Lecture 9 (Dirk Mohr, ETH, Rate-Dependent Plasticity and Fracture) PDF

151-0735: Dynamic behavior of materials and structures
Lecture #9:
• Physical basis of metal plasticity
• Rate-dependent metal plasticity models
by Dirk Mohr
ETH Zurich,
Department of Mechanical and Process Engineering,
Chair of Computational Modeling of Materials in Manufacturing
© 2015
2/15/2016
D. Mohr Lecture #9 – Fall 2015 1 1 1
Physical Basis of Metal Plasticity
2/15/2016
Crystal Structures
In metals, the atoms are arranged in periodic crystal structures. The
most frequent crystal structures are the FCC, BCC and HCP
configurations:
• Face centered cubic (FCC) crystal structure
2/15/2016
Crystal Structures
• Body centered cubic (BCC) crystal structure
2/15/2016
Crystal Structures
• Hexagonally closed packed (HCP) crystal structure
2/15/2016
Crystal Structures
Metal Crystal Structure

Iron BCC
Tungsten BCC
Copper FCC
Lead FCC
Aluminum FCC
Nickel FCC
Titanium HCP
Zinc HCP
2/15/2016
Polycrystalline Microstructure
When a melted (liquid) metal begins to solidify, multiple crystals “grow”
simultaneously. Each crystal continues to grow until it impinges upon a
neighboring crystal. The result is a polycrystalline microstructure.
In metals, crystals are also often referred to as grains. The main

difference between neighboring grains is the orientation of the crystal
lattice. The size of the grains can be influenced through the
temperature history during cooling. Rapid cooling usually results in
fine grain structures, while large grains are often the result of slow
cooling.
2/15/2016
The orientation of crystal lattices can be conveniently measured using
diffraction techniques such as the EBSD (Electron Back Scatter
Diffraction) module of a Scanning Electron Microscope.
www.ebsd.info
2/15/2016
Steel Crystal Structures

• Steel Phases
Ferrite Cementite Austenite Martensite
• Pure iron with BCC • Hard & brittle ceramic • Pure iron with FCC • BCT crystal structure
crystal structure (nonmetallic solid) crystal structure supersaturated with
• Soft & ductile • Chemical compound • Exists above eutectic carbon
of Fe3C (6.7%C, temperature • Non-equilibrium phase
93.3Fe) • Soft, moderate formed at very high
• Orthorhombic crystal strength cooling rates
structure • Hard & brittle
• Two-phase microstructures • Non-equilibrium phase

Pearlite Bainite Martensite
• Fine lamellar structure • Fine non-lamellar • Highly strained ferrite
of ferrite (88wt%) and structure of ferrite and supersaturated with carbon (rapid
cementite (12 wt%) cementite cooling prevents carbon diffusion)
• Non-equilibrium phase
• Forms from austenite during
cooling
2/15/2016
Dual Phase Steels

Dual Phase (DP) steels feature a soft ferrite matrix with strong martensite
islands. The carbon content of dual phase steels varies from 0.06 to 0.15
wt%. DP steels belong to the family of so-called Advanced High Strength
Steels (AHSS) which are widely used in automotive enginerring.
10mm
ferrite martensite
2/15/2016
The interatomic distance in metals is of the order of 0.1nm at room
temperature. From metal to metal, it varies from about 0.1nm to 0.3nm.
The grain size in widely used engineering materials varies typically from
1 to 100mm. ~0.1nm
Grain size
In other words, even in small grains there
are more than 10’000 atoms along the
grain width.
EBSD picture of Inconel 600 microstructure (Eden Instruments)
2/15/2016
Strength of Metals
The real strength of metals is much lower than the ideal strength associated with
the bond between atoms!
This significant difference is attributed to dislocations. Dislocations are
crystallographic defects that can move in the crystal structure upon mechanical
loading, thereby causing plastic deformation at stresses well below the theoretical
strength.
Defect-free lattice Lattice with edge dislocation
2/15/2016
Dislocation Motion
The motion of dislocations is primarily driven by shear stresses.
t
2/15/2016
Dislocation Motion - kinematics
• Consider dislocation movement as the only agent for

