Wear261(2006)
46–52
Influence of surface topography on friction characteristics
in wet clutch applications
P¨a r Nyman a,Rikard M¨a ki a,∗,Richard Olsson b,Bager Ganemi c
a Lule˚a University of Technology,Division of Machine Elements,SE-97187Lule˚a,Sweden
b Haldex Traction Systems AB,Box501,SE-26124Landskrona,Sweden
c Statoil Lubricants R&D,Box194,SE-14922Nyn¨a shamn,Sweden
Received8September2004;accepted13September2005
Available online4November2005
Abstract
In heavily loaded wet clutches,such as in limited slip differentials,sintered friction materials are somet
imes ud due to their resilience at high loads and high temperatures as well as their competitive cost in comparison to alternative friction materials.
During the lifetime of the clutch,changes in the friction materials’topography occur.The changes will influence the friction characteristics of the clutch,and therefore affect the anti-shudder performance of the transmission system.This paper investigates the influence of,and classifies, changes in the topography of the sintered friction material.
The topography is measured by utilizing vertical scanning interferometry.Different parameters are investigated in order tofind relevant parameters correlating to the wear of the material.
Results show that changes in the topography of the friction material do indeed influence the friction characteristics of the clutch and that it is possible to calculate relevant topography parameters that describe the amount of wear the material has been subjected to.
©2005Elvier B.V.All rights rerved.
Keywords:Wear;Topography;Wet clutch;Stick-slip;Sintered friction materials
1.Introduction
The Haldex LSC AWD system features a multiple disc wet clutch that consist of clutch plates covered with a sin-tered friction material running against parator plates made from hardened steel.The clutch pack distributes drive torque to the rear axle of the vehicle.By electronically controlling the wet clutch,controllability of the torque transfer is gained. Hence it is possible to electronically control the drive train in order to optimize the function of different electronic driving aid systems.
Characteristic operating conditions for wet clutches in this type of application include low sliding velocities and high clutch disc pressure,typically up to0.5m/s and10MPa.Under the conditions it is common that stick-slip or shudder ari.This
∗Corresponding author.
E-mail address:Rikard.Maki@ltu.(R.M¨a ki),
Richard.(R.Olsson),sbga@statoil.(B.Ganemi).behaviour has been investigated and described in a number of papers,both experimentally[1–5]and theoretically[6–8].
The general opinion on this matter states that in order to avoid vibrations,the friction–velocity(µ−v)rel
ationship should prent a low static coefficient of friction(µs)and a dynamic coefficient of friction(µd)that increas as the sliding velocity increas.
The friction characteristics will not be constant throughout the life of the clutch,generally the anti-shudder properties will deteriorate,and at some point cau audible shudder to occur which will determine the end of life of the clutch system.The changed properties can be attributed to degradation of the lubri-cating oil as well as wear or degradation of the friction material [9].A number of different articles dealing with the degradation of organic(or paper-bad)friction materials have been pre-nted[10–12],but on the degradation of sintered sinter material less effort has been made so far.
The aim of this paper is to investigate whether the change in frictional behaviour can be explained by changes in the topog-raphy of the sinteredfiction material and,if that is the ca,try
tuscan0043-1648/$–e front matter©2005Elvier B.V.All rights rerved. doi:10.1016/j.wear.2005.09.020
P.Nyman et al./Wear261(2006)46–5247
tofind relevant surface parameters that can be ud as a mea-surement of the wear or degradation.
timber1.1.Approach
The problem has been approached byfirst running tests in a clutch test equipment in order to investigate whether the change in friction is caud byfluid degradation or by a change in the friction material properties.
The next step is to run a number of clutch plates for different sliding distances,ranging from only a short running-in up to the occurrence of d of life.After the tests,each friction disc is cleaned and then the topography is measured in three dimensions using vertical scanning interferometry.
After that,the topography has been evaluated by calculating many different surface parameters in order tofind parameters that give reasonable values and describe the functional surface in a proper way.One source that caus a lot of complications in this work is the porous nature of the friction materials which makes many standardized surface parameters unstable or cut-off dependent,and therefore unusable in this investigation.
