J j 2 j xk
J = + J ,
j xi j xi xi xk j
we obtain
 
j 2 xk p j 2 xp m l j xk
j
J i
=J p j k i J l m
x x x xi xk xp j
2 xk p j 2 xp l m
=J J = 0,
p j xk xi l m xp xi
i, j, k, l, m, p = 1, . . . , n,

i.e. (2.47) has been proved.

The identity (2.47) is obvious when n = 1 or n = 2. For example, for n = 2
we have from (2.3)

j x3i
J i
= (1)i+j 3j , i, j = 1, 2,
x
with xed indices i and j, and therefore
   
j x3i x3i
J = (1) i+1
= 0, i, j = 1, 2.
j xi 1 2 2 1
An inference from (2.47) for n = 3 follows from the dierentiation of
the cross product of the base tangential vectors r i , i = 1, 2, 3. Taking into
48 2 Coordinate Transformations

account (2.25), we readily obtain the following formula for the dierentiation
of the cross product of two three-dimensional vector-valued functions a and b:

(a b) = i a b + a i b, i = 1, 2, 3.
i
With this formula we obtain
3
3 3
(x j x k ) = x j i x k + xj xk i , (2.49)
i=1
i
i=1

i=1

where the indices (i, j, k) are cyclic, i.e. j = i + 1, k = i + 2, m is equivalent

to m + 3. For the last summation of the above formula, we obtain


3 
3
xj xk i = xk xi j .
i=1 i=1

Therefore, from (2.49),

3

i
(xj xk ) = 0,
i=1

since
xi xj k = xj k xi
and (2.32) implies (2.47) for n = 3.
The identity (2.47) can help one to obtain conservative or compact forms
of some dierential expressions and equations in the curvilinear coordinates
1 , . . . , n . For example, for the rst derivative of a function f (x) with respect
to xi we obtain, using (2.47),
 
f 1 j
= J f , j = 1, . . . , n. (2.50)
xi J j xi

For the Laplacian

f
2 f = , j = 1, . . . , n (2.51)
xj xj
we have, substituting the quantity f /xi for f in (2.50),
   
1 j f 1 j m f
2 f = J = J
J j xi xi J j xi xi m
 
1 f
= Jg mj m , i, j, m = 1, . . . , n. (2.52)
J j

Therefore the Poisson equation

 2.4 Conservation Laws 49

2 f = P (2.53)

has the form  

1 mj f
Jg =P (2.54)
J j m
with respect to the independent variables 1 , . . . , n .

2.4 Conservation Laws

This section utilizes the relations described in Sects. 2.2 and 2.3, in particular
the identity (2.47), in order to describe some conservation-law equations of
mechanics in divergent or compact form in new independent curvilinear coor-
dinates 1 , . . . , n . For this purpose the dependent physical variables are also
transformed to new dependent variables using some specic formulas. The
essential advantage of the equations described here is that their coecients
are derived from the elements of the covariant metric tensor (gij ).

2.4.1 Scalar Conservation Laws

Let A be an n-dimensional vector with components Ai , i = 1, . . . , n, in the

Cartesian coordinates x1 , . . . , xn . The operator

Ai
divx A = , i = 1, . . . , n, (2.55)
xi
is commonly used in mechanics for the representation of scalar conservation
laws, commonly in the form
divx A = F.
Using (2.47) we easily obtain

1 j
divx A = j
(JA ) = F, j = 1, . . . , n, (2.56)
J
j
where A is the jth contravariant component of the vector A in the coordinates
i , i = 1, . . . , n, i.e. in accordance with (2.7):

j j
A = Ai , i, j = 1, . . . , n. (2.57)
xi
Therefore a divergent form of the conservation-law equation represented by
(2.55) is obtained in the new coordinates when the dependent variables Ai
i
are replaced by new dependent variables A dened by the rule (2.57). Some
examples of scalar conservation-law equations are given below.
50 2 Coordinate Transformations

