Tedric A. Harris

Fatigue Failure Progression in Ball Bearings

Typically, the life of a component containing rolling contacts is defined as the time to the initiation of a fatigue spall. Initiation of a spall does not always cause the component to stop performing its designed function. Operating the component past the initiation of a spall increases the heat generation rates and vibrations, eventually leading to total failure. A ball/v-ring test rig was used to initiate and progress spalls on bearing halls where spall progression was measured as a function of time, and vibrations monitored using accelerometers. A spall progression life mathematical model for balls endurance tested in the v-ring rig was created by extending the loannides-Harris fatigue life theory. Also. excessive vibratory loading was determined to be the major cause of total component failure.

On the Accuracy of Rolling Bearing Fatigue Life Prediction

Ball and roller bearings are designed to meet endurance requirements basically determined according to the Standard fatigue life calculation method. This method is based on the Lundberg-Palmgren fatigue life theory as modified by reliability, material, and lubrication factors. As application load and speed requirements have increased, the Lundberg-Palmgren method has resulted in bearings of increased size, adding unnecessarily to the size and weight of mechanisms. This is a critical design situation for weight and size-sensitive components such as aircraft gas turbine engines and helicopter power transmissions. The bearing life prediction method developed by loannides and Harris recognizes the existence of a fatigue limit stress. If the stresses an operating bearing experiences do not exceed the limit stress, the bearing can achieve infinite life. In any case, the method tends to predict longer lives than the Lundberg-Palmgren method. This paper evaluates the life prediction accuracies of the Lundberg-Palmgren and loannides-Harris methods by comparing lives calculated according to these methods and to those actually experienced in 62 different applications. As a result of the investigation, the loannides-Harris method is shown to more accurately predict bearing fatigue endurance.

Lundberg-palmgren fatigue theory : Considerations of failure stress and stressed volume

Load ratings and fatigue life prediction of rolling bearings is based on the Lundberg-Palmgren theory first published in Sweden in 1947. The basic equation of this theory includes the stressed volume of material in the bearing raceway subsurfaces as a major parameter. This volume of material is simplistically determined to have a nearly rectangular subsurface cross-sectional area bounded by the length of the maximum contact area ellipse and the depth at which the maximum failure-causing stress occurs. The latter stress is assumed to be effective over this area. In fact, a distribution of stress occurs, and in this investigation it is demonstrated that the subsurface volume with a potential for fatigue cracking is substantially different from that used by Lundberg and Palmgren. This difference in volume, particularly in the presence of surface shear stresses, can have a profound effect on the method and prediction of bearing fatigue lives.

Tribological Performance Prediction of Aircraft Gas Turbine Mainshaft Ball Bearings

The internal thrust loads in aircraft gas turbines are carried by angular-contact ball bearings, one on each of the compressor-turbine-driven shafts. There are typically two or three concentric shafts, rotating at individual speeds to generate the required aerodynamic loading with a resultant thrust load on each of the ball bearings. The designs of these bearings are based primarily on the required life or fatigue endurance. Estimation of this parameter is currently accomplished using empirically developed, factor-based calculation methods, e.g., material life and lubrication life factors. Experience has demonstrated in many instances that bearing fatigue lives predicted by this method are less than those achieved in practice. A more accurate performance prediction system would allow improved optimization of bearing designs with associated savings in engine performance, weight, and cost as well as improved reliability. As thrust-to-weight ratio increases with new engine designs, the need for more compact ...

Fatigue Failure and Ball Bearing Friction

According to the standard method for calculation of life ratings for rolling contact bearings, failure is considered to occur at the initiation of the first spall in a rolling component surface. While this is acceptable for most bearing applications, in many applications bearing operation may effectively continue for a period of time during which spoiling progresses. As spoiling progresses, bearing friction and vibratory loading increase, eventually resulting in bearing seizure and mechanism failure. To estimate the amount of operating time available for corrective action; i.e., mechanism shutdown and bearing replacement, between initial spoiling and potential mechanism failure, an investigation into the friction occurring in a spalled-unspalled concentrated point contact was conducted using a ball/v-ring test rig to generate and progress spalls and a ball-disk rig to test traction. It was thereby determined for a spalled ball-raceway) contact, that a lubricant film of thickness sufficient to adequately “separate” the contacting surfaces can occur: With continuous ample supply of fluid lubricant, measured traction coefficients were only slightly greater than would be expected for unfailed contacts. Using the experimental results, an empirically based mathematical model was developed to estimate ball-raceway traction as a function of degree of spalling. These data will subsequently be used in ongoing studies to estimate rate of spall progression in bearing applications. Presented as a Society of Tribologists and Lubrication Engineers Paper at the STLE/ASME Tribology Conference in Orlando, Florida, October 11–13, 1999

A Method to Calculate Frictional Effects in Oil-Lubricated Ball Bearings

Friction and heat generation in oil-lubricated ball bearings is mainly the result of sliding in the ball-raceway contacts and agitation of the lubricant in the free space between the balls, cage. and bearing rings. Endurance of a ball bearing is highly dependent on the thicknesses of the oil films which serve to separate the balls from the raceways in a well-lubricated bearing. The film thicknesses, in turn, are dependent upon the lubricants viscous properties. These are functions of bearing temperatures and are determined by the balance between the rates of frictional heat generation and heat dissipation. Therefore, in the design of a ball bearing application such as the high speed rotor and low speed rotor support bearings in an aircraft gas turbine, it is important to be able to predict the bearing frictional heat generation rates with reasonable accuracy. This paper presents a method to perform the required calculations considering hearing loading and speeds, realistic lubricant rheological properties, and a relatively simple heat transfer system between the hearing rings, halls. And lubricant. The results of the analysis are shown to compare favorably with experimental data. Presented as a Society of Tribologists and Lubrication Engineers Paper at the STLE/ASME Tribology Conference in San Francisco, CA October 21–24, 2001

On the Causes and Effects of Roller Skewing in Cylindrical Roller Bearings

Cylindrical roller bearings which support thrust and/or moment loading in addition to predominant radial loading have flanges on both the inner and outer rings to accommodate the thrust loading carried by the rollers. The roller thrust loading causes roller tilting and friction moments about the rotating axis perpendicular to both the bearing and roller axis. The friction moments in turn give rise to roller skewing which increases roller-raceway loading affecting bearing friction heat generation and fatigue endurance. This investigation shows the means of including the skewing effects in the analysis of overall bearing performance.

Predicting Micro-Pitting Occurrence in Wind Turbine Gearbox Roller Bearings

The standard rolling contact fatigue life calculations currently in use by the rolling bearing industry is based on the first occurrence of subsurface-initiated spalling of a raceway or roller surface. However, wind turbine gearbox roller bearings have been suffering from another damage mode, which manifests itself as micro-pitting. The micro-pitting, which is spalling on a micro scale, by itself can be tolerated in its early stages; i.e. the roller bearing will still function properly. As the damaged bearing continues to operate, the micro-pitting propagates and at the later stages, often termed peeling, the pitting becomes deep enough to reach the appearance of traditional subsurface-initiated spalling. To better understand the phenomenon micro-pitting and its causes, this study was conducted to review published literature on the topic as it relates to bearing operation. The key findings were the need for a low specific lubricant film thickness parameter, and some component of sliding velocity in the contacting surface. With this knowledge, a proposed test scheme including these variables could be created from which a method to predict the risk of micro-pitting may be determined.Copyright