Tensor and Tensor
Algebra
by
Dr. Tariq JamilReferences
+ Ottosen, N.S. and Ristinmaa, M., 2005. The aii FI li
Mechanics of Constitutive Modeling. Elsevier. Pelee ing hI
+ https://en.wikipedia.org/wiki/TensorTensor
+ Tensors, defined mathematically, are simply arrays of numbers, or functions, that transform according to certain rules
under a change of coordinates.
+ Zeroth-Order Tensors (Scalars) e.g. pressure of gas =P
+ Asealar is a single function (e., one component) which is invariant under changes of the coordinate systems.
‘Such quaity needs only one number to define, that has no particular arrangement to write. Thus no suffix is
needed to locate that single number. These quantity are called as zero order tensor or zero dimensional array.
f,
+ First-Order Tensor (Vectors)e.g. Force F= (FF, F.]
+ Second-Order Tensor e.g. Stress Oy, 1213) [Fxx Oxy xz] [Oxx Oxy xz
Oj =|F21 F22 F2al=|Pyx yy Syz] =| %yx Syy yz
31 932 833) LPzx F2y Fez. Tye zy OzzIndex Notation Rules and Coordinate Rotation
2 oan
yo
lA
Vv
igre 81 Rotation of cota anes
+ Key to classifying scalars, vectors, or tensors is how their components change if the
coordinate axes are rotated to point in new directions
+ Ascalar (temperature, density, etc.) is unchanged by rotation ~ it has the same value
in any coordinate system, which is @ defining characteristic of a scalar
+ Vectors and tensors change with rotating coordinate systemTensor
+ Index notation, also commonly known as subscript notation or tensor notation, is an extremely useful
tool for performing vector algebra.
+ Tensors were conceived in 1900 by Tullio Levi-Civita and Gregorio Ricci-Curbastro,
+ Tensor is the mathematical tool used to express physical quantities
+ In mathematics, a tensor is an algebraic object that describes a (multilinear) relationship between sets
of algebraic objects related to a vector space.
+ Tensors are important in physics because they provide a concise mathematical framework for
formulating and solving physics problems in areas such as mechanics (stress, elasticity, fluid
mechanics, moment of inertia, ...), electrodynamics (electromagnetic tensor, Maxwell tensor,
permittivity, magnetic susceptibility, ..), or general relativity (stress-energy tensor, curvature tensor,
and others.
)Continue...
* Order/rank of tensor depends on how many separate entities are necessary to describe the
matrix and varies according to 3°.
+3054 scalar
+3153 vector (F =F,i + Fy + Fk)
+ This terminology is suggestive. Why stop at rank 1? Why not go onto rank 2, rank 3, and so
on.
+ Dyad: Tensor of rank 2. (magnitude and two directions — 3? = 9 components) e.g. Stress
+ Triad: Tensor of rank 3. (magnitude and three directions — 3° = 27 components)
+ Etcetera...Continue...
+ Tensors are the notations which are used to represent a scalar, vector or matrixes. (may have
columns, rows and tubes; 3-way tensors have 3 modes NxMxK dimensions)
+ Equations governing 3D mechanics problems can be quite lengthy, short hand notation called index
notations.
# DG EK, xt x
+ Algebric expression > Pi tsisn
+ In tensors above expression can be written as simply > ax, 1sisn
+ (if the limits of index is not defined, than by default it would be considered as n=1,2,3
G4, Xj =AYXy HANSA,TENSOR
+ Tensors are often displayed as a matrix
+ The transpose of a tensor is obtained by
interchanging the two indicies, so the
transpose of T, is T,
+ Tensor Q; is symmetric if Q; = Q;
+ Tensor is antisymmetric if it is equal to the
negative of its transpose, Rj = -Ry
+ Any arbitrary tensor T, may be
decomposed into sum of a symmetric
tensor and antisymmetric tensorIndex notation
+ There are two types of index
+ Free
+ Dummy
+ A free index occurs once and only once in each and every term in an equation
+ A dummy or summation index occurs twice in a term
ie ajx, > ‘iis free index and ‘j’ is a dummy indexSummation Convention
+ An important convention is index notation is the so-called summation convention,
which states that if an index is repeated twice then a summation over this index is
implied. An example, the produce bja; , where the index i is repeated twice means
b,aj= b,a,+b2a,+b303Another way to write a vector is in Cartesian form
K=xi+yj+zk = (x,y,z)
The coordinates x, y and z can also be written as x,, x2, Xs. Thus the
vector can be written as
X= (X,XqXa)
oras
X =(x),i=1.3
or in index notation, simply as
where i is understood to be a dummy variable running from 1 to 3
Thus x, x, and x, all refer to the same vector (x,, Xp and Xs) , as the
index (subscript) always runs from 1 to 3ay
a,
‘+ In matrix notations we are restricted to work with one- and two-dimensional arrays. This is not
the case in index notation, for instance ,
@, = 3x3x3 = 27 components (3 order tensor)
y= 81 Components (4 order tensor)
By By Bis
B=[Byl= |B21 Bz2 Bos
Bay Baz BasExample
Expand yi = aix,
Ya = By1X1 + AyaXot ay3X3
Ya = GaiXy + AgaXgt Az3X3
Ya = @giX1 + @gaXgt Ag3X3Examples
Expand ax,
BIR FAX HAN FAN
Expand ajx, (1sj <3)Scalar multiplication: let« be ascalarand A =A, bea vector.
