You are on page 1of 240

Mechanical Behavior

of Materials

Ch. 4
Ch
Strengthening
Mechanisms

By Jae-il Jang
Strengthening
g g Mechanisms
• In this chapter, let us now discuss the various ways by which
materials ((especially
p y metals and manyy ceramics)) could be
strengthened.

• The strengthening
g g in metals and manyy ceramics relates to
processes by which dislocation motion is restricted within the
lattice.

• A number of strengthening mechanisms have been identified


and include dislocation interactions with:
– Other dislocations (strain hardening)
– G i boundaries
Grain b d i (polycrystalline
( l t lli strengthening)
t th i )
– Solute atoms (solid solution strengthening)
– Precipitations (precipitation hardening)
– Dispersoids (dispersion strengthening)

• Additional:
– Metal matrix composites,
composites texture
texture, etc
etc.
Mechanical Behavior
of Materials

Ch. 4
Ch
Strengthening
g g
Mechanisms
4.1 Strain Hardening
g

By Jae-il Jang
Strain ((Work)) Hardening
g
• Strain hardening:
– also
l called
ll d as workk hardening
h d i or cold ld working
ki
– dates back to the Bronze Age
– the first widely used strengthening mechanism for metals
– Hammering and bending of metals to achieve superior
strength!!

• Typical cold-worked commercial products include:


– Cold-drawn piano wire
– Cold-rolled sheet metal

• The hardening results from an increase in the number of


dislocation-dislocation interactions which reduces dislocation
mobility.
mobility
– As a result, larger stresses must be applied in order that
additional deformation may take place.
• The strength
g of a metal approaches
pp extremelyy high
g levels when:
– There are no dislocations present (remember the th!)
– Or when the number of dislocations is extremely high (>1010/cm2)
– Low strength levels correspond to the presence of moderate
numbers of dislocations (~ 103 to 105 /cm2).

Strength of metal crystals as a function of dislocation density


• We already discussed the strain hardening behavior of single
crystals.
Cold worked structure
• Strain hardening is attributed to the interaction of dislocations
with:
– other
th dislocations
di l ti andd
– with other barriers to their motion through the lattice.

• If slip takes place on only a single set of parallel planes:


– e.g., single crystals of fcc metals,
– only a small amount of strain hardening occurs.
occurs

• In reality, extensive easy glide is


– not a general phenomenon even with single crystals,
– not observed with polycrystalline specimens.

• Because of the mutual interference of adjacent grains in a


polycrystalline specimen:
– multiple
lti l slip
li occurs readily,
dil
– and there is appreciable strain hardening.
• Metals and alloys are “cold-worked”:
– if plastic deformation is carried out in a temperature region
and over a time interval such that the strain hardening is not
relieved.

• Such plastic deformation produces:


– an increase in the number of dislocations,
– which
hi h by
b virtue
i t off their
th i interaction
i t ti results
lt in
i a a higher
hi h state
t t
of internal stress.
– an annealed metal contains about 106 to 108 dislocations per
p
cm2
– while a severely plastically deformed metal contains about
1012 dislocations per cm2.
• A metal which has been plastically deformed to a few percent in
strain contains:
– 50,000 km or more of dislocation line in each cubic
centimeter.
ti t

• If this cubic centimeter were enlarged to the size of a large


auditorium:
– these dislocations would be seen to be arranged as an
extremely
t l fi
fine irregular
i l three-dimensional
th di i l spider’s
id ’ web,b with
ith
a mesh spacing varying from 0.1 to 1 mm.

• With this type of structure a moving dislocation can hardly avoid


intersecting other dislocations and traversing the stress field of
other dislocations.
dislocations
Observation

• Strain hardening (or cold-work) can be detected by x-ray


diff ti
diffraction.

• X-ray
X ray line broadening can be due to:
– a decrease in size of the diffraction unit (i.e. grain
fragmentation by cold-work),
– and an increase in lattice strain due to dislocation interaction.

• However
However, detailed analysis of the x-ray
x ray diffraction patterns in
terms of the structure of the cold-worked state is not easy.
• Considerable detailed knowledge on the structure of the cold-
cold
worked state has been obtained from transmission electron
microscopy (TEM).

• In the early stages of plastic deformation slip is essentially on


primaryy glide
p g planes.
p

• As deformation proceeds, cross slip takes place and


multiplication
lti li ti processes operate.
t

• The cold
cold-worked
worked structure forms high-dislocation-density
high dislocation density
regions or tangles, which soon develop into tangled networks.

• Thus, the characteristic structure of the cold-worked state is a


cellular substructure in which high-density dislocation tangles
form the cell walls.
Deformed to 10% strain:
beginning of cell formation
with dislocation tangles.

- The cell structure is usually well


developed at strains of around 10
percent.

- The cell size decreases with strain.

-Reaches a fixed size, indicating


that as strain proceeds the
dislocations sweep across the cells
and join the tangle in the cell walls.

Deformed to 50% strain:


equilibrium cell size with heavy
dislocation density in cell walls.
• The exact nature of the cold-worked structure will depend on
the material, the strain, the strain rate, and the temperature of
deformation.

• The development of a cell structure is less pronounced for:


– low temperature
– high strain-rate deformation
– materials with low stacking-fault energies (so that cross slip
is difficult).
difficult)
Stored energy
• Most of the energy expended in deforming a metal by cold-
working is converted into heat.

• However, roughly about 10% of the expended energy is stored


in the lattice as an increase in internal energy.
– mainly
i l due
d tot the
th generation
ti and d interaction
i t ti off
dislocations during cold-working
– the stored energy range from about 0.01 to 1.0 cal/g of
metal.l

• The magnitude of the stored energy increases:


– with the melting point of the metal and
– with solute additions
– with
ith decreasing
d i temperature
t t off d
deformation
f ti

• For a g
given metal the amount of stored energy
gy depends
p on:
– the type of deformation process, e.g., wire drawing vs.
tension.
Rate of strain hardening
• The rate of strain hardening can be gauged from the slope of
the flow curve.

• Increasing temperature lowers the rate of strain hardening.

• A high rate of strain hardening implies mutual obstruction of


dislocation gliding on intersecting systems.

• This can come about:


– through interaction of the stress fields of the dislocations,
– through interactions which produce sessile locks,
– through the interpenetration of one slip system by another.
• Much attention has been given to developing theories of strain
hardening based on dislocation models.

• The basic equation relating flow stress (strain hardening) to


structure is:

 0   i  Gb 
Changes in properties
• The figure shows the typical variation of strength and ductility
parameters with increasing
p g amount of cold-work.
• In addition to the change in tensile properties,
properties cold-working
cold working
produces changes in other physical properties.
– a small decrease in density of the order of a few tenths of a
percent,
– an appreciable decrease in electrical conductivity,
– and a small increase in the thermal coefficient of expansion.
expansion

• Because of the increased internal energy gy of the cold-worked


state, chemical reactivity is increased.
– leads to a general decrease in corrosion resistance and
– in
i certain
t i alloys
ll introduces
i t d the
th possibility
ibilit off stress-corrosion
t i
cracking.
Changes in grain shape
• Cold-work produces elongation of the grains in the principal
direction of working.

• Severe deformation produces a reorientation of the grains into a


preferred orientation.
Annealing
• Low temperature (<0.5
( 0.5 Tm) plastic deformation (cold
work) of a polycrystal results in:

– Increase in dislocation density (/cm2) from 106 to


108 in the annealed stage to 1012 in the heavily
deformed stage.
stage
– So, strain hardening and reduced ductility
– Changes in grain shape
– Changes in physical properties (increased
electrical resistivity)

• These properties and structures may revert back by


appropriate
i t heat
h t treatment.
t t t
• Rationale: Energy expended in cold work generates:
– heat
– Stored energy
gy in the lattice!!

• Thus: the cold work condition thermodynamically


unstable.
unstable

• Annealing process:
– Recovery
– Recrystallization
– Grain growth
(a) Recovery

• Physical properties recover: electrical and thermal conductivity.

• Some of the stored energy relieved.

• Sometimes, hardness and strength are reduced.

• No migration of high-angle grain boundaries (as during


recrystallization).
• Rearrangement of dislocations to reduce lattice energy:
Polygonization.

The excess dislocations that The rearrangement of the dislocations


remain on active slip planes after polygonization
after a crystal is bent

• Polygonization  glide and climb  Annihilation and


rearrangement of dislocations  sub-boundaries formed.

• Subgrain structure, i.e. low angle grain boundaries form.


(b) Recrystallization
• Formation of a new set of strain-free, equiaxed grains (low
dislocation density characteristic of the pre-cold-worked
pre cold worked
condition).

• Nucleation and growth process:


– Small nuclei form in the deformed matrix and grow by the
advance of high angle boundaries
– The growth involves short range diffusion of atoms across
the boundary
– High angle grain boundaries are known to have high mobility
(for >15~20°)
– Low angle grain boundaries (<10
(<10°):
): mobility is 104 smaller.
• Experiment: isothermal annealing (474K) of copper
(strain=18%).

• The curve shows maximum corresponding to large energy


release.
• This large energy release appears simultaneously with the
growth of an entirely new set of essentially strain-free grains 
recrystallization.
kinetics of recrystallization
• Study the recrystallization process: i.e. the effect of time and
temperature on the R. process.

(1) Deform a number of specimens to a certain level of strain.

(2) Anneal specimens (isothermally) for different lengths of time


at various temperatures.

(3) Measure volume fraction recrystallized, Xv , as a function of


time and temperature using e.g. optical microscopy.

(4) Plot Xv vs. log t:


Xv vs. log t:

Isothermal recrystallization
y curves for pure
p copper
pp ((98% deformation))

• Each curve represents the data for given temperature and


shows the amount of recrystallization as a function of time
time.
• The effect of increasing annealing temperature is shown clearly:
the higher the temperature, the shorter the time needed to
finish the R.
• The S-shaped curves are similar to those of nucleation and
growth processes.
processes
• Draw a horizontal line through the S-curves corresponding to
50% recrystallization.
• The intersection of this line with each of the isothermal
recrystallization curves gives the time at a given temperature
required to recrystallize half of the structure.
• Plot 1/T vs. recrystallization  (Xv=50%)
vs time for half recrystallization, (X 50%)
1   Qr 
 A exp 
  RT 

where 1/ is the rate at which 50% of the structure is


recrystallized,
R iss tthe
e gas co
constant
sta t (8
(8.37
3 J/
J/molK)
o )
Qr is called the activation energy for R.

•  can be
b any number:
b i.e.
i nott necessarily
il 50%.
50%

• Qr is:
– an empirical constant.
– cannot be related to a single process (as opposed to
activation energy for creep: Qc ~ Qd (diffusion)
Recrystallization temperature
• Recrystallization temperature, Tr
– useful
f l parameter
t
– But no physical significance

• There is no fixed Tr, below which recrystallization will not occur.

• Attempt to define Tr :
– Deform a metal to a certain strain level
– Establish
E t bli h T att which
hi h the
th deformed
d f d metal
t l fully
f ll recrystallizes
t lli
in a defined period of time, typically 1 hour.

• Tr – obviously not a material property, but it is regarded as such!


• Example:
p
• when Qr = 200,000 J/mole, Tr = 600K (material fully recrystallized
after 1 hour):

1   Qr 
 A exp 
  RT 

• Using the equation:

At Tr = 590K 600K 610K 620K


 = 2 hours
ho s 60 mins 30 mins 15 mins
Effect of strain on recrystallization
• Plot log1/ vs. 1/T for different strain !
• Example: Zr, 13% and 51% reduction of area
• Assume that  is the time for complete recrystallization

• Larger amount of cold work


accelerates recrystallization
(when annealed at constant T)

• 830K 1.6h vs. 40 h

51% 13%

Tr (1 hour) 840K 900K


• Also, slopes for 51% and 13%
are different.

