Dry Friction Properties of Friction Subsets and Angle Related to Surface Texture of Cemented Carbide by Femtosecond Laser Surface Texturing

Abstract

This paper investigated the use of laser surface texturing (LST) to improve the tribological properties of YG6X cemented carbide. Three different spaced groove textures were processed on the surface of the YG6X carbide samples using a femtosecond laser. Friction experiments and friction simulations were performed under two friction subsets and two friction directions. The testing results showed that when the area density was 46%, the texture surface was beneficial when sliding against Si₃N₄, but not beneficial in reducing the coefficient of friction when sliding against Ti6Al4V titanium alloy. At area densities of 23% and 15.3%, the texture surface was beneficial when sliding against Si₃N₄, but not beneficial when sliding against the Ti6Al4V titanium alloy. When selecting the friction direction at 45° to the area density of 15.3%, the texture surface was not beneficial when sliding against the Si₃N₄ and Ti6Al4V titanium alloy. Sliding with Si₃N₄, the higher the stress value, the more easily the material was destroyed, leading to an elevated coefficient of friction and wear area. Sliding with Ti6Al4V titanium alloy, the higher the stress value of Ti6Al4V titanium alloy, the more easily the Ti6Al4V titanium alloy wore and generated a large number of abrasive chips.

1. Introduction

Titanium alloys have been widely used in the aerospace and biomedical industries due to their high specific strength, corrosion resistance, biocompatibility, and wide operating temperature range [1,2]. However, because of the low thermal conductivity of titanium alloys, the heat generated in the process of sliding with other materials cannot be effectively diffused, which has an adverse effect on the friction material. Second, titanium alloys weld easily with the frictional material during friction due to its high chemical reactivity, resulting in the premature failure of the frictional material [3]. Cemented carbide, mainly used for tool materials but also used as drawing dies and wear-resistant parts, is an alloy material made of hard compounds of refractory metals and viscous metals through powder metallurgy. Commonly used cemented carbides are divided into three types: tungsten-cobalt (YG), tungsten-cobalt-titanium (YT), and tungsten-titanium-tantalum (niobium) (YW). Among them, YG-type cemented carbides have good red hardness, high flexural strength, good thermal conductivity, and poor affinity with titanium alloys, which makes it suitable for processing titanium alloys, and its tribological properties are the key factor affecting its service life.

Surface texturing has remained an attractive area in recent years, which can further improve the wear resistance of materials [4]. There are various methods of micro-texturing such as electrical discharge machining (EDM) [5], lithography [6], micro-milling, and laser texturing (LST) [7]. EDM allows for the micro-fabrication of complex surfaces, but it can only machine conductive materials and has a slow machining speed. Lithography can obtain nanometer-level processing accuracy, however, it is not suitable for single-piece processing due to its high processing cost. Micro-milling can be used for the fine 3D surface machining of materials, but the size of the textures is generally in the micron level due to the size limitation of the milling cutter. Compared to other surface texturing techniques, laser texturing, especially femtosecond laser texturing, offers a smaller heat-affected zone, a precise processing area, flexible processing methods and high processing accuracy [8], and have been widely used for the surface texturing of metallic materials to improve their tribological properties.

