metal-matrix composites

Accepted Manuscript High temperature oxide-fibre/metal-matrix composites S.T. Mileiko PII:

S0254-0584(17)30526-6

DOI:

10.1016/j.matchemphys.2017.07.017

Reference:

MAC 19824

To appear in:

Materials Chemistry and Physics

Received Date: 2 April 2017 Revised Date:

28 June 2017

Accepted Date: 4 July 2017

Please cite this article as: S.T. Mileiko, High temperature oxide-fibre/metal-matrix composites, Materials Chemistry and Physics (2017), doi: 10.1016/j.matchemphys.2017.07.017. 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 proof before it is published in its final 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.

ACCEPTED MANUSCRIPT HIGH TEMPERATURE OXIDE-FIBRE/METAL-MATRIX COMPOSITES S.T. Mileiko Institute of Solid State Physics of RAS, Chernogolovka Moscow distr., 142432 Russia

M AN U

SC

RI PT

Abstract Three families of metal matrix fibrous composites are considered in the paper as potential candidates for future heat resistant materials. Strength, fracture toughness, creep resistance and oxidation resistance of the composites are described and analyzed. Composites with nickel-based matrix are characterised by quasi-plastic behaviour and acceptable oxidation resistance. However, they cannot be used at temperatures above 1200oC, which is higher than that for nickel superalloys but lower than the use temperatures for really prospective thermal machines. Discovering possibilities to reduce essentially oxidation rate of molybdenum reinforced with oxide fibres of special chemical compositions makes molybdenum matrix composites being prospective heat resistant materials with high creep resistance at high temperatures and sufficiently high fracture toughness at low temperatures. A large choice of entropy alloys (HEAs) with a variety of the properties as a matrix and availability of large number of oxide fibres produced by internal crystallisation method make oxide-fibre/HEA-matrix composites highly prospective heat resistant materials. Keywords: metal matrix composites, oxide fibres, strength, fracture toughness, oxidation resistance

1. Introduction

AC C

EP

TE D

The development of a future generation of gas turbines calls for the development of heat resistant materials with the use temperature much higher than that of nickel-based superalloys now is use and those under the improvement. The mainstream of the development mentioned is now molybdenum alloys containing silicide particles [1]. As for high strength metal alloys, an enhancement of their strength and creep resistance yields a decrease in fracture toughens. Hence, fibrous composites, which can be characterized by an increase in both strength and fracture toughness simultaneously [2,3], look as the most prospective materials for high temperature heavily loaded structural elements. In the present papers, the microstructure and mechanical properties of three families of heat resistant fibrous composites are reviewed. The review mainly based on the paper presented at the international conference on high-entropy materials (ICHEM 2016) is completed with experimental results obtained recently. The first family contains oxide-fibre/nickel-based matrix composites. These composites cannot be used at temperatures higher than about 1200oC; hence, composites with matrices of higher melting points are wanted. The second family is that of molybdenum matrix composites. This family will have a practical importance provided necessary coatings are have been developed since mechanical properties of the composites such as strength, damage tolerance and creep resistance are sufficiently high at room temperature and high temperatures up to at least 1300oC. The third family of high temperature metal matrix composites appeared on the technical horizon due to the invention of high entropy alloys (HEAs) by Jeh [4]. Indeed, a unique 1

ACCEPTED MANUSCRIPT

combination of physical, chemical and mechanical properties of HEAs make them nearly ideal materials matrices for fibrous composites. It should be noted that for heat resistant composites three characteristics are of a primary importance, those being creep resistance, fracture toughness and oxidation resistance. So Therefore, it what follows the accent will be done on these characteristics.

2. Internal crystallisation method (ICM)

RI PT

Composites of the first and the third families are reinforced with oxide fibres produced by the internal crystallisation method (ICM) [2,5], composites of the second family are obtained directly by ICM. Hence, we start with a brief description of ICM and then will proceed with a discussion of the composite microstructure and properties.

M AN U

SC

A schematic of the method is presented in Fig. 1. A molybdenum carcass with continuous channels in it, which is prepared by diffusion bonding of an assemblage of the wire and foil (step 1 in Fig. 1) is infiltrated with an oxide melt (steps 2 and 3) by the capillary force. The melt is then crystallised in the channels to form fibres in an oxide/molybdenum block (step 4). This is the main scheme of the ICM desribed in details in a number of the publications (see, i.e., [2,6]). A large part of the experiments to be discussed below were conducted by using a sheme with the crucible located at the top of the molybdenum carcass.