plastic deformation
• Magnitude of Burgers vector (offset caused by dislocation): b
• Mobile dislocation density: m
• Average velocity of mobile dislocations: v
• Average plastic strain rate: 
Orowan (1940) relation 

   mbv
2/15/2016
Average dislocation velocity
104
102 • Orowan relation

v
1    mbv
10-2 • Constitutive function
10-4 v  v [t ,  , s ]
Critical slip resistance
10-6
Resolved shear Temperature
10-8 stress
10 100 1000
Applied shear stress [MPa]
Johnston and Gilman (1957)
2/15/2016
Slip system
The “slip” (motion) of a dislocation usually takes place on closely packed atomic
planes, i.e. hypothetical planes in a crystal that feature the highest planar density
of atoms. Such planes are called slip planes. On a slip plane, dislocation motion is
preferred along the directions of highest linear density which are called slip
directions. The ensemble of slip plane and slip direction is referred to as slip
system.
• FCC crystal
• 4 slip planes
• 3 slip directions per plane
= 12 slip systems
An BCC crystal features 48 potential slip systems.
2/15/2016
Effect of temperature on slip resistance
Temperature effect #1 (modulus)

• Critical slip resistance is governed by elastic interaction on the
atomic scale of mobile dislocation segments with microstructural
state
• Increase in temperature results in decrease of slip resistance due to
decrease of elastic moduli
Temperature effect #2 (short-range obstacles)

• With increasing temperature, the local energy barriers to dislocation
motion due to short-range obstacles can be overcome at a lower
applied stress with the help of thermal fluctuations
Key reference: Kocks, Argon and Ashby (1975), Thermodynamics and kinetics of slips
2/15/2016
Simple model
Resolved
shear stress
sth
sa
Temperature
0 c
s  sa ( microstructural state)  sth ( microstructural state)
Resistance due to athermal Resistance due to thermally activatable

obstacles (mostly long-range) obstacles (mostly short-range)
Temperature and rate-dependency

2/15/2016
Athermal barriers to slip
• Dislocation groups
• Large incoherent precipitates
• Inclusions
• Grain boundaries
2/15/2016
Thermally activated short-range barriers to slip

An increase in temperature corresponds to an increase of the vibration of atoms.
As a result, a lower shear stress is required to overcome short-range barriers to
dislocation motion such as
Mobile Forest
dislocation dislocation
Solute atoms
Mobile dislocation meets
v v forest dislocation
• Forest dislocations (dominant effect in FCC and HCP)

interactions with other non-coplanar dislocations intersecting the slip plane
• Peierls-Nabarro forces (dominant in BCC)

The force needed to move a dislocation is related to the shear modulus and the
atomic bond strength which both decreas as a function of temperature
• Solute atoms
2/15/2016
Activation free enthalpy as function of stress

• “Thermal shear stress”: Resolved shear stress in excess of
athermal resistance
t
t th  t  sa
t th
sth
sa
0 c
• Required activation free enthalpy as function of “thermal shear
stress”
t th
p q
sth
t th  sth t th  0  t  
G  F 1   th  
  sth  
F
Distance
2/15/2016
Statistical mechanics argument

• Probability that a thermal fluctuation at temperature  can supply the
energy G required for a shear increment is p B
 G 
pB  exp   G
 k B  
• Rate at which dislocations overcome these obstacles is

 G 
f 0 exp  
 k B 
f 0 is a characteristic frequency of the order of 1012 s 1
• Average advance of a mobile dislocation segment: l

• Average velocity of mobile dislocations
 G 
v  l f 0 exp  
 k B 
2/15/2016
Thermally activatable short-range barriers to slip

• Average plastic shearing strain rate
 G 
  0 exp   with 0   mbl f 0
 k B 
 0 
G  k B ln  
  
• Recall
p q
 t  
G  F 1   th  
  sth  
• Temperature and rate sensitivity of the thermal shear stress
p 1/ q
 t th   k B  0 
   1   ln 
 sth   F   
2/15/2016
Temperature and strain rate dependence
of resolved shear stress
p 1/ q
 t th   k B  0 
t  t th  sa with    1   ln  
 sth   F   
t
. 1  2  3
 1
sth 2
3
sa
1  2 3 temperature
• t increases with strain rate

