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

Подождите немного. Документ загружается.

Microscopy 25

Wen et al. (2000). It is generally believed that the key to high strength in

fully lamellar TiAl-based alloys lies in refining the lamellar size rather than

the grain size. Quenching from the α phase field and further ageing can be

applied to refine lamellar thickness to less than 10 nm (Yamaguchi et al.,

2000). Transmission electron microscopy (TEM) of the fully lamellar structure

is needed to establish fully the relationship between the grain size and the

lamellar thickness in these alloys. This will be discussed in the next section.

2.2.3 Summary

Heat treatment of the Ti-Al alloy at 1450 °C followed by cooling leads to the

formation of the fully lamellar microstructure which consists of γ lamellae

mostly and of a small amount of α

lamellae. Air cooling and water quenching

lead to a refinement of the grain size of the fully lamellar microstructure

compared with furnace cooling. The grain size of the lamellar structure falls

into the hundreds of micrometer range and is refined with increasing cooling

rate. Increasing the holding time makes the grain structure coarser. The

lamellar thickness varies from 0.2 to 2 micrometers.

2.3 Transmission electron microscopy of

microstructural evolution

Section 2.2.2 and Chapter 3 describe the effect of different heat treatments

on the grain size and α

and γ phases distribution, respectively. This section

discusses transmission electron microscopy of the microstructure of the Ti-

46Al-1.9Cr-3Nb alloy, in particular the lamellar thickness, before and after

heat treatment.

2.3.1 Chemical composition

Table 2.2 summarises and Fig. 2.12 illustrates the composition of α

and γ

phases in the forged state and in the alloy after heat treatment. Table 2.2 does

Table 2.2

Energy dispersive X-ray mapping of the elemental composition of the α

and γ phases before (primary equiaxed structure) and after (lamellar structure)

heat treatment

Analysed phase Al (at.%) Ti (at.%) Cr (at.%) Nb (at.%)

Primary equiaxed α

38.7 55.7 2.6 3.0

Primary equiaxed γ 49.1 46.9 1.2 2.9

Lamellar

(a)

42.1 53.2 2.1 2.6

Lamellar γ

(b)

53.2 42.1 1.4 3.3

(a)(b)

The positions in Fig. 2.12 where the analyses were made.



Titanium alloys: modelling of microstructure26

not present an average composition for the two phases, but the local

composition. Place-to-place variations of the composition are quite significant

here.

The α

phase is supersaturated with aluminium in a lamellar structure in

comparison with the forged state, which is fairly close to the equilibrium

0 0.5 1 1.5 2 2.5

Distance (m)

× 10

–6

Counts

3000

2500

2000

1500

1000

500

2.12

Transmission electron microscopy bright field micrograph

showing α

and γ lamellar structure for the Ti-46Al-1.9Cr-3Nb alloy

cooled from 1450 °C at 5 °C/min and corresponding energy

dispersive X-ray profile of each element. Letters (a) and (b) indicate

the positions at which the elemental composition for α

and γ phases

is shown in Table 2.2.

→



Microscopy 27

concentration of about 37 at.% aluminium for the α

phase. The supersaturation

of aluminium in the lamellar α

phase, as well as a high aluminium

concentration in γ lamellae in comparison with the composition in the forged

state, is indicative of a slow diffusion process. The heat treatment at 1450 °C

before further furnace cooling, which was applied to the present alloy, lies in

the α + β phase field and does not completely equilibrate the composition.

The same correlation, although with slightly different values of the elemental

composition, was reported by Chen et al. (2001) and Qin et al. (2000).

Nearly equal values of aluminium content (and titanium balance) for both

phases, received by energy-dispersive X-ray (EDX) analysis of rapidly cooled

samples of Ti-45Al-2Nb-2Mn alloy (Prasad et al., 2002) might be explained

only by the very high cooling rates that were applied.

As opposed to earlier results (Chen et al., 2001; Prasad et al., 2002; Qin

et al., 2000) that stated that niobium is almost homogeneous in both phases,

the results in Table 2.2 and Fig. 2.12 clearly show that niobium is enriched

in the γ lamellar phase, while chromium has the totally opposite tendency.

