Talk Page



A bearing is a device to allow constrained relative motion between two or more parts, typically rotation or linear movement. Bearings may be classified broadly according to the motions they allow and according to their principle of operation, as well as by the directions of applied loads they can handle.

2.0.1 Load External Type Internal Type
4.2.1 Alignment
4.9.1 Crush
4.9.2 Torque
5.3.7 Vibration

Bearing Design and Construction

Rotating mechanical equipment in industry today is faster, more efficient, and carries greater loads than the machinery in use twenty years ago. Component parts of today’s machines are designed and built to minimize the necessity of frequent replacement and repairs. Bearings, by the nature of the service they perform in a rotating device, are expendable and only have a finite life span. A thorough understanding of the operation, maintenance, and installation of different bearing types is essential to ensure they will last.

Maintenance technicians can spend countless hours and several thousands of dollars to replace damaged bearings that failed due to any number of reasons. Many of the reasons for failure can be avoided by using the information contained within the following sections.

This section discusses the various types, capabilities, and fundamental applications for many commonly used bearings. The following topics are discussed:

Bearing Requirements

There are three basic functions a bearing must perform:

  1. Support a load (including the weight of the components)
  2. Assist in maintaining alignment (axial and radial)
  3. Reduce the friction that exists between parts in relative motion to each other

A properly designed and installed bearing provided with the correct lubricant will perform the above-mentioned functions.

Bearings are expendable parts. The cost of replacing a bearing is much less than the cost of manufacturing a new shaft. Bearings are subjected to load and friction. These two parameters directly affect the operational life of bearings.


Load is defined as "the force a bearing must support." When the equipment is in operation, the load is considered a dynamic load. During the idle period when the machine is stopped, a static load condition exists, due solely to the weight of the component parts of the machine.

The two types of loads a bearing must withstand are radial loads and thrust loads. Radial loads are forces that occur at right angles to the shaft. These forces tend to make the shaft move side to side or up and down. This movement is called radial runout. Figure 1 shows the relationships of radial loads to a shaft.

Figure 1: Radial Loads Acting on a Shaft

Thrust loads are forces that are directed axially along the length of the shaft. These forces tend to make the shaft move back and forth along its axis. This movement is called endplay. Figure 2 shows the relationships of thrust loads to a shaft.

Figure 2: Thrust Load Acting on a Shaft


Friction is defined as "the resistance to motion between two surfaces in contact." The two types of friction that exist between moving bodies are rolling and sliding friction. Rolling friction exists when one body rolls around or within another, while sliding friction exists when one body slides over another. Anti-friction bearings are designed to take advantage of the fact that rolling friction is much less than sliding friction under comparable load conditions. Figure 3 shows the differences between rolling and sliding friction.


Bearing Types

There are dozens of different types of bearings being manufactured today, each with specific design and operating characteristics. All bearings fall into two basic categories:

  • Anti-friction (rolling contact)
  • Friction (sliding contact)

The terminology used to describe bearings varies from manufacturer to manufacturer. Anti-friction bearings are referred to as rolling contact or roller types. Friction bearings are referred to as sliding contact, journal, sleeve, or plain bearings.

Anti-Friction Bearings

Anti-friction bearings consist of some type of rolling element confined within an inner and outer ring. This greatly reduces friction by creating a rolling action. Rolling friction is still present, but it is far less than sliding friction. When properly lubricated, anti-friction bearings work well for many requirements.

Anti-friction bearings come in two basic types--ball or roller--which differ in the shape of the rolling elements and the type of contact they exhibit. Figure 4 shows that the basic designs of both are very similar. Each has an inner and outer ring separated by the rolling elements. The function of the separator is to maintain the proper spacing between the balls and rollers.

Figure 4: Two Basic Anti-Friction Bearings

Anti-Friction Ball Bearings

Anti-friction ball bearings come in many different configurations depending on load requirements. The most common uses for ball-type bearings are in pumps and motors. The different types that are covered in this section are:

  • Deep-groove (Conrad)
  • Maximum capacity
  • Angular contact
  • Self-aligning

Deep-Groove Ball Bearings

The standard deep-groove, or Conrad, ball bearing is the most simple from the standpoint of design and construction. At the same time, it has the greatest all-around use. Although it is primarily a radial bearing, it is capable of handling moderate thrust loads from either direction and operating at relatively high speeds.

Where heavy thrust or radial loads are encountered in either direction or where great axial rigidity is required, a deep-groove ball bearing having two rows of balls is often used. Figure 5 shows a cross-section of the single-row type.

Figure 5: Deep-Groove Ball Bearing Cross-Sections

The bearing rings have symmetrical, deep-grooved raceways without filling slots or a counter-bore. The raceways are precision ground to conform closely to ball curvature, consistent with minimum friction, maximum capacity, and practical manufacturing techniques. The balls are selected for uniformity to ensure optimum internal load distribution.

The bearing uses the maximum number of balls which can be inserted between the raceways by eccentrically displacing the inner and outer rings. Balls are spaced by a two-piece, machined bronze cage. For higher speeds and when other operating conditions warrant, the cage may be made of other materials.

Maximum Capacity Ball Bearings

The maximum capacity ball bearing is a variation of the standard deep-groove ball bearing. It can handle a greater radial capacity than the single row deep-groove bearing, but can take thrust in one direction only. If thrust is applied in the wrong direction, the rolling elements may be displaced via the loading groove.

This type of bearing has an inner ring with a deep groove, but the outer ring is counter-bored, substantially reducing the raceway shoulder on one side. The counter-bore allows a maximum complement of balls in a one-piece, machined bronze cage to be assembled in the bearing. The outer ring is thermally expanded and slipped over the cage, ball, and inner ring assembly. After cooling, the bearing is non-separable. Figure 6 shows a cross-section view of the maximum capacity ball bearing.

Figure 6: Maximum Capacity Ball Bearing

Since the thrust capacity of the maximum capacity bearing is uni-directional, the preferred mounting is in opposed pairs. It may also be opposed by a standard deep-groove bearing or some other type of axially-located bearing. The center distance between bearings should be held to a minimum to help prevent thrust on the shallow shoulder or preloading.

Angular Contact Ball Bearings

The angular contact ball bearing is used to support heavy thrust loads in one direction with moderate radial loads. A high shoulder on one side of the outer ring and a similar high shoulder on the opposite side of the inner ring allows the bearing to accept a thrust load. Figure 7 shows a cross-section of an angular contact bearing.

Figure 7: Angular Contact Ball Bearing Cross-Section

Angular contact bearings may be mounted singly or in tandem for constant load in one direction. They can also be mounted in pairs for combined loads in the following combinations:

Face-to-Face Mounting

Angular contact ball bearings mounted face-to-face are shown in Figure 8. In this mounting, the back surfaces of the outer rings are supported against the housing shoulders and the faces are together. This mounting is used when the outer races are clamped in the housing and is less rigid in the presence of angular deflection.

Figure 8: Face-to-Face Mounting

Back-to-Back Mounting

In back-to-back mounting, the back surfaces of the bearings are mounted together. This mounting is used when the bearing unit is allowed to float axially and the outer races are clamped only to accept axial thrust. The back-to-back mounting is more rigid in the presence of angular deflection than the face-to-face mounting. Figure 9 shows angular contact ball bearings mounted back-to-back.

Figure 9: Back-to-Back Mounting

Figure 10 shows the tandem mounting of bearings face-to-back. The back surfaces of the outer rings are supported against the housing shoulder and will carry thrust in one direction only. This provides for a maximum thrust capacity for axial loads applied along the shaft in one direction.