2.4.1.1 Mass Conservation Law

As an example of the application of (2.56), we consider the equation of con-

servation of mass for steady gas ow

ui
= 0, i = 1, . . . , n, (2.58)
xi
where is the gas density, and ui is the ith component of the ow velocity
vector u in the Cartesian coordinates x1 , . . . , xn . With the substitution Ai =
ui , (2.58) is transformed to the following divergent form with respect to the
new dependent variables and ui in the coordinates 1 , . . . , n :

(Jui ) = 0, i = 1, . . . , n. (2.59)
i

Here ui is the ith contravariant component of the ow velocity vector u in

the basis xi , i = 1, . . . , n, i.e.

i
ui = uj , i, j = 1, . . . , n. (2.60)
xj

2.4.1.2 ConvectionDiusion Equation

Another example is the conservation equation for the steady convection

diusion of a transport variable , which can be expressed as
 

i  i + (ui ) = S, i = 1, . . . , n, (2.61)
x x xi

where and  denote the density and diusion coecient of the uid, respec-
tively. Taking

Ai = ui  i ,
x
we obtain, in accordance with the relation (2.57),

j kj
A = uj  g .
k

Therefore, using (2.56), the convectiondiusion equation (2.61) in the curvi-

linear coordinates 1 , . . . , n is expressed by the divergent form
  
kj
J u g j
= JS, j, k = 1, . . . , n. (2.62)
j k
 2.4 Conservation Laws 51

2.4.1.3 Laplace Equation

Analogously, the Laplace equation

f
2 f = = 0, j = 1, . . . , n, (2.63)
xj xj
has the form (2.55) if we take
f
Ai = , i = 1, . . . , n.
xi
Using (2.57), we obtain

j f
A = g ij , i = 1, . . . , n.
i
Therefore the Laplace equation (2.63) results in
 
1 ij f
f=
2
Jg = 0, (2.64)
J i j

since (2.56) applies.

2.4.2 Vector Conservation Laws

Many physical problems are also modeled as a system of conservation-law

equations in the vector form

Aij
= F i, i, j = 1, . . . , n. (2.65)
xj
For the representation of the system (2.65) in new coordinates 1 , . . . , n in a
form which includes only coecients derived from the elements of the metric
tensor, it is necessary to make a transition from the original expression for
ij
Aij to a new one A . One convenient formula for such a transition from the
ij
dependent variables Aij to A , i, j = 1, . . . , n,

ij i j
A = Akm , i, j, k, m = 1, . . . , n. (2.66)
xk xm
km
This relation between Aij and A is in fact composed of transitions of the
kind (2.57) for the rows and columns of the tensor Aij . In tensor analysis
ij
the quantity A means the (i, j) component of the second-rank contravariant
tensor (A ) in the coordinates 1 , . . . , n .
ij

Multiplying (2.66) by (xp / i )(xl / j ) and summing over i and j we

ij
also obtain a formula for the transition from the new dependent variables A
to the original ones Aij :
52 2 Coordinate Transformations
i
km x xj
Aij = A k
, i, j, k, m = 1, . . . , n. (2.67)
m
Therefore we can obtain a system of equations for the new dependent variables
ij
A by replacing the dependent quantities Aij in (2.65) with their expressions
(2.67). As a result we obtain
 
Aij i
km x x
j
= A
xj xj k m
km  j
A xi km x
2 i
km x
i
x
= + A + A
m k k m k xj m
= F i, i, j, k, m = 1, . . . , n.

The use of the formula (2.45) for dierentiation of the Jacobian in the sum-
mation in the equation above yields
km
Aij A xi 2 i
km x 1 km xi J
j
= m k
+A k m
+ A = F i,
x J k m
i, j, k, l, m = 1, . . . , n.