• The slope is related to the


activation energy
energy, Qr.
Qr

• Thus, the activation energy


depends on the amount of
cold work (strain).
The recrystallized grain size
• Plot the grain size at the end of recrystallization (before the grain
growth) as a function of the amount of deformation.
• With decreasing
de e ing amount
mo nt of strain,
t in
grain size increases  grain size
control through cold working.

• Therefore, large amount of cold work


(strain) not only accelerates the R but
also results in a smaller grain size.

• If strain is too small, recrystallization


impossible.
poss b e
 critical amount of cold work is
necessary to allow the specimen to
recrystallize within a reasonable time.
Effectt off prior
Eff i cold-work
ld k on the
th
recrystallized grain size of brass
• The critical amount of cold work is not
a material property (depends on the
type of deformation).
• The recrystallized grain size at
the end of recrystallization
y is
independent of T.

• ii.e.
e changing temperature will
change the time to achieve
the full R  but the end
results
l in terms off the
h final
f l
grain size is the same for a
given amount of cold work.

It is possible to gain an insight for “SPD (severe plastic


deformation) process” in order to produce ultra-fine-grained
materials.
ECAP (Equal-channel angular pressing)
a.k.a.
k ECAE (Equal-channel
(E l h l angular
l extrusion)
t i )
Al

T.G.
T G Landon
Landon, Mater.
Mater Sci.
Sci Eng.
Eng A 462 (2007) 3.3
R.Z. Valiev, T.G. Langdon, Prog. in Mater. Sci., 51 (2006) 881.

HPT ((High
g Pressure Torsion)) ECAR ((ECA Rolling)
g)
Ni

M.A. Meyers et al., Prog. in Mater. Sci., 51 (2006) 427.


Purity of the metal –
effect on recrystallization
• Al cold-rolled to 80% strain.

• The impurities affect the


recrystallization temperature.

• Solute atoms retard the motion


of g
grain boundaries and hence
retard the recrystallization.

• As a result
result, Tr goes up as
impurity level goes up.
Initial grain size – effect on R.

• Nucleation of new recrystallized


y g
grains occurs preferentially
p y in
the vicinity of grain boundaries.

• The smaller the grains of the metal before cold workwork, the
greater the number of possible sites of nucleation, the smaller
the recrystallized grain size (for a given strain).
Laws of Recrystallization
(Practical Hints)
• Increasing
g the annealing
g time,, decreases the temperature
p
required for R.

• Some minimum (critical) deformation necessary to initiate R.


R

• The smaller the degree of deformation the higher T required to


initiate R.

• The
h final
f l grain size depends
d d on the
h applied
l d strain. The
h greater
the amount of strain, the smaller the grain size is.

• The final grain size depends on initial grain size. The smaller the
initial grain size, the smaller the recrystallized grain size is.
((c)) Grain g
growth
• When the recrystallization is completed, i.e. the growing crystals
h
have consumed d allll the
h strain
i energy
 the polycrystals can lower its energy by reducing its total
area
a ea o
of g
grain
a surfaces
su aces (g (grain
a bou
boundary
da y a
area).
ea)

• The role of temperature:

– e.g. Al at 500K: velocity of grain boundary = 20 mm/sec


300K: 10-33 mm/sec
• The role of time on grain size:

D ~ t0.5
Mechanical Behavior
of Materials

Ch. 4
Ch
Strengthening
g g
Mechanisms
4.2. Grain Boundary y Strengthening
g g
(Polycrystalline Strengthening)

Byy Jae-il Jang


g
intro…

Single crystal Polycrystal


intro…
• We’ve been considering the plastic deformation of single
crystals in terms of:
– the movement of dislocations and
– the basic deformation mechanisms of slip.

• The simplifications which result from the single-crystal


single crystal condition
helps in describing the deformation behavior in terms of
crystallographic structure.

• However, the single crystals are rarely used for engineering


applications because of limitations involving their strength, size,
and production.
p

• Commercial metal products invariably are made up of a


y
tremendous number of individual crystals or g
grains.

• Because of the restraining effect of the surrounding grains:


– The individual grains of the polycrystalline aggregate deform
differently compared to the single crystals
Grain boundary ?
• The boundaries between grains in a
polycrystalline aggregate

• G.B. is a region of disturbed lattice


onlyy a few atomic diameters wide.

• In general, the crystallographic


orientation changes abruptly in
passing from one grain to the next
across the grain boundary.

• a region of random misfit between the


adjoining crystal lattices.
• As the difference in orientation between the grains on each side
of the boundary decreases, the state of order in the boundary
increases.

• For the limiting case of a low-angle boundary where the


orientation difference across the boundaryy mayy be less than 1
degree, the boundary is composed of a regular array of
dislocations.
• The figure (a) schematically
illustrates the structure at a high-
g
angle grain boundary.

• Note the unorganized structure


structure,
with a few atoms belonging to
both grains, while most belong to
neither.
h

• Those atoms that belong to both


grains are called coincidence sites.
• This
h grain boundary
b d structure
contains grain-boundary
dislocations, figure (b).

• These are not mobile dislocations


producing extensive slip; rather,
rather
their chief role is that they group
together within the boundary to
f
form t or grain-boundary
a step i b d
ledge.

• As the misorientation angle of the


grain boundary increases, the
density of the ledges increases.
increases

• Grain-boundaryy ledges
g are effective
sources of dislocations.
• High-angle
High angle grain boundaries are boundaries of rather high
surface energy.

• Because of their high energy, grain boundaries serve as


preferential sites for solid-state reactions such as diffusion,
phase transformation,, and precipitation
p p p reactions.

• The high energy of a grain boundary usually results in a higher


concentration
t ti off solute
l t atoms
t att th
the b
boundary
d th
than iin th
the
interior of the grains.

• This makes it difficult to separate the pure mechanical effect of


grain boundaries on properties from an effect due to impurity
segregation.
segregation
Grain size measurement
• Grain size can be measured with a light optical microscope:
– by counting the number of grains within a given area,
– by determining the number of grains (or grain boundaries)
that intersect a given length of random line, or
– by comparing with standard-grain-size charts.

• Most grain-size measurements involve assumptions relative to


the shape and size distribution of the grains and, therefore,
must be interpreted with some degree of caution.

• Most grain-size measurements seek to correlate the interaction


of grain boundaries with a specific mechanical property.
• Thus, a measurement of the grain-boundary area per unit
volume Sv is a useful p
parameter.

• Smith and Guttman have shown that Sv can be calculated:


– without assumptions concerning grain shape and size
distribution
– from measurements of the mean number of intercepts of
random test lines with grain boundaries per unit length of
test line, NL.

Sv  2 N L

where Sv is the grain-boundary


grain boundary area per unit volume
NL is grain boundaries per unit length of test line.

* C.S. Smith & L. Guttman, Trans. AIME, vol. 197, p. 81, 1953.
• If a mean grain diameter D is required from Sv , this may be
obtained by:
- assuming constant
constant-size
size spherical grains and
- noting that each boundary is shared by two adjacent grains

grain surface area


grain boundary area per unit volume x grain volume 
2

4 ( D / 2) 2
2Sv 
4 / 3( D / 2) 3

or
3 3
D 
Sv 2 N L
• An average grain size may also be obtained from measurements
of the number of grains per unit area on a polished surface, NA.

• F
Fullman
ll h shown
has h that
h the
h mean area on a plane
l off polish
li h
passing through spheres of constant size is:

2
2 D 
A      D2
3 2 6

• Thus, measurements of NA , the reciprocal of A, give a grain size


defined as
6
D
N A

* R.L. Fullman, Trans. AIME, vol. 197, p. 447, 1953.


• A very common method of measuring grain size in the U U.S.
S is to
compare the grains at a fixed magnification with the ASTM
(American Society for Testing and Materials) grain size charts.

• The ASTM grain-size number n is related to N*, the number of


grains per square inch at a magnification of 100X by the
relationship:

N *  2 n 1

* “Determining the standard grain size,” ASTM standard E112.


• Table below compares the ASTM grain-size numbers
with several other useful measures of grain size.
Tilt and Twist Boundary

Low-angle tilt boundary Low-angle twist boundary

- Spacing between dislocation


dislocation, D
b
2 b
D 
 
sin( )
2
Energy of a simple tilt boundary
Etotal = Ecore + Estrain

Gb 2 R b
E strain  l  
 ln R~D
4(1  )  r0  
Gb 2  b 
E total  E core  l 
 ln 
4(1  )  r0 
1 
- Number of dislocations per unit area, 
D b
 Gb 2 b 
E  E total   {E core  [l    ln
[ln l ]}
b 4(1  )  r0  b
E()  E 0(A  ln ) (R d – Shockley
(Read Sh kl equation)
i )
Gb
where E0 
4(1  )
Taylor’s Model
• Taylor’s
y analysis:
y
- Condition of continuity (compatibility of strain)
- All grains undergo homogeneous strain
- Requirement for 5 independent slip systems in each grain
(xx, yy, zz, xy, xz, yz,: strain in each directio must be same
across the interface volume, xx+yy+zz=0:
interface. For constant volume
6-1=5)
- Selection of slip p systems
y
• For a polycrystal, the Schmid factor (M=1/m where m=Taylor
factor = orientation factor) varies from grain to grain and it is
necessary to determine some average m value (m = 3 3.1
1 for
fcc;2.73 for bcc). n

 
n n d i
 i 1

d   i d i   C d i
i 1

 i 1
d
m

 d 2 d
  m  m
m d d
* G.I. Taylor, J. Inst. Met., vol. 62, p.307, 1938.
Grain Size Effects
• Hall-Petch Relation:
– A general relationship between yield stress (and other
mechanical properties) and grain size was proposed by Hall
and extended by Petch.
1
 y  0  k d 2

– where y = the yield stress


0 = the “friction stress”, representing the overall
resistance
i t off th
the crystal
t l llattice
tti tto di
dislocation
l ti
movement
k = the ‘lockingg parameter”,
p , which measures the
relative hardening contribution of the grain
boundaries
d = grain diameter
* E.O. Hall, Proc. Phys. Soc. London, vol. 643, p.747, 1951.
* N.J. Petch, J. Iron Steel Inst. London, vol. 173, p. 25, 1953.
1
 y  0  k d 2

• The Hall-Petch equation was



originally based on yield-point
measurements in low-carbon steel.
k
• As grain size decreases, yield 0
strength increases!
d-1/2
1/2
• Also holds true for:
Flow, fracture stresses, and fatigue strength of
– fcc, bcc, and hcp metals,
– Solid solutions
– Precipitation and dispersion systems

• But:
ut tthe
eeexponent
po e t o
of d ca
can range
a ge from
o –1 to –1/3.
/3
• The original dislocation model for the H-P equation was based
on the concept that grain boundaries act as barriers to
dislocation motion.

• Consider a dislocation source at the center of a grain d which


send out dislocations to pile-up at the grain boundary.

• The stress at the tip of this pile-up must exceed some critical
shear stress to continue slip past the grain boundary barrier.
• Limitations of the “pile
“pile-up”
up” approach:

• While H
H-P
P equation is a very general relationship, it must
be used with some caution.

• For example,
F l if the
h H-P
H P equation
i were extrapolated
l d to a
very small grain size (say 5 nm), it would predict strength
levels close to the theoretical shear strength.

