Temperature Rise at the sliding Interface between a Carbon Steel and DLC Film

2016 STLE Annual Meeting & Exhibition May 15-19, 2016 Bally s Las Vegas Hotel and Casino Las Vegas, Nevada, USA Temperature Rise at the sliding Interface between a Carbon Steel and DLC Film Shuji YAMAMOTO Sankei Giken Kogyo Co., Ltd, Tokyo, Japan Abstract: Temperature rise at the sliding interface between carbon steel (SUJ2 ASTM E52100) and diamond-like carbon (DLC) was simulated using the friction energy and the estimated real contact area. The friction energy was measured by the tribometer. The heat, wear, strain, plastic deformation and chemical reaction energies of which the friction energy consists were quantified and most of the friction energy was transformed into heat. The temperature rise simulation using the friction energy was compared to the experiment data and good accordance of both data was demonstrated. The real contact area was estimated by temperature indication paint method. The poor wear resistance of the carbon steel against the DLC is caused due to the hardness degradation induced by the high temperature rise. Category: Tribotesting Key words: temperature rise, friction energy, real contact area 1. INTRODUCTION It is important to consider the tribological behaviour based on the temperature rise at the contact interface during sliding because of the hardness degradation of the sliding materials [1] or the formation of a tribolayer. However, friction heat is still poorly understood and has been investigated for years. The heat simulation is a useful and convenient tool to estimate the temperature rise at the rubbing surfaces. The friction energy generated between the sliding surfaces is one of the most important parameters for simulating the temperature rise. Yamamoto et al. [2] introduced the friction energy measurement method using the frictional coefficient - sliding distance chart. The energy consumption rate for the friction heat needs to be identified for analysing the temperature rise. In this study, the tribological behaviours of the DLCs against an alumina and the steels against the DLC were investigated. The wear, elastic strain, plastic deformation and chemical reaction energy were evaluated. The energy consumption rate for the friction heat was estimated by subtracting the wear and the strain energies from the friction energy. The real contact area is required to analyse the temperature distribution around the sliding surface. The real contact area was estimated by the temperature indication paint method. The tribological behaviour of SUJ2 was discussed based on the temperature rise simulated using the friction energy and the real contact area. 2. EXPERIMENT The DLCs were deposited onto WC-9%Co alloy by rf plasma CVD for the wear tests and onto single crystal silicon wafers for evaluating the thickness, hardness, Young s modulus, and the surface energy. The tribological behaviour of the DLC and the nitrogenated DLCs against an alumina ball with size of Φ 4.8 mm was evaluated by the ball-on-disk because the alumina ball is not worn with respect to the DLCs, by which can be assumed the wear energy was consumed only to the DLCs. A sliding velocity of 100 mm/s, and a load of 9.8 N under lubricant-free condition were used for all the tests. For the test environment, the temperature was approximately 20 degrees C and the humidity was approximately 10 ~ 20%. The friction energy was calculated using the friction force recorded in the tribometer [3]. The wear volumes of the DLCs were measured by a laser microscope. The wear volume of the steel ball was obtained using the measured diameter of the wear scar. The surface energy of the DLCs on Si wafers was measured by the micro Vickers indentation method [4]. The wear particles size distribution of the DLCs was evaluated by image processing to estimate the surface area of the wear particles [5]. The temperature rise was measured by the ϕ 80 μm thermocouple at 0.3 mm above the contact point of the steel ball. The 3D model of the steel ball, of which the number of nodes and elements were 4870 and 1209, respectively, was simulated with a different friction energy in watts.