A number of studies have found that texture morphology, geometric parameters, distribution, and the friction lubrication state have a significant effect on the performance of the texture. Fang et al. [9,10,11] presented different surface patterns with two different geometric structures that were processed on the contact surface of cemented carbide. The results showed that the linear pattern could have higher friction, while the pit pattern could effectively reduce the friction. Lian et al. [2] believed that a pit texture had better frictional properties than a groove texture, but that the pit texture could reduce the adhesion of titanium alloy under the condition using a solid lubricant. Mikhail et al. [12] found that the efficiency of tool turning of an Fe-Ni alloy could be improved by machining surface texturing on the cutting tool surface. Liu et al. [13] found that because the adhesion of the workpiece material on the surface of tools was obviously reduced, the cutting performance of micro-textured fluted carbide tools could be improved during the turning process of a Ti6Al4V titanium alloy. He et al. [14] studied the effect of circular pit, square pit, and ring groove micro-textured surfaces on the tribological properties of cemented carbide by the finite element simulation method and experiment. The results showed that the circular pit surface had a better friction performance compared to other types of surface micro-textures and smooth surfaces. Dan et al. [15] used a femtosecond laser to process four kinds of V-grooves on the surface of cemented carbide, with texture densities of 5.05%, 9.5%, 13.02%, and 15.2%, respectively. It was found that the surface friction coefficient of the sample with a texture density of 9.5% was the smallest and most stable. Zhang et al. [16] analyzed the tribological performance of YG8 with four types of groove-shaped textures under dry friction and solid lubricant, and corroborated that reasonable dimensional characteristics and distribution could reduce friction. Meng et al. [17] researched the dry machining of austenitic steels with cemented carbide tools and found that the groove texture captured wear debris under dry sliding conditions, thus reducing wear on the carbide. The beneficial effect of micro-scale grooves becomes higher with increasing groove density. Tong et al. [18] prepared micro-textures with variable distribution density on the carbide tool surface for high-speed milling tests with a titanium alloy and analyzed the effect of micro-texture parameters on tool wear, showing that the most determining factors were the diameter and spacing of the micro-textures. In the above study, it can be found that most of the studies considered only one kind of friction substrate and friction direction, but the process of friction is complex, and it is difficult to select only one kind of friction substrate [19] and friction direction [20] for an in-depth study of the actual mechanism of wear reduction.

This study prepared a variety of groove-textures on the surface of cemented carbide using a femtosecond laser, and the effects of texture spacing, friction balls, and friction directions on the tribological behavior of cemented carbide were analyzed to explore the influence mechanism of wear.

2. Experiment

2.1. Materials and Laser Surface Textured

In this experiment, silicon nitride ((Si₃N₄) balls, Ti6Al4V (TC4)) titanium alloy balls, and YG6X cemented carbide were used. The diameter of the balls was 6 mm, and the size of the bulk cemented carbide was 10 mm × 10 mm × 4 mm. The main components of YG6X cemented carbide are tungsten carbide (WC) and cobalt (Co). Before the micro-nano processing of the alloy surface, the surface of the YG6X alloy was abraded using 2000-mesh sandpaper, and then polished into mirror using diamond abrasive paste. The surface textures were processed by the femtosecond laser system (FemtoYL-50, Anyang Laser Technology Co., Ltd., Anyang, China). Grooves were created on the samples using a femtosecond laser with a maximum average power of 88.3 W, a pulse width of 362 (fs), a wavelength of 1035.43 nm, and a spot size of 73 μm. The laser processing parameters were optimized by regulating the frequency, speed, and processing times to obtain straight grooves and reduce the influence of the groove edge effect. The orthogonal test parameters are listed in

Table 1. Laser texture parameters.

Group	Frequency (kHz)	Speed (mm/s)	Processing Times (Times)
T1	25	125	1
T2	25	250	3
T3	25	500	5
T4	25	1000	7
T5	100	125	3
T6	100	250	1
T7	100	500	7
T8	100	1000	5
T9	300	125	5
T10	300	250	7
T11	300	500	1
T12	300	1000	3
T13	800	125	7
T14	800	250	5
T15	800	500	3
T16	800	1000	1

. The groove profiles were observed by an optical microscope (GX51+DP26, Olympus, Tokyo, Japan). Under the optimized process parameters, groove-textures with different spacing were machined on the YG6X alloy, and their texture distribution is shown in Figure 1. We set the texture spacing (L1) to 0.25 mm, 0.5 mm, and 0.75 mm. L2 was the width of the machined grooves. The samples corresponding to these three texture spacings were named G1, G2, and G3 in that order. The untextured sample was named G0. Samples were ultrasonically cleaned with absolute ethanol before and after laser treatment, and then dried with compressed air.