Figure 10: Face-to-Back Mounting

Multiple Mounting

When it is desirable to have a maximum axial thrust capacity as in the tandem mounting but also provide for thrust in the other direction, multiple mountings may be used. In this mounting, two bearings are mounted back-to-face with a third in the back-to-back configuration. This mounting is shown in Figure 11.

Figure 11: Multiple Mounting

Self-Aligning Ball Bearings

Whenever it is impossible or impractical to maintain alignment of bearing seats within normal limits, the use of self-aligning ball bearings is extremely helpful. Two common types of ball bearings having this feature are external self-aligning and internal self-aligning.

External Type

The external type (Figure 12) is a regular deep-groove ball bearing, the outer ring has a spherical surface and fits into another mating ring. The entire bearing assembly is free to swivel within this extra outer ring, thus allowing axial alignment.

Figure 12: External Self-Aligning Ball Bearing

Internal Type

The internal self-aligning ball bearing has two rows of balls in separate grooves in the inner ring, but has only a single spherical raceway in the outer ring (Figure 13). Since this spherical portion has its center at the center of the bearing, the balls and inner assembly can align themselves freely about the center.

The great advantage of this feature is that neither misalignment nor shaft deflection under load is harmful, since the balls will always be in contact with the spherical raceway, as long as they do not reach its edges. The load-carrying capacity per ball in this bearing suffers because of the great difference in curvature between the outer race and the balls. Because of the large number of balls, the bearing can take a high radial load and a considerable thrust load.

Figure 13: Internal Self-Aligning Ball Bearing

Anti-friction Roller Bearings

Roller bearings are very similar in function to ball bearings except that various types of rollers are substituted for balls as friction-reducing elements. Because of the greater area of contact between rollers and raceways, these bearings can generally support a much higher load. They endure much harder service for a given size than ball bearings.

Roller bearings are also more resistant to shock and overloads than ball bearings because the cylinders undergo less deformation under load than do balls. They are frequently manufactured with a tapered sleeve which, when forced on the mating shaft, reduces internal clearance in the bearing and maintains accurate shaft position. The five types of roller bearings that are covered in this section are:

  1. Cylindrical
  2. Spherical
  3. Needle
  4. Tapered
  5. Journal

Cylindrical Roller Bearings

In the cylindrical roller bearing (Figure 14), the rollers are true cylinders that are usually guided by flanges on the inner and outer ring. A cage keeps them in uniform distribution around the circumference of the rings. A cylindrical roller bearing is ideal for use in high-load applications. Cylindrical roller bearings support little or no thrust. The roller assembly may be separable from both inner and outer rings, or it may be assembled with either.

Figure 14: Cylindrical Roller Bearing

Cylindrical roller bearings are normally manufactured in single row designs which differ in the arrangement of the flanges on each ring. A cylindrical roller bearing having two integral flanges on the outer ring and a flange-less inner ring is an example of a separable cylindrical roller bearing. The term integral flange means that the flanges on the ring are actually part of the ring itself and not a separate component. Figure 15 shows two basic types of this bearing.

Figure 15: Separable Single Row Cylindrical Roller Bearing

The outer raceway and rollers act as one unit and are removed together, leaving the inner race secured to the shaft. This design is commonly used in applications such as electric motors. When removing the motor’s end housing, the outer race and cylindrical roller unit separate from the inner raceway.

Another style of separable cylindrical roller bearing has two integral flanges on the inner ring and a flange-less outer ring. Each of the separable cylindrical roller bearings permit axial displacement of the machine’s housing relative to the machine’s shaft within certain limits in both directions. These two types are used as non-locating bearings.

Figure 16 shows an example of a cylindrical roller bearing having two integral flanges on the outer ring and one on the inner ring. This bearing can be used in machine applications where axial location on the shaft is required in one direction.

Figure 16: One Direction Separable Cylindrical Roller Bearing

Another style of cylindrical roller bearing having two integral flanges on the outer ring is shown in Figure 17. The inner ring has one integral and one loose flange. This type of bearing can be used to locate the machine’s shaft in both axial directions.

Figure 17: Two Direction Separable Cylindrical Roller Bearing

Spherical Roller Bearings

Spherical roller bearings have two rows of rollers which run in a common spheroid raceway in the outer ring. The two inner ring raceways are inclined at an angle to the bearing axis. Spherical roller bearings carry heavy radial loads, and the bearing can also accommodate axial loads acting in both directions. This type of bearing can be found as the fixed or float bearing on large fan shafts. Figure 18 shows a spherical roller bearing.

Figure 18: Spherical Roller Bearing

The bearing consists of a number of long, symmetrical rollers of large diameter giving the bearing very high load carrying capacity. The rollers are guided by raceways, the cage, and a non-integral guide ring which is positioned between the two rows of rollers.

Needle Roller Bearings

Needle roller bearings are considered a special design of the roller bearing. The needles have a length-to-diameter ratio which is much greater than that found in square cylindrical roller bearings. Due to their proportions, they cannot be manufactured as accurately. Also, the friction developed by these bearings is several times that of cylindrical roller bearings. They are commonly found in levelers and universal joints.

Needle roller bearings (Figure 19) are used when oscillation of the shaft is encountered or where the radial load is intermittent and allows the needles to relocate themselves. Commonly, the needles run directly on the shaft, which is harder and round. They are also available for applications where radial space is limited.

Figure 19: Needle Roller Bearing

Tapered Roller Bearings

Tapered roller bearings are designed for both heavy radial and thrust loads. As shown in Figure 20, there are four basic parts to all tapered roller bearings: (1) the cage, which retains and guides the rollers, (2) the cup, which acts as the outer race, (3) the rollers, which roll freely between the cup and cone, and (4) the cone, which serves as the inner race.

Figure 20: Principle of Tapered Roller Bearings

The apex of the angles formed by both the rollers and raceways, when extended, meets on a common axis. This feature allows the roller to follow the tapered raceways without skidding or binding. In Figure 21, an overall view of a standard tapered roller bearing shows the basic parts and design features.

Figure 21: Tapered Roller Bearing Components

There are many types of tapered roller bearings in single-row, double-row, and four-row designs. All tapered roller bearings are suitable for carrying combined radial and axial loads. For these reasons, they are commonly used in gear reduction boxes. The cone, rollers, and cage are assembled into one complete unit and cannot be separated. The cone assembly fits into the cup and the rollers revolve freely between the cup and cone.

Figure 22 shows a tapered roller bearing having a single row of rollers. This bearing can carry axial loads in one direction only. A radial load directed on the bearing presents an induced axial load which must be counteracted. It is generally adjusted against a second bearing in the machine.

Figure 22: Single-Row Tapered Roller Bearing

Some single-row tapered roller bearings are available as matched pairs in a face-to-face arrangement. These are specially made. When mounted adjacent to each other, the pair-determined axial internal clearance will be obtained and an even radial load achieved. Paired bearing arrangements are used where the load carrying capabilities of a single bearing is inadequate or when the bearing arrangement must carry axial loads in both directions. These can be commonly found in steel mill work rolls. A two-row tapered roller bearing having a double-cup with the single cones is shown in Figure 23.

Figure 23: Two-Row Tapered Roller Bearing

Another type of two-row tapered roller bearing is shown in Figure 24. This design has a double cup with a spacer positioned between the two single cones. The use of the spacer indicates that the correct running clearance was either built into the bearing at the time of bearing manufacture, or the selected spacer was provided by the equipment owner at the time the machine was assembled. The cones are usually clamped against a shoulder on the shaft with an end plate and cap screws or a threaded collar.

