Rolling Bearing Tribology: Tribology and Failure Modes of Rolling Element Bearings

By Gary L. Doll

Rolling Bearing Tribology: Tribology and Failure Modes of Rolling Element Bearings discusses these machine elements that are used to accommodate motion on or about shafts in mechanical systems, with ball bearings, cylindrical roller bearings, spherical roller bearings, and tapered roller bearings reviewed. Each bearing type experiences different kinds of motion and forces with their respective raceway, retainers and guiding flanges. The material in this book identifies the tribology of the major bearing types and how that tribology depends upon materials, surfaces and lubrication. In addition, the book describes the best practices to mitigate common failure modes of rolling element bearings.

• Discusses important tribological implications surrounding the performance and durability of rolling element bearings
• Describes how the different types of roller bearings work
• Explores the reasons behind the failure of roller bearings and presents information on how to mitigate those failures

• Language: English
• Publisher: Elsevier Science
• Release date: Aug 20, 2022
• ISBN: 9780128221747

Gary L. Doll

Dr.Gary L. Doll teaches classes at the University of Akron on the design, manufacture, and application of rolling element bearings. He is one of the world’s recognized authorities on rolling element bearing tribology. As the Chief Technologist of Tribology at the Timken Company, his responsibilities included overseeing the efforts in identifying and mitigating common failure modes experienced by rolling element bearings in multiple and world-wide market applications. Over his career, Dr. Doll has published over 350 articles and book chapters, edited numerous proceedings, received more than 28 US Patents, and was elevated to Fellow in the Society of Tribologists and Lubrication Engineers and the ASM International.

Book preview

Preface

Rolling element bearings are machine elements that are used to accommodate motion on or about shafts in mechanical systems and are classified according to the geometry of their rolling elements. Each bearing type experiences a different kind of motion and forces with their respective raceway, cages, and guiding flanges. The range of motion and forces between the rolling elements and the other parts of the bearing that they contact comprise a complex tribology that in a number of cases can lead to infantile failures of the bearings.

In this book, I have attempted to identify the tribology of the major bearing types, and how that tribology depends upon materials, surfaces, and lubrication. The book will also describe what I believe are the best practices to mitigate common failure modes of rolling element bearings. The material in this book is primarily based upon a Rolling Bearing Tribology course that I teach at the University of Akron.

Rolling bearings are almost always designed and employed according to their calculated fatigue life, yet millions of bearings are taken out of service annually due to different types of wear damage, which is not considered in fatigue life calculations. Although most types of wear damage that limit the operational lifetimes of rolling bearings can be avoided with a good understanding of the root causes of the wear, it is apparent that many technicians and engineers who design or operate machinery equipped with rolling bearings do not have a sufficient understanding of the tribological causes for wear damage. It is my intention that the material in this book may be beneficial in helping develop a better understanding of the relationships between the causes and effects of wear damage that rolling bearings encounter.

The evolution of the modern rolling bearing is reviewed in Chapter 1, from primitive devices employed by early civilizations to today’s advanced turbine engine bearings. Chapter 2 reviews the various types of rolling bearings and applications where the types of bearings are found. Chapter 3 provides an overview of how the tribology of rolling bearings depends upon the geometries, kinematics, contact stresses, and materials and establishes a foundation upon which a treatment of the tribology of rolling bearings can be based. Chapter 4 contains a treatment of lubricants; it begins with an overview of some of the most important physical properties that enable oils and greases to function as lubricants and how lubricant films are generated in rolling bearings. All rolling bearings experience friction while operating, and Chapter 5 identifies the friction sources and provides a means to estimate friction losses for different bearing types. Within the context of rolling bearing tribology, Chapter 6 attempts to provide an understanding of what rolling contact fatigue is and why it occurs in rolling bearings. Chapter 7 reviews the characteristic forms of rolling bearing damage including fatigue, wear, corrosion, electrical erosion, plastic deformation, and fracture and cracking. Technologies that can be implemented to mitigate or prevent most types of rolling bearing damage are discussed in Chapter 8. Finally, Chapter 9 contains examples of how surface treatments and heat treatment technologies have successfully mitigated common bearing damage modes.

Chapter One: Evolution of the rolling element bearing

Abstract

Although the rolling element bearing is sometimes regarded as a 20th century invention, in reality it is the product of the evolution of technology over thousands of years. From its crude beginnings in Roman times, its development in the Renaissance, and ubiquitous utilization during the Industrial Revolution, the modern rolling element bearing evolved to be one of the most technologically valuable mechanisms of the 20th century. The evolution of the modern rolling element bearing benefited from significant contributions from da Vinci, Osborne Reynolds, Beauchamp Tower, Heinrich Hertz, Richard Stribeck, John Goodman, and H.L. Heathcote.

