Experimental and Numerical Studies on the Tribological Properties of Bearing Steel 20GrNi2MoV Against W2Mo9Cr4VCo8 Steel Under Dry Sliding Process

Abstract

In this paper, the wear characteristics of 20GrNi2MoV bearing steel under different working conditions were investigated by finite element simulation considering microscopic grain size and pin-on-disk friction experiments, and the wear mechanism during friction and wear was explained, along with a finite element model that took initial grain size and material organization into account to predict the process of subsurface crack initiation during friction. The results show that high-speed and large-load conditions have a significant effect on the wear characteristics of dry friction of pinned disks. The effect of high speed and load will greatly reduce the time of the grinding stage, and the friction coefficient can quickly reach a stable state; the roughness of the surface of the friction pair increases with the increase in load, but the roughness shows a tendency to first increase and then decrease with the increase in sliding speed. Martensitic phase transformation occurs in the friction subsurface, and the decrease in Mn element content is one of the causes of cracks on the subsurface of the material; with the increase in load and speed, the damage form of the sample disk material is changed from abrasive wear and adhesive wear to the mixture of three kinds of wear: abrasive wear, adhesive wear, and cracks. In addition, the simulation of crack initiation and growth agrees well with the experiment, which proves the accuracy of crack prediction. This study provides a reference for crack initiation prediction in the study of pinned disk friction vises.

1. Introduction

Bearings are widely used in various fields of machine manufacturing, such as aerospace, machine tools, agriculture, etc. However, bearings are often subjected to high-speed friction and heavy load impacts during operation, and these factors tend to cause early bearing failure [1]. Common forms of bearing failure during operation include wear failure, cracking, and pitting. And wear is the main form of bearing failure [2,3]. Therefore, there is an urgent need to study the form of wear failure and the failure mechanism of bearing materials in order to delay bearing wear and fatigue failure.

20GrNi2MoV is a typical metal material with good hardness and ductility [4] and is widely used in the manufacturing and processing of inner rings of bearings. Ji et al. [5] investigated the effects of load and rotational speed on the coefficient of friction and wear under dry friction conditions, and the results showed that the effect of rotational speed on the coefficient of friction is more obvious, and the effect of load on the wear is more significant; however, the study did not explore on the wear failure mechanism of the material. Sander et al. [6] investigated the wear mechanism of plain bearings under starting and stopping conditions by means of elastic hydrodynamic simulation and obtained the characteristics of wear depth and contact pressure distribution of plain bearings under different rotational speeds of the bearings, but the study did not consider the interactions between wear and cracks at the same time. Ardila et al. [7] investigated the effect of load on friction coefficient and wear by means of fixed ball microwear experiments. They explored the effect of material hardness and friction differences on the wear of GCr15 material by means of fixed ball microwear experiments. It was found that friction between different friction partners produced a large variability in particle entrainment and wear rate. Mirzababaei et al. [8] studied the wear of automotive friction materials in terms of ambient humidity and dryness. The results of the study showed that the wear was significantly reduced when humidity increased and that the same wear mechanism could be observed at different humidity levels. However, the study did not analyze the effect of specific operating conditions on material wear.

However, the wear process of materials is often accompanied by subsurface deformation [9]. From the microscopic point of view, grain refinement, uniform distribution of carburization, and strengthening of intergranular interactions are the root causes of fatigue retardation through reasonable processing [10,11]. Masoudi et al. [12] investigated the effects of load direction and wear phenomena on rolling contact fatigue crack extension. Liu et al. [13] investigated the effects of wear on the crack extension of short cracks under micromovement conditions by using a numerical method and studied the role of material wear on crack extension. However, the study did not carry out any relevant research on the microscopic grain on material crack initiation. Chen et al. [14] investigated the effects of grain size and loading conditions on the microstructure and properties of magnesium alloys through tensile and compression tests. In this study, the effect of grain size on microcrack extension was obtained, but the working conditions were not considered to have an equally important role in the process. Li et al. [15] explored the mechanism of martensite influence on the fatigue and small crack extension behavior of stainless steel and found that the martensite slat is the basic unit affecting the fatigue performance and the small crack extension behavior. However, the effect of grain size on cracking was neglected in this study. Sun et al. [16] investigated the effect of grain size on the organization evolution of Q345 steel plate under high-speed impact and analyzed the microstructure evolution of steel plates with different grain sizes and found that the small-grain steel plate had an obvious grain boundary strengthening effect, which retarded the crack initiation. However, the role of material organization in crack extension was not considered.

