HOMEWORK

Calculate Young’s modulus

for Al and Cu in the <111>,
<100>, and <110> directions.

Module #5

Origins of elasticity
Elastic constants in isotropic materials
Elastic constants in anisotropic materials

READING LIST
DIETER: Ch. 2, Pages 47-60

Pages 21-36 in Hosford ; Ch. 2 in Roesler ;

Ch. 8 in Nye ; Ch. 3 in McClintock and Argon
Ch. 7 in Sines ; Ch. 6 in Wagoner and Chenot
Ch. 11 in Sadd
 Elasticity
• Important because most engineering design is done
in the elastic region. (Remember, plastic deformation
generally constitutes failure.)

• Macroscopically, most polycrystalline materials are

elastically isotropic.

• Microscopically, elastic behavior is inherently

anisotropic for individual grains.

• Polycrystalline materials can be anisotropic if they

exhibit strong crystallographic textures.
 Elasticity – cont’d
• Derived from atomic U Repulsion

bonding forces originating

from long range attractive
ro
forces which draw atoms r
together until short range
repulsive forces become
large enough to balance
them out. Attraction
F is maximum at point
of inflection in U/r
• As such, elastic properties F
curve

are an aggregate effect of F = -dU/dr

individual deformations of ro
r
interatomic bonds.
 Bulk Elastic Behavior
• Applied force is transmitted by the network of bonds
constituting the material.

• Thus, elastic behavior depends quantitatively on the

magnitude of the interatomic forces.

• Elastic properties do not depend on microstructure of

the material.
 Forces Between Atoms
• Bonds form when atoms either share (e.g., covalent
and metallic bonds) or transfer (e.g., ionic bonds)
electrons between atoms.

• This occurs to achieve a minimum in potential energy

for the system corresponding to a stable electron
configuration.

• This has been modeled for ionic systems as follows:

 Ionic Bond
• Electron transfer from one
atom to another occurs due
to electrostatic interaction
between oppositely charged r
ions (e.g., Na+ and Cl- in
NaCl).
+ -
• Force of attraction between
ions is Fattract q2
Fattract 
4 o r 2
• Work is done to draw ions q  charge of ion
together. It is given by:
  permittivity of free space
o

r r  ionic separation distance

q 2 A
U attract   Fattract dr   m

4 o r r
• Each atom maintains an electron cloud surrounding it. The outer
surfaces of each atom are both negatively charged.

• Thus, at close distances, the electron clouds overlap and repel.

The repulsive force can be written as:
B
U repel 
rn

• The total energy of the system is thus the sum of the attractive and
repulsive components. It can be expressed in general as:

A B
U total  U attract  U repel  m
 n
r r

• This is the well known Lennard-Jones potential. You might recall

this from your chemistry or physics courses.
U Repulsion
• The minimum energy point
corresponds to the equilibrium
separation (i.e., the equilibrium
bond length, ro).
ro
r
• The force between atoms is
simply the derivative (i.e., the
slope) of the energy versus
distance plot.
dU total
Attraction F
dr

F • The bond stiffness is the

derivative of the force versus
distance plot.
ro
r dF d 2U total
S 
dr dr 2
• At the equilibrium bond length, ro, the variation of F with r is essentially
linear, which means that the stiffness is essentially constant at small
distances from ro.
dF
So 
dr ro

• Using this expression, the force to “stretch” n bonds in a solid is:

r
F  n  So dr  nSo  r  ro 
ro

where n is the number of bonds.


• The applied stress is:

F nSo  r  ro 
ro  
A nro2

ro ro
• The strain can be expressed as: The derived expression
r  ro shows that the modulus is

ro dependent upon bonding.

• Thus, stress becomes:

Modulus, and other elastic
F nSo  r  ro  So  r  ro  So
  2
   properties, are
A nro ro ro ro
structure insensitive.

