WANG Yan-shuang, ZHANG Luo-ping, YANG Bo-yuan

School of Mechanoelectrical Engineering, Henan University of Science and Technology, Luoyang 471003, China, E-mail: hkd_wang_yan_shuang@126.com

THE DEVELOPMENT OF A LUBRICANT TRACTION MEASUREMENT SYSTEM*

WANG Yan-shuang, ZHANG Luo-ping, YANG Bo-yuan

School of Mechanoelectrical Engineering, Henan University of Science and Technology, Luoyang 471003, China, E-mail: hkd_wang_yan_shuang@126.com

A ball-disc traction test rig is improved through the development of a lubricant traction measurement system, consisting of a resonance force sensitive quartz sensor, a circuit of the sensor and a rigid bracket jointed by a frictionless hinge. The traction coefficients of a kind of domestic lubricating oil were measured at various normal loads, rolling velocities, lubricant inlet temperatures and slide-to-roll ratios on a ball-disc traction test rig equipped with this traction measurement system. The results show that the traction test curve is smooth and the data enjoy a good accuracy, which indicates that the design of the lubricant traction measurement system is reasonable and the precision of the system is high, especially, at high temperature, high rolling speed and small slide-to-roll ratio.

lubricating oil, traction force, sensor, traction measurement system

Introduction

During the operation of a lubricated rolling bearing, an Elastohydrodynamic Lubrication (EHL) film will be formed. The traction force applied on a rolling element interface will accelerate or decelerate the rolling element and influence the dynamic performance of the rolling bearing, its life and reliability. So the traction behavior of the lubricant is an important issue for research. Crook (1963), Johnson and Cameron (1967), Bair the Winner (1977), Dyson (1970), McCool (1999), Gupta (1991) and others tested the traction force of different lubricants in elastohydrodynamic lubrication systems. Now, experimental and theoretical studies in that field are still very active[1-13]. Traction data needs to be obtained by test for engineering applications because no theory is good enough to predict them accurately. Traction test devices can be divided into two types: the two disc traction test rig and the ball-disc traction test rig. The

Fig.1 Ball-disc test rig

1. Working principle of test rig

The traction experiment was carried out on an improved ball-disc test rig designed and built by Wang et al.[14], as shown in Fig.1. The ball-disc test rig includes a ball and a disc specimen, a power system, an oil feed device of the loading system and the hydrostatic bearing, a lubricating oil feed system, a temperature controller, a cooling system, a traction measurement system, a normal load measurement system and a data acquisition system. The principle of the test rig is as follows: the combined rolling and sliding action of the rolling bearings can be simulated by our test rig. The ball and the disc specimens are driven by a horizontal Motor I and a vertical Motor II, respectively, controlled by two frequency changers. The surface velocities of the ball and the disc specimens are thus allowed to vary continuously to generate continuous slide-to-roll ratios. The load is applied to the rotating ball and the disc through a hydrostatic shaft which supports the Motor I, and is measured by a load sensor fixed under the hydrostatic shaft. The rolling motion of the ball and the disc draws the lubricant into the contact conjunction and develops a film with a certain thickness. The sliding velocity is established by adjusting the rotating speeds of Motor I and Motor II, gives this oil film a shear motion and results in the traction force, which causes the ball together with Motor I to deflect around the axis of the hydrostatic shaft to push a traction measurement system mounted on the machine frame so that the traction force can be picked up. The lubricant is fed by an oil jet located in the proximity of the rotating surfaces of the ball and the disc. The lubricant inlet temperature is controlled by a thermostat and an oil heater. A computer system is used to sample and process the data of the load, the traction force and the speeds of the ball and the disc. The rolling velocity can be determined asU=(U1+U2)/2 (whereU1andU2are the linear velocities on the surfaces of the ball and the disc specimens, respectively), and with the slide-to-roll ratioS=(U1?U2)/U, the traction forcesFcan be measured at various slide-to-roll ratios under fixed normal loadW, rolling speedUand lubricant inlet temperatureT, and consequently, the curves describing the traction behavior of the lubricants can be obtained in the form of the traction coefficient against the slide-to-roll ratio under various operating conditions.