• t decreases with temperature
2/15/2016
Summary
• Thermally activated obstacles to dislocation motion are

responsible for the effect of strain-rate and temperature on
the flow stress in metals
• Short range obstacles mostly fall into the category of thermally
activated barriers, while long range obstacles are mostly athermal
• The Peierls-Nabarro stress is the rate-controlling mechanism in
BCC metals (e.g. steel with high ferrite content)
• Forest dislocations control the rate sensitivity in FCC metals (e.g.
aluminum)
• Basic statistical mechanics arguments suggest that the flow
stresses increases as a function of strain rate and decreases as a
function of the temperature
2/15/2016
Rate-dependent metal plasticity models
2/15/2016
MODELING FRAMEWORKS
1. Mechanism-based polycrystal plasticity models, e.g.

• Balasubramanian-Anand model
2. Mechanism-inspired macroscopic models, e.g.
• Zerilli-Armstrong model
• Rusinek-Klepaczko model
3. Empirical models, e.g.
• Cowper-Symonds model
• Johnson-Cook model
• Roth-Mohr model
2/15/2016
Polycrystal Plasticity
In polycrystal plasticity, the material response is evaluated at two length
scales. At the microscopic level, the heterogeneous polycrystalline
microstructure is taken into account and the stress-strain response of
each slip system is computed.
At the macroscopic level, the material is considered as a
homogeneous solid. The special feature of polycrystal plasticity models
is that the macroscopic material response is estimated based on the
analysis of a representative volume of the polycrystalline
microstructure.
This estimation requires not only an accurate mechanical models of
individual crystal, but also micro-macro relationships (homogenization
models) that relate the macroscopic stress and strain tensor to their
counterparts at the microscopic level.
2/15/2016
Model by Balasubramanian & Anand (2002)
Balasubramanian and Anand (2002) make use of a Taylor model, i.e.

they assume that the stress field is uniform at the microscopic level and
equal to the applied macroscopic stress.
Their representative microstructure model includes 400 grains of distinct
orientations which are distributed in a spatially isotropic manner.
n
m
2/15/2016
Model by Balasubramanian & Anand (2002)
• Kinematics
F  FeF p
• Constitutive equation for stress
σ  C : Ee  A (   0 ) 
• Flow rule
Lp   p m  n

2/15/2016
POLYCRYSTAL PLASTICITY
• Resolved shear stress for slip system  t

n
t   σ : m  n 
m
• Constitutive function for slip system 
p  p [t  ,  , s ]
• Evolution of slip system resistances

(self-hardening & latent-hardening)
s   h p


2/15/2016
• Variation of elastic lattice moduli for pure Al (FCC crystal) over
temperature
Simmons and Wang (1971)
2/15/2016
• FCC structure with twelve slip systems (per grain)
2/15/2016
• Selected results
Experimental data from Senseny et al. (1978)
2/15/2016
• Strain-rate jump tests on Al 1100-O
Experimental data from Senseny et al. (1978) Simulations by Balasubramanian & Anand
(2002)
2/15/2016
MODELING FRAMEWORKS
• Mechanism-based polycrystal plasticity models, e.g.

• Mechanism-inspired macroscopic models, e.g.
• Empirical models, e.g.
• Cowper-Symonds model
• Roth-Mohr model
2/15/2016
MECHANISM-INSPIRED CONSTITUTIVE MODELS

Recall the stress/strain rate/temperature relationship for a single slip system:
p 1/ q
 t th   k B  0 
t  t th  sa with    1   ln  
 sth   F   
In mechanism-inspired models, the expressions for the resolved shear stress

and plastic shearing rate in the single crystal model are substituted by
equivalent stress and strain measure:
1 
t s:s   2ε : ε
p p
ε p  
2 σ
Examples for mechanism-inspired models are:
‒ Zerilli-Armstrong model (1990)

‒ Rusinek-Klepaczko model (2002)
2/15/2016
MODELING FRAMEWORKS
• Mechanism-based polycrystal plasticity models, e.g.

