CONTINUUM MECHANICS Dr: Saib Ch.

Chapter 1 : Tensor calculation reminders

 Introduction
 Vector algebra
 Einstein's notation
 Delta kronecker: Properties
 Levi-cevita symbol: Properties
 Transformation law
 Matrices
 Derivative of tensors
 Gauss and stokes integral theorems

I. Introduction :
The continuum mechanics: is the part of physics which studies the mechanics of
deformable solid bodies, it is a physical and mathematical framework allowing a concrete
problem to be modeled. Once the mathematical model has been established, it can be
solved by an analytical or numerical method.…
II. Vector algebra :
When we study physical phenomena we encounter three kinds of magnitudes, the scalars,
vectors and tensors.
A scalar is defined by a single number independently of the example reference axes:
(mass, temperature, volume, pressure....etc.).
A vector is inseparable from the concept of direction; it is defined by its direction, its
length, its point of application and by its supporting axis [1].
A Tensor: is a generic word of mathematical entities that designates a physical quantity.
Tensors are the generalization of scalars, vectors and matrices.
The description of a tensor depends on the reference, we interested in Cartesian tensors,
and the reference system is an orthonormal reference which can be rotated by
transformation matrix and change into another reference [2].

Order Entity Example [1] Dimension Nbr components Notation

0 Scalar Temperature 2D 20=1 
3D 30=1
1 Vector Displacement 2D 21=2 𝑣
3D 31=3
2 Matrix Stress 2D 22=4 𝑎
3D 32=9

II.1 Coordinate system, base vectors:

A vector may be defined with respect to a particular
coordinate system by specifying the components of the vector
in that system. The choice of coordinate system is arbitrary, but
in certain situations a particular choice may be advantageous.
Course 1 page : 1

Figure .1.
CONTINUUM MECHANICS Dr: Saib Ch.
The reference system of coordinate axes provides units for measuring vector magnitudes
and assigns directions in space by which the orientation of vectors may determinate.
The well-known rectangular Cartesian coordinate system is often represented by the
mutually perpendicular axes, 𝑂, 𝑋 , 𝑋 , 𝑋 ) shown in Figure .1.
The most frequent choice of base vectors for rectangular Cartesian system is the set of unit
vectors 𝑒⃗ , 𝑒⃗ , 𝑒⃗ along the coordinate axes as shown in Figure .1.
The orthonormal vector basis 𝑂, 𝑒⃗ , 𝑒⃗ , 𝑒⃗ ), must satisfy the following conditions:

 ‖𝑒⃗ ‖ ‖𝑒⃗ ‖ ‖𝑒⃗ ‖ 1,

 𝑒⃗ . 𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ . 𝑒⃗ 0 ,
 𝑒⃗ . 𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ . 𝑒⃗ 1 and the following :
 𝑒⃗ 𝑒⃗ 𝑒⃗ 𝑒⃗ 𝑒⃗ 𝑒⃗ 0⃗ ,
 𝑒⃗ 𝑒⃗ 𝑒⃗ , 𝑒⃗ 𝑒⃗ 𝑒⃗ ,
 𝑒⃗ 𝑒⃗ 𝑒⃗ , 𝑒⃗ 𝑒⃗ 𝑒⃗ ,
 𝑒⃗ 𝑒⃗ 𝑒⃗ , 𝑒⃗ 𝑒⃗ 𝑒⃗ ,

II.2 Position vector :

In a reference space we define an orthonormal
vector basis 𝑂, 𝑒⃗ , 𝑒⃗ , 𝑒⃗ ). A point P in space can be
defined as a position vector 𝑂𝑃⃗ , which gives the
position of the point P in space with respect to the
defined reference.
The position vector can be expressed by: 𝑂𝑃⃗ 𝑋 𝑒⃗
𝑋 𝑒⃗ 𝑋 𝑒⃗ ;
where: 𝑋 , 𝑋 , 𝑋 are the components of the vector 𝑂𝑃⃗
in the orthonormal basis 𝑂, 𝑒⃗ , 𝑒⃗ , 𝑒⃗ ).
Figure .2. Projection of a point P defined in an
orthonormal vector basis 𝑂, 𝑒⃗ , 𝑒⃗ , 𝑒⃗ ).