Thefinal step is then to correlate the identified surface param-eters to the sliding distance in order tofind parameters that describe the wear of the friction material.Ultimately this should yield a method
that enables the engineer to predict the residual life of the friction disc by only performing a few topography measurements.
1.2.Limitations
The sinteredfiction material is much softer as compared to the opposing hardened steel parator plates,as revealed by much larger deformations in the sinter material.Therefore,this study has primarily investigated the change in topography of the fric-tion discs and not that of the parator plates.
The topography has been studied at micron he oil grooves,as can be en in Fig.2,have not been included in the study.The surface parameters are only calculated using the part of the surface actually influencing the friction characteristics, pically the fraction of the surface within8m below the highest summits.
ctm什么意思All tests have been performed using the same type of trans-missionflhe commercially ud Statoil LSCfluid.The study covers two different generations of one commercially available sintered friction material(e Section2.2).
2.Method
2.1.Equipment
The investigated clutch plates have been tested utilizing the Limited slip clutch test rig,with the exception of a number of clutch plates originating from full-scale vehicle tests.This equip-ment is described below.
The surface topography has been measured using a vertical scanning interferometer called NT1100from Wyko;this method has also been briefly described
below.
Fig.1.Simplified cross-ction of the Limited slip clutch test rig.
2.1.1.The Limited slip clutch test rig
fgfgThe friction measurements have been carried out using the LSC test rig(Fig.1)[13].
The speed can be varied between0.5and125rpm (2.5–600mm/s on the mean radius).The normal force on the test clutch is applied by a double acting hollow piston cylin-der(4).In this ca the normal force was limited to30kN by a pressure limiting valve.
In the magnified view of the clutch housing in Fig.1,the friction disc(10)and the parator discs(11)can be en.The friction disc is connected to the driving shaft(9),and the para-tor discs are connected to the torsion bar(6).During operation, the shaded parts in Fig.1are rotating.
When a normal force is applied to the clutch by the hydraulic cylinder(4),torque is transmitted from the driving shaft(9)to the torsion bar(6).The transmitted torque is then measured by the torque transducer(7).The applied normal force is measured by the load cell(12).This is possible due to the slider system(8) which allows the torsion bar and torque cell to move freely in the axial direction.Both
the force and torque transducers are full bridge,strain gauge type with built-in amplifiers.The accuracy of the measurements from this rig is well within±1%[13].
During this investigation thermocouples have been installed in the oil sump to measure oil bulk temperature,and in the p-arator disc to measure friction surface temperature.
2.1.2.Wyko NT1100
The Wyko NT1100utilizes vertical scanning/white light interferometry.The technique us the bright and dark pattern which is the result of the splitting of a light beam where one part is reflected against an internal smooth reference surface and the other off the sample.After reflection the beams recombine in the interferometer and a pattern of constructive and destructive inter-ference occurs.The pattern is photographed by a CCD-camera analyzed by using a computer.A piezoelectric transducer moves the focus down and another picture is taken continuing until the specified scanning depth is reached.Then the pictures are ana-
48P .Nyman et al./Wear 261(2006)
46–52
Fig.2.Investigated clutch discs.Separator disk of hardened steel to the left and friction disc with sintered brass friction material to the right.
borglyzed in the computer which store the height data and create a three-dimensional picture.After this the measurement data can be analyzed.The software ud for the analysis was Vision32.The optics in u has a large influence on the accuracy of the topometer.In this investigation a magnification of
1.25have been ud.As a result of this the smallest measurable area is 6.57m ×6.57m.Height resolution is t to 1nm.Addi-tionally there is a built-in uncertainty due to problems with the reflection properties of some surfaces.The total area of measurement is 3.8mm ×5.0mm and covers more than a full square of the groove pattern of a sintered lining friction disc (e Fig.2).
