CEE 770 Meeting 11 Objectives of This Meeting

Introduce Issues and Approaches for Non-Linear Fracture Mechanics (NLFM) What makes NLFM more difficult than LEFM? NLFM Approaches under Development A NLFM Approach for Thin Metallics: CTOAc A NLFM Approach for Many Other Situations: Cohesive Fracture Mechanics
215

The Origin of the Cohesive Crack Concept

The Dugdale (1960)-Barenblatt (1962) Model

y
yld

The effective crack is being opened by far-field stresses that cause K = Kff Over the distance, , the effective crack is being closed by local traction, yld, which causes a negative K = -K

x Set Kff = -K, and (1) the effective crack will indeed be closed (2) the singular field disappears, and (3) you can solve for :

Physical Crack Effective Crack

K ff = 8 yld

216

The Generalized D-B Model

No elastic singularity Arbitrary, softening, cohesive zone Toughness is area under traction-COD curve Planar process zone y ft

y ft

Gf c

x Fictitious Crack Tip

Cohesive Traction, y(x)

True Crack

Process Zone
217

General, 3D Coupled Cohesive Zone Model

t = t t c

t a

1.2 1.0 0.8 0.6 0.4 0.2 0 0 1

(t c , c )
Gf

= c
Curve Shape Fixed c, tc, Gf (2 of 3 are independent) Curve Shape Not Critical For Ductile Metals
2 = 2 ( s21 + s22 ) + n 2 2 2 t = 2 (t s + t ) + t s2 n 1

Mode I: t = t n ; = n
218

Why a Cohesive Fracture Model? (for practically every situation!)

Seems to mimic many process zone effects in many materials. Isolates fracture process model from surrounding continuum constitutive model. Condenses the process zone onto a curve (2D) or surface (3D). The fracture process model is simple: generalized traction vs generalized crack displacement. A step towards micromechanical modeling of the process zone: introduces a local length scale c. Easy to implement in the FEM using so-called interface elements.

219

A Possible Explanation for Cohesive Behavior in Cementitious Composites

Crack Growth

Stress-Free Macrocrack

Damaging Process Zone

220

Eg: The Fictitious Crack Model

Hillerborg et al. (1976)
Tension Softening Relationship In Planar Process Zone

Stress-Free Macrocrack
c

ft

Cohesive Stress

K 2 Ic Gf = E

221

A Cohesive Fracture View of Crack Growth in Many Metals

Far-Field Loading

Larger Inclusions Voids

a
Smaller Inclusions

20-100 m

Nucleation of Voids at Cu, Fe... Inclusions Void Coalescence to Form New Crack Surfaces Narrow Layer of Damage

222

Sometimes Cohesive Mechanism Is Obvious

SEM micrographs of damage Micromechanisms in solid rocket propellant. From Bencher et al. (1995). Arrow shows direction of crack growth.

The propellant is a particle filled elastomerrubber filled with metallic particles.

223

2-D, Quadratic Order, Interface Element Used in FRANC2D

tn , n

Q8 or T6 Interface-6 Q8 or T6

ts , s
6-node Interface Element Mixed Modes (1,2) Zero Initial Thickness Isoparametric Formulation

224

3-D, Linear Order, Interface Elements

tn , n
Hex-8

t2 , 2

t1, 1
Interface-8 Hex-8 8-node Interface Element Mixed Modes (1,2,3) Zero Initial Thickness Isoparametric Formulation

225

Example: A NLFM Driving Force Problem-Cohesive Crack Model in Concrete Beam with FRANC2D/L

226

Example: A NLFM Resisting Force problem Coupled Cohesive Zone Model for Fatigue Crack Initiation tp
Tvergaard-Hutchinson Coupled Cohesive Zone Model tp - Strength k0 c - Critical displacement Gc k0 - Initial Stiffness c Gc - Critical Energy Relative Displacement Release Rate Interface Finite Elements Placed between all grains Follow CCZM
Interface elements Traction

No explicit introduction of cracks. Interface elements allow grain boundaries to decohere to form cracks.
227

Fatigue Loading Simulation

n
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.0 1.0 2.0 3.0

0.5 mm

Applied Strain (%)

Cycle

Hill (MPa) E1 E2 y x G 72,000 5% 40,000 5% 10,000 5% G tp k0 c

CZZM 250 5% Nm/m2 500 5% MPa 4e8 1 m

xx yld 500 5% yy yld 450 5% xy yld 400 5%

228

Fatigue Simulation Results

500

y x

A
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0
k

yy
100 A C

500

Applied Strain (%)

yy
100

Opening
De Clo co sed he si o n

0.5

Cycle

1.5
50

Cr ac

yy (MPa)
B
0
229

Fatigue Simulation Result

y x

Applied Strain (%)

0.6 0.4 0.2 0 0 1 2 Cycle

3 2 1

2
3

Opening
De Clo co sed he si o n

500

Cr ac

yy (MPa)
100

230