Fracture behaviour of cast in-situ TiAl matrix composite reinforced with carbide particles

Author’s Accepted Manuscript Fracture behaviour of cast in-situ TiAl matrix composite reinforced with carbide particles J. Lapin, M. Štamborská, T. Pelachová, O. Bajana

www.elsevier.com/locate/msea

PII: DOI: Reference:

S0921-5093(18)30294-6 https://doi.org/10.1016/j.msea.2018.02.077 MSA36164

To appear in: Materials Science & Engineering A Received date: 27 January 2018 Revised date: 19 February 2018 Accepted date: 20 February 2018 Cite this article as: J. Lapin, M. Štamborská, T. Pelachová and O. Bajana, Fracture behaviour of cast in-situ TiAl matrix composite reinforced with carbide p a r t i c l e s , Materials Science & Engineering A, https://doi.org/10.1016/j.msea.2018.02.077 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Fracture behaviour of cast in-situ TiAl matrix composite reinforced with carbide particles J. Lapin*, M. Štamborská, T. Pelachová, O. Bajana Institute of Materials & Machine Mechanics, Slovak Academy of Sciences, Dúbravská cesta 9, 845 13 Bratislava, Slovak Republic Abstract Three-point bending tests, Charpy impact tests and numerical simulations were carried out to study fracture behaviour of in-situ TiAl matrix composite reinforced with Ti2AlC particles prepared by centrifugal casting of Ti-44.5Al-8Nb-0.8Mo-0.1B-5.2C (at.%) alloy. The brittle fracture behaviour of the in-situ composite includes crack deviation, microcrack formation, carbide fragmentation, delamination on the matrix-carbide interfaces and pull-out of the carbide particles from the TiAl matrix. The crack initiation and propagation is related to applied load, deflection and acoustic emission events measured during three-point bending tests. A critical stress leading to a crack initiation in the notch region is numerically calculated for quasi-static loading conditions using finite element analysis (FEA). The measured fracture toughness values are comparable to those of some in-situ TiAl matrix composites prepared by casting and reactive processing. Keywords: intermetallics; composites; fracture toughness; finite element modelling 1. Introduction Cast TiAl-based alloys are attractive for high-temperature structural applications [1– 4]. However, inherent poor room-temperature ductility and insufficient strength at high temperatures (above 800 °C) limit their wide-scale applications. Intermetallic matrix composites may improve the deficiency of these lightweight alloys at high temperatures because of good combination of the properties of intermetallic matrix and reinforcement [5– 11]. Recent studies have shown that in-situ TiAl matrix composites reinforced with homogenously distributed carbide particles can be prepared by melting followed by gravity or centrifugal casting of TiAl based alloys with various content of carbon [9-10]. Besides the primary carbide particles, the additional strengthening of TiAl matrix composites can be achieved by fine secondary needle-like Ti3AlC (P-phase) and plate-like hexagonal Ti2AlC (H-phase) precipitates similarly to that reported for several low carbon TiAl-based alloys [1215]. The carbide particles improve fracture toughness of the in-situ TiAl matrix composites but alloying with carbon deteriorates their room-temperature tensile ductility [14-15]. The increase of the fracture resistance was attributed to crack trapping and crack bridging mechanisms [16]. In spite of the previous studies [7-11], only limited information is available about the effect of Ti2AlC particles on fracture behaviour of in-situ TiAl matrix composites prepared by centrifugal casting. The centrifugal casting represents widely applied technology

