Sha W., Malinov S. Titanium Alloys: Modelling of Microstructure, Properties and Applications

Подождите немного. Документ загружается.

Differential scanning calorimetry and property measurements 75

A polythermal section of the phase diagram (Fig. 4.5) traces the change of

the equilibrium phase composition of the alloy with temperature. In the

calculations, the chromium and niobium compositions are kept constant, and

the aluminium concentration and the temperature are varied. In the same

figure, we depict (see dashed line) the alloy composition. From these

calculations, one can suggest that the equilibrium phase compositions change

in the following sequence on increasing the temperature from room temperature

to 1600 °C:

+ γ + B2 → α + α

+ γ + B2 → α + γ + B2 →

α + γ → α → α + β + L → β + L → L. [4.1]

Further, the amounts of phases are shown (see Fig. 4.6). Though, usually, the

kind of sharp change of the equilibrium mole percent curves of phases around

700 °C in Fig. 4.6 is caused by the change of phase equilibrium, Fig. 4.5

(°C)

1600

1500

1400

1300

1200

1100

1000

900

800

700

600

40 42 44 46 48 50

Al (at.%)

L + β

α + β

α + β + L

L + α

α + γ

α + γ + B2

γ + B2

α + α

+ γ + B2

+ γ

+ γ + B2

+ γ

4.5

Calculated polythermal sections of the quaternary Ti–Al–Cr–Nb

system with the composition of the alloy plotted on them. The

calculations are for Cr = 1.9 at.% and Nb = 3 at.%. For clarity, only

the range of compositional interest for gamma titanium aluminides is

plotted.



Titanium alloys: modelling of microstructure76

clearly shows that this is not the case in the present system. Based on the

above analysis we suggest that:

(i) The thermal effect TE5, which is the major peak in the calorimetry

curves (Fig. 4.2), is due to the γ to α phase transformation at continuous

heating. Microstructures of the alloy heated to different characteristic

temperatures in the calorimetry curve, viz. 900, 1020, 1160, 1200,

1220 (Fig. 4.7a) and 1390 °C (Fig. 4.7b), prove this. These temperatures

are purposely chosen as the temperatures of the ends of the different

thermal effects. After heating to and cooling from 1220 °C, which is

about the start temperature of TE5 at heating with the rate of 20 °C/min

(see Fig. 4.2), the microstructure (Fig. 4.7a) remains similar to the

initial microstructure (Fig. 2.5) in respect to the grain morphology. We

might expect some changes, since thermal effects TE1–TE4 are passed.

However, these changes are possibly at a finer microstructure level,

and are not observable by optical microscopy. On the other hand, the

microstructure of the alloy heated to the next temperature (1390 °C),

which is about the end temperature of TE5 at heating with rate 20 °C/

min (see Fig. 4.2), is totally different. The microstructure is fully lamellar,

which is typical for the alloy after cooling from temperatures above α-

transus. This is an experimental proof that the thermal effect TE5 is a

result of the γ to α phase transformation in the alloy.

600 800 1000 1200 1400 1600

(°C)

Mole % phase

100

Liq

4.6

Calculated equilibrium mole percentage of phases versus

temperature for Ti-46Al-1.9Cr-3Nb-0.17O alloy.



Differential scanning calorimetry and property measurements 77

(ii) The thermal effect TE4, which is just before TE5, is due to the

disappearance of the B2 phase, where the alloy moves from the α + γ

+ B2 to α + γ phase equilibrium ranges. This effect is small because of

the small amount of the B2 phase. The thermodynamic calculations

200 µm

(a)

200 µm

(b)

4.7

Microstructure of Ti-46Al-1.9Cr-3Nb alloy after heating with a rate

of 20 °C/min to (a) 1220 and (b) 1390 °C and cooling with a rate of

20 °C/min.



Titanium alloys: modelling of microstructure78

show that the equilibrium temperature for this transformation is 1130 °C,

but in the calorimetry curves, this effect is observed at about 1200 °C.

Probably, overheating above the equilibrium temperature is necessary

for this transformation to take place. The disappearance of the B2

phase at 1200 °C is also detected by high-temperature X-ray diffraction

in isothermal conditions at elevated temperatures (see Chapter 3 and

Fig. 4.8).