AC C

EP

TE D

Dissolving molybdenum in a mixture of acids yields a bundle of the fibres. The fibre has a cross-sectional shape close to that shown shematically in Fig. 1, step 1; a characteristic size of it is usually 50 to 100 µm.

Fig. 1. The main scheme of the internal crystallisation method. 3. Oxide-fibre/nickel-based matrix composites These composites have been studied thoroughly in a work described in Refs. [7,8,9]. The composites have two important advantages, those being quasi-plastic behaviour at room temperature and high oxidation resistance. However, these composites have two 2

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

drawbacks. The first drawback is a problem with organizing strong fibre/matrix interface. The problem can be, perhaps, solved by choosing appropriate chemical compositions of the fibre and matrix. Creep-resistance/fibre-volume-fraction dependence for such composites is nearly linear, with a slope depending on the effective fibre strength, up to a maximum value, σ*. Fibre volume fraction Vf corresponding to ∗ ∗ σ* depends on interface strength
 , Hence, increasing the value of
 , yields an increase in Vf , that corresponds to an increase in σ*. Results of studying creep resistance of composites with various combination of the fibres and nickel-based matrices (Fig. 2) confirm this statement. The second drawback is of an inherent nature caused by a rather low melting point of the matrix. Still, oxide-fibre /nickel-matrix composites will find their proper place in the family of heat resistant materials of the future.

TE D

Fig. 2. Creep resistance (stress to yield 1% of creep strain at 100 h) of oxide-fibre/Nibased-alloy matrix composites versus fibre volume fraction at 1150oC. The matrix materials are nickel alloys M1 and M2 standing for VKNA-4 and VKNA-25 (Russian trade mark), respetively. The fibre materials are sapphire and oxide eutectics Al2O3Al5Y3O12 (AY), Al2O3-Er5Y3O12 (AEr), Al2O3-Al5Y3O12-ZrO2 (AYZ), LaAl11O18AlLaO3 (ALa).

EP

4. Molybdenum matrix composites

AC C

Molybdenum matrix composites obtained directly by ICM were considered at first as either model material [10] or a source to obtain of oxide fibres [6]. However, recently a possibility to decrease essentially oxidation rate of molybdenum by choosing as reinforcing fibres those containing oxides, which can react with molybdenum oxide to form a molybdate that has saturated vapor pressure much lower than that of molybdenum oxide, was discovered [11,12,13], Fig. 3. This finding makes molybdenum matrix composites be a candidate for future heat resistant materials. To convert the candidate into a real material there should be studied a large number of mechanical, physical, chemical, and technological characteristics.

3

RI PT

ACCEPTED MANUSCRIPT

SC

Fig. 3. А relative decrease in the mass of molybdenum and oxide/molybdenum specimens with a surface of 20 cm2 versus time of the exposure in air at 1000 and 1300оС.

M AN U

Strength (creep strength) and fracture toughness are the most important mechanical characteristics of structural materials for high temperature usage. Hence, in the framework of the present paper written on the basic of the author`s presentation on the international conference on high-entropy materials (ICHEM 2016) corresponding experimental data for some Mo-matrix composites will be presented. The data presented at ICHEM 2016 are now complemented with experimental data obtained in the last half a year.

EP

TE D

Three remarks are to be done before presenting the experimental results. First, a choice of yttrium-containing fibres to reinforce molybdenum is justified by the oxidation data of the composites given above (Fig. 3). Secondly, mullite-containing fibres look as an attractive reinforcement because (i) mullite-zirconia ceramics crystallised from the melt are known as a strong material up to a temperature of 1400oC [14] and (ii) the presence of silicon promises a reduction in the oxidation rate of molybdenum matrix. Thirdly, since linear fracture mechanics can hardly be used to describe quantitavely fracture behaviour of fibrous composites, it is useful to measure not only value of critical stress intensity factor but also notch sensitivity given by ratio ∗ ⁄∗ where ∗ and ∗ are strength values of specimens with a notch and without notch, respectively [3].

