Abstract
In this work, a nanocrystallisation surface layer with an average grain size of 29 nm was fabricated on the 12Cr2Ni4A steel by the supersonic fine particles bombarding (SFPB). The vacuum carburising process was carried out on the original and the SFPB pre-treatment samples. X-ray diffraction, scanning electron microscopy and hardness tester were employed to study the phase constituents, grain size, hardness and residual stress of the two carburised samples. Experimental results showed that compared to the coarse-grain carburised layer (without SFPB), the carburised layer with SFPB pre-treatment has smaller martensite and carbide, and its thickness, hardness, compressive residual stress and wear resistance have been significantly improved. The dominant wear mechanisms of the two carburised samples were both abrasive wear and fatigue wear.
Introduction
Carburising is one of the most adopted heat technologies for surface hardening, wear resistance and fatigue strength. It not only can harden the surface layer but also keep the toughness of the core. Especially, vacuum carburising, which has been studied for many decades, is an advanced method to offer higher quality parts. Compared with the conventional gas carburising process, the vacuum carburised layer has no surface oxidisation and lower CO2 gas emission [1,2]. However, the vacuum carburising process is typically performed at the temperature above 920°C, and it also needs long durations to complete carburisation. To solve these problems in the vacuum carburising process, some methods, such as designing the composition of steel with different alloy elements and introducing rare earth elements to the process of carburisation [3], have been applied, which accelerate carburising rate, reduce the carburising temperature, and improve the quality of the carburised layer.
Owing to excellent structure and unique properties, nanomaterials have attracted great attention in materials and engineering fields. Hence, taking advantage of the nanomaterial characteristics, nanograins have been introduced into the surface of materials by severe plastic deformation in many studies [4–8]. Preliminary experimental results indicated that nanocrystallisation not only can improve surface hardness, tensile, wear, corrosion resistance and fatigue properties but also can be beneficial for the speed and quality of chemical reaction through proving atomic diffusion channels in materials [9–18]. For example, Wang et al. [12] found that by surface mechanical attrition treatment (SMAT) the diffusivity of Cr in the nanocrystalline Fe with the grain size of about 10–25 nm is 7–9 order of magnitude higher than that in typical Fe lattice. Wang et al. [13] noted that the supersonic fine particles bombarding (SFPB) treatment can form a deformation layer with the surface grain size of approximately 30 nm, which can significantly increase the sulphurising thickness and the FeS content ratio. Yuan et al. [15] discovered that the fast multiple rotation rolling technology significantly enhance the penetrating rate of Cr-rare earth-boronised sample at 650°C, which is about 1.5 times higher than that of the original sample. Tong et al. [16] compared the nitriding process of Fe assisted by the surface nanocrystaline with the traditional nitriding process of Fe, and they found that the mechanically induced surface nanocrystaline of Fe can greatly enhance nitriding kinetics and reduce the nitriding temperature at 300°C, which is much lower than conventional nitriding temperatures. Results in Sun et al. [17] showed that the nitriding sample by SMAT pre-treatment has a unique transition zone and much thicker compound layer, which could obviously enhance surface hardness and improve tensile strength.
The 12Cr2Ni4A low-carbon alloy steel is widely used for producing heavy-duty gears that requires better mechanical strength, wear resistance and small deformation after carburisation. Therefore, it puts forward higher demands of process control of carburising. Even though surface nanocrystallisation was confirmed an effective method to improve the quality and atom diffusion rate of heat treatment, until now relatively few efforts have been made to perform surface nanocrystallisation treatments on carburising. In this work, the nanostructured surface layer was obtained by the SFPB technologySFPB on the surface of the 12Cr2Ni4A low-carbon alloy steel, followed by vacuum carburising. The microstructure and wear property of the vacuum carburised and SFPB + vacuum carburised 12Cr2Ni4A steel were studied to display the influence of SFPB pre-treatment on the carburisation process at 905°C.