In summary, the above studies have not comprehensively considered the effects of working conditions on frictional wear and the effects of grain and material organization on crack initiation. However, material wear and crack initiation are the result of the combined effect of the above factors, so it is necessary to consider the effects of working conditions and micrograins and material organization on wear and cracking.

In this paper, dry friction experiments were carried out on pin disks of different materials using a pin disk-type friction and wear tester. The friction coefficient, wear loss, surface and subsurface damage morphology and characteristics of the specimen disk material were analyzed. The friction behavior of the friction vice of W2Mo9Cr4VCo8 high-speed steel pins and 20CrNi2MoV specimen disks was investigated, as well as the influence of different loads and rotational speeds on the wear performance of the 20GrNi2MoV material under dry friction conditions; the wear mechanism in the dry friction process was studied. Through in-depth analysis, the study identified the underlying physical and chemical processes governing the wear behavior of the material, which is crucial for optimizing its performance and durability. A finite element model considering the initial grain size and material organization was established, and the crack initiation during the friction process was simulated by finite element simulation, the crack initiation and expansion laws were obtained, and the accuracy of the finite element simulation was verified by experiments. This predictive capability is of great importance for ensuring the reliability and safety of components made from this material in practical engineering applications. The research results provide a reference for the optimization of steel–steel friction partners.

2. Experiment and Finite Element Simulation

2.1. Materials

The pin bar material in this paper is W2Mo9Cr4VCo8 high-speed steel, which is used to ensure that the friction process maintains good red-hot hardness and wear resistance. Since this material is widely used in the aerospace field, this paper selected this material for the pin to utilize its outstanding performance when used as a wear-resistant pin. Bearing steel 20CrNi2MoV was selected as the test ring material for the specimen part, and its chemical composition is shown in

Table 1. Chemical composition of experimental materials w/%.

Materials	C	Mn	Si	Cr	Ni	Mo	S	P	V	N	Fe
W2Mo9Cr4VCo8 [17]	0.96	0.36	0.19	1.46	0.08	0.02	0.006	0.01	/	/	Bal.
20CrNi2MoV [18]	0.2	0.61	0.25	0.56	1.77	0.26	0.002	0.007	0.21	/	Bal.

. The initial hardness of three kinds of test rings was measured as 60.3HRC, 49.3HRC, and 57.7HRC by using the HRS-150 Rockwell hardness tester of Shangcai Testermachine (SCTMC), Shanghai, China. Moreover, the average hardness values of the three groups of test rings were 60.5HRC, 49.5HRC, and 57.5HRC, respectively, and the standard deviations of the three groups of test rings were all 0.17HRC.

Cutting and CNC machining were carried out to obtain the test ring workpiece, with an outer diameter of 31.7 mm, thickness of 10 mm, and radial opening with a diameter of 2.3 mm through the hole placed in the temperature probe for the detection of friction in the process of temperature change; in order to facilitate sample mounting and dismounting, the bottom edge of the test ring C1.5 underwent a chamfering process. The diameter of the test pin was 1.8 mm and the length was 12.7 mm. Before the friction test, the test ring and the pin were polished to meet the requirements of the friction test. The surface roughness of the pin after polishing was Ra3.2, and the surface roughness of the test ring was Ra0.4. There were three pins on each pin plate, and the friction contact area of each pin and the test ring was 2.54 mm².

2.2. Experimental Procedure

The dry friction test on the pin disk was performed on the MMW-1B (Sunmao Company, Jinan, China) multifunction testing machine, as shown in Figure 1a. During the test, the carrier table and the test ring were positioned by pins and kept relatively stationary; the pin bar and the pin disk were fixedly connected by set screws and rotated with the pin disk at different rotational speeds, as shown in Figure 1b. For the different working conditions of the test ring, the load and power supply mode were consistent, as shown in Figure 1c.