2

allowing processing near net shape industrial components from TiAl-based alloys which might facilitate potential applications of the in-situ composites [9]. The purpose of the present work is to investigate room-temperature fracture behaviour of in-situ TiAl matrix composite reinforced with carbide particles prepared by centrifugal casting of Ti-44.5Al-8Nb-0.8Mo-0.1B-5.2C (at.%) alloy. Quasi-static three-point bending and dynamic Charpy impact tests are carried out to study crack initiation and propagation using V-notch specimens. A blunt notch is used because a fatigue pre-cracking was found to be inconvenient to create an appropriate crack in the studied in-situ composite. Acoustic emission and interrupted three-point bending tests combined with microstructural observations and numerical calculations are performed to identify a critical load and stress leading to a crack initiation during quasi-static loading conditions. 2. Experimental material and procedures Samples of the in-situ TiAl matrix composite reinforced with carbide particles were prepared by a high frequency induction melting of Ti-44.5Al-8Nb-0.8Mo-0.1B-5.2C (at.%) alloy followed by a centrifugal casting into a graphite mould. In order to remove casting porosity, the as-cast samples with a diameter of 15 mm and length of 150 mm were subjected to hot isostatic pressing (HIP) at a temperature of 1250 °C and applied pressure of 200 MPa for 4 h in argon. The HIP-ed samples were consequently annealed at a temperature of 900 °C for 25 h and furnace cooled to room temperature in air. Room-temperature tensile tests were carried out on threaded-head cylindrical specimens with a gauge diameter of 6 mm and gauge length of 30 mm prepared by lathe machining. The surface of the gauge section was polished to a roughness better than 0.3 m. The tests were carried out at an initial strain rate of 1×10−4 s−1 using universal testing machine. Elongation was measured by a laser extensometer. Three-point bending V-notch specimens with a thickness of 10 mm, width of 10 mm, length of 55 mm, notch length of 2 mm and notch tip radius of 0.25 mm were prepared by wire electrical discharge machining and grinding. Due to the brittle nature of the in-situ composite

Fig. 1. Schema of three-point bending test with two acoustic sensors attached to the specimen: 1, 2 – AE sensors.

Fig. 2. Initial mesh of V-notch threepoint bending specimen created by FEA.

3

no fatigue pre-crack was made. The quasi-static three-point bending tests were carried out at a cross head speed of 8.33x10-6 m/s at room temperature. The acquisition of load-deflection data was done electronically during the tests. Two acoustic emission (AE) sensors were attached to the test specimen, as seen in Fig. 1. The AE Vallen System sensors with a peak resonance frequency of 600 kHz were used. Acoustic emission events detected during testing were amplified by a pre-amplifier and analysed by the Vallen AE-Suite Software. The threshold was set to 50 dB which meant that events of amplitude smaller than 50 dB were not recorded. Standard Charpy V-notch specimens with a thickness of 10 mm, width of 10 mm, length of 55 mm and notch length of 2 mm were prepared by wire electrical discharge machining and grinding. The dynamic Charpy impact tests were conducted at room temperature with an impact speed of 5.23 m/s. Load-deflection curves were recorded by the instrumented Charpy impact system attached into the impact tester. Dynamic maximum and fracture loads were determined from the load-deflection curves. Microstructural evaluation was carried out by scanning electron microscopy (SEM) and SEM in back scattered electron (BSE) mode. Metallographic preparation of the samples consisted of standard grinding on abrasive papers and polishing on diamond pastes with various grain sizes up to 0.25 μm. SEM samples were etched in a reagent of 150 ml H2O, 25 ml HNO3 and 10 ml HF. TEM samples were thinned mechanically to a thickness of about 40 µm and subsequently electro-polished on a Struers TenuPol-5 twinjet electro-polisher at 20 V and 233 K. The electropolishing solution consisted of 65% ethanol, 30% butanol and 5% perchloric acid. Volume fraction of phases and size of carbide particles were measured by computerised image analysis using digitalised micrographs and measured data were treated statistically. Software Ansys Workbench was used for numerical simulation. A geometry model of the three-point bending specimen was built as a 3D cuboid with a notch and the supporting and loading pins were modelled as 3D cylinders. Material model was built-up using the achieved experimental data from the three-point bending tests. Deformable hexahedral elements with a size of 0.5 and 0.1 mm were used to mesh the notch specimen and cylinders, respectively, as seen in Fig. 2. In total, the mesh created by finite element analysis (FEA) was composed of 485 117 hexahedral elements, adding up to 2 007 867 nodes in total. 3. Results and discussion 3.1. Microstructural characterisation Fig. 3 shows the typical microstructures of the studied in-situ composite prepared by the centrifugal casting of Ti-44.5Al-8Nb-0.8Mo-0.1B-5.2C (at.%) alloy after HIP at 1250 °C/200 MPa followed by annealing at 900 °C for 25 h. It is clear that the applied HIP and annealing were not sufficient long to homogenise fully the in-situ composite. The microstructure consists of the matrix composed of dendritic (TiAl) enriched in Nb and Ti (bright colour phase) and interdendritic i(TiAl) enriched in Al (black colour phase).