• Such extrapolation is “problematic” because:


• the
th equation
ti ffor th
the stresses
t iin a pile-up
il on which
hi h the
th
H-P equation is based were derived for large pile-ups
containing more than 50 dislocations.

• Furthermore, high SFE metals and alloy obey H-P relation:


however no pile
however, pile-ups??
ups??
• Recent observation of inverse H-P relation in
nano-crystalline
lli materials
i l

For nanocrystalline (Fe, Co)33Zr67 alloys


(H. Albes et al., Mater Sci. Forum 225-226 (1996) 769) For nanocrystalline Zn
(C.C. Koch, J. Narayan, MRS Symp. Proc. 634 (2001) B5.1.1)
 Recent observation of inverse H-P relation in nano-
crystalline
lli materials
i l

For nanocrystalline Cu
(J. Schiotz et al., Nature 391 (2003) 561)
• A more general model proposed by Li avoids the description of
the stresses at grain boundaries.

• IInstead
d concentrates on the
h iinfluence
fl off grain
i size
i on theh
dislocation density, and hence, on the yield or flow stress.

• The flow stress is given in terms of dislocation density by:

 f   0    G  b   0.5 (1)

Where 0 has the same meaning as the original H-P


equation,
ti
 is a numerical constant generally between 0.3
and 0.6,
 is the dislocation density.

• Based on the experimental observation that  is an inverse


function of the grain size.
* J.C.M. Li, Trans. AIME, vol. 227, p. 239, 1963.
• Thus,
Th  = 1/d,
1/d
 f   0    G  b   0.5
  0    G  b  d  0. 5
  0  k   d 0.5
Conrad: work hardening
g theorem of p
polycrystals
y y

 f   0  1  G  b   0.5 (1)  i  constant

   2    b  x (2) Orowan equation

x  mean free path of a dislocation

Conrad' s assumption : x   3  d
Conrad (3) d  grain size


from (2) & (3) :  (4)
 2  3  b  d

( ) :  f   0    G  b 0.5   0.5  d 0.5


(4) to (1)

where,  0 and    2   3 are constants


* H. Conrad, J. Iron Steel Inst. London, vol. 198, p. 364, 1961.
or :  f   0   4   0.5  d 0.5

where  4   0.5 is parameter k of H - P equation


where,

(+):
– this equation predicts: f ~ d-0.5
– Incorporates strain

( ):
(-):
– k does not necessarily increase when strain  increases.
Ashby Model of Deformation of a
Polycrystal
• When a single crystal is deformed in tension:
– deform on a single slip system for a good part of the
deformation
– change its orientation by lattice rotation as extension takes
place.
place

• However, individual grains in a polycrystalline specimen are not


subjected to a single uniaxial stress system when the specimen is
deformed in tension.

• In a polycrystal, because the boundaries between the deforming


crystals should remain intact, continuity must be maintained.

• Although each grain tries to deform homogeneously, the constraints


imposed by continuity cause considerable differences in the
deformation between neighboring grains and within each grain.
* M.F. Ashby, Phil. Mag., ser. 8, vol. 21, p. 399, 1970.
• Studies of the deformation in coarse
coarse-grain
grain aluminum showed
that:
– the strain in the vicinity of a grain boundary usually differs
markedly
k dl from
f the
th strain
t i att th
the center
t off th
the grain.
i

• Although the strain is continuous across boundaries


boundaries, there may
be a steep strain gradient in this region.

• As the grain size decreases and strain increases, the


deformation becomes more homogeneous.
• Because of the constraints imposed by the grain boundaries slip
occurs on several systems, even at low strains.
– this causes slip to occur on non-close-packed planes in
regions near the grain boundary.
boundary
– Slip in polycrystalline aluminum has been observed on the
{100}, {110}, and {113} planes.

• Since more slip systems are usually operative near the grain
boundary, the hardness usually will be higher near the boundary
th iin the
than th center
t off a grain
i 

• As the g
grain diameter is reduced more of the effects of grain
g
boundaries will be felt at the grain center 

• Thus
Thus, the strain hardening of a fine grain size metal will
be greater than in a coarse-grain polycrystalline
aggregate.
…Ashby Model…

• The deformation continuity leads to


more complex deformation modes
in polycrystals than in single
crystals.

• Ashby suggested a dislocation


model for polycrystal deformation.

* M.F. Ashby, Phil. Mag., ser. 8, vol. 21, p. 399, 1970.


• In Ashby’s model:

• Fi
First,
t th
the polycrystal
l t l iis
disassembled into constituent
grains (i.e. assume that there is no
constraint from the continuity)

• and allowed each grain to deform


(slip) according to Schmid’s law
independently (figure b).

• This process generates statistically


stored dislocations.
• However
However, this also generates
overlaps or voids between
the grains.

• Now, each of these


discrepancies
p is corrected byy
the introduction of
appropriate geometrically
necessary dislocations (figure
c) until the grains again fit
together.
• The reassembled polycrystal
is shown in figure d.
Ashby: grain boundary – dislocation interaction

• The role of grain boundary in plastic deformation is not


discussed byy Conrad.

• The development of voids or cracks is not permitted.

• There exists a condition of compatibility between neighboring


grains.

• The deformation in each grain must be accommodated by its


g
neighbor.

• This leads to the generation of geometrically necessary


dislocations near grain boundaries,
boundaries which are strengthening
factor.

• The smaller
smalle the grain
g ain size,
si e the stronger
st onge the effect is
is.
 explanation of H-P equation alternative to pile-up model.
Indentation Size Effect (ISE)

Indentation size effect


ardness, H

Macroscopic hardness, H0
Conventional plasticity
Ha

Characteristic length, h*
h

Indentation depth, h
- McElhaney et al
al., JJ. Mater
Mater. Res
Res. 13
13, 1300 (1997) - Ma & Clarke,
Clarke JJ. Mater
Mater. Res
Res. 10
10, 853 (1995)
Nix-Gao Model

Nix and Gao, J. Mech. Phys. Sol. 46, 411 (1998)

H  3  3 3  3 3b  S   G
Tabor’s factor Taylor
y model
3 tan 2 
G  (Assuming hemi-sphere volume)
2bh

H h*
 1
H0 h
2
3 tan θ
H 0  3 3αμb ρ s h 
*

2bρ s
Macroscopic hardness Characteristic length
Mechanical Behavior
of Materials

Ch. 4
Ch
Strengthening
g g
Mechanisms
4.3. Solid Solution Strengthening
g g

Byy Jae-il Jang


g
• There are two types of solid solutions
solutions.

• Substitutional solid solution:


– If the solute and solvent atoms are roughly similar in size,
the solute atoms will occupy lattice points in the crystal
lattice of the solvent atoms.
atoms

• Interstitial solid solutions:


– If the solute atoms are much smaller than the solvent atoms,
they occupy interstitial positions in the solvent lattice.
– Carbon,
Carbon nitrogen,
nitrogen oxygen,
oxygen hydrogen and boron are the
elements.
Solid Solutions

substitutional interstitial
Interstitial Solid Solutions

 Solute atoms in interstitial alloys must be


small in size (C,
(C H,
H N,
N O)

 Small solute atoms dissolve more readily in


transition metals than in other metals;
Fe Ti
Fe, Ti, Zr
Zr, Ni
Ni, V
V, Cr
Cr, Mn
Mn, Mo,
Mo W,
W U…
U
Substitutional Solid Solutions

 Hume-Rothery Rules

(1) Atomic size factor


- Difference in atomic radius should be
less than 15%

(2) Electro-negativity factor


- Electronegativity should be similar
each other (see Pauling’s table)

(3) Relative Valency factor


S b i i
Substitutional
l Solid
S lid Solutions
S l i
 Pauling’s electronegativity table
Solid Solution Strengthening?

- Interactions between solute atom and dislocations -

(a) Yield Point Phenomenon


(b) Strain Aging
(c) Dynamic Strain Aging
(d) Solid Solution Strengthening
(a) Yield-Point Phenomenon
• Many metals, particularly low carbon
steel, show a localized, heterogeneous
type of transition from elastic to plastic
deformation which produces a yield
point in the stress-strain curve.

• Rather than having a flow curve with a


gradual transition from elastic to plastic
behavior metals with a yield point have
behavior,
a flow curve similar to the figure.

• Load increases steadily with elastic strain


strain, drops suddenly
suddenly,
fluctuates about same approximately constant value of load,
and then rises with further strain.

• The load at which the sudden drop occurs is called the upper
yield point. The constant load is called the lower yield point. The
elongation which occurs at constant load is called the yield-point
yield point
elongation.
Elastic interaction between Disl. & solute atom
• Let us consider the edge dislocation shown below.
below

r

• The region above the slip plane contains the extra half-plane
forced between the normal lattice planes, and is in compression.
• The region below is in tension.
tension
• Interaction energy between the positive edge dislocation and a
point defect is
i 
sin i 
sin i 
sin
E i  4Gba  3
A  B
r r r
a 'a - Ei < 0:
0 attractive!
where 
a - Ei > 0: repulsive!
1. Sharp (well-defined) yield point:

Cottrell:

• Carbon or nitrogen atoms in iron readily


diffuse to the p
position of minimum
energy - just below the extra plane of
atoms in a positive edge dislocation.

• The elastic interaction is so strong that:


– the impurity atmosphere becomes
completely saturates and
– condenses into a row of atoms
along the core of the dislocation.
– Cottrell atmosphere

• Dislocations in annealed steels are locked by solute atoms (C


and N interstitials):
interstitials):.

* A.H. Cottrell & B.A. Bilby, Proc. Phys. Soc. London, vol. 62A, p. 49, 1949.
• The stress required to free dislocations (i.e. the breakaway stress
required to pull a dislocation line away from a line of solute
atoms) is higher than the stress necessary to move them
them.

• When the dislocation line is pulled free from the influence of the
solute
l atoms, slip
l can occur at a lower
l stress.

• The unlocked dislocations move at a very high speed, because


the stress required to unlock them is much higher than the
stress required to move them, until they are stopped at GBs.

• The stress concentration due to the disl. pile-up at GB unlocks


the disl. in the neighboring grains.

• Alternatively, where the dislocations are strongly pinned and


cannot breakaway, such as by carbon and nitrogen in iron, new
di l
dislocations
ti mustt be
b generated
t d to
t allow
ll th
the fl
flow stress
t tto d
drop.

• This explains
p the origin
g of the upper
pp yield
y stress (the
( drop
p in
load after yielding has begun).
• Hence, A > B
Johnston and Gilman:
(J. Appl. Phys. vol. 30, p. 129, 1959; vol. 31, p. 687, 1960)
• The dislocation locking by interstitial atoms (by Cottrell)  a
mechanism of the yield point.

• Subsequent research showed that a yield-point phenomenon


was a very general behavior that was found in such diverse
materials as LiF and Ge crystals and copper whiskers.

• In these materials the dislocation density is quite low and


impurity pinning cannot explain the effect.

• A more general theory has been advanced for all materials


which exhibit a yield drop  J-G !

• Therefore, the impurity locking becomes a special case of yield


point behavior.
• The relation between the strain rate imposed in the material by
the test and the dislocation motion is:
    b  v
where
h
 is the density of mobile dislocations
v is the average dislocation velocity

• So, using this Orowan equation:

If 1 and 2 (strain rates before and after yielding) kept constant.

1  1  b  v1   2  b  v2  2 (1)

• 1, 2 ~ mobile dislocation densities before and after yielding.