• Mechanism-inspired macroscopic models, e.g.
• Empirical models, e.g.
• Roth-Mohr model
2/15/2016
EMIRICAL CONSTITUTIVE MODELS
• Elasticity σ  C : (ε  ε p )  A (   0 ) 
• Yield condition f  k 0 with k  f [ p ,  ]g[ p ]h[ ]

• Flow rule ε p   p
σ
+ loading/unloading conditions
2/15/2016
EMIRICAL CONSTITUTIVE MODELS

Multiplicative decomposition of the yield stress
k  f [ p ]g[ p ]h[ ]
strain hardening temperature sensitivity

strain rate sensitivity
Examples:
‒ Johnson and Cook model (1983)

‒ Klopp-Clifton-Shawki model (1985)
‒ Roth-Mohr model (2014)
2/15/2016
JOHNSON-COOK MODEL
k  f [ p ]g[ p ]h[ ]

• Strain hardening: f [ p ]   0  B pn
  p 
• Strain rate sensitivity: g[ p ]  1  C ln 

 0 
m
   r 
• Temperature sensitivity: h[ ]  1  
    
 m r 
2/15/2016
JOHNSON-COOK MODEL
• Hopkinson bar test results
2/15/2016
TAYLOR IMPACT TEST

• Model validation based on Taylor impact test
v=190m/s
Chung and Taejon Zerilli and Armstrong (1987)
2/15/2016
Roth-Mohr Plasticity Model

k  f [ p ]g[ p ]h[ ]

• Strain hardening: f [ p ]   A pl n

 
  0    1    k0  Q 1  e   pl

Combined Swift-Voce
 
 pl 
• Strain rate sensitivity: g[ p ]  1  C ln 
 0 
same as Johnson-Cook
m
• Temperature sensitivity: h[ ]  1      r 

 
 
 m  r 
• Non-associated flow rule (only associated flow considered in class)
2/15/2016

k
• Explicit temperature-displacement analysis dT   d p
 Cv
• Temperature is treated as internal state variable!

dT   [ p ] k  d p 0 ( isothermal )    1 ( adiabatic )
 Cv
0 for  p  it
 

 [ p ]  
 p  it  3 a  2  p  it 
2
for it   p  a

 a  it 3
1 for a   p

Isothermal adiabatic
• For ease of model calibration conditions conditions
it  0
 Explicit analysis with temperature

softening capability it a
2/15/2016
• Illustration of material response for different ω at different

strain rates
ω= 0( isothermal) ω= ω[ε. pl ] ω= 1 (adiabatic)
2/15/2016
Calibration experiment
• Notched Tension R=20mm (NT20)
2/15/2016

• Notched Tensile R=20 - model calibration
NT20 - slow NT20 - NT20 - fast

intermediate
A 0 n k0 Q   C 0 Troom Tmelt m

[MPa ]  []  [] [MPa ] [MPa ] [] [] [] []  [K ]  [K ]  [] 
1031 0.00128 0.2 350 324.25 25.92 0.65 0.0137

0.64 0.0148 0.00116
0.0015 293 1674
1674 0.8567
0.921
E  cp   a h
[GPa ] []  [ J / kg K ] [kg / m3 ] [1 / s] []
210 0.33 420 7850 1.379 0.9
2/15/2016 (185)
D. Mohr
(0.33) (420) (7850)
Lecture #9 – Fall 2015
(1.343) 48 (0.9)
48 48
Hosford-Coulomb Ductile Fracture Model

(recall from Lecture #7)
• General form
von Mises equivalent plastic
strain to fracture
f
 f   HC [ ,  , a, b, c ]
Stress 3 material
triaxiality Lode angle parameters
parameter
• Detailed expressions 
1
 1 c  n

 HC  b 
 HC
g [ ,  ] 
g HC   | f I  f II |  | f II  f III |  | f I  f III | 
1
1
2
a 1
2
a 1
2
a a
 c2  f I  f III 
2   2   2  
f I [ ]  cos  (1   ) f II [ ]  cos  (3   ) f 3 [ ]   cos  (1   )
3 6  3 6  3 6 
2/15/2016
Incorporation of the effect of strain rate