• From Hooke’s law:

  E The directionality of elastic
properties does depend
• Thus, upon atomic arrangement
So 1 d 2U (crystallography) in single
E 
ro ro dr 2 r  ro
crystals. They tend to
average out in polycrystals.
 Bonding Energy
Material, eV/Atom, Ion, Melting
Bonding Type Substance kJ/mol Molecule Temperature °C
Ionic NaCl 640 3.3 801
MgO 1000 5.2 2800

Covalent Si 450 4.7 1410

C (diamond) 713 7.4 >3550

Metallic Hg 68 0.7 -39

Al 324 3.4 660
Fe 406 4.2 1538
W 849 8.8 3410

van der Waals Ar 7.7 0.08 -189

Cl2 31 0.32 -101

Hydrogen NH3 35 0.36 -78

H2O 51 0.52 0

[From W.D. Callister, Jr, Materials Science and Engineering: An Introduction, 7th Ed., (Wiley, New York, 2007) p.28]
 Metallic Covalent or
bonding Ionic bonding

Steel WC

Young’s modulus, GPa Copper

CFRP

Alumina

Aluminum GFRP
PEEK
Zinc Glass Fibreboard
PP
Lead
PTFE

Metals Polymers Ceramics Composites

Courtesy of Granta Design

A bar chart illustrating Young’s moduli for some common examples of the primary
classes of materials (i.e., metals, ceramics, polymers and composites). Ceramics
tend to exhibit the highest elastic moduli and polymers the lowest. The elastic
moduli of composites are between those of their constituents.
 stress is related to strain for an isotropic solid

• Hooke’s Law
Unload
– Isotropic Solids (properties
are the same in all
directions)
Stress

Load
Slope = E (Modulus of elasticity)

1
   E or   
E
0
0 Strain – Anisotropic solids
(properties are directional)
Stress-strain plot for a linearly elastic material

Need different definition

– Let’s consider both cases

 First consider the Poisson effect
F

x y
unloaded
F

loaded
x 
 ( z , x)   and  ( z , y )   y
z z

A tensile stress along the z axis causes the material to

stretch along the z axis and to contract along the x and
y axes.
lateral strain
Poisson's Ratio,   
longitudinal strain
 Material Class Material 
Room- Metallic solid Ag 0.38
Temperature (crystalline) Al 0.34
Poisson’s Ratio Au 0.42
for Selected Cu 0.34
Solids -Fe 0.29
Ir 0.26
Ni 0.31
W 0.29
Covalent solid Ge 0.28
(crystalline) Si 0.27
Al2O3 0.23
TiC 0.19
Covalent Ionic Solid MgO 0.19
Covalent Glass Silica glass 0.20
Network Polymer Bakelite 0.20
Ebonite (hard rubber) 0.39
Elastomer Natural rubber 0.49
Chain Polymer Polystyrene 0.33
Polyethylene 0.40
Van der Waal’s Solid Argon (at 0 K) 0.25
 A B C
 yy
 yy

 xx  xx
 xx  xx

 yy
 yy
The strains resulting from this stress state must be the sum of the strains
associated with loading along each individual axis.
 xx  xx
B :  xxB  ,  yyB   xxB  
E E
 yy  yy
C:  C
yy  ,   
C
xx
C
yy  
E E
 xx  yy There’s a similar
A  B+C :  xx      B
xx
C
xx  expression for εyy
E E
 RECALL: Elastic Stress-Strain Relations
Strain in the Strain in the Strain in the
Stress x-direction y-direction z-direction

 xx  xx  xx
 xx  xx   yy    zz  
E E E
 yy  yy  yy
 yy  xx    yy   zz  
E E E

 zz  zz  zz
 zz  xx    yy    zz 
E E E

1
 xx   xx    yy   zz  
E 
Variation in
1
 yy   yy    zz   xx   elastic strain for
E an isotropic solid
1
 zz   zz    xx   yy   because of the
E Poisson effect
 xy  G xy ;  yz  G yz ;  xz  G xz
Other Elastic Stress-Strain Relations for Isotropic Solids
• Bulk Modulus (K), also known as the volumetric elastic
modulus:

hydrostatic pressure m  p 1
K   
volume change produced   B

where:
- p  hydrostatic pressure,
 xx   yy   zz
m 
3
  dilatation (i.e., volume change)   xx   yy   zz ,
B  compressibility.
Other Elastic Stress-Strain Relations for Isotropic Solids
• Now we can relate the elastic constants.
1
 xx   xx   yy   zz  
E
1
 yy   yy   zz   xx  
E
1
 zz   zz   xx   yy  
E