1 v2
from (1) :  (2)
 2 v1
• J-G on lithium fluoride (LiF) crystals using etch pit technique:

v  k  M (3)

where M  1 ~ 100

• From (2) and (3):

1/ M
 A  2 
   (4)
 B  1 

Therefore A > B
• Therefore,

• Note: Yield drop due to solute atmosphere (Cottrell) is sharper


than that due to dislocation multiplication (J-G).
2. Plateau in stress-strain curve:

• The deformation occurring


th
throughout
h t ththe yield-point
i ld i t
elongation is heterogeneous.

• For example, in polycrystalline low


C steels, sharp yield point is
followed by a plateau.
plateau
A :
– At the upper yield point a discrete band of deformed metal
metal,
often readily visible with the eye, appears at a stress
concentration such as a fillet.
– i.e.,
i e the deformation starts at positions of stress
concentrations as discrete bands, called Lüders bands.
– This formation of the band is coincident with the load drop to
the lower yield point.
point
• B: stress required to propagate L.B.

• The band then propagates along the


length of the specimen, causing the
yield-point elongation.

• In the usual case several bands will


form at several points of stress
concentration.
concentration

• These bands are generally at about


45° to the tensile axis.

• When L.B. cover the entire gauge


section  strain hardening “normal”
stress-strain behavior.
• Experimental observation:

• If only one LL.B.


B moves – no fluctuations in the yield
elongation zone.

• Two (or more) L.B. moves – the flow curve during the
yield-point elongation will be irregular. And each jog
corresponds to the formation of a new band. The
fluctuation of the stress is due to the formation and
interactions between (or among) LBs.

• Spread of Lüders bands – striations on the surface 


roughened surface.
• Technological Consequences

• Low-carbon
Low carbon steel sheets (automobile bodies)
– Lüders bands form during stamping.
– Sheet thickness irregularities

• Prevention:
– Use carbide formers (V(V, Nb
Nb, Ti
Ti, Al
Al, B) and form carbides (or
nitrides), i.e. remove interstitials from solid solution.
– Prestraining above yield point strain before stamping (why?:
strain aging): strains during stamping occur in the strain-
hardening region.
(b) Strain aging

• Strain aging is a type of behavior:

– Usually associated with the yield point phenomenon


– The strength of a metal is increased
– The ductility is decreased
– O heating
On h ti att a relatively
l ti l low
l temperature
t t
– After cold working
• Case I: tensile stress-strain
stress strain
curve for a metal with a sharp
yield point (a low carbon steel).
(1) stress through the yield point
(2) unload: during the unloading
stage,
g , the stress-strain curve
follows a linear path parallel to
the original elastic portion of
the curve.
(3) re-loading immediately:
on reloading within a short
time (hours), the specimen
behaves elastically to
approximately point X, and
then deforms plastically.

• Yield point does not return: Unloading-reloading does not


affect stress-strain
stress strain curve.
curve
• When the specimen is unloaded and reloaded without
appreciable delay or any heat treatment:

– On reloading the yield point does not occur


– Since the dislocations have been torn away from the
atmosphere of carbon and nitrogen atoms.
• Case II: repeat (1) and (2), then
• Re-load
l d after
f a ffew monthsh at RT: the
h specimen is allowed
ll d
to age at RT.

• Yield point returns!: the aging period has raised the stress
at which the specimen yields and, as a result, the specimen is
strengthened and hardened.
• This type of phenomenon , where a metal hardens as a result of
aging after plastic deformation,
deformation is called strain aging (or
static strain aging).

• Manifests itself by: y, UTS, 


• The yield point that reappears during strain aging is associated
with the formation of solute atom atmospheres around
dislocations.

• The reappearance of the yield point is due to:


– the diffusion of carbon and nitrogen atoms to the dislocations
during the aging period
– Formation of new atmospheres of interstitials anchoring the
dislocations.

• Because solute atoms must diffuse through the lattice in order


to accumulate around dislocations, the reappearance of the
yield point is a function of time.
time

• It also depends on the temperature, in as much as diffusion is a


temperature-dependent function.

• The higher the temperature the faster the rate at which the
yield point will reappear.
• bcc Fe: pronounced strain aging,
aging interstitial C and N

• fcc: limited strain aging effects observed in fcc metals, e.g.


austenitic steels

• IIntuition
i i + observations:
b i
– The higher the aging T, the faster the yield point reappears.
– Also,
Also at sufficiently high T,
T the dislocation
dislocation-solute
solute atoms
interaction should occur during deformation, i.e. aging
occurs during deformation  Dynamic Strain Aging!!
((c)) Dynamic
y Strain Aging
g g (DSA)
( )
• As indicated previously, the higher the temperature the faster
the
h yield
i ld point
i reappears.
• At a sufficiently high temperature, the interaction between the
impurity
pu ty atoms
ato s and
a d the
t e dislocations
d s ocat o s sshould
ou d occu
occur du
during
g
deformation.
• When aging occurs during deformation, the phenomenon is
called dynamic strain aging.

• There are manyy p physical


y manifestations of DSA:
(1) Jerky flow: Portevin - Le Chatelier effect
(2) y(T)
(3) Strain rate sensitivity (SRS)
(4) R(T) (blue brittleness)
(5) Work hardening
(1) Jerky Flow
• Inside DSA temperature range:
– plastic flow tends to become unstable.
– Irregularities in stress-strain curves: sharp drops or,
or typically,
typically
serrations.
• Phenomenon discovered and described in Al-alloys by Portevin-
Le Chatelier (’20s), thus called “Portevin-Le Chatelier Effect.”
• Serrated yielding or Jerky Flow:
RT
373 K The effect as
observed in Fe (small
473 K conc. of C)
623 K at 350K < T < 600K.
• Explanations:

– Solute atoms sufficiently mobile to migrate to moving


dislocations.

– Dislocations repeatedly break away from Cottrell


atmospheres.

– Moving dislocations readily collect atoms and become locked.

– N
New di
dislocations
l ti are being
b i created
t d with
ith the
th formations
f ti off
repeated minute yield points.

– Thus, it occurs within a specific range of temp. and strain


rate: The solute atom ‘chases’ the dislocation. The former is
affected by temp
temp. and the latter is by the applied strain rate
rate.
• First
First, it is important to note:
– that DSA tends to occur over a wide temperature range and
– that the temperature regime in which it is observed depends
on the strain rate.
• Increasing strain rate raises both the upper and lower
temperature limits associated with the DSA.
DSA

p
• For example:
• In steel deformed at a normal cross-head speed of 0.02 in. per
min., DSA effects are observed between approximately 370K
and 630K.
630K
• However, at a strain rate 106 faster, these same limits occur at
about 720 and 970K.
(2) Y(T)
• When
h DSA occurs, the
h yield
ld stress tends
d to become
b
independent of temperature.

Commercial purity Ti

The variation of the 0.2% yield stress with the temperature for commercial
purity titanium.

Note that in the dynamic strain aging interval (600K~800K) the yield
stress
t b
becomes approximately
i t l constant.
t t
(3) Strain Rate Sensitivity (SRS)

• At the same time, the flow stress becomes almost independent


of the strain rate.
• In many metals the flow stress can be related to the strain rate
by a simple power law:
  A( ) n
• Where A is a constant and the exponent n is called the strain
rate sensitivity.

• This equation can be written in:

n
 2  2 
  
 1  1 
• Taking logarithms of both sides we have:

2
logg e
1
n  SRS

log e 2
1

• In most testing machines very rapid changes in the applied


strain
t i rate
t may b be made
d during
d i a tensile
t il test.
t t

• If this is done and the corresponding values of strain rate and


flow stress (measured just before and after the rate change) are
substituted into the equation, one is able to determine a value
of n.
• If there is no DSA, the value n tends to rise linearly with the
absolute temperature
temperature.

• When DSA is observed, the value of n becomes very low in the


temperature range of strain aging:
– Flow stress becomes nearly independent of the strain rate.

Aluminum
Al min m alloys
llo aree
subject to strong DSA
effects near RT.

Note that the SRS of


aluminum alloy shows a
minimum SRS near RT.
RT
(4) R(T):
• In the center (approx.) of DSA T-range, the ductility (elongation)
becomes very low or passes through a minimum (ductility
t
troughs).
h )

• Blue Brittleness:
– At 500<T<650K in carbon steels (when a blue oxide film is
formed on the surface of a steel sample)
– Significant decrease in elongation.
– Ductility troughs, another manifestation of DSA, associated
with dislocation
dislocation-solute
solute interaction.
interaction
(5) Work hardening

• Work hardening rates can become abnormally high during DSA:


– At the same time, it can become strain-rate and temperature
dependent
– O
Onee of
o the
t e most
ost ssignificant
g ca t aspects o
of DSA
S
– Observed primarily in metals containing interstitial solutes.
• In the temperature range of
DSA, the work hardening rate
may become strain-rate
dependent.
p

• The figure shows the stress-


strain curves for three Ti
specimens deformed at 760K
at different rates.

• Note that the specimen


deformed at the intermediate
rate was subject to a much
higher degree of work
hardening than either of the
specimens deformed at a rate
20 times slower or 10 times
faster.

• This illustration suggests that, at the temperature, there is a maximum


work hardening rate corresponding to a specific strain rate
rate.
• A similar maximum work hardening rate will be observed if the
temperature is raised or lowered, provided that the strain rate is
adjusted accordingly.

• If the temperature is raised, the strain rate at which maximum work


hardening is observed also rises.
• Commercial importance of strain aging:

• (-)
• Decrease in ductility in cold-forming operations.
• Ugly stretcher marks in pressing operations.
• T
Troughs
h in
i ductility.
d tilit

• (+)
• Static strain aging: strengthening of pressed car body
components.
• Dynamic strain aging: to increase strength at moderated
temperature.
((d)) Solid Solution Strengthening
g g
• Solid solution strengthening (SSS):
– The introduction of solute atoms into in the solvent-atom
lattice produces an alloy which is stronger than the pure
metal.
eta

• Early studies of the increase in hardness resulting from solid-


solution
l ti additions
dditi showed
h d that:
th t
– the hardness increase varies directly with the difference in
the size of the solute and solvent atoms,,
– or with the change in lattice parameter resulting from
the solute addition.
• The usual result of solute additions is to raise the yield stress
and the level of the stress-strain curve as a whole as shown in
the figure below.
• Solute atoms also
frequently produce a yield-
point effect.

• The solid-solution alloy


additions affect the entire
stress-strain curve.
• Thus, solute atoms have
more influence on the
frictional resistance to
Effect of solute alloy additions on dislocation motion than on
stress-strain curve the static locking of
dislocations.
dislocations
• To investigate solid solution strengthening:

• measure CRSS, 0 (single crystal) or yield stress (polycrystals),


as a function of Cat of solute atoms
atoms.

d 0
• The measure of strength:
g
dC at
dC

• 0 is not a linear function of Cat

0 0

d 0
 constant
((X)) ((O)) dC at

Cat Cat
The effect of foreign atoms on S.S.S.

Solvent Type of foreign Typical value of Arbitrary value


atoms d0/dCat of Cat
Al (fcc) substitutional G/10 10-2 (1%)

Cu (fcc) substitutional G/20 10-2 (1%)

Fe (bcc) substitutional G/16 10-2 (1%)

Nb (bcc) substitutional G/10 10-2 (1%)

Ni (fcc) ** interstitial, C G/10 10-2 (1%)

Fe (bcc) interstitial, C 9G 10-4 (0.01%)

Nb (bcc) interstitial, N 5G 10-4 (0.01%)

Cu Self-interstitial 5G 10-4 (0.01%)


(fcc,irradiated)
• Solute atoms fall into two broad categories with respect to their
relative strengthening effect.

