黑料网

TiAl based alloys represent an important class of alloys that exhibit a unique combination of physical and mechanical properties which include low density of 3.7 g/cm³-4.7 g/cm³, high elastic modulus, high creep resistance and good resistance against oxidation and corrosion which promote this binary alloy as a serious candidate to replace conventional structural materials in energy, automotive and aerospace industries [1-3]. However, after more than three decades of intensive research, very few TiAl based alloys did obtain a satisfactory level of maturity to be used for critical components in turbines and aircraft engines, mainly due to reduced toughness and low ductility at room temperature in comparison to those of conventional structural materials [4].

In order to improve mechanical properties and expand the application of TiAl based alloys, corresponding relations between mechanical properties and microstructure have been investigated widely, in the same context, a great number of studies have been made on the fracture mechanism of TiAl based alloys [5-11]. However, studies on the compressive behaviour are relatively rare. A large difference of deformability between tensile and compression tests was revealed, this variance is caused by the difference of stress and strain conditions. CT Liu showed that the yield strength was a function of the test temperature measured by compression tests at various test temperatures [11]. Pu, reported that the compression strength of Ti-Al-Cr-V alloy will increase at 1170°C when the strain rate changes from 0.001 s^-1 to 100 s^-1 [12].

Several approaches have been made to overcome the insufficiencies of TiAl based alloys. Among these methodologies, element alloying, which consist of introducing innovative elements into TiAl based alloy composition, has received considerable attention in the purpose of meet the optimal composition [13,14].

In this work, microstructure of heat treated, Ti45Al5Nb4V0.1Si and Ti45Al5Nb4V (Mo, Si (at % alloys were studied and hot deformation behaviours were discussed.

The alloys with different compositions have been prepared from high purity elements (>99.8% purity). The prepared mixtures were arc melted with an induction furnace at least three times to ensure good homogeneity of the samples. The melting was conducted in a pure argon gas atmosphere.

As titanium and aluminium are very prone to oxygen contamination that may seriously affect the result of our elaboration, pure elements grits were cleaned in nitric acid (HNO3:1M), rinsed and dried, then weighed in stoichiometric proportions before introduction to the elaboration furnace.

Samples were heat treated after furnace elaboration, considering the affinity between titanium and aluminium with oxygen, the heat treatment was done under argon atmosphere. Samples were heated to 1200°C for 16 hours followed by slow cooling inside the furnace. Table 1 gives a summary of chemical composition and processing of prepared alloys.

Table 1: Chemical composition (% at) and processing of prepared alloys.

In order to perform metallographic observations, samples were sectioned, polished and observed by optical and Scanning Electron Microscopy (SEM). Alloys microstructures were studied using Olympus LEXT OLS4100 laser scanning microscope and Hitachi SU 8230 scanning electron microscope equipped with EDX (Energy Dispersive X-ray) analysis used to determine phase composition (semi quantitative).

In order to evaluate compressive behaviour, samples with 5 mm in diameter and 8 mm in height were cut from the prepared mass. Isothermal compression tests were conducted on a quenching dilatometer DIL 805A/D machine with argon atmosphere at temperature of 1323 K and strain rates of 0.05 s^-1.

Before the compression, each Specimen was heated at a rate of 5 k/s and maintained for 5 min at target temperature for homogenization. The maximum strain obtained in the tests was 50%. Once compression ended, the specimens were quenched to room temperature immediately with cooling rate of 30 C/s. True stress curves were obtained.

Starting microstructures

Several studies have confirmed that initial microstructure of TiAl alloys, before deformation, has a significant role in final mechanical properties, following this aspect, microstructural assessment was carried out for the casted alloys [15-19].

Figure 1 expose the starting microstructure heat treated of Ti-45Al-5Nb-4V-0.3Si alloy. The material shows a nearly lamellar structure composed by large colonies of γ/α₂ along with coarse γ phase, Energy Dispersive Spectra (EDS) pointed at +1, +2 and +3 and listed in Table 2 summarize concentrations of addition elements in the alloy. Nb as addition element is homogenously spread across both the γ/α₂ and the γ phase while vanadium and silicon preferentially positioned in the γ phase. XRD pattern where γ-TiAl and α₂-Ti₃Al phases are confirmed with the identification of β phase in significant less extend.

Figure 1: Scanning Electron micrographs (SE mode) of Ti-45Al-5Nb-4V-0.3Si alloy.

Table 2: Results of energy dispersive spectra of the alloys.