3. RESULTS AND DISCUSSION The energy consumption modes are classified as 1) wear energy: E w, 2) heat energy: E h, 3) elastic strain energy: E e, 4) plastic deformation energy: E p and 5) chemical reaction energy: E c. From this, the friction coefficient may be described as: µ = E NL = E + E + E + E + E NL = μ + μ + μ + μ + μ (1) where µ w is the wear factor, µ h is the heat factor, µ e is the elastic strain factor, µ p is the plastic deformation factor, µ c is the chemical reaction factor [6]. The wear energy was estimated by a product of the surface energy and the newly created surface area. Table 1 shows the wear energy of the DLC of which the surface energy was 274 J/m 2 obtained by the indentation test. Table 1 Wear energy of the DLC Friction energy (J) Surface area of particles (m 2 ) Wear energy (mj) 126 1.17 10-6 0.32 177 1.80 10-6 0.50 235 3.47 10-5 0.97 The estimation results indicate that the consumed wear energy was approximately 10-5 times smaller than the corresponding the friction energy. The wear coefficients of the wear volume-friction energy equation [1] for the DLCs obtained from the wear volume-friction energy chart were approximately 10-5. The index number of the wear coefficients was almost the same as the wear energy/friction energy ratio obtained by the surface energy approach. This accordance affirms that the wear coefficient of the equation indicates the energy consumption rate of the wear to the friction energy. Figure 1 shows the wear coefficient of the SUJ2 against the DLC. The wear coefficient of the steel was approximately 10-7 which indicated wear energy consumption rate to the friction energy for the steel. The strain energy in the ball induced by the applied normal was estimated less than mj. The plastic deformation energy of the steel was estimated using the stress-strain curve. The energy ratio to the corresponding friction energy was less than ppm. The chemical reaction energy was less than 1% of this total and is excluded from the energy consumption totals because the chemical reaction is not endothermic. To clarify, most of the friction energy goes towards the generation of heat. Table 2 Energy consumption rate in each mode Energy mode Consumption rate Wear: E w less than ppm Elastic strain: E e less than ppm Plastic deformation: E p less than ppm Heat: E h more than 99% The temperature distribution in the steel balls were simulated by ANSYS using the friction energy measured by the ball-on-disk test. Figure 2 shows the maximum temperature dependence obtained by the simulation on the friction energy of the SUJ2 with various contact diameters. The maximum temperature was proportional to the friction energy and inversely proportional to the diameter of the contact area. Fig. 2 Tempurature rise simuration results on the sliding surface of SUJ2 [3] Fig.1 Dependence of the wear coefficient of SUJ2 on the friction energy Figure 3 shows the experimental data at 0.3 mm above the contact area, ANSYS data at the same location and ANSYS data calibrated by the heat partition rate of the steel. The calibration data of steel was in good agreement with the experiment data. The temperature rise with considering the real contact area estimated using the temperature indication paint method was simulated for SUJ2 and SUS440C as shown in Fig.4. The temperature

REFERENCES Fig.3 Comparison between calibrated ANSYS data and experiment data [3] rise of SUJ2 was approximately 240 at sliding velocity of 100 mm/sec while that of SUS440c was approximately 50. The poorer wear resistance of SUJ2 than SUS440C was caused due to the hardness degradation induced by the higher temperature rise on the sliding surface. Fig.4 Temperature rise simulation results of SUJ2 and SUS440C with considering real contact area [1] S.Yamamoto: Tribological Mechanisms of Steels with respects to Diamond-like Carbon in terms of Energy Input, Tribol. Trans., 2014, 57, 1001-1006 [2] S.Yamamoto, A.Kawana and C.Masuda: Tribological behaviour of stainless steel with respect to that of DLC in terms of energetic aspects, Tribology Materials, Surfaces & Interfaces, 2013, 7, (4), 161-167 [3] S. Yamamoto, T. Okuaki, M. Egashira, K. Kondoh, and C. Masuda: Evaluation of the temperature distribution in steel balls induced by friction generated during the tribo-test against diamond-like carbon coatings, Tribology Materials, Surfaces & Interfaces, 2015, 9, (1), 33-40 [4] Nastasi, M., Kodali, P., Walter, K. C., Embury, J. D., Raj, R., Nakamura, Y.: Fracture toughness od diamond-like carbon coatings. J. Mater. Res. 14, 2173-2180 (1999) [5] S. Yamamoto, M. Egashira, K. Kondoh, and C. Masuda: Evaluation of the wear energy consumption of nitrogenated diamond-like carbon against alumina, Tribol. Lett., 2014, 55, 279-288 [6] S. Yamamoto, M. Egashira, K. Kondoh, and C. Masuda: Quantification of the wear, strain and heat energy consumption rates for sliding steel ball against diamond-like carbon coatings, Tribology Materials, Surfaces & Interfaces, 2015, 9, (2), 71-76 4. CONCLUSIONS The most of the friction energy goes towards the generation of heat. The temperature rise simulation on the sliding surface proved the poor wear resistance of SUJ2 was caused due to the higher temperature rise. 5. ACKNOLEGEMENTS The author wishes to thank Dr. Chitoshi Masuda (retired Waseda University) for giving the opportunity to study tribology and thanks Mr. Egashira (retired National Institute for Materials Science) for helpful discussions of this work.

