Formulation of Critical Questions for Understanding Brittle Fracture PowerPoint PPT Presentation

presentation player overlay
1 / 101
About This Presentation
Transcript and Presenter's Notes

Title: Formulation of Critical Questions for Understanding Brittle Fracture


1
Formulation of Critical Questions for
Understanding Brittle Fracture
  • J. J. Mecholsky, Jr.
  • Materials Science Engineering
  • University of Florida
  • Gainesville, FL 32611

MS T 2007
2
Bond Breaking Leads to Characteristic Features
Log v
U
s
Kc
e
r
Log K Log (Ys c ½)
3
Fracture In Materials
There are several fundamental questions that need
to be answered
How do bonds break?
Once a bond breaks, how do the ensembles of
broken bonds propagate?
Is there a mathematical formulation which
permits insight into the fracture process at
all scales?
4
Hypothesis Fracture is a fractal process.
Implications ? Bond breaking is the
primary factor controlling


the formation of fracture surface
structure at the macroscopic length scale.
? Fracture topography is statistically
self-similar (or self-affine).
? Fracture topography scales from atomic to
macroscopic
lengths.
5
Questions guide presentation
  • What is fractal geometry?
  • How is fractal geometry related to fracture?
  • What do we know about fractal fracture?
  • What do we need to do in the future to determine
    the extent of the effect of fractal fracture?

6
What Is A Fractal?
Benoit Mandelbrot A rough or fragmented
geometric shape that can be subdivided in parts
that are similar in shape (self-similarity), each
of which is (at least approximately) a
reduced-size copy of the whole shape (scale
invariance).
z z2 C
Defined by a Fractional Dimension (i.e. 1.3, 2.4,
etc. where D 0.3,0.4,etc.)
7
Fractal Geometry A Renewed Math
  • Characteristics
  • Non-differentiable
  • Defined by a Fractional Dimension (i.e. 1.5, 2.4,
    etc.)
  • Self-similar features
  • Scale invariant features
  • http//math.rice.edu/lanius/frac/

8
Characteristic Markings Are Observed on the
Fracture Surface
Hackle
Mist
Mirror
Fracture origin
9
The Mirror Region Is Not Smooth
10
Mist and Hackle Appear Similar in Shape
11
Mist and Hackle Appear Similar in Shape
mist
mist
hackle
12
Many Observations Lead To The Conclusion Of
Fractal Fracture
Scaling Behavior Self-Similarity In A
Plane Scale Invariance
13
FRACTAL DIMENSION IS MEASURED ALONG CONTOUR
Log Length (A-B)
Slope -D
A
Log (Ruler Length)
B
A-B Slit Island Contour
14
(No Transcript)
15
Fractal Dimension Varies For Different Materials
J. J. Mecholsky, Jr., Fractography, Fracture
Mechanics and Fractal Geometry An Integration,
Ceram. Trans. 64, in Fractography of Glasses and
Ceramics III, eds. J. P. Varner, V.D. Frechette,
G. D. Quinn, Am. Ceram. Soc. (1996).
16
Mandelbrot, Passoja Paullay first applied
fractal geometry to fracture

Fractal character of fracture surfaces of metals
Benoit B. Mandelbrot Dann. E. Passoja Alvin J.
Paullay Nature 308, 721 - 722 (1984).
17
There are over 10,000 references for fractal
fracture!
  • Fractal fracture mechanicsA review Engineering
    Fracture Mechanics V. 51, Issue 6, Aug 1995, pp
    997-1033
  • Evidence of fractal fracture. T Chelidze, Y
    Guguen, International Journal of Rock Mechanics
    and Mining Sciences 2733, 223-225, Pergamon
    Press, 1990.
  • On the Interpretation of the Fractal Character of
    Fracture Surfaces , Dauskardt, R H Haubensak, F
    Ritchie, R O Acta Metall. Mater. V. 38, no. 2,
    pp. 143-159. Feb. 1990
  • Surfaces Formed by Subcritical Crack Growth in
    Silicate Glasses, Guin and Wiederhorn and,
    Inter. J. of Fracture 140(1-4)15-26 (2006)