Figure 24: Two-Row Tapered Roller Bearing with Spacer

Because cone spacers are specially made for one particular bearing, they cannot be modified or interchanged with any other bearing. Care should be taken that parts are not mixed. The bearing adjustment in spacer bearings is determined by the bearing manufacturer after consideration is given to speed, load, temperatures, and lubrication. The spacer is ground to a pair-determined end play amount. Figure 25 shows a two-row tapered roller bearing having a double cone with two single cups.

Figure 25: Two-Row Tapered Roller Bearing with Two Single Cups

Figure 26 shows a two-row tapered roller bearing having a double cone with a cup spacer located between the two single cups. Like the cone spacer, the cup spacer provides the correct amount of running clearance. This is pair-determined by the bearing manufacturer.

Figure 26: Two-Row Tapered Roller Bearing with Cup Spacer

Tapered Roller Thrust Bearing

Taper roller thrust bearings offer the highest possible capacity of any thrust bearing of its size. Its design of the races are that one is perfectly flat, while the other has a tapered raceway that matches the rollers. This bearing was designed for screw-down applications in metal rolling mills where thrust loads exceeding one million pounds are common. Figure 27 shows a cross-section of this tapered roller thrust bearing.

Figure 27: Tapered Roller Thrust Bearing

Journal Bearings

Plain bearings, sometimes called sleeve bearings or journal bearings, were among the first types used to support machine shafts. Their use dates back to the times when hardwoods were the only material from which bearings could be made. Plain bearings still have many uses, and their materials now range from plastics to fine grades of metal and various alloys.

The following paragraphs describe the critical dimensions of plain bearings. The basic parts of the bearing and the function of each part are given in detail. Plain bearings may have to be re-lined from time to time, and the techniques and materials used in re-lining are explained.

The term plain bearing refers to a bearing that has no rolling elements, such as balls or rollers (bearings using balls or rollers are called anti-friction bearings). A very basic type of plain bearing appears in Figure 28. Notice that the bearing closely resembles a simple sleeve, which gives it its other name, sleeve bearing.

Figure 28: Plain-Bearing Critical Dimensions

During operation, a plain bearing supports the shaft along the inside surface, or bore, of the bearing. As the shaft rotates, friction develops between the shaft OD (outside diameter) and the bearing bore. This is the critical area in a plain bearing. Proper types and quantities of lubricants must be supplied to this area, and the materials used in the bearing bore must be able to carry the load imposed.

To prevent metal-to-metal contact, an oil film upon which the shaft journal rides must be sufficiently thick to maintain continuous fluid friction. The mechanism of an oil film is shown schematically in Figure 29.

Figure 29: Oil Film Schematic

It is assumed that the layer of oil molecules immediately in contact with each of the two metal surfaces is partly absorbed by the metal surfaces. The lubricating value of this absorbed oil film is termed "oiliness." The balance of oil molecules adhere to the two surfaces and form a "boundary" film.

If there is an insufficient supply of lubricating oil and the oil film is limited to two thin layers, "thin film" or "boundary" lubrication takes place. This may occur in machinery when starting and stopping or when a sudden impact load is applied to a bearing. It may cause either intermittent metal-to-metal contact or actual film breakdown, with consequent hot spots on the metal and seizure or welding. Some factors that affect this condition are:

  • Roughness of the surface
  • Amount and nature of the oil supply
  • Material composing the surfaces
  • The presence of foreign material
  • Properties of the lubricant

Bearing Maintenance

Competition in the manufacturing sector has reached an unprecedented level, forcing many organizations to seek ways to cut production costs and minimize equipment downtime. One of the easiest and least painful solutions to this problem is to reduce the cost of repair parts.

Good bearing maintenance and installation practices can go a long way toward reducing costs when one considers the fact that millions are spent annually on replacement bearings. Rotating equipment repairs are inevitable but using a few simple techniques will minimize these repairs. This section provides guidance in the following major areas:

Bearing Lubrication

Lubrication is the key to extended life and proper performance of bearings. Too much, too little, or the incorrect types of lubricants can cause several bearing related problems, including catastrophic equipment failure.

Making generalizations about bearing lubrication presents some difficulty because so much variation exists from one application to another. For instance, bearings with two shields and/or two seals, as well as some kinds of pillow blocks, receive lubrication at the factory. If you add more lubricant to these bearings when installing them, they will run hot. This, in turn, can lead rapidly to failure.

In their manuals and engineering data, bearing manufacturers supply a great deal of information regarding the use of oil and grease. Usually however, they do not specify oil or grease by manufacturer. Instead, tables are provided that detail the required characteristics (viscosity, stiffness, etc.) of lubricants for various applications. Suppliers and distributors of industrial lubricants can then recommend a specific lubricant that will meet the requirements of your particular application.

You are best advised to consult with the bearing manufacturer, the lubricant supplier, and your own supervisor to make sure that you apply the correct amount of the right lubricant for the job. To guess about the proper lubricant invites early bearing failure.

Anti-Friction Bearings

Anti-friction bearings must be lubricated to prevent metal contact between the rolling elements, raceways, and cage. Additionally, lubrication protects the bearing against corrosion and wear, helps dissipate heat, aids in sealing, and reduces bearing noise.

The best operating temperature for a rolling bearing is obtained when the minimum of lubricant necessary is used to ensure lubrication. The quantity of lubricant used will also depend on other functions required of the lubricant, such as cooling and sealing.

Anti-friction bearings can be lubricated with grease or oil. The choice of lubricant depends on conditions such as operating temperatures, rotating speeds, loads, and environmental conditions.

Grease Lubrication

Under normal operating conditions, anti-friction bearings can be grease lubricated. Grease has several advantages when compared to oil:

  • Easy to add additional grease
  • Offers better adhesion capabilities
  • Provides protection against moisture and contaminants
  • Easily retained in the bearing housing

The free space in the housing and bearing should only be partially filled with grease. Thirty to fifty percent grease in the bearing housing is considered adequate. Overfilling may cause a rapid rise in temperature, particularly at high speeds. Bearings operating at slow speeds and those that require corrosion protection may have their housings completely full of grease. Additionally, overfilling may prevent a bearing that is designed to float in its housing from operating properly.

The period during which a grease-lubricated bearing will function satisfactorily without re-lubrication depends on bearing size, type, speed, operating temperature, and the grease used.

When operating conditions of equipment are such that re-lubrication can be only carried out at infrequent intervals, it is sufficient, if the bearing housing can be opened to remove as much used grease as possible from the bearing. Then, repack fresh grease between all the rolling members from one side only.

If frequent re-lubrication can be performed on the equipment, some provision is made for re-greasing, usually in the form of a grease nipple fitted to the bearing housing. A grease gun adds fresh grease to the bearing and replaces the old grease.

The lubrication duct in the housing should either feed the grease adjacent to the outer ring face, or, preferably, into the bearing by means of the lubrication groove. After numerous relubrications, the bearing housing should be opened and the used greased removed before fresh grease is added.

Oil Lubrication

Oil lubrication is used when high speeds or high operating temperatures prohibit the use of grease. Oil will transfer frictional heat away from a bearing or adjacent machine parts effectively.

Oil bath systems are suitable for low shaft speeds. Oil is picked up by the rotating bearing elements and after circulating through the bearing, it drains back to the oil reservoir. When the bearing is at rest, as shown in Figure 30, the level of the bath should come to just below the center of the bottom rolling element.

Figure 30: Oil Bath Lubrication

To avoid having many oil changes, because of high operating temperatures causing oil aging, lubrication can be provided to ball and roller bearings by an oil circulation system. A positive-displacement oil pump sends pressurized oil through the bearing housing and into the roller bearing. After the oil’s passage through the bearing, the oil is filtered and possibly cooled before being returned to the bearing.

