1、NATIONALADVISORYCOMMIITEEFOR AERONAUTICSTECHNICAL NOTE 4406LOW TIRE FRICTION AND CORNERING FORCES ON A WET SURFACEBy Eziaslav N. IIarrinLangley Aeronautical LaboratoryLangley Field, Va.WashingtonSeptember 1958-Provided by IHSNot for ResaleNo reproduction or networking permitted without license from
2、IHS-,-,-TECH LIBRARY KAFB, NMx.NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS I;lllllllllllll!lllllllllllllIICIL?18LTECHNZCAL NOTE 4406LOW TIRE FRICTION AND CORNERING FORCES ON A W13TSURFACEBy Eziaslav N. HsmrinSUMMARYAn exploratory investigationwas madeCommittee for Aeronautics to study typicalto dete
3、rmine the mechanisms by which waterby the National Advisorytire behavior on wet runways,on runways reduced tire forces,.“a.and to determine the extent of this tire-force reduction. A specially -constructed the treadmill served as a tire test vehicle which allowedeasy control of such pertinent ,param
4、etersas water depth on the treadwayor belt which served as the runway, tire inflation pressure, belt veloc-ity, braking load, yaw angle, and tire-tread pattern. A strain-gagebalance mounted on the wheel chassis of the treadmill measured both thebraking friction forces and cornering forces while tach
5、ometers recordedthe wheel and belt velocities. Measurements of these parameters weremade in testing a smooth-treaded and a diamond-treaded 3.(X) x 7 tire.Observations during the tests and evaluation of data have indicatedthat, under certain conditions of tire pressure, velocity, and waterdepth, the
6、smooth-treaded tire stops rotating and begins to plane evenwithout the application of brakes. For example, for a tire inflationpressure of 13* lb/sq in. gage ad a 0.09-inch water depth, this planingcondition occurred for the smooth tire at a velocity of 76 feet per sec-ond. The tire with the diamond
7、 tread behaved in a similar manner exceptthat, instead of stopping ccqletely, the wheel rotated at about 50 per-cent of the belt or treadmill speed. With both tires, the tire maximumand full-skid (lockedwheel) braking friction coefficients decreasedrapidly with increase in belt velocity, and for cer
8、tain operating con-ditions of the smooth tire the full-skid braking-coefficient values fellbelow that of the rolling friction. In smne of the cases for smooth tiresthe values of tire maximum braking friction coefficient dropped so lowthat the free-rolling friction became the maximum friction force.
9、Underthese conditions any slight and momentsxy braking forced the tire intoa stable full-skid condition.Cornering-force coefficients of both the smooth- and diamond-treadtires also decreased with belt velocity and reached nearly zero valuesat planing velocities. Braking had the effect of greatly dec
10、reasing thecornering-force coefficients even at relatively slow velocities.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-2 NACA TN 4406.INTRODUCTIONLA number of accidents have been reported in which fighter aircraftskidded off the runway in landing
11、 during, or immediately after, a heavyrain. In these accidents the braking and directional control werereported as extremely poor. It appeared that in these cases relatively thick water layers and high speeds were primary factors in the loss oftraction of the tires on the runway surface. Several inv
12、estigationshave been made of tire friction coefficients on wet surfaces (refs. 1to 6) but, in general, the test conditionswere limited in speed, ordegree of surface wetness, or both. The present investigationwas there-fore undertaken to study the effects of water depth, speed, and otherfactors on we
13、t-surface tire characteristics. Tests were made with smallsmooth-treadedand diamond-treaded tires (size 3.(X)x 7) running on atreadmill or endless belt apparatus. The measurements covered speedsfrom 23 feet per second (approximately 14 knots) to 94 feet per second(approximately55 knots), tire inflat
14、ionpressures of 6 to 30 pounds persquare inch, and water depths on the treadmill of 0.02, 0.06, and 0.09 inch. Measurements of braking.tire friction.and cornering forcewere made at yaw angles of 0 and 4.*APPARATUS AND TESTS“The equipment was designed to be as small and simple as possibleconsistentwi
15、th the requirement of reproducing the very poor brakingconditions on wet runways reparted by various agencies. Elementarytheoretical considerations suggested that the poor braking conditionresults from a wedge of water forcing the tire away from contact withthe runway and that the major parsmeter in
16、 this condition is the ratioof the square of speed to the tire inflationpressure. Accordingly, from tire characteristicsand landing speeds involved in the poor brakingincidents, it was estimated that with a commerciallyavailable 3.00x7(2-ply rating, approximately 12-inch outside diameter) tire with
17、an infla-tion pressure of 13 pounds per squme inch, belt speeds and water speedsof 70 to 90 feet per second should be sufficientto produce the very lowfriction condition. This speed requirement and the small tire size madeit possible to use commercially available belts and pulleys for thetreadmill a
18、nd normal fire-hydrantwater pressure.The treadmill (figs. 1 and 2) consisted of a comerchlly available10-inch wide, five-ply rating, power transmissionbelt that ran over12-inch-diameterpulleys. Along the center line of the belt on the pathof the tire, a l-inch-wide coat of bonding agent impregnatedw