Keywords

Rolling element bearing; Technological evolution; Reynolds; Hertz; Stribeck; Goodman; Heathcote

Nomenclature

C1, C2, C3 constants

d 

ball diameter (mm)

F 

load (N)

Fmax 

maximum load (N)

n 

number of balls

N 

rotational speed of bearing (rpm)

1.1: Introduction

Bearings were developed and are employed to provide rotational freedom and/or transmit a load between a shaft and a housing. Evidence of the use of bearings has been found in early civilizations (c.3500 BC) where simple bearings were used in drills for making fire and drilling holes, wheels for the fabrication of clay pottery, and wheels for vehicles. These sliding contact bearings were made of stone or wood, and were lubricated with materials derived from vegetation and animals. Although sliding contact or plain bearings played important roles in our technological evolution, it was not until the end of the 19th century that a serious study of the physics of these types of bearings occurred. Using data provided by Beauchamp Tower [1], a British inventor and railway engineer, Osborne Reynolds developed the theory of hydrodynamic lubrication and provided the first proof that a viscous liquid could physically separate two sliding surfaces by hydrodynamic pressure [2].

The two basic conditions for the creation of hydrodynamic lubrication are: two surfaces must move relative to each other with sufficient velocity for a load-carrying film to be generated, and the surfaces must not be parallel to each other, although there are exceptions to the second condition. With several simplifying approximations, Reynolds was able to derive a mathematical expression for hydrodynamic lubrication that has been used to design and understand the workings of modern hydrodynamic bearings.

It was not until the 20th century that the workings of rolling element bearings became well understood. Unlike plain or hydrodynamic bearings, a modern rolling element bearing is a unit that includes two rings with hardened raceways upon which hardened balls or rollers roll. Fig. 1.1 illustrates the basic components of a rolling element bearing. In most cases, rings and rolling elements are fabricated from steel alloys that can be sufficiently hardened (e.g., 58–65 Rockwell C scale hardness). To minimize inertial loading, components may also be fabricated from ceramic materials such as silicon nitride. The rolling elements are usually held in an angularly spaced relationship by a cage, also referred to as a separator or retainer. Cage materials are generally required to be relatively soft, and must possess a good strength-to-weight ratio. Mild steel, brass, bronze, aluminum, polyamide (nylon), polytetrafluoroethylene (PTFE), fiberglass, and plastics filled with carbon fibers are used as cage materials.

Fig. 1.1

Fig. 1.1 The primary components of a rolling element bearing are the inner and outer rings, the rolling elements, and the retainer or cage.

Harris and Kotzalas [3] compared the advantages and disadvantages of rolling element bearings with hydrodynamic bearings. A rolling element bearing has advantages over plain or hydrodynamic bearings that include:

•Rolling element bearings operate with much less frictional torque than hydrodynamic bearings which means that they experience less power loss and frictional heating.

•Unlike hydrodynamic bearings, the starting frictional torque of rolling element bearings is only slightly greater than their moving torque.

•Rolling element bearings require much smaller quantities of lubricants during operation and are less sensitive to lubricant interruption than hydrodynamic bearings.

•Many types of rolling element bearings can simultaneously support combinations or radial and thrust loads.

•Rolling element bearings can operate over wide ranges of load and speed.

•Rolling element bearings are less sensitive to fluctuations in load, speed, and operating temperatures than hydrodynamic bearings.

Rolling element bearings also have some disadvantages when compared to hydrodynamic bearings that include:

•Rolling element bearings experience rolling contact fatigue that is statistically variable.

•Larger radial envelopes are required for rolling element bearings than for hydrodynamic bearings that have similar load bearing capacity.

•Rolling element bearings have low damping capacity and generate more noise than hydrodynamic bearings.

•Rolling element bearings have more severe alignment requirements than hydrodynamic bearings.

•Rolling element bearings typically cost more than hydrodynamic bearings.

1.2: Evolution of rolling element bearings

Precision rolling element bearings are 20th-century creations and are used to reduce friction and wear in rotating machinery. The concept of rolling element bearings emerged during Roman times, development regressed during the Middle Ages, but was revived during the Renaissance, and steadily evolved in the 17th, 18th, 19th, and into the 20th centuries. One can trace the origins of rolling element bearings to early civilizations. There is a preponderance of evidence that the tribological advantages of rolling over sliding were recognized in the earliest stages of recorded history. For example, writing and the wheel appear to have both arisen about 3200 BC in Mesopotamia. Archeological finds in Egypt, the Indus Valley, and China also provide evidence of later developments of the wheel.

