disorder on the fracture of materials Elisabeth Bouchaud Solid State Physics Division (SPEC) CEA-Saclay, France MATGEN IV, Lerici, Italy September 19-23, 2011
Dec 31, 2015
Effect of disorder on the
fracture of materialsElisabeth Bouchaud
Solid State Physics Division (SPEC)CEA-Saclay, France
MATGEN IV, Lerici, ItalySeptember 19-23, 2011
Irradiation defects in solids
Vacancy Interstitial Frenkel
Other « defects » Good compromise ofmechanical properties
Metallic alloys(ONERA)
ToughenedPolymer(ESPCI)
Fiber composite(Columbia)
Tough ceramics(Berkeley)
How to estimate the properties of a composite?
s
s
Ecomposite F E + F E
sr= F sr + F sr
Include the effect of heterogeneities in a statistical description
- Rare events statistics- Strong stress gradients in the vicinity of a crack tip
OUTLINE1. Elements of LEFM2. Effect of disorder on the morphology
and dynamics of the crack front3. Experimental observations4. Discussion
1- Elements of LEFM
s
s
a
A crude estimate of the strength to failure
s=EDxa
Failure : Dx≈a sf ≈ E
sf ≈ E/100
Presence of flaws!
Stress concentration at a crack tip (Inglis 1913)
s
s
2b
2a
A
sA > s: stress concentration
)21(b
aA
a
b
aA
2
)21(
1- Elements of LEFM
Infinitely sharp tip:
A
A
a
,0
2s
s
r ij Irwin (1950)
)(2
ijij fr
K
K=stress intensity factor
)(f2W
aaK
a
W
Sample geometry
s (r
)r
r
ar )(
Strong stress gradient
A
1- Elements of LEFM
Mode IIIn-plane, shear,
slidingKII
Mode ITension, opening
Mode IIIOut-of-plane, shear
TearingKI KIII
Mixed mode, to leading order:
)()()(2
1
IIIijIII
IIijII
IijIij fKfKfK
r
1- Elements of LEFM
Griffith’s energy balance criterion
Elastic energy'
22
E
BaUE
Surface energy BaU S 4
Total change in potential energy:
SE UUU
Propagation at constant applied load: 0da
Ud2a
B
s
strain plane1
'
stress plane'
2
E
E
EE
1- Elements of LEFM
Happens for a critical load:lengthCrack
constant Material'2
a
EC
Or for a critical stress intensity factor:
'4 EKK C Fracture toughness
' ;
2
E
KG
dA
dUG E Energy release rate
)2
M(1
, IfR
ccC V
KKKVKK
1- Elements of LEFM
q
KII=0
q
Crack path: principle of local symmetry
1- Elements of LEFM
)1( ,G
VVKK RC
Onset of fracture: 4'
2
E
Kc
Beyond threshold:
PMMA Glass
22IcI KKV
(E. Sharon & J. Fineberg, Nature 99)
1- Elements of LEFM
Heterogeneities
Rough crack front
Uneven SIFs
Heterogeneous path
Steady crack morphology?Dynamics?
(JP Bouchaud & al, 93J. Schmittbuhl & al, 95D. Bonamy & al, 06)
2- Effect of disorder…
2- Effect of disorder…
2D 3D
)()( 0 zKKzK III
))(')'(
)()'(
2
11()( 2
20 fodz
zz
zfzfPVKzK II
(Meade & Keer 84, Gao & Rice 89)
Stabilizing term
2- Effect of disorder…
2threshold
2 )()()( zKzKMV II M(f(z),z)
))')'(
)()'((
21()(
2
202 dzzz
zfzfPVKzK II
))'(),(()',())'(,'())(,(
;0))(,(
)))(,(2
11()(threshold
zfzfzzzfzzfz
zfz
zfzKzK IcI
))(,()')'(
)()'((
2)( 22
20220 zfzKdzzz
zfzfPVKKK
t
zfIcIIcI
2- Effect of disorder…
tensionLine))(,(),(
zfzF
t
tzf F
2
2 ),(
z
tzf
Edwards-Wilkinson model
)')'(
)()'((
2dz
zz
zfzfPV
220IcI KK
Non local elastic restoring force
2- Effect of disorder…
))(,()')'(
)()'((
2)( 22
20220 zfzKdzzz
zfzfPVKKK
t
zfIcIIcI
Depinning transition:• order parameter V• control parameter KI
0
(KI0-KIc)q
(KI0-KIc)
V
KI0
KIc~
tzz
tzf
,
),(
Stable
Propagating
2- Effect of disorder…
Depinning: line in a periodic potential
f(x=0,t=0)=0
x
f0
)cos( fFFt
fm
F
Pulling force
Obstacle forceO
bst
acl
e f
orc
ef
f=0
F
1
m
m
F
FF
2)( 2fF
dfdt
m
T?