18
Many others observed fractal fracture
19
Bond Breaking Leads to Characteristic Features
Log v
U
s
Kc
e
r
Log K Log (Ys c ½)
20
Characteristic Markings Are Observed on the
Fracture Surface
s M / r1/2
Hackle
Mist
Mirror
Fracture origin
s KC/ (Y c1/2)
21
Fracture Mechanics Fractography Provide A
Framework for Quantitative Analysis
KIC Y s c 1/2 Crack Boundary KB1 Y1 s
r11/2 Mirror-Mist Boundary KB2 Y2 s
r21/2 Mist-Hackle Boundary KB3 Y3 s
r31/2 Crack Branching Boundary
c/rj constant
22
D Is Related to Mirror-to-Flaw Size Ratio
Mecholsky Freiman J ACerS 74123136 (1991)
23
D Is Related To Flaw-to-Mirror Size Ratio
KB1 Y s r11/2
KC D1/2 E a01/2
Mecholsky Freiman J ACerS 74123136 (1991)
24
Toughness Increases With Fractal Dimension
Kc E a01/2 D1/2
g ½ a0 ED
25
Fractal Geometry Has Been Used In Failure Analysis
  • Silicon J. Mater Res 6,1248 91
  • Ocala Chert J Mater Sci Ltrs 7, 1145 88
  • Intermetallics J. Mater. Sci.6,1856 91
  • Si3N4 J. Mater. Sci. 32 6317 97
  • Dental Glass Cer. JACerS 78 3045 95
  • Epoxy Resin Scanning 20 99 98

2 g a0 ED Kc a01/2 ED 1/2 D
c/r1
26
Bond Breaking Leads to Characteristic Features
Log v
U
s
Kc
e
r
Log K Log (Ys c ½)
27
Bell Dean Model Used for MO Calculations
cf. Varshneya, Fundamentals of Inorganic
Glasses (After Bell and Dean, Nature 212, 1354
1966)
28
cf. Varshneya, Fundamentals of Inorganic Glasses
29
Groups of atoms at crack tip behave in a similar
manner along crack front
s22
Crack tip
30
MO Simulates Bond Breaking At The Crack Tip
2
1
Simulated SiO2 D displ. 1 Å
31
2
1
32
2
2
1
1
33

2

1
34
2
1
35
2
1
36
2
1
37
2
1
38
2
1
39
2
1
40
Strain Can Be Measured In Model
a0 a / e c a / c-c
a
a
c
c'
J. Non-Crystalline Solids 260 (1999) 99-108.
41
Energy Geometry Are Related In The Fracture
Process
g fracture energy E Elastic Modulus D
Fractal Dimensional Increment a0 Characteristic
Structural Parameter KC Fracture Toughness
42
Modeling and Experimental Results Agree
a0 2g/ (ED)
J. K. West, J. J. Mecholsky, Jr, and L. L.
Hench, The Quantum and Fractal Geometry of
Brittle Fracture, J. Non-Crystalline Solids 260
(1999) 99-108.
43
Notice Similarity of Fracture Energy Equation To
Freiman et al.
  • g ½ D E a0 fractal approach
  • g k E d 0 classical
    approach

44
a0 Is Related To Structure
a0 2 g / ( ED)
45
Characteristic Markings Are Observed on the
Fracture Surface
Hackle
c / r constant
Mist
Mirror
Fracture origin
c
r
46
(No Transcript)
47
Fracture Is A Fractal Process
  • Fracture transcends many length scales
    Self-similar (or self-affine), scale invariant
    characterized by D.
  • Hypothesis 2g a0 E D
  • Observations seem to support the
    hypothesis. c/r D g ? ED
  • (D is a geometric energy scaling factor.)
  • (a0 is a fracture surface structural element)
  • MD MO modeling provide framework for
    understanding macroscopic observations

48
Atomic Scale Structure Expands To Fracture Surface
2 g a0 ED
Log v
U
s
Kc
e
r
Log K Ys c 1/2
49
Implications of fractal fracture need to be
considered
  • If Fracture is a fractal process
  • Geometry at atomic scale related to macroscopic
    scale
  • An algoritm can be found to recreate a fracture
    surface
  • Relatively simple rules should govern the
    fracture process
  • Quantum mechanics should be related to LEFM

50
CONCLUSIONS
  • The atomic basis of fracture is a series of
    bond reconfigurations resulting in the
    production of more space.
  • Fractal fracture implies that the same fracture
    process should be able to be observed at all
    length scales.
  • Experimental data acquisition analytical
    modeling have to be interactive to be
    successful.
  • Analytical models have to explain fractal
    nature of fracture, mirror, mist hackle and
    crack branching.