In Greece, Aristotle cited that frictional forces were smallest for rounded objects, and the historian Herodotus wrote about the use of rollers to transport vessels over land and in military engines for throwing stones. Writings of the Roman architect and engineer Marcus Vitruvius Pollio contain descriptions and illustrations of the incorporation of wheels and axles in vehicles used to move and erect stone columns in the construction of the Temple of Artemis [4]. However, the most compelling evidence for the early application of rolling element bearings came from the discovery of two sunken ships recovered from Lake Nemi in 1929 and 1930. It was determined that the ships were built between AD 44 and 54 and belonged to the Roman Emperor Caligula. In 1927, the Italian dictator Benito Mussolini engaged the engineer Guido Ucelli to drain the lake, recover and reconstruct the ships. Upon reconstruction, it was found that each ship contained a rotating platform. One platform was designed to rotate on trunnion-mounted bronze balls, which is the earliest evidence of a thrust ball bearing. The platform on the second ship appeared to be supported by trunnion-mounted wooden rollers with wooden rings which Ucelli viewed as an early tapered roller thrust bearing [5].

Oak hubs from a four-wheeled cart found in West Jutland, Denmark in 1881, contained bronze collars with 32 axial grooves where some of the grooves contained wooden rollers. It is clear that about 2000 years ago, the makers of the cart were intent upon replacing plain hub bearings with bronze roller bearings with cylindrical wooden rollers [6]. Annular bronze rings with internal compartments dating from the second century BC were found in Shansi, China. Following the discovery of granular iron dust in the compartments, it is believed that the compartments of these rings housed balls or rollers and were used to support a rotating shaft [7].

The first archival documentation of rolling element bearings can be found in the Renaissance, most notably in the works of Leonardo da Vinci. Leonardo’s sketches of rolling element pivot bearings can be found in the Codex Madrid I, which was discovered in the Biblioteca Nacional de España in Madrid in 1965 by Dr. Jules Piccus [8]. Leonardo’s drawings also show that he appreciated the concept of a separator, retainer, or cage in a rolling element bearing. Although Leonardo’s writings dominate the history of rolling element bearings in the Renaissance, there were other indications of the growing use of rolling motion to reduce friction and wear. For example, cast-iron balls were manufactured and used as low-friction supports for moving heavy gun carriages in the sixteenth century [9]. The autobiography of the Florentine goldsmith Benvenuto Cellini references the use of a thrust ball bearing containing four wooden spheres as the rolling elements that were used to move a statue of Jupiter in 1534. An astronomical clock constructed by Eberhardt Baldwin in 1561 contained roller bearings in the gear train.

The 17th and 18th centuries contained significant advancements in the development of roller bearings, mostly pertaining to wagons and carriages. Roller bearing development during this period benefited from the contributions of Hooke, de Mondran, and Rowe. In a discourse on carriages to the Royal Academy in 1685, Robert Hooke cited the advantage of using roller bearings instead of plain bearings to prevent gnawing, rubbing and fretting [10]. Later, the Academy of Science in Paris approved a carriage design by de Mondran citing that because of the rolling contact friction, one horse could easily do the work which could hardly be accomplished by two. Jacob Rowe was awarded a patent in 1734 for friction wheels that were made of wood, hooped with iron, and rotated on iron axles mounted in iron or metal balls. Rowe further attempted to quantify the economics of his invention. He calculated that if all carriages in England were equipped with his friction wheels, the use of fewer horses could produce a potential annual savings to the economy of £747,500.

The English inventor C. Varlo described advantages of rolling motion over sliding motion in carriage bearings in 1772. Varlo’s design of a full complement ball bearing showed an inner ring mounted on a nonrotating shaft, and balls with diameters of 25.4, 31.6, and 38.1 mm to be used for carrying increasing loads. He further specified that the hardest materials should be used in the rolling elements. Weather vanes atop old Trinity Church in Lancaster, Pennsylvania (built in 1794) and Independence Hall in Philadelphia, Pennsylvania (built in 1770) were supported by roller thrust bearings comprising copper rollers running against bronze rings. Count Marin Carburi developed a system containing linear ball bearings for transporting a granite block from Finland to St. Petersburg in 1777. A full-complement thrust ball bearing constructed from cast iron rings and balls was discovered in the aftermath of a fire that destroyed the Sprowston windmill (Fig. 1.2) constructed around 1730 near Norwich, England [11]. The inner and outer ring diameters were 61 cm and 86 cm, respectively, and the 40 balls had a diameter of 57 mm. The diameter of the grooves in which the balls ran was 70 mm, and the ratio of groove to ball radius was 1.22, slightly larger than that employed in modern ball bearings. Other notable achievements during this period include Henry Sully’s introduction of antifriction rollers in a chronometer in 1714 [12], John Garnett’s introduction of the essential features of rolling element bearings including rollers of various geometries and early forms of cages in 1787, Watkin George’s roller bearing patent in 1787 for use in destroying friction in axles and shafts, the first patented full complement ball bearing for light and heavy wheeled carriages in 1794 by Philip Vaughan, and Charles Greenway’s 1840 patent covering a wide range of rolling element bearings with cages for carriage wheels.