4

Fig. 3. Micrographs showing the typical microstructure of the in-situ composite before mechanical testing: (a) uniform distribution of carbide particles in the matrix composed of dendritic (TiAl) and interdenritic i(TiAl), BSE; (b) morphology of primary Ti2AlC in the (TiAl) matrix, BSE, (c) fine secondary Ti2AlC particles formed within the (TiAl) grains and along the (TiAl) grain boundaries, TEM. 1 – primary plate-like Ti2AlC, 2 – primary regular shaped Ti2AlC, S - secondary Ti2AlC precipitates. The centrifugal casting leads to a relatively uniform distribution of coarse carbide particles within the (TiAl) matrix, as seen in Figs. 3a and 3b. The coarse carbide particles belong to Ti2AlC phase and are formed during solidification according to phase transformation pathway L + TiC1-x  L + Ti2AlC [9]. This solidification pathway affects the morphology of the carbide particles. The Ti2AlC particles, which were formed directly by transformation of TiC1-x particles present in the melt, are frequently nearly spherical and occasionally irregular shaped ones. The particles which are formed directly from the melt in the L + Ti2AlC phase region are plate-like, as shown in Fig. 3b. The volume fraction of the carbide particles is measured to be (21.5 ± 0.7) vol.%. This volume fraction Vp is in a very good agreement with that of 21.9 vol.% calculated according to relationship determined by Lapin et al. [9] for Ti-44.5Al-8Nb-0.8Mo-0.1B-xC (at.%) alloys (x is ranging from 1.4 to 4.8 at.%) in the form:

5

Vp  4.22aC

(1)

where aC is the content of carbon in at.%. The mean length and diameter of plate-like and regular shaped Ti2AlC particles are measured to be (17.3 ± 0.7) m and (9.8 ± 0.4) m, respectively. Fig. 3c clearly indicates that the intermetallic (TiAl) matrix is reinforced by bimodal sized carbide particles. Besides the coarse primary carbides, the in-situ composite contains fine secondary Ti2AlC particles which are formed during the heat treatments. These fine carbides with an average width of (29.5 ± 0.7) nm and length of (84.7 ± 4.3) nm precipitate preferentially within the coarse  grains (grain size of 28 µm) and along the  grain boundaries during the dissolution of α2(Ti3Al) phase in the form of continuous lamellae and irregular particles. In, addition the HIP leads to the formation of numerous small recrystallized  grain in the vicinity of the coarse primary carbide particles. 3.2. Tensile tests The studied in-situ composite shows brittle behaviour without yielding during the room-temperature tensile tests. The stress increases nearly linearly with the strain up-to a fracture. Average ultimate tensile strength (5 tested specimens) is measured to be (330 ± 15) MPa. Leguillon et al. [17,18] have reported that the tensile strength of brittle materials is higher than flexural one because fewer randomly distributed defects in the material are involved in bending than in tension in case of specimens of the same size. However, such effect is minimize in the present work by selecting the size of tensile specimens which is close to the size of the tested three-point bending specimens stressed in tension. 3.3. Three-point bending tests Fig. 4 shows the typical example of load-deflection curve of three-point bending test with the standard V-notch specimen. It is obvious that a large number of AE events are produced by both AE sensors during the quasi-static loading. The frequency of these events increases with increasing load and deflection. Most of the AE events of amplitudes ranging from 70 to 90 dB occur at a deflection higher than 0.031 mm. The crack propagation is connected with the events of amplitudes up to 100 dB. The fracture of the specimen is observed at a maximum load of 1580 N and deflection of 0.036 mm. 3.3.1. Static fracture toughness The fracture toughness of notched bending specimen can be calculated according to relationship [19]

K S 1.5

FLa 0.5 1.99  a / w(1  a / w)(2.15  3.93a / w  2.7(a / w) 2 )    bw2  (1  2a / w)(1  2a / w)1.5 

(2)

where F is the fracture load, L is the span of the loading, b is the thickness of specimen, w is

