Current–Voltage–Friction Characteristics of Grease in Electromechanically Loaded Sliding Contacts

Tribological contacts subjected to electromechanical loading conditions are found in many applications. The presence of electrical loads, in addition to mechanical loads, causes surface damage and accelerates the lubricant loss by evaporation [1,2]. Surface damage, such as ablation pits, three-body abrasion, adhesion, and corrugation patterns, is caused by arc erosion in the presence of electric current [3–5]. The temperatures in the vicinity of the arc erosion zone are predicted to reach beyond the lubricant degradation temperature [6,7]. One of the machine elements frequently found failing due to the presence of electrical loads is the bearings used in traction motors and transmission systems of electric vehicles. The flow of parasitic currents occurs when the voltage exceeds the breakdown strength of the lubricant film formed between the contacting surfaces in rolling elements and raceways in the bearings [1,8–10]. The increase in friction coefficient occurs in electromechanical contacts and has been attributed to electro-adhesion and grease decomposition [11–13].

Breakdown voltage, the voltage at which current can pass through an insulator, is reported for a few bearing types using full-bearing tests [14–16]. Multiple contacts prevail in rolling element (ball) bearings such as inner race-rolling element, outer race-rolling element, cages-rolling element, and outer race-motor housing, as shown in Fig. 1(a). Therefore, the full-bearing tests do not reveal an accurate picture of the local failure due to the passage of current. The contact between the rolling element-on-outer raceway subjected to voltage in addition to the mechanical loads can be reduced to a ball-on-disk system where it is easier to evaluate the individual effects (Fig. 1(b)). Although the contact between the ball and the raceway is a rolling contact governed by rolling friction, researchers still use the ball-on-disk, in which there is sliding contact, to characterize lubricants used in bearings as the worst operating scenario [13,17,18]. Greases are preferred for lubrication in motors due to their sealed-for-life nature. Grease thickeners, base oils, and additives play a major role in the performance [19,20]. For electromechanical contact applications, greases need to be able to resist electric current flow to have less oxidation and degradation [21].

The static dielectric breakdown voltage of greases is estimated according to ASTM D149, whereas the standard ASTM D877/D1816 is used to characterize the dielectric field strength (V/mm) of oils used in transformers. The composition of base oil, thickener, and additives influences the dielectric strength of the grease [22,23]. Four different types of greases were studied, and it is reported that when grease undergoes repeated discharges, the chemical changes manifested as a decrease in the concentration of additives and thickener in the Fourier transform infrared (FTIR) spectra due to ionization and high temperature [22]. The conducting channel formed due to the chemical decomposition leads to a reduction in breakdown voltage. The film thicknesses under highly stressed elastohydrodynamic lubrication conditions are of the order of a few hundred nanometers. The estimation of dielectric strength in a dynamic contact is difficult, and limited research has been carried out. Sunahara et al. [24] reported the breakdown voltages at different film thicknesses for grease in a modified ball-on-disk configuration. The authors also reported that the discharge would occur at random locations in the contact zone. The debris and adhesion products that form in dynamic sliding contact influence the dielectric field strength.