2.2. Characterization of Mechanical Properties

A hardness tester (HV-1000, AICEYI Opto-electronic Technology Co., Ltd., Shenzhen, China) was used to measure the hardness of all samples. The detection force was 0.2 N, the loading time was 10 s, and four points were measured for each sample and the average value was taken. Tribological investigations were performed using a friction and wear testing machine (HSR-2M, Lanzhou Zhongke Kaihua Technology Development Co., Ltd., Lanzhou, China) with a ball-on-disc contact configuration adopted under unlubricated ambient reciprocating sliding conditions. The silicon nitride balls, TC4 titanium alloy balls, and YG6X cemented carbide were subjected to friction experiments. The loading force was 50 N, the friction rate was 66.67 mm/s, and the diameter was 4 mm. The friction test lasted for 30 min with 15,000 reciprocations. The samples before and after the friction experiment were ultrasonically cleaned with absolute ethanol for 5 min. The morphology of the abrasion marks was observed by laser scanning confocal microscopy (LSCM) (OLS5000, Olympus, Tokyo, Japan) and scanning electron microscopy (SEM) (SUPRA 40, Zeiss, Oberkochen, Germany).

2.3. Finite Element Analysis of the Tribological Behavior of Groove-Textured YG6X Alloy

The tool was replaced by a cemented carbide, and the workpiece was replaced by a titanium alloy ball. The ball-block test model was established, and the finite element analysis of the ball-block test was carried out, and the stress distribution diagram was obtained. Here, the volume of the YG6X and TC4 simulation models was consistent with the actual volume of each material in the friction experiment, the groove width was set to 115 μm, and the groove width was set to 40 μm. The TC4 simulation model used a C3D4T four-node thermally coupled cell, and the YG6X simulation model used a C3D8RT eight-node thermally coupled hexahedral cell using reduced integration. The element size of the specimen and the ball were 0.4 and 0.3, respectively. The YG6X simulation model kept the bottom surface fixed and defined the TC4 simulation model’s pressure and velocity. Pure elasticity calculation was conducted using an explicit formula. Their computations were made under the purely elastic deformation assumption and stresses may have be overestimated. The coefficient of friction was 0.4 and was obtained from friction experiments with TC4 by non-textured carbide. The focus of this study was to explore the stress on the contact surface during the friction process, the simulation of different grinding ball materials to ensure that the contact stress had the same law, so only the TC4 and YG6X friction pairs were selected for simulation. The simulation models of YG6X cemented carbide with different textured spacing are shown in Figure 2. F denoted the applied load, R denoted the radius of the grinding ball, and V denoted the sliding speed of the grinding ball.

The finite element simulation of the friction process was carried out using Abaqus software. The material properties are shown in

Table 2. Material properties of the block and ball samples.

Materials Name Material Properties	Block Sample (YG6X)	Ball Sample (Ti6Al4V)
Density (kg/m3)	1.485 × 1020	4.43 × 1021
Modulus of elasticity (GPa)	640	114
Poisson’s ratio	0.22	0.33
Conductivity (W/m·k)	79.6	6.6
Thermal expansion coefficient (×10−6/°C)	4.7	9
Specific heat capacity (J/(kg·°C))	176	670

. The temperature-displacement coupling analysis step was established in the analysis step module, and the total time was set to 0.0042 s. The whole process was set up so that the friction ball moved from one end of the groove textures to the other and moved a small distance over the original carbide surface. Then, the YG6X was fixed and restrained, and a pressure of 1.67 MPa was applied to the surface of the titanium alloy hemisphere. Finally, a velocity/angular velocity analysis type was established for TC4, and the velocity was set to 66.67 mm/s according to this analysis step. Previous work by numerous researchers has shown that sliding orientation plays a vital role in tribological performance. We selected a sample according to the simulation results to study the effect of a friction direction of 45° on wear resistance.