Experimental
The parent material used in this work was an annealed 12Cr2Ni4A gear steel, which was cut into square samples with a size of 14 mm × 14 mm × 14 mm and polished to a surfaces roughness of Ra 0.08 µm. Before carburisation, the 12Cr2Ni4A steel was treated by the SFPB to achieve a nanocrystal structure on the surface. The main parameters of the SFPB process were as follows: the airflow pressure was about 1.5 MPa, the processing duration was 240 s and the bombardments angle was 90° by a large number of 40∼60 µm diameter Al2O3 balls with the Kinetic-3000 particle bombardment system. The following carburising treatment of the SFPB sample was carried out in vacuum (1 × 10−5 Pa) circumstance at 905°C for 4 h. The process of carburising was departed into 25 pulse cycles, which will produce an even distribution of carbon concentration. As a comparison, coarse-grain sample was carburised under the same carburising condition.
The wear tests were carried out using the MMS-2A ring-on-flat reciprocating rolling tester (Jinan Yihua Tribology Testing Technology Co., Ltd, China) under the oil lubrication condition at room temperature. The lubricants were dropped on the samples before testing, and then the samples were covered by the oil. The slider of 40 mm in diameter was made of wolfram carbide fixed at the core of the samples. The slider rotation speed of 200 rev min−1, the test load was 200 N and the duration time of each test was 30 min. The wear mass loss of two carburised samples was evaluated by electronic scales with 0.01 mg.
The microstructures of the samples were examined by optical microscopy (OM) and scanning electron microscopy (SEM, FEI Quanta 200, FEI Company, Netherlands). The corrosive fluid of OM and SEM was a mixture liquid consisting of 96% nitric acid and 4% alcohol. The grain size of surface layer was measured by the X-ray diffraction using a Cu Ka radiation (X'Pert Pro MPD, PANalytical B.V. Company, Netherlands). Residual stresses of only the SFPB treatment layer, the coarse-grain carburised layer and the SFPB + carburised layer were measured by the X-350 X-ray diffraction system with Cr–Kα radiation. The thin layers of material were removed repeatedly by electro-polishing in NaCl solution to measure the residual stresses along the depth. The hardness variation along the depth of the SFPB and carburising samples along a section was measured by a microhardness tester with a load of 1000 g and a holding time of 10 s.
Results and discussions
Characterisation of the SFPB
The OM and SEM photographs of the cross-section of the SFPB sample are shown in Figure 1. The deformed layer with refined microstructure in the surface can be seen by OM (Figure 1(a)). There is a gradient variation from the surface to the matrix; however, the dividing line between the matrix and the surface deformed layer does not exist. The severe deformation layer is marked in dotted line, which is about 70 µm beneath the treated surface. To further study the changing of surface microstructure by the SFPB treatment, the SEM image was obtained. In Figure 1(b), the radial deformation flows produced by high-energy particles bombardment are observed clearly on the metal surface. The irreversible permanent micro-pits forms and plastic deformation layer generates below it when the bombardment force exceeds the yield limit of the metal. As the repeated effect of the shots enhances, severe deformation layer with a lots of boundaries and dislocation forms [19,20].
Cross-section morphology of the SFPB sample: (a) optical morphology of overall view and (b) SEM observation of the detailed examination.
Figure 2 shows the XRD patterns of the 12Cr2Ni4A stainless steel before and after the SFPB treatment. It is observed that the diffraction peaks of (110), (200) and (211) belong to martensite and there is no phase transformation after the SFPB treatment. However, the strength of diffraction peaks increase, and the significant broadening of the (110) peak can be found in the SFPB sample, which is the result of grain refining. According to the Scherrer–Wilson equation, the integral width of the broadening peak is measured to approximately calculate the grain size of the SFPB layer. Thus, the average grain size of the SFPB surface layer is 29 nm. However, XRD penetration depth is 5 µm, so the average value of the grains of the surface layer is smaller than that of the measurement data. This also suggests that a large number of grain boundaries exist in the SFPB-treated layer.