The electrical equivalent of an elastohydrodynamic lubricated contact in bearings is a resistor-capacitor network [25–27]. When bearings experience an increasing voltage, their electrical condition transitions from a non-conductive state (capacitor) to a conductive state (resistor). The transition voltage depended on the contact pressure and rotation speed and was evaluated using an oscilloscope in a full-bearing test [15]. Degradation of the grease and morphological damages were found to occur in bearings when the applied voltage exceeds the breakdown voltage [16]. The reason for the transition is that the lubricant film behaves as a capacitance at high film thickness and a resistor at low lubricant film thickness due to increased metal-to-metal contact and intense electrical discharges. Current flows through (i) the metal-to-metal asperity contacts and (ii) a very thin lubricant film when the applied voltage exceeds the lubricant breakdown voltage. In the simplified experiments, the ball and the disk behave like a sphere-plane electrode system. The lubricant in between acts as a dielectric medium. The electrical equivalent of the ball-on-disk system is shown in Fig. 1(b). The electrical resistance in the simplified system includes the resistance of the ball (R_ball) and the disk (R_disk), while R_lub includes the resistances due to metal-to-metal contact, breakdown of the lubricant [28], and the wear debris present. When the lubricant does not allow the current to flow, it behaves like a capacitor instead of a resistor. The transition behavior of lubricant with increasing voltage was discussed in recent works by Li et al. [29,30]. The authors reported that the lubricant film thickness was affected in a glass disk versus steel ball contact subjected to increasing voltages. In a steel-on-steel contact, it is not possible to observe film thickness, but other methods, such as contact resistance [31] and ultrasonic transducers [32], can be used to estimate the film thickness. Current–voltage (I–V) characteristics can reveal the nonlinear behavior of the electrical contact resistance due to varying film thickness caused by an increase in voltage. If the I–V curves of bearings can be reduced to a single contact, it will aid in screening lubricants for use in traction motor bearings.

In this study, an experimental methodology to measure the I–V characteristics of grease under dynamic conditions is reported using a simplified ball-on-disk system. The lubricant is subjected to increasing voltage, and the current and friction behaviors are discussed under dynamic conditions. The current–voltage (I–V) and friction–voltage (µ–V) curves can together offer a safe voltage limit that could be considered for the bearing operation under an electric environment. Knowing the voltages at which significant currents can pass through a lubricated contact can help develop appropriate mitigation measures for the bearing current problem.

Grease-lubricated motor bearings are widely used in traction motor bearings as they are one-time filled and maintenance-free. National Lubricating Grease Institute (NLGI) Grade II lithium soap thickener with mineral base oil grease, used in motor bearings, is considered in the current study. The ball and disk samples are made of bearing steel, SUJ2 (JIS equivalent of AISI 52100). The disks were ground and polished to a surface roughness (R_a) in the range of 33 nm ± 7 nm. The balls were used in as-received condition with a surface roughness (R_a) in the range of 35 nm ± 4 nm. The ball and disk samples were ultrasonically cleaned with acetone for 10 min before the experiment. Properties of the test materials are given in Table 1. The rheological properties of the grease are measured using a rheometer (Anton Paar MCR 52, Austria) with a concentric cylinder geometry. Dynamic viscosity was measured in a rotation test with a strain rate varying from 1 s⁻¹ to 1000 s⁻¹. Storage and loss modulus were measured in an oscillatory test with strain amplitude varying from 0.1 to 100%. The measurements were carried out at two different temperatures, 40 °C and 80 °C.

To study the contact characteristics under combined electrical and mechanical loads, an in-house developed ball-on-disk electromechanical tribometer is used (Fig. 2). The ball specimen is mounted in the stationary ball holder, and the disk, mounted on the disk holder, is rotated using a servomotor at the set speed. The positive terminal is connected to the ball holder, and the negative terminal is connected to the disk holder via a carbon brush arrangement. A regulated DC power supply was used to supply the voltage to the contact. The test rig can measure the instantaneous contact voltage and contact current. The contact resistance is calculated using the measured contact voltage and current. In addition, the instantaneous friction force is acquired using a high-precision load cell and a data acquisition system. The sliding contact studied represents the worst-case scenario in bearings, although rolling/sliding contacts occur in rolling element bearings. In addition, the measurement of electrical parameters is simplified in the case of sliding contact as it requires a single contact brush instead of two carbon brushes in the case of rolling contact. A detailed description of the test device used can be found in Ref. [33].