6

Fig. 4. Three-point bending load-deflection curve and corresponding AE events detected with AE sensors 1 and 2. the width of specimen and a is the length of the notch. Since the ASTM specifications for fracture toughness testing concerning notch geometry and pre-cracking were not rigorously followed in the present study, the toughness values are denoted as KS. The measured value of fracture toughness for the studied in-situ composite of KS = (11.6 ± 0.6) MPam0.5 can be compared with those of 13.7 MPam0.5 for Ti-44.5Al-8Nb-0.8Mo-0.1B-4.8C in-situ composite prepared by centrifugal casting [9], 7.8 MPam0.5 for TiAl/Ti2AlC in-situ composite prepared by reactive hot pressing [20], 7.3 MPam0.5 for TiAl/Ti3A1C2-Ti2AlC synthesised by hotpressing sintering [21] and 12–17.8 MPam0.5 measured for in-situ TiAl-Ti2AlC composites produced by reactive processing [22]. The incorporation of TiAl2C particles increases the fracture toughness of the in-situ composite compared those of 8.0 MPam0.5 and 7.0 MPam0.5 reported by Rao et al. [16] and Wang et al. [23] for TiAl and Ti2AlC, respectively. The studied in–situ composite suffer from a lack of intrinsic toughening mechanisms due to absence of mobile dislocations which can form a plastic zone ahead of crack tip. Hence, the improvement of the fracture toughness can be related to extrinsic toughening mechanisms [24] namely to crack deviation, zone shielding by microcrack toughening and bridging by coarse lathe-shaped carbide particles. However, the effectiveness of these extrinsic toughening mechanisms depends strongly on the applied processing techniques, volume fraction and size of the carbide particles [9, 20-22]. As shown by Yang et al. [20] and Ai et al. [21], the incorporation of the Ti2AlC particles into TiAl matrix has not improved fracture toughness of the in-situ composites prepared by hot-pressing sintering compared to that of TiAl or Ti2AlC [16,23].

7

Fig. 5. SEM micrographs showing crack initiation and propagation during three-point bending test: (a) crack intiation in the notch tip region; (b) propagation of the main and arrest of the secondary cracks by Ti2AlC particles.

Fig. 6. The typical brittle fracture of the in-situ composite after three-point bending test: (a) smooth fracture surface, SEM; (b) fracture surface in the vicinity of the notch tip, SEM; (c) step-like fracture surface of the (TiAl) matrix with embeded fine carbide particles (S), BSE; (d) brittle fracture of primary plate-like (1) and regular shaped (2) Ti2AlC particles, BSE.

8

3.3.2. Fractography Fig. 5 shows the tip region of the V-notch specimen after interrupted three-point bending test. The initiation of short cracks in the vicinity of the notch tip is observed at a load of 1418 N which corresponds to a deflection of 0.031 mm and AE events with an amplitude of about 90 dB, as seen in Figs. 4 and 5a. These cracks initiated preferentially in the (TiAl) matrix are arrested by the carbide particles. Further increase of the load to 1580 N results in a fast propagation of the main crack through the specimen in the zig-zag manner. The main crack is accompanied with several short cracks which are usually initiated along the grain boundaries, as seen in Fig. 5b. It is clear from Fig. 5 that the primary Ti2AlC particles serve as the effective obstacles to the crack propagation, deviate the main crack and arrest the short secondary cracks. Fig. 6 shows brittle fracture mode of the in-situ composite after three-point bending test. The fracture surface is macroscopically smooth, as seen in Fig. 6a. The crack propagates along the crystallographic planes of some  grains, grain boundaries and interfaces between the  matrix and Ti2AlC particles, as seen in Figs. 6b. The step-like propagation of the crack and tearing is observed in the (TiAl) matrix with lamellar / type of microstructure, as shown in Fig. 6c. The / lamellae are pinned by fine incoherent Ti2AlC precipitates formed during the HIP at 1250 °C followed by the annealing at 900 °C [9,12,13], as seen in Fig. 6c. The fracture surface indicates pull-out of the primary carbides from the matrix. The coarse primary Ti2AlC with layered structure are fractured along their crystallographic planes. Fig. 6d clearly indicates the cleavage fracture surface of the regular shaped and delamination along the crystallographic planes leading to a step-like fracture of the plate-like particles. The laminated tearing of the primary carbides and the kink boundaries indicate that these particles contribute to the damage tolerance of the in-situ composite when compared with a single TiAl matrix. The embedded carbides in the matrix retard the crack propagation and increase the fracture energy. As shown by Rao et al. [16] for ductile TiNb particles in TiAl matrix and Ramaseshan et al. [22] for Ti2AlC particles in lamellar Ti3Al+TiAl matrix, the increase of the fracture toughness of the composites results from crack trapping and its re-nucleation. The fracture toughness can be increased by increasing strength of the reinforcement. In the case of the studied in-situ composite, the phase composition of the primary carbide particles can be influenced by the applied processing routes. Significantly harder TiC particles can be preserved within the coarse irregular shaped Ti2AlC ones during casting and appropriate heat treatments which would improve the fracture toughness of the in-situ composite [9]. 3.3.3. Numerical simulations During the quasi-static three-point bending tests the notch creates a stress concentrator and increases local stresses in the specimen. The condition for the crack initiation are fulfilled when a local stress in the vicinity of the notch tip achieve a critical value corresponding to a flexural strength of the material [18]. The analysis of fracture behaviour shows that an average critical load of F = (1418 ± 37) N is required to initiate a crack in the notch tip

