車輛工程外文翻譯-測(cè)試工業(yè)制動(dòng)器襯片摩擦特性【中文3700字】【PDF+中文WORD】,中文3700字,PDF+中文WORD,車輛,工程,外文,翻譯,測(cè)試,工業(yè),制動(dòng)器,摩擦,特性,中文,3700,PDF,WORD TECHNICAL ARTICLETesting the Friction Characteristics of Industrial Drum BrakeLiningsJ.Van Wittenberghe,W.Ost,and P.De BaetsDepartment of Mechanical Construction and Production at Ghent University,Ghent,BelgiumKeywordsDrum Brake,Friction,Testing,Coefficient ofFriction,TemperatureCorrespondenceJ.Van Wittenberghe,Department of Mechanical Construction andProduction at Ghent University,Ghent,BelgiumEmail:Jeroen.VanWittenbergheUGent.beReceived:December 7,2009;accepted:August 30,2010doi:10.1111/j.1747-1567.2010.00675.xAbstractIn the present study a new brake setup was developed to test drum brake liningson an industrial brake with drum diameter of 30?.During the tests performedon the setup,the brake undergoes a series of cycles in which the drum isslowed down from service speed to standstill.In each cycle the same amount ofenergy is dissipated as during a realistic safety stop.This was obtained by addinga flywheel in the setup so that the systems kinetic energy at service speedmatches the energy of the hoisting system dissipated during an emergencystop.Two different brake lining materials were characterized.Both materialswere subjected to two test series to study the changes in coefficient of frictionover a number of cycles.It was observed that the coefficient of friction of bothlinings was dependent on the drum temperature.The coefficient of friction ofthe first material decreased with increasing drum temperature,while the latterhad the opposite behaviour.IntroductionSpring applied,electrically released drum brakes areused in industrial environments,such as steel mills,tocontrol the movement of travelling cranes as well asthe hoisting apparatus of the crane.Such cranes aretypically powered by an electromotor,but althoughthe hoist motors are normally geared to producegreater torque and reduce the output speeds to anacceptable level for lifting and lowering heavy objects,it remains nevertheless possible for the motor to bedriven by a heavy object in case of an electrical failureduring lifting.This dangerous situation is referred toas block drop.To stop the motor in case of blockdrop,spring applied,electrically released drum brakesare used.These brakes contain heavy springs whichpush the brake shoes against a drum that rotateswith the motor or the transmission output shaft.Toretract the springs,a built-in electric solenoid has tobe powered.The solenoid is generally wired in themotors electrical circuit,so when power is lost tothe motor,the solenoid also loses power allowing thesprings to thrust the brake shoes against the drumand hence preventing the motor to turn freely.Whenblock drop appears,the drum brake is closed,stoppingthe lifted load to fall down and keeping it at its height.But before trying to solve the failure of the electricalpower circuit of the crane,it is important to putthe load safely on the ground.Normal procedure isthen to use a backup circuit with manual control toopen the brake for a moment.To prevent a too fastrate of descent,the brake is closed after a moment,stopping the load again.These actions are repeatedseveral times until the load is lowered completely.During this procedure,the drum brake material isseverally put to the test because total load has to beslowed down repeatedly without the help of hoistingapparatus traction.The drum brakes braking power depends not onlyon the force applied by the springs,but is alsodetermined by the frictional properties between thematerial used in the braking shoes and the drumof the brake.The behaviour of this friction materialduring its service life has to be known because a lackof friction can cause the brake to slip due to heavyloads.Nevertheless,a coefficient of friction(COF)thatis too high can overload the drum axle and can causehigh drum temperatures and high dynamic loads onthe drum which can lead to cracks at the drum surfaceExperimental Techniques 36(2012)4349 2010,Society for Experimental Mechanics43Friction of Drum Brake LiningsJ.Van Wittenberghe,W.Ost,and P.De Baetsor even drum fracture.Nowadays,a wide range offriction materials is available,but as is known fromZhang and Wang1the behaviour of those material ishighly dependent on their composition and serviceconditions.Through a series of tests on small-scalesamples,they found the friction performances andwear resistance of the same material to be changingwith load,sliding speed,and temperature.In anotherstudy2they showed that also the drum material canhave an impact on the tribological behaviour of thebrake because of changes in specific heat capacityand thermal conductivity.Hence when new brakematerials are developed,it is still necessary to performexperimental tests to characterize the lining materialin combination with the drum material.In additionto this it is known that the pressure distribution is notevenly spread across the surface of brakes due to bothgeometrical deviations of drum