The magnitude or the module of the position vector is given by: 𝑂𝑃⃗ 𝑋 𝑋 𝑋
Or: 𝑂𝑃⃗ . 𝑂𝑃⃗ 𝑋 𝑋 𝑋

_Unit vector : is a vector of length “magnitude, modulus” equal to one (unit).

‖𝑒⃗‖ 1.
_ We can extract a unit vector from a non-zero vector, as follows:
⃗
𝑒⃗ , or: 𝑢⃗ ‖𝑢⃗‖. 𝑒⃗ .
‖ ⃗‖

_ Null vector: is a zero-length vector. 0⃗ , or: 0⃗ 0.

Course 1 page : 2
CONTINUUM MECHANICS Dr: Saib Ch.

II.3 Vector operations :

Let be the non null vectors: 𝑢⃗, 𝑣⃗ and 𝑤⃗,

 𝑢⃗ 𝑣⃗ 𝑣⃗ 𝑢⃗ (commutative),
 𝑢⃗ 𝑣⃗ 𝑤⃗ 𝑢⃗ 𝑣⃗ 𝑤⃗ (associativity),
 𝑢⃗ 0⃗ 0⃗ 𝑢⃗ 𝑢⃗ (null element),
 𝑢⃗ 𝑢⃗ 0⃗ (opposing element).

 Multiplying vectors by a scalar:

Let be the scalars ,  and the vectors 𝑢⃗, 𝑣⃗ ;

** ( 𝑢⃗) = ()𝑢⃗ (associativity),

** (+)𝑢⃗= 𝛼𝑢⃗ β𝑢⃗ (distributive scalar addition),
** 𝛼 𝑢⃗ 𝑣⃗ 𝛼𝑢⃗ 𝛼𝑣⃗ (addition distribution vector),
** 1. 𝑢⃗ 𝑢⃗. 1 𝑢⃗,
** 0. 𝑢⃗ 0⃗,

II.4 The dot or inner product or the scalar product of the vectors: The result is a scalar
(number).
Let be the vectors 𝑢⃗, 𝑣⃗ and their successive components 𝑢 , 𝑢 , 𝑢 and 𝑣 , 𝑣 , 𝑣
 𝑢⃗ 𝑢 𝑒⃗ 𝑢 𝑒⃗ 𝑢 𝑒⃗ , 𝑣⃗ 𝑣 𝑒⃗ 𝑣 𝑒⃗ 𝑣 𝑒⃗ ; then :

** 𝑢⃗. 𝑣⃗ ‖𝑢⃗‖. ‖𝑣⃗ ‖. cos 𝜃 𝑢 𝑣 𝑢 𝑣 𝑢 𝑣

** The angle  between the two vectors 𝑢⃗, 𝑣⃗ can be determined by:
𝑢⃗. 𝑣⃗
cos 𝜃
‖𝑢⃗‖. ‖𝑣⃗ ‖
Properties :
** 𝑢⃗. 𝑣⃗ 𝑣⃗. 𝑢⃗ (commutative),
** soit le vecteur 𝑤⃗ : 𝑢⃗. 𝑣⃗ 𝑤⃗ 𝑢⃗. 𝑣⃗ 𝑢⃗. 𝑤⃗ (distribution)
** soit le scalaire  : 𝛼 𝑢⃗. 𝑣⃗ 𝛼𝑢⃗. 𝑣⃗ 𝑢⃗. 𝛼𝑣⃗
** 0⃗. 𝑣⃗ 0,
** 𝑣⃗. 𝑣⃗ ‖𝑣⃗ ‖ ,

II.5 The cross or vector product: The result of the cross product of two vectors is a
vector perpendicular to the two vectors.

