Chapter 8
Chapter 8
1
Analysis of Tensile Stress and Modulus of Unidirectional FRC
Assumption : Fiber : elastic & plastic
Matrix : elastic & plastic
2
Stage III : fiber & matrix plastic
Strength c f V f m Vm
d f d m
Modulus Ec V f Vm
d f d m
UTS cu fu V f m Vm
fu : ultimate tensile strength of fiber
m : flow stress of matrix at the fracture strain of fiber
3
Effect of Fiber Volume Fraction on Tensile Strength
(Kelly and Davies, 1965)
Assumption : Ductile matrix ( f ,fiber f ,matrix ) work hardens.
All fibers are identical and uniform. same UTS
If the fibers are fractured, a work hardenable matrix counterbalances the loss
of load-carrying capacity.
In order to have composite strengthening from the fibers,
cu fu V f m (1 Vf ) mu (1 Vf )
UTS of composite UTS of matrix after fiber fracture
4
In order to be the strength of composite higher than that of monolithic matrix,
cu fu V f m (1 Vf ) mu
UTS of pure matrix
As fu , Vcrit .
As mu m , Vcrit .
degree of work hardening
5
6
8-2. Compressive Strength of Unidirectional Fiber Reinforced Composites
Compression of Fiber Reinforced Composite
Fibers - respond as elastic columns in compression.
Failure of composite occurs by the buckling of fibers.
2
2E d
where dc is diameter,
16 l l is length of column.
7
2 Types of Compressive Deformation
1) In-phase Buckling : involves shear deformation of matrix
Gm Em
c Gm (or Em)
Vm 2(1
m ) Vm
for isostropic matrix, Gm Em
2(1 m )
predominant at high fiber volume fraction
9
1) Single Fracture
- predominant at high fiber volume fraction
- all fibers and matrix are fractured in same plane
- condition for single fracture
fu V f mu Vm m Vm
2) Multiple Fracture
- predominant at low fiber volume fraction
- fibers and matrix are fractured in different planes
- condition for multiple fracture
fu V f mu Vm m Vm
10
2. Debonding, Fiber Pullout and Delamination Fracture
Fracture Process : crack propagation
11
Fracture of Continuous Fiber Reinforced Composite
i
f
12
Force Equilibrium
rf2 f 2rf il
rf2 fu 2rf ilc ( lc : critical length of fiber )
fu lc
2i r f
fu lc lc
4i 2r f d
13
Fracture Process of Fiber Reinforced Composites
Real fibers - non-uniform properties
3 steps of fracture process
1) Fracture of fibers at weak points near fracture plane :
Wd
2) Debonding of fibers :
3) Pullout of fibers : W p
Load
W fracture W d W p
Outwater and Murphy
WP
Displacement
Wd
14
Energy Required for Fracture & Debonding
2fu d2
Wd x x : debond length
Ef 24
elastic strain E. volume
W p,ave 0
dk
lc 2 2 24
15
Fracture of Discontinuous Fiber Reinforced Composite
l
If a fiber is located within a distance, c , from crack plane, pullout
2
Probability for pullout of a fiber with length, l lc
l
Average energy to pullout per fiber with length, l
2
lc idlc
Wp,ave
l 24
probability for pullout energy required for pullout
Energy for Fiber Pullout vs Fiber Length(l)
yy
xx
17
If xx > interfacial tensile strength delamination
crack deflection
18
8-4. Statistical Analysis of Fiber Strength
Real fiber : nonuniform properties need statistical approach
Brittle fiber (ex. ceramic fibers) - nonuniform strength
Ductile fiber (ex. metal fibers) - relatively uniform strength
: probability density
f Lthat
Probability
1function
the fiber
L is between
exp strength and .
f
: statistical parameters
L : fiber length d
,
19
Let, Mk 0
k f ()d Mk : kth moment of statistical distribution function
Coefficient of Variation
S 1 2 / 2 11/
1/ 2
1 1/
1
( f() only, for 0.05 0.5 )
- 0.92
20
1
(L )1/ (1 ) 1/ vs L plot
As L , . "Size Effect
As , . is less dependent on L.
If , spike distribution function (dirac delta function)
uniform strength independent on L
Glass fiber
Boron, SiC fibers 0.1, 11
0.2 0.4, 2.7 5.8
21
Strength of Fiber Bundle
Bundle strength Average strength of fiber n
# of fibers
<
Assumption : Fibers - same cross-sectional area
- same stress-strain curve
- different strain-to-fracture
Let F() : The probability that a fiber will break before a certain value of is
attained.