and brake shoes anddynamic effects.This means that extrapolations ofresults of small scale tests on friction material cannotalways be used to make reliable predictions on thebehaviour of the full-scale brake.Hence in most casesfull-scale tests are the only option to get accurateinformation about the performance of the brake.Full Scale Test SetupPrinciples of the drum brake setupDuring previous studies,setups were developedmainly to quantify the frictional behaviour duringcontinuous braking.3In that case,the local fric-tion intensity of an imaginary friction lining segmentchanges during braking.This process is called ther-moelastic instability(TEI)and causes,over a criticalspeed,a steady-state regime with harmonic changesin friction.The TEI can be predicted accurately byfinite element analyses.4However,in the case ofblock drop and the procedure of safely lowering theload afterwards,the transient regime is the regionof interest because the steady-state regime is notreached.For this purpose,a new setup was designedto simulate the block drop situation in a better way.In the new setup the brake undergoes a series ofcycles in which the drum is slowed down from servicespeed to rest.Of course to have a realistic situation,there should be an equal amount of energy dissipatedduring one cycle as in a real safety stop.To obtainthis,the systems mass moment of inertia was chosenin such a way that the kinetic energy of the system atservice speed would match the maximum energy tobe dissipated during an emergency stop.In the following paragraphs,firstly,the test setupdetails are presented together with a calculatingmethod to obtain the COF.Later the test data of thetwo different brake lining materials will be discussed.Test setup descriptionSchematicdrawingsofboththefrontalandthesectionview of the setup are shown in Figs.1 and 2.A viewof the total setup is given in Fig.3.The setup consistsof a spring applied and electrically released Igranicsafety brake type M 30?,whose drum(1)is drivenby an electrical DC compound 100 kW(at 5000 rpm)motor(17).The braking force is applied by the spring(4)that pushes the brake shoes(2)against the drum.Different friction lining materials(3)can be mountedin the brake shoes to test their behaviour during brak-ing.The braking pressure can be set by adjusting thespringcompressionwiththebolt(5)andcanbevariedbetween 0 and 16.6 N/cm2.The latter corresponds toa maximum braking torque of approximately 10 kNmfor a COF between the drum and the friction liningof 0.6.To open the brake the solenoid(6)is poweredpulling part(7)to the left and compressing the spring.To obtain a system that contains enough kineticenergy to simulate a realistic block drop situation,a drive wheel(8)is added to increase the inertia ofthe system.Drum(1)and drive wheel(8)are carriedby the main axle(10).The drive wheel is connectedto the main axle using two locking assemblies(9).The main axle is supported by two self-aligning ballbearings(11)and is connected to the DC motor by aflexible jaw coupling(12).Drum and drive wheel have the same diameter of30?or 760 mm.The moments of inertia of the differ-ent rotating parts of the setup are given in Table 1.Drum,drive wheel,and main axle are the parts thatcontribute the most to the moment of inertia of thesystem.Since a jaw coupling is used,the rotor of theDC motor rotates with the other rotating parts of thesetup and its inertia of 6 kgm2has to be taken intoaccount.This gives the setup a total moment of iner-tia of 95.1 kgm2resulting in a total kinetic energyof 422 kJ at the service speed of 900 rpm.Becausethe total brake shoe area is 0.28 m2,the mean energydensity during each braking cycle is approximately1500 kJ/m2.In a previous study by Severin5a brakewith a 25?drum dissipating 168 kJ during each brak-ing cycle starting from a service speed of 900 rpmwas used,giving an energy density of approximately1100 kJ/m2.Hence the setup of this study is able toapply a much higher energy density into the materialstarting from the same service speed.During a braking cycle,the drum and the drivewheel are brought up to service speed,while thebrake is open.Once the speed of 900 rpm is reached,44Experimental Techniques 36(2012)4349 2010,Society for Experimental MechanicsJ.Van Wittenberghe,W.Ost,and P.De BaetsFriction of Drum Brake Linings1FGe324567131415abFL0.500mMBFigure 1 Schematic front view of the drum brake setupto the motor1112111810916Figure 2 Schematic section view of the drum brake setupExperimental Techniques 36(2012)4349 2010,Society for Experimental Mechanics45Friction of Drum Brake LiningsJ.Van Wittenberghe,W.Ost,and P.De Baets18117813Figure 3 Drum brake setupTable 1 Properties of the rotating parts of the setupPartMass(kg)Inertia(kgm2)MaterialDrum32028.8Cast ironDrive wheel70056.7Structural steelMain axle603.642CrMo4 alloy steelCoupling90.01Steel+elastomer spiderTwo locking assemblies50.02Steelthe power of the motor is switched off and the brakeis closed.When finally the drum has come to rest,thebrake is opened again and the