51
What should be done in the future?
  • Raman spectroscopy should be related to atomic
    scale modeling as a check
  • Models of fracture should include quantum
    mechanics
  • Continuum vs. discrete phenomena should be
    investigated (fracture is a discrete event and
    yet can be modeled using a continuum approach)
  • There has to be an interactive cooperation
    between analysis (modeling) and experimentation.

52
Many Have Contributed To This Work
  • Dr. Dann Passoja Dr. Alvaro Della Bona
  • Dr. Jon West Dr. Tom Hill
  • Dr. Steve Freiman Jacques Cuneo
  • Dr. Joe Simmons Prof . Jeff Thompson
  • Dr. Jim Runt Dr. Ananth Naman
  • Dr. Larry Hench Jim Plaia
  • Prof. Tom Mackin Dr. Zheng Chen
  • Karen Feinberg-Ringel
  • Former PSU Students Dr. Ben Tsai
  • Colleagues Prof. Jason Griggs
  • Dr. Luis Carney
  • Former UF Students

Partial Support ONR United Technologies
53

University of Florida, Gainesville FL







54
(No Transcript)
55
Technical Approach
Modeling e.g., Molecular Orbital (MO)
Modeling of Fracture - determines a0. a0 is a
characteristic structural parameter that measures
strained bonds.
Experimental determination and comparison of
parameters obtained in modeling, i.e., a0 and D
D is the fractal dimensional increment
56
The Mirror/Mist Region Is Not Smooth
57
Mandelbrot, Passoja Paullay first applied
fractal geometry to fracture

Fractal character of fracture surfaces of metals
Benoit B. Mandelbrot Dann. E. Passoja Alvin J.
Paullay Nature 308, 721 - 722 (1984).
58
Stresses at Crack Tip Are Near Theoretical Values
s22
Crack tip
59
Theoretical Strength Can Be Obtained Using
Fractal Geometry
  • (1a) sij K / (2 p r)1/2 f(q) LEFM
  • (1b) s22 Kc / (2 p r)1/2
  • (2) gc g0 ½ a0 ED
    Fractal Geometry
  • (3a) st E a0 ½ D ½ / (2 p a0)1/2 combine
    (1) (2)
  • let r a0
  • (3b) st E D ½ / (2 p )1/2 No
    potential energy expression needed!

60
Theoretical Strength Predicted Between E/10 and
E/p
61
CONCLUSIONS
  • Fractal fracture implies that the same fracture
    process should be able to be observed at all
    length scales.
  • Experimental data analytical modeling

    have to be interactive to
    be successful.
  • All models should be compared to (real)
    experimental data.
  • Analytical models have to explain fractal
    nature of fracture, mirror, mist hackle and
    crack branching.

62
Critical Questions Need To Be Asked
  • What are the energetic geometric steps to
    fracture?
  • Is a flat fracture (of primary bonds) possible
    above absolute zero? What is bond rupture?
  • Is roughness a meaningful parameter in
    fracture?
  • How does energy scale?
  • How does a crack propagate at all length
    scales?

63
Relationship Holds For Large Size Stress Range
s r1/2 constant
J.J. Mecholsky, Jr., Fractography of Optical
Fibers, in ASM Engineered Materials Handbook, 4,
Ceramics and Glasses, Section 9 Failure
Analysis, (1992).
64
FSA Can Be Applied To Single Crystals
65
Slate Fracture Surface Shows Characteristic
Features
Courtesy of Prof. Yet-Ming Chiang
66
Fracture Markings Can Last 4000 Years
Titis Sarcophagus - Egypt c. 2500 BC Prof.
Greenhut
67
Many Tools Are Needed for Unified Fracture Theory
Model Scale Experiment
MO sub-atomic fractoemission Ab initio
atomic AFM Quantum nano
STM Mechanics Raman Fluorescence

Fractal Geometry
MD micro crack velocity Monte
Carlo SEM Finite diff.
Meso AFM
FEM Macro fractography
fracture mechanics

