Cr₃C₂–20%Ni cermets prepared by high energy milling and reactive sintering, and their mechanical properties

Introduction

Chromium carbides present three crystallographic structures: cubic Cr₂₃C₆, hexagonal Cr₇C₃ and orthorhombic Cr₃C₂, while the last one possessing the best mechanical properties.^{Citation1,Citation2} Cr₃C₂–Ni cermets are prospective materials to operate in corrosive and abrasive environments.^{Citation3–6} Combination of such important properties as high hardness, good strength, medium fracture toughness and excellent oxidation resistance results in growing interest to Cr₃C₂-based cermets.^Citation7 The sintering method of cermets is usually carried out in the presence of liquid phase at temperatures range from 1473 to 1573 K. This liquid-phase sintering process could greatly enhance the relative density of the cermets because of the liquid reaction theory.^Citation8

Cr₃C₂–Ni cermets have been exploited for numbers of applications particularly in chemical industries due to their unique properties, such as high resistance to wear, erosion and corrosion at elevated temperatures. The application of chromium carbide cermets includes shaft bearings, seals, valve components, measuring tools and die parts operating at high temperatures.^{Citation9–12}

The main disadvantages of Cr₃C₂–Ni cermets are their relatively low mechanical properties for their coarse-grained structure (the Cr₃C₂ grain size is usually over 4 μm).^Citation13 Duran and Eroglu^Citation8 successfully prepared the Cr₃C₂ cermets and researched their microstructure and mechanical properties. The average grain size of Cr₃C₂ is about 10 μm for the 75 wt-% Cr₃C₂ sample sintering for 30 min. The hardness and transverse rupture strength are about 85.2HRA and 620 MPa, respectively. In order to decrease the grain size of Cr₃C₂, Hussainova et al.^Citation14 added some amount of Mo into the Cr₃C₂ cermets. The grain sizes of Cr₃C₂ with and without Mo addition are about 3–8 and 4–15 μm, respectively. The decreases of the grain size of Cr₃C₂ effectively increase the mechanical properties of the Cr₃C₂ cermets.

The milled powder is the foundation to prepare the Cr₃C₂–Ni cermets with excellent mechanical properties. Mean grain size of the milled powder is the most important factor determining the structure and properties of Cr₃C₂–Ni cermets. However, there is a lack of information about the effect of the appropriate milled-powder preparation technique on the microstructure and mechanical properties of Cr₃C₂–Ni cermets. Meanwhile, there is a lack of theoretical foundation about the optimal sintering temperature which is often achieved by experimental methods.

The aim of present work was to investigate the influence of milling time on the morphology and distribution of the milled powder and research the optimal sintering temperature through subsequent sintering technique.

Materials and experimental procedures

Pure chromium (>99.0 pure, 10–20 μm), amorphous carbon (carbon black, >99.0 pure, 1–5 μm) and nickel (>99.0 pure, 1–5 μm) powders were weighed with an electronic scale and then milled in a planetary ball mill at 400 rpm. All powders were supplied by XSF Technology Co., Ltd, Shenzhen, Guangdong Province, PR China. The Ni content was of 20 wt-% and Cr/C = 3:2 in atomic ratio. Agate balls (diameter 5 mm) and vial were used, and the ball-to-powder ratio was 10:1. Seven groups of milled powders (milled for 1, 3, 5, 7, 9, 11 and 13 h) were obtained.

The as-milled powders were cold-isostatically pressed at 200 MPa into cylindrical compacts (about 30 mm diameter × 5 mm length) and then sintered for 1 h at 1453, 1503 and 1553 K in a vacuum furnace with graphite heaters. The heating rate and the cooling rate of each sintering process were about 5 and 5 K min⁻¹, respectively.

Phase compositions were identified by X-ray diffraction (XRD) on a Rigaku D/Max-2400 diffractometer with Cu Ka radiation at 40 kV and 200 mA as an X-ray source. For microstructural analysis, samples were polished using standard metallographic techniques and then examined by a scanning electron microscope (SEM) equipped with an energy-dispersive X-ray spectrometer (EDS). Chemical composition distributions of the phases in the microstructure were observed by JXA-8230 electron probe microanalysis (EPMA) with an acceleration voltage of 15 kV. Particle size of the milled powders was observed by JEM-200CX transmission electron microscopy (TEM) with an acceleration voltage of 120 kV. A small drop of the milled powders re-dispersed by ethanol was dropped on a carbon film-coated copper grid. The relative densities of sintered samples were measured by the Archimedes' method.^Citation15 Each test result was the average of at least five values.

After the XRD test of the milled powder with different milling times, full width at half maximum (FWHM) of the diffraction peaks was obtained by the XRD analysis software (MDI Jade 5.0). The average grain size was then calculated according to the FWHM result and Scherrer equation.^Citation20

Rockwell hardness (HRA) was measured using a diamond cone indenter with an angle of 120° at a load of 60 kg and a loading time of 10 s. Each test result was the average of 20 hardness values measured on each sample to ensure the precision of the hardness. According to test methods,^{Citation16,Citation17} sintered samples were cut and polished to the final dimensions of 3 mm × 4 mm × 20 mm, the bending strength were measured using a three-point bending test (INSTRON-1195) with a crosshead speed of 0.5 mm min⁻¹. Each test result was the average of at least five values.