Fig. 1.2

Fig. 1.2 Photograph of the Sprowston windmill c.1920. (https://en.wikipedia.org/wiki/File:Sprowston_Mill.jpg.)

The years 1850–1925 saw the emergence of modern forms of rolling element bearings, and the employment of rolling element bearings in general industrial applications became widespread. However, it was the bicycle that was largely responsible for the critical developments in the ball bearing. During the 20-year period from 1890 until 1910, more than 600 British patent applications were filed on ball bearings for bicycles. It was the bicycle’s demand for large numbers of steel balls with reliable quality that led to the formation of precision-bearing manufacturers. Geometrical accuracy and uniformity of balls was of paramount importance since one ball larger than the others would carry a disproportionate amount of load and suffer premature failure. Additionally, the lack of a suitable bearing material limited ball bearing performance and life. During this time the importance of a material’s ability to resist plastic flow and fatigue was found to correlate with ball bearing life, and with contact stresses in excess of 1.4 × 10⁹ Nm− 2 (1.4 GPa), cast iron balls were too brittle to provide desirable or adequate bearing life. The geometrical and materials requirements of ball bearings for bicycles drove revolutionary improvements in precision manufacturing methods and in the development of specialty steels.

To withstand the severe contact conditions described above, it was necessary to harden ball bearing steels. Beginning about 1900, hypoeutectic steels with about 0.15% carbon were placed in a high temperature, carbon containing environment to facilitate the diffusion of carbon atoms into the steel, forming a thin, hard case. While this process considerably enhanced the performance of bearing steels, it was difficult to ensure homogeneity with turn of the century heat-treating technology. Consequently, most bearing manufacturers adopted a through-hardening process utilizing hypereutectic steel alloys that contained about 1% carbon and 1½% chromium. Later, steel alloys with 12%–15% chromium were utilized as bearing materials for more severe operating environments.

Initially, balls were produced by either casting or turning on a lathe from a steel bar. However, to meet the high output demand for balls, a different process was needed. The process development started with an automated cold-heading of steel wire for smaller size balls and hot-pressing for larger size balls. Material protruding around the equator of the balls was removed by tumbling or rough grinding; then the balls were placed between cast-iron disks and flat grinding wheels arranged in a manner that ensured random motion of the balls. By the end of the 19th century, steel balls with less than ± 12 μm variances in diameters could be achieved with this process.

Prior to 1900, manufacturers of machines made their own rolling element bearings, although many outsourced the ball manufacturing. The beginning of the 20th century saw the emergence of the rolling element bearing industry. Sven Wingquist, a plant engineer with the Gamlestadens Fabriker textile company, was tasked with addressing a large number of bearing failures due to misalignments and shaft deflections. His solution was the creation of a self-aligning, double row ball bearing and a new bearing company, A.B. Svenska Kullagerfabriken (SKF), which was established in Göteborg in 1907. The company expanded rapidly and factories were soon established in England (1911), Germany (1914), the United States (1916), and France (1917).

Henry Timken, owner of a carriage-building business in St. Louis, Missouri became focused on developing axle bearings capable of operating under combinations of high radial and axial loads. Timken’s solution was the design of a tapered roller bearing for vehicles (Fig. 1.3). The Timken Roller Bearing and Axle Company was created in St. Louis in 1898, and relocated to Canton, Ohio in 1902 to be nearer to the emerging automotive industry. In 1909, the Bearing and Axle divisions separated, and the Timken-Detroit Axle Company (later renamed Rockwell then Meritor) was formed. In 1917, the Timken Company began its steel- and tube-making operations to vertically integrate and maintain better control over the steel used in its bearings.

Fig. 1.3

Fig. 1.3 Schematic of the tapered roller bearing for vehicles. (From the H. Timken and R. Heinzelman US Patent No. 606,635.)

The rapidly growing automobile and precision-manufacturing industries led to the formation of other specialist bearing companies such as the Fafnir Bearing Company of Connecticut in 1911 and Japan’s Nippon Seiko in 1916.

Concomitant with the emergence of the precision rolling bearing companies, important scientific studies of contact mechanics, ball bearings, and rolling friction were performed. After viewing a presentation on Newton’s rings at a meeting of the Physical Society of Berlin, Heinrich Rudolph Hertz was moved to consider how glasses deform when pressed together. Hertz developed the mathematics that solved the problem of contact stresses between elastic solids and the 23-year-old engineer presented his work to the Physical Society of Berlin in 1881 [13] Hertz’s mathematics described the deformations and contact stresses within elliptical regions of contact between arbitrarily shaped elastic solids when loaded together, and have provided the basis for the analysis of contact stresses in rolling element