AC C

4.1 YAG-fibre/Mo-matrix The microstructure and mechanical properties of single crystalline YAG fibre grown from the melt in the channels of a molybdenum carcass were studied in Ref. [15]. The crystallisation process in that work was performed with a seed. The seed was oriented in such a manner as to provide the <111> crystallographic orientation to the fibre axis, which is characterised by the largest creep resistance [16]. The seed was not used in the present experiments and the crystallographic orientation of the fibres were grown in the direction different from was not the optimal . Pulling up rates in the fibre crystallisation process were 50 and 250 mm/min. These values are close to the crystallisation rates. Strength of composite specimens obtained at the pulling up rates mentioned at 20 and 1400oC is presented in Fig. 4. One can see that high temperature strength of the composite is decreasing with crystallisation rate increasing; however, room temperature strength is slightly increasing. Reasons for such behaviour are not clear at the present time. Perhaps, a large crystallisation rate leads to the formation of a high density of the defects in the fibre volume, which yields a decrease in the fibre strength. A limited degree of plasticity at high temperatures can make the defects not so dangerous. 4

RI PT

ACCEPTED MANUSCRIPT

Fig. 4. Flexure strength of single-crystalline-YAG-fibre/Mo-matrix composites versus pulling up rate of the oxide/molybdenum block during crystallisation of the fibre.

AC C

EP

TE D

M AN U

SC

The crystallisation rate does not affect fracture toughness of the specimen measured according to ASTM 399, Fig. 5. The values of K* for composite specimens measured in accordance to the standard developed for testing metal alloys should be called the apparent critical stress intensity factor because the usage of linear fracture mechanics for fibrous composites, in which a fracture zones at the crack tip can be much larger than the plastic zones in metals, can hardly be fully justified [3]. It should be noted that the value of K* of the molybdenum matrix undergone to heating up to nearly 2000oC and, hence, to recrystallisation, does not reach 10 MPa·m1/2. Reinforcing recrystallised molybdenum with the oxide fibre does actually double the value of K* of the matrix.

Fig. 5. Critical stress intensity factor YAG-fibre/Mo-matrix composites and the Momatrix versus pulling up rate of the oxide/molybdenum block during crystallisation of the fibre. A bit more reliable estimation of the crack resistance is based on evaluating notch sensitively. Corresponding data obtained along the way described in Ref. [3] are presented in Fig. 6, which demonstrates a correlation between ∗ ⁄∗ and K*. Note that some specimens are characterized by the value of ∗ ⁄∗ close to 1. Obviously, the notch sensitivity cannot be used for calculation of the ultimate load of a composite structural element containing crack-like defects. However, the same is true for the stress intensity factor measured for composites since linear fracture mechanics can hardly be used is not applicable for composites under consideration. An important problem to be solved in future is to make clear reasons for a rather large scatter of the data of both 5

ACCEPTED MANUSCRIPT

RI PT

ratio ∗ ⁄∗ and K*. The problem being solved will open a possibility to optimise the microstructure of the composites to increase fracture toughness even more.

SC

Fig. 6. Correlation between critical stress intensity factor and notch sensitivity for YAGfibre/Mo-matrix composites.

AC C

EP

TE D

M AN U

High strength of the composites at a temperature of 1400oC promises high creep resistance. Creep experiments are to be performed at the next stage of the work. Here we present (Fig. 7) the creep data for Al2O3-YAG/Mo-matrix composite obtained earlier in comparison with those for a molybdenum alloy developed for the use temperature of about 1300oC [17]. Fibre strength determines creep resistance of such composites unlike the case of metal alloys, in which creep resistance is determined by dislocation motions around particles. This leads to an increase in a difference between the values of stress rupture of the composite and alloy with increasing the loading time (see extrapolations in the graph).