Summation yields:
1  2 1  2
 xx   yy   zz 
E
  xx   yy   zz  
E
3 m  


m E
K 
 3 1  2 
Other Relations Between Isotropic Elastic Constants

In terms
Elastic
of
E, ν E, G K, ν E, G λ, μ
constants

9K   3  2 /  
E [elastic] =E =E = 3 1  2  K = =
1  3K / G 2 1   /  
E 1  2G / 3K 1
ν =ν = 1  =ν = =
2G 2  2G / 3K 2 1   /  
E 3 1  2  K
G [shear] = =G = =G =μ
2 1    2 1   
E E 2
K [bulk] = = =K =K = 
3 1  2  9  3E / G 3
E E 1  2G / E  3K 2G
λ = = = = K =λ
1   1  2  3 E / G 1  3
E 3 1  2  K
μ = =G = =G =μ
2 1    2 1   

λ, μ = Lame constants
 STRAIN ENERGY
Energy stored in a body due to deformation.
• Work to deform a body elastically is stored as elastic strain
energy. It is recovered when the applied forces are released.

• Strain energy is proportional to the area under the load-

deformation (stress-strain) curve.

• We will elaborate on this

concept on the next Load F
few pages. 1
U F
2

Deformation δ
 Strain Energy
• The first law of thermodynamics states:
dU   Q   W  TdS  PdV
where,

dU = internal energy
δQ = heat
δW = mechanical work done

• During elastic deformation, the amount of heat generated

is negligible.

• Thus the work done on the body is converted into internal

energy and is fully recovered upon unloading.
 ELASTIC STRAIN ENERGY – cont’d
• Consider an elemental cube that is subjected to only an elastic
tensile stress along the x-axis. The elastic strain energy, U, is:
1
U  (load)(deformation)
z 2
y
1 1 1
x dU  Fdu   xx A   xx dx    xx xx   Adx 
2 2 2
 total elastic energy absorbed by the material
A
element.

 xx Since Adx is the volume, the elastic strain energy per

F
unit volume (i.e., the strain energy density) is:
 xx 
A
du
 xx  dU 1
dx Uo    xx xx  .
Adx 2
 ELASTIC STRAIN ENERGY – cont’d
 xx
From Hooke's law,  xx = .
E
Therefore:
z

1  xx2 1 2
  xx E [in tension]
y
Uo 
x 2 E 2

1  xy 1 2
2
1
U o   xy xy     xy G [in pure shear]
A
2 2G 2
 xx
In three dimensions, the general expression for elastic
 xx 
F strain energy is:
A
du 1
 xx 
dx U o   xx xx   yy yy   zz  zz   xy xy   xz  xz   yz  yz  .
2
 ELASTIC STRAIN ENERGY – cont’d

In tensor notation, these general expressions become:

1  ij 1 2
2
1
U o   ij  ij    ij E.
2 2 E 2
If we substite into these expressions, the variation in elastic strain
for an isotropic solid, accounting for the Poisson effect, we get:
1  1
Uo 
2E
    
2
xx
2
yy
2
zz  E
 xx yy   yy zz   xx zz   2G
 xy xz yz 
 2
  2
  2

[for a 3D isotropic solid]

Expressed in terms of strains and elastic constants:

1 2 1
U o    G   xx   yy   zz   G   xy2   xz2   yz2 
2 2 2

2 2
 ELASTIC STRAIN ENERGY – cont’d
1
Uo 
2
  xx xx   yy yy   zz zz   xy xy   xz xz   yz yz 

The derivative of U o with respect to any strain component yields the

corresponding stress component, or:
U o
   2G ij   ij
 ij

Similarly, the derivative of U o with respect to any stress component

yields the corresponding strain or:

U o
  ij
 ij

You can use these terms to calculate stresses and strains in elasticty.
Techniques for doing this include Castigliano's theorem, the theorem
of least work, and the principal of virtual work.
 Anisotropy and Single Crystals
• Forces between atoms are directional.

(110)BCC

• They act along the “lines” connecting the atoms together.

 Relations between stress and strain for
anisotropic crystals
• Since force varies with direction, elastic constants will
also vary.

• In general, we need to relate every stress component to

every strain component.

