12

Dispersion Strengthening by
Phase Transformations and
Heat Treatment
Have You Ever Wondered?
9 Who invented and flew the first controllable airplane?
9 How do engineers strengthen aluminum alloys used in aircrafts?
9 Why do some steels become very hard upon quenching from high temperatures?
9 What alloys are used to make orthodontic braces?
9 Would it be possible to further enhance the strength and, hence, the dent resistance of sheet
steels after the car chassis is made?

In Chapter 11, we examined in detail how
second-phase particles can increase the strength
of metallic materials. In this chapter, we further
discuss dispersion strengthening as we describe
a variety of solid-state transformation processes
including precipitation or age hardening and the
eutectoid reaction. We also examine how non-
equilibrium phase transformations—in particular,
the martensitic reaction—can provide strength-
ening.
As we discuss these strengthening mecha-
nisms, keep in mind the characteristics that pro-

357
358 CHAPTER 12 Dispersion Strengthening by Phase Transformations and Heat Treatment

duce the most desirable dispersion strengthening
as discussed in Chapter 11:
9 The matrix should be relatively soft and
ductile and the precipitate, or second
phase, should be strong;
9 the precipitate particles should be round
and discontinuous;
9 the second-phase particles should be small
and numerous; and
9 in general, the more precipitate phase we
have, the stronger the alloy will be.
As in Chapter 11, we will concentrate on how
phase transformations influence the strength of
the materials and how heat treatments can influ-
ence other properties. Since we will be dealing
with solid-state phase transformations, we will
begin with a discussion on the nucleation and
growth of second-phase particles in solid-state
phase transformations.

12-1 Nucleation and Growth in Solid-State Reactions

In Chapter 9, we discussed nucleation of a solid nucleus from a melt. We also discussed
the concepts of supersaturation, undercooling, and homogeneous and heterogeneous
nucleation. Let’s now see how these concepts apply to solid-state phase transformations
such as the eutectoid reaction. In order for a precipitate of phase b to form from a solid
matrix of phase a, both nucleation and growth must occur. The total change in free
energy required for nucleation of a spherical solid precipitate from the matrix is:
DG ¼ 43 pr 3 DGvða!bÞ þ 4pr 2 sab þ 43 pr 3 e ð12-1Þ
The first two terms include the free energy change per unit volume (DGv ), and the en-
ergy change needed to create the unit area of the interface (sab ), just as in solidification.
However, the third term takes into account the strain energy per unit volume (e), the
energy required to permit a precipitate to fit into the surrounding matrix during the
nucleation and growth of the precipitate, introduced when the precipitate forms in a
solid, rigid matrix. The precipitate does not occupy the same volume that is displaced,
so additional energy is required to accommodate the precipitate in the matrix.

Nucleation As in solidification, nucleation occurs most easily on surfaces already

present in the structure, thereby minimizing the surface energy term. Thus, the precip-
itates heterogeneously nucleate most easily at grain boundaries and other defects.

Growth Growth of the precipitates normally occurs by long-range di¤usion and redis-
tribution of atoms. Di¤using atoms must be detached from their original locations
(perhaps at lattice points in a solid solution), move through the surrounding material
to the nucleus, and be incorporated into the crystal structure of the precipitate. In some
cases, the di¤using atoms might be so tightly bonded within an existing phase that the
detachment process limits the rate of growth. In other cases, attaching the di¤using
atoms to the precipitate—perhaps because of the lattice strain—limits growth. This
result sometimes leads to the formation of precipitates that have a special relationship
to the matrix structure that minimizes the strain at the interface between the parent
phase and the precipitate particles. In most cases, however, the controlling factor is the
di¤usion step.
 12-1 Nucleation and Growth in Solid-State Reactions 359

Figure 12-1
Sigmoidal curve showing the rate
of transformation of FCC iron at a
constant temperature. The incuba-
tion time t0 and the time t for 50%
transformation are also shown.

Kinetics The overall rate, or kinetics, of a transformation depends on both nucleation

and growth. If more nuclei are present at a particular temperature, growth occurs from
a larger number of sites and the phase transformation is completed in a shorter period
of time. At higher temperatures, the di¤usion coe‰cient is higher, growth rates are
higher, and again we expect the transformation to be completed in a shorter time, as-
suming an equal number of nuclei.
The rate of transformation is given by the Avrami equation (Equation 12-2), with
the fraction of the transformation, f , related to time, t, by
f ¼ 1 " expð"ct n Þ ð12-2Þ
where c and n are constants for a particular temperature. This Avrami relationship,
shown in Figure 12-1, produces a sigmoidal, or S-shaped, curve. This equation can de-
scribe most solid-state phase transformations. An incubation time, t0 , during which no
observable transformation occurs, is the time required for nucleation to occur. Initially,
the transformation occurs slowly as nuclei form.
The incubation period is followed by rapid growth as atoms di¤use to the growing
precipitate. Near the end of the transformation, the rate again slows as the source of
atoms available to di¤use to the growing precipitate is depleted. The transformation is
50% complete in time t; the rate of transformation is often given by the reciprocal of t:

Rate ¼ 1=t ð12-3Þ

Effect of Temperature In many phase transformations, the material undercools below

the temperature at which the phase transformation occurs under equilibrium condi-
tions. Recall from Chapter 9 the undercooling of water and other liquids and other
supersaturation phenomena. Because both nucleation and growth are temperature-
dependent, the rate of phase transformation depends on the undercooling (DT). The
rate of nucleation is low for small undercoolings (since the thermodynamic driving
force is low) and increases for larger undercoolings as the thermodynamic driving force
increases at least up to a certain point (since di¤usion becomes slower as temperature
decreases). At the same time, the growth rate of the new phase decreases continuously,
because of slower di¤usion, as the undercooling increases. The growth rate follows an
Arrhenius relationship (recall, Equation 5-1):
" #
"Q
Growth rate ¼ A exp ð12-4Þ
RT
where Q is the activation energy (in this case for the phase transformation), R is the gas
constant, T is the temperature, and A is a constant.
360 CHAPTER 12 Dispersion Strengthening by Phase Transformations and Heat Treatment

Figure 12-2 The effect of temperature on recrystallization of cold-worked copper.

Figure 12-2 shows sigmoidal curves at di¤erent temperatures for the recrystalli-
zation of copper; as the temperature increases, the rate of recrystallization of copper
increases, because growth is the most important factor for copper.
At any particular temperature, the overall rate of transformation is the product of
the nucleation and growth rates. In Figure 12-3(a), the combined e¤ect of the nuclea-
tion and growth rates is shown. A maximum transformation rate may be observed at a
critical undercooling. The time required for transformation is inversely related to the
rate of transformation; Figure 12-3(b) describes the time (on a log scale) required for the
transformation. This C-shaped curve is common for many transformations in metals,
ceramics, glasses, and polymers. Notice that the time required at a temperature corre-
sponding to the equilibrium phase transformation would be y (i.e., the phase transfor-
mation will not occur). This is because there is no undercooling and, hence, the rate of
homogenous nucleation is zero.
In some processes, such as the recrystallization of a cold-worked metal, we find that
the transformation rate continually decreases with decreasing temperature. In this case,
nucleation occurs easily, and di¤usion—or growth—predominates (i.e., the growth is
the rate limiting step for the transformation). The following example illustrates how the
activation energy for a solid-state phase transformation such as recrystallization can be
obtained from data related to the kinetics of the process.

Figure 12-3 (a) The effect of temperature on the rate of a phase transformation is the product
of the growth rate and nucleation rate contributions, giving a maximum transformation rate at a
critical temperature. (b) Consequently, there is a minimum time (tmin ) required for the
transformation, given by the ‘‘C-curve’’.
 12-1 Nucleation and Growth in Solid-State Reactions 361

EXAMPLE 12-1 Activation Energy for the Recrystallization of Copper

Determine the activation energy for the recrystallization of copper from the
sigmoidal curves in Figure 12-2.

SOLUTION
The rate of transformation is the reciprocal of the time t required for half of
the transformation to occur. From Figure 12-2, the times required for 50%
transformation at several di¤erent temperatures can be calculated:

T (˚C) T (K) t (min) Rate (minC1 )

135 408 9 0.111

119 392 22 0.045
102 375 80 0.0125
88 361 250 0.0040

The rate of transformation is an Arrhenius equation, so a plot of ln (rate)

versus 1/T (Figure 12-4 and Equation 12-4) allows us to calculate the constants
in the equation. Taking natural log of both sides of Equation 12-4:

Q
ln(Growth rate) ¼ ln A "
RT

Thus, if we plot ln(Growth rate) as a function of 1=T, we expect a straight line

that has a slope of "Q=R.
2 3

"Q 6 7
6 D lnðrateÞ 7 ½ðlnð0:111Þ " lnð0:004ÞÞ*
Slope ¼ ¼6 " # 7¼ ( )
R 4 1 5 1 1
D "
T 408 361

Figure 12-4
Arrhenius plot of
transformation rate
versus reciprocal
temperature for
recrystallization of
copper (for Example
12-1).
362 CHAPTER 12 Dispersion Strengthening by Phase Transformations and Heat Treatment

Q=R ¼ 10;414
cal
Q ¼ 20;693
mol
0 1
"20;693 cal=mol
8 0:111 ¼ A expB" # C
@ cal A
1:987 ' ð408 degÞ
deg " mol
A ¼ 0:111=8:21 ' 10"12 ¼ 1:351 ' 10 10 s"1
" #
10 "20;693
8 rate ¼ 1:351 ' 10 exp
RT
In this particular example, the rate at which the reaction occurs increases as
the temperature increases, indicating that the reaction may be dominated by
di¤usion.

12-2 Alloys Strengthened by Exceeding the Solubility Limit

In Chapter 11, we learned that lead-tin (Pb-Sn) alloys containing about 2 to 19% Sn can
be dispersion-strengthened because the solubility of tin in lead is exceeded.
A similar situation occurs in aluminum-copper alloys. For example, the Al-4% Cu
alloy (shown in Figure 12-5) is 100% a above 500! C. The a phase is a solid solution
of aluminum containing copper up to 5.65 wt%. On cooling below the solvus tempera-

4
Figure 12-5 The aluminum-copper phase diagram and the microstructures that may develop
during cooling of an Al-4% Cu alloy.