19、ith sandwas laid to approximate the texture of a concrete surface. Water wasapplied in a smooth sheet across the entire width of the sanded surfaced*Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN ti06 3by means of a specially designed and bu
20、ilt water nozzle in order tosimulate standing water on a runway. The depth of this sheet of waterwas not measured as such. It was assumed to be equivalent to the thick-ness of shim stock that could be slipped in between the upper water noz-zle block (whose height was adjustable) and the sanded belt
21、surface whichserved as the lower water nozzle block. The exit end of the nozzle blockwas located within inchesof the tire so that the water depth wasassumed to be the same at the tire as at the water nozzle exit. A gatevalve located in the water feed pipe controlled the water pressure sothat the wat
22、er velocity at the nozzle exit matched the belt speed. Thewater velocity at this point was determined by means of a pitot tubelocated in the sheet of water at a position approximately 1/2 inch down-stream of the nozzle exit.The braking and cornering forces were measured by a strain-gagebslance which
23、 consisted of a vertically mounted beam (fig. 3) withbending-moment strain gages bonded onto its four faces. The brakingand cornering forces were measured in the plane and normal to the planeof the wheel, respectively. The data frmn the strain gages were recordedon photographic film by means of stan
24、dard NACA recording galvanometers.Tire braking was obtained by means of a small hydraulically operatedaircraft tisk brake whereas cornering forces were obtained by rotatingthe wheel carriage and strain-gage balance about the vertical axis toprovide a 4 yaw angle. Wheel speeds and belt speeds were ob
25、tained fromtachometer generators mounted on the axles of the test wheel d on theidler pulley of the belt. Lag characteristics of the tachometer genera-tor smd recorder systems were such that accurate wheel velocities duringdeceleration were not obtaimble so that data could not be presented asa funct
26、ion of slip ratio. Standard WJA tachomter recorders were usedto record velocities.In general, the data in this paper were obtained with the tiresloaded to 100 pounds. The static tire footprint characteristicswhichwere obtained frcm imprints of csrbon-smesred tires upon a white cud-board surface for
27、the two tire types under this load condition ard pre-sented in figure 4. At this load condition the recommended inflationpressure was 13 lb/sq in. gage. Some exceptions to this load conditionwere a few tests with the diamond-treaded tire loaded with 50, 70, 90,and 100 pounds of weights at three tire
28、 inflation pressures and with abelt velocity of about 82 feet per second. These exceptions sre notedin this paper.The smooth tire was tested at yaw angles of either 0 or 4, inwater depths of 0.02 inch and 0.09 inch for each angle, and at threetire inflation pressures for each water depth (approximat
29、ely7, 13,and 28 pounds per square inch).Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-4 NACA TN 44-06.All diamond-treaded-tiretests were made at one of two yaw angles(0 or 4). At the 0 yaw angle, the tire was run in 0.02-, 0.06-, and0.09-inch depth
30、s of water on the belt for each of the following tireinflation pressures: ( (half normal 6 to 7lb/sq in.), normal 13to.16 lb/sq in.) ( ), twice normal 2to 30 lb/sq in. . At the 4 yaw angle,the tire was run in 0.02- and 0.09-inch depths of water. For the testsin which the tire operated at a 4 yaw ang
31、le and in 0.02 inch of water,only the tire inflation pressure of l% lb/sq in. was used. As the result -of damage incurred by the equipment during this test, the programed testsat tire inflation pressures of 7 and 28 lb/sq in. gage were not made.Tests with the tire running at a yaw angle of 4 and in
32、0.09 inch ofwater were made at tire inflation pressures of 7, 13, and 28 lb/sq in.gage.Belt velocities up to 93 feet per second (55 knots) were attained.However, some tests were limited to lower speeds since the available water pressure was insufficient at times to have matching water and beltveloci
33、ties at higher belt speeds. .For the most pert, the instrument accuracies are believed to be ofthe order of 3 percent. Sample records are shown in figure 5. *DEFINITIONS OF COEFFICIENTSThe coefficients used in this paper are defined as follows:Maximum braking friction coefficient: coefficient obtain
34、ed bydividing the maximum braking force at a given velocity by thevertical forceFull-skid or locked-wheelbraking friction coefficient: coeffi-cient obtained by dividing the friction force of a nonrotatingwheel at a given belt velocity by the vertical forceRolling friction coefficient: coefficient ob
35、tainedby dividingthe friction force of a freely rotating wheel with no brakingby the vertical force . .JRree-rollwheel velocity: velocity of an unbraked wheel as obtainedby multiplying the wheel rotational speed by the measured tirerolling radiusJCornering-forcecoefficient: coefficient obtained by d
36、ividingthe side force, measured Perpendicti= to the pne of the wheel,by the vertical force.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN MC6 5All the coefficients are based on the vertical force and not onthe static load (weight on wheel).
