30.04.2006

L. Huang, H.Y. Wang, F. Qiu and Q.C. Jiang ^,

Key Laboratory of Automobile Materials of Ministry of Education and Department of Materials Science and Engineering, Jilin University at Nanling Campus, No. 142 Renmin Street, Changchun 130025, PR China

Received 3 October 2005; accepted 14 February 2006. Available online 29 March 2006.

1. Introduction

Self-propagating high-temperature synthesis (SHS), also termed combustion synthesis (CS) which was developed by Merzhanov and Borovinskaya in the late 1960s [1] and [2], has been expanding during the several decades due to its attractive merits of high purity of products, low processing cost, time and energy efficiency, no high-temperature furnace process and non-polluting traits, etc. [1], [2] and [3]. And now it is a new technique of synthesizing ceramic, intermetallics, ceramic matrix composites, metal matrix composites (MMCs) [1], [2], [3], [4], [5] and [6] and so on. However, a major limitation in the application of SHS reactions for the production of advanced material components is the inherent high porosity, e.g. typically 50% of the theoretical density [1], in the synthesized materials. Therefore, it is necessary to combine SHS reactions with other processing techniques such as quasi-isostatic pressing [7], hot isostatic pressing [8], extrusion [9], rolling [10], reaction infiltration [6], [11] and [12] and centrifugal casting [13], etc. to fabricate relatively dense materials.

In the past years, many investigations were widely done on TiC and/or TiB₂ particulate reinforced Al, Fe, Cu, Ni and Ti MMCs [14], [15], [16], [17], [18], [19], [20], [21], [22], [23] and [24], due to the high melting points, high hardness and modulus as well as good abrasive resistance of TiC and TiB₂ ceramics [14], [15] and [25]. Among these MMCs, TiC–Ni composite integrating metallic properties (ductility and toughness) with ceramic characteristics leads to greater strength in shear and compression as well as to higher service temperature capabilities [16]; therefore, it has obtained more and more attention recently.

Dunmead et al. [20] first reported that the kinetics and mechanisms of SHS reactions in the Ti–C and Ti–C–Ni systems by measuring the activation energies from the relation of combustion front velocity and combustion temperature, and they concluded that the product was controlled by the dissolution of carbon into a Ti–Ni melt, and that the products were formed by precipitation of TiC from the melt. Wong et al. [24] studied the phase transformations of CS in Ti–C–Ni system using time-resolved X-ray diffraction method, and found that the first step in the combustion process was the melting of titanium and nickel particles, then the molten Ti reacted with the solid C to form TiC successionally, and in the course of reaction there were Ti–Ni alloys including TiNi₃ or TiNi existing and persisted as a final reaction product. Han et al. [26] combined the SHS with quasi-isostatic pressing to investigate TiC–xNi cermets, and found that the products were TiC and Ni without Ni–Ti compounds. And more recently, Xiao et al. [16] studied the mechanism of TiC–Ni cermet SHS reactions by means of combustion front quenching method, and the results showed that the combustion reaction started with a local formation of a Ti–Ni liquid solution and could be described with a dissolution–precipitation mechanism.

However, the researches on Ni matrix composites are mainly focused on the ones fabricated from Ni–Ti–C system [16], [20], [24] and [26]; whereas the studies on Ni–Ti–B [11] and Ni–Ti–B₄C [12] systems are still limited; moreover, the composites synthesized only by SHS reactions are usually porous and hence with poor mechanical properties. Therefore, in the present work the feasibility of synthesizing Ni matrix composites with diverse reinforcements through SHS reactions, arc melting and then densifying the composites by suction casting process was investigated from Ni–Ti–C, Ni–Ti–B, Ni–Ti–B₄C and Ni–Ti–C–B systems, respectively. Furthermore, the microhardness and wear resistance of them are also tested and evaluated. It is expected that the preliminary results could be concernful in the development of the particulate reinforced Ni matrix composites.

2. Experimental

2.1. Starting materials

The starting materials were made of commercial powders of Ni (99.8% purity, not, vert, similar 5 μm), Ti (99.5% purity, 38 μm), B (98.0% purity, 3 μm), graphite (99.9% purity, 38 μm) and B₄C (98.0% purity, 3.5 μm), respectively, and the designed ratios of the products as well as the compositions of materials were shown in Table 1. The powder blends were mixed sufficiently by ball milling for 6 h and then pressed into cylindrical preforms (20 mm diameter and 15 mm length) using a stainless steel die with two plungers. The green preforms were pressed uniaxially at the pressure of 80–85 MPa to obtain densities of 62 ± 5% theoretical density. Furthermore, in order to compare with the composites, it was also designed a preform with the single composition of Ni.

Table 1.