At high shaft speeds, oil must penetrate the interior of the bearing to remove the excess heat. An effective method for doing this is injecting oil into the bearing. The speed of the oil being injected must be high enough to ensure that sufficient oil penetrates the air vortex created during bearing rotation. Figure 31 shows an oil injection unit.

Figure 31: Oil Injection Unit

How often the lubricating oil has to be changed depends upon the operating condition and the quality of the oil.

For oil bath systems, the oil should be changed more often if the operating temperature of the oil exceeds 120 degrees Fahrenheit, or if the machine operates in an environment where abrasive and fluid contamination is great.

With circulation, the time for an oil change can be determined best by inspecting the oil quality. An oil analysis can determine:

  • If abrasive particles exist
  • If oxidation is present
  • If oil additive breakdown has occurred

Friction Bearings

It is important to monitor the lubricating oil performance, as this can effect the bearing’s service and life.

There are numerous methods used on industrial equipment to supply sufficient lubricant to the journal bearings. Unlike roller bearings, journal bearings just use oil as a lubricant. The more common of these methods are described below.

Oil Ring Lubrication

Oil is supplied to the bearing by a ring in contact with the rotating shaft. The ring rotates and will, within reasonable limits, supply enough oil to the bearing to maintain hydrodynamic lubrication. If the shaft speed is too low, minimal oil will follow the ring to the bearing. If the rotational speed is too high, the ring speed will have difficulty keeping pace with the shaft. A high-speed ring can lose oil by the action of centrifugal force. For best results, the peripheral speed of the shaft should be between 200 and 2,000 feet per minute. Figure 32 shows a ring oiler in a journal bearing.

Figure 32: Ring Oiler

Oil Bath Lubrication

In an oil bath lubrication system, the bushing is partially or fully submerged in oil. It can be a practical method of lubrication if the housing can be made oil tight, and if the shaft speed is not great enough to cause excessive churning of the oil.

Splash Fed Lubrication

Splash fed is a term applied to a variety of intermittently lubricated bushings or journal bearings. The bearings could be splattered with oil from the action of various moving parts regularly dipped in the lube oil. Like oil bath systems, splash feeding is practical when the housings can be positively oil-tight and when the rotating parts do not churn up the oil too much.

Wick Lubrication

Wick lubrication delivers oil to a bushing or bearing by the capillary action of a wick. The amount of oil delivered is proportional to the size of the wick. This system is recommended for small, slow turning shafts and where the loads are light. This system does not provide for any great amount of hydrodynamic lubrication.

Pressure Lubrication

In pressure lubrication, oil under pressure is fed to the journal bearing from a positive displacement pump by a central oil groove, single or multiple holes, or axial grooves. The flowing oil assists in flushing contaminants from the bearing. It also helps to dissipate the heat from the bearing. The oil supply pressure required for journal bearings carrying medium loads depends on the shaft speed, load, and type of oil used. For most journal bearing installations, 50 psi will be adequate.

Bearing Inspections

Inspection of bearings is a very critical step in the performance of maintenance. Performing these inspections will help to prevent bearing failures. The items that are of particular concern in this section are alignments, clearances, and dimensions.


Correct alignment of all devices in contact with the bearing shaft cannot be overemphasized. All the components of the shaft must work together in correct alignment to prevent damage to any of them. A misaligned motor shaft will eventually cause abnormal wear of the bearing, which will cause even greater misalignment. This can be an even greater hazard if a driving gear set is attached to the shaft. Not only can this misalignment be transferred to the gears but it may also go on to cause bearing damage and misalignment to the components being driven.

Along the same lines of alignment, a soft foot within a large bearing housing or pillow block can cause uneven bearing wear. Prior to aligning a unit, a soft foot check should be performed on the bearing housing. This check must be completed to guarantee the proper reaction of any realignment shim changes that will be done.


The clearance designed into a bearing by the manufacturer is another very important aspect of bearing life. In both roller and journal bearings, the correct clearances will help to ensure proper operation. Not to use the manufacturer's specifications for clearances will almost guarantee premature bearing failure. In this section, we will discuss how these clearances work and methods of checking them for both roller and journal bearings.

Roller Bearings

Bearing clearance is the amount one bearing ring moves in relation to the other in a radial or axial direction. For radial bearings, the radial clearance is of primary importance. As a general rule, the radial clearance in ball bearings should be close to zero when the bearing has been mounted and has reached its operating temperature. The following are clearances that may be required to be measured in radial bearings:

  • Clearance - The space between the rolling member and the outer ring of a ball or roller bearing measured with the bearing standing in an upright position.
  • Initial Clearance - The clearance in the bearing before it is heated or mounted.
  • Installed Clearance - The remaining clearance in the bearing after it has been installed on a shaft or in a housing, or the clearance left in a tapered bore bearing after it has been pushed up the taper.
  • Running Clearance - The clearance in the bearing once the machine reaches its operating temperature and adequate lubrication has been applied.

In Figure 33, the most common method for determining the clearance in a ball or roller bearing is shown. The steps involved in checking these clearances are:

  1. Rotate the inner ring to properly seat the rollers.
  2. Measure the internal clearance in the bearing by inserting progressively larger feeler blades the full length of the roller between the uppermost-unloaded roller and the outer ring sphere. Do not roll over the feeler blade, slide it through the clearance.
  3. Record the measurement on the largest sized blade that will slide through. This is the unmounted internal clearance.
  4. Repeat this procedure in two or three other locations by resting the bearing on a different spot on its O.D. and measuring over different rollers in one row.
  5. Repeat the above procedure for the other row of rollers or measure each row alternately in the procedure described above.

Figure 33: Measuring Internal Clearances

The inner rings of bearings with tapered bores are always mounted with an interference fit. The amount of interference depends on how far the bearing is driven up on its tapered seat. The initial clearance of the bearing will be reduced gradually as the inner ring expands.

The manufacturer of the bearing should have clearances listed for the amount of reduction. If no radial clearances are given, a rule of thumb for reducing the radial clearance on spherical roller bearings mounted on tapered seats is to reduce the initial clearance by 50%.

In some applications, the manufacturer may require other than normal internal clearance. Ball and roller bearings with special internal clearances are designated by a suffix after the bearing number as follows:

  • C1 - Radial clearance is less than C2
  • C2 - Radial clearance is less than normal
  • C3 - Radial clearance is greater than normal
  • C4 - Radial clearance is greater than C3
  • C5 - Radial clearance is greater than C4

When mounting self-aligning ball bearings on tapered seats, the radial clearance can be checked by rotating and swiveling the outer raceway during the tightening procedure. A light drag on the bearing will be felt when the initial clearance is reduced. It must be possible to rotate the outer ring with ease, but there should be a resistance to swiveling.

Journal Bearings

Establishing correct bearing clearance is essential for reliable performance of journal bearings. Excessive bearing clearance will result in poor load distribution within the bearing, increased fatigue, and excessive shaft deflections. Insufficient bearing clearance may produce excessive operating temperatures, thermal growth, increased journal and bearing wear, poor oil flow, and eventual bearing failure and equipment seizure.

Journal bearings may be manufactured with an initial clearance. The initial clearance can be altered by the fits used on the shaft and/or housing and by the thermal growth developed during operation. The clearance remaining in the journal bearing during full operation is often termed running clearance. Running clearance requirements vary with the nature of the application, load, speeds, and type of bearing material.

Bearings for high-speed applications are designed for greater running clearances. The additional clearance allows for less predictable thermal differentials. Low-speed equipment often involves heavier loads, and the journal bearings are designed for reduced running clearances to obtain optimum load distribution and increased fatigue life.

As a general rule, for any journal bearing assembly with a constant one direction load and operation, the bearing clearance can be in the medium to loose fit range. For equipment that reverses rotation periodically, and where fluctuating loads are present, the bearing clearances can be in the medium to tight fit range. For high-speed internal combustion engine bearings, using a force feed lubrication system, medium fits are recommended.