9

region of the studied in-situ composite during the interrupted three-point bending tests. This experimentally measured critical load is used to calculate stress distribution within the notched specimen and to determine a critical stress required for the crack initiation. Fig. 7 shows the distribution of numerically calculated equivalent (von-Mises) stresses within the notched specimens corresponding to a load of 1418 N. A maximum local stress of 347 MPa is calculated by FEA in the tip notch region shown in detail (Fig. 7). This calculated maximum local stress is close to the measured average ultimate tensile strength of (330 ± 15) MPa of the studied in-situ composite.

Fig. 7. Equivalent (von-Mises) stress distribution calculated at a critical load corresponding to a crack initiation during three-point bending test. 3.4. Charpy impact tests Fig. 8 shows the typical examples of dynamic Charpy impact load-time curves of the studied in-situ composite. The measured Charpy impact value of EId = (6.1 ± 0.3) kJ/m2 is lower than those ranging from 6.4 to 7.6 kJ/m2 reported for Ti-44.5Al-8Nb-0.8Mo-0.1B-xC (x is ranging from 1.4 to 4.8 at.%) in-situ composites [9] or 12.5 kJ/m2 determined by Tetsui et al. [25] for Ti-42Al-10V (at.%) alloy. However, it is higher than that of 2.0 kJ/m2 measured by Matsugi et al. [26] for spark sintered binary TiAl-based alloy with duplex microstructure. 3.4.1. Dynamic fracture toughness During impact test a crack initiation generally occurs in brittle materials at a maximum load. As reported by Sahraoui and Lataillade [27], the Charpy impact tester generates oscillations, which can disturb significantly an effective load at the fracture initiation. In the case of the brittle materials such as TiAl matrix in-situ composites, dynamic fracture toughness can be calculated by applying linear fracture mechanics. The dynamic fracture toughness KId can be calculated according to relationship [28] 2 3 4  FL a a a a  0.5 K Id 1.5 2 a  1.11  1.55  7.71   13.53   14.23   w bw  w  w  w   

(3)

10

Fig. 8. The examples of load–time curves of the studied in-situ composite obtained from instrumented Charpy impact tests. Eq. (3) can be applied for a quasi-static loading conditions or in case the fracture occurs after time sufficiently long to achieve a quasi-static loading conditions in the specimen. Sahraoui et al. [27,28] have proposed that the maximum load determined from the load-time curve can be related to the fracture deflection and time to fracture. Hence, the fracture load F can be calculated from the fracture deflection u and the stiffness of each specimen ko according to relationship

F  kou

(4)

For the studied in-situ composite, fracture loads are determined from the measured load-timedeflection curves and stiffness of the specimens according to Eq. (4). The dynamic fracture toughness value of KId = (12.2 ± 0.7) MPam0.5 calculated according to Eq. (3) is comparable to the static one of 11.6 MPam0.5 measured by three-point bending test. Kim et al. [29] have shown that the fracture toughness decreases as the notch root radius decreases until a critical radius. The fracture toughness is independent of notch radius below the critical value and normalised static fracture toughness corresponds to the dynamic one. The notch root radius of 0.25 mm applied in the present work is well below a critical value of 0.83 mm calculated for brittle casting steel by Kim et al. [29]. The values of EId and KId measured for Ti-44.5Al-8Nb-0.8Mo-0.1B-5.2C (at.%) insitu composite are lower than those reported for the centrifugally cast in-situ Ti-44.5Al-8Nb0.8Mo-0.1B-(1.4-4.8)C (at.%) composites [9]. Generally, the decrease of the EId and KId with increasing carbon content in Ti-44.5Al-8Nb-0.8Mo-0.1B-(1.4-5.2)C (at.%) in-situ composites can be explained by microstructural changes of the matrix which changes from γ(TiAl)+α2(Ti3Al)+β/B2 in low carbon to γ(TiAl) type in high carbon systems [9].