XRD patterns of the untreated sample and the SFPB sample.
In Figure 3, it is clear that the SFPB impact has a powerful effect on the hardness and residual stress of the 12Cr2Ni4A steel. The hardness along the depth of the SFPB sample is significantly higher than that of the untreated sample in the range of 400 µm. When the range is at 0∼100 µm, the hardness decreases sharply, and at the range of 100∼400 µm, the decrease trend become tardiness until the hardness of the SFPB sample is the same as that of the base material. Combining with Figure 1(a), we know that the severe deformation layer is about 70 µm beneath the treated surface; however, the influenced zone by the SFPB treatment is about 400 µm. In this range, the increase and gradient change of hardness is attributed to the refined grain and the high density dislocations [21,22]. The maximum compressive residual stress (−577 MPa) is below 50 µm from the SFPB specimen surface, and the residual stress gradually decreased along the depth from 50 to 400 µm. So, the SFPB sample has high compressive residual stress. According to the analysis of structure, the increased hardness and residual compressive stress of the SFPB sample is the combination effects of grain refinement and high density of dislocations by severe plastic deformation, which are beneficial to improve the properties and the effects on the following carburising process [22–24].
Residual stress and hardness variation along the depth of the SFPB sample.
Characterisation of carburised layer
Figure 4 shows the surface SEM morphologies of the coarse-grain and the SFPB carburised layers. As shown in Figure 4, the carburised layers are composed of lots of tempered martensite, retained austenite (RA%) and dispersed carbides. In contrast with the coarse-grain carburised sample, the smaller grain size of martensite is obtained after the SFPB pre-treatment carburised sample. To observe better, the partial magnified pictures of carburised layers are adopted in Figure 4(c,d). There are some granular and dispersed carbides existing around the martensite of both carburised layers. As a result, the diameter and ratios of carbides are measured by the Nano measurer software.
Surface microstructure of carburised layer for coarse-grain and the SFPB carburised samples: (a) coarse-grain and (b) SFPB.
The diameter distribution of the carbides in these two carburised specimens is shown in Figure 5. It can be seen that the average equivalent diameters of the carbides in the surface of both carburised layers show similarity, which are 0.18 and 0.19 µm, respectively. However, the number of diameters above 0.2 µm is about 28% in the SFPB surface carburised layer, which is the 2.5 times of coarse-grain carburised specimen (11.2%). Thus, it is manifested that there are no obvious differences in the diameters of carbides of the two surface carburised layers, excepting more amount of larger carbides in the SFPB carburised layer. Subsequently, the ratios of the long axis to the short axis of carbides in both samples are measured. The ratios of the carbides in the coarse-grain carburised surface are 1.32 ± 0.16, and that is 1.26 ± 0.15 for the SFPB carburised sample. More rounder carbides in the SFPB carburised layer than in the coarse-grain carburised layer are demonstrated.
Diameter carbide particles of top surface of carburised layers: (a) coarse-grain and (b) SFPB.
Figure 6 shows the carbide particle's average diameters along the depth of carburised layers for coarse-grain and SFPB treatment specimens according to the above method, which are the different position average values of six images in horizontal direction. Some SEM morphologies in the critical position of the carburised layer have been exhibited in terms of the experimental observations. It can be clearly seen that the diameter of carbide in the two carburised layers decreases along the depth and the smaller carbide size in SFPB carburized samples than that of coarse-grain carburized samples except the top surface. On the other hand, the size of both martensites increases.
Average diameter of carbide particles along the depth of carburised layers.
Figure 7(a) displays the residual stress and content of RA% in coarse-grain and the SFPB carburised layer, respectively. It can be seen that the RA% in the SFPB carburised sample is higher than that of the coarse-grain carburised sample within 125 µm from the top surface. The RA% of the SFPB carburised sample is less than that of the coarse-grain carburised sample and reduces gradually when the depth is above 125 µm. In the section characterisation of the SFPB, there are higher compressive residual stress in the SFPB pre-treatment sample, which can restrain austenite transform in the carburising. So, the higher RA% exists in shallow location of the SFPB carburised sample. However, a higher surface residual compressive stress was introduced in the SFPB carburised sample compared to the coarse-grain carburised specimen, maintaining this tendency within 300 µm. Experimental reports have demonstrated that the higher RA% in shallow location and large compressive residual are beneficial for improving the property of layer [25,26].