• Thus we must define two new elastic constants:

– C  stiffness
– S  compliance

• We also re-write Hooke’s law as:

 ij  Cijkl  kl or  ij  Sijkl kl
 3 3
Stiffness is a  ij   Cijkl  kl
4th rank tensor k 1 l 1


34 = 81 terms
 ij  Cijkl  kl

  xx   Cxxxx Cxxyy C xxzz C xxyz C xxzx Cxxxy Cxxzy Cxxxz C xxyx    xx 

   C C yyyy C yyzz C yyyz C yyxz C yyxy C yyzy C yyxz C yyyx    yy 
 yy   yyxx  
  zz   Czzxx Czzyy Czzzz Czzyz Czzxz Czzxy Czzzy Czzxz Czzyx    zz 
   C C yzyy C yzzz C yzyz C yzxz C yzxy C yzzy C yzxz C yzyx    yz 
 yz   yzxx
  zx    Czxxx Czxyy Czxzz Czxyz Czxxz Czxxy Czxzy Czxxz Czxyx    zx 
    
  xy   Cxyxx Cxyyy C xyzz C xyyz C xyxz C xyxy C xyzy C xyxz C xyyx    xy 
  zy   Czyxx Czyyy Czyzz Czyyz Czyxz Czyxy Czyzy Czyxz Czyyx    zy 
    

 xz   Cxzxx C xzyy Cxzzz Cxzyz Cxzxz Cxzxy Cxzzy Cxzxz Cxzyx    xz 
   C C yxyy C yxzz C yxyz C yxxz C yxxy C yxzy C yxxz C yxyx    yx 
 yx   yxxx
Stress Elastic Strain
tensor Stiffness tensor
 Allow us to
EQUILIBRIUM
CONSIDERATIONS reduce the
number of
 ij   ji  kl   lk constants from
Cijkl  C jikl Cijkl  C jikl
Cijkl  C jikl Cijkl  Cijlk
81 to 36

  xx   C xxxx C xxyy C xxzz C xxyz C xxxz Cxxxy    xx 

   C C yyyy C yyzz C yyyz C yyxz C yyxy    yy 
 yy   yyxx  
  zz   C zzxx C zzyy C zzzz C zzyz C zzxz C zzxy    zz 
    C C yzyy C yzzz C yzyz C yzxz C yzxy    yz 
 yz   yzxx
  xz   Cxzxx Cxzyy C xzzz Cxzyz Cxzxz Cxzxy    xz 
    
  xy   C xyxx C xyyy C xyzz C xyyz C xyxz Cxyxy    xy 
  xx   Cxxxx C xxyy C xxzz Cxxyz Cxxxz C xxxy    xx 
   C C yyyy C yyzz C yyyz C yyxz C yyxy    yy 
 yy   yyxx  
  zz   Czzxx C zzyy C zzzz Czzyz Czzxz Czzxy    zz 
    C C yzyy C yzzz C yzyz C yzxz C yzxy    yz 
 yz   yzxx
  xz   Cxzxx Cxzyy Cxzzz Cxzyz Cxzxz Cxzxy    xz 
    
  xy   Cxyxx Cxyyy Cxyzz Cxyyz Cxyxz Cxyxy   xy 

 ij  Cijkl  kl Cijkl is the STIFNESS matrix

  xx   S xxxx S xxyy S xxzz S xxyz S xxxz S xxxy    xx 

   S S yyyy S yyzz S yyyz S yyxz S yyxy    yy 
 yy   yyxx  
  zz   S zzxx S zzyy S zzzz S zzyz S zzxz S zzxy    zz 
    S S yzyy S yzzz S yzyz S yzxz S yzxy    yz 
 yz   yzxx
  xz   S xzxx S xzyy S xzzz S xzyz S xzxz S xzxy    xz 
    
  xy   S xyxx S xyyy S xyzz S xyyz S xyxz S xyxy   xy 

 ij  Sijkl kl Sijkl is the COMPLIANCE matrix

 Matrix/Contracted Notation
• We often replace the indices with matrix (contracted) notation
for simplicity [Voigt, Lehrbuch der Kristallphysik (Teubner
VErlag, Berlin, 1910)].
xx  11  1 yy  22  2 zz  33  3
yz  23  4 xz  13  5 xy  12  6