A measurement uncertainty analysis was performed which showed that 10measurements on each friction disc is sufficient in order to obtain an interval of confidence level of about 90%.2.2.Friction and parator discs
In this study two different sintered friction materials have been investigated,one is an older composition and the other a newer modified material.The investigated friction materials consist of a dispersion sintered lining with a brass ba applied to a hardened steel disc.The materials are able to withstand higher stress and temperatures compared to paper-bad mate-rials and are fairly cheap to manufacture contrary to carbon-fibre materials.The parator disks are manufactured of hardened steel.
A friction pair can be en in Fig.2.The outer diameter of the friction lining is 108mm,and the inner diameter is 76mm.The area of contact when the oil grooves have been accounted for is approximate
ly 2250mm 2.The grooves facilitate oil distribution to the area of contact,and help in lowering the temperature in the clutch by enhanced oil flow.
The friction materials consist of a brass ba containing a small amount of solid carbon-bad lubricants.In addition to this,they also contain solid friction increasing fractions such as silicon oxides.Their heat conductivity is ∼15.7W/m ◦K,their heat capacity is ∼471J/kg ◦K and their E-module is 2000–5000MPa (porosity-dependant).
The older type of friction material appears darker in colour compared to the newer material.More importantly,the life of the material differs by a factor 4in favour of the new composition
under the test conditions chon in this study.The older material managed to sustain its frictional properties for about 8h com-pared to 32h for the new material.The test is accelerated and the values should therefore not be regarded as actual values but rather rve as an indication of the performance of the materials under very high load conditions.The older material,in compar-ison with the newer material,contains less copper and tin,and a larger fraction of zinc and solid carbon-bad lubricants.The topography of the sintered face of the friction disc is very porous;this enables the lubricant to reach to and flow through the contact zone providing lubrication and transporting the frictional heat away.
When measuring the topography,the high porosity makes the measurement complicated due to the fact that at the bottom of the pores,if there is one,the material doesn not reflect light;this means that the ratio of measured data points to actual points usu-ally is about 50–70%compared to about 99%for a homogenous steel plate.The problem was solved by normalizing the mea-surement data to contain only height information 3m above and 5m below the most frequent height of each measured sur-face.Another advantage of this is that most of the points left in the datat have participated in the contact between the points that actually experience a change in topography.The topography of a new friction disc is too rough to be characterized using the topometer.2.3.Lubricant
In this study a commercially ud mi-synthetic Statoil LSC fluid designed for this application have been ud.This fluid has a viscosity of 35cST at 40◦C and 6.7cST at 100◦C,and meets the requirements for a fill-for-life rvice in the vehicles.The lubricant is further described in [5].3.Friction 3.1.Test cycle
Each tested clutch disc is run at a different number of load cycles.A load cycle consists of two parts,unloaded and loaded.The unloaded part lasts 30s and the loaded load cycle/min.During the loaded part,a normal force of 15kN,a
P.Nyman et al./Wear261(2006)46–52
49
Fig.3.The change in friction characteristics during clutch life. force that accelerates the testing without exceeding the maxi-mum force in road application,is applied to the clutch.As the clutch engages an overshoot of the force is prent yielding around20kN for thefirst1–2s before levelling out at15kN. The sliding velocity is always held constant at25rpm and the cooling oilflow between the friction surfaces is200ml/min.
Some of the discs have also been tested in order to determine the friction characteristics.In this test the load is20kN,and same oilflow as before.The velocity isfirst held constant at1rpm, and then the velocity is linearly incread from1to100rpm in 10s while the friction is monitored yielding the friction–velocity characteristics.This speed ramp is runfive times each at tem-peratures50,40and30◦C.
3.2.Changes in friction characteristics during clutch life
The friction characteristics change during the life of the clutch.After run-in,as en in Fig.3,the friction characteristics stabilizes and changes very little over sliding distance.
In order to verify that the friction discs play an important part in the clutch performance,a t of fresh friction discs and fresh lubricants were driven until shudder occurred.The fric-tion characteristics were measured,and after this the oil was changed.The clutch was run for another hour and the friction characteristics were measured once again.As the result in Fig.4 shows,the curves are quite similar,indicating that the friction material rather than thefluid is responsible for the change in fric-tion during clutch life.Thus the friction material has a vere impact on clutch life.A friction disc of the newer type lasted typically about32h in the test rig compared to about8h for the older material(e Figs.9and10).