• Solute atoms which produce non-spherical distortions:


– such as most interstitial atoms (interstitials in bcc, self
interstitials)
– strong interactions (between solute atoms and dislocations)
– giving rise to rapid strengthening
– have a relative strengthening effect per unit concentration of
about 3G,

• Solute atoms which produce spherical distortion (symmetrical


distortion) :
– such as substitutional atoms and interstitials in fcc,
fcc
– Weak interactions
– have a relative strengthening of about G/10.
Solute atoms-dislocations interaction
• Solute atoms can interact with dislocations by the following
mechanisms:
– Elastic interaction
– Modulus interaction
– Stacking-fault interaction
– Electrical interaction
– Short-range order interaction
– Long-range
Long range order interaction

• Of these various mechanisms, the elastic interaction, modulus


interaction, and long-range order interaction are long range, i.e.
they are relatively insensitive to temperature and continue to
act to about 00.6T
6Tm.

• The other three interactions constitute short-range barriers and


only contribute strongly to the flow stress at lower temperatures.
Elastic interaction

• Elastic interaction between solute atoms and dislocations:


– arises from the mutual interaction of elastic stress fields
of misfitting solute atoms and the core of edge dislocations
– the most important contribution:

• The strengthening
g g due to elastic interaction is directlyy
proportional to the misfit of the solute.

• Substitutional solutes (which produce spherical fields and


thus do not have shear component) only impede the motion
of edge dislocations. (Remember screw dislocations have
only shear component)

• However, interstitial solutes have both dilatation and shear


components and can interact with both edge and screw
components,
dislocations.
• Stress field of a straight edge dislocation

Gby (3x 2  y 2 )
 xx 
2 (1  ) ( x 2  y 2 ) 2

Gby ( x 2  y 2 )
 yy 
2 (1  ) ( x 2  y 2 ) 2

Gbx ( x 2  y 2 )
 xyy   yyx 
2 (1  ) ( x 2  y 2 ) 2

 zz   ( xx   yy )

 xz   zx   yz   zy  0
• Stress Field of a straight screw dislocation

 xx   yy   zz   xy   yx  0

Gb y Gb sin 
 xz   zx   
2 ( x  y )
2 2
2 r

Gb x Gb cos 
 yz   zy  
2 ( x 2  y 2 ) 2 r
Elastic interaction:
(Edge) Dislocation-
Dislocation (Sym)
(S m) point defect inte
interactions
actions

• Isolated solute atoms and vacancies are centers of elastic


distortion just as are dislocations.

• Therefore, point defects and dislocations will interact elastically


and exert forces on each other  changes the elastic strain
energy of the crystal.
crystal
• To a good approximation the strains
around
d a point defect
d f distort
d the
h lattice
l ra
spherically: VH

– jjust as though
h h an elastic
l i sphere
h
(defect) of radius ra(1+) with a
volume of Vs had been forced into
ra(1+)
– a spherical hole of radius ra with a
Vs
volume of VH in an elastic continuum.

– where  is the resulting strain.

– If th
the point
i td defect
f t iis a vacancy, th
the
radius ra is the radius of the atom
normally at the lattice site,

– while if the defect is an interstitial


atom, ra corresponds to the average
radius of an empty interstitial site.
• G,  ~ the same
Vh
(the case: symmetrical distortion).
• Misfit volume (when <<1):
ra(1+) ra
Vmis  Vs  VH  4ra3 (1)

ra(1+) 4 3 4 3
Vh  ra (1  )  ra  4ra3 (2)
3

3 3
By Eshelby:
(1  ) (1  )
Vh  Vmis (3)   ( 4)
3(1  ) 3(1  )
• The misfit volume Vmis is the volume change produced by the
point defect.
• In the presence of dislocation:
• Extra work to insert the solute atom
• requires work against the stress field of the dislocation

* J.D. Eshelby, Proc. Roy. Soc. London, vol. A241, p. 376, 1957.
• Because we are concerned only with spherical distortions (i.e.
radial displacement only), the interaction only occurs with the
hydrostatic component of the dislocation stress field.
field

• Thus need to work against


g onlyy the hydrostatic
y portion
p of the
stress field: p

• Th
The interaction
i t ti energy b
between
t th
the di
dislocation
l ti and
d the
th point
i t
defect is:
E  p  Vmis (5)

* J.D. Eshelby, Proc. Roy. Soc. London, vol. A241, p. 376, 1957.
• The hydrostatic
y g dislocation at r,,  from the
stress of a edge
dislocation is:

r
 x
b

1 (1  )Gb sin
i 
p ( xx   yy   zz )   (6)
3 3(1  ) r
• So that the interaction energy is:

from (1) ~ (6)


4(1  )Gbra3 sin  sin 
E   4Gbra  
3

3(1  ) r r

• Since the equation is derived from elasticity theory, it is not


strictly correct near the core of the dislocation where linear
elasticity theory is no longer applicable.

• Thus, this expression includes only the energy external to the


point defect.

• The above equation provides only an estimate.


• A negative value of interaction energy indicates attraction
between the point defect and the dislocation.

• A positive value denotes repulsion.

• A solute atom larger than the hole (or solvent atom)


(>0) will be repelled from the compression side of a positive
edge dislocation (0<<) and will be attracted to the tension
side
id (
( <<2).
<<2 )

• An atom smaller than the hole (or solvent atom) ( <0)


will be attracted to a site on the compression side of a positive
edge dislocation.

• Similarly vacancies will be attracted to regions of compression


and interstitials will collect at regions
g of tension.
• Point defect-screw dislocation interaction:

• If the strain fields around a point defect are spherically


symmetrical:
ti l

– there is no hydrostatic stress field around a screw


dislocation which can be relaxed by the presence of a point
defect.

– such a defect will produce no net force on a screw


dislocation.

• However, some point defects (such as interstitial carbon atoms


in a bcc lattice) produce a non-spherical distortion:
– there will be an interaction energy between a screw
dislocation and the defect.
• Since the energy associated with a point defect is affected by its
proximity to a dislocation, we should expect the
concentration of defects to be different in the vicinity of
a dislocation
di l i li
line.

• The concentration in the vicinity of a dislocation C is related to


the average concentration C0 by the relationship:

E
C  C0 exp 
 kT 
• The concentration of point defects around a dislocation
exceeds the average value when E is negative and is less
than the average when E is positive.
• Remember: A negative value of interaction energy indicates
attraction between the point defect and the dislocation; a
positive value denotes repulsion.
p p
• An increased concentration of solute atoms around a dislocation
is called an impurity atmosphere or an impurity cloud.

• Interaction of dislocations with solute atoms is important for the


explanation of yield-point behavior, strain aging, and solid-
solution strengthening.
strengthening
Modulus interaction
• Case: even if the point defect has the same volume as the
atoms in the lattice:
– an interaction
i t ti withith a dislocation
di l ti can occur
– if the defect has different elastic constants than the matrix.

• A modulus interaction occurs if the presence of a solute atom


locally alters the modulus of the crystal  important when the
misfit
i fit is
i small.
ll

• If the solute has a smaller shear modulus than the matrix (i (i.e.
e if
the point defect is elastically softer)
– the energy of the strain field of the dislocation will be
reduced
d d (remember
( b that
th t E can b ib d as Gb
be described
d Gb2)
– and there will be an attraction between solute and matrix.

• if the point defect is elastically harder - a repulsion


• The modulus interaction is similar to the elastic interaction.

• But
But, because a change in shear modulus is accompanied by a
local change in bulk modulus, both edge and screw dislocations
will be subject to this interaction.
Stacking-fault interaction (Suzuki effect)

• Stacking-fault interactions arise because solute atoms


preferentially segregate to the stacking faults contained in
extended dislocations.

• For this to happen the solute must have a preferential solubility


in the hcp structure of the stacking fault (in fcc crystal).

• As the concentration of the solute within the stacking fault


increases it lowers the stacking-fault energy and increases the
separation
ti off th
the partial
ti l dislocations.
di l ti

• Therefore, the motion of the extended dislocations is made


more difficult and additional work must be done to constrict the
pair of partial dislocations.
Electrical interaction

• Electrical interaction arises from the fact that some of the


charge associated with solute atoms of dissimilar valence
remains localized around the solute atom.

• The electrical interaction of solute atoms is much weaker than


the elastic and modulus interaction.

• The interaction becomes important only when there is a large


valence difference between solute and matrix and the elastic
misfit
i fit is
i small.
ll
Short-range order interaction (Fisher effect)

• Short-range order interaction arises from the tendency for


solute atoms to arrange themselves so that they have more
than the equilibrium number of dissimilar neighbors.

• The opposite of short-range order is clustering, where like


solute atoms tend to group together in regions of the lattice.

• Strengthening occurs because the movement of a dislocation


through a region of short-range order (or clustering) reduces
th d
the degree off llocall order.
d

• This process of disordering will cause an increase in the energy


of the alloy and, to sustain the energetically unfavorable
dislocation motion, extra work (represented by the interaction
energy) must be provided
provided.

* J.C. Fisher, Acta Metallurgica, vol. 2, p. 9, 1954.


Long-range order interaction

• Long-range order interaction arises in alloys which form


superlattices (ref: Reed-Hill Chapter 11.7).

• In a superlattice there is a long-range periodic arrangement of


dissimilar atoms.

• The movement of a dislocation through a superlattice creates


regions of disorder called anti-phase boundaries (APB) because
the atoms across the slip plane have become “out of phase”
with respect to the energetically preferred superlattice structure.

• The dislocation dissociates into two pairs of ordinary


dislocations separated by APB.

• The details of the geometry and the dislocation reactions are


quite complex.
• The width of the APB is the result of a balance between the
elastic repulsion of the two dislocations of the same sign and
the energy of the APB.

• The stress required to move a dislocation through a long-range


region is:

0 
t

• where
h  is
i the
h energy off the
h APB and d
t is the spacing of the APBs.
Overall effects
• The resistance to dislocation motion that constitutes solid-
solution strengthening can come from one or more of these
factors.
• At first it might appear that dislocations would not be impeded
by the interactions from solute atoms since on the average as
manyy interactions will tend to ppromote motion as retard it.
• In a random solid solution, provided the dislocation remains
straight, there will be no net force on the dislocation since the
g
net effect of all interaction energies will be zero.

Straight dislocation line in random solid solution


• Mott and Nabarro provided the key theory by pointing out that
dislocation lines generally are not straight.

• R
Rather,
h dislocation
di l i lines
li are flexible:
fl ibl
– so that the entire line does not have to move simultaneously
and
– can take up lower energy positions by bending around
regions of high interaction energy.
– Net effect – not zero!

Flexible dislocation line

* N.F. Mott, F.R.N. Nabarro, Report on Conference on the Strength of Solids (Phys. Soc. London, 1948), p.1.
• The smallest radius of curvature that the dislocation line can
accommodate
d t tto under
d a llocall stress
t i att a sloute
l t atom
t iis:
Gb
R
2i
2

• The degree of interaction that the dislocation will have will


depend on the average spacing  of the barriers.
• For individual solute atoms distributed through the crystal lattice,
 is
i very smallll and
d is
i given
i by
b
a
 1

c 3

where a is the interatomic spacing and


c is the atomic concentration of solute
solute.