Temperature Rise at the Sliding Interface Between a Carbon Steel and DLC Film 2016.5.18 Sankei Giken Kogyo Co., LTD. Shuji Yamamoto μ

Objective Propose the evaluation method of the tribological properties in terms of the friction energy Breakdown the friction energy into heat, wear, elastic, plastic and chemical energies Investigate the tribological behavior of the steels against the DLC based on temperature rise µ 2

Outline 1. What is friction energy? 2. Breakdown the friction energy generated by sliding between steels and DLCs wear energy elastic energy plastic deformation energy chemical reaction energy heat energy 3. Temperature rise simulation in the steel ball 4. Energetic analysis of the wear behavior of the steels against DLC µ 3 5. Conclusions

Tribological behaviors of steels and alumina against the DLC on ball-on-disk Ball-on-disk test (ball size 4.8 mm) Load: 9.8 N Sliding velocity: 100 mm/s Sliding distance: 500 m Temp. 20~25, Humidity 10~20% μ Fig.1 Friction coefficient of SUJ2, 440C and alumina against the DLC SUJ2 440C Alumina Fig.2 Wear scars of the balls against DLC after tribotest Table 1 Ball materials: Ball ASTM Material Hardness (GPa) SUJ2 E52100 Carbon steel 8.3 SUS440C 440C Stainless steel 8.3 Alumina Al 2 O 3 14 4

1. What is friction energy? 2. Breakdown the friction energy generated by sliding between steels and DLCs wear energy elastic energy plastic deformation energy chemical reaction energy heat energy 3. Temperature rise simulation in the steel ball 4. Energetic analysis of the wear behavior of the steels against DLC 5. Conclusions µ 5

1. What is friction energy? Objective Introduce the measurement method of the friction energy during the tribotest. Derive the wear volume friction energy equation from Holm-Archard equation. μ 6

Measurement method of friction energy Friction energy E f (joule) =Friction force distance = F i l i = μndl = N μ dl i F: friction force, μ: frictional coefficient, N: load, l: sliding distance μ Fig.3 Friction coefficient sliding distance Friction energy is calculated using recorded data of the tribometer Friction energy is provided from the motor of the tribometer Data recorded during tribotest Time (s) Friction Force Sampling (mn) distance(m) F i l i 0 0 0.01 0 0.1 2.38 0.01 0..0238 0,2 1.54 0.01 0.0154 499.9 1.05 0.01 0.0105 500 1.52 0.01 0.0152 E= F(i) l

Wear volume Friction energy equation Holm-Archard equation V = k NL σ V: wear volume, k: wear coefficient, N: load, L: sliding distance, σ: hardness Dimension analysis for Holm-Archard equation V L = k NL = k (L N ) L k is dimensionless Hardness is pressure unit k can be described as μk because μ is dimensionless μ Wear volume - Friction energy equation μnl V = k σ = k E f σ 8