The instantaneous current waveforms are monitored using a digital oscilloscope (Keysight DSOX1204G, Malaysia). The oscilloscope measures the magnitude of the contact current by measuring the voltage across the shunt resistor. The waveforms were initially sampled at 250 kHz for a duration of 1 s, and the average readings were considered. At each voltage, two waveforms of length 2000 were collected. The NI 6229 data acquisition system (DAQ) acquires the average current and voltage values every second while sampling at a rate of 2 kHz. Since the contact in dynamic sliding conditions continuously varies due to film formation and damage in the contact zone, the electrical resistance varies continuously. Hence, the contact voltage and current measured are varying. The current measurements from the DAQ and the average current from the oscilloscope waveform were typically found to be within standard deviation of each other. An in-house designed electronic measurement module measures the current supplied to the contact using a shunt resistor and the voltage at the contact. The voltage supplied by the power source is termed V_app. The contact voltage (V_con) is the voltage drop that occurs in the ball and disk contact (predominantly) and also includes minimal additional resistances from the carbon brush and the wires leading to terminals of DAQ. The instantaneous friction force is continuously measured using the high-precision load cell (Fig. 2), and the coefficient of friction is estimated. The frictional data is collected at a sampling rate of 500 Hz and averaged for 1 s.

The measurement of current–voltage characteristics was carried out during the sliding test. At the start of the test, a thin layer of grease was applied uniformly on the disk surface. The contact is operated in starvation conditions. The film thickness of the lubricant, which is proportional to the ratio of sliding speed and load [34], plays a major role in the passage of the currents. In this study, the experiments were carried out at five different loads: 5, 7.5, 10, 12.5, and 15 N, applied on the stationary ball. The disk rotational speed was chosen to achieve a constant sliding velocity of 200 mm/s between the ball and the disk. The film thickness estimated using the Hamrock–Dowson equation is in the range of 85–93 nm for the loads in the range of 15–5 N, respectively (pressure viscosity coefficient is considered as 25 GPa⁻¹ for mineral oil) [35]. The composite roughness of the ball and disk samples is around 77 nm, resulting in a lambda ratio (λ) in the range of 1.1–1.2, which is on the borderline of boundary and mixed film regime. In the current investigations, the ball is stationary, and only the disk rotates. This set of experiments will be referred to as “dynamic” or “sliding” conditions.

In all the experiments, the voltages were applied after a run-in period of 1 h (sliding distance = 720 m). Following the run-in period, the voltage was increased from 0 V in steps of 0.1 V, with each voltage being applied for 60 s to measure the voltage at which the ball-on-disk contact begins to conduct. The experiments were conducted till the applied current reached 3 A due to the power rating limitation of the shunt resistor (10 W) used. Two oscilloscope readings were taken in each period. Application of ramped voltage loads causes only minimal progressive damage due to current passage. A schematic of the testing protocol is shown in Fig. 3. The test conditions employed are summarized in Table 2.

A run-in period is considered to prevent the surface roughness effects and facilitate the formation of uniform film thickness (Fig. 4). As seen from the typical experiment run (Fig. 4), the coefficient of friction in the run-in period reaches a maximum value of around 0.5, which is close to the metal-to-metal friction value of 0.49–0.71 [36]. Once the steady-state is reached, the voltage required for the passage of current represents the voltage needed to cross the barrier of the lubricant film. The coefficient of friction decreased to a range of 0.07–0.12. Figure 5 shows a typical variation of the coefficient of friction and current at varying contact voltages based on three tests carried out.

Tests were also conducted to measure contact resistance in a lubricated static condition. A thin film of grease (like the case of sliding condition) was applied on the surface, and the load was applied to the ball without disk rotation (i.e., static condition). In a static test, metal-to-metal contact occurs and it operates in a “resistive condition.” This test is carried out to compare the effect of film formation under dynamic conditions on the electrical parameters. The voltage ramp was performed like the dynamic experiment. Each voltage was applied for 20 s as the values did not fluctuate in this case due to their non-dynamic nature. This set of experiments is referred to hereafter as “static” or “stationary” conditions.