Let be the vectors 𝑢⃗, 𝑣⃗ and their successive components

𝑢 , 𝑢 , 𝑢 and 𝑣 , 𝑣 , 𝑣  𝑢⃗ 𝑢 𝑒⃗ 𝑢 𝑒⃗
𝑢 𝑒⃗ , 𝑣⃗ 𝑣 𝑒⃗ 𝑣 𝑒⃗ 𝑣 𝑒⃗ ; then :

** 𝑢⃗ 𝑣⃗ ‖𝑢⃗‖. ‖𝑣⃗ ‖. sin 𝜃 . 𝑛⃗

Course 1 Figure .3. page : 3

CONTINUUM MECHANICS Dr: Saib Ch.

Or :

𝑢⃗ 𝑣⃗ 𝑢 𝑣 𝑢 𝑣 𝑒⃗ 𝑢 𝑣 𝑢 𝑣 𝑒⃗ 𝑢 𝑣 𝑢 𝑣 𝑒⃗

Properties :
** 𝑢⃗ 𝑣⃗ 𝑣⃗ 𝑢⃗ ,
** 𝑢⃗ 𝑣⃗ 0⃗  the two vectors 𝑢⃗ and 𝑣⃗ are linearly dependent,
** let be the scalar  : 𝛼𝑢⃗ 𝑣⃗ 𝑢⃗ 𝛼𝑣⃗ 𝛼 𝑢⃗ 𝑣⃗ ,
** let be the vector 𝑤⃗ : 𝑢⃗. 𝑣⃗ 𝑤⃗ 𝑣⃗. 𝑤⃗ 𝑢⃗ 𝑤⃗. 𝑢⃗ 𝑣⃗ ,
** 𝑢⃗ 𝑣⃗ 𝑤⃗ 𝑢⃗ 𝑣⃗ 𝑢⃗ 𝑤⃗ 𝑢⃗ 𝑣⃗ 𝑢⃗ 𝑤⃗,
** 𝑢⃗ 𝑣⃗ 𝑤⃗ 𝑢⃗. 𝑤⃗ 𝑣⃗ 𝑣⃗. 𝑤⃗ 𝑢⃗,

III. Einstein’s “notation” summation convention:

In ordinary physical space a basis is composed of three axes, noncoplanar vectors, and so
any vector in this space is completely specified by its three components.
𝑂𝑃⃗ 𝑋 𝑒⃗ 𝑋 𝑒⃗ 𝑋 𝑒⃗

Can be written in the form: 𝑂𝑃⃗ ∑ 𝑋 𝑒⃗ ;

Where 𝑂𝑃⃗ is the position vector and 𝑋 are the components 𝑋 , 𝑋 , 𝑋 .
By convention we can write:
𝑂𝑃⃗ 𝑋 𝑒⃗ ,
with i as index and can take the values 1, 2, 3.
Accordingly the symbol 𝑋 is understood to represent the three components 𝑋 , 𝑋 , 𝑋 and
𝑒⃗ is understood to represent the three vectors 𝑒⃗ , 𝑒⃗ , 𝑒⃗ [3].

So, an index which occurs twice in an expression is summed over. Such an index is called
dummy. The rule is called the Einstein summation convention. So, now we write simply
[4]:

For a range of three on both indices, the symbol 𝐴 with i as dummy indice represents the
sum of 𝐴 𝐴 𝐴 [3].
The i-component of the vector 𝑂𝑃⃗ we write 𝑋 . This index is a free index. It is not
summed over. It is very important to distinguish between free indices and dummy indices.
 Look at the expression: 𝑎 𝑏 𝑐
Here, j occurs once only, so it is a free index. But “i” occurs twice. “i” must be a dummy
index and is summed over. So this expression has one free index.
The order of the factors can be changed arbitrarily: writing out the expression explicitly,
we have: 𝑎 𝑏 𝑐 𝑎𝑐𝑏 𝑎𝑐 𝑏 𝑎 𝑐 𝑎 𝑐 𝑎 𝑐 𝑏 [5].