Then 7
oA = aA, = (a.A,,0:A,,0A5)
is a vector.
Dot or scalar productof two vectors results in a scalar:
AeB=AB,+A,B, +A,B, = scalar
In index notation, the dot product takes the form
BoB AB-TAB,-DAB, -
Einstein summation convention: if the same index occurs twice, always
sum over that index. So we abbreviate to
AeB=AB,=A,B, = AB,
There is no free index in the above expressions. Instead the indices are
paired (e.g. two i's), implying summation. The result of the dot product
is thus a scalar.Magnitude of a vector:
Atensor can be constructed by multiplying two vectors (not scalar
product):
AB, A,B, A.B,
AB, A,B, A,B,
AB, AB, A,B,
Two free indices (i, j) means the result is a second-order tensor
Now consider the expression
AAB,
This is a first-order tensor, or vector because there is only one free
index, i (the j's are paired, implying summation).
AAB; = (A.B, + A,B, + 3B, )(A,,A2,As)
That is, scalar times vector = vector.
AB, =(AB,),iKronecker delta 5,
(1 0 0)
{ # i=i_|y 4 9 — ;
Of 14] |) g 4) 827 8a =Bar a= Br
+ Since there are two free indices, the result is a second-order tensor,
or matrix. The Kronecker delta corresponds to the identity matrix.
+ Itis used in tensor calculus to annihilate-non-diagonal terms.
Example:
Expand xx
GK = Bap # ya iXy + DyaMiXa + OaiKaXy + OarKoXp + OzaXoXs + Oy iXaX + OaaXaXp + OyaXaXa
5 (1) x24 (1) x2 (1) xe?
CPt Kt A
= HK OF KK,A.useful manipulation in tensor notation can be used to change an index
in an expression:
Su, =u,
This manipulation works because the Kronecker delta 6, = 0 except when
i=j, in which case it equals 1Example
+ Expand and approximate 5,a,
ya, = Syaq+ Bi2aq + Gysas + B21 + 52a + By33 + 5514) + O3242 + Os3as
dyaj= ai tay +as
Problem 1.6: Solve the following equations:
1) 5,6y
Solution: 548 y = (6. + 8xy +533 Ny, + 6m, +533) =3X3=9
2) baibar5n
Solution: — Sab 51 = 651 = 5 =1
NOTE: Note that the following algebraic operation is incorrect 5,5, #6,, =3 #54 =1,
since what must be replaced is the repeated index, not the number aExpand
i= 511 + 5 a0 + Bag= 3
(= 1,2, 3)Practice Problem
Expand and proof By, = Bx
Expand 6) +Fi=0
Example
Recall the strain-splacement relations, Eqns. 1.2.19, which in fll eadPractice Problems
‘Simply he expressions 5,5,5,,5,. Solution:
{iD)KBkpopi = BIOKBKi = 6 B= Oi = 9.
Write Equetion for directional cosines n tensor notations,
costa + costa + costy
Solution:
Let cosa=axscose
Hence a,0;+ azasray
Solve the equations, a, = C,), b = oD,
Solution:
Let =k 6D, > 4, = Cy4.DePractice Problems
‘Problem It: Rewrite the fllonring eqtions sing india notation:
Dag sane rats,
Suto: az, i=123)
er)
b
»,
fs ays=b,
Sabo
founsecseeeueh ae (ash
Vig segntenn