Bearing noise due to increased friction in electric vehicles is a major concern as the drive system has fewer moving parts than engine-driven vehicles. The dynamic viscosity, storage modulus, and loss modulus of the test grease at varying temperatures are shown in Fig. 6. With increasing temperature, the grease viscosity is found to decrease (Fig. 6(a)). The grease is found to have a high structural strength, as seen from the intersection of the loss and storage modulus curves in Fig. 6(b), and is found to decrease with an increase in temperature [37,38]. The lubricant film thickness is an important parameter that influences the coefficient of friction and is affected by electrical load. The coefficient of friction under mechanical load is related to the film thickness [39]. The lubrication mechanism in grease is due to the release of oil from the grease thickener. At the contact region, between the ball and disk, there will be two layers of lubricant: the base layer is the grease thickener, and the top layer is the released oil [40]. The oil release from the grease is high at high loads due to an increase in shear load and results in a low friction coefficient [41]. This phenomenon of lubricant release is essential for understanding the I–V characteristics and is discussed in the following sections. The wear scar images of the tested ball samples at different applied loads are shown in Fig. 7. An increase in the wear scar diameter with increasing load is observed. When current passes through the contact by electric discharges, spot welds form between the ball and the disk. The localized spot welds shear due to the sliding action, and the remnants are seen at high magnification. A similar observation of adhesive products is reported in Ref. [13]. In addition, oxide debris is formed due to the excess temperature, as reported by Cao-Romero-Gallegos et al. [18]. Three-body abrasion results in scratch marks in the wear scar.

The current–voltage (I–V) characteristics of grease under lubricated dynamic and static conditions are shown in Fig. 8. The current values are plotted in log scale on the y-axis for better understanding. The electrical behavior of the contact in the static and dynamic conditions differs significantly at low voltages. Larger fluctuations in the current flow are observed in dynamic (sliding) contact as opposed to static contact due to the variation in the contacting surfaces of the ball and disk. In static experiments, the metal-to-metal contact offers a path of least conduction and allows a stable current flow. Even at 0.1 V, a current flow between the surfaces is observed. The electrical behavior in dynamic experiments can be divided into three regimes. In regime 1, no current flow occurs at the contact. The voltage is not sufficient for the current to pass through asperities separated by the boundary film. In regime 2, the current starts to flow, but the voltage is not sufficiently high enough for the entire region to conduct the current. As the disk rotates, the regions in which the asperity gaps are closest conduct, whereas there is no current passage through other regions. In regime 3, most of the contact has experienced an electrical breakdown. The lubricant film can no longer resist the current flow. Both the lubricant and the contact are now exposed to a severe electrical environment that can cause lubricant degradation and morphological damage [12,42]. The transition from the first to the second regime depends on the applied stresses. A high contact load causes the transition to occur at a low voltage. The late onset of the flow of current in the 5-N contact load case indicates that although the oil release is less than higher loads, the gap between the local asperities is high. Hence, a high voltage is required for the current to flow. In the case of the 15-N load test, although there is an increased oil release, less gap between the asperities is expected compared with the 5 N load. Therefore, the current flow occurs at a lower voltage. However, in regime 3, the current magnitudes are found to be lower at higher loads. The excess oil released from the grease at high load shields a larger part of the contact compared to a lower load from electrical breakdown.

The electric contact resistance can be obtained by the ratio of the contact voltage and current and is plotted in Figs. 9(a) and 9(b) for the dynamic and static cases, respectively. A dependence of the contact resistance on the contact voltage is observed. Due to the joule heating-induced temperature rise, the resistivity of oil decreases [43]. In addition, the viscosity of the lubricant decreases, which leads to decreased film thickness (as seen in Fig. 6(a)), which in turn reduces the lubricant's electrical resistance. The resistance at the contact is the equivalent of the metal-to-metal asperity contact resistance and the resistance offered by the lubricant. The resistance offered by the lubricant to current flow is affected by temperature rise at the contact and electrical breakdown. Significant metal-to-metal asperity offers a pathway with the least resistance in static contact at low voltages. At higher voltages, the lubricant resistance changes due to the temperature rise at asperities, which, coupled with electrical breakdown, leads to a decreased resistance. In the case of dynamic contact, the resistance at the contact is high at low applied voltages. With an increase in voltage, a large volume of the lubricant breaks down electrically, providing more pathways for current flow [44]. As seen from Fig. 9(a), the electrical resistance in the dynamic contact decreases rapidly and tends to the static value with increasing voltage.