Friction coefficient based on friction energy μ Fig.4 Friction coefficient sliding distance chart μ ave = μds L = N μds NL L: total sliding distance = E f NL

Section 1: Conclusions Friction coefficient and wear volume are described in terms of the friction energy V = k E σ μ = E NL μ 10

2. Breakdown the friction energy generated by sliding between steels and DLCs Objective Quantify each consumption energy in the friction energy 1. Heat energy 2. Wear energy 3. Elastic strain energy 4. Plastic deformation energy 5. Chemical reaction energy μ 12

Energy balance in sliding bodies Wear energy Plastic deformation energy Chemical reaction energy Elastic strain energy Steel Heat energy DLC WC-9%Co μ 13 Fig.5 Energy balance of friction energy

Specimens: Nitrogenated DLCs 1. DLC C 6 H 6 [Benzene] 2. N 5.4 DLC N: 5.4 wt% 3. N 10.7 DLC N: 10.7 wt% for wear energy evaluation Specimen Table 2 Properties of DLCs Thickness (μm) Hardness (GPa) Young s modulus (GPa) DCL 0.49 29.2 263 N 5.4 DLC 0.59 24.7 221 N 10.7 DLC 1.1 20.1 173 μ 14

Evaluation of wear energy for DLCs Wear energy = net surface area of the wear particles surface energy Evaluate surface energy Specimens were aligned on the micro-vickers table to form the radical cracks along the < 011> of the silicon substrates crack Indentation mark 500 g 500 300 g 1000 200 g 1000 100 g 2000 Fig. 6 SEM images of the cracks generated from edges of indentation mark μ 16

Evaluation of surface energy for DLCs Table 3 Surface energy of nitrogenated DLCs Film Young s modulus (GPa) Si (substrate) 180 gradient γ (J/m 2 ) DLC 263 5.66 x 10-5 274 N 5.4 DLC 221 1.02 x 10-5 410 N 10.7 DLC 173 5.66 x 10-5 374 μ Equation for fracture energy using the crack length G f = 2γ Surface energy C 0 : crack length of Si C : crack length for NDLC d : DLC thickness Gs : Fracture energy of Si Gf : Fracture energy of NDLC Es : Young s modulus of Si Ef : Young s modulus of NDLC 17

Wear particles size distribution sample of wear particles Measurement of wear particle size using image processing worn region DLC film Fig. 7 Wear particle size distribution for the DLC obtained by image analysis μ 18

Evaluation of wear energy Table 4 Wear energy of the DLC Film Volume (μm 3 ) Surface area (μm 2 ) Surface area/volume DLC 4.71 73.2 15.5 N 5.4 DLC 6.61 170 25.7 N 10.7 DLC 6.05 105 17.3 Total wear volume ratio of surface area to wear volume = total surface area of wear particles Friction energy (J) Wear volume (m 3 ) Surface area (m 2 ) Wear energy (mj) Wear/friction energy ratio 126 7.52 10-14 1.17 10-6 0.321 2.54 10-6 177 1.16 10-13 1.80 10-6 0.496 2.80 10-6 235 2.24 10-12 3.47 10-5 0.956 4.07 10-6 μ surface energy surface area = wear energy The ratio of wear energy to the corresponding friction energy was approximately several ppm 19

Wear volumes of DLCs against alumina Slide distance: 500 m V = k E f σ Wear volume Energy equation Fig.8 Wear volume of DLCs against alumina as a function of friction energy μ 20

Comparison between Wear coefficient and ratio of wear energy / friction energy Table 5 Wear coefficient of DLCs Specimen DLC N 5.4 DLC N 10.7 DLC Wear coefficient 2 x 10-5 3 x 10-5 2 x 10-5 Wear energy /friction energy 0.3 x 10-5 1 x 10-5 1 x 10-5 The wear coefficients are equivalent to the ratios of wear energy/ friction energy Wear coefficient of wear volume friction energy equation indicates the energy consumption rate for the wear μ 21

