SIM / AM transformations de phases / Phase transformations

Tutorials

1. Free energy curves and phase stability

G(To)
g

a
d
b

Xo X1 Solute concentration

We consider a phase a at a temperature To, in which several phases can be formed: b, g, and d. Classify
these 3 phases in terms of relative stability. Which one is the stable phase in equilibrium with a?
At temperature To, what phases can be thermodynamically formed if the concentration of the a-phase is
Xo ? X1 ?

2. Free energy diagrams & phase diagrams

The figure below gives the free energy diagrams for two phases a and b at three temperatures (T1<T2<T3).

From these diagrams, plot the limits of the phase diagram in the graph below. Describe the single and two-
phase fields.

Give the physical reason why the free energy curves are, for a given concentration, a negative function of
temperature.

Give the physical reason why the solubility in both phases decreases with temperature.

, the negative temperature dependence of free energy curves, particularly the Gibbs free energy, is
primarily due to the temperature-dependent behavior of entropy. As temperature rises, the entropy
term becomes more dominant in the Gibbs free energy equation, leading to more negative values of
G for processes, including phase transitions and chemical reactions, which are thermodynamically
favored at higher temperatures.
G(T)
Temperature T1
b
a
Temperature T2

Temperature T3

Solute concentration
Temperature

T3

T2

T1

Solute concentration

3. Heterogeneous precipitation

We consider the nucleation of a spherical precipitate on a planar grain boundary of the a-phase of interfacial
energy gGB. The interfacial energy between the precipitate and the matrix is g and the chemical driving force
of the reaction is Dgv.

Write the energy balance for the formation of the precipitate; demonstrate that the nucleation of the
precipitate on the grain boundary can be considered in the same way as homogeneous nucleation, with a
modified surface energy geff. Calculate this modified surface energy.

Calculate the ratio between the heterogeneous nucleation barrier and the homogeneous one, as a function
of the ratio gGB/g.

4. Nucleation in the solid state

We consider the precipitation of small clusters of B atoms (b phase) in a A-B solid solution (a).
The free energy change per unit volume associated to the formation of b precipitates is denoted Dgv and
the a/b interfacial energy per unit area is denoted g.
1- Derive the expressions for the critical radius R* and the related energy barrier DG*.

2- In order to illustrate the influence of the temperature on the nucleation process, derive the expression of
Dgv , R* and DG* as a function of the enthalpy of formation of the b phase in the a solid-solution (Dh), the
temperature T, and the temperature TE of the b solvus.
How do these parameters vary with temperature?
We consider now the possibility that two distinct phases b and b' may precipitate in the a solid solution:
- for the metastable b' phase : g = 80 mJ.m-2, TE = 200°C, Dh = -1.109 J.m-3.
- for the stable b phase : g = 300 mJ.m-2, TE = 450 °C, Dh = -1.4.109 J.m-3.
Calculate R* and DG* (eV) for two nucleation temperatures : 100 °C and 170 °C. Conclusion?

#$ ∗
3- Neglecting the Zeldovich factor, the nucleation rate is given by: 𝐽 = 𝑁! 𝛽 ∗ 𝑒𝑥𝑝 (− %&
*

where N0 is the atomic site density = W -1, ( W being the volume per atom = 0.016 nm3)
'() ∗" *+#
and: 𝛽 ∗ = ,$
-
D is the diffusion coefficient, 𝐷 = 𝐷! 𝑒𝑥𝑝 (− %&* with E=1.3 eV and D0=108 m2s-1
X0 is the initial solute content of the solid solution (atomic fraction) : X0= 0.04
a is the lattice parameter : a = 0.4 nm
Calculate the nucleation rate of b and b' at 100°C and 170°C. Conclusion?