1
2
)(0
0 2
f
m fF
dfT V (F-Fm)
2- Effect of disorder…
z2D=0.39 (A. Rosso & W. Krauth & O. Duemmer)
z z+Dz
Df(Dz)
x2/12))()(()(
zzfzzfzf
Dzzf 2)(
t
t+Dt
ttfttftft
2/12))()(()(
In plane projection of crack front
')'(
)',(),(
2
32
22),(
2
000 dz
zz
zxhzxhK
x
hKKzxK II
IIII
(Movchan, Gao & Willis 98)
Out of plane projection of crack front
X
Z
f(z) ')'(
)()'()(
2
1)()(
200 dz
zz
zfzfzKPVzKzK III
zy
h(z)
2- Effect of disorder…
Local symmetry principle
KII=0
Crack trajectory
)),(,,(')'(
)'()(
)2(
3212
20
0
zxhzxdzzz
zhzh
K
K
x
h
I
II
)(.),(/1
x
zfxzxh
1 if
1 if1)(
u
u
uuf
≈ 0.4 ≈ 0.5k ≈ / ~ 0.8 (Bonamy et al, 06)
2- Effect of disorder…
f(z)
Out-of-plane
Projection on theyz plane
In-plane
Projection on thexz plane
3- Experiments 3D
P.Daguier et al. (95)
x
z2D ≈ 0.55-06
3- Experiments 3D
Aluminiumalloy
z=0.773nm0.1mm
(M. Hinojosa et al., 98)
Profiles perpendicular to the directionof crack propagation
3- Experiments 3D
z= 0.78from 5nm to
0.5mm
Dz
Profiles perpendicular to the direction of crack propagation
(Dz)
(µm
)
z = 0.77
Zm
ax(D
z) (
µm
)
Dz (µm)
Aluminum alloy (SEM+Stereo)
)(/1
x
zfxh
1 if
1 if1)(
u
u
uuf
h/
x
z/ x1/k
3- Experiments 3D
A+
B+ΔxΔz
= 0.75 = 0.6k = / ~ 1.2
x
y
z
h/
x
z/ x1/k
Mortar
Quasi-crystal(STM)
h (
Å)
Δh2D(Δz, Δx) = (<(h(zA+Δz, xA+Δx) - h(zA, xA))2>A)1/2
z/ x1/k
h/
x
Exceptions…Sandstone fracture surfaces
log(P(f))
log(f)
z≈0.47
(Boffa et al. 99)
dzz
P(D
h)
Dh/(dz)z
(Ponson at al. 07)
z≈0.4
3- Experiments 3D
« Model » material : sintered glass beads (Ponson et al, 06)
Porosity 3 to 25%Grain size 50 to 200 mmVitreous grain boundaries
Exceptions…
3- Experiments 3D
ζ=0.4 ± 0.05β=0.5 ± 0.05
k=ζ/β=0.8 ±0.05
2 independent
exponents
« Universal » structurefunction
+
Roughness at scales> Grain size
1/
(Ponson et al. 06)
3- Experiments 3D
Rc increases with time
x1
x
S. Morel & al, PRE 2008
Rc(x1)
=0.75
=0.4
z=0.79
z=0.4
Rc(x1) Rc(x2)
X2
Rc(x2)>Rc (X1)
3- Experiments 3D
Mortar specimens
(K.J. Måløy & al)3- Experiments: interfacial fracture
z(mm)
x(mm)
3- Experiments: interfacial fracture
z2D=0.63 z2D=0.37
100µm
log(D
h(D
z)/d
0);
d0=
1µm
log(Dz/d0); d0=1µm
F 200µm
(S. Santucci et al, EPL 2010)
z
xx
x
<v>=28.1µm/s; a=3.5µm
(K.J. Måløy et al. 06)
Waiting time matrix:t=0 W(z,x)=0t>0 Wt+dt(z,x)=1+Wt(z,x)
if front in (z,x)
Front location
Spatial distribution of clusters (white) v(z,x)>10 <v>
t
a
xzwxzv
),(
1),(
3- Experiments: interfacial fracture
0.39µm/s≤<v>≤40µm/s1.7µm ≤a≤10µm
C=3
Cluster size distribution
Slope -1.6
(K.J. Måløy et al. 06)
(D. Bonamy & al., 08)
3- Experiments: interfacial fracture
(S. Santucci & al., 08)
3- Experiments: interfacial fracture
•Disorder roughnening •Elastic restoring forces rigidity
Undamaged materialTransmission of stresses throughundamaged material :long range interactions (1/r2) very rigid line