The wear tests of the specimens were conducted in strict accordance with the GB/T 12444-2016 standard [19], which were promulgated by the Standardization Administration of China.

The experimental conditions are summarized in

Table 2. Experimental conditions and parameters.

Tests	D4-1	D4-2	D4-3	D8-1	D8-2	D8-3	D12-1	D12-2	D12-3
Load/N	30	60	90	30	60	90	30	60	90
Speed/rpm	400	400	400	800	800	800	1200	1200	1200
Cycle number	5 × 104	5 × 104	5 × 104	5 × 104	5 × 104	5 × 104	5 × 104	5 × 104	5 × 104

. In order to study the frictional wear and crack initiation under dry friction conditions, the pin-on-disk friction tests were carried out under dry conditions throughout the process, and 9 different sets of tests were carried out on the materials at the rotational speeds of 400 rpm, 800 rpm, and 1200 rpm, and the pressures of 30 N, 60 N, and 90 N, corresponding to the numbering of 1, 2, and 3, respectively, under dry conditions. These tests are designated D4-1-2-3, D8-1-2-3, and D12-1-2-3, respectively.

The temperature rise in the friction area of the test ring was measured during the dry friction experiment using a temperature probe with an accuracy of ±2 °C. No cooling device was used in the contact area, and the test ring was subjected to heat exchange by air convection. Before the experiment, the pin bar and test ring were polished, and the initial mass was recorded. After 50,000 dry friction cycles, the test ring was removed from the tester, and wear was measured.

A torque transducer with a capacity of 20,000 N·mm and a relative error of ±0.2%FS was used to measure the friction torque and calculate the coefficient of friction. At the end of the test, the amount of wear on the test disk was determined by using an electronic balance. A scanning electron microscope (SEM, JEOL, JSM-6610LV, Shanghai, China) and an energy-dispersive spectrometer (EDS) were used to observe the worn surface. Disk sections were cut from the center of the track, then ground, polished, and prepared to observe cracks and deformed microstructures under the wear surface. The surface and cracks in the contact area of the test ring were observed using a three-dimensional microscope with super depth of field (KEYENCE, VHX-6000, Shanghai, China). The number and depth of cracks were recorded for the specimen.

2.3. Finite Element Model

In this paper, the fatigue crack initiation under different sliding velocities and contact loads is determined by finite element analysis. The Maxs Damage and the cohesive unit were used to simulate the crack initiation and extension behavior. By comparing the finite element results with the experimental results, the influence mechanism of crack initiation and extension of 20GrNi2MoV under different working conditions was investigated. The commercial finite element software Abaqus 2021 was used to generate Tyson polygons to simulate the finite element model of 20CrNi2MoV material and phase composition by generating polygons through Voronoi, as shown in Figure 2a. A 20 × 20 mm area was selected as the research object, and the initial average grain size was 18.95 μm. The statistical results are shown in Figure 2b, and the grain size satisfies the normal distribution law.

Figure 3 shows the finite element model developed. This model is a 2D plane model, with the bottom edge fixed and pressures of 200 MPa and 500 MPa applied to the top edge, which are directed perpendicular to the surface towards the interior of the material. By inserting the global cohesive unit between grains, this model simulates the material inter-grain interaction forces and crack initiation. The Maxs Damage is set to be consistent with that in the literature [20], and the normal maximum fracture stress is 2000 MPa.

The finite element model of the critical stress micro-region is set as an elastic–plastic body with a total of 74,810 meshes. The meshes are divided into three categories: CPS4R, COH2D4, and CPS3, with 59,149 quadrilateral CPS4R meshes, 1560 triangular CPS3 meshes, and 14,101 cohesive COH2D4 cells. The bottom edge of the CSM model is set as a fixed boundary.

