UDK 669.04:548.4	ISSN 1580-2949
Original scientific article/Izvirni znanstveni članek	MTAEC9, 48(6)831(2014)
EFFECT OF Ti2AlC PARTICLES ON THE MICROSTRUCTURE AND ELEVATED-TEMPERATURE-DEFORMATION PROPERTIES
OF 7-TiAl ALLOYS
VPLIV DELCEV Ti2AlC NA MIKROSTRUKTURO IN DEFORMACIJSKE LASTNOSTI ZLITIN y-TiAl PRI POVIŠANI
TEMPERATURI
Tomaš Cegan, Michal Cagala, Miroslav Kursa, Petr Kawulok, Stanislav Rusz,
Jan Jurica, Jirina Vontorova
VŠB-Technical University of Ostrava, Faculty of Metallurgy and Materials Engineering, Department of RMSTC, 17. listopadu 15,
708 33 Ostrava - Poruba, Czech Republic tomas.cegan@vsb. cz
Prejem rokopisa - received: 2013-09-30; sprejem za objavo - accepted for publication: 2014-01-14
The y-TiAl intermetallics with and without amount fractions x = 0.3 % of yttrium and with different carbon amount fractions ranging from 0.9 % to 1.6 % were prepared by vacuum-induction melting in a graphite crucible. Different amounts of carbon were obtained by varying the time of melting. The resulting samples had large rod-like carbides with volume fractions 0.6 % to 3.7 % Ti2AlC in the lamellar matrix consisting of the alternating lamellas of y-TiAl and a2-Ti3Al phases. It was found that the grain size of the matrix decreases with the increasing volume fraction of Ti2AlC carbides from 430 ^m to 35 ^m. In addition, finer carbides with a similar composition were observed. Their size was below 1 ^m and they appear mainly at the grain boundaries and also inside the lamellar grains. Compression testing revealed that all the samples had a relatively high yield strength at elevated temperature, from 917 MPa to 623 MPa at 600 °C and from 608 MPa to 547 MPa at 800 °C, but the best strength and the best yield strength were found for the sample with the lowest amount of the large rod-like Ti2AlC carbides. From these results it is concluded that the strengthening at an elevated temperature is a result of C in the solid solution and small carbide precipitates rather than of the rod-like Ti2AlC carbides.
Keywords: titanium aluminides, mechanical properties, microstructure, carbides, carbon
Intermetalne zlitine y-TiAl z mnozinskim deležem itrija x = 0,3 % in brez njega ter z različno vsebnostjo množinskega deleža ogljika od 0,9 % do 1,6 % so bile izdelane z vakuumskim indukcijskim pretaljevanjem v grafitnem lončku. Različne vsebnosti ogljika so bile dobljene pri različnih časih taljenja. Dobljeni vzorci so imeli volumenski delež Ti2AlC od 0,6 % do 3,7 % velikih paličastih karbidov v lamelarni osnovi iz lamel faz y-TiAl in a2-Ti3Al. Ugotovljeno je bilo, da se z večanjem volumenskega deleža Ti2AlC-karbidov velikost zrn osnove zmanjšuje od 430 ^m do 35 ^m. Opazilo se je tudi, da imajo drobnejši karbidi podobno sestavo. Njihova velikost je manjša od 1 ^m, pojavljajo se večinoma po mejah zrn in tudi v lamelarnih zrnih. Tlačni preizkus je odkril, da imajo vsi vzorci relativno visoko mejo tečenja pri povišanih temperaturah: od 917 MPa do 623 MPa pri 600 °C in od 608 MPa do 547 MPa pri 800 °C. Najvišjo trdnost in najvišjo mejo tečenja je imel vzorec z najmanjšo vsebnostjo velikih, paličastih Ti2AlC-karbidov. Iz rezultatov je razvidno, da je utrjevanje pri povišanih temperaturah bolj posledica ogljika v trdni raztopini in drobnih karbidnih izločkov kot pa velikih paličastih Ti2AlC-karbidov.