Multiplying this system by p /xi and summing over i we obtain, after

simple manipulations,
1 ij 2 xl i kj i
(JA ) + A =F , i, j, k, l = 1, . . . , n, (2.68)
J j k j xl
where
i i
F = Fj , i, j = 1, . . . , n,
xj
is the ith contravariant component of the vector F = (F 1 , . . . , F n ) in the
basis x1 , . . . , xn . The quantities ( 2 xl / k j )( i /xl ) in (2.68) are the
i
space Christoel symbols of the second kind kj . Thus the system (2.68) has,
i
using the notation jk , the form

1 ij i kj i
(JA ) + kj A =F , i, j, k = 1, . . . , n. (2.69)
J j
We see that all coecients of (2.69) are derived from the metric tensor (gij ).
Equations of the form (2.69), in contrast to (2.65), do not have a conser-
vative form. The conservative form of (2.65) in new dependent variables is
obtained, in analogy with (2.56), from the system
1 j
(JAi ) = F i , i, j = 1, . . . , n, (2.70)
J j
j
where Ai is the jth component of the vector Ai = (Ai1 , . . . , Ain ) in the basis
xj , j = 1, . . . , n, i.e.
 2.4 Conservation Laws 53

j j
Ai = Aik , i, j, k = 1, . . . , n. (2.71)
xk
In fact, (2.70) is the result of the application of (2.56) to the ith line of (2.65).
Therefore in the relations (2.65), (2.70), (2.71) we can assume an arbitrary
range for the index i, i.e. the matrix Aij in (2.65) can be a nonsquare matrix
with i = 1, . . . , m, j = 1, . . . , n.
Though the system (2.70) is conservative and more compact than (2.69),
it has its drawbacks. In particular, mathematical simulations of uid ows are
generally formulated in the form (2.65) with the tensor Aij represented as

Aij = B ij + ui uj , i, j = 1, . . . , n,

where ui , i = 1, . . . , n, are the Cartesian components of the ow velocity. The

transformation of the tensor ui uj by the rule (2.71),

j
ui uk = ui uj , i, j, k = 1, . . . , n,
xk
results in equations with an increased number of dependent variables, namely
ui and uj . The substitution of ui for uj or vice versa leads to equations whose
coecients are derived from the elements xi / j of the Jacobi matrix and
not from the elements of the metric tensor (gij ).

2.4.2.1 Example

As an example of (2.65) we consider the stationary equation of a compressible

gas ow

p ui
(u i j
u ) + = F i , i, j = 1, . . . , n, (2.72)
xj xi xj xj
where ui is the ith Cartesian component of the vector of the uid velocity u,
is the density, p is the pressure and, is the viscosity. The tensor form of
(2.65) is given by

ui
Aij = ui uj + ji p , i, j = 1, . . . , n.
xj
From (2.66) we obtain in this case

ij ul i j
A = ui uj + g ij p , (2.73)
xk xl xk
where ui is the ith component of u in the basis xi , i.e. ui is computed from
the formula (2.60). It is obvious that

xl
ul = uj , j, l = 1, . . . , n. (2.74)
j
54 2 Coordinate Transformations

Therefore
  m
ul p x
l

= u
xk m p xk
up xl m 2 xl m
= m p k
+ up p m .
x xk
Using this equation, we obtain, for the last term of (2.73),
 i 
ul i j u
k l k = g mj + i
pm u p
, i, j, m, p = 1, . . . , n,
x x x m

since (2.37) applies. Thus (2.73) has the form

 i 
ij u
A = u u + g p g
i j ij mj i p
+ pm u , i, j, m, p = 1, . . . , n, (2.75)
m

and, applying (2.69), we obtain the following system of stationary equations

(2.72) with respect to the new dependent variables , ui , and p and the inde-
pendent variables i :
   i 
1 u
J u i j
u + g ij
p g mj
+ i
pm u p
J j m
  k 
mj u i
+ kj u u + g p g
i k j kj k
+ pm u p
= F ,
m
i, j, k, m, p = 1, . . . , n. (2.76)

The application of (2.70) to (2.72) yields the following system of stationary

equations:
  
1 j ui kj
i j
J u u + p kg = F i ,
J j xi
j
uj = uk k , i, j, k = 1, . . . , n. (2.77)
x
Now, as an example of the utilization of the Christoel symbols of the
i
second kind kj , we write out the expression for the transformed elements of
the tensor  i 
u uj
ij = + , i, j = 1, . . . , n, (2.78)
xj xi
in the coordinates 1 , . . . , n , obtained in accordance with the rule (2.66). This
tensor is very common and is important in applications simulating deforma-
tion in the theory of elasticity and deformation rate in uid mechanics. Using
the notations described above, the tensor ij can be expressed in the coordi-
nates 1 , . . . , n through the metric elements and the Christoel symbols of
the second kind. For the component ij we have
 2.5 Time-Dependent Transformations 55

i j
ij = mk m k
x x
 i j

jl u il u jl i il j p
= g +g + (g pl + g pl )u ,
l l
i, j, l, p = 1, . . . , n. (2.79)