3. Results

3.1. Friction Coefficient and Wear Loss

The variation in the coefficient of friction with the friction time under the dry friction conditions of the pin disk is shown in Figure 4. From Figure 4, it can be seen that the variation in the friction coefficient under different rotational speeds and pressures is different, but all of them can be divided into two main phases, which are the break-in phase and the steady wear phase [21]. At the beginning of the experiment, the friction coefficient in the break-in phase increases rapidly due to the difference in the initial surface state between the contacting pairs. It is worth noting that from the comparison in Figure 4, it can be seen that the effect of high rotational speed and large load will greatly reduce the time of the first stage, i.e., shorten the break-in stage, so that the coefficient of friction quickly reaches the steady state. The effects of different rotational speeds and contact loads on the coefficient of friction vary. As shown in Figure 4a, when the rotational speed is 400 rpm, the coefficient of friction of the break-in stage increases and then decreases with the increase in contact load. Figure 4b shows that when the rotational speed is increased to 800 rpm, the friction coefficient of the break-in stage gradually decreases with the increase in the contact load, and the change in the friction coefficient of the break-in stage at high rotational speed is opposite to that at low rotational speed; that is, the friction coefficient of the break-in stage decreases and then increases with the increase in the contact load. This may be due to the fact that the newly formed rough peaks between the friction pairs are quickly smoothed out at high speed, and when the contact load increases, the rough peaks cannot be worn out quickly, resulting in an increase in the coefficient of friction [22]. In addition, it can be seen from Figure 4c that the growth of the friction coefficient during the break-in period shows a non-linear variation, where the friction coefficient first increases to the maximum and then slowly decreases to near the stabilized value, which may be caused by the rapid increase in the surface roughness followed by the slow wear of the rough peaks of the contacting pair [23].

In the steady wear stage, the friction coefficients under different load conditions tend to stabilize, which is due to the fact that the surface micropeaks between the friction partners are gradually worn away and the contact surface enters a completely sliding state. The average coefficients of friction at various speeds and loads during the steady wear stage are shown in Figure 4d. It can be seen that the average friction coefficient at 400 rpm is 0.9–1.12 and decreases and then increases with the increase in contact pressure; the contact pressure increases from 30 N to 60 N when the friction peaks between the friction plates are generated faster than the wear rate; when the contact pressure increases to 90N, the wear rate of the rough peaks is higher than the rate of their generation, so the friction coefficient decreases. At 800 rpm and 1200 rpm, the average coefficient of friction is 0.9–1.12. At 800 rpm and 1200 rpm, the average coefficient of friction is between 0.98 and 1.15 and 0.96–1.57, respectively, and decreases as the contact load increases. It is worth noting that the coefficient of friction rapidly increases to 1.57 at high speed and high load of 1200 rpm 90 N. This may be due to the rapid generation of roughness peaks at high speed and heavy load [24].

The cumulative wear losses of the test rings after the dry friction sliding experiments under different working conditions are shown in Figure 5. The cumulative wear of the test rings at 400 rpm ranged from 0.08 to 0.13 g, that of the test rings at 800 rpm ranged from 0.04 to 0.06 g, and that at 1200 rpm ranged from 0.04 to 1.1 g. This indicates that the sliding speed and contact load have a combined effect on the wear resistance of the test rings during the dry friction sliding experiments. This indicates that the sliding speed and contact load have a combined effect on the wear resistance of the test rings during the dry friction sliding experiments. It can be seen from Figure 5 that, in general, the wear of the test ring shows an increasing trend with the increase in load. At a low speed of 400 rpm and a high speed of 1200 rpm, the cumulative wear amount of the test ring gradually increases with the increase in contact load. When the rotational speed is 800 rpm, the cumulative wear amount of the test ring increases and then decreases with the increase in contact load, and this result may be related to the change in friction coefficient in the stable wear stage [21]. It is worth noting that under the dry friction experimental conditions of 600 N at 800 rpm, the wear of the test ring is minimized, which is due to the fact that the tribological performance of the friction pair can be improved to some extent after a certain degree of wear, but the sliding speed and contact load need to be controlled within a certain range to ensure a balance between performance and life [25].

3.2. Surface Damage

In order to further study the friction process and damage mechanism between W2Mo9Cr4VCo8 pin/20GrNi2MoV disks, this section analyzes the surface damage of friction disk specimens under nine different working conditions. The optical microscopy results and 3D scanning results of the specimen surface after 50,000 cycles under different working conditions were analyzed and studied. As shown in Figure 6, the wear surface was analyzed under different operating conditions.