11

3.4.2. Fractography Fig. 9 shows the typical brittle fracture of the in-situ composite after the dynamic Charpy impact test. The fracture surface is macroscopically smooth, as seen in Fig. 9a. The transgranular fracture with radial cracks is a predominant mode of failure demonstrating that the composite exhibits low ductility at room temperature, as shown in Fig. 9b. During the specimen impact numerous short cracks are initiated along the grain boundaries and carbide-matrix interfaces in the vicinity of the notch tip. Once a crack is initiated with the direction coinciding with the crack propagation direction, it can easily propagate through the whole grains and carbide particles. The fracture surface dominates by cleavage fracture connected with numerous delamination along the crystallographic planes of the regular shaped and plate-like carbide particles and pull out of the primary carbides from the

Fig. 9. The typical brittle fracture of the in-situ composite after Charpy impact test: (a) smooth fracture surface, SEM; (b) fracture surface in the vicinity of the notch tip with radial cracks, SEM; (c) fracture mode of the regular shaped (1) and plate-like (2) Ti2AlC particles, BSE; (d) step-like fracture mode of the (TiAl) matrix with embeded secondary carbides (S), BSE.

12

matrix, as shown in Fig. 9c. In addition, numerous short radial cracks along the grain boundaries and matrix-carbide interfaces are formed during the propagation of the main crack. The measured maximum load of 12860 N leading to the fracture of V-notch samples at an impact speed of 5.23 m/s is significantly higher that of 1580 N measured during the three point bending carried out at a cross head speed of 8.33x10-6 m/s. In spite of some common features observed on the fracture surfaces of the studied in-situ composite after quasi-static three-point bending and dynamic Charpy impact tests (Figs. 6 and 9), the static and dynamic loading conditions result in some particular differences. The dynamic loading leads to the formation of numerous short radial short cracks within the (TiAl) grains and along the (TiAl) grain boundaries, as seen in Figs. 9b and 9c. This short crack formation contributes to the crack zone shielding by microcrack toughening [24]. The high impact speed loading leads also to a more intensive fragmentation of the primary carbide particles and their delamination along crystallographic planes, as seen in Fig. 9c. This intensive fragmentation of the coarse reinforcing carbides can be related to a decrease of their strength with increasing strain rate similarly to that reported by Salama et al. [30] for (Ti,Nb)2AlC. Fig. 3c clearly indicates that the fine plate-like secondary Ti2AlC particles formed during a gradual transformation of α2 lamellae [31] are well distributed along the lamellar / interfaces. Fig. 9d indicates that several parallel short cracks propagates along the / lamellar interfaces pinned by these fine precipitates forming step-like fracture by shear-ligament bridging within some coarse (TiAl) grains oriented favourably to the principal crack propagation. The parallel cracks are formed ahead of the principal crack and are preserved on the fracture surfaces. On the other hand, the step-like fracture surfaces show no evidence of such remaining parallel cracks formed along the lamellar / interfaces after the fracture at quasi-static loading conditions, as seen in Fig. 6c. 4. Conclusions The static and dynamic fracture behaviour of the in-situ TiAl matrix composite reinforced with carbide particles was studied. The achieved results can be summarised as follows: 1. The microstructure of the in-situ composite prepared by centrifugal casting followed by HIP and annealing consists of the TiAl matrix reinforced with bimodal sized Ti2AlC particles. 2. A critical load and deflection for a crack initiation in the notch tip region of test specimens is determined by the acoustic emission measurements combined with the metallographic analysis during interrupted quasi-static three-point bending tests. 3. The stress distribution in the notched specimen and critical stress leading to a crack initiation in the notch tip region are numerically calculated using finite element analysis

13

4.

5.

software ANSYS. The calculated critical stress is close to the measured ultimate tensile strength of the studied in-situ composite. The fracture toughness values measured on V-notch specimens using quasi-static threepoint bending and dynamic instrumented Charpy impact tests are comparable to those of some in-situ TiAl matrix composites prepared by casting and reactive processing. The brittle fracture behaviour of the in-situ composite includes crack deviation, zone shielding by microcrack toughening, carbide fragmentation, delamination on the matrix-carbide interfaces and pull-out of the carbide particles from the TiAl matrix. Acknowledgements

This work was financially supported by the Slovak Research and Development Agency under the contract APVV-15-0660 and the Slovak Grant Agency for Science under the contract VEGA 2/0125/16. The experimental work was carried out thanks to the infrastructure supported by the Competence Centre for New Materials, Advanced Technologies and Energy under the contract ITMS 26240220073. References [1] [2]

[3]

[4] [5]

[6]

[7]

[8]

[9]