Four methods can be used to determine the clearance in a journal bearing: feeler gauge, dial indicator, plastigage, and lead wire.

Feeler Gauge Method

With the two bearing halves properly tightened and torqued to required specifications, initial clearances can be found on some equipment by inserting feeler gages between the journal and the shimmed bearing cap. This method works well if the bearing is open at either end, allowing the feelers to be inserted.

Dial Indicator Method

When a journal bearing is shielded at either end by housings, gears, etc., the simplest method to check for bearing clearance is to mount a dial indicator on a magnetic base and have the indicator point contact the shaft. By carefully prying or lifting up on the shaft and closely watching the dial, you can determine the bearing clearance. It is recommended that jacks be used at both ends to lift the shaft evenly for an accurate dial reading. If only one jack is used, the shaft may tilt within the journal bearing and give a false dial reading.

Figure 34 shows the dial indicator method for determining the clearance within a journal bearing. A dial reading will indicate the total clearance in the bearing but will not identify the high and low spots. Any wear checks must be performed visually.

Figure 34: Checking Clearance With a Dial Indicator

Plastigage Method

Plastigage is a soft material, which will easily deform when squeezed between the journal and the bearing with the cap bolts properly torqued. A small length of plastigage is placed on the bearing half. The plastic sits on the center-bottom portion of the bearing parallel to the shaft’s axis. Be careful not to roll the shaft.

After removal of the bearing cap, the flattened material is compared with a prepared gage chart and the tolerance can be read in thousandths of an inch or hundredths of a millimeter directly from the chart. Figure 35 shows how a plastigage reading is taken from the bearing half and compared to the standard chart.

Figure 35: Checking Clearance With Plastigage

Lead Wire Method

In the lead wire method, lengths of lead wire are placed across the shaft in several positions with the bearing cap removed. The cap is torqued to the recommended amounts, and then removed again.

The thickness of the squeezed lead wire is measured with a micrometer to determine the amount of bearing clearance. Care must be taken when removing the cap because some wire may adhere to the cap and some to the journal.

Record the thickness of each compressed wire to help determine if there are any high and low spots in the bearing/journal assembly. Use a small diameter soft lead wire to avoid damaging the bearing material.

Roller Bearing Installation

The installation process used for ball or roller bearings is one of the most critical steps associated with increasing bearing life. It will do no good to take the time to check clearances and perform alignments if damage is done to the bearing upon installation.

Bearing Installation Precautions

Listed below are some precautions to take before installing roller bearings:

  • Check that the bearing number stamped on the package of the new bearing agrees with the designation given on the drawing or parts list.
  • Be acquainted with the procedure for mounting. Study and understand the machine drawing and design details of the application and assembly sequence.
  • Preserve new bearings in their original packing with the anti-corrosive lubricant. Washing out this lubricant is not required when mounting the bearing.
  • Never attempt to modify the bearing. Any drilling, machining, or cutting on the bearing will result in premature failure.
  • Protect disassembled bearings and components from foreign contaminants.
  • Keep the work area clean, dry, and well removed from any welding or machining operations.
  • Do not clean bearings with compressed air.
  • Never hold a roller bearing and spin the outer race with compressed air. Friction and heat will reduce the radial clearance causing the bearing to quickly seize or explode.
  • Thoroughly clean the shaft, housing, and any other mating parts.
  • Flush out all lubrication holes, passages, and ports if possible, and ensure that they will freely allow passage of the clean lubricant.
  • Ensure that the vent holes in the housing are free of obstructions. Seal failure and lubrication overheating may occur.
  • Ensure the bearing seats on the shaft or sleeve and in the housing are free of burrs, nicks, and sharp edges.
  • Shaft and housing tolerances should be checked. A fit that is too loose may cause the inner ring to creep on the shaft. The working accuracy of the machine will also be affected. Excessively tight fit leads to a reduction in radial clearance and this may result in detrimental preload and hot running of the bearing.
  • Check the seating areas for diameter and out of roundness. Measure the shaft and housing diameter from top to bottom, and from side to side. This may indicate out of roundness. Positioning the shaft between centers on a lathe and running a dial indicator on the bearing seat area may also indicate if there are any high spots or out of round conditions. The walls of the bearing rings are relatively thin, any high spot or out of roundness on the shaft or in the housing is transmitted to the bearing raceways. Additionally, any mismatch between the bearing cap and base of split housings will be transferred to the bearing. This can be indicated by vibration analysis.
  • Check the seating area for surface finish and contact area. The shaft should have a surface finish of 63 RMS and provide at least 80% contact to the bearing bore.

Interference Fits

An interference fit, sometimes referred to as a tight fit, is a fit where the bearing inner race inside diameter (ID) is actually smaller than the shaft outside diameter (OD) it rides on. In this situation, the inner race is the tight ring.

The interference produced by tight fits expands the inner ring and contracts the outer ring. This reduces the initial radial clearance. The radial clearance should be adapted to the fit. Table 1 and Table 2 are provided by bearing manufacturers that indicate maximum and minimum shaft and housing diameters for specific bearing ID and OD sizes.

Table 1

Table 2

Methods of Mounting Bearings

Bearings mounted to a shaft or within a housing may be mounted hot or cold depending on the type of bearing and required fit. In this section, we will cover bearing heating and cold mounting, along with tapered bore bearing mountings.

Bearing Heating

A method of mounting a bearing to a shaft is by heating it. This process is generally used on medium to large sized bearings since smaller bearings are usually pressed on cold. The heating of bearings should only be accomplished by one of the three following methods:

When heating bearings, accurate temperature control is essential to ensure no damage is done to the bearing. If the temperature exceeds 250 degrees Fahrenheit, there is a risk of altering the bearing grain structure resulting in a drop of hardness and wear life. Additionally, sealed bearings should never be heated but cold mounted only.

Never place a flame directly on a ball or roller bearing’s surface. The flame’s temperature far exceeds the maximum heating temperature recommended for anti-friction bearings. Heating a bearing with a torch is almost a guarantee that it will fail much earlier than designed.

Bearing Heaters

Bearings that must be mounted with an interference fit on the shaft can be safely heated fast and without contamination, using an electric heater. The two types that are in use are the induction bar-type and the cone-type. Each comes with a number of control options such as heating a bearing for a specified time or to a specified value.

In the bar-type induction heater, shown in Figure 36, the bearing is positioned between the two supports of the heater by using one or more metal bars through the bore of the bearing. After a bearing is heated on this unit, it is automatically de-magnetized. The cone-type heater can support a much larger variety of bearings due to its varying OD as you move down the cone. A disadvantage of the cone-type heater is that it uses electric coils for heating and therefore can take up to twenty minutes to perform adequate heating.

Figure 36: Bearing Heaters

Electric Ovens

An oven using an electric coil or a heat lamp can be used to heat bearings. This is a simple, reliable, clean method of heating a bearing. The oven must be free from contamination, and it is best if the bearings are suspended. Use a thermometer to check the heat inside the oven. A temperature control switch may be incorporated to shut off the oven at a specified maximum temperature.

Hot Oil Bath Bearing Heaters

Heating a bearing in oil is also a simple and reliable method. A clean oil tank with a cover should be used. The bearing should not be in direct contact with the sides or the bottom of the receptacle, but should be placed on some type of support or suspended in the oil bath. Figure 37 shows a bearing being heated in an oil bath. After the bearing is heated, it should be carefully wiped off using a lint-free cloth.

Figure 37: Hot Oil Bath

Installing a Heated Bearing

When the hot bearing is started on the shaft, it must be moved quickly to its position to prevent seizing. The shaft must be prepared and measured to help avoid detrimental effects. If the bearing should start to bind, it must be removed quickly and reheated.