As shown in Figure 6, ring-shaped scratches appear on the surface of the specimen, and the surface of the pin-disk friction area is rough. With the increase in load, the color of the specimen disk surface deepens, which is related to the increase in oxide content on the specimen surface. The wear depth of the friction surface of the specimen disk in (d–f) is greater than that of the specimen in (a–c), and the surface oxide content decreases significantly, which is due to the increase in rotational speed that will accelerate the wear of the oxide layer on the surface of the specimen ring in the course of the pin-disk friction test so that the surface of the specimen shows a metallic luster, but the surface scratches are significantly enhanced in density and depth. With the increase in rotational speed and contact load, the scratches and wear furrows on the surface of the specimen disk gradually increase; in addition, on the disk specimen wear surface appear black agglomerates, which may be caused by the adhesive wear between the friction areas of the pin and disk under the conditions of high speed and heavy pressure.

The optical micrographs (1000×) of the friction contact surface of the specimen disk at the rotational speeds of 400 rpm, 800 rpm, and 1200 rpm and the contact loads of 30 N, 60 N, and 90 N are shown in Figure 7. From Figure 7a–c, it can be seen that at 400 rpm, with the increase in contact pressure, the friction contact surface appears to be adherent metal, and with the friction, the adherent metal is continuously generated and constantly worn. After 50,000 cycles at 400 rpm and 90 N, tiny cracks appear on the surface of the specimen disk, as can be seen in Figure 7d–f. The adhesive wear between the friction partners of the pinned disk increases at the speed of 800 rpm, which may be due to the rapid increase in temperature caused by the increase in speed [26]. Figure 7g shows the optical micrograph of the friction contact area of the disk specimen at 1200 rpm, 30 N for 50,000 cycles, in which it can be observed that after 50,000 cycles, the surface scratches on the specimen disk increase significantly, the surface groove density increase, and at this time, some of the metal debris adheres to the inside and surface of the grooves. Figure 7h shows that when the rotational speed is 1200 rpm and the contact load is increased to 60 N, the surface grooves of the specimen disk are further extended and deepened, and tiny cracks appear on the surface. As shown in Figure 7i, when the contact load is increased to 90 N, the surface of the specimen disk adheres to the debris and forms a large flaky structure.

Figure 8 shows the three-dimensional super-depth-of-field diagram of the wear surface of the specimen disk under different working conditions. From the figure, it can be seen that with the increase in contact load, the height of the surface roughness peaks of the disk specimen increases gradually, and the number of surface roughness peaks increases. In addition, from Figure 8a–c, it can be seen that when the rotational speed is 800 rpm and the contact load is increased from 30 N to 60 N, the depth of the friction surface wear marks on the specimen disk decreases from 4 μm to 3.4 μm and then increases to 12 μm; the three-dimensional surface inspection of Figure c shows that obvious friction furrows appeared on the friction subsurface. From Figure 8d–f, it can be seen that when the rotational speed is 1200 rpm and the contact load is increased from 30 N to 90 N, the depth of abrasion marks on the surface of the specimen disk decreases from 5 μm to 3.6 μm and then increases to 6.8 μm.

Analysis of Figure 8 shows that, when the rotational speed is lower, a large number of furrows are produced on the friction subsurface due to the effect of abrasive wear, and the furrow effect is deepened gradually as the friction is deepened further. However, it is worth noting that when the relative rotational speed between the friction pair increases, the furrows generated on the surface of the material are quickly smoothed out and disappear, and the tiny debris generated during the wear process adheres to the surface of the material pair to form adhesive wear. This change in the wear form may be due to the large amount of frictional heat generated during the high-speed friction process [26].

It can be observed that with the increase in contact load, the rough peaks on the surface of the friction subsurface are gradually smoothed, and new rough peaks are generated, but the density of new rough peaks is not as much as the initial number, and the peak of rough peaks decreases gradually. Figure 8a,c are compared. With the increase in load, the friction subsurface gradually produces a furrow effect. When the rotational speed is increased to 1200 rpm, the friction-generated furrows are gradually smoothed out, and due to the friction process generating a large amount of friction heat, the friction subsurface produces a large number of adhesives, and the wear process changes from the initial abrasive wear to adhesive wear [27].