Only in the forged state does niobium not exhibit any preferential partitioning

between the two phases.

The EDX analysis undoubtedly proves that the dark and the bright lamellae

in the TEM bright field image have different chemical composition,

corresponding to the composition of the α

and γ phases, respectively.

2.3.2 Microstructural analysis

Microdiffraction of the fully lamellar structure of Ti-46Al-1.9Cr-3Nb after

the heat treatment is shown in Fig. 2.13, which shows the example of a

sample cooled at 5 °C/min. The electron diffraction data prove directly that

the dark and the bright lamellae in the TEM bright field image belong to

different crystal structures – ordered α

(Ti

Al with the DO

structure based

on an hcp lattice) and γ (near-cubic face-centred tetragonal L1

crystal

structure), respectively. This is in strict correspondence with studies by Gupta

and Wiezorek (2003).

The lamellar phase thickness is defined as the edge-to-edge dimension

(measured perpendicular to the phase boundary) of the adjacent lamellae for

a given phase. The existence of γ variant is not seen in the TEM micrographs

(Figs. 2.12 and 2.13) and therefore each variant is not discriminated. The

thickness of the bright regions is measured, whatever the number of variants

(invisible), so the measured thickness is equivalent to inter-α

thickness or

inter-spacing. The average lamellar thicknesses of the fully lamellar

microstructure for each phase after different cooling rates are given in Table

2.3. Figure 2.14 shows log-normal distribution of the thicknesses of γ and α

lamellae.

A number of authors have studied the lamellar thickness of fully lamellar



Titanium alloys: modelling of microstructure28

structure regardless of lamellar phases at different conditions, and they have

reported the lamellar thickness to be in the range of 0.1–2 µm, depending on

heat-treatment, processing and alloy composition (Zhang et al., 2000). Some

authors distinguish γ and α

phases (Gupta and Wiezorek, 2003). There are

data on the dependence of mean lamellar thickness on cooling rate (Beschliesser

et al., 2002; Cao et al., 2000; Wen et al., 2000). The lamellar thickness of

both α

and γ phases is refined with increasing cooling rate.

Figure 2.15 shows typical fully lamellar microstructure, revealing an

important aspect of characterising gamma alloys: the lamellae develop in a

regular fashion with alternating γ and α

laths. This occurrence is common

2.13

Transmission electron microscopy bright field micrograph

showing α

and γ lamellar structure for the Ti-46Al-1.9Cr-3Nb alloy

cooled from 1450 °C at 5 °C/min. The insets show selected area

diffraction (SAD) patterns corresponding to (a) the [010] axis of α

phase, (b) the [111] axis of γ phase, (c) the [110] axis of γ phase and

(d) the overlapped pattern with [111] γ and [010] α

axes.

Table 2.3

γ and α

phases’ lamellar thickness distribution in Ti-46Al-1.9Cr-3Nb alloy

after furnace cooling with two different cooling rates

γα

Cooling rate Average Standard deviation Average Standard deviation

(°C/min) (µm) (µm) (µm) (µm)

5 0.88 0.73 0.61 0.55

40 0.29 0.19 0.21 0.14



Microscopy 29

05 °C/min

40 °C/min

Lamellar thickness (nm)

(a)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

Occurrence

05 °C/min

40 °C/min

Lamellar thickness (nm)

(b)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

Occurrence

2.14

Lamellar thickness distribution for the Ti-46Al-1.9Cr-3Nb alloy

furnace cooled from 1450 °C with 5 and 40 °C/min cooling rates:

(a) γ lamellae, (b) α

lamellae.



Titanium alloys: modelling of microstructure30

for the quaternary alloy composition and the processing scheme illustrated

in this chapter.

Each α

lamella (bright in Fig. 2.15) may vary significantly in thickness,

or even terminate, which implies considerable interfacial curvature of the

/γ boundaries. The curvature and its change are largest around the end of

a terminated α

. Fine, fragmented or terminated α

lamellae are frequently

observed along with thicker α

and γ lamellae. The frequent presence of

terminating γ and α

lamellae and of significantly curved interface segments

implies that the instability mechanism of termination migration is chiefly

responsible for the coarsening of the fully lamellar microstructure. Each

event of boundary splitting of a lamella produces two terminations, which

then can propagate relatively rapidly by diffusional processes at the curved

migrating interface segments and coarsen the fully lamellar structure (Gupta

and Wiezorek, 2003).