For a tight fit of the outer ring, the housing should be brought up to mounting temperature. This may be inconvenient for large bearing housings. A bearing may be contracted by cooling in a mixture of dry ice and alcohol or with liquid nitrogen.

Although the bearing or housing may be hot, it may be necessary to use a press during installation depending on the total interference of the fit. It is also possible to cool a shaft and heat a bearing to provide for the maximum amount of clearance during installation.

Cold Mounting

Another method used in the mounting of roller bearings is cold mounting. Ball and roller bearings vary greatly in type, design, and size and there is no universally applicable mounting procedure. For non-separable bearings, the mounting and dismounting forces must be applied directly to the ring that is being fitted. When a bearing is mounted on a shaft, the pressure must be applied to the face of the inner ring. If the pressure is directed through the wrong ring, damage to the rolling members and raceways is likely to occur. Never hammer directly on the hardened bearing rings.

The sequence for mounting bearings is important. If a tight fit is required for the inner ring of a non-separable bearing and a loose fit for the outer ring, the bearing, as shown in Figure 38, will first be mounted on the shaft, then the shaft and bearing will be pushed into the housing.

Figure 38: Mounting a Non-Separable Bearing

Figure 39 shows how mounting separable bearings is simpler because the inner and outer raceways can be mounted separately.

Figure 39: Mounting a Separable Bearing

Small bearings may be mounted with the aid of a mounting sleeve as shown in Figure 40. Place the sleeve against the inner ring and use a standard steel hammer to drive the bearing in place.

Figure 40: Mounting Sleeve

Hammers with lead or other soft metal faces are unsuitable for bearing work since metal fragments can break off and enter the bearing. The bearing must start straight and square onto the shaft.

The safest and most acceptable method of cold mounting cylindrical bore bearings is to use an arbor or hydraulic press. Care must be taken to properly support the inner race with spacers, and the ram must push the shaft squarely. Figure 41 shows an example of a hydraulic mounting press.

Figure 41: Hydraulic Mounting Press

A major advantage of using a press for mounting small bearings is that the fitter can feel the resistance offered as pressure is applied to push the bearing onto the shaft. If the resistance feels too great, it could be an error in shaft size, or a damaged shaft surface causing abnormal interference.

Mounting Tapered Bore Bearings

Bearings with a tapered bore can be either fitted directly onto a tapered shaft journal or, if the shaft is cylindrical, the bearings are mounted on an adapter sleeve or a withdrawal sleeve. Apply a thin oil coat to the seating areas of the bearing, shaft, and sleeve. This helps to reduce friction and facilitate mounting.

Pushing a bearing onto a tapered seat expands the inner ring, reducing the radial clearance. To find the required reduction in radial clearance, it is necessary to determine the initial radial clearance before mounting, and then check the clearance repeatedly during mounting until the proper amount of reduction and fit is obtained.

Figure 42 shows how a tapered-bore spherical roller bearing can be driven up a tapered shaft seat using a threaded shaft nut. This is an accurate method for reaching the desired reduction of radial clearance. The nut is tightened with a C-spanner.

Figure 42: Shaft Nut Mounting

Figure 43 shows how a spherical roller bearing can be mounted onto a tapered sleeve. A specially designed shaft nut is used to press the withdrawal sleeve into the space between the shaft and bearing inner ring.

Figure 43: Withdrawal Sleeve Mounting

Figure 44 shows how the spherical roller bearing can be removed from the withdrawal sleeve. By slackening off the shaft nut, then using a removal nut threaded onto the withdrawal sleeve, the sleeve is extracted from the bearing bore and off the shaft.

Figure 44: Withdrawal Sleeve Removal

Figure 45 shows how a tapered adapter sleeve nut is used to push the tapered bore spherical roller bearing up onto the tapered seat of the sleeve. A C-spanner is used to tighten the nut. A major advantage of using adapter sleeves for mounting ball and roller bearings is that the bearing can be mounted at any point on a shaft.

Figure 45: Taper Adapter Sleeve Mounting

Figure 46 shows dismounting of an adapter sleeve mounted bearing. Loosen the adapter sleeve nut lock washer and slightly back off the nut. Drive the inner ring from the sleeve. Use a soft steel drift to avoid damaging the bearing. Brass drifts should not be used since fragments can enter the bearing and cause damage.

Figure 46: Taper Adapter Sleeve Removal

Preloading Duplex Bearings

The main purpose of duplex bearings is to achieve greater axial and radial rigidity than is possible with one single row bearing. The extra stiffness in these bearings is obtained by preloading. Preloading is a predetermined value that is achieved by grinding a certain amount of material off the inner or outer ring faces. When the bearings are mounted and the faces clamped together, an internal preload occurs by one bearing opposing the other. This preload decreases deflection from external loads applied to the clamped-up pair.

For each bearing size, manufacturers have established standard preload levels which are considered proper for most applications. Special preloads can also be achieved through the use of spacers or by grinding the bearing faces. Care must be used when changing the preload because although it can provide greater rigidity, it reduces bearing life and increases power consumption.

A typical axial deflection curve for a non-preloaded bearing is shown in Figure 47 as curve A. As axial thrust increases, deflection increases. However, if thrust is doubled, deflection also increases. If thrust is doubled, deflection is not necessarily doubled since the greatest amount of deflection occurs initially. Curve B shows deflection of a preloaded set to T1 and Curve C shows that same set with the preload doubled to T2.

Figure 47: Axial Deflection Curve

Setting Tapered Roller Bearings

One of the advantages of tapered roller bearings is that they can be set to meet specifications and performance for the particular machine in which they are used.

The setting of tapered roller bearings depends on the flexibility or rigidity of the housings and shaft, the lubrication, speed, and bearing spread. There are many machine variations in design, and the variation will govern the setting on the tapered roller bearing.

One of the two following general rules should be followed when setting the bearing; otherwise, the machine manufacturer should be contacted.

1. In general, the tapered roller bearings located in gear reduction units, transmissions, wheels, rolls, agitators, screw conveyors, mill bearings, and similar applications are set with a free running clearance.
This clearance or endplay may vary from .002 inch to .005 inch on small diameter bearings to .007 inch to .010 inch on larger diameter bearings.
It may be necessary to hold shafts, gears, or spindles absolutely rigid to obtain proper operation. With these types, the tapered roller bearings are set with a necessary preload without damaging the bearing. The amount of preload depends on the machine in which the bearings are used.
Preloaded bearings are often used in machine tool spindles, heavy equipment pinions and differentials, automotive differentials, or other machinery where shafts must be held rigid. The amount of preload is usually given in inch-pounds (in./lbs).
2. The general rule for pinion bearings in these applications is 10 to 20 in./lbs torque and 25 to 35 in./lbs torque for differential bearings. Setting tapered bearings on machine tool spindles depends on design and speed. Where spindle speeds are slow, preloads are often used, but where speeds are high, the bearings are set with a slight amount of endplay to compensate for thermal expansion.

Cone-Adjusting Devices

Setting devices for tapered roller bearings are classified as either cone-adjusting devices or cup-adjusting devices. Figure 48 shows a slotted hex nut and a cotter pin on a threaded shaft. The nut is tightened up until the required free running clearance or preload is reached. The nut and washer should be large enough to provide adequate force to the cone.

Figure 48: Slotted Hex Nut

Two cotter pins spaced at 90 degrees are used to obtain numerous locking positions per revolution of the nut and subsequently a closer bearing setting.

A bearing arrangement where shims and an end cap, held in place by cap screws, are used to set the cone is shown in Figure 49.