Variations of residual stress and hardness along the depth for samples: (a) residual stress and RA% and (b) hardness.
Figure 7(b) shows that the hardness of coarse-grain and the SFPB carburised decrease as the distance increases, but the hardness of SFPB carburized samples is much higher than that of the coarse-grained carburised sample. On the one hand, the hardness of the material is inversely proportional to the grain size according to the Hall–Petch relationship. The tiny martensite and smaller carbides in the SFPB carburised layer are beneficial to hardness. The hardness of retained austenite is lower than that of martensite, the lesser RA% as the depth above 125 µm in the SFPB carburised sample also contributed to the increase in hardness. So, the hardness of the SFPB carburised layer is higher than that of the coarse-grain carburised layer. Meanwhile, the hardness distribution of cross-section carburised specimens not only shows the depth of carburised layer but also is related to carbon gradient. The depth of carburised layer (550 HV1) exhibits an obvious higher thickness in the SFPB sample (approx. 1.1 mm) compared to the coarse-grain sample (approx. 0.97 mm). Therefore, the SFPB pre-treatment can improve the hardness of carburised layer and is beneficial to carbon diffusion.
Friction and wear analysis
Figure 8 shows the friction coefficient and mass loss of carburised samples investigated under the same experimental condition. It can be seen that from Figure 8(a), there is no obvious distinction of the friction coefficient in both carburised layers (approx. 0.11) at the stable wear stage, but the friction coefficient of coarse-grain carburised sample is floating as larger as that of the SFPB carburised sample during the wear test. Some relative larger hard phases (carbides) exist in coarse-grain carburised layer compared to the SFPB carburised layer; this may be the main reason for this phenomenon. Figure 8(b) manifests the wear mass loss of carburised samples. In this case, the mass loss of coarse-grain and the SFPB carburised samples are 13.9 and 12.3 mg, respectively. The SFPB carburised specimens exhibit much improved wear resistance when compared with the untreated specimen.
Friction coefficient and mass loss of both carburised samples: (a) friction coefficient and (b) mass loss.
Figure 9(a–d) shows the worn surface morphologies of the carburised layers, and it is clear that both the worn surfaces of carburised samples are smooth, with little grooves. Compared with Figure 9(c) and (d), some deeper and severer grooves, cracks and delamination can be observed on the coarse-grain carburised sample. These evidently indicate that the abrasive wear and fatigue wear were the two primary wear mechanisms for the both carburised samples.
Wear surface morphologies of the carburised layers: (a) and (c) SFPB carburised; (b) and (d) coarse-grain carburised.
Discussion
Enhancement mechanism of the SFPB on carburising behaviour
In contrast with the coarse-grain carburised sample, the carburised layer pre-treated by the SFPB has larger carburised depth, which are constituted by smaller organisations (martensite and carbides). We all know that the SFPB treatment is able to form a severe deformed layer with vast grain boundaries, sub-boundaries and dislocations [27]. But the small grains are unstable when the temperature is above the austenitising temperature. Because the grain size of austenite is the key factor to decide the maximum size of martensite plates in the phase transition process, we can consider that the grain size of austenite in the SFPB treatment surface is smaller than that of coarse-grain surface. Therefore, there is also high storage energy and activity in the SFPB-treated surface. In that way, the SFPB surface exhibits stronger adsorption ability and higher chemical activity than the coarse-grain surface during carburising. According to Fick's second law, the diffusion of carbon atoms in the sample can be considered as follows:
The SFPB pre-treatment impact on wear property of carburising
According to the above results, compared to the coarse-grain carburised sample, the SFPB pre-treatment carburised surface has lower float of friction coefficient and less mass loss under oil lubrication. Obviously, the surface nanocrystallisation pre-treatment by the SFPB on the 12Cr2Ni4A steel can improve the wear property of carburised layer greatly, which is attributed to the finer structure and high hardness of the SFPB pre-treatment sample.