Transmission of stressesthrough a « Swiss cheese »: Screening of elastic interactions low rigidity
4- Discussion
Gradient percolation(A. Hansen & J. Schmittbuhl, 03)
Z
X
Damage RFM gradient percolation process
z3D=b3D= 2n/(1+2n)=4/5(nRFM/3D=2)
4- Discussion
r « Rc r » Rc
Rc
Damage zonescale
Large scales:elastic domain
z=0.75, b=0.6 z=0.4, b=0.5
?4- Discussion
3 regions on a fracture surface:1 Linear elastic region z=0.4 b=0.5/log2 Intermediate region: within the FPZ
Damage = « perturbation » of the front (screening)z=0.8 b=0.6 direction of crack propagation3 Cavity scale: isotropic region
1 2 3
- Size of the FPZ- Direction of crack propagation within FPZ- Damage spreading reconstruction
Fracture of an elastic solid is a dynamic phase transition
4- Discussion
Questions•A model in the PFZ? How to reconcile line model and percolation gradient model ?
•Size of FPZ? Reliable measurements?
•Direct measurement of the disorder correlator
•Dynamics of crack propagation in 3D?
• Radiation damage?
•Breaking liquids…
cuttu ]v)()][(v)([)vt(
Thank you!
2/12
/1
/1
)()(
)()()(
)()()(
x
k
x
k
k
k
x
k
k
xhxxh
xhxxhxR
xhxxhxh
k
Gk
k
R
/1
21
2
(S. Santucci et al., 07)
Rk(
Dx)/
RkG
Log10(Dx/d0)
PMMAL ≈ 50µm
Dx/d0
Dh
k(D
x)/
RkG
PMMAb≈0.6
3- Statistical characterization of fracture
3.2- Interfacial fracture
(Salminen et al, EPL06)
Peel-in(paper)
3- Statistical characterization of fracture
Gutenberg-Richter exponent
3- Statistical characterization of fracture
3.2- Interfacial fracture
Omori’s law
Slope -1
3.2- Interfacial fracture
3- Statistical characterization of fracture
vv
tip
Au(111) film(~150 nm)
mica plate
Sample holder
Z-piezo
It
wedge
preamplifier
feedbacksystem of STM
PC
Vibration isolation system
Ut
(A. Marchenko et al., 06)
3.2- Interfacial fracture
3- Statistical characterization of fracture
Humid air
n-tetradecane
Humid air Tetradecane
3.2- Interfacial fracture
3- Statistical characterization of fracture
smv
vvvP
O
O
/10
)/exp()(6
9.2)(
vvP
Magnitude
10-4 10-2 1 102
Approximate energy radiated (1015J)
104
102
1
103
10
Num
ber
of
eart
hquake
s
San Andreas fault
(J. Sethna et al)
3.2- Interfacial fracture
3- Statistical characterization of fracture
AE measurements on mortar (B. Pant, G. Mourot et al., 07)
Energy distribution
Log(E/Emax)
Log(N
(E))
P(E)E-1.41
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
P(E)E-1.49
P(E)E-1.40
AE measurements on polymeric foams (S. Deschanel et al., 06)
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
Al alloy Ni-platedBS SEM
(E.B. et al., 89)
r/x
C(r)r-z
z≈0.8
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
Profiles perpendicular to the directionof crack propagation
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