4- We are now concerned with the evolution of the nucleation parameters with the solute content in the
matrix X0, at constant temperature T.
/
The solvus line for the a-b equilibrium is given by an equation of the form : 𝑋. = 𝑒𝑥𝑝 (− %& *
Derive the expression for the ratio T/TE as a function of X0 and Xe (the equilibrium concentration at
temperature T).
Compare the values of the energy barrier DG* for two solute contents (4% and 2% ) if Xe=1%.
Assuming a constant value of the critical radius R*, compare the nucleation rates for the b' phase at 100°C.
5- The precipitates formed during the nucleation process may grow by long-range diffusion of B atoms.
0) *(+3+% )
Their growth rate is given by : 01
= )(53+% )

Show that the plot of the growth rate dR/dt versus temperature exhibits a maximum and derive the
expression for the corresponding temperature (Tmax) : assume that Xe<<1 and X~4%.
Calculate Tmax for b and b' from the knowledge of their solvus temperatures TE.

5. Capillarity effects

The objective of this exercise is to evaluate the magnitude of the capillarity effect.
The Gibbs Thomson effect expresses for a binary system in the regular solid solution model as (see course
notes):
2𝛾Ω
𝑋6 (𝑅) = 𝑋. 𝑒𝑥𝑝 0 6
𝑅𝑘𝑇

Give the expression of the radius that would double the equilibrium solubility. What value takes this radius
in the case where T=200°C, Xe(T=200°C)=0.01, W=16 Å3, in the two following cases:
- g = 0.05 J.m-2 (coherent phase)
- g = 0.5 J.m-2 (incoherent phase)
- Now we consider an alloy having an overall solute content of Xo=0.04. The current solute content is
/
X=0.02, and the material contains precipitates of R=5 nm. We assume that 𝑋. = 𝑒𝑥𝑝 (− %&*. Calculate Q
if Xe=0.01 at 200°C.
One raises the temperature suddenly. At what temperature would the precipitate start to dissolve in the
matrix in the case where g = 0.5 J.m-2 ?
What would be the dissolution temperature for precipitates of infinite radius with the same matrix solute
content X=0.02? Compare these two temperatures (finite and infinite radius) with the solvus of the alloy.

6. Growth of an interfacial reaction: the case of pearlite

We consider an Fe-C eutectoid which is subjected to an austenite à pearlite reaction. The average grain
size of the austenitic grain, f, is 50 µm.
Pearlite is observed to nucleate exclusively on grain boundaries, and progressively invades the austenite
grains, according to the scheme below. The eutectoid temperature is TE=720°C.

At a time t after quenching at temperature T lower than TE, f is the transformed fraction of austenite, and v
is the pearlite growth rate.

Grain boundary

Pearlite

1) If the nucleation time of pearlite is neglected, and assuming a spherical shape of the austenite grains,
give the relationship existing between t, f, f, and v. At T = 650°C, the transformation is completed in 40
seconds. Find out the value of v at this temperature.

2) One wants to establish the TTT curves for this transformation. Recall their definition, and propose an
experimental method for determining them.

3) What is the equation of these curves, if one uses the kinetic model of Zener ? Calculate the temperature
of the nose of these TTT curves. What is its value if the activation energy of carbon diffusion is 1.4 eV?

4) Find out the time to reach the nose of the TTT curve corresponding to a transformed fraction of 1%.
If one wants to avoid the formation of 1% pearlite in order to obtain a martensitic structure, what cooling
rate would be necessary? Is it an achievable cooling rate?
The maximum cooling rate achievable on this sample is 40°C/s, and one admits to have a pearlite fraction
of 10%. On what material characteristic is it possible to play to reach this goal?

5) One tries now to obtain the finest possible pearlite without forming martensite, so as to obtain good
mechanical properties with a good ductility. What type of cooling would you propose, knowing that the
MS temperature of the alloy is 400°C?
Calculate the time needed by the steel to transform at 450°C before it can be cooled safely (transformed
fraction of 90%). The initial austenite grain size is taken as 50 µm.