•  << R and the dislocation will move in lengths much greater


than .
Mechanical Behavior
of Materials

Ch. 4
Strengthening
Mechanisms
h i
4.4. Particle Strengthening I
- Precipitation Hardening

Byy Jae-il Jang


g
Intro…
• There are many binary alloys (two component) with limited
solubilities of the component metals in each other.
• Curves
C like
lik that
th t in
i figure
fi are quite
it common.
• This figure (= temperature
temperature-composition
composition curve = phase diagram)
shows the solubility of carbon in alpha-iron.
• Solubility relationships of the type shown in previous figure have
great practical significance:
• they make possible the precipitation hardening (or age
hardening) of metals.
• PH is an extremely important method of hardening of metals.

• This type of hardening is used most often in commercial


strengthening of nonferrous alloys, especially aluminum and
magnesium alloys.
• Summary:

• A suitable alloy is heated to a temperature at which a second


phase (usually percent in small quantities) dissolves in the more
abundant phase.

• The metal is left at this temperature until a homogenous solid


solution is attained, and then it is quenched to a lower
t
temperature
t tto create
t a supersaturated
t t d condition.
diti

• This heat
heat-treating
treating cycle is known as the solution treatment,
treatment
while the second stage, which is to be discussed, is called the
aging treatment.
Development of precipitates
• How the precipitates form and grow during precipitation is a subject of
considerable technical importance.

• The precipitation hardening process becomes very complex, particularly


in commercial alloys, which may contain more than one solute that
contributes to the age hardening.

• Fortunately, however, a great deal of research has been performed on


simple binary alloy systems containing only a singles solute.

• Among these systems are alloys formed by adding either copper, silver,
or zinc to aluminum.

• In the first of these systems, that is Al-Cu, the solute atom (copper)
has a diameter about 12% larger g than the aluminum atom.

• In the other two systems, the solute atom differs in size from that of
the solvent byy onlyy about one p
percent.
• A significant feature of precipitation hardening,
hardening even in the
binary systems, is that:
– the precipitating phase normally does not originate in its
fi l stable
final t bl form.
f

• In other words
words, the precipitate often passes through
several stages before a final stable structure appears.

• Thus, for example, an alloys of aluminum containing 4% copper


may pass through three different intermediate precipitation
g before the final stable  p
stages phase (CuAl
( 2) is attained.
An example:
aging of Al-Cu alloys at temperatures above 100°C

• If the aging temperature is raised to 100°C (373K) or above:


– tthe
e formation
o at o o of G
GP zones
o es iss not
ot tthe
eoonly
y stage o
of
precipitation hardening in an Al-Cu alloy.
– several intermediate stages,
– followed
f ll d by
b the
th appearance off the
th final
fi l stage
t  (or
( CuAl
C Al2)
phase.
• A summary of stages in the development of the precipitates in
the Al-Cu precipitation hardening alloys:

– Supersaturated alpha phase.


– GP zones found independently by Guinier and Preston in 1938).
This is the final precipitate below 100°C.
– The ” intermediate precipitate structure.
– The 
’ intermediate precipitate structure.
– The stable  phase, CuAl2, in a matrix of the equilibrium alpha phase.
• The effect of aging above 100°C is clearly shown in the figure.
figure

• A set of isothermal aging


curves
e of an n aluminum
l min m alloy,
llo
2014-T4, are shown.

• The yield stress of the metal is


plotted as a function of the
aging time in hours.

• This alloy contains, in addition


to 4.4% Cu,, 0.8% Si,, 0.8% Mn,,
and 0.5% Mg.

• The principal
Th i i l hardening
h d i agentt off this thi alloy
ll is
i its
it copper and d while
hil it is
i
not strictly a binary Al-Cu alloy, its aging behavior is largely
representative of the binary Al-Cu alloys.
• The room temperature curve:
– hardening due only to the formation of GP zones,
– shows only a slight rise in the yield stress for aging times in
excess of 100 hours
hours.
– The yield stress up to an aging time of 10,000 hours also
shows no indication of decreasing.
• On the other hand, specimens
aged at temperatures above 212°F
(100°C) show:
– a yyield stress that q
quicklyy rises
to a maximum and then falls
rapidly at longer aging times.

• Aging at temperatures above room


temperature is normally called
artificial aging.

• A significant feature of the artificially aged curves in the figure


is that they show it is possible to attain a higher strength (or
hardness) by aging above 100°C.

• This gain of strength can be retained by controlling the time


that the metal is held at the higher aging temperature.
• The maximum increase in
strength, in this set of curves,
occurs in the 300°F (149°C)
curve.
curve

• As the aging temperature is


i
increased:
d
– the maximum in these
curves moves to a shorter
ti
time and
d
– decreases in height.

• The figure above shows the difference between aging at room


temperature and aging above 100°C.

• Let us now look closely at the development of the intermediate


stages
g of aging
g g from an isothermal aging
g g curve of a binaryy Al-
Cu alloy.
• In this figure
g the diamond pyramid
py hardness (DPH)
( ) is plotted
p
against the time of aging measured in days.

- Al-4.0% Cu
- agedd for
f various
i times
i
- at 130°C.

• The development
p of GP zones:
• The initial rise of the curve that begins after the start of
aging and extends to a time of one hour,
• the plateau that follows this rise.
rise
• However,, x-rayy diffraction studies have revealed that the second
rise is accompanied by the formation of a new structure.

• O
Originally,
i i ll this
thi intermediate
i t di t structure
t t was called
ll d GP(2),
GP(2) but
b t later
l t
authors have tended to identify it by the symbol ”:
• a three-dimensional ordered phase.
p
• It also consists of plates that
li along
lie l aluminum
l i {100}
planes.

• but these plates now have a


thickness of several atomic
layers.

• the sizes or diameters of the ”


plates are larger than those of
the GP zones.
• In this specific alloy they may become at least four to five
times larger in diameter than the GP zones.
zones

• The GP zone and ” structures can be seen to overlap each


other
th ffor a short
h t partt off th
the iisothermal
th l aging
i curve.
• A greater amount of overlap
is apparent between the
next intermediate structure,
’, and the ” structure.

• The maximum hardness (or


strength)
st e gt ) o
of tthe
eaalloy
oy iss
attained when the ”
constituent is close to its
maximum amount.

• The ’ precipitate:
• has a composition that is the same or very close to that of the
stable  (CuAl2) phase and
• its crystal structure is also tetragonal like that of the  phase.

• The primary difference between these two structures is that ’


precipitates are partially coherent with the lattice of the
aluminum
l i matrix,
t i while th precipitates
hil the  i it t are incoherent.
i h t
• Figure shows that the growth of the 
’ particles is accompanied
by a loss in hardness.

• This is a result of the fact that as the particles grow in size:


– they decrease in number and
– there is a general loss in the strength of the coherency
strain.

• Eventually, with increasing aging time:


– the stable and incoherent  phase replaces the coherent ’
precipitate and
– the alloy becomes softened to well below its maximum
strength.
• Overaging:

– softening resulting from prolonged aging.

– it is connected with the continued growth of precipitate particles.


particles

– in some age-hardening alloys, it appears concurrently with the loss


of coherency by the precipitate.
precipitate

– growth will continue as long as the metal is maintained at a fixed


temperature.

– certain particles (the larger ones) continue to grow, while others


(the smaller ones) disappear.

– as aging progresses, the size of the average particle increases, but


the number of particles decreases.

– Maximum hardening is associated with an optimum small-particle


size and a corresponding large number of particles, while overaging
is associated with few relatively large particles.
• The interrelations of these structures are not well understood.

• That is, it has not always been possible to determine whether


th various
the i precipitate
i it t structures
t t evolve
l directly
di tl from
f each
h
other in sequence or whether they are separately nucleated.

• In the case of the intermediate structure ”, which replaces the


GP zones, the size of this precipitate particle is much greater
than that of the GP zones
zones.

q
• As a consequence, , manyy of the GP zones must dissolve and
release their solute atoms in order to form one of the ”
particles.

• Thus, the question of whether a ” platelet forms from a GP


zone or not is somewhat academic.
• In addition,, it was shown earlier byy Guinier that:
– depending on the degree of supersaturation,
• the final stable precipitate may form as a result of a
di t ttransformation
direct f ti ffrom ththe ’ structure
t t or
• nucleate directly from the matrix solid solution.

• Also, it has been deduced as a result of the work of Silcock,


Heal, and Hardy that:
– given the correct conditions,
– the first structure that appears may be either GP zones, ”,
or 
’

• This may be interpreted as evidence that all these structures are


capable of being independently nucleated.

* J.M. Silcock, T.J. Heal, and H.K. Hardy, J. of Inst. of Metals, vol. 82, p. 239, 1953-1954.
Theories of precipitation hardening
• The crystallographic nature of the precipitate particles that form
during
g the various stages
g of precipitation
p p are now much better
understood than they were a few years ago.

• However
However, the exact nature of the hardening process is still not
completely resolved.

• It appears that there are several hardening mechanisms, and


that what is predominant in one alloy is not necessarily
important in another.
another
• In general, however, it may be said that:
– an increase in hardness is synonymous with an increased
difficulty of moving dislocations.

• Either a dislocation must cut through the precipitate particles in


its path, or it must move between them.

• In either case,
case it can be shown that:
– a stress increase is needed to move the dislocations through
a lattice containing precipitate particles.
• System investigated:
– size of particles < a few hundred nm
– Volume fraction of particles < 10%

• The strengthening mechanism of precipitation and dispersion


h d i discussed
hardening di d together
h as “particle
“ i l hardening”.
h d i ”

• Particle hardening:

– Interactions between dislocation and particles.


– Depending on the nature of the second phase and its
crystallographic relationship to the matrix: different
interactions.
interactions
• There are several ways in which fine particles can act as
barriers to dislocations.

• They can act as strong impenetrable particles through which the


dislocations can move only by sharp changes in curvature of the
dislocation line.
line

• On the other hand, they can act as coherent particles through


which dislocations can pass, but only at stress levels much
above those required to move dislocations through the matrix
p
phase.

• Thus, second-phase particles act in two distinct ways to retard


th motion
the ti off di
dislocations.
l ti
• Two limiting cases:

– The particle will be sheared along with the matrix


– The particle will NOT be sheared along with the matrix, and
the dislocations are forced to bypass them.

• It is up to the interface between matrix and second phase


p
particle.
• Impenetrable particles:
– Dislocation bows around the particle and bypasses it.
– Leaving a dislocation loop.
– Orowan Mechanism.

* E. Orowan, Symposium on Internal Stresses (Inst. Of Metals, London, 1947), p. 451.


• Penetrable particles.
p
– Dislocation shears the particle.
– The interface has to be coherent.
– Cutting Mechanism.

• Both mechanisms to be disc


discussed
ssed in te
terms
ms of the inte
interaction
action of
a single dislocation with a linear array of particles.
• CRSS to be calculated.
• The degree of strengthening resulting from second-phase
second phase
particles depends on the distribution of particles in the ductile
matrix.

• The second phase dispersion can be described by specifying the


volume fraction
fraction, average particle diameter
diameter, and mean particle
spacing.

• These factors are all interrelated so that one factor cannot be


changed without affecting the others.

• For example, for a given volume fraction of second phase,


reducing the particle size decreases the average distance
b t
between particles.
ti l

• For a given size particle


particle, the distance between particles
decreases with an increase in the volume fraction of second
phase.
• A critical parameter of the dispersion particles is the inter
inter-
particle spacing .

• Inter-particle spacing has been subject of many interpretations


and expressed by many parameters.