Fracture Process
68

Epoxy Fracture With Glass Fiber As Origin
69
Glass Fiber Fractures Within Epoxy Matrix
70
Fracture In Steel Shows Characteristic Features
AISI 4340
s 790MPa
71
Strain Can Be Measured In Model
a0 a / e c a / c-c
a
a
c
c'
J. Non-Crystalline Solids 260 (1999) 99-108.
72
(No Transcript)
73
Brittle Fracture Is Observed At Many Length Scales
74
Brittle Fracture Can Be Observed At Many Length
Scales
Courtesy of Dr. Darryl Butt UF
75
Characteristic Markings Are Observed on the
Fracture Surface
Hackle
Mist
Mirror
Fracture origin
76
(No Transcript)
77
(No Transcript)
78
Replicas Provide Multiple Slit Islands for
Analysis
79
Energy Geometry Are Related In The Fracture
Process
g fracture energy E Elastic Modulus D
Fractal Dimensional Increment a0 Characteristic
Structural Parameter KC Fracture Toughness
80
Fracture Behavior Appears Related To Material
Class
a0 2g/ (ED)
81

Fracture Surfaces Polished Surfaces
3BaO5SiO2

Glass Kc 0.7 I D 0.1
82
Fractal Geometry Has Been Used In Failure Analysis
  • Silicon J. Mater Res 6,1248 91
  • Ocala Chert J Mater Sci Ltrs 7, 1145 88
  • Intermetallics J. Mater. Sci.6,1856 91
  • Si3N4 J. Mater. Sci. 32 6317 97
  • Dental Glass Cer. JACerS 78 3045 95
  • Epoxy Resin Scanning 20 99 98

2 g a0 ED Kc a01/2 ED 1/2 D
c/r1
83
FSA Can Be Applied To Single Crystals
Single Crystal Silicon
84

Simulated Fracture Can Form Fracture Surface

85

Simulated Fracture Can Form Fracture Surface

86

Simulated Fracture Can Form Fracture Surface

87

Simulated Fracture Can Form Fracture Surface

88

Surface Can Be Created From MD Simulation

89

Slit Island Contour Can Be Made From 3-D Map

90
FRACTAL DIMENSION IS MEASURED ALONG CONTOUR
Log Length (A-B)
Slope -D
A
Log (Ruler Length)
B
A-B Slit Island Contour
91
Summary
  • At the atomic level, quantum mechanics describes
    the fracture process as a ring contraction
    process dictated by minimum energy and
    availability of free volume.

On the molecular scale, MD modeling describes
creation of the fracture surface.
  • On the macroscopic scale, mirror, mist hackle
    form c/r D
  • At all length scales, 2g a0 E D

92

Potentials Depend on Bond Behavior
If r lt a, then Stillinger-Weber potential is
used. If r gt a, then Born-Mayer potential is
used.

Following bond breakage at r a, each Si atom
will gain one positive charge for each lost
neighbor.
93

Standard MD Techniques Are Used

94

95
(No Transcript)
96
Fractal Dimension Varies For Different Materials
J. J. Mecholsky, Jr., Fractography, Fracture
Mechanics and Fractal Geometry An Integration,
Ceram. Trans. 64, in Fractography of Glasses and
Ceramics III, eds. J. P. Varner, V.D. Frechette,
G. D. Quinn, Am. Ceram. Soc. (1996).
97
MD Simulations Experimental Results Agree
Y. L. Tsai, T. P. Swiler , J. H. Simmons and J.
J. Mecholsky, Jr., in Computational Modelling of
Materials and Processing, J. H. Simmons, et al.
(eds) The Am. Cer. Society, Ceram. Trans. 69
(1997) 217.
98
Outline
Fundamental Questions About Fracture -
How does a crack propagate at all length
scales?
  • Analytical Tools - Quantum Mechanics (QM),
  • Molecular Dynamics (MD), ab initio, Monte Carlo,
    FEM, FD

Experimental Tools - Fractography (FSA) ,
Fracture Mechanics (FM) Fractal Analysis (FA)
Fractoemission (FE), Crack Velocity
Measurements
  • Conclusions - FSA, FM, FA, MD QM combine to
    form model of the scaled fracture process.

99
Characteristic Features Aid Failure Analysis
KC Y s (c)1/2 KBj Y s (rj)1/2 c
(a b )1/2 rj / c constant
100
(No Transcript)
101
Modeling Has to Explain Several Observations
  • Scaling
  • topography (mirror, mist, etc.)
  • self similarity (self affinity)
  • Fracto-emission
  • Chaotic (Velocity) Behavior
Write a Comment
User Comments (0)
About PowerShow.com