Designed compositions of products and the needed materials as well as corresponding content of every element in the four samples of Ni–Ti–C, Ni–Ti–B, Ni–Ti–B₄C and Ni–Ti–C–B systems, respectively

Sample Composition Mass (g) Designed composition

Ni content (vol.%) Ceramic composition Molar ratio (TiC:TiB₂) Ceramic content (vol.%)

1 Ni 32.3 70.0 TiC – 30.0

Ti 6.1

C 1.6

2 Ni 32.9 70.0 TiB₂ – 30.0

Ti 4.9

B 2.2

3 Ni 32.8 70.0 TiC and TiB₂ 1:2 8.4 and 21.6

Ti 5.2

B₄C 2.0

4 Ni 32.7 70.0 TiC and TiB₂ 1:1 13.1 and 16.9

Ti 5.4

C 0.7

B 1.2

2.2. Arc melting and suction casting

The high vacuum arc melting furnace was vacuumized until the vacuum degree was up to 5 × 10⁻³ Pa after the preform was put onto the copper crucibles in it, followed by filling 0.05 MPa argon, and then the preform was melted by the electric arc under the argon atmosphere for 5 min applying the current of about 250 A with the alternative face melted in the same way successionally for the sake of the uniformity of the compositions. Then the composite melt was suctioned into a cavity with Ø7 × 120 mm³ in the copper mould with Ø110 × 125 mm³ rapidly to produce the sample in vacuum. Fig. 1 illustrates the equipment schematic of the arc melting and suction casting.

(12K)

Fig. 1. Equipment schematic of the arc melting and suction casting.

2.3. Microhardness test

Vickers microhardness values of the matrix and TiC particulates in the samples as well as pure Ni were tested under a load of 50 g with an indentation time of 10 s on Buehler Omninet Vickers hardness tester. This experiment was conducted on every sample for 10 times, and then the average of respective sample could be calculated.

2.4. Wear test

Dry sliding wear experiments were conducted in air at room temperature on a pin-on-disk machine using 600 grit abrasive papers with the load levels of 20 and 35 N as well as the sliding distance of 9 m, and the pin was loaded against the disk by a dead-weight loading system. The weight losses during the sliding tests were calculated from the weight differences of pin specimens before and after the tests to the nearest 0.1 mg, which were taken for three times, and then the average weight loss data gained were diverted to the volume losses per unit length.

2.5. Metallographic observations

Metallographic samples were prepared in accordance with standard procedures used for metallographic preparation of metal samples, and etched with about 15 vol.% HNO₃ alcoholic solution for 10–15 s at 25 °C. The microstructure analysis of the composites was investigated by scanning electron microscopy (SEM) (Model SHIMADZU SSX–500, Japan), and phase analysis was conducted by X-ray diffraction (XRD) using Cu Kα radiation (Model D/Max 2500PC Rigaku, Japan).

3. Results and discussion

3.1. SHS reaction and solidification

According to the phenomena observed, the SHS reactions of the preform were ignited firstly by the electric arc and the ceramic particulates were formed primarily, and then followed immediately by the product remelted at the high temperature of the arc. The final composites were fabricated after suction casting successionally. Therefore, although the SHS reactions have played an important part in them, the ultimate microstructures of the products, such as morphology, size, distribution, type and so on, are mainly determined by the solidification process. And the microstructures formed in this experiment are different from the ones synthesized only by SHS [11], [12], [16] and [26], which will be discussed in the following parts.

3.2. XRD results

Fig. 2 shows the XRD patterns of the four samples as well as Ni fabricated by arc melting and suction casting. It can be observed that the phases displayed in Fig. 2(a) and (b) are mainly Ni and Ni + TiC, respectively, as expected. However the products fabricated from Ni–Ti–B system are mainly composed of Ni, Ni₃Ti, Ti₂Ni and Ni₄Ti₃ as well as a small amount of TiB₂ shown in Fig. 2(c), which is different from designed compositions (Table 1). And the ones in Fig. 2(d) and (e) are mainly composed of Ni, Ni–Ti compounds (including Ni₃Ti, Ti₂Ni, and Ni₄Ti₃) and TiC as well as a relatively small amount of TiB₂.

(8K)

Fig. 2. XRD patterns of the materials fabricated by SHS reactions, arc melting and suction casting from (a) Ni, (b) Ni–Ti–C, (c) Ni–Ti–B, (d) Ni–Ti–B₄C and (e) Ni–Ti–C–B systems, respectively.

