Nelson H. Forster

Viscoelastic Effects in MIL-L-7808-Type Lubricant, Part I: Analytical Formulation

Analytical formulations for the computation of lubricant film thickness and traction in a high-speed rolling-sliding contact are presented with the objective of investigating the viscoelastic response of the MIL-L-7808-type lubricant Two types of relations are used to model the viscous shear strain rate. In the Type I model, a hyperbolic sine relation is used to model the viscous effect which becomes significant when the shear stress reaches a critical value. The Type II model employs a limiting shear stress, which the lubricant can withstand, and an inverse hyperbolic tangent function is considered to model the viscous behavior. Both models are based on three fundamental properties: lubricant viscosity, shear modulus and a critical shear stress. While the viscosity relations may be obtained by direct measurements, estimates of shear modulus and critical shear stress may be derived by curve-fitting the model predictions to experimental traction data. Presented at the 45th Annual Meeting in Denver, Colorad...

Rolling Contact Fatigue Life and Spall Propagation of AISI M50, M50NiL, and AISI 52100, Part I: Experimental Results

This article is the first part of a three-part series that investigates the rolling contact fatigue (RCF) initiation and spall propagation characteristics of three bearing materials, namely, AISI 52100, VIM-VAR AISI M50, and VIM-VAR M50NiL steels. Although there is substantial prior work published on the rolling contact fatigue initiation of these materials, little has been published on their spall propagation characteristics after spall initiation. It is recognized that rapid spall growth can lead to catastrophic bearing failure. Hence, understanding the spall growth phase and factors that may cause accelerated growth rates is key to achieving a reliable and robust bearing design. The end goal is to identify control parameters for optimizing bearing materials for improved spall growth resistance. This first part study features the experimental results from 208-size (40 mm bore) angular-contact ball bearings endurance life tested at maximum Hertzian contact stress levels of 3.10 GPa and bearing outer race temperatures up to 131°C. Spall propagation experiments were conducted on new and life tested bearings at 2.10 and 2.41 GPa maximum contact stress. Spall propagation experiments show that all materials exhibit a rapid or critical spall growth rate after undergoing an initial low-rate spall growth period. The time-to-critical growth rate is dependent on contact stress and was swiftest in AISI 52100 steel. To better understand the underlying physics, the driving factors, and failure mechanisms, the state of stress is modeled using finite element analysis in Part II and an in-depth microstructural analysis of selected bearings is presented in Part III.

Rolling Contact Fatigue Life and Spall Propagation Characteristics of AISI M50, M50 NiL, and AISI 52100, Part III: Metallurgical Examination

This is the third part of a three-part series that investigates the rolling contact fatigue initiation and spall propagation characteristics of three bearing materials, namely, AISI 52100, VIM-VAR M50, and VIM-VAR M50 NiL steels. Though there is substantial prior work published on the rolling contact fatigue initiation of these materials, little is known about their spall propagation characteristics after spall initiation. In Part III, 208-size, 40-mm-bore bearings are examined for changes in appearance of the microstructure as well as residual stress as a function of depth in the circumferential direction. The correlations between the experimental results from Part I and computer modeling in Part II are made.

Rolling Contact Fatigue Life and Spall Propagation of AISI M50, M50NiL, and AISI 52100, Part II: Stress Modeling

This is the second part of a three-part series that investigates the rolling contact fatigue initiation and spall propagation characteristics of three bearing materials, namely, AISI 52100, VIM-VAR M50, and VIM-VAR M50NiL steels. A systematic investigation of the effects of rolling contact fatigue (RCF) on the evolution of material properties and microstructural changes resulting in spall propagation rate has not been conducted. As a first step toward understanding spall propagation, we present stress distribution in the neighborhood of a spall initiated by RCF by an indent on a hybrid 40-mm ball bearing. Elastic and elastic-plastic subsurface stress fields are computed using finite element models that incorporate the full three-dimensional (3D) ball-raceway geometry. The stress fields predicted indicate extensive yielding around the spall edges. The spall is shown to widen first axially across the width of the raceway, causing the ball to unload as it enters the spall. The effect of the potential impact on the spall trailing edge is also considered. The proposed scenario involves repeated application of contact stress resulting from a combination of ball static and the impact loads results in extensive plastic deformation of the spall trailing edge, leading to degradation and release of material resulting in spall propagation.