• A simple expression for the linear mean free path is:

4(1  f )r

3f

where f is the volume fraction of spherical particles of radius r.


d

Interaction of a single dislocation with a row of obstacles having a distance


d between neighbors. The upper part of the sketch demonstrates the forces
acting on the dislocation at an obstacle. (Check the equations for ‘line tension’)

• Dislocation bends between particles, with a bending angle .


• In equilibrium (assumption d>>2R):

F  2T sin  (1)
F  bd (2)

•  – the increase in the applied shear stress (in the slip plane,
parallel
ll l to b).
b)

• Depending on the origin of the interaction:


– There exist a maximum force Fm which a particle can
sustain.
• If Fm is reached before the bending angle become 90°:
– Cutting: assuming this initiates plastic deformation.

  c (CRSS)

• from (2)

Fm
coherent particle : c  (3) CUTTING
bd
• If  becomes 90° before Fm is reached: Orowan mechanism.

F  2T sin  (1) F  bd ( 2)


incoherent particle : from (1) and (2),
2T
c  ( 4) Orowan
bd
(I) Cutting of spherical coherent particles

• When the particles are small and/or soft, dislocations can cut
and deform the particles.
p

Strengthening effect Mechanism

1. Chemical S. Creation of a new interface area during


cutting.
2 Modulus S.
2. S Differences in elastic moduli
moduli, affecting
the line energy of dislocation.
y
3. Coherency-Strain S. Small misfit creates an elastic strain
field.
4. Atomic Order S. Creation of an antiphase boundary.

5. Stacking-Fault S. Difference in SFE repels or attracts


dislocation.
1. Chemical S.

• When a particle is sheared by a dislocation, a step is produced


at the particle-matrix interface.

• Since this process increases the surface area of the particle


there is an associated increase in surface energy.

2 6 f
  s
 r

where
h s is
i the
th particle-matrix
ti l t i surface
f energy.

• For veryy flat coherent pprecipitates


p (GP
( zones in Al-Cu):
)
• particles consists mainly of interfaces.
• S/V is high and
• Chemical
h l S. is important.

* A. Kelly and R.B. Nicholoson, Prog. Mater. Sci., vol. 10, p.151, 1963.
2. Modulus S.

• Dislocation line energy (E) depends linearly on the local


modulus.
modulus

• Particles which have a modulus differs significantly from the


matrix will raise or lower the energy of a dislocation as it
passes through them.

• Dislocation line energy changes if GMGP (EMEP).


G 3 G 1/ 2 r
  2 [ ] [0.8  0.143 ln( )]3 / 2 r1/ 2 f 1/ 2
2 Gb b
• In many alloys, however, there is not enough of a modulus
diff
difference between
b the
h matrix i and
d the
h particle
i l to produce
d a
strong strengthening effect.

• However, a large G will be produced when voids are present


and these can contribute to the strengthening.
3. Coherency-Strain S.

• The strain field resulting from the mismatch between a particle and
th matrix
the t i wouldld be
b a source off strengthening.
t th i

• Byy Mott and Nabarro:   2Gf


• Difference in the lattice parameter: a  
a
• Eshelby:   (1  ) 
3(1  )

  0.66
•  – misfit strain.

* N.F. Mott, F.R.N. Nabarro, Report on Conference on the Strength of Solids (Phys. Soc. London, 1948), p.1.
* J.D. Eshelby, Proc. Roy. Soc. London, vol. A241, p. 376, 1957.
4. Atomic Order S.

• Strengthening due to ordered particles is responsible for the


good high
high-temperature
temperature strength of many superalloys.

• If the particles have an ordered structure then anti-phase


boundaries (APBs) are introduced when they are sheared.

2  APB 3 / 2 1/ 2 1/ 2
  ( ) r f
E b
where APB is the anti-phase boundary energy.

* L.M. Brown and R.K. Hamm, Strengthening Methods in Crystals (Elsevier. London, 1971), p.9.
5. Stacking Fault S.

• For precipitates which have stacking-fault energies significantly


different from the matrix:
– the
th interaction
i t ti b between
t th
the di
dislocations
l ti and
d particles
ti l can be
b
dominated by the local variation of fault width when glide
dislocations enter the particles.

• The increase in flow stress is proportional to the difference in


stacking fault energy between the particle and the matrix
matrix.

• By Hirsh and Kelly:


m  p 3k () ln(  m /  p )
  C( ){ }F1f 2 / 3
b E
where m and p are stacking fault energies of the matrix and the particle,
respectively, k() is the partial dislocation separating force times the separation
distance and F1 is a complex function of stacking
distance, stacking-fault
fault width and particle size.
size

* P.B. Hirsh and A. Kelly, Phil. Mag., vol. 12, p. 881, 1965.
(II) The Orowan mechanism

• proposed FOR OVERAGED INCOHERENT PRECIPITAES:

– to explain the interaction of dislocations with precipitate


particles

– that have grown large enough for dislocation segments to


be able to bend and pass between adjacent particles. It is
accordingly applicable to the later stages of aging.
• When a moving dislocation encounters precipitates it will not,
not in
general, be able to cut through them because the precipitates
are generally stronger than the matrix.
• Consequently, the dislocation will have to bow between the
precipitates and around them, leaving a dislocation loop around
tthe
e pa
particle.
tce
• The stress required for this process is approximately:

Gb


• where  is the distance of closest approach between the
particles, i.e., the bowing of the dislocation between the
precipitates is exactly analogous to the bowing out of the
dislocation segment.
segment
• When the dispersion of the particle are very fine, very large
stresses must be applied before dislocation motion can occur,
and the material exhibits a high yield strength.
• In this mechanism the dislocation is assumed to form expanding
l
loops around
d the
th precipitate
i it t particles,
ti l which
hi h cancell as in
i a
Frank-Read source.

• This cancellation permits the dislocation to continue its motion,


but leaves a dislocation ring around the particle whose stress
field adds resistance to the motion of the next dislocation.
Strengthening by a random array of particles

• Recall: 2T
c  Orowan
bd
Fm
c  Cutting
bd

• Implicit assumption: straight dislocation.


dislocation
• d to be replaced by  – average spacing between particles along
dislocation line.
• Two limiting cases:
• (1) “Weak” particles: <<90°, cutting, coherent.
Strong particles: 
• (2) “Strong” =90°,
90 , Orowan, incoherent.
• The average g, .
g spacing,
p
• Particle radius, R. 1
2
2

• Average particle radius in slip plane.    R  R (random cutting)


3
• Volume fraction occupied by particles f.

2 2
• Area by 1 particle R  2 R 2
3

2R 2
f  2 (1) d 2  area pper 1 pparticle
d

2R
d (2)
f

d
• Orowan mechanism - CRSS
  d,

2R 2R
(2)  
from (2), (3) d (2)
f f

2T 2T f
c   ( 4)
b 2bR

since T  Gb 2

2Gb f
c   2  G  b  (5)
 R
• Cutting mechanism – CRSS (not covered here in detail)
Fm R f
c   const  G 
bL b

• Cutting vs.
vs Orowan depends on
- particle radius
- interface energy

* V. Gerold and H. Haberkorn, Phys. Stat. Solidi, vol. 16, p. 675, 1966.
• Cutting vs. Orowan (cutting vs. Orowan) depends on
- particle radius R
- interface energy 

• For coherent particle,  is high, and thus only very small


particle (R < 25 nm) will be cut.

• With aging, R and  increases and Orowan decreases,


but cutting increases.

• Thus, Orowan mechanism works for the overaged


particles.
l
(Example Problem)

• An aluminum
aluminum-4%
4% copper alloy has a yield stress of 600 MPa.
Estimate the particle spacing and particle size in this alloy.
(G=27.6 GPa, b=0.25 nm)

• At this strength level we are dealing with a precipitation-


hardening g alloyy that has been aged
g beyond
y the maximum
strength. The strengthening mechanism is dislocation bypassing
of particles.

• Use Gb 4(1  f )r
0  
 3f

( is the interparticle spacing)


• 0=600/2=300 MPa.

Gb
0 

27.6 109 Pa (0.25  10 9 m)
  interparticle spacing 
3 108 Pa
 2.3 10 8 m  0.023 m

• We now know the interparticle spacing. To estimate the particle


size ((radius)) we first need to determine the volume fraction of
precipitate particles.

4(1  f )r

3f
• The maximum volume of precipitate would be obtained on slow
cooling from the melt.
• We can estimate this quantity from the phase diagram for the
aluminum-copper system.
54  4
wt %   phase ( Al )   93.5
54  0.5
4  0.5
wt %   phase (CuAl2 )   6.5
54  0.5

93.5 g
volume off    34 .6 cm 3

2.70 g / cm3
6. 5 g
volume of    1 .5 cm 3

4.43 g / cm3

volume ffraction off   0.96


volume fraction of   0.04

• Assuming that the particles are spherical:

3 f 3(0.04)(0.023)
r   0.0007 m
4(1  f ) 4(1  0.04)
Mechanical Behavior
of Materials

Ch. 4
Ch
Strengthening
g g
Mechanisms
4.5. Particle Strengthening
g g II
- Dispersion Strengthening

Byy Jae-il Jang


g
Intro…
• Two methods for producing a dispersed second phase.

• Precipitation hardening:

– Heat treatment of an appropriate two- or multi- component


system

– Precipitation from supersaturated solid solution (steels and


age-hardened non-ferrous alloys)
Intro…
• Dispersion hardening:

• Diffusion reaction techniques:


– internal oxidation alloys are suitable for this treatment
– a dilute solid solution of a base metal (e
(e.g.
g Si,
Si Al) in a more
noble metal (e.g. Cu).

– Wh
When the
th alloy
ll heated
h t d underd oxidizing
idi i conditions:
diti
• Oxygen diffuses into the alloy and produces a dispersion
of the oxide of the base metal in a matrix of the noble
metal.
• Cu + 0.1% Si – SiO2
• Cu + 0.1% Al – Al2O3

• Powder metallurgy techniques:


– mixing
i i powdered
d d metalt l and
d oxides
id tot produce
d composites
it
• Powder metallurgy technique:

– The hard particles are mixed with matrix powder and


consolidated

– There (generally) is no coherency between the second


second-
phase particles and the matrix.

– It is possible to produce an almost infinite number of


dispersion-hardened systems by mixing finely divided
metallic powders and second
second-phase
phase particles (oxides
(oxides,
carbides, nitrides, borides, etc.) and consolidating them.
• Differences between the Precipitation strengthening and
the Dispersion strengthening:

– Both involve strengthening by dispersing second phase


particles into a crystalline
p y matrix.

– Precipitation strengthening: the second phase is formed by


precipitation
i it ti from
f solid
lid solution.
l ti

– Dispersion strengthening: is a more general term and refers


also to second-phase strengthening when the second phase
is formed by a process other than solid state precipitation
(such as the formation of oxide particles in a matrix by
internal oxidation).
• Advantages:

– thermally stable at very high temperatures.

– these alloys are much more resistant to recrystallization and


grain
i growth h than
h single-phase
i l h alloys:
ll because
b off the
h finely
fi l
dispersed second-phase particles

– Because there is very little solubility of the second-phase


constituent in the matrix, the particles resist growth or
overaging to a much greater extent than the second-phase
second phase
particles in a precipitation-hardening system.
• The most common examples of this type of strengthening are
found in metals and alloys that derive their strength from the
presence of dispersed oxides.

• For example, thoria dispersed (TD)-nickel and TD-nichrome are


high-temperature
high temperature structural materials in which strength is
derived from a dispersion of ThO2 particles.