4.Topography
4.1.Measurement method
Before measurement every disc has to be cleaned.This is per-formed by ultrasonic cleaning.A magnification of1.25allows measurement over a full square of the groove pattern on the
fric-
Fig.4.Friction characteristics before and after oil change.
tion disc while maintaining high accuracy.The scanning depth is t to well cover the8m depth required for data analysis.
汉堡包用英语怎么说
4.2.Method for recognizing the contact area
If the histogram curve describing the height distribution of an unworn surface of a sintered friction disc is known,it is possible tofind the lowest height that has been in contact with the counter surface.This is of great interest since every point higher than that actually has been subjected to wear.Still the measurements in this work are normalized to contain data points below the lowest contact height since it is believed that they affect friction characteristics.
The method is performed by trying tofind a point in the right hand side of the histogram curve which contains height information from at least three data points,everything higher than that can be considered noi and should not be taken into account.The next step is tofind a point in the histogram curve where the negative slope suddenly increa,this point is denoted ‘decreasing negative slope’in Fig.5,and a point clo to it where the negative slope decrea,this point is denoted‘Increasing negative slope’,going from right to left in Fig.5.The
heights Fig.5.Histogram curve with cursors enclosing the points that has been subjected to wear.
50P.Nyman et al./Wear261(2006)
46–52
Fig.6.The surface described in Fig.5before removal of data points.White spots contain no height information.
between the two points reprent the points on the surface that has only partly participated in the contact between the friction surfaces.The increasing slope is caud by the fact that all the summits higher than this point has been worn down and by that has been moved to this area in the histogram curve.The area where the slope decreas reprents the points on the surfaces that has been subjected to the most wear.The slope decrea becau the points that ud to be in this area has been worn down and thus been moved to a lower level in the histogram curve, i.e.the area between the two described points.Summarized the method assumes that the points to the right in the histogram curve when the disc is new moves to the left in the histogram curve when they are subjected to wear;this in turn caus the slope in the curve to change in two distinct points as described. Figs.6and7display a measured surface before and after u of the method described.
4.3.Investigated parameters
fuxinfernoAll of the parameters investigated are included in the Birm-ingham14parameters,a proposition of standardization of sur-face topography measurement.The parameters chon are S sk, surface skewness which describes the skew of the height distri-bution of the surface and is usually referred to as an amplitude parameter.This parameter describes the distribution of heights but it contains no information of where on the surface a cer-tain height is located.The next parameter is S dq or S d,as
it Fig.7.The same surface as in Fig.6after removal of surface points which have not been subjected to wear.White spots contain no height information.
only if
is referred to sometimes.It describes the average slope of the in a way it relates a specific height to its clost neighbours,and thereby it is dependant of lateral resolution since this determines the distance to the neighbours and the height resolution since this determines the smallest measurable height difference.In that manner it is similar to the next parameter S sc, the arithmetic mean summit curvature of a surface.S sc is highly resolution nsitive since it is only calculated for summits.Both S dq and S sc are often referred to as hybrid parameters.Thefinal parameter of this study is S bi,the surface bearing index,it is included in the investigation becau it claims to describe the load-bearing capacity of the surface,something that is consid-ered important.S bi is a so-called functional parameter.
4.4.Changes in topography during clutch life
The topography of the sintered friction lining changes during its life time,as shown in Fig.8.When new,it is very porous and rough as a result of the manufacturing method.The run-in process is characterized by its high wear rate,which is a result of a number of process,one of which is the high contact pressure applied to the highest summits.This results in wearing down of the high summits and fresh material coming into contact from below.The wear rate then decreas with the decreasing pressure applied to the summits caud by an increa of the real area of contact.This e
福田网页设计ventually leads to a state of constant wear rate, which is referred to as steady state.Both on the sintered
face Fig.8.New surface to the left and a worn surface to the right.