11 12 13   1 6 5 
 22 23    2 4
   
 33   3 

• Thus, Hooke’s law is often re-written in matrix form as:
 p  C pq q or  p  S pq q

row column
 Additional note about contracted notation

• Numbers 1-3 are associated

with normal stresses and 3
strains. 5
4
4
5 2
6
• Numbers 4-6 are associated 1 6

with shear stresses and

strains. z

• On the next slide I attempt to y

show you which stresses and x
strains are related to each
elastic constant.
  ,  ,
TENSOR  Cxxxx C xxyy C xxzz C xxyz Cxxxz C xxxy 
C C yyyy C yyzz C yyyz C yyxz C yyxy 
 yyxx 
 C zzxx Czzyy Czzzz Czzyz C zzxz Czzxy 
Cijkl  
 C yzxx C yzyy C yzzz C yzyz C yzxz C yzxy 
 C xzxx C xzyy Cxzzz C xzyz C xzxz Cxzxy 
 
 Cxyxx C xyyy C xyzz C xyyz Cxyxz C xyxy 
 ,  ,

 ,  ,
CONTRACTED  C11 C12 C13 C14 C15 C16 
C C22 C23 C 24 C 25 C 26 
 21 
C C32 C33 C34 C35 C36 
Cijkl   31 
 C41 C42 C43 C44 C45 C46 
 C51 C52 C53 C54 C55 C56 
 
 C61 C62 C63 C64 C65 C66 
 ,  ,
 Contracted vs. Tensor Notation
• Contracted notation is simple and convenient
for conducting matrix inversions.

►However, for coordinate transformations or

invariant determination, it is more useful to use
the stiffness and compliances in tensor notation.
Note the change   1   C11 C12 C13 C14 C15 C16   1 
from tensor    C C22 C23 C24 C25 C26    2 
notation to matrix  2   21  
notation. See   3   C31 C32 C33 C34 C35 C36    3 
    C C42 C43 C44 C45 C46    4 
Chapter 8 in Nye  4   41
for details.   5   C51 C52 C53 C54 C55 C56    5 
    
  6   C61 C62 C63 C64 C65 C66    6 

 p  C pq  q Cpq is the STIFNESS matrix

 1   S11 S12 S13 S14 S15 S16    1 

   S S22 S 23 S24 S 25 S26    2 
 2   21  
  3   S31 S32 S33 S34 S35 S36    3 
    S S42 S 43 S44 S 45 S46    4 
 4   41
  5   S51 S52 S53 S54 S55 S56    5 
    
  6   S61 S62 S 63 S64 S65 S66    6 

 p  S pq q Spq is the COMPLIANCE matrix

Stiffness and compliance matrices are also symmetric about the main diagonal

 C11 C12 C13 C14 C15 C16   C11 C12 C13 C14 C15 C16 
C C22 C23 C24 C25 C26    C22 C23 C24 C25 C26 
 21   
 C31 C32 C33 C34 C35 C36     C33 C34 C35 C36 
C 
 41 C42 C43 C44 C45 C46      C44 C45 C46 
 C51 C52 C53 C54 C55 C56       C55 C56 
   
 C61 C62 C63 C64 C65 C66        C66 

 S11 S12 S13 S14 S15 S16   S11 S12 S13 S14 S15 S16 
S S 22 S 23 S 24 S 25 S 26    S 22 S 23 S 24 S 25 S 26 
 21   
 S31 S32 S33 S34 S35 S36     S33 S34 S35 S36 
S  
 41 S 42 S 43 S 44 S 45 S 46      S 44 S 45 S 46 
 S51 S52 S53 S54 S55 S56       S55 S56 
   
 S61 S62 S63 S64 S65 S66        S66 

Required number of components reduces to 21

The number can be further reduced by considering crystal symmetry
 Neumann’s Principle
• “The symmetry of any physical property of a crystal
must include the symmetry elements of the point
group of the crystal.”

• Point groups are the group of macroscopic symmetry

elements possessed by the structure of a crystal.*

• Crystals exhibit specific point symmetries.

• Crystal “properties” will include the point group but

can exhibit more symmetry than the point group.

* Review pages 20-25 and 279-288 in Nye for a more complete treatment.
Possible Macroscopic Symmetry Elements

• Center of symmetry

• Mirror plane

► 1-, 2-, 3-, 4- or 6-fold rotation axes

• 1-, 2-, 3-, 4- or 6-fold inversion axes

 Crystal systems
Symbolism for the symmetry elements of the
32 point groups
• The 32 crystal
classes (i.e., point
groups) are
conventionally
grouped into seven
crystal systems.

• The requirements
for membership in a
given crystal
system is that the
symmetry of the
class should
possess a certain
minimum
characteristic.