• Dispersion strengthened materials (and dispersion strengthening


mechanism) are important because such materials have
superior
p strength
g at temperatures
p that correspond
p to the
hottest sections of a gas turbine engine (1200°C).
• There are two important features of the mechanism of
dispersion strengthening.

• The first is that the particles can act as obstacles to dislocation


motion in the same way as was described for precipitates.

• In that case, the equation below may be used to describe the


strengthening process.

Gb


• Another contribution to dispersion strengthening is more
complex and cannot be explained simply in a quantitative way.

• The dispersed particles usually act as sources of dislocations,


and consequently the crystalline matrix itself becomes strain-
hardened.
hardened

• In fact, dispersion-strengthened materials will be stronger if


they have been mechanically processed (deformed at an
intermediate temperature, such as 500-600°C for TD-nickel).

• Furthermore, the second-phase particles inhibit softening


processes such as recovery and recrystallization and thereby
permit
it th
the alloys
ll to
t retain
t i their
th i strength
t th even as the
th
temperature approaches the melting temperature.
(Example)
• The tensile yield strength of pure polycrystalline silver is 0=69
MPa. Estimate the strength of polycrystalline silver that contains
5 volume
l percentt MgO
M O in
i the
th form
f off inclusions
i l i 1000Å in
i
diameter. For silver, G=43.6 GPa and b=2.9Å. The alloy is well
annealed after the oxide inclusions are formed.
Mechanical Behavior
of Materials

Ch. 4
Ch
Strengthening
g g
Mechanisms
4.6. Composite
p Strengthening
g g
(Fiber Strengthening)

Byy Jae-il Jang


g
Intro…
• Materials of high strength, and especially high strength-to-
weight ratio, can be produced by incorporating fibers in a
d il matrix.
ductile i

• The fibers must have:


– high strength and
– high elastic modulus.

• The matrix must be:


– ductile
d til andd
– non-reactive with the fibers.

• Because of their very high strength, whiskers of materials such


as Al2O3 have been used in many applications.
• Many fiber-strengthened
fiber strengthened materials use fibers of boron or
graphite or metal wires such as tungsten.

• The fibers may be long and continuous, or they may be


discontinuous.

axial matrix
transverse
fiber
• Metals and polymers have been used as matrix materials.

• Glass-fiber-reinforced polymers are the most common fiber-


strengthened materials.

• Fiber-reinforced materials are an important group of materials


generally known as composite materials
materials.

• An important distinction between fiber-strengthened and


dispersion strengthened metals is that:
dispersion-strengthened
– in fiber strengthening the high modulus fibers carry
essentially all of the load.

• The matrix serves:


– to
t transmit
t it the
th load
l d to
t th
the fib
fibers,
– to protect fibers from surface damage,
– to separate the individual fibers, and
– to blunt cracks which arise from fiber breakage.
• Because the fibers and the matrix have quite different elastic
moduli:
– a complex stress distribution will be developed when a
composite body is loaded uniaxially in the direction of the
fibers.
• A fairly rigorous analysis shows that shear stresses develop
at the fiber
fiber-matrix
matrix interface.
interface
• The distribution of this shear stress  and the axial tensile stress
in the fiber  along the length of the fiber is given in the figure
below.

The variation of stresses along a fiber


Strength and Moduli of Composites
• To a reasonable approximation the modulus and strength of a
fiber-reinforced composite are given by a rule of mixtures.

• Using the model shown in the previous figure, if we apply a


tensile force P in the direction of the fiber,, we can assume that
the fiber and the matrix will strain equally, i.e.,

e f  em  ec

P   f A f   m Am
• where Af and Am are the cross-sectional areas of fiber and
matrix
matrix.
• The average composite strength is

P
c 
Ac
where Ac  A f  Am

P  f A f  m Am
c   
Ac Ac Ac

• But Af /Ac represents the fraction of total cross section taken up


by fibers, and if we multiply by the total length of composite
this represents the volume fraction of fibers, ff.

• In a similar way Am/Ac represents fm, and ff+fm=1. Therefore,

 c   f f f   m f m   f f f  (1  f f ) m
• Since we are dealing with elastic behavior,

 f  E f e f and  m  Em em

• Therefore,
P   f A f   m Am
Pc  E f e f A f  Em em Am
• and
Pc  Ec ec Ac
• Assuming g that ec=ef=em, and transforming g from area ratio to
volume fraction, we arrive at a rule of mixtures relation for the
modulus of the composite.

Ec  E f f f  Em f m  E f f f  (1  f f ) Em
• (Example)
• Boron fibers, Ef=380 GPa, are made into a unidirectional
composite with an aluminum matrix, Em=60 60 GPa. What is the
modulus parallel to the fibers for 10 and 60 vol%?

Ec  E f f f  (1  f f ) Em

• ff=0.1
=0 1 Ec = 380(0
380(0.1)
1) + 0.9(60)
0 9(60) = 92 GPa
• ff=0.6 Ec= 380(0.6) + 0.4(60) = 252 GPa
• The uniaxial tensile stress-strain curves for unidirectional
continuous fiber composites show four stages stages.

Stages in stress-strain curves of the fiber,


p
matrix and fiber-reinforced composite
material
• In stage
g 1: both fibers and matrix undergo g elastic deformation.
Young’s modulus for the composite Ec can be determined from a
simple “rule of mixtures” addition of the elastic modulus of the
matrix and the fiber.
• In stage 2: the matrix deforms plastically and the fiber still is
elastic. This stage begins at approximately the yield strain of the
matrix
t i material
t i l without
ith t fibers.
fib

• To calculate Ec in stage 2,
2 Em must be replaced by the slope of
the matrix stress-strain curve.
 d 
Ec  E f f f    fm
 d  m
• Since the slope of the plastic part of the matrix stress-strain
curve is less than Em, the last term in the above equation is
small and we can express the composite modulus as

Ec  E f f f
• The composite responds in a quasi-elastic manner in stage 2.

• When the composite is unloaded:


– the fibers return to their original length but
– the matrix is deformed into compression.
• Stage 3 occurs when both the fibers and the matrix undergo
plastic deformation.

• Since many of the high strength-high modulus fibers like boron


are brittle, they fracture on entering stage 3, but metal wire
fibers show this region
region.

• Finally, in stage 4 the fibers fracture and the composite as a


whole soon fractures.
• Assuming that the fibers all have the same strength (which
usually is not true), the ultimate tensile strength of the
composite
p is g
given byy

 cu   fu f f   'm (1  f f )

• where fu is the ultimate tensile strength of the fiber and


• ’m is the flow stress in the matrix at a strain equal to the fiber
breaking stress.
• To obtain any benefit from the presence of fibers
fibers, the strength
of the composite must be greater than the strength of the
strain-hardened matrix, i.e.,

 cu   mu

 fu f f   'm (1  f f )   mu

• Which leads to a critical fiber volume which must be


exceeded for fiber strengthening
g g to occur.

 mu   m'
f crit 
 fu   m'
Anisotropy
• A unidirectional array of fibers in a matrix is a highly anisotropic
material.

• When such a composite is loaded at an angle to the fibers,


three strength parameters must be considered.

• The stress required to produce failure by flow parallel to the


fibers is c, given by:


 cu   fu f f   'm (1  f f )

• `m is the flow stress in the matrix at a strain equal to the fiber
breaking stress.
• Three failure criteria are plotted in the figure below
below.

• The shear stress


required to produce
failure by shear in the
matrix
t i or att th
the fib
fiber-
matrix interface is s

• s is the tensile stress


required to produce
failure of the
composite in a
direction normal to the
fib
fibers.
Variation of composite strength with
angle between fibers and tensile axis.
• The tensile stress to produce failure of the composite by
fracture of the fibers is
   c sec 2 

• If failure
a u e occurs
occu s by sshear
ea in tthe
eddirection
ect o oof tthe
e fibers
be s o
on a
plane parallel to the fibers, the failure stress is

  2 s cosec22

• Failure by flow of the matrix transverse to the fibers or tensile


failure of the interface requires a stress

   s cosec2
• The strength of a unidirectional composite falls off significantly
at small departures from the fiber orientation.

• One of the consequences of the anisotropy of fiber composites


is that they display shear coupling.

• This means that an axial stress produces shear strains and a


shear stress produces axial strains.

• In an isotropic material a uniaxially applied load produces only


axial and transverse normal strains
strains.

• But,, in a fiber-reinforced material,, in addition to those normal


strains resulting from a uniaxial load there is a shear strain.
Mechanical Behavior
of Materials

Ch. 4
Strengthening
Mechanisms
4.7. Other Strengthening

- Martensite Strengthening
- Texture Strengthening
g g

Byy Jae-il Jang


g
Martensite Strengthening
g g

• The transformation of austenite to martensite:


– is
i a diffusionless
diff i l transformation
t f ti
– happens during quenching of steel
– is one of the most common strengthening processes used in
engineering materials.

• Although martensitic transformations occur in a number of


metallurgical systems, only the alloys based on iron and carbon
show such a p pronounced strengthening
g g effect.
• Figure shows how the hardness of martensite varies with carbon
content and compares this degree of strengthening with that
achieved in dispersed aggregates of iron and cementite.
• This high strength of martensite implies that there are many
strong barriers to dislocation motion in this structure.

• Th
The complexity
l i off the
h system allows
ll for
f considerable
id bl
controversy and associated hardening mechanisms.

• It appears that there are two main contributions to the high


strength of martensite.

• Part of the high strength of martensite arises from the effective


barriers to slip provided by the high dislocation density.

• One type of martensite structure is a block martensite


containing a high dislocation density of 1011 to 1012 dislocations
pe square
per sq a e centimete
centimeter, compa
comparable
able to that in a highl
highly deformed
defo med
metal.
• The second important contribution to the strength of martensite
comes from the carbon atoms.

• The previous figure shows that the hardness of martensite is


very sensitive to carbon content below 0
0.4
4 percent
percent.

• On rapidly transforming from austenite to ferrite in the quench,


the solubility of carbon in iron is greatly reduced.

• The carbon atoms strain the ferrite lattice and this strain can be
relieved by redistribution of carbon atoms by diffusion at room
temperature.
• One result is that a strong binding is set up between
dislocations and the carbon atoms.

• We have already seen that this restricts the motion of


dislocations.

• Another result is the formation of carbon atoms clusters on


{100} planes.

• This is very similar to the GP zones discussed earlier in


conjunction with the age hardening of aluminum alloys.
alloys

• The contribution to strength


g from the barriers in the martensite
structure is essentially independent of carbon content, while the
strengthening due to carbon atom clustering and dislocation
interaction increases approximately linearly with carbon content.
content
• Actually, in a martensitic steel, all the strengthening mechanisms
are working together!

• e.g.)
g ) Byy Norstom (for
( lath martensitic steel),
),

1 / 2
 YS  i  k c  k l d  Gb 

Peierls stress
S lid l i
Solid-solution Work hardening
strengthening
GB strengthening

* L.A. Norstrom, Scand. J. Metall., vol.5, p.159, 1976


e.g.) Advanced 9-
9-12% Chromium Ferritic Steels
Extremely fine
f and complicated microstructures
 Dispersion Strengthening: MX
 Precipitation Strengthening: M23C6, Laves phase
 Solid
Solid--solution Strengthening: W, Mo
 Dislocation Strengthening: Martensitic structure
 Several
S Boundaries: Lath, Block, Packet, Prior  grain

Jang et al., JMR, vol. 22 (2007) 175


Preferred Orientation ((Texture)) Strengthening
g g

• Read textbook
- Dieter’s Chapter 6-17
- Meyers & Chawla’s Chapter 6-5.

You might also like