Fig. 7. Creep strength of an Al2O3-Y3Al5O12/Mo composite with fibre volume fraction of 40% and that of most creep resistant three-phase Mo-Si-B alloy (acquired from experimental results in Ref. [17]. Linear extrapolations to rapture time 1000 h is shown by the dash lines. 4.2 Mullite-ZrO2-fibre/molybdenum-matrix The microstructure of fibres in the mullite-zirconia system depends on a particular ratio mullite:zirconia determined by the ZrO2:SiO2:Al2O3 ratio in the a raw mixture of the oxides. A variety of the microstructures of crystallised oxides of this system [18] calls for crystallising fibres of various compositions; corresponding results will be published elsewhere. Here we will focus on one composition that is an off-eutectic composition crystallised from the following raw mixture of the oxides: 67% ZrO2 -10% SiO26

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

23%Al2O3 (mass percent). Typical microstructures of the fibre are presented in Fig. 8, X-ray spectra of the fibres are shown in Fig. 9. It is interesting to note that (i) the microstructures of a set of the fibres in one specimen can vary, perhaps, as a result of a special way of crystallising the fibres in the channels of a molybdenum matrix; (ii) tetragonal phase of ZrO2 exists in the fibres despite no doping to stabilise this phase has been used in the raw oxide mixture.

AC C

EP

TE D

(a)

(b)

7

SC

RI PT

ACCEPTED MANUSCRIPT

M AN U

(c)

Fig. 8. Characteristic microstructures of the fibres obtained at crystallisation rates 50 (a and b) and 250 (c) mm/min. “White” phase is zirconia, “black” phase is mullite. C0067:Al-Zr-Si

v

l

x-Aliminum Oxide Mullite (00-089-2645) l-tetrag.ZrO2 (00-079-1769) v-monokl.ZrO2 (00-037-1484) o-Al2O3 (00-089-3072) l

TE D

Intensity, a.u.

l

v v

x

x

v

o v

EP

x

10

20

vv

x

v

v

l

v

x

v

xo v

30

xv

v xo vv

40

v

v

x

v o

x

50

o

v

v vv

60

v o x x

70

2 theta, degree

AC C

Fig. 9. X-ray spectra of the fibres obtained at crystallisation rate 50 mm/min. Strength/temperature dependence for the composites with fibre volume fraction

  0.40  0.05 is presented in Fig. 10. One can see that increasing the fibre crystallisation rate from 50 to 250 mm/min leads to a slight increase in the composite strength at all testing temperature, from 20 to 1400oC. The composite strength depends nearly linearly on temperature in the temperature interval from 20 to 1400oC. If to calculate effective values of the fibre strength taking into account a strength/temperature dependence of the recrystallised molybdenum matrix [10] we will see also a nearly linear temperature dependence of the fibre strength, Fig. 11. This is an unexpected result since usually oxide fibres grown from the melt retain their room temperature strength up to a temperature of about 1000oC. Note that the calculated data presented in Fig. 11 were obtained by using flexure strength values for the composites and tensile data for the matrix. At the present time, the tensile strength data for the composites are not available; the appropriate bending loading of matrix with unfilled channels in it is impossible. This means that the real strength of the fibre is slightly less than the 8

ACCEPTED MANUSCRIPT

RI PT

calculated values, but it does not affect the conclusion on the linearity of strength/temperature dependence.

TE D

M AN U

SC

Fig. 10. Flexture strength versus test temperature for the Mullite-ZrO2fibre/Molybdenum-matrix composites obtained at pulling up rates in fibre crystallisation process 50 and 250 mm/min.

EP

Fig. 11. Temperature dependence of tensile strength of the recrystallised molybdenum matrix and calculated flexure strength of the mullite-zirconia fibre.

AC C

The apparent value of critical stress intensity factor of the composites is increasing with fibre crystallisation rates increasing, Fig. 12. As in the case of the composites with YAG fibre we see an essential increase in the fracture toughness of the composites as compared with that of the matrix. Measuring notch sensitivity of the composites gives ∗ ⁄∗  0.73  0.18, which looks acceptable for structural materials.

9

ACCEPTED MANUSCRIPT

Fig. 12. Critical stress intensity factor of Mullite-zirconia-fibre/Molybdenum-matrix composites versus pulling up rates in fibre crystallisation process. 5. Oxide-fibre/HEA-matrix composites

RI PT

Fibres of the Al2O3-Al5Y3O12-ZrO2 eutectics were used as reinforcements. The microstructure and properties of the fibres obtained by ICM are described in details in Ref. [19]. A high entropy alloy containing Fe, Co, Ni, Cr and W with melting point of about 1400oC was chosen as the matrix. The alloy is composed of two phases, namely a the base BCC phase and inclusions of µ-phase of a Fe7W6-structure containing all the elements of the alloy, but their rations differ from that of the base phase. Such microstructure determines high strength and ductility of the alloy [20].