This formula is obtained rather easily. For this purpose one can use the relation
(2.74) for the inverse transition from the contravariant components ui to the
Cartesian components uj of the vector u = (u1 , . . . , un ) and the formula
(2.37). By substituting (2.74) in (2.78), carrying out dierentiation by the
chain rule, and using the expression (2.37), we obtain (2.79).

2.5 Time-Dependent Transformations

The numerical solution of time-dependent equations requires the application
of moving grids and the corresponding coordinate transformations, which are
dependent on time. Commonly, such coordinate transformations are deter-
mined in the form of a vector-valued time-dependent function

x(t, ) : n Xtn , n , t [0, 1], (2.80)

where the variable t represents the time and Xtn is an n-dimensional domain
whose boundary points change smoothly with respect to t. It is assumed that
x(t, ) is suciently smooth with respect to i and t and, in addition, that
it is invertible for all t [0, 1]. Therefore there is also the time-dependent
inverse transformation
(t, x) : Xtn n (2.81)
for every t [0, 1]. The introduction of these time-dependent coordinate trans-
formations enables one to compute an unsteady solution on a xed uniform
grid in n by the numerical solution of the transformed equations.

2.5.1 Reformulation of Time-Dependent Transformations

Many physical problems are modeled in the form of nonstationary conserv-

ation-law equations which include the time derivative. The formulas of
Sects. 2.3 and 2.4 can be used directly, by transforming the equations at every
value of time t. However, such utilization of the formulas does not inuence
the temporal derivative, which is transformed simply to the form

i
+ , i = 1, . . . , n,
t t i
so that does not maintain the property of divergency and its coecients are
not derived from the elements of the metric tensor.
56 2 Coordinate Transformations

Instead, the formulas of Sects. 2.3 and 2.4 can be more successfully applied
to time-dependent conservation-law equations if the set of the functions x(t, )
is expanded to an (n + 1)-dimensional coordinate transformation in which the
temporal parameter t is considered in the same manner as the spatial variables.
To carry out this process we expand the n-dimensional computational and
physical domains in (2.80) to (n + 1)-dimensional ones, assuming

n+1 = I n , X n+1 = t Xtn .
t

Let the points of these domains be designated by 0 = ( 0 , 1 , . . . , n ) and

x0 = (x0 , x1 , . . . , xn ), respectively. The expanded coordinate transformation
is dened as
x0 ( 0 ) : n+1 X n+1 , (2.82)
where x0 ( 0 ) = 0 and xi ( 0 ), i = 1, . . . , n, this coincides with (2.80) with
0 = t.
The variables x0 and 0 in (2.82) represent in fact the temporal variable t.
For convenience and in order to avoid ambiguity we shall also designate the
variable 0 in n+1 by and the variable x0 in X n+1 by t. Thus x0 ( 0 ) is the
(n + 1)-dimensional coordinate transformation which is identical to x(, ) at
every section 0 = .
The inverted coordinate transformation

0 (x0 ) : X n+1 n+1 (2.83)

satises
0 (x0 ) = x0 , i (x0 ) = i (t, x), i = 1, . . . , n,
where t = x0 , x = (x1 , . . . , xn ), and i (t, x) is dened by (2.81). Thus (2.83)
is identical to (2.81) at each section Xt .

2.5.2 Basic Relations

This subsection discusses some relations and, in particular, identities of the

kind (2.45) and (2.47) for the time-dependent coordinate transformations
(2.80), using for this purpose the (n + 1)-dimensional vector functions (2.82)
and (2.83) introduced above.