7. Metallurgical structure of steel : 100WC10

2 sample of the steel 100WC10 are austenized, and then are subjected to various treatments. What will be
the metallurgic structure of each of these after the following treatments? (Nature of phases, spatial
distribution, hardness)?
a) Austenized at 850°C during 30 min, and quenched into liquid N2
b) Same as (a) and then heated 2 minutes at 600°C and quenched at 0°C
c) Austenized at 850°C during 30 min, quenched at 600°C, maintained for 1 hour at this temperature and
quenched
d) Austenized at 850°C during 30 min, quenched at 600°C, maintained for 15s at this temperature and
quenched
e) Comment on the shape of the TTT diagram. Why are there two noses on the right-hand side? Explain
the variation of hardness as a function of temperature of heat treatment.
8. Phase transformations in Ti alloys

Titanium exists in two crystal structures: a, of hexagonal structure, stable at room temperature until
882.5°C ; and b, of Body Center Cubic structure, stable from this temperature until the melting point.

1. What is the name of the transformation that transforms a into b or the other way round ? Is it a
continuous or discontinuous transformation? One cools pure Ti from the b field into the a field.
What mechanism controls
2. the growth velocity of the interface?

Ti based alloys contain many different solute elements (Al, O, Mo, V, Cu, Mn, …). Some of these elements
are a-extenders, some others b-extenders. Figure 1 shows two example of Ti-X phase diagrams.

(a) a+b
b
Temperature

Temperature

b
a

a a+b
(b)
Ms
X concentration X concentration

2. Identify the diagram corresponding to an a-extender and that corresponding to a b-extender.

In the case (a), explain what is the « Ms » line . Describe briefly what happens below this temperature.

One considers now an alloy of type (b). The alloy is at an X concentration Xo, at a temperature To so that
the alloy is in the b field. The alloy is now cooled at temperature T1.

3. What phase transformation takes place? Is the

driving force large or small? What type of nucleation To
a+b
site can therefore be activated for the new phase? T1 b
Température

Draw a concentration profile across a grain initially of

b-phase: a
- at the beginning of the transformation T2
- when the transformation is finished
(b)
4. The alloy is now cooled (still from the To temperature)
until temperature T2. Describe the sequence of events leading Xo
to the thermodynamic equilibrium in the two following cases: Concentration in X
(a) slow cooling
(b) instantaneous cooling. What is the name of the transformation in this last case?
5. We consider now a type (a) alloy.

We will not consider the presence of the Ms curve. (a)

The micrographs (a) (optical) and (b) (electron microscopy)

Température
show the microstructure after cooling, respectively from
(a) the b phase ; (b) the (a+b) field. b
Comment the two microstructures. Under what
morphology, which enables to increase the transformation
speed, does the phases appear during cooling? Explain in what a a+b
respect this morphology enables a faster transformation rate.
Ms
Draw schematically concentration profiles in the two cases.
Concentration en X

(a) (b)

6. An alloy in a-phase is strongly plastically deformed, and then heated at a temperature (T1) situated in
the single phase a field just below the limit of the phase diagram. The alloy recrystallises.

(a) What happens if one waits for a very long time ?

(b) Instead of temperature T1, one uses a slightly higher temperature T2, in the two phase a+b field. If one
assumes that the phase transformation is faster than recrystallisation, describe the sequence of events that
occurs. If one waits for a very long time at this temperature, can the grain size become very large? What
will limit the grain growth and what mechanism can still enable the grain size increase?
9. Recrystallisation

We aim at estimating semi-quantitatively the parameters influencing static recrystallisation and its
relationship with recovery, after a plastic deformation at room temperature.

1. Definitions
Recall the definitions for static & dynamic recrystallisation and recovery.

An aluminium alloy is deformed at room temperature.

