SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Process consisting of grinding raceways to oversize but original quality condition and installing new oversize balls or bearings restores wornout ball and roller bearings to original quality, thereby doubling their operating life. Evaluations reveal process results in restoration of 90% of replaced bearings at less than 50% of new-bearing costs.
The elastic distortion of the inner ring of an experimental roller bearing is investigated analytically. The geometry of the experimental bearing is unusually complex, and for this reason a bearing with an axially symmetric inner ring is also analyzed. Only the inner ring and shaft are considered using a two dimensional finite element program that can accommodate interference between these components. The results suggest that the elastic distortions are modest in relation to the design clearances associated with the experimental bearing. The variation of the radial deflection of the raceway may, however, be significant in some circumstances.
The elastic distortion of the inner ring of an experimental 3,000,000 DN roller bearing is investigated analytically. The geometry of this bearing is unusually complex and for this reason a bearing with an axially symmetric inner ring and shaft is also analyzed. Only the inner ring and shaft are considered using a two-dimensional finite element computer program which enables interference between these components to be accommodated. The results for the experimental bearing suggest that elastic distortions are modest in relation to the design clearances. However, the variation of the radial deflection of the raceway may be significant for some circumstances and the interference fit adopted between the ring and shaft appears to be questionably low.
The CYBEAN (Cylindrical Bearing Analysis) was created to detail radially loaded, aligned and misaligned cylindrical roller bearing performance under a variety of operating conditions. Emphasis was placed on detailing the effects of high speed, preload and system thermal coupling. Roller tilt, skew, radial, circumferential and axial displacement as well as flange contact were considered. Variable housing and flexible out-of-round outer ring geometries, and both steady state and time transient temperature calculations were enabled. The complete range of elastohydrodynamic contact considerations, employing full and partial film conditions were treated in the computation of raceway and flange contacts. Input and output architectures containing guidelines for use and a sample execution are detailed.
The operating characteristics of large diameter rolling-element bearings in the ultra high speed regimes expected in advanced turbine engines for high performance aircraft were investigated. A high temperature lubricant, DuPont Krytox 143 AC, was evaluated at bearing speeds to 3 million DN. Compared to the results of earlier, similar tests using a MIL-L-23699 (Type II) lubricant, bearings lubricated with the high density Krytox fluid showed significantly higher power requirements. Additionally, short bearing lives were observed when this fluid was used with AISI M50 bearings in an air atmosphere. The primary mode of failure was corrosion initiated surface distress (fatigue) on the raceways. The potential of a case-carburized bearing to sustain a combination of high-tangential and hertzian stresses without experiencing race fracture was also investigated. Limited full scale bearing tests of a 120 mm bore ball bearing at a speed of 25,000 rpm (3 million DN) indicated that a carburized material could sustain spalling fatigue without subsequent propagation to fracture. Planned life tests of the carburized material had to be aborted, however, because of apparent processing-induced material defects.
Corrosion tests were run with AISI 52100, AISI M-50 and AMS 5794 under conditions that simulate the crevice corrosion found in aircraft ball and roller bearings rejected at overhaul for corrosion. Test specimens were fabricated that simulated the contacts of balls or rollers and the raceways. Corrosion cells were assembled in the presence of a lubricant contaminated with water and chloride ions. The cell was then thermally cycled between 339 K (150 F) and 276 K (37 F). The corrosion observed after 14 cycles was that of crevice and pitting corrosion typically found in aircraft bearings. AMS 5749 showed a very slight amount of corrosion. No appreciable differences were noted between AISI 52100 and AISI M-50, but both showed much greater corrosion than AMS 5749. The corrosion pits observed in AISI M-50 appeared to be fewer in number but generally deeper and larger than in AISI 52100.