(iii) The thermal effect TE3 is due to the transition of α

+ γ + B2 to α + γ

+ B2 by crossing the four phase α + α

+ γ + B2 region. In the

quaternary Ti–Al–Cr–Nb system, this transition is in a narrow temperature

range. The effect is relatively small, because of the small amount of the

phase. The thermodynamic calculations show that the equilibrium

temperature range for this transformation is 1067–1070 °C. On the

calorimetry curves for continuous heating, it is observed at higher

temperatures – around 1170 to 1180 °C. Again, this difference can be

due to kinetics reasons. Appreciable increase in the amount of α phase

at temperatures higher than 1100 °C has been shown for the same alloy

(Chapter 3), which is in agreement with what is suggested above.

Finally, effects like TE4 and TE3 have not been observed in differential

33.0 33.1 33.2 33.3 33.4 33.5 33.6 33.7 33.8

100

150

200

250

300

350

{011} B2

20 °C

1000 °C

1200 °C

Intensity (c.p.s.)

2θ (°)

20 °C

1000 °C

1200 °C

{111} γ

{002} α(α

)

31.5 32.0 32.5 33.0 33.5

2θ (°)

Intensity (c.p.s.)

10000

8000

6000

4000

2000

4.8

In situ

synchrotron radiation diffraction patterns of Ti-46Al-1.9Cr-

3Nb alloy at different temperatures.



Differential scanning calorimetry and property measurements 79

thermal analysis and differential scanning calorimetry curves of other

gamma titanium aluminides where the B2 phase does not exist (Ohnuma

et al., 2000).

(iv) The broad thermal effect TE2 is associated with gradual change of the

ratios between the α

, γ and B2 phases. The ratios between these phases

change upon heating to above 900 °C (Fig. 4.6), where the amount of

the B2 phase increases, while the amounts of α

and γ phases decrease.

Such an increase in the amount of B2 phase is experimentally validated

for the same alloy (see Fig. 4.8 and Chapter 3). The amount of B2

phase increases when the temperature is increased from room temperature

to 1000 °C (Fig. 4.8). Further increase of the temperature to 1100 and

1200 °C results in decrease and disappearance of the B2 phase. At

higher heating rates, as earlier stated, TE1 and TE2 are overlapped.

Even at low heating rates where these peaks are separated, the start

temperature of TE2 seems to be before the end of TE1.

(v) It is not possible to explain the appearance of TE1 on the basis of the

calculated phase equilibria. We therefore suggest that TE1 is due to

transformation towards equilibration of the initial thermomechanically

forged alloy. Though X-ray analysis shows the presence of α

, γ and

B2 phases in the forged state, the phase composition in terms of the

amounts of phases may differ from the equilibrium one. Heating to

above 900 °C would allow diffusion processes to take place that would

lead to equilibration of the phase constitution. In addition, in the

temperature range where TE1 and TE2 exist, homogenisation diffusion

processes in the γ phase occur. These are shown by X-ray measurements

(Chapter 3 and Fig. 4.8). An experiment with two heating cycles to a

temperature just above the TE1 further proves this. TE1 is clearly

present during the first heating, but it is not present during cooling or

during repeat heating. Hence, we suggest that this peak is related to, in

the main, the non-equilibrium structure after thermomechanical

processing.

The thermodynamic calculations suggest that, for the composition, the

homogeneous α phase (after completion of TE5) at heating would transform

directly through peritectic transformation of α to α + β + L without the α to

α + β phase transition before that. The α to α + β transition occurs in the

binary Ti–Al phase diagram. Our calorimetry data show that the thermal

effect TE6 could be assigned only to the solid state α to α + β phase transition,

and TE7 involves the liquid phase. Indications for this are the appearance

and shape of the TE6 peak as well as the temperature range where it appears.