M AN U

SC

Specimens of a length of 60 mm and diameter of 4 mm were produced by pressure liquid infiltration [21]. A casting mold made of quartz had a molybdenum foil of a thickness of about 25 microns dividing quartz and the matrix melt to prevent interactions between the highly reactive HEA and quartz. The casting mold filled with the fibres is placed in a chamber for pressure casting, which is pumping out to a pressure of about 10-1 mm Hg. The matrix material locating in a crucible at the top of the mold is then heated up to a temperature of 1530oC argon gas under a pressure of 1.5 atm lets to the chamber. The infiltration time was 0.5 min. It is important to emphasize that the necessary gas pressure is much lower than that in case of infiltrating oxide fibres with superalloys, the latter being no less than 5 atm. [22]. This is an evidence of good wetting in this system, which is to provide sufficiently high strength to the fibre/matrix interface [23].

EP

TE D

A typical microstructure of the composite is presented in Fig. 13a. The micrograph shows that the matrix microstructure is composed of two phases. Results of the X-ray microanalysis of the matrix are presented in Table 1. It can be seen that the base phase (called “grey” in the Table) contains all the elements in nearly equi-atomic ratios, tungsten being an exception. Molybdenum occurs as a result of partial dissolution of the molybdenum foil used to prevent an interaction of the molten matrix alloy and silica in the mold. The composition of an inclusion (“white” phase in the Table) is close to Fe1(x+y+z)CrxCoyNizW6. A deviation of the measured composition from the exact composition of µ-phase is partially a result of capturing of a part of the base phase by the ionization pear caused by the electron beam.

AC C

Table 1. Results of the X-ray microanalysis of the matrix in the composite shown in Fig. 13a (in at%). Element

“Grey” phase

“White” Phase

Cr Fe Co Ni Mo W

21.3 22.0 23.3 24.6 2.1 6.89

15.9 15.3 19.0 10.7 0 35.71

10

AC C

EP

TE D

M AN U

(a)

SC

RI PT

ACCEPTED MANUSCRIPT

(b)

11

M AN U

(c)

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 13. SEM-micrograph of a part of the cross-section of a specimen with Al2O3Y3Al5O12-eutectic fibre and the HEA-matrix in the as received state (a) and after heat treatments 1400oC – 1 h (b) and 1400oC – 6 h (c).

AC C

EP

TE D

Composite specimens have been tested in 3-point bending to measure strength at 20 1300oC. Temperature dependence of the strength is presented in Fig. 14. Two interesting observations [21] should be pointed out.

Fig. 14. Temperature dependence of the bending strength of the composites reinforced with Al2O3-Y3Al5O12-ZrO2-eutectic fibre. The fibre volume fractions are between 40 and 50%. First, the values of composite strength at room temperature and at 1200oC are about the same. This can be a result of high fibre/matrix interface strength, which yields loading the fibre up to high stress. An attempt to relate composite strength to the fibre strength at of a critical length determined by Kelly [24] has no meaning here since the ratio of an effective strength of oxide fibre produced by ICM in the matrix to the fibre strength measured on separate fibres can be as high as 5 [23]. A particular value of the ratio depends on the fibre/matrix interface microstructure, which determines the interface strength. 12

ACCEPTED MANUSCRIPT

SC

RI PT

Secondly, composite specimens are fracturing in a non-brittle manner (Fig. 15) despite they have a large, up to 50%, volume fraction of the brittle fibres. Hence, they behave as a typical brittle-fibre/ductile-matrix composites, boron/aluminium [2] and oxide/nickel fibrous composites [7], being examples.

M AN U

Fig. 15. A typical load/displacement curve obtained in 3-point bending of a specimen with Al2O3-Y3Al5O12-eutectic fibre and the HEA-matrix at room temperature.