B.P. Bewlay, S. Nag, A. Suzuki, M.J. Weimer, TiAl alloys in commercial aircraft engines, Mater. High Temp. 33 (2016) 549–559. doi:10.1080/09603409.2016.1183068 W. Shouren, G. Peiquan, Y. Liying, Centrifugal precision cast TiAl turbocharger wheel using ceramic mold, J. Mater. Process. Technol. 204 (2008) 492–497. doi:10.1016/j.jmatprotec.2008.01.062. J. Lapin, M. Nazmy, Microstructure and creep properties of a cast intermetallic Ti46Al-2W-0.5Si alloy for gas turbine applications, Mater. Sci. Eng. A 380 (2004) 298– 307. doi:10.1016/j.msea.2004.05.011. J. Lapin, T. Pelachová, M. Dománková, Long-term creep behaviour of cast TiAl-Ta alloy, Intermetallics 95 (2018) 24–32. doi:10.1016/j.intermet.2018.01.013. J. Lapin, L. Ondrúš, O. Bajana, Effect of Al2O3 particles on mechanical properties of directionally solidified intermetallic Ti-46Al-2W-0.5Si alloy, Mater. Sci. Eng. A 360 (2003). doi:10.1016/S0921-5093(03)00445-3. H.Z. Niu, S.L. Xiao, F.T. Kong, C.J. Zhang, Y.Y. Chen, Microstructure characterization and mechanical properties of TiB2/TiAl in situ composite by induction skull melting process, Mater. Sci. Eng. A 532 (2011) 522–527. doi:10.1016/j.msea.2011.11.017. X.J. Song, H.Z. Cui, N. Hou, N. Wei, Y. Han, J. Tian, Q. Song, Lamellar structure and effect of Ti2AlC on properties of prepared in-situ TiAl matrix composites, Ceram. Int. 42 (2016) 13586–13592. doi:10.1016/j.ceramint.2016.05.152. X. Song, H. Cui, Y. Han, N. Hou, N. Wei, L. Ding, Q. Song, Effect of carbon reactant on microstructures and mechanical properties of TiAl/Ti2AlC composites, Mater. Sci. Eng. A 684 (2017) 406–412. doi:10.1016/j.msea.2016.12.069. J. Lapin, A. Klimová, Z. Gabalcová, T. Pelachová, O. Bajana, M. Š, Microstructure and mechanical properties of cast in-situ TiAl matrix composites reinforced with (Ti,Nb)2AlC particles, Mater. Design 133 (2017) 404–415. doi:10.1016/j.matdes.2017.08.012.