Microstructure of second alloy Ti-45Al-5Nb-4V-0.5Mo-0.1Si is depicted (Figures 2 and 3). Similarly to previous material, the microstructure exhibit two main regions. XRD analysis, presented in Figure 4, confirmed composition is mainly γ/α₂ along with γ phase and β phase. In this case, the addition of molybdenum and reduction of silicon content, comparatively to previous alloy, had qualitatively enlarged lamellar γ/α₂ and decreased γ phase ratio. It is generally admitted that Mo addition has a positive impact on ductility [20]. EDS results are summarized in Table 2.

Figure 2: Scanning Electron micrographs (SE mode) of Ti-45Al-5Nb-4V-0.5Mo-0.1Si alloy.

Figure 3: Scanning Electron micrographs (SE mode) of Ti-45Al-5Nb-4V-0.5Mo-0.5Si alloy.

Figure 4: XRD patterns of (a). Ti-45Al-5Nb-4V-0.3Si, (b). Ti-45Al-5Nb-4V-0.5Mo-0.1Si, (c). Ti-45Al-5Nb-4V-0.5Mo-0.5Si.

This depict microstructure of Ti-45Al-5Nb-4V-0.5Mo-0.5Si, elevation of silicon content to 0.5% (at %) comparatively to 0.1% (at %) in previous second alloy impacted initial, heat treated microstructure with formation of long precipitation chain, EDS analysis summarized revealed high silicon concentration and confirmed precipitation nature of the ribbons that have subsisted to the heat treatment.

Deformation behaviour

Compression proprieties: The true stress strain curves of various TiAl based alloys less than 1323 K and 0.05 s^-1 are illustrated in Figure 5. It can be seen that most of flow stress trends increase to peak value before decreasing until a steady state is reached. The Ti-45Al-5Nb-4V (Mo, Si) samples deformed at 1323 K and 0.05 s^_-1 exhibit greater effect of DRX softening where Ti-45Al- Ti-45Al-5Nb-4V-0.3Si specimen shows more steady state curves.

Figure 5: True stress true strain curves of tested alloys under 1323 K and 0.05 s^-1, WH stage refers to work hardening stage, DRX stage refers to dynamic recrystallization stage.

It is well known that deformation mechanisms at hot temperatures rely on dynamic softening and work hardening processes. The hot compression process can be separated into three stages: At initial stage of deformation, flow stress increases rapidly due to increasing density of dislocations till a peak stress is reached; at this stage, work hardening exceeds dynamic softening which generates the sharply increasing tendency. Successively, the deformation energy accumulated with the growth of deformation strain will offsets the work hardening and brings forward dynamic softening mechanism which diminishes the dislocation density, inducing decrease in true stress. This deceasing tendency at this stage is characteristic of dynamic recrystallization over dynamic recovery during deformation [21,22]. When the work hardening and dynamic softening reach a dynamic equilibrium, the dislocation density remains relatively constant, and a steady state flow stress is obtained as shown on the curves.

As the true stress of the alloys increases with the increase in the strain rate in the first work hardening stage. The alloys exhibit significant difference in the ultimate true stress, it is observed that Ti-45Al-5Nb-4V (Mo, Si) alloys shows higher strength than Ti-45Al-5Nb-4V-0.3Si. Because the alloys have similar overall starting microstructures, the differences in the compression behaviour could be mainly due to the difference in solid solution of alloying elements such as Mo and V and difference in dislocation density.

Microstructures after deformation: This shows microstructures of hot compressed alloys, deformed at 1323 K and 0.05 s^-1.

The stressed materials exhibit deformed microstructure and heavily elongated γ/α₂ colonies in direction perpendicular to force application in all three alloys.

In Figure 6, Ti-45Al-5Nb-4V-0.3Si lamellar colonies are shown elongated perpendicularly to compression force. Different dislocation movement and slips systems were activated to process the deformation resulting twisting and bending in multiples orientations. The lack of clear dynamic recrystallisation manifestation suggests that compression at 1323 K and 0.05 s^-1 didn’t provide sufficient energy to reach recrystallization stage.

Figure 6: Scanning electron micrographs of Ti-45Al-5Nb-4V-0.3Si alloy after the compression test.

With addition of molybdenum, Ti-45Al-5Nb-4V-0.5Mo-0.1Si alloy shows, in Figure 7, deformed microstructure with evidence of dynamic recrystallization. Both α₂ and γ phases undergo driving force for recrystallization resulting from the energy of deformation [23]. Recrystallization driving force are dependent to deformation subjected, therefore, in γ/α₂ colonies with higher deformation (i.e. higher forces) dynamic recrystallization took place and fine DXR grains are formed [24,25].