[Table from J.F. Nye, Physical Properties of Crystals, (Oxford University Press, Oxford, 1985) p. 280]
• We can describe shapes
and arrangements of Cubic
points within shapes in
terms of the symmetry
exhibited by them. Orthorhombic

120°
Hexagonal
• In crystal systems, the
symmetry of a class of
crystals should possess a Monoclinic

certain characteristic
symmetry (i.e., certain Tetragonal
minimum symmetry
elements). Triclinic

Rhombohedral
We describe crystal symmetry relative to specific crystallographic
axes and symmetry elements operating on those axes

The order of Hermann-Mauguin symbols in point groups

Crystal System Primary Secondary Tertiary
Triclinic   
Monoclinic [010], unique axis b  
[001], unique axis c
Orthorhombic [100] [010] [001]
Tetragonal [001] [100], [010] [1 10], [110]
Trigonal, Rhombohedral Axes [111] [1 10], [01 1], [101]
Trigonal, Hexagonal Axes [001] [100], [010], [1 10]
Hexagonal [001] [100], [010], [1 10] [1 10], [120, [2 10]
Cubic [100], [010], [001] [111], [1 1 1], [11 1], [1 11] [1 10], [110], [01 1],
[011], [101], [101]
Table adapted from R. Tilley, Crystals and Crystal Structures, John Wiley & Sons, Hoboken, NJ, 2006, p. 73.

The concept of crystal symmetry is illustrated on the next six viewgraphs.

You should review this and related literature on your own time.
 • Rotation about the center
point.
1

• 180° rotation moves point

“1” to “2” but produces
same pattern.

• Second 180° rotation

moves original point
(labeled with a “1”) back
to its starting point.
2

• 2-fold rotation.
43m 4-fold inversion
Tetrahedron axes parallel to
inscribed in a cube x,y,z
(aka, <100>)

3-fold rotation axes

parallel to cube
diagonals
(aka, <111>)
Mirror plane
parallel to face
diagonals
(aka, <110>)

Symmetry elements in a regular tetrahedron. (a) tetrahedron inscribed in a cube

showing three Cartesian axes; (b) 4-fold inversion axes along x [100], y [010]
and z [001]; (c) 3-fold rotation axes passing through vertex and center of
opposite triangular face (i.e., [111]); (d) mirror planes parallel to [110]. Adapted
from Fig. 4.6 in R. Tilley, Crystals and Crystal Structures, John Wiley & Sons,
B* Hoboken, NJ, 2006.
43m 4-fold rotation axes
Octahedron parallel to x,y,z
(aka, <100>) with
inscribed in a cube perpendicular
mirror plane.

3-fold inversion
axes parallel to
cube diagonals
(aka, <111>) R. Tilley, Crystals and
Crystal Structures,
John Wiley & Sons,
Hoboken, NJ, 2006.

2-fold rotation axes

parallel to face
diagonals
(aka, <111>) with
perpendicular
mirror plane.

B*
We describe lattice shapes and
arrangements of lattice points in
the same way.
 Symmetry of the Cubic P-Lattice
Letter
symbol
 Four 3-fold
P4m 3 2m rotation axes
   parallel to cube
100 111 110 diagonals
(aka, <111>)

Primary Secondary Tertiary

Adapted from W. Borchardt-Ott, Crystallography, 2nd Edition, (Springer, Berlin,

Germany, 1995) p. 99
 Table 21
Symmetry elements and conventions for the choice
• There are only 32 of axes in the 32 crystal classes
Symbols of centrosymmetrical classes are enclosed in boxes

crystal classes derived Triclinic

from the 7 crystal

systems (i.e., shapes).
Monoclinic

• Each has a certain

characteristic
symmetry. Orthorhombic

• They are presented on

stereograms on this
viewgraph and the two
that follow. [Nye, p. 284]
B* Table 21.3
 Tetragonal Trigonal

Table 21.2 Table 21.3

[Nye, pp. 285-286] B*
 Hexagonal Cubic

Table 21.4 Table 21.5

[Nye, pp. 287-288] B*
 Influence of Crystal Symmetry on Number of Elastic Constants
Axial Number of Increasing
Crystal Interaxial Minimum Symmetry
Relation- Elastic crystal
System Angles Elements
ships Constants symmetry
Four 3-fold rotation
or rotoinversion axes
Cubic a=b=c  =  =  = 90° 3
parallel to body
diagonals
One 6-fold rotation
 =  = 90°;
Hexagonal a=b≠c or rotoinversion axis 5
 = 120°
parallel to z-axis
One 4-fold rotation
Tetragonal a=b≠c  =  =  = 90° or rotoinversion axis 6 or 7
parallel to z-axis

