Title: MicrostructureProperties: II Elastic Effects, Interfaces
1Microstructure-Properties IIElastic Effects,
Interfaces
- 27-302
- Lecture 7
- Fall, 2002
- Prof. A. D. Rollett
2Materials Tetrahedron
Processing
Performance
Properties
Microstructure
3Objective
- The objective of this lecture is to show how
important (a) elastic effects are in controlling
precipitation, and (b) the variety of interface
structures that occur, and their importance in
precipitation. - More specifically, this lecture examines the role
of the interface (coherent vs. incoherent) in
precipitate morphology and growth. - The main concepts are the misfit strain between
two lattices at an interface that determines a
dislocation density, and the (bulk) misfit
parameter that determines the elastic energy
associated with a precipitate.
4References
- Phase transformations in metals and alloys, D.A.
Porter, K.E. Easterling, Chapman Hall.
Chapter 3 is most relevant to this lecture. - Interfaces in Materials (1997), James M. Howe,
Wiley Interscience. - Materials Principles Practice, Butterworth
Heinemann, Edited by C. Newey G. Weaver.
5Elasticity and Interface Structure
- Why consider elasticity and interface structure
together? Both aspects exert a strong influence
over not just the shape of the new phase that
appears during a phase transformation but also
which phase appears in the first place. - The fact that a new phase has a different
composition means that its lattice has a
different set of repeat distances (lattice
parameters), even if the crystal structure is the
same as the parent phase. - These differences mean (a) different elastic
moduli between parent and product phase and (b) a
mismatch in atomic positions at a parent-product
interface.
6Basic results
- For small precipitates (lt 5nm), precipitates are
usually spherical with coherent interfaces in
order to minimize surface energy. - For intermediate sizes, precipitates are often
plates or needles in order to minimize surface
energy in situations where one plane or direction
is atomically similar between parent and product
phases. - Large precipitates (gt 1µm) are often spherical
with incoherent interfaces in order to minimize
volumetric free energy.
7Coherence at interfaces
- Coherent/semi-coherent/incoherent interfaces
these terms are based on the degree of atomic
matching across the interface. - Coherent interface means an interface in which
the atoms match up on a 1-to-1 basis (even if
some elastic strain is present). - Incoherent interface means an interface in which
the atomic structure is disordered. - Semi-coherent interface means an interface in
which the atoms match up, but only on a local
basis, with defects (dislocations) in between.
8Homophase vs. Heterophase
- There is a useful comparison that can be made
between grain boundaries (homophase) and
interphase boundaries (heterophase).Structure
G.B. Interface - atoms no boundary coherentmatch (or, S3
coherent twin in fcc) interface - dislocations low angle g.b. semi- coherent
- disordered high angle g.b. incoherent
- Remember for a grain boundary to exist, there
must be a difference in the lattice position
(rotationally) between the two grains. An
interface can exist even when the lattices are
the same structure and in the same (rotational)
position because of the chemical difference.
9LAGB to HAGB Transition
- LAGB steep risewith angle.HAGB plateau
Disordered Structure
Dislocation Structure
10Read-Shockley model
- Start with a symmetric tilt boundary composed of
a wall of infinitely straight, parallel edge
dislocations (e.g. based on a 100, 111 or 110
rotation axis with the planes symmetrically
disposed). - Dislocation density (L-1) given by1/D
2sin(q/2)/b ? q/b for small angles.
D
11Read-Shockley, contd.
- For an infinite array of edge dislocations the
long-range stress field depends on the spacing.
Therefore given the dislocation density and the
core energy of the dislocations, the energy of
the wall (boundary) is estimated (r0 sets the
core energy of the dislocation) ggb E0 q(A0
- lnq), whereE0 µb/4p(1-n) A0 1
ln(b/2pr0)
12LAGB experimental results
- Experimental results on copper.
Gjostein Rhines, Acta metall. 7, 319 (1959)
13Low Angle Grain Boundary Energy Yang et al.
Scripta Materiala (2001) 44 2735-2740
High
117
105
113
205
215
335
203
Low
8411
323
727
Measurements of low angle grainboundary energy
in 99.98Al
? vs.
14High angle g.b. structure
- High angle boundaries have a disordered
structure. - Bubble rafts provide a useful example.
- Disordered structure results in a high energy.