Example expressions:
 𝐹 𝐴 𝐵 𝐶 , in this expression: i is a free indice however j is dummy one.
 𝐺 𝐻 2 3𝐴 𝐵 𝑃 𝑄 𝐹 : k is a free indice and i,j are dummy indices.
Course 1 page : 4
CONTINUUM MECHANICS Dr: Saib Ch.

IV. Kronecker-delta symbol:

Let us take the scalar product of the unit vectors of a basis 𝑂, 𝑒⃗ , 𝑒⃗ , 𝑒⃗ ), we conclude :
 𝑒⃗ . 𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ . 𝑒⃗ 0 ,
 𝑒⃗ . 𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ . 𝑒⃗ 1.
This allowed Kronecker, for simplification; to use the symbol that bears its name:
Kronecker delta symbol, which is defined:
1, 𝑖𝑓 𝑖 𝑗
𝛿 . and we can also write: 𝛿 𝑒⃗ . 𝑒⃗ ,with i,j = 1, 2, 3.
0, 𝑖𝑓 𝑖 𝑗

The Kronecker delta is sometimes called the substitution operator, since, for example;
𝛿 𝑏 𝛿 𝑏 𝛿 𝑏 𝛿 𝑏 𝑏
Properties:
 𝑢𝛿 𝑢,  𝑢𝑣𝛿 𝑢 𝑣 𝑢 𝑣 , 𝛿 𝛿 𝛿
The dot or the scalar product of two vectors 𝑢⃗, 𝑣⃗ can be written:
𝑢⃗. 𝑣⃗ 𝑢 𝑒⃗ . 𝑣 𝑒⃗ 𝑢 𝑣𝛿 𝑢 𝑣 𝑢 𝑣

V. Levi-Civita permutation symbol:

If we mix the vector products of the unit vectors with the scalar products, such that:
𝑒⃗ 𝑒⃗ . 𝑒⃗ , 𝑒⃗ 𝑒⃗ . 𝑒⃗ and so on
𝑒⃗ 𝑒⃗ . 𝑒⃗ , 𝑒⃗ 𝑒⃗ . 𝑒⃗ and so on
𝑒⃗ 𝑒⃗ . 𝑒⃗ , 𝑒⃗ 𝑒⃗ . 𝑒⃗ and so on, or write: 𝑒⃗ 𝑒⃗ . 𝑒⃗ with : i ,j , k =1, 2, 3.

We will find the results as:

𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 0
𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 0
𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 0
𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 1
𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 𝑒⃗ 𝑒⃗ . 𝑒⃗ 1

This allowed Levi-Civita to define its "epsilon" permutation symbol as follows:

𝑖𝑓 𝑖, 𝑗, 𝑘 𝑎𝑟𝑒 𝑎𝑛 𝑒𝑣𝑒𝑛 𝑝𝑒𝑟𝑚𝑢𝑡𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 1,2,3

⎧1
⎪ 𝑖. 𝑒. 𝑖𝑓 𝑡ℎ𝑒𝑦 𝑎𝑝𝑝𝑒𝑎𝑟 𝑖𝑛 𝑠𝑒𝑞𝑢𝑒𝑛𝑐𝑒 𝑎𝑠 𝑖𝑛 𝑡ℎ𝑒 𝑎𝑟𝑟𝑎𝑛𝑔𝑒𝑚𝑒𝑛𝑡 12312 .
⎪
𝑖𝑓 𝑖, 𝑗, 𝑘 𝑎𝑟𝑒 𝑎𝑛 𝑜𝑑𝑑 𝑝𝑒𝑟𝑚𝑢𝑡𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 1,2,3
𝜀 1
⎨ 𝑖. 𝑒. 𝑖𝑓 𝑡ℎ𝑒𝑦 𝑎𝑝𝑝𝑒𝑎𝑟 𝑖𝑛 𝑠𝑒𝑞𝑢𝑒𝑛𝑐𝑒 𝑎𝑠 𝑖𝑛 𝑡ℎ𝑒 𝑎𝑟𝑟𝑎𝑛𝑔𝑒𝑚𝑒𝑛𝑡 32132 .
⎪
⎪ 𝑖𝑓 𝑖, 𝑗, 𝑘 𝑎𝑟𝑒 𝑛𝑜𝑡 𝑎 𝑝𝑒𝑟𝑚𝑢𝑡𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 1,2,3
0
⎩ 𝑖. 𝑒. 𝑖𝑓 𝑡𝑤𝑜 𝑜𝑟 𝑚𝑜𝑟𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑖𝑛𝑑𝑖𝑐𝑒𝑠 ℎ𝑎𝑣𝑒 𝑡ℎ𝑒 𝑠𝑎𝑚𝑒 𝑣𝑎𝑙𝑢𝑒 .