Figure 7: Scanning electron micrographs of Ti-45Al-5Nb-4V-0.5Mo-0.1Si alloy after the compression test.

Figure 8 show microstructure of Ti-45Al-5Nb-4V-0.5Mo-0.5Si after hot compression test. The long precipitate chains were fractured into small coarse grains. Similarly to previous alloy, dynamic recrystallization process was identified with appearance of new fine DXR grains. When recrystallization driving force are strong enough, the precipitates act as nucleation sites for recrystallization and enhance microstructural refinement [26,27].

Figure 8: Scanning electron micrographs of Ti-45Al-5Nb-4V-0.5Mo-0.5Si alloy after the compression test.

Dynamic recrystallization volume fraction: Dynamic Recrystallization (DRX) commence with the increase of dislocation density which is a thermally activated process [28]. With further plastic deformation heat, more potential DRXed nucleus will form resulting in new recrystallized grains [29]. DRX behavior can be further examined with the calculation of the recrystallization volume fraction using the modified Avrami type equation:

Where X_DRX is the volume fraction of DRXed grains, ε^* is the strain for maximum softening rate; ε_c is the critical strain and k and n are material constants [30].

From equation (1) above, it is shown that critical strain is important in calculation and prediction of microstructural DRX evolution. Several mathematical models have been proposed by researchers to spot the critical condition for the commencement of dynamic recrystallization using flow stress curves. In this work, the work hardening rate (θ=∂σ/∂ε) was calculated [31,32].

The θ-σ curve of Ti-45Al-5Nb-4V-0.5Mo-0.1Si alloy under 1323 K and strain rate of 0.05 s^-1 is shown in Figure 9. The work hardening curve show typically four stages: stage I begins from initial stress to the critical stress (σ_c), where the value of θ decreases sharply, and the inflection point on the θ–σ curve demonstrates the onset of DRX. Stage II covers the transition between the DRX commencements (corresponding to σ_c) to the peak stress, where θ hits zero. Stage III is where θ reaches the lowest value indicates maximum softening and DRX prevalence region, it begins at σ_p and ends at σ^*. Last stage IV is where equilibrium between work hardening and DRX softening occurs and θ=0 [33,34].

Figure 9: Hardening rate curve of Ti-45Al-5Nb-4V-0.5Mo-0.1Si alloy under 1323K and strain rate of 0.05 s^-1.

J Liu proposed that flow stress function of DRX can be expressed as:

Where σ_p is the peak stress and σ_s is the steady state stress.

Based on equation (1), the below is obtained:

With the condition of 1323 K and 0.05 s^-1, the relationship between ln [-ln (1−X_DRX)] and ln [(ε−ε_c)/ε^*] is plotted and the values of n and k are determined to be 1.021 and 0.229, respectively.

Following the findings, volume fraction of Ti-45Al-5Nb-4V-0.5Mo-0.1Si alloy in testing conditions can be expressed as:

DRX volume fraction is shown in Figure 10 where it can be seen that the dynamic recrystallization will not occur at strain less than the critical strain ε_c. The volume fraction of DRXed grains increases as the strain increases.

Figure 10: DRX volume fraction curve of Ti-45Al-5Nb-4V-0.5Mo-0.1Si alloy under 1323 K and strain rate of 0.05 s^-1.

In an effort of assessing addition element impact on mechanical proprieties in intermetallic TiAl based alloys and toward mechanical performance enhancement, alloys with nominal composition of Ti-45Al-5Nb-4V-0.5Mo-0.1Si, Ti-45Al-5Nb-4V-0.5Mo-0.5Si, and Ti-45Al-5Nb-4V-0.3Si (at %) were casted, heat treated, the obtained materials were subsequently mechanically assessed using hot compression test at 1323 K and strain rate of 0.05 s^-1. The resulting microstructures were further evaluated through microstructural analysis.

• The nearly lamellar microstructure exhibits typical compression behaviour of intermetallic TiAl alloys with dynamic flow and softening features seen on the true stress true strain curves.

• Under hot compression of 1323 K and strain rate of 0.05 s, dynamic recrystallization behavior was not obviously reached for Ti-45Al-5Nb-4V-0.3Si suggesting that deformation activation energy of the alloy is larger than the other casted alloys. This observation further indicates that Ti-45Al-5Nb-4V-0.3Si alloy tend to have low hot workability when compared to Ti-45Al-5Nb-4V (Mo, Si) tested alloys for which dynamic recrystallization was more obvious.