The electrical contact resistance shows two different behaviors until regime 3 and in regime 3 in the dynamic test condition. To understand this, two terms: the lowest asperity gap, d_min, and the average gap of asperities, d_avg, are considered. At contact voltages before the transition to regime 3, the highly loaded (15 N) contact exhibits a lower contact resistance than the 5-N loaded contact. At these voltages, the contact relies on the passage of current through the lowest asperity gap between the ball and disk [45]. This d_min in the case of low load (d_min.LL) is expected to be high, as the lubricant film should better resist the contact pressure, whereas in the case of higher load (d_min.HL), the asperities would come much closer to each other, which results in a lower resistance. This is also responsible for the early onset of regime 2. At voltage levels after the transition to regime 3, it is found that the contact resistance is increased with increased load. The mean contact resistance varies from 2.89 Ω to 1.27 Ω in the case of 5 N and 4.06 Ω to 1.63 Ω in the case of 15 N in the voltage range of 2.3–3.3 V. As stated before, the higher oil release in the 15 N case, effectively shields a larger area of the contact from breakdown. In static contact conditions, not much variation is seen at different normal loads investigated. However, beyond a particular voltage, all five load cases show similar resistance values. This is attributed to the saturation in the area in which the breakdown can occur in the contact region. A schematic representation of the mechanisms for current flow in static and dynamic conditions is shown in Fig. 10, which is based on the wear scar observations. In the case of a 5 N load, several small-sized spot welds are observed, whereas in the case of a 15 N load, large-sized spot welds are found.

The typical evolution of the friction coefficient with constant applied voltages at different normal loads is shown in Fig. 11. For all applied loads, the friction coefficient remains constant initially and then increases beyond a particular applied voltage. The current magnitudes at this voltage varied between 35 and 60 mA. Under combined electrical and mechanical loading conditions, the electric current influences the friction coefficient due to electrostriction [46], lubricant viscosity variation due to higher contact temperature (joule heating), micro-bubble formation [47], and lubricant degradation. Another reason that could lead to high friction could be the formation and destruction of the adhesive welding between the ball and the disk due to the electric current flow [18] and is also observed in wear scar images (Fig. 7). The friction coefficient starts to deteriorate at high loads even at low voltages indicating the dependence of normal load on the lubricant breakdown.

The current conduction duration is calculated from the waveforms as the time during which the observed current value is greater than the background noise. The measurement of pulse duration from the oscilloscope waveforms in the 10-N load test is shown in Fig. 12. The total duration of the waveform is T = 1 s. The percentage duration of conduction (%t_toc) was calculated from the waveform files using python® code. For all the other loads, the same procedure is followed, and the mean values of the percentage duration of conduction at different applied loads are plotted in Fig. 13. The sigmoidal trend observed in Fig. 13 closely matches the literature on applied voltages in full-bearing tests [15], albeit for different operating conditions.

The current flow at the contact occurs in three different phases, as observed from the different waveforms in the oscilloscope. In the first phase, no current passes, and in the subsequent phase, the current passes intermittently. In the third phase, the current passes through the contact completely (Fig. 11). The three phases are related to the contact and applied voltage, as contact voltage does not vary linearly with the applied voltage. In a grease-lubricated sliding contact under mechanical loading conditions, there will be oil film, grease thickener, and wear debris at the contact. Under electrical loading conditions, an electric potential is present across the ball and disk. The barrier for the passage of current is the lubricant film. The electric current will normally flow when the voltage across the asperities exceeds the dielectric breakdown strength of the intervening medium. The current will take the path of least resistance to flow between the two asperities. This path is influenced by the presence of wear debris, thickener particles, air bubbles, and other inhomogeneities in the contact.