The performance of the test rig is as follows: (1) The range of the rolling velocity is 10 m/s-50 m/s, and the velocities can be changed steplessly. (2) The slide-to-roll ratioSis in the range of 0.0-0.2.

(3) The maximum Hertzian contact pressurep0ranges from 0.8 GPa to 1.5 GPa. (4) The lubricant inlet temperature is in the range of 20oC-125oC. (5) The ball and disc specimens are both made of GCr15 and their average surface roughness is less than 0.02 μm. (6) The contact zone is in a state of the full elastohydrodynamic lubrication and the surface roughness makes no contribution to the traction force between the ball and the disk. (The minimum film thicknesshmincomputed by Hamrock-Dowson formula is above 0.1 μm, thusλ=hmin/σ≈5＞3).

2. The principle of the traction measurement system and its construction

The design of the traction measurement system determines the measurement accuracy of the traction force. In the design, the traction measurement system consists of a resonance force sensitive quartz sensor, the circuit of the sensor and a rigid bracket jointed by a frictionless hinge as shown in Fig.2.

Fig.2 Traction measurement system

2.1Resonance force sensitive quartz sensor

The strain measurement is a common and relatively mature method to measure force. The strain of an elastomer with its surface adhered with a resistance strain gauge is used to measure the force. The sensor turns strain signals into voltage signals, which, in turns, are transformed into digital signals by data acquisition board. Thus, not only many pipelines and much cost are involved, but also errors. The precision of the measurement and the control will be affected. On the other hand, the structure and the material of the elastomer would greatly influence the precision of the measurement, especially, the elasticity lag would be difficult to avoid. In recent decades, the resonance force sensitive quartz sensor comes into being, with steady physical and chemical properties, good repeatability and high sensitivity and without elasticity lag. Besides, this kind of sensor could output directly digital signals. As compared with traditional sensors, it performs better in signal acquisition and transmission. A new kind of digital sensor- resonance forcesensitive quartz sensor is selected in this article after many tests and observations. It has a good stiffness with insignificant deformation throughout the experiments. The measurement precision can thus be improved. The basic frequency of the sensor is 3.58 MHz, its force measurement range is from 0 N to 70 N.

傳統(tǒng)制造業(yè)管理者大多是60后、70后，傳統(tǒng)的制造理念導(dǎo)致他們對數(shù)字化企業(yè)認(rèn)知不足，或由于固有觀念太深入，致使管理者在企業(yè)轉(zhuǎn)型變革上顧慮太多，放不開手腳。甚至部分企業(yè)只停留在喊口號上，實(shí)際行動(dòng)缺乏動(dòng)力和活力。一些企業(yè)把數(shù)字化轉(zhuǎn)型升級的重任下放到個(gè)別職能部門權(quán)限里，由于職能部門沒有足夠的權(quán)限進(jìn)行全公司資源整合，導(dǎo)致轉(zhuǎn)型升級困難重重、舉步維艱。

The working principles of the resonance force sensitive quartz sensor are as follows. Its sense organ is an AT-cut disc quartz resonator. The force can be obtained by measuring the change of the resonant frequency. The basic part of this kind of sensor is a mechanical vibrator made by quart slice with its surface plated by a metal film. The mechanical vibration will be excited in the mechanical vibrator due to the piezoelectric effect of the crystal in the alternatingelectric field. As the response, the vibration deformation of the vibrator creates electric charges and in turns an alternating current in the external circuit. Because of stable mechanical characteristics of the quartz, its natural frequency will not change if the vibrator is in a steady state. When a radial forceFis acted on the vibrator edge, the inner stress state in the vibrator will change. This would bring about a stiffness increment Δk, which changes the natural frequency. Meanwhile, the resonant frequency is also changed. The increment of the frequency could be related toF.So, the force could be obtained by measuring the frequency.

2.2Circuit of sensor

The traction is transferred from a pole and a rigid lever (Fig.2) to the sensor, then is picked up and processed by the circuit of the sensor and transmitted into the control software. The sensor circuit is composed of an oscillatory circuit, a voltage to frequency conversion circuit, a temperature compensation circuit and a power supply module as shown in Fig.3.