z= 0.78from 5nm to
0.5mmDz
Profiles perpendicular to the direction of crack propagation
(Dz)
(µm
)
(P. Daguier & al., 96)
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
Aluminiumalloy
z=0.773nm0.1mm
z = 0.77
Zm
ax(D
z) (
µm
)
Dz (µm)(M. Hinojosa et al., 98)
Profiles perpendicular to the direction of crack propagation
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
(J. Schmittbuhl et al, 95)
Profiles perpendicular to the directionof crack propagation: granite
z≈0.8
z≈0.85
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
z (µm)direction ofcrack front
x (µm)direction of
crackpropagation
Anisotropy of fracture surfaces
~ 0.8
~ 0.6
Direction ofcrack propagation
Direction of crack front
Log(Δx), log(Δz)
Log
(Δ
h)
L. Ponson, D. Bonamy, E.B. (05)
1 10 102 103
1
10
0.1
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
Exceptions…
Sandstone fracture surfaceslog(P(f))
log(f)
z≈0.47
(Boffa et al. 99)
dzz
P(D
h)
Dh/(dz)z
(Ponson at al. 07)
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
« Model » material : sintered glass beads (Coll. H. Auradou, J.-P. Hulin & P. Vié 06)
Porosity 3 to 25%Grain size 50 to 200 mmVitreous grain boundaries
Linear elastic material
3.3- Fracture of three-dimensional solids
3- Statistical characterization of fracture
Exceptions…
1/z
ζ=0.4 ± 0.05β=0.5 ± 0.05
z=ζ/β =0.8 ±0.05
2 independent
exponents
« Universal » structurefunction
+
Roughness at scales> Grain size
(Ponson et al. 06)
3- Statistical characterization of fracture
Summary
Cracks and fracture surfaces are self-affine:
-Thin sheets b≈0.6 at scales > L
- Interfacial fracture z’ ≈0.6 at scales < L
- 3D solid out-of-plane perpendicular to crack propagation z≈0.75 (5 decades!)
- Out-of-plane parallel to the direction of crack propagation b ≈0.6
- In-plane z’ ≈0.6
3- Statistical characterization of fracture
f(z)
Out-of-plane
Projection on theyz plane
In-plane
Projection on thexz plane
3.3- Fracture of 3D specimens
3- Statistical characterization of fracture
P.Daguier et al. (95)
x
z’≈0.55-06
3.3- Fracture of 3D specimens
3- Statistical characterization of fracture
Out-of-plane roughness measurements
Polishing
3.3- Fracture of 3D specimens
3- Statistical characterization of fracture
Al alloy Ni-platedBS SEM
(E.B. et al., 89)
r/x
C(r)r-z
z≈0.8
3.3- Fracture of 3D specimens
3- Statistical characterization of fracture
(J. Schmittbuhl et al, 95)
Profiles perpendicular to the direction of crack propagation: granite
z≈0.8
z≈0.85
3.3- Fracture of 3D specimens
3- Statistical characterization of fracture
z (µm)direction ofcrack front
x (µm)direction of
crackpropagation
Anisotropy of fracture surfaces
~ 0.8
~ 0.6
Direction ofcrack propagation
Direction of crack front
Log(Δx), log(Δz)
Log
(Δ
h)