Rhombohedral One 3-fold rotation

a=b=c  =  =  ≠ 90° or rotoinversion axis 5, 6 or 7
(Trigonal) parallel to z-axis
Three 2-fold rotation
Orthorhombic a≠b≠c  =  =  = 90° or rotoinversion axes 9
parallel to x,y,z-axes
One 2-fold rotation
Monoclinic a≠b≠c  =  = 90° ≠  or rotoinversion axis 13
parallel to y-axis
- None - Increasing
Triclinic a≠b≠c  ≠  ≠  ≠ 90° One 1-fold rotation 21 # Elastic
or rotoinversion axis const.
 More Reading Regarding the Reduction of
Elastic Constants with Increasing Symmetry

• A clear explanation of the reduction in the number of

elastic constants due to crystal symmetry is provided in
Chapter 6 of Wagoner and Chenot (pp. 196-207).

• Excellent explanations are also provided in Nye (i.e.,

Ch. 8 and Appendix B) and in Reid (Ch. 3).

• You are urged, but not required, to review some of this

material.
Form of the elastic constant matrices for different symmetries.

Triclinic Monoclinic Orthorhombic

Both classes All classes All classes

Tetragonal Trigonal
Classes 4mm, 42m,
Classes 4, 4, 4 / m 422, 4 / mmm Classes 3, 3 Classes 32, 3 m,3m Zero component (s = 0, c = 0)
Non-zero component (s ≠ 0, c ≠ 0)
Equal components
Numerically equal but opposite
 = 2 for s;  = /2 for c
=2(s11 – s22) or (c11 – c22)/2
Diad  x2 Diad  x3
standard
orientation

Hexagonal Cubic Isotropic

All classes All classes All classes
Figure adapted from J.F.
Nye, Physical Properties of
Crystals, Oxford University
Press (1985), p. 140-141.
 Cubic Symmetry and Elastic Constants
3

2
1

• Application of a stress along the [100] direction evokes a certain

elastic response. Stresses applied along the [010] or [001] axes will
evoke an equivalent elastic response. Thus:
– C11 = C22 = C33

• Similarly, applied shear stresses will evoke equivalent responses

along the following axes:
– C44 = C55 = C66; C12 = C13 = C23

• The number of independent elastic constants for a cubic material is 3.

 Isotropy considerations
3

2
1

• Cubic materials are not necessarily isotropic.

• For these systems, anisotropy is defined by the Zener ratio, A:

 C11  C12  / 2C44

NOTE: Some books (Ex., Hertzberg) present an
A or inverted version of this equation. However, the
implications are still the same.
2  S11  S12  / S 44

• When the Zener ratio = 1, the material is isotropic.

 FOR CUBIC SOLIDS
 C11 C12 C12 0 0 0 
  C C12 0 0 0 
 11

   C11 0 0 0  C11  C22
  C44 
   C44 0 0  2
     C44 0 
 
      C44 

 S11 S12 S12 0 0 0 

  S11 S12 0 0 0 
 
   S11 0 0 0 
  S 44  2  S11  S 22 
   S 44 0 0 
     S 44 0 
 
      S 44 

We can assume that most polycrystalline solids are isotropic but not all.
Many parameters are used to describe the elastic
properties of materials

1 1 1 S12
E G   
S11 S44 2  S11  S12  S11

These expressions are for isotropic solids. They are interrelated

E
G
2 1  

See Ch. 8 in Nye, Ch. 13 in Newnham, or Ch. 10 in Gersten for a

more thorough explanation.
Wikipedia also provides a nice discussion.
 Elastic Moduli in Cubic Materials

We can use the different relations among elastic constants

to ascertain elastic moduli in any crystallographic orientation

1 1
 S11  2  S11  S12  S 44   2  2   2 2   2 2 

Ehkl  2 

Where α, β, and γ are the direction cosines between the [hkl]

crystal direction and the [100], [010], and [001] directions
(i.e., the x,y,z axes).