Figure 49: Shaft End Plate and Shims

Minimum shim thickness between the end cap and shaft end makes for a closer bearing setting, while more shimming provides for more clearance. The cap screws are wired together or lock washers can be used to keep the unit secure. A slot in the end plate may be provided to measure the shim gap.

Figure 50 has a cone-adjusting device, which uses two locknuts and a lock washer. The inner nut is tightened until there is a bind on the bearings, then backed off enough to ensure proper running clearances exist. The outer jam nut is then tightened. This adjusting device is used where slotted hex nuts are not practical and it provides an accurate setting.

Figure 50: Two Lock Nuts and Lockwasher

Cup-Adjusting Devices

Figure 51 shows a cup-adjusting device using shims. Shims are located between the cup follower plate and the housing. The shim amount used will provide the proper bearing running clearance recommended for the particular machine. Removing shims makes for a closer bearing setting, while adding shims increases the bearing clearance. The cup follower is held in place by cap screws, which are locked with lockwashers or wired. This clearance can also be achieved by machining either the housing or cup follower plate to set the desired clearance.

Figure 51: Shim Adjusting Device

Figure 52 shows a cup set by a special cup carrier and shim. Shims are located between the cup carrier flange and the housing. Gearboxes often use this type of cup setting device.

Figure 52: Cup Carrier and Shims

Figure 53 shows how a threaded cup follower is used to set the cup. The threaded cup follower is tightened up until the running clearances are reached. The follower is locked in place by a plate and cap screws.

Figure 53: Threaded Cup Follower

Setting Procedure

Threaded shafts with a nut or housing using a cup follower are common bearing setting devices. The recommended procedure is to tighten up the nut while rotating the bearings until there is a noticeable drag or until all the end play has been removed. This procedure assures seating of the bearing parts. If endplay is desired, the nut is backed off. If the bearings require preloading, the nut is torqued to a specific value.

If locknuts or double nuts are used, some allowance must be made for tightening of the setting when the nuts are locked. Clearance or looseness in the threads of the locked nut will push against the bearing and tighten the original setting.

When shims are used, the setting can be made by tightening the end cap or bearing cup carrier until the bearings bind slightly while being rotated, without the shims in place. The gap between the end plate and housing or end plate and shaft is measured with feeler gages. The endplay desired plus the measured gap gives the total shim amount required. If a certain preload is desired, subtract the amount of preload from the gap and the result is the shim pack required.

Lead wire can also be used to determine bearing settings. Place the lead wire in the space where the shims would be located. The end plate or bearing cup carrier is then tightened until all the play is taken up. The wire, which has to be larger than the original gap, is flattened during the tightening procedure. The lead wire is measured with a micrometer and the shim pack required is determined. Desired endplay or preload can also be calculated with this method.

Axial Positioning

An interference fit is required for holding a bearing in position but a positive locking device may also be needed. This is found on equipment where the bearing has been installed with minimum interference for ease of installation and removal.

Methods of Positioning

The following are methods used for positive axial position of the inner race:

  • External snap ring
  • Lock nut and washer holder bearing against a shoulder on a straight shaft
  • Lock nut and lock washer holding tapered bore bearings on tapered shafts
  • Adapter sleeve mountings
  • Withdrawal sleeves with the inner race mounted against a shoulder on the shaft

At times, it may also be desirable to position the outer race axially. The following are the most common methods of this:

  • End covers with or without shims and cap screws
  • Threaded end covers pushing against the outer ring
  • Internal snap ring

Fixed and Floating Bearings

A temperature variation will expand or contract the internal parts of any machine. Because of this condition, it is essential that such parts be permitted to expand or contract without restriction. For this reason, only one bearing on any one shaft should be fixed axially in the housing to prevent axial motion.

All other bearings on that same shaft should have adequate axial clearance in the housing. These are commonly referred to as floating or free bearings. Floating bearings permit the shaft to expand (stretch) and contract without restriction.

Figure 54 A shows how a bearing is fixed in a housing. By using two stabilizing rings, placed on either side of the outer race, the bearing is located axially. Also shown in Figure 54 B is a floating bearing in a housing. The floating bearing is usually located in the center of the outer ring seating area of the housing.

Figure 54: Fixed and Float Bearings

Fixed and floating bearing arrangements do not include axial-control bearings such as tapered roller, and angular contact ball bearings. It is usually best to fix the bearing that is located nearest the drive end of the shaft.

The load on the machine may dictate which bearing is fixed. It may be more of an advantage to fix the bearing that carries the smaller radial load, especially when the fixed bearing is receiving a thrust load of some magnitude.

An arrangement like this tends to equalize the distribution of the load on the bearings. Equipment using flexible couplings connecting two shafts should use a fixed bearing on each shaft end because a flexible coupling usually permits some end motion on both shafts.

Roller Bearing Removal

When removing roller bearings, care should be taken not to unnecessarily damage them. The reasons for this are that it may be desirable to reuse the bearing and the least amount of damage done during removal will aid in failure analysis. If possible, the removal force should always be directed towards the inner ring and never through the rolling elements.

Bearing Pullers

Small bearings can best be removed by using a two or three-legged puller. If the bearing is mounted with an interference fit on the shaft, the puller should engage the inner ring. A center hole should be provided on the end of the shaft as the puller must be accurately centered to help avoid bearing damage and for user safety.

In some cases where it is impossible to engage the inner ring, the puller legs can engage the outer ring but the outer ring must be rotated during removal to avoid bearing damage. Figure 55 shows the use of a bearing puller.

Figure 55: Bearing Removal with a Puller

Bearing Press

Figure 56 shows how a bearing can be removed from a shaft using a press. The inner race is supported by the heavy plate and the ram pushes down on the shaft squarely and evenly. The bearing and press should be covered or caged in case the bearing suddenly jumps or fractures.

Figure 56: Bearing Removal with a Press

Journal Bearing Maintenance

The proper mounting of a friction (journal) bearing to the shaft and machine frame is a critical factor in the performance of the unit. The journal bearing is sensitive to shaft deflections, misalignments, distortions, vibrations, and surface imperfections of the mating machine elements.


When using replaceable journal bearing inserts, it is important that the inserts have positive contact with the housing or seat. To assure this, the diameter of the two matching inserts, when placed together at right angles to the parting line, is slightly larger than the diameter across the parting surface when the bearing is in place; thereby requiring this amount to be compressed when the bearing cap is torqued in place.

Figure 57 identifies two connecting rods where proper crush height exists. The excess height is called the crush and its primary purpose is to permit the insert to be positively locked into the bearing seat. If the bearing has insufficient crush it will not be held securely and will have a slight amount of play during operation. Loose inserts will allow the lubricant to work its way between the back of the bearing and the housing. This reduces heat conductivity and eventually raises the bearing temperature.

Figure 57: Bearing Crush

Insufficient bearing crush can be caused by filing or grinding of the parting surfaces of the shells or because of the presence of foreign particles lodged between the parting faces of the bearings and bearing caps. The dirt acts as a shim, which prevents the faces from coming together.

Under no circumstances are the parting surfaces of the inserts, caps, or saddle to be ground or filed. Do not attempt any operation on the bearing insert other than correcting the spread, and this is done only when necessary. The spread is built into the bearing so that the inserts have to be lightly pressed into position. If the inserts have excessive spread, they can be tapped lightly on the end to close up the spread. If the inserts have insufficient spread placing them on a wooden block with a convex side and gently tapping the insert with a hammer can open them.


It is necessary to use proper torque sequences in addition to recommended bolt torque values and a torque wrench when locking a bearing cap down during assembly. Any variations in torque amounts and sequence can affect the bore size, bearing crush, clearances, and bearing performance.