37、The reason for this is that thepivot-point location of the wheel tow bar is such that the friction forceproduces a load-reducing moment on the tire. A sumation of the staticload and friction force moments about the pivot point determines thevertical force. This force is given by the formulaFv=W- Ff(
38、h/l)whereFv vertical forcew static loadFf friction forceh height of pivot point above tire-belt contact point1 distance of wheel frmn pivot pointPRESENTATION OF RESULTSMaximum braking friction coefficients, full-skid or locked-wheelbraking fhiction coefficients, free-roll friction coefficient, and f
39、ree-roll wheel velocity =e presented in figures 6 to 14 as functions ofbelt velocity. Figures 6, 8, 10, and 13 and figures 7, 9, XL, and 14are paired together to present data at generally the same conditionsbut at 0 and 4 yaw angles, respectively. Figure 12 presents data at0 yaw angle; there is no c
40、mrparable figure at 4 yaw angle. The alter-nate figures have two additional parameters - cornering-fmce coeffi-cients at free roll and at maximum braking - also plotted as functionsof belt velocity. Each of the aforementioned figures presents hforma-tion obtained from a tire operated in a particular
41、 water depth and inmost cases at three tire inflation pressures - approximately 7, 14,and 28 lb/sq in. gage. However, because of equipment -ge, the dawtaken for the tests shown in figure 11 are for the tire inflation pres-sure of 1* lb/sq in. gage only. Data taken at the two water depthsof 0.02 and
42、0.09 inch used in tests of smooth tires are shown in fig-ures 6 snd 7 and figures 8 and 9, respectively. Dismond-tre=led tiredata taken at three water depths (0.02, 0.06, and 0.09 inch) are shownin figures 10 and 11, figure 12, sad figures 13 and 14, respectively.Figures 15 and 16 present the variat
43、ion, which was obtaed at a4 yaw sm.gleand at three different speeds (approximately27, 49,71 ft/see), of the cornering-force coefficientwith braking frictionProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-6 MCA TN 4406force coefficient. Figure 15 pres
44、ents these results for the smooth tireoperated in 0.02- and 0.09-inch water depths and at three tire inflationpressures, whereas figure 16 presents data obtained at the same condi-tions as figure 15 but with a diamond-treadedtire substituted for thesmooth tire. However, for reasons explained previou
45、sly, only thedata for a tire inflation pressure of l% lb/sq in. are available for the0.02-inch water depth.Varying the weight of the static load affects the friction coeffi-cient of the diamond-treadedtire in the manner shown in figure 17.DISCUSSIONSmooth-TireFriction Coefficients“b,An examination o
46、f the results presented in figures 6 to 9 for thesmooth tire operated in either thin or thick films of water shows that,as the belt velocity increased, the values of maximum braking coeffi-cients decreased until at certain high speeds the values either approachedor equaled coefficients of a freely r
47、olling tire (0.05 to 0.0!3). At these -“certain high speeds” of the freely rolling tire, the spinning stoppedwithout any brake application. l%om this it is apparent that water pres- “sure at these condi.ti.onscaused a complete separation of the tire fromthe belt and created torques which stopped tir
48、e spinning. This condition -w be referred to as “tire-planing”and appesrs to be a stable condition.Figures 6 to 9 also show the rapid decrease in values of full-skid(locked-wheel)braking friction coefficient with increasingbelt velo furthermore,not only are thesevalues low for a tire in full skid bu
49、t they are also so low that theydrop below those for the tire in free roll. It was found that underthese friction-coefficientconditd.onsand at belt speeds below that oftire planing, the smooth tire would either not accelerate from a stoppedcondition, or would accelerate very slowly. The importance of thisphenomena lies in the fact that-a nonspinning tire, such as that on a .-landing airc