Endurance tests with ball bearings made of corrosion resistant bearing steel which resulted in fatigue lives much lower than were predicted are discussed. Metallurgical analysis revealed an undesirable carbide distribution in the races. It was shown in accelerated fatigue tests in the RC rig that large, banded carbides can reduce rolling element fatigue life by a factor of approximately four. The early spalling failures on the bearing raceways are attributed to the large carbide size and banded distribution.
Parametric tests were conducted with a 35-mm-bore, split-inner-ring ball bearing with a double-inner-land-guided cage. Provisions were made for through-the-inner-ring lubrication. Test condictions were either a thrust load of 667 N (150 lb) or a combined load of 667 N (150 lb) thrust and 222 N (50 lb) radial, shaft speeds from 32000 to 72000 rpm, and an oil-inlet temperature of 394 K (250 deg F). Outer ring cooling was used in some tests. Tests were run with either 50 or 75 percent of the total oil flow distributed to the inner-ring raceway. Successful operation was experienced with both 50% and 75% flow patterns to 2.5 million DN. Cooling the outer ring had little effect on inner-ring temperature; however, the outer-ring temperature decreased as much as 7% at 2.5 million DN. Maximum recorded power loss was 3.1 kW (4.2 hp), and maximum cage slip was 8.7 percent. Both occurred at a shaft speed of 72000 rpm, a lubricant flow rate of 1900 cu/min (0.50 gal/min), a combined load, and no outer-ring cooling.
Based on analytical calculations assuming isothermal elastohydrodynamic (EHD) lubrication conditions, and on curve-fitting of a thermal correction factor, a new formula is proposed to calculate the sliding traction force developed in a concentrated contact, assuming a nonlinear, viscoelastic lubricant and undirectional sliding. When sliding occurs in the rolling and transverse directions (bidimensional sliding), a simplified numerical method is outlined by which the components of the local shear stress are quickly calculated. The latter method is applied to the calculation of the shear stress and temperature distribution in the ball-raceway contacts of an angular-contact ball bearing. Because of the viscoelastic behavior of the lubricant, a nonzero lateral traction force is obtained.
Crack propagation in a rotating inner raceway of a high speed roller bearing is analyzed using the boundary integral equation method. The method consists of an edge crack in a plate under tension, upon which varying Hertzian stress fields are superimposed. A computer program for the boundary integral equation method was written using quadratic elements to determine the stress and displacement fields for discrete roller positions. Mode I and Mode II stress intensity factors and crack extension forces G sub 00 (energy release rate due to tensile opening mode) and G sub r0 (energy release rate due to shear displacement mode) were computed. These calculations permit determination of that crack growth angle for which the change in the crack extension forces is maximum. The crack driving force was found to be the alternating mixed-mode loading that occurs with each passage of the most heavily loaded roller. The crack is predicted to propagate in a step-like fashion alternating between radial and inclined segments, and this pattern was observed experimentally. The maximum changes DeltaG sub 00 and DeltaG sub r0 of the crack extension forces are found to be good measures of the crack propagation rate and direction.
On 24 January 1985, the four Solar Maximum Mission Reaction Wheels, S/N 102 through 105 were returned to Sperry for postflight analysis. The analysis consisted of visual examination; preliminary electrical checks; performance testing at ambient, hot, and cold temperature environments; and internal pressure measurements. Based on the performance test results and past utilization in space, one reaction wheel was selected for teardown to study lubricant distribution, bearing and reservoir lube loss, bearing raceway condition, and visual examination of conformal coating, soldering, and other internal features.
Lubricant considerations for rolling-element bearings have within the last two decades taken on added importance in the design and operation of mechanical systems. The phenomenon which limits the useful life of bearings is rolling-element or surface pitting fatigue. The elastohydrodynamic (EHD) film thickness which separates the ball or roller surface from those of the raceways of the bearing directly affects bearing life. Chemical additives added to the lubricant can also significantly affect bearings life and reliability. The interaction of these physical and chemical effects is important to the design engineer and user of these systems. Design methods and lubricant selection for rolling-element bearings are presented and discussed.