cycle repeated.During the tests,the rotational speed was measuredusing a tachometer mounted on the motor and thesurface temperature of the drum was continuouslymeasured using an SP i-tec 2005D infrared sensor(see(18)in Fig.3).The control of the systemand measuring of all signals are carried out by acomputer with a Texas Instruments BNC-2110 dataacquisition card and a Labview programme.Speed,surface temperature and force in the loadcell wererecorded at a frequency of five samples/second.In order to measure the brake torque,the brakeis mounted on two inclined surfaces(13)and(14),as can be seen in Fig.1.These two supports aremanufactured in the way that the supporting surfacesare perpendicular to the two construction lines aand b.As the drum rotates in the counter clockwisedirection,the reaction force on the support(14)canbecome negative.To counter this force the part(15)is present,whose contact surface is parallel to thecontact surface of(14).A loadcell(16)with a capacityof 20 kN is mounted 500 mm below the centre ofrotationofthedrum.Duringbrakingthebrakewilltryto rotate with the drum.The loadcell will prevent thisfrom happening and will apply a force FL(N).Becausethe loadcell is rigid,the actual rotation is very smalland the position of the brake on the inclined surfaces(13)will not change significantly.Hence the reactionforcesinthesupportsstayalignedwiththeconnectionlines a and b in Fig.1.This means the vector of thereaction forces goes through the centre of rotation ofthe drum and the reaction forces do not contribute tothe torque equilibrium around this point.Calculating the coefficient of frictionThe COF can be calculated from the applied brakingtorque MB,which can be calculated from the forcemeasured by the loadcell FLby expressing the torqueequilibrium around the centre of the drum(Fig.1):MB=FL0.500 FGe(Nm)(1)with FG(N)the gravitational force of the brake and e(m)the eccentricity of the centre of mass to the centreof rotation of the drum.The gravitational force of thebrake is constant and because the actual rotation ofthe brake is very small,the eccentricity can also beconsidered constant.When the brake is open,nobraking moment is applied,but due to the eccentriccentre of mass of the brake,there is still a force appliedon the loadcell.For this case(MB=0)Eq.1 becomesFL0.500=FGe(Nm)(2)where FLis a measured value.By this way avalue for FGe of 3136 Nm was found.With themass of the brake of approximately 1 tonne,anestimated eccentricity of 0.31 m was obtained.Inthe calculations only the product FGe is used.Theestimated value of the eccentricity is only mentionedas an illustration.From the braking torque MB,calculated from Eq.1,the COF can be calculated as explained in thefollowing section.As is schematically shown in Fig.4,the brakingpressure p(N/m2)multiplied by the COF,integratedover the surface of the brake shoes equals the brakingtorque MB:MB=2b?rprd(Nm)(3)The factor 2 in Eq.3 results from the two brakeshoes that are present.Equation 3 can be simplifiedtoMB=4br2p(Nm)(4)46Experimental Techniques 36(2012)4349 2010,Society for Experimental MechanicsJ.Van Wittenberghe,W.Ost,and P.De BaetsFriction of Drum Brake Liningsa-arpmpFigure 4 Schematic view of the pressure in the brake shoeHence=MB4br2p(5)with b the width of the brake shoes(0.300 m),r the radius of the brake drum(0.380 m),p themean braking pressure during the tests(8.1 N/cm2=8.1104N/m2)and the half angle of one brake shoe(35or 0.611 rad).With the above values Eq.5 becomes=MB(Nm)8574(Nm)(6)Course of a braking cycleDuring each braking cycle,the drum and the drivewheel were brought up to 900 rpm.This took about90 s.Once the drum was at the required speed,dataacquisition started and 2 s later the brake was closed.Two seconds after the drum stopped,data acquisitionwas interrupted and the break opened again,afterwhich the cycle restarted.In order to control thedataflowandavoidrecordingexcessdata,dataloggingwas interrupted when the drum was brought up toservice speed.In Fig.5,the course of a braking cycle is shown.For this cycle the braking time is 2.2 s,in whichthe braking speed is brought from 900 rpm torest.The course of the braking torque is somehowdifferent from what one could expect from small-scale material tests.Common frictional behaviourof braking materials includes a difference in staticand dynamic COF,from which we could expect thebraking torque to have a peak when the brake is010002000300040005000-101234Time sSpeed rpmTorque Nm020406080100Temperature CSpeedTorqueTemperaturebrakingFigure 5 Measured signals during one braking cycleclosed and remain constant until the drum is broughtto a halt.In Fig.5,however,it can be observedthat the braking torque increases linearly for about1.4 s after which the torque reaches a more orless stable value.This linear increase is caused byelectromagnetic effects in the solenoid(6)in Fig.1)of the brake.When the current over the solenoidis removed,the force of the spring(4)in Fig.1)isnot immediately applied on the braking shoes.Dueto the solenoids