3. Results and Discussion

3.1. Optimization of Laser Processing Parameters

A series of groove textures prepared with different frequencies, speeds, and processing times are shown in Figure 3.

The grooves shown in Figure 3b were straight with a clear edge shape, while the grooves shown in Figure 3a were straight with a blurred edge shape, which was a problem of formability. The grooves shown in Figure 3c,d were not straight with a blurred edge shape. The optimized laser texturing parameter corresponding (100 kHz, 500 mm/s, processed 7 times) to the T7 groove was selected here due to the straight groove with a clear edge. The average width value of the T7 groove was 115 μm and the depth of the T7 groove was 4–5 μm, as shown in Figure 4a. The hardness value of 1903.4 HV_0.2 was obtained near the edge of the groove and 1800.0 HV_0.2 was obtained away from the edge of the groove, respectively, indicating that there was an increase in hardness at the edge of the groove during the laser texturing. This was considered here as laser processing creates a heat-affected zone in the material, melting and regenerating and creating a hard layer of the material at the edges of the grooves [21]. Figure 4c displays the morphology of the T7 groove. It could be seen that after femtosecond laser processing, there was no obvious convex structure on the edge of the groove and many micro/nano particle structures were formed inside the trench.

In the process of groove texturing, there was a problem of groove formability, so the morphological values were quoted here to analyze this orthogonal test. The evaluation of the shape value was divided into two parts: the groove textures were straight and the edge of the groove textures was clear. If one of them was met, the value was assigned to 1. If both were met, the value was assigned as 2. The morphological values corresponding to the orthogonal tests are collated in

Table 3. Shape values (the evaluation of the shape value was divided into two parts: the groove textures were straight and the edge of the groove textures was clear; the value 1 was assigned if one criteria was met, and the value 2 was assigned if both criteria were met).

Group	T1	T2	T3	T4	T5	T6	T7	T8	T9	T10	T11	T12	T14	T15	T16
Shape values	1	1	1	1	2	1	2	2	1	1	1	1	1	1	1

. The experimental results were analyzed visually using the analysis of range differences to obtain the mean response table listed in

Table 4. Analysis of range.

	Level	Frequency	Speed	Processing Times
K average value	1 time	-	-	1.00
	3 times	-	-	1.25
	5 times	-	-	1.25
	7 times	-	-	1.25
	25 kHz	1.00	-	-
	100 kHz	1.75	-	-
	300 kHz	1.00	-	-
	800 kHz	1.00	-	-
	125 mm/s	-	1.25	-
	250 mm/s	-	1.00	-
	500 mm/s	-	1.25	-
	1000 mm/s	-	1.25	-
R		0.75	0.25	0.25

. The K average value represents the average value of the morphology value at each level of each factor, reflecting the influence of different levels of the same factor on the test results. R is equal to the maximum value of the K average value minus the minimum value, reflecting the magnitude of change in the test index when the factor level is changed. As shown in Table 4, the range in frequency (0.75) was greater than the range in the scanning speed (0.25) and processing times (0.25).

It can be concluded that the laser processing frequency had the dominant influence on the morphological values of the groove textures and it was easier to obtain groove textures that met the requirements at 100 kHz. Afterward, if further study of other parameters of the grooves is needed, the frequency can be fixed at 100 kHz, thus adjusting the speed and the processing times to achieve it.

3.2. Friction Simulation Result

Stress is an important factor causing surface wear during sliding, and surface texture affects stress distribution [3]. In order to study the effect of groove textures on the stress distribution of the samples, ABAQUS software (version 6.14) as used to simulate the stress field in the friction process between YG6X cemented carbide with different texture spacing and the TC4 titanium alloy. The simulation results of the stress field on the YG6X surfaces when the sliding direction was perpendicular to the groove textures are shown in Figure 5a–d. These simples were named G1-90°, G2-90°, and G3-90°. It was found that the stress and its distribution could be changed by preparing the groove texture on the surface.