14

[10] A. Klimová, J. Lapin, T. Pelachová, Characterization of TiAl based alloys with various content of carbon, in: IOP Conf. Ser. Mater. Sci. Eng., 2017. doi:10.1088/1757899X/179/1/012038. [11] J. Ramkumar, S.K. Malhotra, R. Krishnamurthy, H. Mabuchi, K. Demizu, A. Kakitsuji, H. Tsuda, T. Matsui, K. Morii, Microstructure and dry sliding wear of Ti–50Al alloy and Ti–47Al–3W/Ti2AlC composite produced by reactive arc-melting, Mater. Trans. 44 (2003) 1861–1865. [12] E. Schwaighofer, P. Staron, B. Rashkova, A. Stark, N. Schell, H. Clemens, S. Mayer, In situ small-angle X-ray scattering study of the perovskite-type carbide precipitation behavior in a carbon-containing intermetallic TiAl alloy using synchrotron radiation, Acta Mater. 77 (2014) 360–369. doi:10.1016/j.actamat.2014.06.017. [13] T. Cegan, I. Szurman, Thermal stability and precipitation strengthening of fully lamellar Ti-45Al-5Nb-0.2B-0.75C alloy, Kovove Mater. 55 (2017) 421–430. doi:10.4149/km_2015_2_69. [14] Z. Wu, R. Hu, T. Zhang, F. Zhang, H. Kou, J. Li, Understanding the role of carbon atoms on microstructure and phase transformation of high Nb containing TiAl alloys, Mater. Charact. 124 (2017) 1–7. doi:10.1016/j.matchar.2016.12.008. [15] M. Li, S. Xiao, L. Xiao, L. Xu, J. Tian, Y. Chen, Effects of carbon and boron addition on microstructure and mechanical properties of TiAl alloys, J. Alloys Compd. 728 (2017) 206–221. doi:10.1016/j.jallcom.2017.08.211. [16] K.T. Venkateswara Rao, G.R. Odette, R.O. Ritchie, Ductile-reinforcement toughening in -TiAl intermetallic-matrix composites: Effects on fracture toughness and fatiguecrack propagation resistance, Acta Metall. Mater. 42 (1994) 893–911. doi:10.1016/0956-7151(94)90285-2. [17] D. Leguillon, É. Martin, M.C. Lafarie-Frenot, Flexural vs. tensile strength in brittle materials, Comptes Rendus - Mec. 343 (2015) 275–281. doi:10.1016/j.crme.2015.02.003. [18] D. Leguillon, Strength or toughness? A criterion for crack onset at a notch, Eur. J. Mech. A/Solids. 21 (2002) 61–72. doi:10.1016/S0997-7538(01)01184-6. [19] R. Cao, M.X. Lei, J.H. Chen, J. Zhang, Effects of loading rate on damage and fracture behavior of TiAl alloys, Mater. Sci. Eng. A 465 (2007) 183–193. doi:10.1016/j.msea.2007.02.026. [20] C. Yang, F. Wang, T. Ai, J. Zhu, Microstructure and mechanical properties of in situ TiAl/Ti2AlC composites prepared by reactive hot pressing, J. Eur. Ceram. Soc. 40 (2014) 8165–8171. doi:10.1016/j.jeurceramsoc.2014.07.021. [21] T.T. Ai, F. Wang, X.M. Feng, M.M. Ruan, Microstructural and mechanical properties of dual Ti3A1C2-Ti2AlC reinforced TiAl composites fabricated by reaction hot pressing, Ceram. Int. 40 (2014) 9947–9953. doi:10.1016/j.ceramint.2014.02.092. [22] R. Ramaseshan, A. Kakitsuji, S.K. Seshadri, N.G. Nair, H. Mabuchi, H. Tsuda, T. Matsui, K. Morii, Microstructure and some properties of TiAl-Ti2AlC composites produced by reactive processing, Intermetallics 7 (1999) 571–577. doi:10.1016/S09669795(98)00069-7. [23] P. Wang, B. Mei, X. Hong, W. Zhou, Synthesis of Ti2AlC by hot pressing and its mechanical and electrical properties, Trans. Nonferrous Met. Soc. China 17 (2007) 1001–1004. doi:10.1016/S1003-6326(07)60215-5. [24] M.E. Launey, R.O. Ritchie, On the fracture toughness of advanced materials, Adv. Mater. 21 (2009) 2103–2110. doi:10.1002/adma.200803322. [25] T. Tetsui, K. Shindo, S. Kobayashi, M. Takeyama, Strengthening a high-strength TiAl alloy by hot-forging, Intermetallics 11 (2003) 299–306. doi:10.1016/S09669795(02)00245-5.

15

[26] K. Matsugi, T. Hatayama, O. Yanagisawa, Impact properties of spark sintered titanium aluminides at elevated temperatures, Intermetallics 7 (1999) 1049–1057. doi:10.1016/S0966-9795(99)00014-X. [27] S. Sahraoui, J.L. Lataillade, Analysis of load oscillations in instrumented impact testing, Eng. Fract. Mech. 60 (1998) 437–446. doi:10.1016/S0013-7944(98)00024-1. [28] S. Sahraoui, A. El Mahi, B. Castagnède, S. Saheoui, A. El Mahi, B. Castagnède, S. Sahraoui, Measurement of the dynamic fracture toughness with notchede PMMA specimen under impact loading, Polym. Test. 28 (2009) 780–783. doi:10.1016/j.polymertesting.2009.06.005. [29] J.H. Kim, D.H. Kim, S. Il Moon, Evaluation of static and dynamic fracture toughness using apparent fracture toughness of notched specimen, Mater. Sci. Eng. A 387–389 (2004) 381–384. doi:10.1016/j.msea.2004.01.134. [30] I. Salama, T. El-Reghy, M.W. Barsoum, Synthesis and mechanical properties of Nb2AlC and (Ti,Nb)2AlC, J. Alloys Compd. 347 (2002) 271–278. doi:10.1016/S09258388(02)00756-9. [31] L. Wang, U. Lorenz, M. Münch, A. Stark, F. Pyczak, Influence of alloy composition and thermal history on carbide precipitation in γ-based TiAl alloys, Intermetallics 89 (2017) 32–39. doi:10.1016/j.intermet.2017.05.006.