The flow stress as a function of plastic strain can be written in exponential form:

s -s o æ e ö
= 1 - expçç - ÷÷
s s -so è eo ø
where s is the flow stress, so is the yield stress=100 MPa, ss=400 MPa is the saturation stress, eo is a
reference strain=0.2. In the following we will use Ds= ss-so.
On the other hand one knows that s-so= K r1/2, where r is the dislocation density, and K»Maµb,
where M~3 is Schmid’s factor, a~0.3 is a constant, µ is the shear modulus (Aluminium 25 GPa).

2. Calculate the dislocation density for a given strain. Deduce the stored energy in the form of
dislocations in the material by unit volume Dgv. (the line energy of a dislocation is El = ½ µb2
(b=0.286 nm).

3. The surface energy of a grain boundary equals to gGB ~ 0.5 J.m-2. Estimate the nucleation radius R* of
a new grain for a very large strain.

4. The nucleation rate of new grains can be written:

æ DG * ö
J = J o expç - ÷
ç kT ÷
è ø
Estimate the influence of temperature and initial strain on the grain size. What happens additionally
for high temperatures and long times?

5. How does static recovery act on recrystallisation? Do critical temperatures and critical strains exist to
activate recrystallization?

6. Estimate, at very large strains and the temperature of 350°C, the value of DG*, the activation barrier
for the formation of a new grain. Does it seem reasonable to you? What other phenomenon could
substitute to homogeneous nucleation of new grains?
Exam 2020-2021
Exercice 1 : crystallization of an amorphous film

Figure 1 : (a) View of a crystalline film of silicon in between two films of amorphous silicon ; (b) Two successive images of the
formation of silicon crystals in a matrix of amorphous silicon

Figure 2 : Amorphous Si film on a substrate of silica

A film of amorphous silicon is deposited on a substrate of silica (SiO2). One aims at controlling the crystallisation kinetics of
this film.
The free energy difference between the amorphous and crystalline phases is given by:
Dga/c=1.36. 109 J.m-3 (approximation: independent of temperature)
The interfacial energy between amorphous and crystalline silicon is: ga/c=0.3 J.m-2
Shear modulus of Silicon: µ=72 GPa
Molar mass : 28.09 g/mol, Molar volume: 12.05 cm3/mol, Atomic volume= 20 Å3
Boltzman’s constant : k=1.38 10-23 J.K-1 Charge of the electron e=1.6. 10-19 C
Activation energy for breaking Si-Si bonds : 1.8 eV
1. a) Calculate the critical radius for nucleation at 700°C. How many Si atoms are involved in a nucleus? Is this value consistent
with the images (result of simulations) of figure 1? Calculate the activation barrier for nucleation (in eV).
b) The density of amorphous Silicon is 2.45 g.cm-3 and that of crystalline silicon is 2.33 g.cm-3. What type of stresses will appear
in both phases during crystallisation? Using the course formulae, calculate the elastic energy per unit volume that will appear in
a crystalline nucleus (assuming a spherical shape), and compare it to Dga/c. Is it important to take this elastic term into account?
2. It is observed that nucleation of the crystalline Silicon grains occurs essentially in a heterogeneous way, at the SiO2/Si interface.
Using the course notes, recall the geometry of this nucleation process, as well at the associated reduction of activation barrier,
as a function of the wetting angle. Indicate the condition that applies on the interfacial energies between SiO2 and crystalline
Silicon (gSiO2/c) and between SiO2 and amorphous Silicon (gSiO2/a) in order to have:
- perfect wetting (formation without energy barrier), or
- only homogeneous nucleation.
Actually (gSiO2/c)=0.5 J.m-2, (gSiO2/a)=0.3 J.m-2. What is the reduction of energy barrier as compared to homogeneous nucleation?
At 700°C, give the ratio between the term exp(-DG*/kT) for heterogeneous and homogeneous nucleation.
3. We will now consider the growth of Si crystals into amorphous silicon.
a) Describe the mode that controls the interface movement. What is the limiting mechanism?
b) If the capillary effects (Gibbs Thomson) are neglected, how does the growth rate vary with time, once crystals have nucleated
(assume that there are no interactions between crystals).
Explain why this rate is modified in the first stages of growth that follow the apparition of a nucleus at the critical size. Write, in
the case of homogeneous nucleation, the difference in free energy per unit volume that exist between a crystal of size R and the
amorphous matrix, taking into account both the chemical and interfacial energy terms. At what size does this difference becomes
equal to 90% of its equilibrium value Dga/c ?
c) Using the data in the exam subject, and based on the mechanism controlling the movement of the interface, give a likely value
of the activation energy for the growth interface velocity. The growth velocity is 1 nm/s at 700°C. What becomes this velocity
at room temperature? How much can the interface travel at this temperature in 10 years?