It could be hypothesized that there exists a dynamic dielectric breakdown strength, which is the equivalent dielectric strength of the contact. This dynamic dielectric strength is expected to depend on the static breakdown strength of the lubricant. However, estimating dielectric breakdown strength in dynamic conditions is difficult. This is because when a contact operates under starvation conditions, the film thickness deteriorates to a steady-state value of around 20–50 nm [48]. Different types of electrical conduction mechanisms can occur at these small length scales and are not only limited to electrical breakdown but also to others, such as tunneling effects, A-fritting, and B-fritting [28]. Due to this, the electrical breakdown could occur at a voltage less than what is predicted from a static test. This is shown schematically in Fig. 14(a).

The flow of current in contact with different levels of applied voltage is shown in Fig. 14(b). When the applied voltage is below V_init, the contact operates as a capacitor and, will not conduct any current. However, the lubricant in the vicinity of the asperities is polarized. Since there is no flow of current, the waveform obtained is like when no voltage is applied. Upon further increase of current, at the local asperity level, the applied voltage is just sufficient to breach the dielectric breakdown strength of the lubricant. However, due to the dynamic sliding effect, the gap between the asperities changes instantaneously. Subsequently, the area in which the breakdown occurs changes. So, large fluctuations in the current values are observed.

With a further increase in voltage, larger areas of the lubricated contact come under the influence of the applied voltage. The voltage is sufficient to breach the breakdown strength in a wider area. As a result, the flow of current in the contact is continuous and not intermittent, as in the previous two cases. However, there is still a deviation observed in the waveforms as different areas still show different levels of asperity contacts. This can also be seen in the fact that all the waveforms show periodic behavior because of the effect of disk rotation. Beyond a particular voltage, nearly the entire contact comes under the influence of the voltage. As a result, the current flow becomes stable. Further, the resistance has reached a steady-state value in this region. The breakdown resistance can be believed to be inversely proportional to the area in which electrical breakdown occurs. Beyond a particular threshold, slight variations in area do not have a significant effect on the current. However, with the passage of current through a larger volume of lubricant in the contact, the average temperature in the lubricant at the contact can increase drastically. This large increase in temperature will affect the lubricant properties.

The decrease in resistance observed with increasing voltage is due to two reasons. One is the decrease in film thickness, and the second is an increase in the area where the breakdown occurs. The two reasons ultimately lead to one point, which is that there is a large amount of current that can flow. There could be various reasons for a decrease in film thickness. The large current flow can induce micro-bubble formation [47] and increased temperatures due to joule heating. The low film thickness will further increase the area of breakdown. These effects can pile up together to cause significant damage to the contact by the occurrence of metal-to-metal contact and lubricant degradation. The results from this study agree well with the recent work by Li et al. [29], with the difference being that the present method utilizes a steel-versus-steel contact that occurs in bearings and uses contact resistance measurement instead of film thickness measurement.

Parameters that will assist in choosing the lubricant for electromechanical contact will be helpful. The following are considered important in the grease-lubricated contact (as shown in Fig. 15 for the 15 N load):

• V_init: The initialization voltage at which the contact begins to conduct electricity. For a conservative design, this is the voltage beyond which the current can pass through a dynamic contact under a given load.

• V_10%TOC: While the current begins to pass at V_init, it may still not be harmful enough for the bearing performance. From the percentage time of conduction graph, the voltage level at which a 10% time of conduction is observed can be used. This criterion is described in Ref. [15].

• V_90%TOC: When the percentage time of conduction reaches around 90, the contact is no longer in a state where it can resist the current flow. This criterion is described in Ref. [15].

• V_friction: The voltage level at which a change in friction is observed in the short-duration testing. Encountering this level of voltage would be harmful as the very purpose of the lubricant to reduce friction is defeated. The current (I_friction) observed at this value could also be of interest. In the present study, this value was found in the range of 35–60 mA at the different loads applied. Evaluating this from a full-bearing test is complex.