Course 1 page : 5
CONTINUUM MECHANICS Dr: Saib Ch.

Or write: 𝜀 𝑒⃗ 𝑒⃗ . 𝑒⃗ with: i, j, k =1, 2, 3.

It represents the orientation of the coordinate system as
shown in Figure.4.

Properties: Figure .4.

 𝜀 𝜀 𝜀 ;
 𝜀 𝜀 𝜀 𝜀 ,
The cross or the vector product of two vectors 𝑢⃗, 𝑣⃗ can be written:
𝑢⃗ 𝑣⃗ 𝑢 𝑒⃗ 𝑣 𝑒⃗ 𝑢 𝑣 𝜀 𝑒⃗

Relationship between permutation symbol and Konecker-delta:

𝛿 𝛿 𝛿 𝛿 𝛿 𝛿
𝜀 det 𝛿 𝛿 𝛿 , et : 𝜀 𝜀 det 𝛿 𝛿 𝛿
𝛿 𝛿 𝛿 𝛿 𝛿 𝛿

VI. Basis Change - Transformation law:

Let the two orthonormal (rectangular) Cartesian
coordinate systems 𝑂, 𝑒⃗ , 𝑒⃗ , 𝑒⃗ ) and 𝑂′, 𝑒⃗′ , 𝑒⃗′ , 𝑒⃗′ with
common origin at an arbitrary point O as shown in Figure
.5.
The primed system may be imagined to be obtained from
the unprimed by a rotation of the axes about the origin, or
by a reflection of axes in one of the coordinate planes, or
by a combination of these. If the symbol 𝑎 denotes the Figure .5.
cosine of the angle between the ith primed and jth unprimed
coordinate axes,
i.e. 𝑎 𝑒⃗ . 𝑒′⃗ cos 𝑒⃗ , 𝑒⃗′ cos 𝜃 , Figure .6., the
relative orientation of the individual axes of each system
with respect to the other is conveniently given by the table :

𝑒⃗ 𝑒⃗ 𝑒⃗
𝑒′⃗ 𝑎 𝑎 𝑎
𝑒′⃗ 𝑎 𝑎 𝑎
Figure .6.
𝑒′⃗ 𝑎 𝑎 𝑎

Course 1 page : 6
CONTINUUM MECHANICS Dr: Saib Ch.

Or alternatively by the transformation tensor:

𝑎 𝑎 𝑎
𝑎 𝑎 𝑎 𝑎
𝑎 𝑎 𝑎

From this definition the unit vectors 𝑒′⃗ along the 𝑋′ axis is given by:
𝑒′⃗ 𝑎 𝑒⃗ 𝑎 𝑒⃗ 𝑎 𝑒⃗
An obvious generalization of this equation gives the arbitrary unit base vector 𝑒′⃗ as:
𝑒′⃗ 𝑎 𝑒⃗
In component form, the arbitrary vector 𝑣⃗ may be expressed in the unprimed system by the
equation: 𝑣⃗ 𝑣 𝑒⃗ and in the primed system by: 𝑣⃗ 𝑣′ 𝑒′⃗ 𝑣′ 𝑎 𝑒⃗ by comparison
we can conclude that the vector components in the primed and unprimed systems are
related by the equations: 𝑣 𝑎 𝑣′
This expression is the transformation law for first- order Cartesian tensors, by
interchanging the roles of the primed and unprimed base vectors in the above
development, the inverse is found to be: 𝑣′ 𝑎 𝑣
It is important to note that the free index on 𝑎 appears as the second index in the second
equation, the free index appear as the first index.
By an appropriate choice of dummy indices may be combined to produce: 𝑣 𝑎 𝑎 𝑣
Since vector 𝑣⃗ is arbitrary, which allow to write: 𝑎 𝑎 𝛿 , this last equation is known
as the orthogonality or orthonormality conditions on the direction cosines 𝑎 .
According to the transformation law, the unprimed vector 𝑢′𝑣′⃗ which has components in
the primed coordinate system given by:
𝑢′ 𝑣′ 𝑎 𝑢 𝑎 𝑣 𝑎 𝑎 𝑢 𝑣
In an obvious generalization, any second-order Cartesian tensor 𝑇 obeys to the
transformation law: 𝑇′ 𝑎 𝑎 𝑇 ;
With the help of the orthogonality conditions it is a simple calculation to invert:
𝑇 𝑎 𝑎 𝑇′ .
Working with tensors requires knowledge of matrix operations.

VII. Matrices, Matrix representation of Cartesian tensors:

A rectangular array of elements, enclosed by square brackets and subject to certain
laws of combination, is called matrix.
An MN matrix is one having M (horizontal) rows and N (vertical) columns of
elements. In the symbol 𝐴 used to represent the typical element of a matrix, the first
subscript denotes the row, the second subscript the column occupied by the element, said
of two dimensions (M, N) noted for example:

Course 1 page : 7
CONTINUUM MECHANICS Dr: Saib Ch.

𝑎 𝑎 𝑎
𝐴 𝑎 𝑎 𝑎 𝑎 , with: i=1, 2, 3 and j=1, 2, 3.
𝑎 𝑎 𝑎
M=3 (three rows), N=3 (three columns), number of array elements: MN=3X3=9 scalar
elements.
• Generally, in Continuum mechanics, M=N, in this case the matrices are said to be
“square”.

Matrix operations:
Let the square matrices of the same dimensions be: A, B, C,
 𝐴 𝐵 𝐵 𝐴 ; commutative addition,
 𝐴 𝐵 𝐶 𝐴 𝐵 𝐶 ; associative addition,
 𝐴 0 0 𝐴 𝐴 , the neutral matrix 0 is a matrix where the elements are null
(known as zero elements).
 𝐴 𝐴 0, for each matrix A, there is a unique matrix –A.
 Let be the scalar  then: 𝛼 𝐴 𝐵 𝛼𝐴 𝛼𝐵,
 𝐴 𝐵 𝐶 𝐴𝐶 𝐵𝐶 : distribution on the addition,
 Transposed from a matrix noted AT : 𝐴 𝑎 𝐴 𝑎 , with :
𝐴 𝐴.
Properties:
 𝐴 𝐵 𝐴 𝐵 and  𝐴𝐵 𝐴 𝐵 .

Matrix Inverse: if a square matrix is non-singular, it possesses a unique inverse matrix.

A square matrix whose determinant is zero is called a singular matrix.

The inverse matrix is a matrix denoted by 𝐴 , such that:

𝐴𝐴 𝐴 𝐴 𝐼
Where I is the “identity” unit matrix, having ones on the principal diagonal elements and
zeros elsewhere.
Properties:
 𝐴 𝐴 and 𝐴𝐵 𝐴 𝐵 ,
• The matrix inverse can be calculated using the cofactor method:
1
𝐴 . 𝑐𝑜𝑚 𝐴
det 𝐴
𝑐𝑜𝑚 𝐴 1 det 𝐴 ,
If A is a matrix which size is N then 𝐴 , is the square sub-matrix of size N–1 deduced
from A by deleting the i-th row and from the j-th column (its determinant is therefore one
of the minors of A).

Course 1 page : 8
CONTINUUM MECHANICS Dr: Saib Ch.