AC C

EP

TE D

Composites of a type under consideration are to be developed for high temperature use. Hence, creep properties of the composites are of importance. Although at the present stage of the work no creep tests have been performed, the results of the strength tests together with a previous experience of the author in the field of creep of fibrous composites allow formulating some expectations. Theory of creep behaviour of metal matrix composites [9,25] and a large bulk of the experimental work [7,8,25] have shown that the most important structural parameter of composites with oxide fibres produced via liquid infiltration way is the fibre/matrix interface strength determined by a degree of wetting in the system. If a matrix melt wets a fibre then the matrix heals the surface defects that yields an increase in the effective fibre strength by a factor reaching 5 in case of the fibre obtained by the internal crystallisation method [23]. The effective fibre strength determines creep resistance of composites at temperatures, at which fibre creep is negligible. For oxide eutectics used to reinforce the HEA, a temperature of 1200oC is sufficiently low in this sense. Therefore, a high creep resistance of the composites described here at temperatures up to 1200oC is to be expected. Obviously, such ICMoxide fibres as YAG [15] and mullite [26,27] can effectively reinforce HEAs of a higher melting points than that used in the work just described. This makes oxide-fibre/HEAmatrix composites prospective candidates for future heat resistant composites. An observation of the evolution of the fibre/matrix interface and matrix microstructures during heat treatment of a composite specimen at a temperature of 1400oC in vacuum reveals an effect of chromium on wetting oxide fibres with a Cr-containing matrix. The composite microstructure changes even after 1 h of heat treatment (see Fig. 13b.). With time increasing the changes are becoming more pronounced (Fig. 13c). Chemical composition of the phases at the interface and away of it in the matrix is also changing, Tables 2 and 3 and Fig. 16.

13

ACCEPTED MANUSCRIPT

SC

RI PT

Table 2. Results of X-ray microanalysis of the in the specimen on Fig. 13b after heat treatment 1400oC – 1 h (in at%). “White” Matrix Matrix “Black” phase on the phase on (“white”) (“grey”) interface the phase phase interface Points → 1 2 4 5 6 Al 0.57 0.22 33.10 33.52 0.63 Cr 11.72 11.75 24.32 40.41 12.15 Fe 15.57 20.86 4.30 6.66 21.66 Co 18.68 23.71 5.64 6.22 22.86 Ni 13.94 30.32 3.67 8.45 30.69 Y 0.00 0.00 11.30 1.53 0.00 Zr 0.00 0.00 7.28 1.13 0.00 Mo 7.89 3.89 2.96 0.64 4.98 W 31.62 9.12 7.44 1.03 7.03

“Black” phase 10 19.94 0.14 14.35 1.20 1.19 1.40 0.96 0.93 0.12 0.16

AC C

EP

TE D

Points → Al Si Cr Fe Co Ni Y Zr Mo W

M AN U

Table 3. Results of X-ray microanalysis of the phases on the interface in the specimen shown in Fig.13c after heat treatment 1400oC – 6 h (in at%).

14

“White” phase 11 8.41 0.00 11.45 17.40 19.50 19.38 0.86 1.06 4.26 17.69

RI PT

ACCEPTED MANUSCRIPT

SC

Fig. 16. Dependence of the chromium and nickel content in two matrix phases in the Al2O3-Y3Al5O12-ZrO2-fibre/FeCoNiCrW-matrix composite on heating time at a temperature of 1400oC in vacuum. One We can see that

6. Conclusions

TE D

M AN U

1. With heat treatment time increasing, the size of white inclusions (µ-phase) in the matrix is increasing. 2. The heat treatment makes a smooth interface to become rough; at the interface new inclusions of the (AlCr)2O3 phase occur. The size of the inclusions becomes 2 to 3 microns after 6 h of heating. 3. The content of chromium at the interface is going up. The matrix phases are depleting with chromium. It is accompanied with a corresponding increase in nickel content in the matrix.

The analysis of the data obtained in experiments with three families of metal matrix composites reinforced with oxide fibres produced by the internal crystallisation method yields the following main conclusions:

AC C

EP

1. Composites with nickel-based matrix are characterised by quasi-plastic behaviour and acceptable oxidation resistance. However, they cannot be used at temperatures above 1200oC, which is higher than that for nickel superalloys but lower than the use temperatures for really prospective thermal machines. 2. Discovering possibilities to reduce essentially oxidation rate of molybdenum reinforced with oxide fibres of special chemical compositions makes molybdenum matrix composites being prospective heat resistant materials with high creep resistance at high temperatures and sufficiently high fracture toughness at low temperatures. 3. A large choice of entropy alloys with a variety of the properties as a matrix and availability of large number of oxide fibres produced by internal crystallisation method make oxide-fibre/HEA-matrix composites highly prospective heat resistant materials.