self-induction,the original magneticfield only decreases gradually and hence,the brakingtorque is applied over a certain period of timeinstead of instantaneously.In this cycle the maximumbraking torque is 4045 Nm,from which a COF of=0.47 can be calculated according to Eq.6.Thedrum temperature increases here from 27C beforethebrakingtoamaximumof47Cduringthebraking.Experimental TestsIn following sections the results of the test seriesperformed on two different composite brake liningswith a different composition is presented.Bothmaterials were subjected to two test series on thenew setup.First,a short test series was conducted,where the objective was to test until the mean surfacetemperature of the drum saturated.The short testseries was stopped after 50 cycles.Second,a long testseries was conducted,consisting of 250 successivecycles to study the integrity of the lining materialwhen subjected to a high number of braking cycles.The conducted tests are summarized in Table 2.Thenoted numbers for the materials and tests will be usedaccording to this table in the rest of this article.Testseries 1 and 3 are the short test series,2 and 4 are thelong series.Experimental Techniques 36(2012)4349 2010,Society for Experimental Mechanics47Friction of Drum Brake LiningsJ.Van Wittenberghe,W.Ost,and P.De BaetsTable 2 Summary of the tests short test seriesMaterial 1Material 2TestSeries 1TestSeries 2TestSeries 3TestSeries 4Number of cycles5025050250Final speed(rpm)900900900900Environmenttemperature atstart(C)22.520.421.020.8Drum temperatureat start(C)31.222.827.221.1Mean drumtemperature atend(C)63.664.869.964.9Coefficient of frictionlast cycle0.470.490.350.31Short test seriesThe results for the short test series of materials 1and 2 are shown in Figs.6 and 7.For both materials,the COF together with the minimum,maximum,andmean temperatures are plotted as a function of thecycle number.For both materials it can be seen thatthe mean temperature saturates at about 65C afterapproximately 30 cycles.At this point the minimumand maximum temperatures are also saturated,witha minimum drum temperature of about 50C forboth materials.The maximum drum temperaturesare different for both materials,as can be seen inFig.6,the maximum drum temperature with liningmaterial 1 can reach peak values of about 118C,while only 104C for lining material 2(Fig.7).Thisdifference is caused by the difference in COF betweenthe two materials.The COF of material 1 is higherthan that of material 2,which means the brakingtime will be shorter for material 1.Consequently,thesame amount of kinetic energy has to be transferredfrom the drum to the friction lining in a shortertime,resultinginhigherpeaktemperatures.However,because the actual braking time(about 2.5 s)is shortin comparison to the total cycle time of about 96 s,the minimum and mean drum temperatures for bothfriction linings are practically the same.Additionally,it can be observed from Fig.6 thatthe COF shows a slight increase with increasingtemperature:the COF started at a value of 0.44 fora mean drum temperature of 36.0C and increasedto a value of about 0.47.The opposite behaviourwas observed for material 2(Fig.7).Here the COFdecreased with increasing drum temperature:at startCOF=0.47 and the mean drum temperature was27.2C,while the COF=0.35 after 50 cycles.02040608010012001020304050Number of CyclesTemperature C0.000.100.200.300.400.500.60Coefficient of Friction-Min.TemperatureMax TemperatureCoefficient of FrictionMean TemperatureFigure 6 Coefficient of friction and temperatures during test series 1(short)on material 102040608010012001020304050Number of CyclesTemperature C0.000.100.200.300.400.500.60Coefficient of Friction-Min.TemperatureMax.TemperatureCoefficient of FrictionMean TemperatureFigure 7 Coefficient of friction and temperatures during test series 3(short)on material 2Long test seriesDuring the long-term tests,the temperature depen-dency of both materials was confirmed.In Fig.8,theresults of the long test series on material 1 are shown.Again it can be seen that the COF increases withincreasing drum temperature.It is noted that at cycle25 a short interruption took place during which thedrum temperature dropped to about 8C.This tem-perature drop is also clearly visible as a drop in thepath of the COF at this cycle.In Fig.9,the results of the long test series on mate-rial 2 showacleardecreaseoftheCOF with increasingdrum temperature.Even though the most important change in drumtemperature and COF for both lining materials tookplace during the first 30 brake cycles,a small changeappeared during the subsequent cycles,resulting in aslight COF increase for material 1(to 0.49 at cycle250)and a slight COF decrease for material 2(to 0.31at cycle 250).48Experimental Techniques 36(2012)4349 2010,Society for Experimental MechanicsJ.Van Wittenberghe,W.Ost,and P.De BaetsFriction of Drum Brake Linings020406080100120050100150200250Number of CyclesTemperature C0.000.100.200.300.400.500.60Coefficient of Friction-Coefficient of FrictionMean Temperature