Matrix determinant: is a scalar given by:

𝑑𝑒𝑡 𝐴 𝜀 𝐴 𝐴 𝐴
𝑑𝑒𝑡 𝐴 𝜀 𝜀 𝐴 𝐴 𝐴 [6]
Principal values and principal directions of matrix or second-order tensors:
Let be the matrix 𝐴 𝑎 , there is associated with each direction(specified by the
𝑛
𝑛
unit normal 𝑛 ) an arbitrary vector 𝑛⃗ , a vector given by inner product 𝑣⃗ where its
⋮
𝑛
components will be 𝑣 𝑎 𝑛 .
Here 𝑎 may be envisioned as a linear vector operator which produces the vector 𝑣
conjugate to the direction 𝑛 . If the direction is one for which 𝑣 is parallel to 𝑛 , the inner
product may be expressed as a scalar multiple of 𝑛 . For this case,
𝐴𝑛⃗  𝑛⃗
And the direction 𝑛 is called a principal direction, or principal axis of 𝑎 . With the help
of the identity 𝑛 𝛿 𝑛 can be put in the form:
𝐴 δ 𝑛⃗ 𝑂⃗ or 𝑎 𝛿 𝑛 0.
Which represents a system of three equations for four unknown 𝑛 and , associated
with each principal direction.
Note first that for every , the trivial solution 𝑛 0 satisfies the equations. The
purpose here, however, is to obtain non-trivial solutions. Also, for the homogeneity of the
system it follows that no loss of generally is incurred by restricting attention to solutions
for which:
𝑛𝑛 1,
and this condition is imposed from now on.
To have a non trivial solution, the determinant of coefficients must be zero, that is;
𝑎 𝛿 𝑛 0
Expansion of this determinant leads to a cubic polynomial in , namely;
 𝐼 𝐼𝐼 -III 0
Which is known as the characteristic equation of 𝑎 , and for which the scalar
coefficients,
𝐼 𝑎 𝑡𝑟 𝑎 (trace of a )
𝐼𝐼 1 2 𝑎 𝑎 𝑎 𝑎

𝐼𝐼𝐼 |𝑎 | 𝑑𝑒𝑡 𝑎

Are called the first, second and third invariants, respectively of 𝑎 . The three roots of the
cubic characteristic equation, labeled  ,  ,  , are called the principal values of 𝑎 .
For a symmetric tensor with real components, the principal values are real; and if these
values are distinct, the three principal directions are mutually orthogonal.
When referred to principal axes, both the tensor appears diagonal form as:

Course 1 page : 9
CONTINUUM MECHANICS Dr: Saib Ch.
 0 0
𝐴∗ 𝑎∗ 0  0
0 0 
The principal values are ordered as in    [5].
𝑛
𝑛
An eigenvector 𝑛∗⃗ of components associated with the eigenvalue  must satisfy the
⋮
𝑛
relation:
𝑛
𝑛
𝐴 I 𝑛∗⃗ 𝑂⃗  𝐴 I 𝑂⃗
⋮
𝑛
If the matrix A admits  eigenvalues, distinct two by two, the 𝑛∗⃗ associated
eigenvectors are linearly independent.
For principal axes labeled 𝑂, 𝑋 ∗ , 𝑋 ∗ , 𝑋 ∗ ) the transformation from 𝑂, 𝑋 , 𝑋 , 𝑋 ) axes is
given by the elements of the table:
𝑋 𝑋 𝑋
∗
𝑋 𝑎 𝑛 𝑎 𝑛 𝑎 𝑛
∗
𝑋 𝑎 𝑛 𝑎 𝑛 𝑎 𝑛
𝑋∗ 𝑎 𝑛 𝑎 𝑛 𝑎 𝑛
In which 𝑛 are the direction cosines of the jth principal direction.