The simulation results of the stress field on the surface of TC4 are shown in Figure 6a–d. The maximum stress of TC4 during the friction process also changed continuously, as shown in Figure 6e. The relationship between the areal density and the average stress of G-90°, G2-90°, and G3-90° is shown in Figure 6f. The average stress value first increased and then decreased with the decrease in areal density. The average stress value of G1-90°, with a 46% areal density, was the minimum. The larger the stress of TC4, the easier TC4 was removed, resulting in G2-90° and G3-90° having more wear debris than G1-90°, so G1-90° had the best wear resistance.

Subsequently, G3 was selected to research the influence of sliding orientation between the groove and grinding ball on the stress field of YG6X and TC4. When the sliding orientation was 45°, the simulation results of stress field of the groove-textured YG6X and TC4 ball can be seen as exhibited in Figure 7a,b, respectively. This sample was named G3-45°. The stress value during the friction process also changed continuously, as shown in Figure 7c,d, and the average stress values of YG6X and TC4 were 444.04 MPa and 286.15 MPa, respectively. Compared with the results of G3-90°, the average stress value of the groove-textured YG6X and TC4 ball both increased when the sliding orientation was 45°.

3.3. Friction Characteristics

In order to explore the friction and wear behavior of the textured samples under dry friction, using Si₃N₄ and TC4 as the sliding ball, friction and wear tests were carried out when the sliding direction was perpendicular to the groove textures. Figure 8a displays the coefficient of friction (COF) curves of samples sliding with the Si₃N₄ ball when the sliding direction was perpendicular to the groove textures. The COF is the ratio between the tangential and the normal load Ft/Fn. The COF of all samples initially increased, then remained stable with increasing time, and the average value of COF was G2-90° < G3-90° < G0 < G1-90°. The results demonstrate that the groove textures had the potential for reducing the friction coefficients under dry friction. Figure 8b,c demonstrates the wear cross section and the wear cross-sectional area of YG6X with different texture spacings, respectively. The wear area of G1-90° (8508.480 μm²) was the largest, which was increased by 38.17% compared to G0 (6157.948 μm²). However, the wear area of G2-90° (5510.097 μm²) and G3-90° (5858.555 μm²) were separately decreased by 10.52% and 4.86% compared to G0.

The SEM and LSCM morphologies of the worn YG6X surface, sliding with Si₃N₄, are shown in Figure 9. The results showed that the surface of YG6X started to peel off after several passes of low-speed reciprocating friction under high load and had the same worn-out appearance. This was mainly a tearing and peeling phenomenon, but the surface peeling phenomenon was more obvious in G1-90° and less so in G2-90° and G3-90°. This was consistent with the previous simulation results. During the wear process, by the action of pressure and shear force, G1-90°, with a high stress concentration value, was prone to sprout micro-cracks, and crack expansion led to the occurrence of spalling. Because of the wear debris-trapped effect [22], as shown in Figure 10, the depth of the wear scars of G2-90° and G3-90° was shallower than that of G0 because the groove texture with a certain surface density could effectively store the abrasive particles generated during the wear process and reduce the grinding on the surface of the material. When sliding with the Si₃N₄ ball and a sliding direction perpendicular to the groove textures, the groove textures, whose area densities were 15.3% and 23%, could obviously decrease the wear resistance of the material surface, while the groove texture improved the wear resistance of the material surface when the area density was 46%.