Results and discussion

The XRD patterns of Cr, C and Ni powders with different milling times are shown in Fig. 1. From this figure, it can be known that no new phase appears at milling time from 1 to 13 h, indicating that there is no mechanical alloying reaction between Cr, C and Ni powders. This phenomenon is consistent with the result reported by Sharafi and Gomari.^{Citation18,Citation19} Meanwhile, the position of Cr and Ni peaks has not shifted. From Fig. 1, it can be seen that the peaks of C powder do not appear mainly because of the C powder used possessing an amorphous structure.

1 XRD patterns of Cr, C and Ni milled powder after different milling times

Display full size

Figure 2 shows the variation trend of the FWHM of diffraction peaks of Cr and Ni with the milling times. From this figure, it can be seen that between 1- and 11-h milling time, the FWHM broadens gradually due to grain refining and lattice distortion which is caused by high energy milling. However, with longer milling time from 11 to 13 h, the FWHM has no obvious change. This phenomenon illustrates that grain size of milled powder has no distinct change as the milling time exceeds to 11 h. Obviously, 11-h milling time is the optimal factor for powder preparation. According to FWHM as shown in Fig. 2 and Scherrer equation,^Citation20 the average grain sizes of Cr and Ni after 11-h milling are about 21.4 and 21.5 nm, respectively.

2 FWHM of diffraction peak of Cr and Ni after different milling times

Display full size

In order to further indicate the average particle size of the milled powder after milling 11 h, TEM measurement was carried out on the milled powder, as shown in Fig. 3. The milled powder shows good dispersion and is mainly composed of spherical or near-spherical particles with an average particle size of 20–30 nm, which agrees well with the calculation result obtained by Scherrer equation.

3 TEM micrograph of the milled powder after milling 11 h

Display full size

Figure 4 a–e shows the SEM morphologies of the powders milled for different times. As it is shown in this figure, with the milling time increasing from 1 to 11 h, the morphologies of the milled powder change from large flakes to fine particles and the distributions become more and more uniform (Fig. 4 a–d). Because the powders of Cr and Ni have excellent ductility under the serious impact and crush by the milling balls, the morphology of the milled powder presents large flakes at the initial milling stage such as milling 1 h. With increasing the milling time from 1 to 11 h, the powders become more and more brittle because of the work-hardening, resulting in the morphologies of the powders change from large flakes to fine particles. However, with further milling, the powders hardly become finer due to the limitation of the milling condition such as process parameters adopted in the test. So, the morphologies of the powders have no obvious change after milling 11 h (Fig. 4 d and e).

4 SEM morphologies of the milled powder after different milling times a 1, b 5, c 9, d 11 and e 13 h

Display full size

In order to observe the distribution of C powders inside the Cr and Ni powders, the milled powders with different milling times were mounted using the mounting resin, respectively. The milled-powder samples were grinded orderly on the coarse and fine abrasive papers, and then polished slightly until there was no scratch in the different areas observed by the optical microscope. The cross-sections of the milled powder after milling 1, 5, 9, 11 and 13 h are prepared and the cross-section figures are shown in Fig. 5 a–e. As it is shown in this figure, the Cr, Ni and C phases are labelled by arrows. With increasing milling time, the distribution of the milled powder becomes gradually uniform first, and then has no obvious change as the milling time exceeds 11 h. After 1-h milling, the C powder lies outside the Cr and Ni powders (Fig. 5 a), indicating that there is no mixing phenomenon between Cr, C and Ni powders mainly because of the shorter milling time. With increasing ball milling time from 1 to 11 h, the powders become more and more brittle because of the work-hardening, causing the morphologies change distinctly, which has been explained above. From Fig. 5 b–d, it can be seen that the grain sizes of Cr and Ni decrease obviously mainly because of the adequately milling of the three different powders. Furthermore, the distribution of the milled powder becomes more and more uniform. However, with milling time exceeding 11 h, the distribution of the milled powder has no obvious change (Fig. 5 d and e). The result further illustrates that the 11-h milling as defined above is reasonable to improve the possibility of successfully preparing Cr₃C₂–20%Ni cermets.

5 Distributions of the milled powder after different milling times a 1, b 5, c 9, d 11, e 13 h and d₁–d₃ EDS point analysis at three different points of Fig. 5d

Display full size

In order to further support the explanation on the SEM images in Fig. 5, EDS point analysis was used on the milled powder after milling 11 h, as shown in Fig. 5 d₁–d₃ and Table 1. From Fig. 5 d₁–d₃ and Table 1, it can be seen that every test point contains three different chemical elements (Cr, Ni and C), even though one of them is the main chemical element. This phenomenon further illustrates that the milled powder shows uniform distribution after milling 11 h.