Effect of structure
Comparing with the coarse-grain carburised sample, the martensite on the SFPB sample has finer size, and the carbide particles have smaller and rounder structure that make the carburised layer much denser. First, fine austenite grain forms small-sized martensite and restrains vast microcrack occurring in martensite. Generally, martensite is composed of dislocation and twin, so the smaller the martensite size, the more the grain and twin boundaries. Abundant grain and twin orientations of the martensite can resist the propagation of the fatigue crack [28]. Second, the carbon concentration in the location close to and away from the undissolved carbides in the carburised layer is different. The large carbide particles lead to the great difference in carbon concentration, while the small carbide particles lead to the small difference in carbon concentration [29]. The small carbon concentration difference in the material helps rolling contact fatigue. Generally, stress concentration easily occurs at the sharp edges of larger carbides, which is the source of the initial cracks [30–32]. These fine undissolved spherical carbides and smaller martensite in the SFPB carburised sample offer resistance to abrasive wear and promote the transformation of RA% [29,33]. Third, RA% can transform into martensite during the wear test, which leads to the absorption of more strain energy. The higher RA% in shallow location (Figure 7(a)) of the SFPB carburised layer can delay the crack initiation and propagation during the wear test [26].
Effect of hardness and residual stress
As shown in Figure 7(a), by contrast, the hardness and compressive residual stress in the SFPB carburised sample are higher than that in the coarse-grain carburised sample. In this experiment, the smaller martensite and carbon concentration are the mainly factors to bring about the high hardness in the SFPB carburised layer. A lot of different grains and twin orientations in the martensite, and the more carbides on account of higher carbon concentration both increase wear resistance [28]. Generally speaking, the delamination caused by crack extension is the mainly reason of mass loss for carburised layer in wear process. The compressive residual stress caused by the different phase transformation sequences of the surface and centre is beneficial to crack closed and improve the wear resistance of material in the sliding wear [34]. Experimental reports have demonstrated that the higher hardness and compressive residual stress, which form in the carburised layer within a certain range, can increase the wear resistance and fatigue further [26,35].
Therefore, the wear resistance of the SFPB carburised sample is better than that of the coarse-grain sample due to the smaller martensite in the top surface, the finer carbide dispersion, the reasonable RA% distribution, the higher hardness and residual compressive stresses in the carburised layer.
Conclusions
In this paper, the vacuum carburising behaviour of low-carbon steel with a nanocrystalline surface layer induced by the SFPB was investigated in comparison with that of the original sample after carburising at the same conditions. The following conclusions were obtained:
The SFPB process could produce a severe deformation layer about 70 µm with a surface grain of 29 µm, and higher distribution of hardness and compressive residual stress along the depth. In comparison with the coarse-grained carburised sample, a carburised layer composed of fine martensite, rounder and smaller carbides, formed on the SFPB carburised sample, resulting in relatively higher hardness and compressive residual stress. In addition, the SFPB pre-treatment before carburised is beneficial to carbon diffusion that significantly increased the thickness of carburised layer (approx. 1.1 mm) compared to the coarse-grain sample (approx. 0.97 mm). Wear tests showed that the SFPB carburised layer has lower float of friction coefficient and less mass loss (12.3 mg) relative to that of coarse-grain carburised layer under oil lubrication. The excellent wear performance of the SFPB carburised layer is mainly attributed to the following reasons: the smaller martensite in the top surface, the finer carbide dispersion and the reasonable RA% distribution in carburised layer, the higher hardness and residual compressive stresses in the carburised layer.
Footnotes
Disclosure statement
No potential conflict of interest was reported by the authors.