Figure 11a displays the COF curves of the samples sliding with the TC4 ball when the sliding direction was perpendicular to the groove textures. The average value of COF was G2-90° < G1-90° < G3-90° < G0. The results demonstrated that the groove textures had a potential for increasing the COF between YG6X and TC4 under dry friction. Figure 11b,c demonstrate the wear cross section and the wear cross-sectional area of YG6X under sliding with TC4 at different texture spacing conditions, respectively. The wear area of G2-90° (455,536.012 μm²) and G3-90° (34,781.901 μm²) were separately increased by 39.64% and 6.66% compared to G0 (32,610.684 μm²). The wear area of G1-90° (30,362.702 μm²) was decreased by 6.90% compared to G0. When sliding with the TC4 ball and a sliding direction perpendicular to the groove textures, the groove textures, whose area densities were 15.3% and 23%, could obviously improve the wear resistance of the material surface, while the groove texture decreased the wear resistance of the material surface when the area density was 46%.

The SEM and LSCM morphologies of the worn surface on the YG6X sliding with TC4, are shown in Figure 12. These mainly showed tearing and spalling phenomena, which was typical of adhesive wear characteristics. From the aforementioned simulation results, it could be seen that when the surface area density was 23% (G2), the stress concentration on the surface of TC4 ball was the largest, which might be susceptible to sprouting micro-cracks and crack expansion and cause the occurrence of spalling by the action of pressure and shear force in the sliding process. In addition, the soft and tough TC4 wear chips could not be collected effectively due to the lower area density and adhered to the surface of YG6X, producing an obvious tearing phenomenon.

From the results above-mentioned, G3 had better friction properties when sliding with Si₃N₄ and TC4 balls under a sliding direction perpendicular to the groove textures. Subsequently, the effect of sliding orientation on the friction and wear behavior of G3-90° was studied. Figure 13 shows the friction performance of G3 when it slid with Si₃N₄ under the sliding orientation of 45°. At the beginning of the friction, the COF increased more compared to G3-90°, as shown in Figure 13a. The wear area was 8481.196 μm², greater than G3-90° (5858.555 μm²), as shown in Figure 13b. The YG6X surface had more serious peeling and produced multiple deep pits compared to G3-90°, as shown in Figure 13c. Figure 14 shows the friction performance of G3 when it slid with TC4 under the sliding orientations of 45° and 90°. The COF of G3-45° had an increasing trend to be greater than that of G3, as shown in Figure 14a. The wear area was 34,884.417 μm² greater than G3-90° (34,781.901 μm²), as shown in Figure 14b. Figure 14c also showed a large number of TC4 abrasive chips adhered on the YG6X surface. It was also concluded that the friction properties of the textured samples did not improve with this friction orientation, but rather increased the wear of the material surface.

4. Conclusions

Three different spacing groove textures were machined on the YG6X cemented carbide by a femtosecond laser. The effect of the groove texture spacing, friction subsets, and friction directions on friction wear was analyzed by ball-block reciprocating friction simulations and tests. The main conclusions obtained can be summarized as follows:

(1)

Using the optimized laser texturing parameter corresponding to 100 kHz, 500 mm/s, processed 7 times could obtain a straight groove with a clear edge on the surface of the YG6X carbide.

(2)

Sliding with hard material Si₃N₄ and the sliding direction perpendicular to the groove textures, when the area densities of the groove textures were 15.3% and 23%, the groove textures could reduce the friction coefficient and wear area of cemented carbide at the same time, while the friction coefficient and wear area of the YG6X cemented carbide increased when the area density of the groove texture was 46% compared with the smooth surface. When the area density of the groove texture was 23%, it reduced the wear resistance, changing the direction of friction to 45° compared with the original.

(3)

Sliding with the TC4 titanium alloy and a sliding direction perpendicular to the groove textures when the area densities of the groove textures were 15.3% and 23% showed that the groove textures could increase the friction coefficient and wear area of cemented carbide at the same time, while the groove textures could reduce the wear area of cemented carbide when the area density of the groove texture was 46% compared with the smooth surface. When the area densities of the groove texture were 23%, it reduced the wear resistance, changing the direction of friction to 45° compare with the original.

(4)

The simulation results were consistent with the experimental results. For the texture samples, the higher the average stress, the easier it was to wear.