The formation of γ phase within the α

lamellae requires a change in the

crystal structure as well as a change in concentrations of titanium, aluminium

and alloying elements. One of the possible mechanisms of the transformation

of the α

structure to γ structure is a ledge migration mechanism. The process

of continuous coarsening of γ is carried out then by the movement of Shockley

partial dislocations. This diffusion controlled migration process is supported

by the observation of the ledges along α

–γ interfaces (Beschliesser et al.,

2002). Although the decomposition process had already started during cooling

with relatively slow cooling rates, as applied to the alloy here, long-term

exposure should be applied in order to reveal all the details of the process.

2.15

Scanning transmission electron microscopy (STEM) dark field

micrograph showing α

(bright) and γ (dark) lamellar structure for the

Ti-46Al-1.9Cr-3Nb alloy cooled from 1450 °C at (a) 5 °C/min and

(b) 20 °C/min.

(a) (b)



Microscopy 31

Transfer of slip either via dislocations or twinning in the γ phase is apparently

not always possible since fracture has been observed frequently at α

–γ

interfaces. It is possible that specific dislocation sources are required at these

interfaces in order for slip to take place in the α

before fracture intervenes.

The relatively coherent boundaries that are characteristic of lamellar structures

are not likely to contain a high density of such sources.

2.3.3 Summary

The dark and the bright lamellae in the TEM bright field image belong to the

different crystal structures – ordered α

(Ti

Al with the DO

structure based

on an hcp lattice) and γ (near-cubic face-centred tetragonal L1

crystal

structure), respectively. The lamellar thickness of both α

and γ phases varies

significantly, falling into the hundreds of nanometres range and is refined

with increasing cooling rate.

2.4 References

Beschliesser M, Chatterjee A, Lorich A, Knabl W, Kestler H, Dehm G and Clemens H

(2002), ‘Designed fully lamellar microstructures in a γ-TiAl based alloy: Adjustment

and microstructural changes upon long-term isothermal exposure at 700 and 800 °C’,

Mater Sci Eng A, 329–331, 124–29.

Cao G, Fu L, Lin J, Zhang Y and Chen C (2000), ‘The relationships of microstructure and

properties of a fully lamellar TiAl alloy’, Intermetallics, 8 (5–6), 647–53.

Chen S H, Schumacher G, Mukherji D, Frohberg G and Wahi R P (2001), ‘The effect of

local composition on defect in a near-γ-TiAl alloy with duplex microstructure’, Phil

Mag A, 8 (11), 2653–64.

Guo F A, Ji V, Zhang Y G and Chen C Q (2001), ‘A study of mechanical properties and

microscopic stress of a two-phase TiAl-based intermetallic alloy’, Mater Sci Eng A,

315 (1–2), 195–201.

Gupta A and Wiezorek J M K (2003), ‘Microstructural evolution of PST-TiAl during low-

rate compressive micro-straining at 1023 K in hard and soft orientations’, Intermetallics,

11 (6), 589–600.

Hu D (2001), ‘Effect of grain refinement on continuous cooling phase transformation in

some TiAl-based alloys’, in: Winstone M R (ed), Titanium Alloys at Elevated

Temperature: Structural Development and Service Behaviour, London: IoM

Communications, 263–75.

Hu D and Botten R R (2002), ‘Phase transformations in some TiAl-based alloys’,

Intermetallics, 10 (7), 701–15.

Kobayashi S, Nakai K and Ohmori Y (2001), ‘Analysis of phase transformation in a Ti-

10 mass % Zr alloy by hot stage optical microscopy’, Mater Trans, 42 (11), 2398–405.

Prasad U, Xu Q and Chaturvedi M C (2002), ‘Effect of cooling rate and manganese

concentration on phase transformation in Ti–45 at.% Al based alloys’, Mater Sci Eng

A, 329–331, 906–13.