3.3. Subsurface Damage

In addition to the complex wear changes on the friction subsurface during the friction process of the pinned disk, in this study, SEM inspection was performed after cutting the samples along the direction of the male tangent of the specimen disk scratches after 50,000 cycles of 1200 rpm 60N and 1200 rpm 90N, respectively, as shown in Figure 9a,b.

After analysis, it can be seen that after 50,000 friction cycles, the subsurface of the specimen disk with lower relative hardness has undergone drastic changes. After 1200 rpm 60 N cycles for 50,000 times, a certain depth from the specimen surface produces small cracks. Comparing Figure 9a,b, it can be seen that under the same relative sliding speed, with the increase in contact load, the crack density of the subsurface increases and the cracks show an expansion trend. That is, the cracks gradually expand from small cracks and gradually increase from a single crack to multiple cracks. This may be related to the change in elemental content of the specimen disk during the friction process [28].

On the other hand, SEM and EDS measurements of the specimens after 50,000 cycles at 1200 rpm, 60 N and 90 N, respectively, were also performed in this section, as shown in Figure 10. The corresponding energy spectra of Figure 10a shows that the region is enriched in Fe, Si, Mn, Mo, and C. From the analysis of Figure 10b, it is clear that the 1200 rpm 90 N specimens have the same elemental enrichment. However, it is noteworthy that the Mn content decreases as the load increases. The degree of Mn elemental enrichment in the friction process has a certain influence on the friction performance of the friction substitute [29], and the decrease in Mn content may be one of the causes of the rapid crack initiation after 50,000 cycles at 1200 rpm 60 N.

3.4. Phase Transition in Dry Friction

Figure 11 shows the X-ray diffraction (XRD) spectra of D4-1, D4-2, D4-3, and the original sample. From Figure 11, it can be seen that XRD can only detect austenite and martensite. However, due to the small size and quantity of the carburized body, the diffraction peaks of the carburized body cannot be seen in the spectra. It is worth noting that during the wear process, intense frictional heat and residual stress lead to a martensitic phase change in the organization. The martensite and austenite contents at different contact loads were determined by XRD. In Figure 11, the XRD spectra of the initial specimen and D4-3 after 5000 cycles show austenite peaks in the (225) plane and martensite peaks in the (139) plane. As frictional wear and stress transform austenite to martensite, the number of martensitic phases (γ) increases, and the number of austenitic phases (α) decreases in the material.

The initial austenite content of the specimen disk on Figure 11 is 27.5%. After 50,000 cycles of frictional wear under 30 N, 60 N, and 90 N contact loads, respectively, the austenite (α) residuals are 8.8%, 9.6%, and 12.8%, respectively. The conversion of D4-2 to martensite is much higher in D4-1 compared to D4-3.

From Figure 12, it can be seen that during the wear process, the coupling of contact stress and frictional heat causes the surface of the specimen disk to undergo a martensitic phase transformation, with martensitic peaks at diffraction angles of 44°, 65°, and 82°, and diffraction peaks at diffraction angles of 51° and 76° for austenitic grain diffraction peaks. When the rotational speed is 800 rpm, with the increase in contact load, the martensite peak gradually increases, while the austenite peak gradually disappears. The increase in intensity of the martensite diffraction peaks indicates that the content of martensite in the material has increased, so the martensite phase transformation occurs during the friction of the pinned disk [30].

From Figure 13, it can be seen that the austenite peak appeared in the 42° diffraction peak after 5000 cycles at 1200 rpm under 90N working conditions. In Figure 13, the XRD spectra of the initial specimen and D12-3 after 5000 cycles include an austenite peak in the (225) plane and a martensite peak in the (139) plane. As frictional wear and stress transform austenite to martensite, the number of martensitic phases (γ) increases, and the number of austenitic phases (α) decreases in the material.

3.5. Prediction of Crack Initiation

A localized region on the surface of the inner ring of the bearing with an initial average grain size of 20 μm was established by commercial finite element analysis software Abaqus. The model was set to apply a radial load of 250 MPa to the contact area, and the rotational speeds were 4000 rpm and 12,000 rpm. The effects of different rotational speeds on crack formation under the same load were obtained and are shown in Figure 14.