Ključne besede: titanovi aluminidi, mehanske lastnosti, mikrostruktura, karbidi, ogljik
1 INTRODUCTION
tion hardening or precipitation of fine carbide particles.2 3 Moreover, TiAl composites with higher shares of carbi-
The advantages of gamma TiAl-based alloys such as des also exhibit excellent strength and oxidation resi-specific modulus, specific high-temperature strength and stance at elevated temperatures4. Powder metallurgy has
oxidation resistance makes them attractive candidates for	been the most frequently used method for the production
the use as high-temperature structural materials in the	of TiAl alloys with carbon contents. Additions of T2C
automotive, aerospace and power industries'. However, a	powder' graphite or Al/TiC master alloy were tested.2
problematic preparation of the final products and their	However, melting in a graphite crucible could be a much
high price present the biggest problem for their higher	cheaper flt!rnative for the preParation of this material.In
utilisation. Precise casting is the most cost-effective me-	thilwork TiAl allo/s w;th different amounts of Ti2AlC
thod for the production of the components from TiAl- carbides were produced using a graphite crucible and
based alloys1. Unfortunately, gamma TiAl-based alloys	their mechanical properties were evaluated. cannot be melted in conventional refractory crucibles
and skull-melting processes are required to prevent con-	2 EXPERIMENTAL WORK tamination.
It was demonstrated that a small amount of carbon (x The melting of the alloys was performed in a LEY-
< 0.5 %) increases the strength of TiAl-based materials	BOLD - HERAEUS IS1/FFF vacuum induction furnace
at room and elevated temperatures due to the solid-solu-	using an extruded graphite melting crucible with the
average density of 1.75 g/cm3, the porosity of 17-20 % and the maximum granularity of 0.8 mm. The charge was composed of formed bits of Ti of 3N purity, bits of Al of 4N purity and in the case of alloys alloyed with yttrium also of bits of yttrium of 3N purity. The alloying with yttrium was used in order to assess the influence of this element on the microstructure. Afterwards four alloys were melted with the nominal compositions of Ti-47Al and Ti-47Al-0.3Y (amount fractions, x/%), which were then cast into a graphite mould. All these processes ran under an atmosphere of argon of 4N5 purity. Different amounts of carbon and carbides were achieved with different times of stabilising the melt ranging from 30 s to 90 s. The final products were the castings with a diameter of 10 mm and a length of 100 mm. The samples for a metallographic investigation using light microscopy (LM), X-ray diffraction analysis (XRD) and scanning electron microscopy in the mode of back-scattered electrons (BSE) and secondary electrons (SE) were taken from the castings. The chemical composition was determined with an energy dispersive analysis (EDS). The amounts of oxygen and carbon were measured with the thermo-evolution method using the ELTRA ONH-2000 and ELTRA CS-2000 analysers. The volume fraction of carbides and grain size were determined with a computerised image analysis.
Mechanical-compression tests were performed on the cylindrical samples with a diameter of 8 mm and a length of 12 mm, which were prepared by electro-spark machining and then turned to the required size on a HAAS- ST10 CNC turning lathe. An HDS-20 simulator of hot deformation was used for mechanical testing at the strain rate of 0.05 s-1. The heating to the required temperature (800, 600, 400) °C was performed at a rate of 3 °C/s and the sample was deformed after a dwell time of 15 s at the required temperature.
3 RESULTS AND DISCUSSION
3.1 Chemical compositions and the amounts of interstitial elements
Table 1 presents the analyzed chemical compositions of the prepared alloys. The compositions of the alloys do not show any significant deviations from the nominal composition. The prepared alloys show a comparatively low amount of oxygen and, in this respect, they satisfy
the requirements for the applications of these alloys in the industry, where the maximum admissible mass fraction of oxygen is approximately 0.05 %6. An analysis of the carbon amounts in the prepared castings showed that during the melting of TiAl intermetallics in graphite crucibles the amount of carbon increased as a result of its absorption from the walls of a graphite crucible into the melt. The determined values of the carbon amounts in individual samples were, nevertheless, quite different. The highest carbon mass fraction of 0.5 % was found in alloy 1, for which the longest time of stabilisation of the melt was used, namely 90 seconds. The lowest mass fraction was found in alloy 4, namely 0.295 %, for which the time of stabilisation of the melt was 30 s.