Low angle boundarywith dislocation structure
15Energy of High Angle Boundaries
- No universal theory exists to describe the energy
of HAGBs. - Abundant experimental evidence for special
boundaries at (a small number) of certain
orientations. - Good fit for special boundaries based on good fit
of a certain fraction of the atoms at the
interface. - Mathematically, special orientations are easier
to characterize in terms of coincidence of
lattice points between the two lattices (by
imagining that they interpenetrate) leading to
the Coincident Site Lattice (CSL). - Each special point (in misorientation space)
expected to have a cusp in energy, similar to
zero-boundary case but with non-zero energy at
the bottom of the cusp. - Special boundaries defined by a sigma number
which is the reciprocal of the fraction of
lattice points (not atoms!) that coincide. For
cubic materials, this number is always odd, so we
have S1, S3, S5, S7.
16Exptl. vs. Calculated G.B. energies for 99.89 Al
lt100gtTilts
Twin
lt110gtTilts
Hasson Goux
17Dislocation models of HAGBs
- Boundaries near CSL points expected to exhibit
dislocation networks, which is observed.
lt100gt twist boundariesin gold.
S5 twist Boundary in gold, showing dislocation
structure
18Heterophase boundaries coherent interfaces
- Coherent interfaces have perfect atomic matching
at the boundary. - See figures 3.32, 3.33 in PE.
- The indices of the planes comprising the boundary
do not have to be the same in each phase. - In general, a coherent interface is based on an
orientation relationship. This relationship is
specified crystallographically in terms of a pair
of planes and directions as in hklA//hklB
with ltuvwgtA//ltuvwgtB. - For example, the hcp k phase precipitate in fcc
copper has an almost perfect match through the
orientation relationship, 111fcc//(0001)hcp and
lt110gtfcc//lt11-20gthcp. The interfacial energy is
estimated to be as low as 1 mJ.m-2. - Even in the case of perfect atomic matching,
there is always a chemical contribution to the
interface energy.
19Strained interfaces
- Unlike the case of grain boundaries, there is an
important elastic aspect of coherent interfaces. - Small differences in lattice parameter are
accommodated by elastic strain. - Given lattice parameters specified as interplanar
spacings, dA and dB, the misfit parameter, d is
given by the following simple formula d
(dB - dA)/ dA - See figs. 3.34, 3.35.
- We shall see later, that the misfit that can be
accommodated by elastic strain is limited.
20Semi-coherent interfaces
- The logical next step in typing interfaces is to
note that too-large misfit strains can be
accommodated (i.e. lower energy interfaces
constructed) by replacing uniform elastic strains
with dislocations (which localizes the strain
into the dislocation cores), fig. 3.35. - The dislocation spacing in 1D is given by D
dB/ d .For small enough misfits, this can be
written as D d/ d. - The Burgers vector, d, of the interface
dislocations is given by d (dAdB)/2
21Semi-coherent interfaces, contd.
- Just as for low angle grain boundaries, the
interface energy is proportional to the
dislocation density at small misfits and then
following a logarithmic dependence at larger
misfits. - In two dimensions, a network involving more than
one Burgers vector may be required to accommodate
the misfit. - The limit to dislocation-based structures is at
d 0.25, corresponding to one dislocation every
four plane spacings (where the cores start to
overlap). - Interface energies are in the 200-500 mJ.m-2
range.
22Complex semi-coherent interfaces
- It can often happen that an orientation
relationship exists despite the lack of an exact
match. - Such is the case for the relationship between bcc
and fcc iron (ferrite and austenite).
Note limited atomic match for the NS relationship
23Orientation relationships in iron
- There are two well-known orientation
relationships for fcc-bcc iron. - The Nishiyama-Wasserman (NW) relationship is
specified as 110bcc/111fcc,
lt001gtbcc//lt101gtfcc. - The NS relationship only gives good atomic fit in
8 of the boundary area. - The Kurdjumov-Sachs (KS) relationship is
specified as 110bcc/111fcc,
lt111gtbcc//lt101gtfcc. - These two differ by only a 5.6 rotation in the
interface plane. - Better atomic matching is possible for irrational
planes used.
24Orientations from KS OR
- Based on a particular orientation relationship
(OR), the orientations of new grains of the
product phase can be predicted, as derived from a
product phase. - Illustrated for the Kurdjumov-Sachs (KS)
relationship for iron, with the 001lt100gt
starting orientation for the fcc (austenite)
phase. - The different new orientations are called
variants.
001 pole figure, showing 001 poles in the
eight variant positions of the ferrite phase for
the KS OR, starting from a single austenite
crystal in (001)100 position.
25Incoherent Interfaces
- Not surprisingly, incoherent interfaces have a
disordered structure similar to high angle grain
boundaries. - Their energies range up to 1 J.m-2.