Analyzing Figure 14 showing the effect of rotational speed on the cracks in the contact area, it can be seen that the maximum stress occurs at the contact position between the pin and the disk friction vice, the stress in the subsurface region is concentrated, and the intergranular stress is larger than the intragranular stress, which also indicates that the crack initiation often tends to occur along the region of grain boundaries instead of inside the grains [31]. A comparison of Figure 14a,d shows that after 5.1 × 10⁵ cycles, two more obvious cracks with dense microcracks appear on the subsurface. In addition, the crack locations are distributed at an angle, and the crack extension region is located in the tangential direction of the load. With the increase in relative velocity, the subsurface cracks emerge and extend to the material surface [32], and the crack density increases. The above analysis shows that under the same load, when the rotational speed is low, the surface of the inner ring of the bearing is not easily cracked. When the relative speed is higher, the crack density is higher.

In order to comparatively analyze the crack initiation mechanism of bearings under different loads, the model is set to the relative speed of the friction contact surface at 12,000 rpm, and the loads of 200 MPa and 500 MPa are applied for comparative analysis. Figure 15 shows the effect of load on crack initiation.

As shown in Figure 15, at the same rotational speed, the internal and surface crack density of the material increases as the load increases. Under high-load conditions, more cracks appear in the subsurface region of the inner ring of the bearing, and the crack initiation rate is non-linear, indicating that after crack initiation, its expansion rate is affected by a combination of factors [33].

The above analysis shows that the effect of load on crack initiation is more significant than the effect of rotational speed on crack initiation and extension, and when the load increases, the rate of crack initiation and extension increases. Moreover, the crack initiation rate under high-speed and heavy-load conditions is faster than that under low-speed and light-load conditions.

4. Discussion

During the dry friction process, a large number of furrow scratches are produced in the friction contact area of the test pin disk under different load and rotational speed conditions, as shown in Figure 7. When the contact load is increased from 30 N to 90 N, the surface wear of the specimen disk deepens, and metal wear debris appears near the scratches and adheres to the surface of the specimen disk, which leads to an increase in the roughness of the contact surface during the fire-adapted frictional contact process, causing the friction coefficient to increase rapidly. At the same time, the friction contact area of the pin-disk friction conditions is higher; the friction temperature rise increases rapidly, increasing the friction area of the sample disk martensitic phase transformation process.

During the friction process, the metal particles shed by abrasion on the friction contact surfaces make the cutting action increase, thus increasing the wear rate of the material. As can be seen in Figure 5, when the rotational speed is increased, the rate of abrasive chips produced between the friction pair is accelerated, and the accumulation time for the formation of adhesion is shortened. As a result, the rate of adhesion generated on the surface of the friction pair is accelerated, which reduces the wear rate of the specimen disk. When the rotational speed is increased to 1200 rpm, the wear rate of metal chips adhering to the surface of the specimen disk is greater than the rate of adhesion due to the accelerated speed, and therefore, the material wear rate increases [28].

SEM images of the surface of D4-3, D8-3, and D12-3 specimen disks are shown in Figure 16. From Figure 16, the following conclusions can be obtained: The contact surface damage of the friction vice of the pinned disk mainly includes surface scratches, shedding, and adhesion. With the increase in temperature, the surface shedding point is when the trend of the first enhancement and then reduction occurs, while the adhesion wear shows a trend of first weakening and then strengthening. The main forms of damage on the surface of the specimen disk include scratches, furrows, surface pitting, and a small amount of abrasive and adhesive wear. And with the increase in temperature, the wear forms on the surface of the specimen disk change from abrasive wear and adhesive wear to a mixture of abrasive wear, adhesive wear, and cracks. In the process of intense friction of the specimen disk material due to contact pressure and impact action, broken and small amounts of metal particles are dislodged and first appear on its surface, freely participating in the friction between the friction surfaces [34].