The effect of strain rate and temperature on ductile fracture is still an ongoing
research topic today. In the absence of new results, it is recommended to follow
the Johnson-Cook approach by multiplying the strain to fracture for isothermal
static loading with strain rate and temperature dependent functions:
 f   HC [ ,  ]g[ ]h[ ]
 1 for  p  0

g   p 
bg [-]
1   ln    for  p  0
  0
with 0 taken from the plasticity model
2/15/2016
Rate-dependent Fracture Model

The resulting rate-dependent Hosford-Coulomb model (Mohr and Roth, 2014)
then predicts an increase in ductility as a function of the strain rate.
2/15/2016
Reading Materials for Lecture #9

• Kocks, Argon and Ashby (1975), Thermodynamics and kinetics of slips
• M.A. Meyers, Dynamic behavior of Materials
• C. Roth and D. Mohr (2015), “Ductile fracture experiments with locally
proportional loading histories”, Int. J. Plasticity,
http://www.sciencedirect.com/science/article/pii/S0749641915001412
2/15/2016

Lecture 9 (Dirk Mohr, ETH, Rate-Dependent Plasticity and Fracture) PDF

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Lecture 9 (Dirk Mohr, ETH, Rate-Dependent Plasticity and Fracture) PDF

Hochgeladen von

Copyright:

Verfügbare Formate

151-0735: Dynamic behavior of materials and structures

Physical Basis of Metal Plasticity

• Face centered cubic (FCC) crystal structure

• Body centered cubic (BCC) crystal structure

• Hexagonally closed packed (HCP) crystal structure

Metal Crystal Structure

In metals, crystals are also often referred to as grains. The main

Steel Crystal Structures

• Two-phase microstructures • Non-equilibrium phase

Dual Phase Steels

Defect-free lattice Lattice with edge dislocation

Dislocation Motion - kinematics

• Consider dislocation movement as the only agent for

Orowan (1940) relation 

Average dislocation velocity

102 • Orowan relation

10-2 • Constitutive function

An BCC crystal features 48 potential slip systems.

Effect of temperature on slip resistance

Temperature effect #1 (modulus)

Temperature effect #2 (short-range obstacles)

s  sa ( microstructural state)  sth ( microstructural state)

Resistance due to athermal Resistance due to thermally activatable

Temperature and rate-dependency

Athermal barriers to slip

Thermally activated short-range barriers to slip

• Forest dislocations (dominant effect in FCC and HCP)

• Peierls-Nabarro forces (dominant in BCC)

Activation free enthalpy as function of stress

Statistical mechanics argument

• Rate at which dislocations overcome these obstacles is

• Average advance of a mobile dislocation segment: l

Thermally activatable short-range barriers to slip

• t increases with strain rate

• Thermally activated obstacles to dislocation motion are

Rate-dependent metal plasticity models

1. Mechanism-based polycrystal plasticity models, e.g.

Balasubramanian and Anand (2002) make use of a Taylor model, i.e.

• Constitutive equation for stress

• Resolved shear stress for slip system  t

• Evolution of slip system resistances

s   h p

Simmons and Wang (1971)

• FCC structure with twelve slip systems (per grain)

Experimental data from Senseny et al. (1978)

• Strain-rate jump tests on Al 1100-O

• Mechanism-based polycrystal plasticity models, e.g.

MECHANISM-INSPIRED CONSTITUTIVE MODELS

In mechanism-inspired models, the expressions for the resolved shear stress

‒ Zerilli-Armstrong model (1990)

• Mechanism-based polycrystal plasticity models, e.g.

EMIRICAL CONSTITUTIVE MODELS

• Yield condition f  k 0 with k  f [ p ,  ]g[ p ]h[ ]

EMIRICAL CONSTITUTIVE MODELS

strain hardening temperature sensitivity

‒ Johnson and Cook model (1983)

strain hardening temperature sensitivity

TAYLOR IMPACT TEST

Chung and Taejon Zerilli and Armstrong (1987)

Roth-Mohr Plasticity Model

strain hardening temperature sensitivity

• Strain hardening: f [ p ]   A pl n

• Temperature sensitivity: h[ ]  1      r 