2.5.2.1 Velocity of Grid Movement

The rst derivative x , x = (x1 , x2 , . . . , xn ), of the transformation x(, ) has

a clear physical interpretation as the velocity vector of grid point movement.
Let the vector x , in analogy with the ow velocity vector u, be designated
by w = (w1 , . . . , wn ), i.e. wi = xi . The ith component wi of the vector wi in
the tangential bases xi , i = 1, . . . , n, is expressed by (2.7) as
 2.5 Time-Dependent Transformations 57

i xj i
wi = wj j
= , i, j = 1, . . . , n.
x xj
Therefore
w = w i x i , i = 1, . . . , n, (2.84)
i.e.
xi xi
wi = = wj j , i, j = 1, . . . , n.

Dierentiation with respect to 0 of the composition of x0 ( 0 ) and 0 (x0 )
yields
i x0 i xj
+ = 0, i, j = 1, . . . , n.
x0 0 xj 0
Therefore we obtain the result
i xj i
= = wi , i, j = 1, . . . , n. (2.85)
t xj

2.5.2.2 Derivatives of the Jacobian

It is apparent that the Jacobians of the coordinate transformations x(, )

and x0 ( 0 ) coincide, i.e.
 i  k
x x
det = det = J, i, j = 0, 1, . . . , n, k, l = 1, . . . , n.
j l

In the notation introduced above, the formula (2.45) for dierentiation of the
Jacobian of the transformation

x0 ( 0 ) : n+1 X n+1

is expressed by the relation

1 2 xk m
J = , i, k, m = 0, 1, . . . , n, (2.86)
J i i m xk
diering from (2.45) only by the range of the indices. As a result, we obtain
from (2.86) for i = 0,
 k m
1 x x
J= m k
= divx , k, m = 0, 1, . . . , n,
J x

and, taking into account (2.84),

 k
 m
1 j x
J= m w j
J xk
wm 2 k
i x m
= + w , j, k, m = 1, . . . , n.
m j m xk
58 2 Coordinate Transformations

Now, taking advantage of the formula for dierentiation of the Jacobian (2.45),
in the last sum of this equation we have
1 wm 1 J
J= + wj j , j, m = 1, . . . , n,
J m J
and consequently
1 1
J= (Jwj ), j = 1, . . . , n. (2.87)
J J j

2.5.2.3 Basic Identity

Analogously, the system of identities (2.47) has the following form:

 
j
J i = 0, i, j = 0, 1, . . . , n. (2.88)
j x
Therefore for i = 0 we obtain
 
j
(J) + j J = 0, j = 1, . . . , n, (2.89)
t

and, taking into account (2.85),

J j (Jwj ) = 0, j = 1, . . . , n, (2.90)

which corresponds to (2.87). For i > 0 the identity (2.88) coincides with (2.47),
i.e.  
j
J i = 0, i, j = 1, . . . , n.
j x
As a result of (2.89), we obtain, in analogy with (2.50),
   
f 1 j 1
= J f = (Jf ) (Jw k
f ) ,
t J j t J k
j = 0, 1, . . . , n, k = 1, . . . , n. (2.91)

2.5.3 Equations in the Form of Scalar Conservation Laws

Many time-dependent equations can be expressed in the form of a scalar

conservation law in the Cartesian coordinates t, x1 , . . . , xn :

A0 Ai
+ = F, i = 1, . . . , n. (2.92)
t xi
Using (2.88), in analogy with (2.56), this equation is transformed in the co-
ordinates 0 , 1 , . . . , n , 0 = to
 2.5 Time-Dependent Transformations 59
 
1 j
(JA0 ) = F, j = 0, 1, . . . , n, (2.93)
J j
j
where by A0 we denote the jth contravariant component of the (n + 1)-
dimensional vector A0 = (A0 , A1 , . . . , An ) in the basis x0 / i , i = 0, 1, . . . , n,
i.e.
j j
A0 = Ai i , i, j = 0, 1, . . . , n. (2.94)
x
j
We can express each component A0 , j = 1, . . . , n, of the vector A0 through the
i
components A and wk , i, k = 1, . . . , n, of the n-dimensional spatial vectors
A = (A1 , . . . , An ) and w = (w1 , . . . , wn ) in the coordinates l , l = 1, . . . , n,
where A is a vector obtained by projecting the vector A0 into the space Rn ,
i.e. P (A0 , A1 , . . . , An ) = (A1 , . . . , An ). Namely,

j j j j
A0 = A0 + Ai i = A A0 wj , i, j = 1, . . . , n,
t x
using (2.85). Further, we have