TE6 is not completed for heating with rates of 30, 40 and 50 °C/min, probably

because of kinetic reasons. Next, TE7, which has just started at about

1440 °C for small heating rates (5 and 10 °C/min), most probably is due to



Titanium alloys: modelling of microstructure80

the beginning of partial melting (α + β to α + β + L). In these samples, after

the calorimetry experiments, there were partially melted parts around the

edges. The calorimetry measurement is not an equilibrium process. It is

possible for the calorimetry results to deviate from the thermodynamic

calculation. The alloy composition of each phase cannot reach uniformity at

any time due to the diffusion barrier. The transformations α → α + β → α

+ β + L may be interpreted according to the non-equilibrium reason.

Table 4.1 summarises these effects.

4.1.4 Kinetics of the γ to α phase transformation

As discussed in the previous section, a variety of phase transformations take

place during continuous heating. The different phase transformations appear

with different thermal effects, affected differently by the heating rate.

Since the γ to α phase transformation is the major phase transformation

and has practical importance in the heat treatment of the gamma titanium

aluminides, here, we pay a special attention to it. The transformation involving

γ at continuous heating starts at TE3, where the α

phase disappears. However,

in some temperature range, from TE3 to TE4, it proceeds with the presence

of another phase – B2. In its pure form, where only the two phases exist, the

γ to α phase transformation starts after TE4, when the B2 phase also disappears.

The pure γ to α phase transformation is represented by TE5. This phase

transformation does not start with 100% γ phase but with a mixture of about

66% γ + 34% α phases (at the end of TE4), and completes with homogeneous

100% α at the α-transus temperature.

In order to reveal the kinetics of the γ to α phase transformation from the

calorimetry signal, we extract the signal corresponding to pure γ to α phase

transformation only (Fig. 4.9), without the presence of other phases. This is

done by baseline connecting the end of TE4, where the B2 phase disappears

Table 4.1

Thermal effects in the differential scanning calorimetry curves

at continuous heating and corresponding phase transformations of Ti-

46Al-1.9Cr-3Nb alloy

Thermal Phase transformation

effect

TE1 Equilibration and homogenisation of the alloy

TE2 Change of phase ratios between α

, γ and B2 phases. Increase

of B2 in respect to α

and γ

TE3 α

+ γ + B2 to α + γ + B2

TE4 α + γ + B2 to α + γ (disappearance of B2)

TE5 γ + α to α

TE6 α to α + β

TE7 α + β to α + β + L



Differential scanning calorimetry and property measurements 81

and pure γ to α phase transformation starts, and the end of TE5, where this

transformation completes with 100% of α phase. The calculated enthalpy

values using the processed calorimetry curves are similar for the different

heating rates (Table 4.2), confirming that these calorimetry signals correspond

to one same phase transformation at different heating rates.

The degree of transformation is calculated (Fig. 4.10), using the method

detailed in Chapter 7. The calculated curves trace the course of the γ to α

phase transformation in the Ti-46A1-1.9Cr-3Nb alloy at continuous heating.

Transformed fraction does not mean the amount of the α phase. At zero

transformed fraction, the phase composition is 66% γ + 34% α, while at

50 °C/min

40 °C/min

30 °C/min

20 °C/min

10 °C/min

5 °C/min

endo. ⇒

Heat flow (W/g)

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

1200 1220 1240 1260 1280 1300 1320 1340 1360 1380

(°C)

4.9

Processed calorimetry curves for the γ to α phase transformation

in Ti-46Al-1.9Cr-3Nb alloy at different heating rates.

Table 4.2

Enthalpy of the γ + α to α phase transformation

at continuous heating with various heating rates

Heating rate (°C/min) Enthalpy (J/g)

50 65.9

40 69.6

30 68.1

20 70.6

10 72.0

5 78.2



Titanium alloys: modelling of microstructure82

transformed fraction = 1, the phase composition is 100% α phase. Based on

this, we calculate the amounts of the α phase at different temperatures and

heating rates. The iso-lines in Fig. 4.11 show the dependence of the amount

of the α phase on the heating rate. Such a thermo-kinetic diagram can be

used to trace the course of transformation in real products, where the heating

rate varies significantly from the surface to the core. This is important for

titanium-based materials that have relatively low heat conduction coefficients.