Wear coefficient of Steels against DLC Fig.9 Wear coefficient of steels against DLC as a function of friction energy The energy consumption rate of the steel was less than ppm μ 22

Evaluation of elastic energy in steel ball Fig. 10 Elastic energy induced by friction force Ur: Elastic strain energy G: Transverse elastic modulus a: Initial contact point b: thickness Load (N) Table 6 Elastic strain energy in steel ball induced by friction force b ( mm) Integral value Stress energy Ur (mj) 9.8 0.062 256 0.03 19.6 0.098 157 0.07 29.4 0.129 97 0.1 Amount μ of elastic energy is apploximately same to the wear energy 24

Estimation of plastics deformation energy Plastics deformation energy of wear particles: Tensile strength (N/m 2 ) fracture strain (% ) wear volume (m 3 ) Table 7 Tensile strength properties of the SUJ2 Steel Tensile strength (MPa) Fracture strain % SUJ2 1960 0.5 μ Friction energy ( J) Fig. 11 Stress strain chart Table 8 Plastics deformation energy of wear particles of SUJ2 Wear volume Plastic deformation (m 3 ) energy (mj) 395 5.40 10-14 0.053 1.34 10-7 972 7.96 10-14 0.078 8.03 10-8 1749 1.05 10-14 0.100 5.88 10-8 2961 1.22 10-14 0.120 4.04 10-8 Plastic deformation energy /Friction energy Ratio of plastics deformation energy to the corresponding friction energy was less than ppm 26

Chemical reaction energy of wear particles Chemical reaction energy of iron oxidation is excluded because of exothermic 4Fe + 3O 2 Fe 2 O 3 + 412 kj/mol μ Table 9 Chemical reaction energy of iron wear particles oxidation for SUJ2 Friction Energy (J) Wear volume (m 3 ) Chemical energy (J) Chemical/ friction energy 8 0.02 x10-12 0.06 8 x10-3 50 0.04 x10-12 0.13 3 x10-3 395 0.05 x10-12 0.16 0.4 x10-3 972 0.08 x10-12 0.26 0.3 x10-3 1749 0.1 x10-12 0.32 0.2 x10-3 2961 0.12 x10-12 0.39 0.1 x10-3 28

Heat energy in the friction energy Table 10 Each energy consumption rate of the steel to the friction energy Consumption energy Energy consumption rate to friction energy Wear energy: E w Less than ppm Elastic strain energy: E e Less than ppm Plastics deformation energy: E p Less than ppm Chemical reaction energy: E c Less than 1%. of exothermic reaction Friction coefficient factors μ = E NL = E + E + E + E + E NL Eh =μ h + μ w + μ e + μ p + μ c μ h : heat factor, μ w : wear factor, μ e : elastic strain factor, μ p :plastic deformation factor, μ c : chemical factor Fig. μ12the ratio of heat energy to the friction energy 30

Section 2: Conclusions 1. The wear coefficient of wear volume friction energy equation indicates the energy consumption rate to the friction energy. 2. The most of the friction energy is transformed into heat. μ 31

3. Temperature rise simulation in the steel ball Objective Simulate the temperature rise on the sliding surface between steel ball and DLC using the heat energy μ 33

Initial condition for temperature distribution simulation 3D ball model W = de dt = μn dl dt = μnv t: time v: sliding velocity Contact diameter:a 60, 90, 120, 300 µm μ Thermal conductivity (W/mm ) Heat transfer coefficient (W/mm 2 ) The contact layer is metal or metal oxide Fig.13 3D model for temperature distribution simulation Table 11 Initial condition of the temperature simulation for SUJ2 Initial temperature ( ) 4.6 x 10-2 4.65 x 10-6 22 Layer 1 22 Layer 2 22 Layer 3 22 Body Watt (W) 0.2 (at 4.9N test) 0.4 (at 9.8N test) 0.8 (at 19.6N test) 34 1.2 (at 29.4N test)