- Little is known about the detailed structure of
such interfaces. - Large differences in crystal structure and
lattice parameter between parent and product
phases tend to mean that the interface must be
incoherent. - Possibilities for partially coherent interfaces
exist even under the latter circumstance, but
better tools are need for prediction of interface
structure and energy (current research topic,
e.g. W. Reynolds).
26Interfaces in precipitates
- In order to present examples of real systems, it
is important to keep in mind that the interface
around a precipitate is not, in general, the same
over the entire surface. - Analogy with grain boundaries the boundary of an
island grain (fully enclosed within another
grain) varies from pure twist at opposite poles,
to pure tilt around its equator. - Thus, some precipitates possess a mixture of
interface types around their perimeter.
Misorientation axis
Tilt boundary on the equator
Twist boundaries on the poles
27Fully coherent precipitates
- One example of the Cu-Si system has been given.
- Precipitation of Co from Cu is another example.
- Guinier-Preston zones in the early stages of
precipitation in Al alloys are another example. - See fig. 3.39 in PE for Ag-rich zones in Al-4Ag.
- In all cases, the crystal structure is the same
in parent and product also the lattice
parameters are similar.
28Partially coherent precipitates
- When only part of the surface of a precipitate
can be coherent, it is said to be partially
coherent. - Typically, one plane is coherent or
semi-coherent. - As fig. 3.40 shows, the shape of the precipitate
can be determined by the Wulff shape through the
inverse ratio of the interfacial energies. Large
coherent facets are terminated by incoherent
edges. - Caution! An anisotropic shape can also be
determined by either growth anisotropy or elastic
anisotropy.
29Widmanstätten morphology
- Widmanstättens name is associated with platy
precipitates that possess a definite
crystallographic relationship with their parent
phase. - Examples - ferrite in austenite (iron-rich
meteors!) - g precipitates in Al-Ag (see fig.
3.42) - hcp Ti in bcc Ti (two-phase Ti alloys,
slow cooled) - q precipitates in Al-Cu - The latter example is based on the orientation
relationship (001)q//001Al, 100q//lt100gtAl.
See fig. 3.41 for a diagram of the tetragonal
structure of q whose a-b plane, i.e. (001),
aligns with the (100) plane of the parent Al.
30Incoherent precipitates
- Al alloys provide many examples of incoherent
precipitates that lack orientation relationships. - CuAl2 (q) in Al - fig. 3.44Al6Mn in AlAl3Fe in
Al - Note that heterogeneous nucleation at grain
boundaries can give rise to precipitates that are
incoherent on one side, and semi-coherent on the
other side. This leads to significant
differences in growth rate, fig. 3.45.
31Elastic Effects
- The effects of elastic interactions between the
matrix and the precipitate can be as important as
for the interfacial energy. - The two effects can compete this is one reason
for changes during growth, such as the loss of
coherency. - Elastic effects can influence precipitate shape.
32Misfit
- Imagine that a certain volume of the parent phase
(matrix) is removed and replaced by a different
volume of product phase (precipitate). The
difference in volume leads to a dilatational
strain which is positive or negative, depending
on the sign of the volume change. - In the case of identical crystal structures and a
coherent interface, the parent and product have
equal and opposite forces at the interface, see
fig. 3.47c.
33Misfit definitions
- Given lattice parameters specified as aA and aB,
the misfit parameter, d is given by the following
simple formula d (aB - aA)/ aA - Note the similarity to the definition of misfit
for coherent and semi-coherent interfaces. - In the case of dilatational/hydrostatic strains,
one can define a constrained misfit or in situ
misfit based on the strained precipitate lattice
parameter, aB e (aB - aA)/ aA
34Elastic strain energy
- The constrained misfit and the unconstrained
misfit are related to each other. For the
simplest case of identical moduli in parent and
product, e 2d/3 - For typical variations in moduli, the range of
values observed is d/2 lt e lt d. - The elastic energy associated with the
dilatational strains is of order d2 V, where V is
the volume. For the simplest case of isotropic
matrix and precipitate, the elastic energy is
independent of shape ?Gs 4µd2V - µ is
the shear modulus.
35Strain energy anisotropy
- The effect of modulus differences is interesting
and asymmetric Precipitate stiffer than matrix
minimum elastic energy occurs for a
sphere.Precipitate more compliant than matrix
minimum elastic energy occurs for a disc (oblate
ellipsoid). - Note that differences in elastic modulus can be
synergistic or antagonistic to effects of
interface structure. - Anisotropic matrix most cubic metals are more
compliant along lt100gt, hence elastic energy
considerations favor discs perpendicular to lt100gt.