The Lundberg-Palmgren equation was modified to show the effect on rolling-element fatigue life of removing by grinding a portion of the stressed volume of the raceways of a rolling-element bearing. Results of this analysis show that depending on the amount of material removed, and depending on the initial running time of the bearing when material removal occurs, the 10-percent life of the reground bearings ranges from 74 to 100 percent of the 10-percent life of a brand new bearing. Three bearing types were selected for testing. A total of 250 bearings were reground. Of this matter, 30 bearings from each type were endurance tested to 1600 hr. No bearing failure occurred related to material removal. Two bearing failures occurred due to defective rolling elements and were typical of those which may occur in new bearings.
In conventional ball bearings, steel to steel contacts between the balls and the raceways are at the origin of microwelds which lead to material transfer, surface roughening, lubricant breakdown, and finally to a loss in the bearing performances. To minimize the microwelding tendencies of the contacting partners it is necessary to modify their surface materials; the solid to solid collisions themselves are difficult to avoid. The use of titanium carbide coated steel balls can bring spectacular improvements in the performances and lifetimes of both oil-grease lubricated and oil-grease free bearings in a series of severe applications.
Current development of high-performance rolling element bearings for aircraft engines (up to 3 million DN, where DN is the product of shaft diameter in millimeters and speed in revolutions per minute) has aroused concern about fatigue crack growth in the inner bearing race that leads to catastrophic failure of the bearing and the engine. A failure sequence was postulated by Srawley, and an analytical program was undertaken to simulate fatigue crack propagation in the inner raceway of such a bearing. A fatigue specimen was developed at NASA by which fatigue data may be obtained relative to the cracking problems. The specimen may be used to obtain either mode 2 data alone or a combination of mixed-mode (1 and 2) data as well and was calibrated in this regard. Mixed-mode fracture data for M-50 bearing steel are presented, and a method for performing reversed-loading tests is described.
Crack propagation in a rotating inner raceway of a high-speed roller bearing is analyzed using the boundary integral method. The model consists of an edge plate under plane strain condition upon which varying Hertzian stress fields are superimposed. A multidomain boundary integral equation using quadratic elements was written to determine the stress intensity factors KI and KII at the crack tip for various roller positions. The multidomain formulation allows the two faces of the crack to be modeled in two different subregions, making it possible to analyze crack closure when the roller is positioned on or close to the crack line. KI and KII stress intensity factors along any direction were computed. These calculations permit determination of crack growth direction along which the average KI times the alternating KI is maximum.
The results are presented of a stress intensity factor analysis of semicircular surface cracks in the inner raceway of an engine bearing. The loading consists of a moving spherical Hertzian contact load and an axial stress due to rotation and shrink fit. A 3-D linear elastic Boundary Element Method code was developed to perform the stress analysis. The element library includes linear and quadratic isoparametric surface elements. Singular quarter point elements were employed to capture the square root displacement variation and the inverse square root stress singularity along the crack front. The program also possesses the capability to separate the whole domain into two subregions. This procedure enables one to solve nonsymmetric fracture mechanics problems without having to separate the crack surfaces a priori. A wide range of configuration parameters was investigated. The ratio of crack depth to bearing thickness was varied from one-sixtieth to one-fifth for several different locations of the Hertzian load. The stress intensity factors for several crack inclinations were also investigated. The results demonstrate the efficiency and accuracy of the Boundary Element Method. Moreover, the results can provide the basis for crack growth calculations and fatigue life prediction.
The SSME high pressure liquid oxygen turbopump (HPOTP) bearings sometimes wear and experience heating and oxidation of the ball and raceway surfaces. So far it has been impossible to measure the temperature of the bearings directly during operation of the turbopumps. However, a method was developed for determining the surface temperature of the bearings from the composition of the oxides using oxidation samples for calibration and Auger Electron Spectroscopy (AES) for chemical analysis.