VIII. Derivative of tensors:

1/ The temporal differentiation of a vector is given by:
𝜕𝑣⃗ 𝜕𝑣 𝜕𝑣 𝜕𝑣
𝑒⃗ 𝑒⃗ 𝑒⃗ 𝑣 , 𝑒⃗ 𝑣 𝑒⃗
𝜕𝑡 𝜕𝑡 𝜕𝑡 𝜕𝑡
Properties:
⃗  ⃗
 𝐴⃗  𝐴⃗ 𝐴⃗
⃗. ⃗ ⃗ ⃗
 . 𝐵⃗ 𝐴⃗. 𝐴⃗. 𝐵⃗ 𝐴⃗. 𝐵⃗
⃗ ⃗ ⃗ ⃗
 𝐵⃗ 𝐴⃗ 𝐴⃗ 𝐵⃗ 𝐴⃗ 𝐵⃗
2/ the spatial differentiation of a vector:
⃗
 𝑒⃗ 𝑣 , 𝑒⃗
⃗
 𝑒⃗ 𝑣 , 𝑒⃗

 𝑣,
3/ the spatial differentiation of a tensor:
 𝑎 ,

 𝑎 ,

Course 1 page : 10
CONTINUUM MECHANICS Dr: Saib Ch.

Several important differential operators appear in continuum mechanics and are given here
for reference.
The divergence: the divergence of a vector is a scalar given by the sum of the partial
derivatives of its components in the three directions of space:
 𝑑𝑖𝑣 𝑣⃗ ∇𝑣⃗ or : 𝜕 𝑣 𝑣,
The divergence of a tensor will produce a vector of coordinates:
 𝑑𝚤𝑣⃗ 𝑎 or : 𝜕 𝑎 𝑎 ,
The gradient: it includes the concept of "slope" which exists in two dimensions. It
describes the variation of a quantity according to the three directions of space.
 𝑔𝑟𝑎𝑑⃗ 𝑓 𝑒⃗ 𝑒⃗ 𝑒⃗ 𝑒⃗ 𝑓, 𝑒⃗
We define the gradient operator applied to a vector. It is about a tensor of order 2, whose
components are:
 𝑔𝑟𝑎𝑑⃗ 𝑈⃗ 𝑈,
The Laplacian: noted, is the sum of the second derivatives in the three directions of
space.it preserves the tensorial order of which it is applied, because it is the composition of
a divergence and a gradient, applied to a scalar, we therefore have:
 ∆𝑓 𝑑𝑖𝑣 𝑔𝑟𝑎𝑑⃗ 𝑓

 ∆𝑈⃗ 𝑑𝚤𝑣⃗ 𝑔𝑟𝑎𝑑⃗ 𝑈⃗ 𝑈⃗,

IX. Gauss and stokes integral theorems:
Integral transformation formulas are mathematical tools used to match an integral
over D (volume integral) to an integral over S (surface integral).

Figure .7.

They are given here without proof, applied to a scalar, vector or to the tensor:
 𝑔𝑟𝑎𝑑⃗ 𝑓 𝑑𝑣 𝑓𝑛⃗𝑑𝑆
 𝑑𝑖𝑣 𝑈⃗ 𝑑𝑣 𝑈⃗. 𝑛⃗𝑑𝑆
 𝑑𝚤𝑣⃗ 𝑇 𝑑𝑣 𝑇𝑛⃗𝑑𝑆

Course 1 page : 11
CONTINUUM MECHANICS Dr: Saib Ch.
These formulas have a computational interest, but have a strong meaning when they
present physical quantities. They make it possible to write the behavior of a quantity on
a domain from its behavior on the surface [2].

Bibliographic references:

[1] Samir DEGHBOUDJ,2016, POLYCOPIE DE COURS « MECANIQUE DES MILIEUX

CONTINUS COURS ET APPLICATIONS ».
[2] Slim KAMMOUN,2020-2021,Plan de la MMC, Ecole nationale d’ingénieurs de Sfax, Tunis
[3] Nicholas Moés, 2011. Mécanique des milieux continus, Ecole centrale de Nantes.
[4] Lars H. Sӧderholm, 2008. BASIC CONTINUUM MECHANICS, KTH Royal Institute of
Technology, Sweden
[5] François Sidoroff,2010. Mécanique des milieux continus, Ecole centrale de Lyon.
[6] http://uel.unisciel.fr/physique/outils_nancy/outils_nancy_ch11/co/apprendre_ch11_20.html

Course 1 page : 12