Mechanical Properties

•  Elastic deformation

•  Plastic deformation

•  Fracture
–  Fatigue
–  Environmental crack growth

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Crack Instabilty

•  The critical crack length for given σa

2
 
−2 K Ic
σT ac = Q  
ß σa 
ßa
r •  Sources of the critical crack
a –  Manufacturing defects
€
–  Crack growth in service
•  Fatigue
•  Corrosion (H-embrittlement)

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Crack Growth to Failure

•  Crack growth mechanisms

–  Fatigue (cyclic load)
–  Corrosive crack growth (hydrogen)

Initial feature •  Characteristic pattern:

–  Initiating flaw
•  Defect or corrosion pit
•  Nucleated defect (fatigue)
–  Crack growth to critical size
•  Identify by characteristic fracture mode
–  Corrosion: often intergranular
Crack growth –  Fatigue: “beach marks”, “striations”
–  Final failure at critical size
•  Crack length a = ac
Unstable crack •  Crack mechanism = expected unstable mode
propagation –  Usually ductile fracture

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Example: Failure of a
High-Strength Steel Spring in Seawater

pit
ductile intergranular

•  Initiation at a corrosion pit

•  Significant 2nd stage growth

–  Intergranular mode

•  Final fracture at expected ac

–  Ductile mode

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Fatigue

Crack nucleation & propagation

σmax

Stress or Strain amplitude

Stress
0

Da
σmin

ma
ge
Final Failure

ac
cu
Δσ(t) Time

mu
lat
ion
Fatigue life curve
Fatigue limit

Number of cycles
•  Phenomenology
–  Cyclic load causes failure at stresses well below ultimate strength
–  Failure is often sudden after a long period of use
–  Material grows “tired” from accumulated wear and tear
•  (Like students and professors, at the tail end of a long semester)

•  Two distinct situations:

–  Growth of a pre-existing crack
–  Nucleation and growth of a fresh crack

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Fatigue Crack Growth

•  Driving force
–  Cyclic applied stress (Δσa)
σT –  Cycles crack tip stress (ΔσT)
ß
Δσa ⇒ ΔσT
ßa •  Growth mechanism
r –  Plastic deformation irreversible
a •  Due to hardening
–  Deformation cycle grows crack
•  LeChateliere’s Principle

•  Implications
–  Crack growth rate ∝ ΔσT
•  Or ΔK = ΔσT√ρ
–  Crack grows in steps
•  Leaves marks on fracture surface
•  “fatigue striations”
•  “beach marks”

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Fatigue:
Microscopic Appearance

•  Fatigue striations in SEM

–  Not always visible - best in low-strength materials
–  Sometimes only one per cycle
•  Can compute crack growth rate and back out stress

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Fatigue Crack Growth Rate

power law
threshold

fracture: Kmax = KIc

•  Crack growth driven by ΔK
da/dn ΔK = Q(Δσ a ) a
–  No growth below threshold (ΔKth)
–  Power law at intermediate ΔK
ΔKt da
ΔK € m - “Paris Law”
= A(ΔK ) - m ~ 2 for steels
dn
ac fracture: a = ac
•  Crack tip acceleration
a
–  As a increases, ΔK increases
€ –  Crack growth rate accelerates
–  Often have very rapid growth near ac
a0 –  Crack is not safe because it is small
MSE 200A n J.W. Morris, Jr.
Fall, 2008 University of California, Berkeley
 Fatigue:
Macroscopic Appearance

•  Crankshaft fatigue in an aircraft engine

–  Pre-existing cracks
–  Visible beach marks
–  Instability and failure

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Fatigue via Crack Nucleation and Growth

Crack nucleation & propagation

σmax

Stress or Strain amplitude

Stress
0

Da
σmin

ma
ge
Final Failure

ac
cu
Time
Δσ(t)

mu
lat
ion
Fatigue life curve
Fatigue limit

Number of cycles

•  Assume no meaningful pre-existing crack

•  Cyclic deformation to failure
–  Life (cycles) decreases exponentially with cyclic stress amplitude
–  For about 90% of life, damage accumulates without cracking
–  At about 90% of life, cracks nucleate and grow to failure
•  Fatigue limit
–  No growth in 108 cycles when Δσ < Δσf

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Fatigue Damage

1 cycle 100 cycle

300 cycle 1000 cycle

P.Lukas et al. Z.Metallkde. 56 (1965) 109

•  Prior to crack nucleation

–  Increase in dislocation density
–  Reconfiguration of dislocations (well-defined dislocation “cells”)
–  Damage is internal, very difficult to detect
•  Eventual crack nucleation at well-developed cell walls

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Low-Cycle Fatigue

•  Crack nucleation, growth and failure in a Ti rod

–  Loaded a few hundred cycles in tension and torsion

MSE 200A J.W. Morris, Jr.

Fall, 2008 University of California, Berkeley
 Defeating Fatigue:
Design for Infinite Life

su
sm •  Cyclic stress below fatigue limit
s –  Asymptote on s-n curve
-sm •  Cyclic s like that in service
Δs
t –  Note s1 is a median value
sf •  s << s1 for confidence

log(n) •  Cyclic stress intensity below threshold

–  Combination of stress and crack size
power law
threshold

–  Requires inspection
fracture: Kmax = KIc

da/dn ΔK = QΔσ a
2
 ΔK 
at = Q−2  t 
 Δσ 
ΔKt
MSE 200A ΔK € J.W. Morris, Jr.
Fall, 2008 University of California, Berkeley

€
 Defeating Fatigue:
Design for Safe Life

Δs •  From s-n curve

–  Restrict allowed cycles to safe value
Δsm –  Problems:
•  counting meaningful cycles
•  no good NDE before cracking

nT n •  From crack growth curve

–  Use NDE
•  Assume worst possible flaw (a0)
ac –  Choose “safe” inspection interval (n’)
–  Use NDE
a •  Restart clock if no flaw detected
•  Retire or repair if flaw detected

a0
ni
MSE 200A n J.W. Morris, Jr.
Fall, 2008 University of California, Berkeley