α = cos([hkl][100]) ; β = cos([hkl][010]) ; γ = cos([hkl][010])

R.W. Hertzberg, Deformation & Fracture Mechanics of Engineering Materials, 4th ed.,
(John Wiley & Sons, New York, 1996) p. 14.
 Example Problem
• Compute the elastic modulus for tungsten in
the <110> direction.

We need the direction cosines for the unit

vectors in the cubic lattice.

<hkl> = <110>

α = 1/√2

β = 1/√2 z <110>

y
γ=0
x
 Example Problem – cont’d

1 1
 S11  2  S11  S12  S 44   2  2   2 2   2 2 

Ehkl  2 

1 1 1
 0.26  2  0.26  (0.07)  (0.66)    0  0 
  
E110  2  4 
 0.26  (0)(1/ 4)

E110  384.6 GPa

 Elastic Moduli in Hexagonal Crystals
[001]
 [hkl]

1
   S11 (1   2 ) 2  S33 4   2 1   2   2 S13  S 44 
[010] Ehkl
[100] Hexagonal

• α = cos([hkl][100]) ; β = cos([hkl][010]) ; γ = cos([hkl][010])

for the hexagonal unit cell.

• In hexagonal crystals, Ehkl depends only on the direction

cosine γ which lies perpendicular to the basal plane.

• As a result the modulus of elasticity in hexagonal crystals is

isotropic everywhere in the basal plane.
 [001]

Fig. 2.10 from Roesler et al.

Orientation dependence of Young’s
modulus for some materials. In each
spatial direction, the distance of the
[010]
[100] surface from the origin is a measure
of Young’s modulus. For example,
Copper is elastically soft along
(a) Titanium carbide (b) Tungsten, A = 1.00 (c) Aluminum, A = 1.23 <100> type directions and is
A = 0.88 elastically stiff along <111> type
directions.

We can describe anisotropy

in terms of the Zener ratio:
2  S11  S12  2C44
A 
(d) Silicon, A = 1.57 (b) Gold, A = 1.89 (f) Iron, A = 2.13 S 44  C11  C12 

When A = 1, the material is

isotropic. As A deviates, the
material becomes more
anisotropic.

(g) Nickel, A = 2.50 (h) Copper, A = 3.22 (i) Zinc

E(0001) / E(1010) = 2.13
 Elastic properties of polycrystals
• In single crystals elastic constants are determined by
bonding between atoms. As such, most single crystals are
mechanically anisotropic.

• In polycrystals, the anisotropic behavior of each grain

“averages out” because: (a) grain orientations tend to be
random; and (b) the deformation of one grain is dependent
on the deformation of its neighbor.

• Exceptions occur when the material is textured (i.e., the

grains exhibit a preferred orientation).

• Page 34 in Hosford lists some possibilities.

Temperature dependence of elastic moduli

• Young’s modulus does vary with temperature.

• For metals and ceramics there are some general

rules of thumb for temperatures lower than half of the
melting temperature.
 T 
Metals: E (T )  E (0 K)   1  0.5 
 Tmp 

 T 
Ceramics: E (T )  E (0 K)   1  0.3 
 Tmp 

• Different rules apply for polymers.

 Origin of T dependence
• Raising T increases the energy of
the atoms by an amount Ut
causing them to oscillate about
U Repulsion
their equilibrium positions.

ro rT1 • The repulsive interaction is short

range in comparison to the
attractive interaction.
r
T1 > 0 K • The mean distance between
Ut
To = 0 K atoms thus grows when
temperature is increased due to
oscillation.
Attraction
• Thermal expansion increases and
E decreases with increasing T.
 Young’s modulus also tends to scale with Tmp for similar reasons.

E (GPa)
500 Ir
Re
Ru W
400

300 Mo
Be
Fe
200
Mn Ni
Cu Ta
Zn Cr
100 Al Ti
Sn Nb
Ag Au
Zr
Na Pb Mg
0
0 500 1000 1500 2000 2500 3000 3500
Tmp (K)
 Influence of alloying
• Alloying generally has little
influence on the stiffness of
207
materials. This is because the
solubility of alloying elements is
generally very small (<10%).
121
• As noted by Roesler et al., the
elastic modulus for Al alloys
varies by roughly 10% while [Roesler, p. 41]

their strengths can be

significantly altered by alloying.

• The Cu-Ni system is an

exception to the rule. Ni is
100% soluble in Ni.