3.2 Microstructures and the solidification behaviour
The microstructures of all the prepared alloys were dendritic (Figure 1). Dendrites were formed of alternating lamellas of the aa- and y-phases. In the interdendritic spaces the y-phase was observed. Figure 2 shows a detail of one of the lamellar dendrites (L) with the chemical composition of (53.29 ± 0.81) % Ti and (46.71 ± 0.81) % Al, and also an interdendritic y-phase, with the average chemical composition of (49.06 ± 1.07) % Ti and (50.94 ± 1.07) % Al (all the results of EDS are given in amount fractions, x/%).
Figure 1: Microstructure of sample 3 (BSE) Slika 1: Mikrostruktura vzorca 1 (BSE)
Table 1: Chemical compositions, amounts of oxygen, carbon, Ti2AlC carbides and Y-rich phases (mass fraction, w/%; volume fraction, amount fraction, x/%)
Tabela 1: Kemijska sestava, vsebnosti kisika in ogljika ter vsebnost Ti2AlC-karbidov in faz, bogatih z Y (masni delež, w/%; volumenski delež (p/%, množinski delež, x/%)
Sample	Measured composition (x/%)			Oxygen (w/%)	Carbon (w/%)	Amount of carbides (p/%)	Amount of Y-rich phases (p/%)
	Ti	Al	Y				
1	52.64 ± 0.06	47.36 ± 0.06	0	0.045 ± 0.002	0.500 ± 0.006	3.73 ± 0.82	-
2	52.42 ± 0.18	47.58 ± 0.18	0	0.038 ± 0.003	0.315 ± 0.004	1.28 ± 0.78	-
3	52.15 ± 0.01	47.60 ± 0.09	0.25 ± 0.08	0.033 ± 0.010	0.386 ± 0.009	2.59 ± 0.67	1.60 ± 0.44
4	53.23 ± 0.42	46.39 ± 0.45	0.38 ± 0.06	0.034 ± 0.003	0.295 ± 0.003	0.64 ± 0.28	1.58 ± 0.40
Figure 2: Lamellar-dendrite detail of sample 2 (BSE) Slika 2: Lamelarni dendriti vzorca 2 (BSE)
Figure 3: Yttrium-rich phase in sample 4 (BSE) Slika 3: Faza, bogata z Y, v vzorcu 4 (BSE)
In the alloys alloyed with yttrium we also observed a certain share of the phases rich in yttrium, shown in Figures 1, 3 and 4 as white areas. They were present in two different forms, particularly in the interdendritic
spaces and at the grain boundaries. They were also observed, though to a lesser extent, inside the dendrites.
The determined chemical composition of smaller phases of a lighter colour analysed with the EDS method was the following: (19.60 ± 3.84) % Y, (31.18 ± 2.88) % O, (22.95 ± 5.65) % Ti and (26.27 ± 1.90) % Al (x/%). The composition of this phase, therefore, corresponds to Y2O3, which was also found with diffraction (not presented), although almost in every particle we also identified the amounts of Ti and Al. This may have been caused by the small size of these particles and the results of the measurements might have been influenced by the environment. The established chemical composition of the larger darker phases rich in yttrium, analysed with the EDS method was the following: (26.44 ± 9.05) % Ti, (55.64 ± 2.15) % Al and (17.92 ± 7.23) % Y (x/%). This phase was identified as YAl2.
The a-phase was determined as the primary crystallisation phase since dendrite arms have a hexagonal symmetry (with an angle between the dendrite arms of approximately 45°), which corresponds with the solidification of the melt through the a-phase6. Figure 4 shows an example of the a symmetry of dendrites.