• V_knee: A piecewise linear fit was made from the raw data used for the I–V curve (Fig. 8). The piecewise fit showed that ohmic behavior occurs beyond a specific voltage, and this is termed knee voltage. Such behavior has also been observed in the bearing tests, as reported in Ref. [49]. The significant amount of current passing through the contact can lead to very high contact temperatures. The amplitudes of current recorded beyond V_knee, which are in the range of 1500–2500 mA in the present study, can show a significant effect on the friction and wear characteristics of the contact, even in short-term tests. These current amplitudes can raise the temperature of the grease beyond 300 °C, significantly decreasing the grease life [18]. Further, the formation of oxidation products due to the rapid depletion of the anti-oxidant additives in the grease is harmful to the bearing [50]. However, while the identification of this point is of scientific importance, it is to be noted that in the real-life context of bearing, encountering this voltage for a significant amount of time can often mean that the bearing is on its way to failure and using V_friction might be a much better parameter for the design.

The V_10%TOC and V_90%TOC criteria described in Ref. [15] were used to characterize the flow of currents passing through bearings in simulated tests. The voltage levels at the different points of interest for the lithium grease used in the present study are shown in Fig. 16. The contact load, which influences the lubrication film thickness, affects all the points of interest. The voltage levels for the different parameters show an increasing behavior. It may be possible that the other lubricants could show a different trend as well. In the full-bearing tests carried out on 608 bearings, the authors [15] observed a 10% time of conduction (V_10%TOC) and 90% time of conduction (V_90%TOC) in the range of 2–4 V for a speed 1000 RPM, respectively, at 4 N load. In the present work, for a single contact, the voltages at 10% time of conduction (V_10%TOC) and 90% time of conduction (V_90%TOC) are in the range of 1.1–2 V for the case of 5 N load and speed of 200 mm/s.

The experimental studies carried out indicate the suitability of using simple ball-on-disk tests for the initial screening of lubricants used in electromechanical environments. The following conclusions are drawn for the lithium mineral oil grease.

• Three regimes were identified from the current–voltage characteristics of the lubricant under dynamic electromechanical contact conditions. A significant current passes through the contact in regime 3, and such voltages need to be avoided to prevent lubricant degradation.

• The static contact results in high current flow at a given voltage due to metal-to-metal contact. With increasing voltage, dynamic contact tends to behave like a static contact as the lubricant breaks down.

• With increasing contact pressure, the voltage required for the current to flow decreases. However, beyond the knee voltage, the contact pressure does not play a significant role as the current amplitude rises sharply.

• The friction coefficient is found to be dependent on the normal load and applied voltage. Friction coefficient rise occurs at a higher voltage in case of low normal loads as compared to higher normal loads. The voltage and current amplitudes are found in the range of 1.6–1.9 V and 35–60 mA, respectively, representing the limiting electrical parameters based on the normal load for the grease investigated.

• From the oscilloscope waveforms, the current flow in the ball-on flat sliding contact shows a transition behavior like that reported in the full-bearing test of grease-lubricated bearings. The simplified methodology described using ball-on-disk contact study can be used as a screening method for the greases used in various motors.

The authors acknowledge the support of M/s NSK Bearings, Japan, for providing samples. The authors acknowledge the support of Ms. Prerna Maheshwar and Dr. Jitendra Sangwai from the Department of Chemistry, Indian Institute of Technology Madras for rheological measurements using Anton Paar MCR 52.

• The Indian Institute of Technology Madras under the Institute of Eminence Grants (Project No. SB22231255MEETWO000104).

There are no conflicts of interest.

The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request.

I =

current, A

T =

time, s

V =

voltage, V

d_avg =

average gap between the asperities of ball and disk

d_min =

the minimum gap between two asperities in ball and disk

d_min.HL =

minimum gap between two asperities in case of high applied load

d_min.LL =

minimum gap between two asperities in case of low applied load

V_app =

applied voltage from the power supply, V

V_con =

contact voltage measured using data acquisition system, V

V_friction =

voltage at which friction is affected, V

V_init =

initialization voltage for current flow, V

V_knee =

voltage beyond which significant current flows for small increase in voltage, V

V_10%TOC =

voltage at which current flows for 10% time, V

V_90%TOC =

voltage at which current flows for 90% time, V

dc =

direct current

%t_toc =

percentage time of conduction

µ =

coefficient of friction