Figure 6 shows XRD pattern of Cr₃C₂–20%Ni sample sintered at 1503 K for 1 h. As known from the pattern, sintered sample consists of two phases including orthorhombic Cr₃C₂ and face-centred cubic Ni. It should be noted that, prior to forming liquid phase, a considerable amount of solid-state sintering takes place through interdiffusion of Cr and C to form Cr₃C₂. When the liquid phase appears, it fills the voids of Cr₃C₂ particles through capillary action, leading to higher relative density. The liquid phase also facilitates Cr₃C₂ particles' transport, which further increases the relative density.

6 XRD pattern of Cr₃C₂–20%Ni sample sintered at 1503 K for 1 h

Display full size

Figure 7 a–c shows SEM photographs of Cr₃C₂–20%Ni samples by reactive carburising sintering. From this figure, it can be concluded that different colour areas indicate Ni phase as the lightest area and Cr₃C₂ phase as the light grey area. From Fig. 7 a, it can be seen that there are some voids in Cr₃C₂–20%Ni cermets labelled by arrows and the average carbide grain size is about 3–4 μm. As the sintering temperature reaches 1503 K, the voids of Cr₃C₂–20%Ni cermets are less than those samples at 1453 K and the average carbide grain size is about 4–5 μm (Fig. 7 b), slightly larger than that mentioned above. Obviously, most of the carbide grains were of the same shape, approaching that of an equiaxed irregular hexagon or tetragon (Fig. 7 a and b). However, a higher sintering temperature of 1553 K increases the trend of the grain boundary diffusion of Cr₃C₂, leading to the merging and growing of carbide grains and the average carbide grain size is about 20–30 μm (Fig. 7 c). There are lots of cracks in the microstructure mainly because of the coarsening of the Cr₃C₂ grains. During cooling from sintering temperature they crack, which has a deleterious effect on the mechanical properties of the cermets. According to Ionkina and Belko,^Citation21 it was found that increasing sintering temperature not only increased the sizes of the carbide grains but also changed their shapes, which grew more and more elongated.

7 SEM photographs of Cr₃C₂–20%Ni cermets by reactive carburising sintering for 1 h at: a 1453, b 1503, c 1553 K and a₁–c₁ EPMA map analysis of a–c

Display full size

In order to confirm the chemical composition distributions of phases co-existing in the microstructure of materials after sintering, EPMA map analysis was used on the samples in Fig. 7 a–c, as shown in Fig 7 a₁–c₁, respectively. Moreover, EDS point analysis was used on the Cr₃C₂–20%Ni sample sintered at 1503 K for 1 h, as shown in Table 2. From Table 2, it can be seen that Cr/C atomic ratio in carbide phase is about 3:2. Besides, a small amount of Cr and C is detected in Ni phase, mainly because of the effect of carbide phases around the Ni phase.

The density and relative density of different sintering temperature are shown in Fig. 8. From this figure, it can be known that the density and relative density increase with sintering temperature increasing from 1453 to 1503 K due to the decrease of the voids. However, lots of cracks were formed in the chromium carbide-based cermets which caused the relative density to decrease slightly to about 98.8% when the sintering temperature exceeding 1503 K.

8 Density and relative density of different sintering temperatures

Display full size

Figure 9 shows the variation trends of the hardness and three-point bending strength of Cr₃C₂–20%Ni cermets with sintering temperature. In this figure, the hardness and bending strength increase first, due to the increase of the relative density, and then decrease when increasing sintering temperature from 1453 to 1553 K. The hardness has no obvious change, mainly because the hardness should depend on its intrinsic structure and characteristic. Therefore, the sintering temperature has a slight effect on the hardness of the cermets. However, the bending strength decreases severely when the sintering temperature above 1503 K. The decline of the bending strength is attributed to the appearance of lots of cracks in the Cr₃C₂–20%Ni cermets. Hence, the relative density has a significant effect on the mechanical properties of Cr₃C₂–20%Ni cermets. The maximum values of the hardness (86.7HRA) and bending strength (1140 MPa) were achieved at 1503 K.

9 Variation of the hardness and three-point bending strength with different sintering temperatures

Display full size

Conclusions

(i) The elemental powders of Cr, C and Ni were mixed with proper ratio (Cr/C = 3:2 atomic ratio) and milled to the nanometre crystallite sizes (about 20–30 nm). Between 1- and 11-h milling time, the morphology of milled powder becomes finer and its distribution becomes more uniform; however, with further milling, the morphology and distribution have no distinct change. Obviously, 11-h milling time is the optimal milling time for powder preparation.

(ii) The density and relative density of the sintered specimens increase at beginning and then decrease slightly with increasing sintering temperature from 1453 to 1553 K. The maximum values of the hardness (86.7HRA) and bending strength (1140 MPa) were achieved at 1503 K.