Exercice 2: HSLA steels

Table 1 : Compositions of the R and Nb grades in wt%.

Nuance C Mn Si Al Nb Fe
R 0.20 1.35 0.40 0.05 - Balance
Nb 0.20 1.35 0.40 0.05 0.04 Balance

850
a b
800 γ
Temperature (°C)

750

700 α+γ

650

600
0,00 0,50 1,00 1,50
Carbon content (wt. %)
Figure 1 : a) Paraequilibirum diagram for ferrite and austenite in grades R and Nb. b) Transmission electron micrograph of NbC carbides
precipitated during the austenite-to-ferrite transformation

a b

Figure 2: TTT diagrams of grades a)Nb et b)R after austenitization at 1300°C for 30 min. The hatched area indicates NbC precipitation.

In the following, the heat treatment of two HSLA steel grades designated R and Nb is considered. Their compositions are shown
in Table 1. Their compositions are identical except for their niobium content: grade R does not contain niobium while grade Nb
contains 0.04 wt.%. The pseudo-binary (Fe,Mn,Si,Al)-C phase diagram, corresponding to para-equilibrium conditions for
austenite and ferrite and valid for both grades, is given in Figure 1a. The respective TTT diagrams for the two grades are provided
in Figure 2.
We are interested in the allotropic transformation of austenite to ferrite in the alloy at 650°C after a 30 min austenitization at
1300°C. This transformation is supposed to take place at para-equilibrium.
1) What are the concentrations of the various alloying elements on either side of the austenite/ferrite transformation interface?
Give a schematic representation of the concentration profiles across the interface at the onset of the transformation.
2) How long does it take at 650°C for the allotropic transformation to be completed for both grades?
It can be experimentally observed that nanometric precipitates of NbC carbides are formed at the transformation interface
giving rise to a succession of planes in the ferrite on which the carbides are arranged as can be seen in Figure 1b. It is also
noticed that the interface stops and starts again during the transformation. Moreover, no precipitates can be found in austenite
far from the interface.
3) NbC carbides have a very high interface energy with austenite. Explain how this parameter prevents the precipitation of
NbC in austenite at the overall alloy concentration, despite the fact that this austenite is effectively supersaturated. Then
explain how the transformation of austenite to ferrite allows precipitation to occur.
4) Taking into account the size of the carbides in Figure 1b, propose by analogy with their effect on a grain boundary an
explanation for the difference in transformation times found between the two grades.
5) The precipitation of NbC leads to a local depletion of carbon in the austenitic matrix. Explain how this phenomenon
destabilizes the austenite and promotes its transformation into ferrite. You can use free energy curves.
6) Based on the previous questions, propose a mechanism leading to the observed microstructure. You can use schematic
concentration profiles across the interface. Comment on the effect of this mechanism with respect to your answer to question 2.
7) If the alloy were held at 650°C after the allotropic transformation was complete, what type of transformation would occur?
Sketch the microstructure that would be obtained.
8) Upon completion of the allotropic transformation at 650°C, one chooses instead to quench the alloy with water then with
liquid nitrogen. What is the transformation that will occur? What is the composition of the product phase(s)? Position, with
justification, the starting temperature of this transformation in relation to the Ms temperature shown in Figure 2. For this
question, you can disregard the effect of NbC precipitation on the carbon composition of the solid solution.