• Molybdenum addition at 0.5% (at %) improve general compression behavior. The Ti-45Al-5Nb-4V-0.5Mo-0.1Si alloy exhibit highest strength, flow stress decreases significantly toward steady state with increasing strain, indicating best hot workability behaviour.

1. Dimiduk DM (1999) . Mater Sci Eng A 263: 281-288.

   [] []

2.  Loria EA (2000) . Intermetallics 8: 1339-1345.

   [] []

3.  Clemens H, Smarsly W (2011) . Adv Mat Res 278: 551-556.

   []

4.  Leyens C, Peters M (2005) Titanium and titanium alloys fundamentals and applications. John Wiley and Sons, 77-79.
5. Zhang DY, Li HZ, Liang XP, Wei ZW, Liu Y (2014) . Mater Des 59: 415-420.

   [] []

6. Thomass S, Michael MM, Francisca S, Thomas, Boryana R, et al. (2015) . Acta Mater 94: 205-213.

   [] []

7. Yokoshima S, Yamaguchi M (1996) . Acta Mater 44: 873-883.

   []

8. Rogers NJ, Crofts PD, Jones IP, Bowen P (1995) . Mater Sci Eng A 192: 379-386.

   [] []

9. Chan KS, Kim YW (1992) . Metall Trans A 23: 1663-1677.

   [] []

10. Chan KS (2000) . Mater Trans A 31A : 71–80.

    [] []

11.  Liu CT, Schneibel JH, Maziasz PJ (1996) . Intermetallics 4: 429-440.

    [] []

12.  Pu ZJ, Wu KH, Shi J, Zou D (1995) . Mater Sci Eng A 192: 780-787.

    [] []

13.  Wang NJ, Zhu J, Wu JS, Du XW (2002) . Acta Mater 50: 1307-1318.

    [] []

14. Brotzu A,  F,  D (2014) . Intermetallics 54: 176-180.

    [] []

15. Schwaighofer E, Clemens H, Lindemann J, Stark A, Mayer S (2014) . Mater Sci Eng 614: 297-310.

    [] []

16. Erdely P, Staron P, Maawad E, Schell N, Clemens H, et al. (2018. Acta Mater 158: 193-205.

    [] []

17. Kim YW (1998) . Intermetallics 6: 623-628.

    [] []

18. Tian S, Lv X, Yu H, Wang Q, Jiao Z, et al. (2016) . Mater Sci Eng A 651: 490-498.

    [] []

19. Dimiduk D (1999) . Mater Sci Eng A 263: 281-288.

    [] []

20.  Appel F, Wagner R (1998) . Mater Sci Eng 22: 178-268.

    [] []

21. Sakai T (1995) . J Mater Process Technol 53: 349-361.

    [] []

22. Ding R, Guo ZX (2001) . Acta Mater 49: 3163-3317.

    [] []

23. Schmoelzer T, Liss KD, Kirchlechner C, Mayer S, Stark A, et al. (2013) . Intermetallics 39: 25-33.

    [] []

24. Imayev R, Oehring M, Appel F (2007) . Intermetallics 15: 451-460.

    [] []

25. Zhang S, Zhang C, Du Z, Hou Z, Lin P, et al. (2016) . Mater Des 90: 225-229.

    [] []

26. Chladil HF, Clemens H, Otto A, Güther V, Kremmer S, et al. (2006) Characterization of a β-solidifying γ-TiAl base alloy. BHM Berg Huttenmann 151: 356–36.
27. Li J, Liu Y, Wang Y, Liu B, He Y (2014) Dynamic recrystallization behavior of an as-cast TiAl alloy during hot compression. Mater Charact 97: 169-177.

    [] []

28. Werner R, Schwaighofer E, Schloffer M, Clemens H, Lindemann J, et al. (2016) . Adv Mater Res 922: 807-812.

    []

29. Wilson PW (2013) . InTech.

    []

30. Lv BJ, Peng J, Shi DW, Tang AT, Pan FS (2013) . Mater Sci Eng A 560: 727-733.

    [] []

31. Liu YH, Ning YQ, Yao ZK, Li H, Miao XP, et al. (2016) . J Alloys Compd 675: 73-80.

    [] []

32. Liu HJ, Ryan ND (2002) . Mater Sci Eng 322: 43-63.

    [] []

33. Shaban M, Eghbali B (2010) . Mater Sci Eng A 527: 4320-4325.

    [] []

34. Liu J, Cui Z, Ruan L (2011) . Mater Sci Eng A 529: 300-310.

    [] []