L. Ponson, D. Bonamy, E.B. (05)
1 10 102 103
1
10
0.1
3.3- Fracture of 3D specimens
3- Statistical characterization of fracture
Béton(Profilométrie)
Glass (AFM)
Alliage métallique (SEM+Stéréoscopie)
Quasi-cristaux (STM)
130mm
Δh2D(Δz, Δx) = (<(h(zA+Δz, xA+Δx) - h(zA, xA))2>A)1/2
h (
nm)
z (nm)
A B
ΔxΔz
L. Ponson et al, PRL 2006L. Ponson et al, IJF 2006
h/
x
z/ x1/ z
)(. /1
x
zfxh
1 si
1 si1)(
u
u
uuf
= 0.75 = 0.6Z= / ~ 1.2
z
Béton(Profilométrie)
Glass (AFM)
Alliage métallique (SEM+Stéréoscopie)
Quasi-crystals(STM)
Δh2D(Δz, Δx) = (<(h(zA+Δz, xA+Δx) - h(zA, xA))2>A)1/2
A B
ΔxΔz 130mm
Quasi-crystalsCourtesy P. Ebert
Coll. D.B., L.P., L. Barbier, P. Ebert
z
z
)(. /1
x
zfxh
1 si
1 si1)(
u
u
uuf
= 0.75 = 0.6
z = / ~ 1.2
h (
Å)
4- Statistical characterization of fracture
Béton(Profilométrie)
Glass (AFM)
Aluminum alloy (SEM+Stereo)
Quasi-crystals (STM)
Δh2D(Δz, Δx) = (<(h(zA+Δz, xA+Δx) - h(zA, xA))2>A)1/2
A B
ΔxΔz 130mm
)(. /1
x
zfxh
1 si
1 si1)(
u
u
uuf
= 0.75 = 0.6
z = / ~ 1.2
h/
x
z/ x1/z
h (
Å)
Coll. D.B., L.P., L. Barbier, P. Ebert
4- Statistical characterization of fracture
Mortar(Profilometry)
Glass (AFM)
Aluminum alloy (SEM+Stereo)
Quasi-crystals(STM)
Δh2D(Δz, Δx) = (<(h(zA+Δz, xA+Δx) - h(zA, xA))2>A)1/2
A B
ΔxΔz 130mm
)(. /1
x
zfxh
1 si
1 si1)(
u
u
uuf
= 0.75 = 0.6
z= / ~ 1.2
h/
x
z/ x1/z
Mortar
(Coll. S. Morel & G. Mourot)h (
Å)
Coll. D.B., L.P., L. Barbier, P. Ebert
Mortar(Profilometry)
Glass (AFM)
Metallic alloy (SEM+Stereo)
Quasi-crystals(STM)
A B
ΔxΔz 130mm
z/ x1/z(lz/lx)1/(z/lz)/(x/lx)1/z
h/
x(
h/l
x)/(x
/lx)
Universal structure functionVery different length scales
h (
Å)
Coll. D.B.,L.P.,L. Barbier,P. Ebert
4- Statistical characterization of fracture
q
Exp
on
en
t
q
1mm
Preliminary results(G. Pallarès, B. Nowakowski et al., 08)
4- Statistical characterization of fracture
Exceptions…
Sandstone fracture surfaceslog(P(f))
log(f)
z≈0.47
(Boffa et al. 99)
dzz
P(D
h)
Dh/(dz)z
(Ponson at al. 07)
ζ=0.4 ± 0.05β=0.5 ± 0.05
z=ζ/β=0.8 ±0.05
2 independent
exponents
« Universal » structurefunction
+
Roughness at scales> Grain size
1/z
(Ponson et al. 06)
4- Statistical characterization of fracture
1D crack in a 2D sample
4.1- Random Fuse Models
L. De Arcanglis et al, 1985
4- Stochastic models of failure
2/1
1
2)(1
)(
L
x
yxyL
Lw
(E. Hinrichsen et al. 91)
b≈0.7=2/3 ?
4.1- Random Fuse Models
4- Stochastic models of failure
(P.Nukala et al. 05)
(P.Nukala et al. 06)(G. Batrouni & A. Hansen, 98)
z=b=0.52
Minimum energy surface z≈0.41
(A. Middleton, 95Hansen & Roux, 91)
=z b≈0.5
Fracture surface=juxtapositionof rough damage cavities
(Metallic glass, E.B. et al, 08)
4.1- Random Fuse Models
s-2.55
P(E)E-1.8
Esa, a≈1.3
Avalanche size distribution (S. Zapperi et al.05)
4.1- Random Fuse Models
4- Stochastic models of failure