Qin G W, Smith G D W, Inkson B J and Dunin-Borkowski R (2000), ‘Distribution

behaviour of alloying elements in α

(α)/γ lamellae of TiAl-based alloy’, Intermetallics,

8 (8), 945–51.



Titanium alloys: modelling of microstructure32

Wang J N and Xie K (2000), ‘Refining of coarse lamellar microstructure of TiAl alloys

by rapid heat treatment’, Intermetallics, 8 (5–6), 545–48.

Wen C E, Yasue K, Lin J G, Zhang Y G and Chen C Q (2000), ‘The effect of lamellar

spacing on the creep behavior of a fully lamellar TiAl alloy’, Intermetallics, 8 (5–6),

525–29.

Winstone M R (ed) (2001), Titanium Alloys at Elevated Temperature: Structural Development

and Service Behaviour, London: IoM Communications.

Yamaguchi M, Inui H and Ito K (2000), ‘High-temperature structural intermetallics’,

Acta Mater, 48 (1), 307–22.

Zhang D, Dehm G, Clemens H (2000), ‘On the microstructural evolution and phase

transformations in a high niobium containing gamma-TiAl alloy’, Z Metallkd, 91 (11),

950–56.



Synchrotron radiation X-ray diffraction

Abstract: This chapter concentrates on the use of high-resolution

synchrotron X-ray diffraction technique for research into phase

transformations in titanium alloys and aluminides. Both room and high

temperature measurements are described, the former for studying the

structure of the alloys after different heat treatments and the latter for in situ

study of the transformations on the sample surface at elevated temperatures.

The phase equilibria of the alloys are calculated and compared with the

molar phase fractions of phases derived after the fitting procedure of the

diffraction patterns at elevated temperatures.

Key words: X-ray diffraction, phase transformation, heat-treating,

synchrotron radiation, crystal structure.

3.1 Introduction

In situ studies of the phase transformations are not common for titanium

alloys. Most of the research on the microstructure evolution and the kinetics

of the phase transformations is performed at room temperature on samples

quenched after heat treatments.

Titanium alloys are appropriate for applications at elevated temperatures.

Understanding the processes and transformations taking place on the alloys’

surface during high temperature exposure is important for reliable exploitation.

High-resolution X-ray diffraction (HR-XRD) available at a synchrotron

radiation source (SRS) is currently a very powerful technique for detailed

studies of structure and phase transformations in materials. The main advantages

are the inherent high brightness and the tuneable monochromatic beam,

which allow the development of a high-resolution powder diffraction

instrument. Further developments allow a variety of experiments by users

from a diverse scientific community (MacLean et al., 2000). With a relatively

small instrumental contribution (∆2

of the order of 0.006°), it can be used

for diffraction line profile analysis of microstructures. These facilities allow

in situ studies of materials phenomena occurring at elevated temperatures.

Berberich et al. (2000) has reported on the potential use of synchrotron-

based methods to study various effects in titanium alloys.

The main aim of this section is to illustrate room- and high-temperature

synchrotron radiation HR-XRD measurements to study the phase

transformations taking place in the most commonly used titanium alloys.

This should lead to an understanding of the processes at the alloys’ surfaces

during high-temperature exposure.



Titanium alloys: modelling of microstructure34

3.2 Measurements at room temperature

The diffraction patterns obtained from the SR measurements at room

temperature for Ti 6-4, Ti 6-2-4-2 and β21s alloys with different heat treatment

conditions are shown in Figs. 3.1, 3.2 and 3.3.

Ti-6Al-4V Room temperature

30 32 34 36 38 40

2θ (°)

Relative intensity

3.5

2.5

1.5

0.5

{100}α(α′)

{002}α(α′)

{101}α(α′)

Water quenching

+ ageing

Water quenching

from 850 °C

Furnace cooling

from 1100 °C

Annealed

{110}β

3.1

Diffraction patterns at room temperature and profile fits (dotted

lines) in the range of 30–40° 2

for Ti-6Al-4V under different heat

treatment conditions. The intensities are given relative to the {101}α

reflection. For clarity, the diffraction patterns are shifted with respect

to each other along the vertical axis.