Aknowledgements The work was supported by Russian Science Foundation, Project #16-19-10624. The author thanks his colleagues, Mrs N.A. Prokopenko, Mr A.A. Kolchin, Dr N.I. Novokhatskaya, Mr A.Ya. Mizkevich, Mr V.A. Chumichev, for their efforts to in obtaining the experimental results. 15

ACCEPTED MANUSCRIPT

References

M AN U

SC

RI PT

[1] M. Heilmaier, M. Krüger, H. Saage, J. Rösler, D. Mukherji, U. Glatzel, R. Völkl, R. Hüttner,G. Eggeler, Ch. Somsen, T. Depka, Metallic materials for structural applications beyond nickel-based superalloys. J. Metals, 61 (2009) 61-67. DOI: 10.1007/s11837-009-0106-7 [2] S.T. Mileiko, Metal and Ceramic Based Composites, Elsevier, Amsterdam, 1997. [3] S.T. Mileiko, Fracture-toughness/notch-sensitivity correlation for metal- and ceramic-based fibrous composites, Composites Part B 116 (2017) 1-6. DOI: 10.1016/j.compositesb.2016.12.060 [4] J.W. Yeh, S.K. Chen, S.J. Lin, J.Y. Gan, T.S. Chin, T.T. Shun, C.H. Tsau, S.Y.Chang, Nanostructured high-entropy alloys with multiple principal elements: Novel alloy design concepts and outcomes. Adv. Eng. Mater. 6 (2004) 299–303. DOI: 10.1002/adem.200300567 [5] S.T. Mileiko, Composites and Nanostructures. Composites and Nanostructures, 1,#1 (2009) 6-37 (in Russian). [6] S.T.Mileiko, Single crystalline oxide fibres for heat-resistant composites, Compos. Sci. and Technol. 65 (2005) 2500-2513. DOI: 10.1016/j.compscitech.2005.05.029 [7] S. T. Mileiko, V.M. Kolchin, N.I. Novokhatskaya, K. V.Van, O.A. Bazyleva, Yu.A. Bondarenko, Creep of oxide/nickel composites, Composites and Nanostructures, 1 #4 (2009) 5-18 (in Russian). [8] N.I. Novokhatskaya, A.N. Tolstun, V.M. Kiiko, A.A. Kolchin, S.T. Mileiko, An effect of non-homogeneous fibre packing on mechanical properties of oxide/nickel composites, Composites and Nanostructures, 3 #1 (2011) 5-17 (in Russian).

EP

TE D

[9] S.T. Mileiko, Speeding up creep tests of novel high temperature composites, in: J.F.S. Gomes and M.A.P. Vaz (Eds.) Proc. of 15th International Conference on Experimental Mechanics, Porto, 22-27 July 2012, INEGI, 2012, pp. 13-14. [10] S.T. Mileiko and V.I. Kazmin,Structure and mechanical properties of oxide fibre reinforced metal matrix composites produced by the internal crystallisation method, Comp. Sci. & Tech. 45 (1992) 209-220. 10.1016/0266-3538(92)90081-D [11] S.T. Mileiko and N.I. Novokhatskaya, On a possibility to make heat resistant composites of high gas corrosion resistance based on refractory metal matrix, Composites and Nanostructures, 4 # 4 (2012) 5-14 (in Russian).