3.3 Carbide precipitation
Coarser particles of the acicular type were observed in all the prepared alloys. They typically achieved a diameter of up to 2 pm and their length varied from 10 pm to 150 pm. They were distributed almost uniformly all over the whole areas of the samples' cross-sections; only in the boundary areas their occurrence was slightly lower. Figure 5 shows an example of these particles. By analysing the chemical compositions (EDS) it was determined that these particles contained x = (44.41 ± 2.34) % Ti, x = (24.74 ± 2.47) % Al and x = (30.85 ± 3.85) % C. The titanium and aluminium amounts were approximately 2 : 1 and a bright display of these particles in the BSE mode indicates the amounts of interstitial elements. For these reasons these particles were identified, probably as Ti2AlC. The X-ray diffraction patterns
Figure 4: Hexagonal symmetry of dendrites in sample 4 (BSE) Slika 4: Heksagonalna simetrija dendritov v vzorcu 4 (BSE)
Figure 5: Coarse Ti2AlC precipitates in sample 1 (SE) Slika 5: Grobi izločki TiaAlC v vzorcu 1 (SE)
Figure 6: Fine TiaAlC precipitates in sample 2 (BSE) Slika 6: Fini izločki Ti2AlC v vzorcu 2 (BSE)
of the selected samples (not presented) also confirmed that the alloys contained, apart from the y-TiAl and a2-Ti3Al phases, also the Ti2AlC phase. However, the quantity and size of acicular carbides differed rather significantly between individual samples. The highest
SSrW;



r'^. S-.-»
Figure 7: Macrostructure of sample 4 Slika 7: Makrostruktura vzorca 4
volume fraction and the largest carbides were found in alloys 1 and 3, in which the highest amount of carbon was also established. In samples 2 and 4 the carbides were observed only in some parts and they were smaller. The amounts of carbides determined with an image analysis are presented in Table 1. Carbides were observed particularly inside the primary phase, indicating, according to McCullough7, together with their bigger size, that they were formed directly from the melt.
In addition to these coarser carbides we also observed finer carbides. Their sizes were up to 1 pm and they were present particularly on the grain boundaries and also inside the lamellar grains. Figure 6 shows an example of fine carbides. The carbides inside lamellar dendrites were observed primarily in y-lamellas. However, their small size indicates that unlike the long acicular carbides they were formed from the solid phase and their formation was caused by a small solubility of C in the y-phase, which was only x = 0.02-0.03 %2.
3.4 Grain size
Figures 7 and 8 show the macrostructures of samples 1 and 4. It is evident already at first glance that alloy 1, which contained (p = 3.73 % of Ti2AlC carbides, has a much smaller grain size than sample 4, in which the amount fractions of Ti2AlC carbides was only p = 0.64 %. The results of determined grain sizes are presented in Table 2. It follows from the results that the grain size decreases with the increasing amount of Ti2AlC carbides. The grain-size refinement is caused by the fact that Ti2AlC carbides are the first to crystallise from the melt and that they form nucleation centres for the formation and growth of a-grains, which then transform into a lamellar structure. Smaller deviations of the grain-size values for the samples with higher amounts of carbides also indicate a more regular shape and smaller differences in the grain size. No significant influence of yttrium on the grain size was observed in the alloys alloyed with yttrium.