36Elasticity vs coherency
- The competition between elastic energy and
interfacial energy is illustrated by reference to
specific examples in Al alloys. - Observation a sequence of precipitation
reactions is observed in Al-Cu alloys (containing
up to, say 5Cu, i.e. the maximum solid
solubility of Cu in Al, at the eutectic
temperature). - The sequence can be explained as the appearance
of successively more stable precipitates, each of
which has a larger nucleation barrier.
37Al-Cu precipitation sequence
- The sequence is a0 ? a1 GP-zones ? a2 q?
a3 q? a4 q - The phase are an - fcc aluminum nth subscript
denotes each equilibriumGP zones - mono-atomic
layers of Cu on (001)Al q - thin discs, fully
coherent with matrix q - disc-shaped,
semi-coherent on (001)q bct. q - incoherent
interface, spherical, complex body-centered
tetragonal (bct).
38Al-Cu driving forces
- Each precipitate has a different free energy
curve w.r.t composition. Exception is the GP
zone, which may be regarded as continuous with
the alloy (leading to the possibility of spinodal
decomposition, discussed later). - PE fig. 5.27 illustrates the sequence of
successively greater free energy decreases and
also successively greater ?G. - PE fig. 5.28 illustrates the point that the
nucleation barriers are much smaller for each
individual nucleation step when the next
precipitate nucleates heterogeneously on the
previous structure.
39Al-Cu ppt structures
GP zone structure
40Nucleation sites, reversion
- The nucleation sites vary depending on
circumstances. - q most likely nucleates on GP zones by adding
additional layers of Cu atoms. - Similarly, q nucleates on q by in-situ
transformation. - However, q can also nucleate on dislocations,
see PE fig. 5.31a. - The full sequence is only observable for
annealing temperatures below the GP solvus. Any
of the intermediate precipitates can be
dissolved, reverted, by increasing the
temperature above the relevant solvus, fig. 5.32.
41Coherency loss
- The growth of the penultimate precipitate, q,
illustrates an important point about the loss of
coherency that commonly occurs during growth. - A precipitate may start out fully coherent but
nucleate interfacial dislocations once it reaches
a critical size. - Illustration large q ppts commonly have
dislocations, see PE fig. 5.30c. - Why? Again, a competition exists between
volumetric elastic energy, and interfacial energy.
42Coherency loss, analysis
- The assumptions of the simple analysis (PE
section 3.4.4) are elastic strain energy is
significant for the fully coherent case, not for
the non-coherent case. Also, the energy of
non-coherent interface is significantly larger
than that of the coherent interface.
Thus ?Gcoherent ?Gelastic
?Ginterface 4µd2 4pr3/3
4pr2gcoherent ?Gnon-coherent ?Gelastic
?Ginterface 0
4pr2gnon-coherent - At some size, the former becomes larger than the
latter. Provided dislocations (or other defects)
can be nucleated, the character of the interface
will change, and coherency will be lost.
43Coherency loss, estimates
- It is possible to estimate the size, and type of
precipitate for coherency loss. - Given a difference in interfacial energy between
coherent and non-coherent (which PE write as gst
gnon-coherent - gcoherent), one can estimate a
critical radius, rcrit. rcrit 3?g/4µd2. - Fig. 3.53 illustrates the requirement for
dislocation loops to be arranged on the perimeter
of the precipitate (similar to a low angle grain
boundary). - Based on a constrained misfit, e, the minimum
stress required to nucleate a dislocation (fig.
3.54a) is of order 3µe and the minimum value of
misfit to exceed the theoretical shear strength
of a matrix is e 0.05. This implies that
precipitates that have a small enough misfit will
never lose coherency as they grow because they
are unable to nucleate the required interfacial
dislocations.
44Impact on Properties
- The most obvious impact of precipitation in
metals is on mechanical strength. - Precipitation was measured by hardness, long
before the structure of the precipitates was
known. - Example age-hardening curves in Al-Cu alloys,
PE fig. 5.37.
45Summary
- This lecture may be summarized by stating that
both differences in elastic properties and
interface structure exert a strong influence on
precipitate morphology. - Through their effect on free energy changes as a
function of size, they also affect which
precipitates actually nucleate under any given
conditions. - The precipitate with the smallest nucleation
barrier (generally) appears first. Small
nucleation barriers are associated with coherent
interfaces (small interfacial energy) and similar
lattices (small elastic energies from misfit).