https://ntrs.nasa.gov/search.jsp?R=19780013526 2019-01-24T18:06:28+00:00Z NASA ContractorR eport 2976 Ball to Separator Contact Forces in Angular Contact Ball Bearings UnderT hrust and Radial Loads Lester J. Nypan GRANT NSG3- 0 G 5 APRIL 1978 NASA I TECH LIBRARY KAFB, NM NASA Contractor Report 2976 Ball to Separator Contact Forces in AngularC ontactB allB earings Under Thrusta ndR adial Loads Lester J. Nypan CaliforniaS tate University, Northridge Northridge, California Prepared for LewisR esearch Center under Grant NSG-3065 National Aeronautics and Space Administration Scientific and Technical Information Office 1978 Table of Contents Pape Introduction 1 Separator Study Machine 2 TestB earings SeparatorF orceT ransducer ForceT ransducerC alibration Experimental Procedure 7 Data Analysis 8 Results and Discussion 11 Conclusion 16 References 17 Appendix I Computer Program 102 iii List of Figures Figure Page 1 SeparaStto urd y Machine 18 2 TesB te ariniSgn e paratoSr t udy Machine 21 3 Ball ContaFco trT c rea nsducer 22 4 Sample ForTc rea nsducCe ar libration 23 5 Example of Photographs Taken 24 6 BaCl lo ntaFc to rces at 4000 rpm 25 7 Ball ContaFc ot rces at 8000 rpm 43 8 Ball ContaFc to rces at 12000 rpm 64 9 Variatioon f Magnitude of Cage Forces 73 with Speed-Radial Loads 10 Variation of Location of Cage Forces 77 with Speed-Radial Loads 11 Variation of Magnitude of Cage Forces 81 with Speed-ThrustL oads 12 Cage toI n ner Race Land ContacF to rce 84 13 Cage tIo n ner Race Land Clearance 86 Variation with Location 14 Cage tSo h aft Speed Ratios 91 iv List of Tables Table Page Sp1eB cei afirci naTgtie osnt s 97 2 Speed, LoadSC paor nin ndSg t ants 98 3 SpCroi nnBgsa tea nandr ti sn gs Used 99 4 Calculated Number of ShRaf etv olutions 100 Between Photographs 5 MFoearscuer e Rda nges 101 V Introduction Currentd evelopments in jet engine technology are placing more stringent demands on gast urbined esign.T here is a constantlyi n- creasing requirement for engines to develop greater thrust outputs. In addition to this increased loading the need to raise the thrust/weight ratio of engines and to improve the fuel consumptionh as led to higher rotors peedsa ndo peratingt emperatures,l ighter componentsa nd corres- pondinglyi ncreaseds tructuralf lexibility.I na nticipationo ft o- morrow's requirements,f urther advanced knowledge ofe ngine component technology must be obtained.(l)* In the case of rolling contact bearings there is a need for a betteru nderstandingo fc age and rolling elementd ynamics, particularly in ultra-high speed applications.(2,3,4,5) Recently developed, advanced bearing theories have resulted in computerized optimization of rolling element bearing designs and in some casesa ccuratep rediction of bearingp erformance.T hesed evelop- mentsa nda dvances byn om eans substitute for testing of rolling element bearings which for many years was the basis for bearing development. To the contrary, the need for more refined data gathering methods hasb e- come obvious. Tests are needed tov erifyt het heories which form the foundationo ft hesec omputerp rograms. Also, performance tests and studies will alwaysb en eeded to refine bearing designs for critical applications. The interaction between the rolling elements of a bearing with the racewaysa nd separators is particularly difficult to measured ue -"" "Numbers in parentheses designate references at end of report. to the rapidity of their motion. The kinematicb ehaviora ndt he resulting forces acting on a rolling element/separator/raceway assembly could in the past be measured only by tests where the operating condi- tions were drastically simplified. Separator Study Machine An optical bearing test rig has been constructed to operate the bearing and make photographic records of the rolling elements and separator behavior. Figure la shows an overall viewo f the machine as it is presently installed in the Dynamics Laboratory of the Engineering Building at CaliforniaS tateU nT%ersity,N orthridge. The machine was originally assembled by IndustrialT ectonics,I nc., Compton, Calif., and has been used in industrial bearing research, and in a ball motion study reported in (6). The bearing test rig is basically a shaft supported by a pair of preloaded ball bearings at onee nda nd the test bearing at the other end. One face of the test bearing