0 0
A0 = Ak = A0 , k = 0, 1, . . . , n.
xk
Therefore (2.93) implies a conservation law in the variables , 1 , . . . , n in the
conservative form
 
1 j
(JA ) + j [J(A A w )] = F, j = 1, . . . , n.
0 0 j
(2.95)
J

2.5.3.1 Examples of Scalar Conservation-Law Equations

As an illustration of the formula (2.95), we write out some time-dependent

scalar conservation law equations presented rst in the form (2.92).

Parabolic Equation

For the parabolic equation

f f
= , j = 1, . . . , n, (2.96)
t xj xj
we obtain from (2.95), with A0 = f and Ai = f /xi , i = 1, . . . , n,
  
Jf f
= j J g jk k + f wj , j, k = 1, . . . , n. (2.97)

60 2 Coordinate Transformations

Mass Conservation Law

The scalar mass conservation law for unsteady compressible gas ow

ui
+ = F, i = 1, 2, 3, (2.98)
t xi
is expressed in the new coordinates as

J J(uj wj )
+ = JF, j = 1, 2, 3. (2.99)
t j
ConvectionDiusion Equation

The unsteady convectiondiusion conservation equation

 

() + (u )
i
 = S, i = 1, . . . , n, (2.100)
t xi xi xi

has the form in the coordinates , 1 , . . . , n

 
kj
(J) + j J(u w ) Jg  k = JS,
j j
j, k = 1, . . . , n. (2.101)

Energy Conservation Law

Analogously, the energy conservation law

(e + u2 /2) + j uj (e + u2 /2 + p/) = Fj uj , j = 1, 2, 3, (2.102)
t x
where
ui
e = e(, p), p=p , i = 1, 2, 3,
xi

3
u2 = (ui )2 ,
i=1

is transformed in accordance with (2.95) to

      
1 1
J e + m k
gmk u u + j J e + gmk u u (u w ) + Jpu
m k j j j
2 2
m
= Jgmk f uk , j, m, k = 1, 2, 3, (2.103)

where, taking into account (2.56),

p=p (Jui ), i = 1, 2, 3.
J i
 2.5 Time-Dependent Transformations 61

Linear Wave Equation

The linear wave equation

utt = c2 2 u (2.104)
arises in many areas such as uid dynamics, elasticity, acoustics, and magne-
tohydrodynamics. If the coecient c is constant than (2.104) has a divergent
form (2.92) with

u
A0 = ut , Ai = c2 , i = 1, . . . , n,
xi
or, in the coordinates , 1 , . . . , n ,

u u k
A0 = u wi , Ai = c2 , i, k = 1, . . . , n.
i k xi

Therefore the divergent form (2.93) of (2.104) in the coordinates , 1 , . . . , n

has the form
     
u u
J u w i i + j J u wj + (c2 g mj wi wj ) i = 0. (2.105)

Another representation of the linear wave equation (2.104) in the coordi-

nates , 1 , . . . , n comes from the formula (2.64) for the Laplace operator and
the description of the temporal derivative (2.91). Taking advantage of (2.91),
we obtain
    
1 u k u
utt = J k Jw
J t t
  
1 u
= J u w i i
J
  
1 i u
Jw u w
k
.
J k i

This equation and (2.64) allow one to derive the following form of (2.104) in
the coordinates , 1 , . . . , n :
  
i u
J u w
i
    
u u
= k Jwk u wj i + c2 k Jg kj j , (2.106)

which coincides with (2.105) if c2 is a constant.