Rolling Contact Testing of Vapor Phase Lubricants—Part III: Surface Analysis©

Infrared microscopy (μ-FTIR), Raman spectroscopy, Auger electron spectroscopy (AES), scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) were used to characterize bearing surfaces lubricated from the vapor phase. Results are presented for surface films generated with a tertiary-butyl phenyl phosphate (TBPP) and a polyphenylether (5P4E). Infrared bands consistent with inorganic poly-phosphates and phosphites were detected in deposition films generated with the TBPP lubricant. Raman shifts consistent with graphite of short range order were detected in films deposited by both lubricants. However, specimens lubricated by the 5P4E had several Raman shifts consistent with Fe2O3 and Fe3O4. Iron oxides were not detected in deposition films generated with the TBPP lubricant. AES depth profiles indicate a predominant mixture of iron and carbon in the TBPP deposition film; whereas, the 5P4E film predominately consists of iron and oxygen. Localized corrosion pitting generated with the TBPP lub...

Stress Field Evolution in a Ball Bearing Raceway Fatigue Spall

The governing mechanisms of fatigue spall propagation in ball bearing inner raceways are investigated through the use of elastic-plastic finite element modeling, X-ray diffraction, and the visual inspection of fatigue spall cracks. The model simulates multiple ball impacts with a fatigue spalls edge in a 208 size ball bearing operating at 10,000 rpm. Ball impacts are shown to cause severe plastic deformation within the spall edge and induce tensile residual stresses. The finite element results are supported by X-ray diffraction measurements and the locations of cracks observed around the edge of a spall.

Solid Lubrication of Silicon Nitride with Cesium-Based Compounds: Part I — Rolling Contact Endurance, Friction and Wear

The high temperature rolling contact endurance, friction, and wear of 16 cesium-based compounds with solid lubricating properties were investigated on silicon nitride (Si3N4). Some were also investigated on bearing tool steels and several state-of-the-art high temperature solid lubricants were investigated for comparison. Experiments were conducted in air at temperatures up to 650°C, contact stresses up to 4.34 GPa, and a pure rolling surface speed of 1.8 m/s. Although all of the cesium-based compounds exhibited self-lubricating properties, the best overall performance was achieved with a cesium silicate reaction film formed in-situ (Cs2O·xSiO2) and a hydrated cesium silicate bonded coating (Cs2O·3SiO2·nH2O). Bonded coatings of cesium oxythiotungstate + tungsten disulfide mixture (Cs2WOS3 + WS2) and cesium hydroxide (CsOH) also performed well. It is hypothesized that high temperature chemical reactions between the cesium-containing compounds and the silicon nitride surface form a lubricious cesium silicate film. Presented at the 55th Annual Meeting Nashville, Tennessee May 7–11, 2000

Viscoelastic Effects in MIL-L-7808-Type Lubricant, Part II: Experimental Data Correlations

Viscoelastic behavior of the MIL-L-7808-type lubricant is modeled by correlating experimental traction data to rheological models based on lubricant viscosity, shear modulus and a critical or limiting shear stress. The two types of traction models used respectively employ a hyperbolic sine and inverse hyperbolic tangent function between the viscous shear strain rate and shear stress in the lubricant. While the viscosity-pressure-temperature relation is determined from viscosity measurements on a high-pressure viscometer, estimates of shear modulus and critical or limiting shear stress are derived by curve-fitting the model predictions to actual traction data obtained from a rolling-disk traction machine. Presented at the 45th Annual Meeting in Denver, Colorado May 7–10, 1990

Modeling of Wear in a Solid-Lubricated Ball Bearing

Computer modeling of wear in a solid-lubricated ball bearing for high-speed turbine applications is considered in terms of local interactions and the overall dynamics of the bearing elements. With prescribed coefficients of wear at the various interfaces between the interacting bearing elements, the computer model ADORE is used to obtain the time-averaged wear rates for the balls, races, and the cage as a function of the operating conditions typical of a gas turbine application. The model presents analytical estimates of wear in the ball pockets and at the guide lands of the cage, and it provides some guidance for optimizing cage design in solid-lubricated ball bearings. Presented at the 41st Annual Meeting in Toronto, Ontario, Canada May 12–15, 1986

Rolling Contact Testing of Vapor Phase Lubricants—Part IV: Diffusion Mechanisms©

Auger depth profile data, obtained from vapor-lubricated T15 bearing steel, were modeled using the solution for diffusion in a semi-infinite pair. A tertiary-butyl phenyl phosphate was used as the vapor lubricant. The primary species undergoing diffusion are iron and carbon. The objective of the experiment was to determine the order of magnitude of the diffusion coefficient and to qualitatively assess what types of diffusion mechanisms are involved. The experimental results indicate that the diffusion profile travels at a velocity equal to the bearing wear rate under dynamic conditions. This is possible if iron diffusions at a faster rate than carbon, i.e., the Kirkendall effect. Analyses of the data were performed using Darkens equations. The results indicate that the diffusion coefficient of iron is of the order of 1 × 10−14 cm2/s at test temperatures of 370° and 430°C. Diffusion is thought to occur via the migration of iron cations through an anionic lattice of polyphosphate and phosphite, i.e., catio...