Fig.3 Block diagram of the circuit of sensor

The oscillatory circuit is a circuit that creates frequency signals proportional to the force. It consists of a series circuit and an integrated inverter. As the basic frequency is very high, it can not be transmitted directly. In order to debase the frequency, there are two oscillators in the conversion circuit. One is connected to a sensitivity resonator which produces a frequency proportional to the force, the other is connected to a reference resonator. The two oscillators give frequencies with a difference, to be dealt with by a difference frequency circuit, which would produce a square wave signal about 2 K in frequency if there is no external force. With an external force, the frequency increases in a rate of 300 Hz/Kg. A shaping circuit makes the frequency signal more regular.

The voltage to frequency conversion circuit is to change voltage signals into frequency signals. A data acquisition board in the computer can only sample frequency signals. So the voltage signals must be transformed into frequency signals by this circuit before they are sampled by the data acquisition board. For the data sampling, the changed frequency signals also need to be treated by a shaping circuit.

The temperature compensation circuit is used to compensate the variation of sensitivity due to temperature in order to insure that the test results are valid at different ambient temperatures. The variation of sensitivity is about 0.392 Hz/Kg/oC, much smaller than the sensor sensitivity, and it should be compensated only when the temperature changes a great deal.

The power supply module is actually a voltagestabilizing chip. Voltage change would affect the measurement, so the chip is to make the voltage stable as the input voltage changes.2.3Rigid bracket

Table 1 physical properties of lubricating oil HKD

The sensor is in a column form and is difficult to be fixed. So a special bracket is designed (Fig.2). The bracket is composed of a pole for transmitting the traction force, a rigid lever and a frictionless hinge. The pole and the rigid lever are made of steel with high stiffness. The frictionless hinge is a cross flexible hinge, composed of a bar and aUchip. The cross direction displacements are all confined, but the rotation is not confined. The frictionless hinge is made of phosphor bronze with good flexibility. The thickness of the hinge is 0.3 mm, determined according to the magnitude of the load. The amplification factor of the lever is 3, determined according to the traction measurement precision. A traction force less than0.01 N could be measured through the lever amplifier and the sensor. The rigid lever deforms very little because of the high stiffness of the bracket and the sensor, to ensure the test precision.

3. Test results

A Chinese aviation lubricating oil Henan Keji Da (HKD) was tested by the traction measurement system on a ball-disc test rig. The physical properties are listed in the following Table 1.

The working conditions are as follows: the lubricant inlet temperatures T: 20oC-125oC, the rolling speeds U: 15 m/s-45 m/s, the slide-to-roll ratios S: 0-0.2, the nominal load W: 20 N-135 N, and the maximum Hertzian contact stresses corresponding to the nominal load p0: 0.8 GPa-1.5 GPa.

For a fixed maximum Hertzian contact stress p0, the rolling velocity U and the lubricant inlet temperatures T. Firstly, the rotating velocities of Motors I and II are adjusted in such a way that the surface velocities of ball U1and disc U2are all equal to the prescribed rolling velocity U, to keep the purerolling condition with no sliding on the contact surfaces of the ball and the disc. The original rotating velocities of Motors I and II (n10, n20) can be obtained as

where R1is the ball’s radius, R2is the distance from the center of the disc to the contact point. Then, the rolling velocity of the ball and the disc is kept constant, while, respectively, the rotating speeds of Motors I and II (n1, n2) are increased or decreased to create a relative sliding velocity ΔU (ΔU=U1? U2). At this moment, the rotating velocities of Motors I and II are, respectively,

The traction forces F is obtained for each relative sliding velocity ΔU. Thus, the traction coefficients μ (μ=F/W ) for each slide-to-roll ratio S (S=ΔU/U) are obtaind, together with the curves describing the traction behavior of the lubricant HKD in the form of the traction coefficient against the slide-to-roll ratio under all operating conditions. Several representational test results are shown in Fig.4.