However, what is more important is that in Fig. 2(b) the peaks corresponding with Ni deviating a little to the left compared with the ones of pure Ni in Fig. 2(a), which reveals that the lattice constant of Ni matrix of sample 1 increased. According to Fig. 2(b) and literature [17], the stoichiometry of TiC_x has been calculated, and the value of x is about 0.8. Therefore, some C should be residual in the final products. However, no C phase is found in the XRD result. In addition to the volatilization of C at the high temperature of arc melting ( not, vert, similar 3773 K), some C may be dissolved into Ni matrix to form an interstitial solid solution, which leads to the increase of Ni lattice constant and correspondingly the peaks of Ni turning to the left a little.

Furthermore, according to Fig. 2(c)–(e), the Ni content reduce dramatically replaced by Ni–Ti compounds, and the content of TiB₂ is relatively small compared with the designed ratio in samples 2, 3 and 4 which may be caused by the volatilization of B phase at the high temperature. Therefore, the matrices of samples 2, 3 and 4 are mainly composed of Ni and Ni–Ti compounds, and the composition change of the matrix will result in the variation of the mechanical property and wear resistance of the materials.

3.3. SEM analysis

Fig. 3 shows the typical SEM of the four samples fabricated by arc melting and suction casting. The gray parts are the matrices and the dark ones are ceramic particulates.

(136K)

Fig. 3. Typical SEM of the samples fabricated by SHS reactions, arc melting and suction casting from (a) Ni–Ti–C, (b) Ni–Ti–B, (c) Ni–Ti–B₄C and (d) Ni–Ti–C–B systems, respectively.

It can be observed from Fig. 3(a) that TiC particulates fabricated include the coarse blocky or dentritic primary phases with an average size less than 10 μm and the Chinese script type eutectic phases. Furthermore, it is interesting to note that also some fine eutectic phases adhere to the primary TiC, which indicates the eutectic TiC is the pre-eutectic phase during the solidification process.

While the typical morphologies of sample 2 are shown in Fig. 3(b), from which it can be seen that there is a very small amount of TiB₂ distributing in the matrix. This result is consistent with the XRD analyses. Moreover, through a large number of observations it is found that the distribution of TiB₂ particulates is agglomerated in some region. The sizes of TiB₂ particulates are of some differences and less than 10 μm, while the shapes of them are irregular.

Fig. 3(c) and (d) show the typical morphologies of samples 3 and 4. The sizes of particulates are almost the same with sample 1 and the distribution of them is relatively dispersed. However, it is difficult to distinguish TiC from TiB₂ in the two SEM due to their sizes and nondescript shapes.

It is worth noting that the shapes of TiC and TiB₂ particulates formed by SHS reactions, arc melting and suction casting are different from the ones formed only by the SHS reactions. In the present study the mechanism of TiC formation is nucleation-growth, and therefore, the shapes of TiC particulates are irregular; while the ones formed only by SHS reactions are usually round with a dissolution–precipitation mechanism [16] and [26]. At the same time, TiB₂ particulates synthesized from SHS reactions are usually hexagonal prisms [11] and [12] and the mechanism needs further investigation. Otherwise the SEM shows the absence of macropores and blowholes as well as micropores, which indicates the near-fully dense composites can be fabricated by SHS reactions, arc melting and suction casting.

3.4. Microhardness and density

Moreover, Vickers microhardness values of the matrix in the four samples were tested, so were the densities of them including both the theoretical and practical ones, which are listed in Table 2.

Table 2.

Microhardness values of matrices and densities of the four samples fabricated by SHS reactions, arc melting and suction casting from Ni–Ti–C, Ni–Ti–B, Ni–Ti–B₄C and Ni–Ti–C–B systems, respectively

Sample Matrix (HV) Theoretical density (g/cm³) Practical density (g/cm³)

1 478 7.70 7.51

2 1298 7.57 8.52

3 1207 7.60 8.06

4 1009 7.63 7.86

From Table 2 it can be noted that the matrix microhardness of the four samples increases to different extents relative to pure Ni (146 HV) with several reasons, among which is the sustaining effect taken by the TiC and/or TiB₂. Furthermore, the phenomenon that the matrix microhardness of sample 1 is 478 HV may be mainly caused by the solid solution strengthening of some C dissolved into Ni matrix, which is consistent with the XRD result (Fig. 2(b)). At the same time, it can be clearly found that the matrix microhardness of sample 2 is the highest in the four ones; therefore, combined with the XRD results it is concluded that there are more Ni–Ti compounds in the matrix instead of Ni matrix with the probability of the volatilization of B phase. Moreover, the matrix microhardness values of samples 3 and 4 are also pretty high compared with sample 1 si

Synthesis of dense ceramic particulate reinforced composites from Ni–Ti–C, Ni–Ti–B, Ni–Ti–B₄C and Ni–Ti–C–B systems via the SHS reaction, arc melting and suction casting

Abstract

Article Outline

1. Introduction