AC C

[12] S.T. Mileiko, N.I. Novokhatskaya, High temperature oxide-fibre/molybdenummatrix composites of improved oxidation resistance, J. Materials Engineering and Performance 24 (2015) 2836-2840. DOI: 10.1007/s11665-014-1305-0 [13] N.I. Novokhatskaya, N.A. Prokopenko, A.Ya. Mizkevich, Oxidation of molybdenum matrix composites containing oxide fibre of a special chemical composition, in: J.F.S. Gomes, S.A. Meguid Eds.) Proc. 5th International Conference on Integrity-Reliability-Failure, 2016, pp. 315-320. [14] R.G. Carvalho, F.J. Oliveira, R.F. Silva, F.M. Costa, Mechanical behaviour of zirconia–mullite directionally solidified eutectics, Materials and Design, 61 (2014) 211216 DOI: 10.1016/j.matdes.2014.04.050 [15] S.T. Mileiko, V.N. Kurlov, A.A. Kolchin and V.M. Kiiko, Fabrication, properties and usage of single-crystalline YAG fibres, J. European Ceramic Society, 22 (2002) 1831-1837. DOI: 10.1016/S0955-2219(01)00505-2 [16] G.S. Corman, Creep of yttrium aluminium garnet single crystals, J. Mater. Sci. Lett., 12 (1993) 379-382. 16

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

[17] V. Jain, K.S. Kumar, Tensile creep of Mo–Si–B alloys, Acta Mater. 58 (2010) 2124–2142 (2010). DOI: 10.1016/j.actamat.2009.11.054 [18] R.G. Carvalho, M.S. Pires, A.J.S. Fernandes, R.F. Silva, F.M. Costa, Directionally solidified eutectic and off-eutectic mullite–zirconia fibres, J. Europ. Ceram. Soc. 33 (2013) 953–963. DOI: 10.1016/j.jeurceramsoc.2012.09.032 [19] A.N. Tolstun, V.M. Kiiko, V.N. Kurlov, A.A. Kolchin, , A.A. Kolchin, N.I. Novokhatskaya, S.T. Mileiko, Deformation and Fracture of Materials, #3 (2007) 12-20 (in Russian). [20] S.A. Firstov, V.F.Gorban, N.A.Krapivka, E.P. Pechkovsky, A.V.Samilyuk, High temperature mechanical properties of cast nano-structured high entropy alloys, in: Proc 4th Intern Conf Deformation and Fracture of Materials and Nanomaterials, Moscow: IMET RAS, 2011, p. 396–398. [21] S.T. Mileiko,.S.A. Firstov, N,A. Novokhatskaya, V.F. Gorban’, N.P. Krapivka, Oxide-fibre/high-entropy-alloy-matrix composites, Composites Part A-Applied Science and Manufacturing, 76 (2015) 131-134. DOI: 10.1016/j.compositesa.2015.05.023 [22] S.T.Mileiko, V.M.Kiiko, A.A.Kolchin, A.V.Serebryakov, V.P.Korzhov, M.Yu.Starostin and N.S.Sarkissyan, Oxide-fibre/Ni-based matrix composites – I: Fabrication and microstructure, Compos. Sci. and Technol., 62 (2002) 167-179. [23] S.T.Mileiko, N.S.Sarkissyan, A.A.Kolchin, V.M.Kiiko, Oxide fibres in a Ni-based matrix – do they degrade or become stronger? Journal of Materials: Design and Applications, 218 (2004) 193-200. [24] A. Kelly, The strengthening of metals by dispersed particles, Proc. R. Soc. Lond. A. 282 (1964) 63-79. DOI: 10.1098/rspa.1964.0214

AC C

EP

TE D

[25] S.T.Mileiko, Oxide-fibre/Ni-based matrix composites – III: A creep model and analysis of experimental data, Compos. Sci. and Technol., 62 (2002) 195-204. DOI: 10.1016/S0266-3538(01)00162-2 [26] C. H. Rüscher, S.T. Mileiko, H. Schneider, Mullite single crystal fibers produced by the Internal Crystallisation Method (ICM), Journal of the European Ceramic Society, 23 (2003) 3113–3117. [27] S.T. Mileiko, A.V. Serebryakov, V.M. Kiiko, A.A. Kolchin, V.N. Kurlov, N.I. Novokhatskaya, Single crystalline mullite fibres obtained by the internal crystallisation method: microstructure and creep resistance, J Europ Ceram Soc, 29 (2009) 337-345. DOI: 10.1016/j.jeurceramsoc.2008.06.022

17

ACCEPTED MANUSCRIPT

Characteristics of oxide-fibre/nickel-based matrix composites are reviewed. Oxide-fibre/molybdenum-matrix composites are produced and characterised. MMC with HEA matrix is obtained and prospects of such composites are discussed.

AC C

EP

TE D

M AN U

SC

RI PT

• • •