4.1.5 Summary

The Ti-46Al-1.9Cr-3Nb alloy after being forged at 1200 °C without further

treatment has a duplex microstructure, consisting of fine equiaxed and lamellar

γ grains, with a small amount of α

plates and particles and about 1 wt.% B2

phase. Differential scanning calorimetry can reveal, reproducibly, several

thermal effects upon heating of the as-forged alloy. These thermal effects are

related to the equilibriation and homogenisation of the alloy, change of

phase ratios between α

, γ and B2 phases (in particular, the increase of B2 in

respect to α

and γ), and the following five phase transformations: α

+ γ +

50 °C/min

40 °C/min

30 °C/min

20 °C/min

10 °C/min

5 °C/min

Transformed fraction

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

1200 1220 1240 1260 1280 1300 1320 1340 1360 1380

(°C)

4.10

Calculated transformed fractions as a function of the

temperature for the γ to α phase transformation in Ti-46Al-1.9Cr-3Nb

alloy at different heating rates.



Differential scanning calorimetry and property measurements 83

B2 → α + γ + B2 → α + γ → α → α + β → α + β + L. The observation of

these transformations by differential scanning calorimetry is largely in

agreement with literature phase diagrams and thermodynamic calculations,

though care is needed to consider the different alloy compositions. Kinetics

of the γ + α to α phase transformation in the Ti-46Al-1.9Cr-3Nb alloy can be

quantitatively derived from the calorimetry data, giving phase compositions

at any point during the transformation upon continuous heating.

4.2 Mechanical properties of

ββ

β21s alloy

This section investigates and shows the effect of heat treatment parameters

with respect to the microstructure and mechanical properties of β-type titanium

alloys. β alloys are the most versatile class of titanium alloys. They offer

very attractive combinations of strength, toughness, and fatigue resistance,

at large cross-sections. The disadvantage compared to α + β alloys is an

increase in density and cost.

β21s (Ti-15Mo-3Al-2.7Nb-0.25Si) is a relatively recently developed

metastable β alloy. Molybdenum, as an alloy agent to titanium, up to 15%,

provides oxidation resistance and improves corrosion resistance. Aluminium

is an α stabiliser, but it is added to β21s to improve ductility and to reduce

the weight. Strip is the main product form, and the material can be economically

rolled to foil.

(°C)

1350

1340

1330

1320

1310

1300

1290

1280

1270

1260

1250

1240

01020304050

Heating rate (°C/min)

Alpha transus temperature

99 % alpha

90 % alpha

80 % alpha

70 % alpha

60 % alpha

50 % alpha

40 % alpha

4.11

Calculated thermo-kinetics diagram plotted as temperature

versus heating rate. Iso-lines show the amounts of the α phase.



Titanium alloys: modelling of microstructure84

β21s is metallurgically stable for at least 1000 hours at temperatures up to

615 °C. However, at elevated temperatures, oxygen absorption at the surface

degrades the tensile ductility of the material.

Five different thermal processing conditions are considered here (Fig.

4.12):

(i) aged (this is the starting state of the alloy, which was processed through

β solution treatment followed by ageing)

(ii) heat treated and then water quenched,

(iii) water quenched and ageing at 480 °C for 8 hours,

(iv) water quenched and ageing at 540 °C for 8 hours,

(v) water quenched and ageing at 595 °C for 8 hours.

The thermal processing has produced a fully equiaxed microstructure with α

phase precipitating at the grain boundaries of the β phase with varying grain

sizes (Fig. 4.13).

The Vickers hardness after water quenching decreases as compared to the

initial aged condition (see Fig. 4.14), due to the formation of mostly metastable

β microstructure during fast cooling from the β homogenisation field. Ageing

after quenching causes an increase of hardness, as compared to the quenched

condition. The reason for this is the formation of fine α precipitates during

ageing. There is a clear trend showing that as the ageing temperature is

increased, the hardness decreases. The reason for this decrease is the formation

of coarser precipitates during ageing at higher temperature, and possibly

relief of the micro-stresses.

The same trend, i.e. decrease at higher temperatures of ageing, is observed

with the ultimate tensile strength (Fig. 4.15), for the same reasons as for

hardness changes. The trend for change of the elongation, a measure of the

6 hours

810 °C

8 hours

Water quench

595 °C

540 °C

480 °C

4.12

Schematic presentation of heat treatments for β21s alloy.