62 2 Coordinate Transformations

2.5.3.2 Lagrangian Coordinates

One of the most popular systems of coordinates in uid dynamics is the La-
grangian system. A coordinate i is Lagrangian if the both the ith component
of the ow velocity vector u and the grid velocity w in the tangent basis
xj , j = 1, . . . , m, coincide, i.e.

ui wi = 0. (2.107)

The examples of gas-dynamics equations described above, which include the

terms wi , allow one to obtain the equations in Lagrange coordinates by sub-
stituting ui for wi in the written-out equations in accordance with the relation
(2.107). In such a manner, we obtain the equation of mass conservation, for
example, in the Lagrangian coordinates 1 , . . . , n as
J
= JF (2.108)

from (2.99). Analogously, the convectiondiusion equation (2.101) and the
energy conservation law (2.103) have the forms in Lagrangian coordinates i
 
kj
(J) + j Jg  k = JS, j, k = 1, . . . , n,

and
  
1 m k m
J e + gmk u u + j (Jpuj ) = Jgmk f uk , j, m, k = 1, 2, 3,
2

respectively.
In the same manner, the equations can be written in the EulerLagrange
form, where some coordinates are Lagrangian while the rest are Cartesian
coordinates.

2.5.4 Equations in the Form of Vector Conservation Laws

Now we consider a formula for a vector conservation law with time-dependent

physical magnitudes Aij
ij
A = F i, i, j = 0, 1, . . . , n, (2.109)
xj
where the independent variable x0 represents the time variable t, i.e. x0 = t.
Let the new independent variables 0 , 1 , . . . , n be obtained by means of
(2.82). As well as (2.69), which expresses the vector conservation law (2.65)
in the coordinates 1 , . . . , n , we nd that the transformation (2.109) has the
form of the following system of equations for the new dependent quantities
 2.5 Time-Dependent Transformations 63
ij
A0 , i, j = 0, 1, . . . , n, with respect to the independent variables 0 , 1 , . . . , n ,
0 = :
1 ij i kj i
(JA0 ) + kj A0 = F 0 , i, j = 0, 1, . . . , n, (2.110)
J j
where
ij i j
A0 = Amn , i, j, m, n = 0, 1, . . . , n,
xm xn
i 2 xl i
kj = k j l , i, j, k, l = 0, 1, . . . , n,
x
i i
F0 = Fj j , i, j = 0, 1, . . . , n.
x
As in the case of the scalar conservation law, we represent all of the terms of
(2.110) through A00 and the spatial components:

ij i j
A = Akm , i, j, m, n = 1, . . . , n,
xk xm
i 2 xl i
kj = k j l, i, j, k, l = 1, . . . , n,
x
i i
F = Fj j , i, j = 1, . . . , n,
x
i
xj i
wi = = , i, j = 1, . . . , n.
t xj
ij
For A0 we obtain
00
A0 = A00 ,
0i i i 0i
A0 = A00 + A0m m = A A00 wi , i = 1, . . . , n,
t x
i0 i0
A0 = A A00 wi , i = 1, . . . , n,
ij 0j i0 ij
A0 = A00 wi wj A wi A wj + A , i, j = 1, . . . , n.
i
Analogously, for kj we obtain
0
kj = 0, k, j = 0, 1, . . . , n,
i
i w
00 = + wl wm lm
i
, i, l, m = 1, . . . , n,
t
i
i i w
j0 = 0j = + wl jli , i, j, l = 1, . . . , n,
j
i i
jk = jk , i, j, k = 1, . . . , n,
i
and for F 0 ,
64 2 Coordinate Transformations
0
F 0 = F 0,
i i
F 0 = F A0 wi , i = 1, . . . , n.
Using these expression in (2.110), we obtain a system of equations for the vec-
tor conservation law in the coordinates , 1 , . . . , n with an explicit expression
for the components of the speed of the grid movement:
0i
(JA00 ) + j [J(A A00 wj )] = JF 0 ,

i0 ij 0j i0
J(A A00 wj ) + j J(A + A00 wi wj A wj A wi )

 i i

00 w l w l j i
+ JA +w + w w lj
l (2.111)
 i 
j0 0j w
+ J(A + A 2A00 wj ) + wl jli
j
lj 0j 0l
+ J(A + A00 wl wj A wl A wj )lji
i
= J(F F 0 wi ), i, j, l = 1, . . . , n.
Another representation of (2.109) in new coordinates can be derived in
the form of (2.95) by applying (2.95) to each line of the system (2.109). As a
result we obtain
   