Cummulative Strength Distribution Function
F() f () d
0
22
Comparison of B and
B
As , . ( : coefficien t of variation)
( )
B
0.1 0.8
0.25 B 0.6
23
8-5. Failure Criteria of an Orthotropic Lamina
Assumption : Fiber reinforced lamina - homogeneous, orthotropic
Failure Criterion of Lamina
1. Maximum Stress Criterion
Failure occurs when any one of the stress components is equal to or greater
than its ultimate strength.
Interaction between stresses is not considered.
Failure Condition
1 X1T or 1 X1C
or 2 XT2 or 2 XC2
or 6 S or 6 S
where XT : ultimate uniaxial tensile strength in fiber direction (>0)
1
: ultimate
X1C uniaxial compressive strength in fiber direction (<0)
: ultimate
XT2 uniaxial tensile strength in transverse direction
: ultimate
XC2 uniaxial compressive strength in transverse direction
S : ultimate planar shear strength
24
ex) If uniaxial tensile stress x is given in a direction at an angle with the fiber axis.
1 x x
[T] 0
2 1
6 0
m2 n2 2mn
[ T ] n 2 m2 2mn
2
2
mn mn m 2
n
Failure occurs when,
1 x m2 X1T
or 2 x n2 XT2
or 6 x mn S
Failure Criterion
X1T
x 2 longitudin al tensile failure
m
X1T Failure occurs by a criteria, which
or x 2 transverse tensile failure
n is satisfied earlier.
S
or x planar shear failure
mn
25
26
2. Maximum Strain Criterion
Failure occurs when any one of the strain components is equal to or greater
than its corresponding allowable strain.
Failure Condition
1 1T or 1 1C
or 2 T2 or 2 C2
or 6 S6 or 6 S6
where 1
T
: ultimate tensile strain in fiber direction
1 : ultimate compressive strain in fiber direction
C
27
3. Maximum Work Criterion
Failure criterion under general stress state
Tsai-Hill
12 12 22 12
2 2 1
X12
X1 X2 S 2
ex) For uniaxial stress x , having angle with the fiber axis
1 x m2
2 x n2 substituting
6 x mn
Failure criterion
m4 n4 2 2 1 1
X2 X2 m n 2
2x 1
S X
2
1 2
28
4. Quadratic Interaction Criterion
For the shear stress components, the reverse sign of shear stress should
give the same criterion.
30
Boron/Epoxy composite
Intrinsic properties
X1T 27.3 MPa, XT2 1.3 MPa, S 1.4 MPa
X1C 52.4 MPa, XC2 6.5 MPa
31
32
8-6. Fatigue of Composite Materials
33
Constant-stress-amplitude Fatigue Test
Damage Accumulation vs Cycles
34
S-N Curves of Unreinforced Plolysulfone vs Glassf/Polysulfone, Carbonf/Polysulfone
35
Fatigue of Particle and Whisker Reinforced Composites
For strain-controlled cyclic fatigue or low cycle fatigue, the composites show
lower fatigue resistance compared to Al alloy, which is attributed to the lower
ductility of the composites.
Particle or short fibers can provide easy crack initiation sites. The detailed
behavior can vary depending on the volume fraction, shape, size of
reinforcement and mostly on the reinforcement/matrix bond strength.
36
Fatigue of Laminated Composites
Crack Density, Delamination, Modulus vs Cycles
i) Ply cracking
ii) Delamination
iii) Fiber fatigue
37
Modulus Reduction during Fatigue
Ogin et al.
Modulus Reduction Rate max
n
1 dE max
2
m
A 2
E0 dN E0 (1 E / E0 ) time
min
where E : current modulus
E0 : initial modulus N : number of cycles
max : peak fatigue stress A, n : constants
1 dE max
2
log vs log 2 plot linear fitting
E0 dN E
0 (1 E / E )
0
38
Integrate the equation to obtain a diagram relating modulus reduction to number
of cycles for different stress levels.
used for material design
39
8-7. Thermal Fatigue of Composite Materials
Thermal Stress
Thermal stresses arise in composite materials due to the generally large
differences in thermal expansion coefficients() of the reinforcement and matrix.
It should be emphasized that thermal stresses in composites will arise even if
the temperature change is uniform throughout the volume of composite.
T
Thermal Fatigue
When the temperature is repeatedly changed, the thermal stress results in the
thermal fatigue, because the cyclic stress is thermal in origin. Thermal fatigue
can cause cracking of brittle matrix or plastic deformation of ductile matrix.
Cavitation in the matrix and fiber/matrix debonding are the other forms of
damage observed due to thermal fatigue of composites. Thermal fatigue in
matrix can be reduced by choosing a matrix that has a high yield strength
and a large strain-to-failure. The fiber/matrix debonding can only be avoided
by choosing the constituents such that the difference in the thermal expansion
coefficients of the reinforcement and the matrix is low.
40