Figure 8: Macrostructure of sample 1 Slika 8: Makrostruktura vzorca 1
Table 2: Dependence of the amount of Ti2AlC particles on the grain
size
Tabela 2: Odvisnost vsebnosti izločkov Ti2AlC od velikosti zrna
Sample	Amount of carbides (p/%)	Grain size (pm)
1	3.73 ± 0.82	34.40 ± 8.10
2	1.28 ± 0.78	255.99 ± 59.66
3	2.59 ± 0.67	139.26 ± 46.28
4	0.64 ± 0.28	430.58 ± 99.63
3.5 Mechanical properties
Figure 9 presents the curves illustrating the relation between the strain and stress determined at the temperature of 800 °C during the compression mechanical tests for all four alloys. As it can be seen in this figure, all the samples have a similar behaviour at this temperature. The established values of the proof stress did not show
Figure 9: Strain-stress curves at 800 °C Slika 9: Krivulja raztezek - napetost pri 800 °C
Table 3: Compressive 0.2 yield strength as a function of the test temperature (MPa)
Tabela 3: Tlačna meja tecenja 0,2 v odvisnosti od temperature (MPa)
Temperature (°C)	Sample 1	Sample 2	Sample 3	Sample 4
800	566	608	547	588
600	645	695	623	917
400	684	734	-	998
Figure 10: Dependence of the amount of carbides on the compressive 0.2 yield strength
Slika 10: Odvisnost vsebnosti ogljika od tlačne meje tečenja 0,2
any significant differences, ranging from 547 MPa for alloy 3 up to 608 MPa for alloy 2. Bigger differences were, however, found between the determined values of the strength limit. The lowest strength was found for alloy 3, namely, 1158 MPa, and the highest strength was determined for alloy 2, namely, 1350 MPa. This alloy also had the highest achieved ductility, up to 23.5 %. Alloy 1 had the lowest ductility, up to 14.98 %. Table 3 presents the values of the proof limit analyzed at all the tested temperatures. It follows from the results that the lowest yield strength was achieved for alloys 1 and 3, which had higher amounts of carbon and carbides despite their much smaller grain sizes. The highest values
of the yield strength were determined for alloy 4, which had the lowest amounts of carbon and carbides.
The diagram presented in Figure 10 illustrates the relation between the proof limit at the temperatures of (400, 600 and 800) °C and the determined amounts of carbides in the alloys. It seems from this diagram that if the amount of carbides is higher than (p = 1 %, the values of the proof limit at (400 and 600) °C are significantly decreased. At the temperature of 800 °C the values of the proof limit are not decreased. From this information it can be, therefore, concluded that larger Ti2AlC carbides do not have a favourable influence on the mechanical properties of y-TiAl alloys at these temperatures. Comparatively high values of the yield strength for alloy 4 were probably caused by the strengthening of the solid solution due to carbon and the precipitation of fine carbides.
4 CONCLUSIONS
The alloys were prepared by vacuum-induction melting in a graphite crucible. The melting in this crucible resulted in a contamination with carbon mass fractions ranging from 0.295 % to 0.5 %. Coarser acicu-lar Ti2AlC precipitates were observed in the alloys as well as finer carbide precipitates. The quantity and size of the acicular precipitates increased with the determined carbon amount. The highest values of the yield strength at the temperatures of (400, 600 and 800) °C were found for the alloys with smaller amounts of acicular carbides.
From these results it is concluded that the elevated-temperature strengthening is a result of C in the solid solution and small carbide precipitates rather than of the rod-shaped coarse Ti2AlC precipitates.
Acknowledgement
This paper was created within projects No. CZ.1.05/2.1.00/01.0040 "Regional Materials Science and Technology Centre" within the frame of the operational programme "Research and Development for Innovations" financed by the Structural Funds and from the state budget of the Czech Republic and SP2013/64 "Specific research in metallurgy, materials and process engineering".
5 REFERENCES
'X. Wu, Intermetallics, 14 (2006), 1114-1122 2F. Perdrix, Intermetallics, 9 (2001), 807-815 3W. H. Tian, Intermetallics, 5 (1997), 237-244
4	J. D. Whittenberger, Intermetallics, 2 (1994), 167-178
5	V. Kevorkijan, S. D. Skapin, Mater. Tehnol., 43 (2009) 5, 239-244
6	J. Zollinger, Intermetallics, 15 (2007), 1343-1350
7	C. McCullough, Materials Science and Engineering, 124A (1990), 83-101