1 j
(JAi0 ) + j J Aik k Ai0 wj = F, j, k = 1, . . . , n. (2.112)
J x
Recall that this approach is not restricted to a square form of the system
(2.109), i.e. the ranges for the indices i and j can be dierent.
As an illustration of these equations for a vector conservation law in the
curvilinear coordinates , 1 , . . . , n , we write out a joint system for the con-
servation of mass and momentum, which in the coordinates t, x1 , x2 , x3 has
the following form:

+ ui = 0, i = 1, 2, 3,
t xi
(2.113)
ui
+ (ui uj + pji ) = f i , i, j = 1, 2, 3,
t xj
where
ui
p=p , ji = 0 if i = j and ji = 1 if i = j.
xi
This system is represented in the form (2.109) with
A00 = ,
A0i = Ai0 = ui , i = 1, 2, 3,
Aij = ui uj + ij p, i, j = 1, 2, 3,
 2.5 Time-Dependent Transformations 65

i.e.

u1 u2 u3
u1 1 1
u u + p u1 u2 u1 u3
(A ) =
ij
u2
,
u2 u1 u2 u2 + p u2 u3
u3 u3 u1 u3 u2 u3 u3 + p
i, j = 0, 1, 2, 3 (2.114)

and

F 0 = 0,
F i = f i , i = 1, 2, 3.

For the coordinate system , 1 , . . . , n we obtain

00
A = ,
0i i0
A =A = (ui wi ), i = 1, 2, 3,
ij
A = (u w )(u w ) + pg ij ,
i i j j
i, j = 1, 2, 3.

Substituting these expressions in (2.111) we obtain a system of equations for

the mass and momentum conservation laws in the coordinates , 1 , . . . , n :

(J) + j [J(uj wj )] = 0, i = 1, 2, 3,


[J(ui wi )] + j [J(ui wi )(uj wj ) + Jpg ij ]
(2.115)
wi wi
+ J + J(2uj wj ) j + J(ul uj + pg lj )lji

i
= Jf , i, j, l = 1, 2, 3.

If the coordinates i are the Lagrangian ones, i.e. ui = wi , then we obtain

from (2.115)

(J) = 0,

(2.116)
ui ui i
J + j (Jpg ij ) + Juj j + J(ul uj + pg lj )lji = Jf .

Note that the rst equation of the system (2.115) coincides with (2.99) if
F = 0; this was obtained as the scalar mass conservation law.
In the same manner, we can obtain an expression for the general Navier
Stokes equations of mass and momentum conservation by inserting the tensor
( ij ) described by (2.78) in the system (2.113) and the tensor ( ij ) represented
by (2.79) in the system (2.115).
66 2 Coordinate Transformations

A divergent form of (2.113) in arbitrary coordinates , 1 , . . . , n is ob-

tained by applying (2.112). With this, we obtain the system

(J) + j [J(uj wj )] = 0,

  
j
(Ju ) + j J u (u w ) + p i
i i j j
= JF i , (2.117)
x
j
uj = uk k , i, j, k = 1, 2, 3,
x
and, in the Lagrangian coordinates,

(J) = 0,
  (2.118)
j
(Jui ) + j Jp i = JF i , i, j = 1, 2, 3.
x

2.6 Comments
Many of the basic formulations of vector calculus and tensor analysis may be
found in the books by Kochin (1951), Sokolniko (1964) and Gurtin (1981).
The formulation of general metric and tensor concepts specically aimed
at grid generation was originally performed by Eiseman (1980) and Warsi
(1981).
Very important applications of the most general tensor relations to the
formulation of unsteady equations in curvilinear coordinates in a strong con-
servative form were presented by Vinokur (1974). A strong conservation-law
form of unsteady Euler equations also was also described by Viviand (1974).
A derivation of various forms of the NavierStokes equations in general
moving coordinates was described by Ogawa and Ishiguto (1987).
http://www.springer.com/978-90-481-2911-9