Fig.4 Variation in traction coefficient (μ) with slide-to-roll (S) ratio at different maximum Hertzian contact stresses when U=35m/s , T=90οC

Fig.5 Variation in traction coefficient with slide-to-roll ratio at different rolling velocities when P0=1.2GPa, T= 90οC

Fig.6 Variation in traction coefficient with slide-to-roll ratio at different lubricant inlet temperatures when P0= 0.8GPa,U=25m/s

As can be seen from Figs.4-6, in the part of low slide-to-roll ratios, the curve is close to a straight line, with a further increase of sliding, the traction coefficient increases non-linearly, and at a certain point, reaches a maximum value. Then it keeps a constant or decreases. The traction coefficient increases with the increase of the maximum Hertzian contact stress at a fixed rolling velocity and lubricant inlet temperature,as shown in Fig.4. It can be seen from Fig.5 that a lower rolling velocity yields a larger traction coefficient at constant maximum Hertzian contact stress and temperature. The traction coefficient decreases with the increase of temperature at fixed maximum Hertzian contact stress and rolling velocity, as shown in Fig.6.

The test results agree well with the traction test data of kin lubricant, tested by Gupta[15]. But from a comparison between Fig.6 and Fig.7, we can see that 5 or 6 test points are obtained by our test rig and only 1 point is obtained by Gupta’s test rig in the range of the slide-to-roll ratio from 0 to 0.01, which indicates that the precision of our traction test system is higher than that of Gupta’s.

Fig.7 Test data from Gupta[15]

4. Conclusion

The traction test system designed in this article can measure the traction force or the traction coefficients of lubricants at various working conditions. The precision of the traction test system is higher than that of other traction test systems especially at high temperature, high rolling speed and small slide-to-roll ratio.

[1] FORD J. M., CHEN K. Speeding up the solution of thermal EHL problems[J].International Journal for Numerical Methods in Engineering,2002, 53(10): 2305-2310.

[2] BAIR S., MCCABE C. and CUMMINGS P. T. Calculation of viscous EHL traction for squalane using molecular simulation and rheometry[J].Tribology Letters,2002, 13(4): 251-254.

[3] STAHL J., JACOBSON B. O. A lubricant model considering wall-slip in EHL line contacts[J].ASME Journal of Tribology,2003,125(3): 523-532.

[4] BAIR S. The nature of the logarithmic traction gradient[J].Tribology International,2002, 35(9): 591-597.

[5] LAFOUNTAN A. R., JOHNSTON G. J. The elastohydrodynamic traction of synthetic base oil blends[J].STLE Tribology Transction,2001, 44(4): 648-656.

[6] DANIEL N., LEGRAND E. Traction behavior of some lubricants used for rolling bearings in spacecraft applications[J].ASME Journal of Tribology,2002, 124(1): 72-81.

[7] GE P. Q., LIU Z. C. Experimental and computational investigation of traction coefficient of a ball traction drive device[J].Tribology International,2002, 35(4): 219-224.

[8] WU Zhao-chun, WANG Dao-zeng. Simulation of the oil slick movement in tidal waterways[J].Journal of Hydrodynamics,2010, 22(1): 96-102.

[9] ATTIA HILI Molka, BOUAZIZ Slim and MAATAR Mohamed et al. Hydrodynamic and elastohydrodynamic studies of a cylindrical journal bearing[J].Journal of Hydrodynamics,2010, 22(2): 155-163.

[10] CANN P. M. Grease degradation in a bearing simulation device[J].Tribology International,2006, 39(12): 1698-1706.

[11] MARTINI A., ZHU D. and WANG Q. Friction reduction in mixed lubrication[J].Tribology Letters,2007, 28(2): 139-147.

[12] KRUPKA I., BAIR S. and KUMAR P. et al. Mechanical degradation of the liquid in an operating EHL contact[J].Tribology Letters,2011, 41(1): 191-197.

[13] MURAKI M., DONG D. Rheological properties of high-viscosity-index mineral base oils for automotive engine lubricants[J].Lubrication Science,2005, 17(2): 185-196.

[14] WANG Y. S., YANG B. Y. An investigation into grease behavior in thermal EHL circular contacts[J].Tribology Transactions,2006, 49(3): 449-453.

[15] GUPTA P. K. Traction coefficients for some lubricants for rolling bearing dynamics modeling[J].Tribology Transactions,2000, 43(4): 647-652.

November 9, 2010, Revised April 24, 2011)

2011,23(4):516-520

10.1016/S1001-6058(10)60144-8

* Project supported by the National 11th Five Year Plan Key Programs for Science and Technology Development (JPDT-115-189), the Creative Talent Foundation in University of Henan Province (Grant No. 2011HAS727).