In addition to the craters and surface scratches formed by the surface detachment, obvious adherent metal is observed on the surface of the disk specimen, which has a good bond strength with the surface of the specimen disk material, protects the matrix tissue, greatly reduces the wear and tear of the substrate, and, to a certain extent, acts as a stabilizer of the coefficient of friction. In the process of sliding friction, a large amount of abrasive debris and frictional heat are generated on the material surface, and as the abrasive debris continues to increase, a bonded layer is eventually formed on the surface of the friction layer under the joint action of friction, pressure, and temperature [35,36]. During the frictional wear process, the bonded material separates the pins of the counter-abrasive material from the disk, which reduces the direct contact between the counter-abrasive materials and also reduces the possibility of adhesive wear of the counter-abrasive materials. During the friction process, the abrasive chips are affected by both friction and load and participate in the formation of adhesive material on the friction surface. In addition, the pressure and impact act simultaneously on the brittle and hard adherent metal and produce cracks, which ultimately lead to pitting corrosion due to the detachment of the surface material [37].

The abrasive particles dislodged from the specimen disk, pin bar, and adhesives end up at the friction surface to participate in friction, cycling the same process throughout the frictional wear process [25]. The metallic cohesive aggregates adhering to the surface of the specimen disk can improve the stability of the sliding process and reduce the wear rate between the specimen disk and the pin bar [22]. As shown in Figure 17, the failure process during the frictional wear of the bobbin is simulated. In this process, the adhesion mechanism between the contact surfaces of the friction pair can be understood. During sliding friction, wear, pitting, and cracking occur on the surface of the less hardened specimen disk due to the interaction of contact pressure and relative sliding. Under conditions of high speed, high pressure, and dry friction, the dislodged particles are subjected to the combined action of various wear mechanisms, repeatedly loaded, compacted, and shaped.

After the formation of metal adhesion on the surface of the specimen disk, a large number of particles are still shed between the interfaces, so the adhesion on the surface of the specimen is constantly cut and worn so that the metal adhesion is constantly broken and shed, and in the continuous compaction and grinding process, the broken particles form new adhesion and finally reach a dynamic equilibrium process [25].

5. Conclusions

In this paper, the influence of different working conditions on the frictional wear and crack emergence of the specimen pin and specimen disk materials in the friction process of the pin and disk is analyzed in terms of the friction coefficient, wear rate, surface damage morphology, and crack emergence. In summary, the analysis results lead to the following main conclusions:

(1)

The coefficient of friction in the break-in stage increases and then decreases with the increase in contact load. The average coefficient of friction at 400 rpm ranges from 0.9 to 1.12 and decreases and then increases with the increase in contact load, and the average coefficient of friction at the speeds of 800 rpm and 1200 rpm increases and then decreases with the contact load.

(2)

With the increase in rotational speed, the wear of the friction partner decreases gradually. The amount of wear increases as the load increases. The cumulative wear of the test ring at 400 rpm ranges from 0.08 to 0.13 g, that of the test ring at 800 rpm ranges from 0.04 to 0.06 g, and that of the test ring at 1200 rpm ranges from 0.04 to 1.1 g.

(3)

After dry sliding friction occurs in the W2Mo9Cr4VCo8 HSS pin/20CrNi2MoV disk friction partner, the depth of wear marks on the specimen disk increases and then decreases with the increase in rotational speed. The depth of wear marks on the surface of the test ring is in the range of 2–4 μm at the rotational speed of 400 rpm, and the depth of wear marks on the surface of the test ring is in the range of 3–12 μm and 2–6.5 μm at the rotational speeds of 800 rpm and 1200 rpm, respectively.

(4)

Due to the interaction of contact pressure and relative sliding, wear, pitting, and cracking occur on the surface of the specimen disk with less hardness. Under the conditions of high speed, high pressure, and dry friction, with the increase in temperature, the wear form of the specimen disk surface goes from abrasive wear and adhesive wear to a mix of three kinds of wear, abrasive wear, adhesive wear, and cracking.

(5)

Grooves and spalls of varying degrees appear in the contact area, and martensitic phase transformation occurs during the frictional wear of the pinned disk.

(6)

The effect of load on crack initiation is more significant than the effect of speed on crack initiation and extension compared to speed, and the rate of crack initiation is faster under high-speed and heavy-load conditions compared to low-speed and light-load conditions.

(7)

The crack location can be predicted using cohesive unit and tensile separation criteria. The crack extension locations predicted by the finite element model are consistent with the experimental results of pin-on-disk friction, which provide a technical reference for the study of crack initiation and extension of bearings with different grain sizes and phase compositions.