is exposed to allow free view of the balls and thes eparator. Radial load was applied to the test bearing by a hydraulic actuator through a cable loop over the bearing housing. The shaft was driven by a 75 hp hydraulic motor through a geared belt drive, giving speeds infinitely variable from 100 to 15,000 rpm. Figure lb is a schematic of the shaft assembly. Lubricating oil for under race cooling and test bearing lubrication is supplied through a series of orifices from the rear of the test bearing. Due to the high tangential velocities present when a bearing is rotated at shafts peeds up to l5,OOO rpm, conventionalp hotograph techniques are inadequate. The difficulty lies ino btaining photo- 2 graphs having sufficient resolution for analysis when very short ex- posures are required to freeze the motfon of the bearing elements. This problem has been overcome by eliminating the gross rotational motion using a derotationp rism. The resulting image thusp resentst hed ifferen- tial motionb etween the separator and the rolling element, enabling ob- servation and photographs to be made of an individual separator pocket. Derotation is accomplished by synchronizing the rotation of a Pechan prism at half-speed with the rotation of the ball separator, thus causing the apparent image rotatbn and the true separator rotation toc oincide. This results in the derotated imageo f thea reao fi nterest being imaged on the film planeo f a camera.L ightr aysf romt hei llumi- nated bearing are collected through the front window of the instrument. This window is optically coated to reject ultraviolet energy produced by ultraviolet lamps, serving the circuit for the prism speed control. From the window, the rays travel through collimator lenses and the Pechanp rismb eforet heyt ravelt hrought hee xitl enses.T heirp ath i s then deflected by mirrors which fold the image in different directions. Figure ICs hows light paths through the scanner. One light trace travels to a beam splitter wherea pproximately 15 % of the light is reflected to the eyepiece optics to provide an image observablet ot heo perator. The balance of lighte nterst he aperture of the pulse camera. Alignmenta nd positioning of the optical elements ensure that the eyepiece observes the same image quality andf ormat as that which the film sees. Another image is foldeda ndd emagnified in the transfer lens assembly before it enters the camera. A derotated image of the luminous painted segment of the test bearing separator is optically folded out and directed toward an 3 image splitting mirror swface wedce which proportions the light enterin5 two photomultipliert ubep hotocathodes. The electronics ignals from these tubes are used by the tracking system to control the prism speed. A trackings ystemh oldst he imageo f a selected point at the test bearings eparatori nt hef ieldo f view. It will accommodate a variation of up to 10% of a fixed ratio ofs eparator-to-shafts peedw ithout loss of the ability to lock onto the proper position within one revolution of the bearing retainer. It is necessary to sense the tracking error in angular position of thed erotationp rismt op rovide an input for thes ervos ystem.T his error signal is provided in the form of the difference in output of two multiplier phototubes. A sector on the bearing retainer is coated with fluorescent paint and illuminatedw ithu ltravioletl ight. An image of thes ectoro fa rc is formed in a plane containing the apex of a mirror surface wedge. The light striking the two surfaces of the wedge is reflected and illu- minatest hep hotocathodeo ft hep hotomultipliert ubes.R otation of the prism results in a displacemento ft he image and a consequenti n- crease in the output ofo ne tube and decrease in the output of the other. Electronic filtering is provided to discriminate between the steady signal due to the ultraviolet excitation of the phosphora nd any intermittent excitation due to strobe lamps to reducea nyi nteraction between the level of light striking the phosphora nd the error signal produced by the photomultipliers. Four Chadwick-HehuthS trobexl amps are flashed at the 16-frame- per-secondc amera rate to illuminate the separator and to stop the images of the protractors on the inner and outer races of the bearing so that angular position information is recorded on thep hotographs. 4