Temperature distribution in the steel ball due to the friction heat Temperature distribution around the contact surface for SUJ2: (a) Top view (b) Cross section (b) (a) Maximum temperature rise at the contact surface was 560 on the condition the contact layer is oxidized μ 35 Fig.14 3D temperature distribution simulation result

Temperature rise on the contact surface of steel ball (b) Fig. 15 Temperature rise on the contact surface for SUJ2 :bottom layer (a) metal (b) oxidized metal μ 36

Measuring temperature rise in the steel ball using thermocouple 0.3 mm Fig. 16 Schematic of friction heat measurement apparatus mounted thermocouple Fig. 17 Temperature rise at the 0.3 mm upward position in SUJ2 ball from the sliding surface during tribotest against DLC Measurement point: 0.3 mm above the contact surface μ 37

Comparison between computer simulation and experimental results of temperature rise Evaluation point: 0.3 mm above the contact surface Partition rate to DLC (%) Partition rate to SUJ2 (%) Fig. 18 Comparison between calibrated ANSYS data and experimental data: SUJ2 Good accordance between calibrated simulation and experiment was obtained μ 38

Section 3: Conclusions The temperature simulation proves that the temperature on the contact surface increases with decreasing the contact diameter or by the existence of the iron oxide layer. μ 39

4. Energetic analysis of the wear behavior of the steels against DLC Objective Investigate the difference of tribological behaviors between SUJ2 and 440C against DLC in terms of friction energy. μ 41

Tribological behaviors of SUJ2 and 440C against the DLC (a) Friction coefficient behaviors of SUJ2 and 440C Fig. 19 Tribological behaviors of SUJ2 and 440C (b) Wear volume as a function of sliding distance Investigate the relationship between the tribological properties and the temperature rise μ Simulate the temperature rise on the sliding surface 42

Prediction of real contact area on the sliding surface for simulation F = 9.8 N An = π (2x10-4 ) 2 Radius of sliding surface: 200 μm 1 E = 1 1 ν E + 1 ν E steel : 200 GPa ν steel : 0.3 E DLC : 290 GPa ν DLC : 0.3 F/(E An) = 6 10-4 Fig. 20 Normalized real contact area versus normalized load: standard deviation of surface height:15 μm [1] E μ [1] R. Jackson and I. Green, On the modeling of elastic contact between rough surfaces, Tribology transactions, 54, (2011), pp.300-314 43

Comparison between the simulation data and experiment result Simulation result : real contact ratio 1/1000 SUJ2 240 Temperatures indication paint method 130 light red blue violet 260 light green red brown 330 light blue gray 440C 50 Fig. 21 Temperature rises of the sliding surface as a function of sliding velocity Nominal contact diameter: 300μm Fig 22 Thermal paint test on SUJ2 Sliding velocity: 100 mm/s μ 44

Dependence of SUJ2 and 440C hardness on temperature Fig. 23 Dependence of the hardness of SUJ2 and 440C on temperature [2], [3] The wear resistance of SUJ2 was poor than that of 440C The hardness of SUJ2 was lowered due to higher temperature rise μ 45 [2] Steel Handbook (1980), [3]Hasegawa, M. (1973), Stainless Steel Handbook,

Section 4: Conclusions The poor wear resistance of SUJ2 was caused due to high temperature rise induced by the friction energy μ 46

5: Conclusions 1. Friction coefficient and wear volume are related with the friction energy V = k E μ = E NL σ 2. The wear coefficient of wear volume friction energy indicates the energy consumption rate to the friction energy 3. Most of the friction energy is transformed into heat. 4. Temperature on the contact surface raise with the smaller contact area and the existence of the oxide layer 5. The poor wear resistance of SUJ2 was caused due to high temperature rise induced by the friction energy µ 48

ACKNOWLEDGMENTS I wish to thank Dr. Chitoshi Masuda (retired Waseda University) for giving the opportunities to study tribology and thanks Mr.Egashira (retired National Institute for Materials Science) for helpful discussions of this work. µ 49