Babbitt Bearing Fitting

Any journal bearing, regardless of its shape, diameter, or material, must be accurately prepared in the following areas:

  • The surface finish and area of journal to bearing contact
  • The running clearances
  • The lubrication entry area, which enables an effective oil wedge to form on hydrodynamic bearings
  • The oil grooves, which help distribute the lubricant

After a bearing has been babbitted, it has to be prepared several ways before it can be operated. Figure 58 shows how the top edge of the bearing half must be chamfered almost to the bearing end to ensure that the lubricant is channeled to the shaft. The equipment manufacturer usually specifies the amount of the chamfer angle.

Figure 58: Bearing Chamfer

For large journal diameters, the chamfer is usually extended on the oil entry side, almost down to the area of contact between the shaft and bearing. The chamfer should not extend into the load area or high-pressure section of the bearing.

To properly check a journal bearing for contact, use either a mandrel or the machine’s shaft. For a good contact impression, the shaft can be lightly coated with non-drying mechanics bluing. As the shaft is slowly rotated in the bearing, the bluing wipes off at the points of contact and transfers to the bearing surface to identify the significant high spots.

Figure 59 A shows a journal bearing which has been relieved and chamfered. The contact points are mostly below the chamfer but not sufficient enough in the bottom area of the bearing.

Figure 59 B shows a journal bearing which has no contact on its bottom half. Only high spots along the top edges appear. The top edges must be scraped down to allow the shaft to make contact with the bottom of the bearing. Fitting for all contact is usually performed after the chamfers and oil grooves have been cut.

Figure 59 C shows a properly fitted bearing where sufficient contact has been made between the shaft and the bearing half. Equipment and bearing manufacturers may recommend a contact area from 75% to 90%. This will depend upon loads, speeds, bearing materials, and the type of lubrication system used. To ensure that a good distribution of lubricant occurs, the bearing cap should also be relieved and chamfered in the same manner as the base.

Figure 59: Bearing Fits

Bearing Scraping

A scraper is used to remove minute amounts of metal to reduce the high spots on the bearing surface that may have previously been made as true as possible by machining.

Scrapers are made of hardened carbon steel and are shaped to suit different scraping and fitting operations. Three types of scrapers in general use is shown in Figure 60.

Figure 60: Bearing Scrapers

Flat Scrapers

Flat scrapers are generally used to remove high spots on flat surfaces. The end of the blade is ground to form an accurate cutting edge. Figure 61 shows the use of a flat scraper.

Figure 61: Using a Flat Scraper

  1. Hold the handle in the right hand, with the thumb and first finger running along the handle in the direction of the blade.
  2. Lay the cutting edge of the blade on the surface and raise the handle to approximately 20 degrees from the horizontal.
  3. Place the left hand approximately half way along the blade, fingers pointing downwards.
  4. Push the scraper forward with the right hand and, at the same time, apply light downward pressure with the left hand. The stroke should be less than one inch (25.4 mm).
  5. Keep the blade in contact with the surface, release the pressure, and draw back the scraper into position for the next stroke.
  6. When the high spots have been scraped in one direction, scrape them a second time at right angles to the original direction until the desired surface finish is reached.
  7. Test the surface frequently with a straight edge or block until the contact markings indicate that the complete surface is in contact.

Curved Scrapers

Curved scrapers and spoons are normally used to scrape hollow surfaces such as the inner surface of a journal bearing. Figure 62 shows an example of a babbitt spoon. The sides and point of curved scrapers and spoons are ground to form an accurate cutting edge.

Figure 62: Babbitt Spoon

  1. Hold the scraper handle in the right hand, thumb and first finger along the handle in the direction of the blade.
  2. Grip the blade with the left hand near the work piece, fingers curled under the blade to allow the cutting edge to contact the bearing surface.
  3. Draw the blade upwards with the left hand with even pressure, and, at the same time, turn the handle slightly with the right hand to keep the cutting edge at the correct angle.
  4. Keep the blade in contact with the surface, move it without applying pressure into position to make the next cut.

The following is an alternative scraping method. Hold the scraper in the same method as previously described, but now rest the left hand on the vice. In this position, the left hand is used as a fulcrum and the right hand controls the cutting action of the scraper by rotary lever movement of the handle.

Metal removal must be done evenly without chatter marks. The bearing surfaces should be checked frequently in conjunction with the part with which it has to be fitted.

Scraping of a journal bearing must be confined to the areas indicated by the visible high spots. A general scraping of a surface will not make the bearing surface any truer and will only remove metal from places where it is needed. Metal cannot be replaced after removal.

Three-Cornered Scrapers

Three-cornered scrapers are triangular in cross-section, giving three cutting edges. These scrapers are generally used for de-burring and chamfering bored and drilled holes on small bushings and bearings and to chamfer journal-bearing edges.

Bearing Care

This section covers some of the considerations in ensuring proper care of bearings. Items covered are: the cleaning of bearings, protection and temporary storage, and handling practices.

Cleaning Bearings

The proper installation of anti-friction bearings is very important for their proper operation. An anti-friction bearing must have running clearance between the balls or rollers and the races. The improper fitting of bearings on the shafts can reduce the normal clearance and cause bearing failures.

The following general items should always be followed when handling bearings:

  • Petroleum-based solvents are recommended for cleaning; leaded gasoline should never be used. The solvent and containers used to clean ball or roller bearings must be clean. Do not put bearings to be cleaned into tanks; clean them separately. Do not let them lie in the bottom of the tank, it should have a wire or mesh platform so that contaminants loosened by the solvents will settle to the bottom of the tank.
  • A new, soft bristle paintbrush is commonly used to assist in cleaning. Cloths or wiping rags must be clean and lint-free. Never use cotton waste because of loose fibers, which can get into inaccessible places and be overlooked when reassembling the bearing.
  • Check the races for scoring, fractures or chipping.
  • The ball or roller bearings should be free of flat spots, spalls or evidence of out of round wear. The ball and roller separators should hold the bearings in positive alignment, not be too loose or show signs of weakness. Shields (grease plates) should also be inspected; these keep the grease in and dirt out of the bearing. Bearings with removable seals should be checked for condition of the lips.
  • These components are discussed and illustrated further along in this article; however, inspecting assembled bearings is very difficult. No economy is achieved in trying to salvage a low-cost bearing.
  • Reassembly should be performed in a clean area if possible. Avoid contamination and exposure to dust-laden air. The workbench should be clean and preferably covered with a shock absorbing type of material such as foam rubber or heavy cardboard. Spread clean newspaper or lint-free shop cloths over this surface.

Protection and Temporary Storage

After cleaning, inspecting, and reassembling a bearing that is to be reused, certain added precautions should be taken. It must be oiled or re-greased immediately, then protected from contamination by airborne dust or dirt, which is always present. The recommended procedure is to wrap and seal it in a heavy grade of oil-treated paper until it is ready to be reassembled into the machine. If long-term storage is contemplated, it should be boxed, preferably in the carton from which a new bearing has been removed, and marked with the correct type, make, and other pertinent information necessary to identify it. This saves from having to unwrap and handle it, or several others, in search of a spare bearing.

Handling Practices

The following are the Do’s and Don’ts of handling bearings.


  • Remove all dirt from bearing housings, before exposing the bearing.
  • Treat a used bearing as carefully as you would a new one.
  • Work with clean tools in clean surroundings.
  • Handle with clean, dry hands, or wear clean, canvas gloves.
  • Use only clean solvents and flushing oils.
  • Lay bearings out on clean paper or lint-free rags.
  • Protect disassembled bearings from dirt and moisture, which can cause rust.
  • Use only clean rags to wipe bearings, never cotton waste or loose fiber cloths.
  • Keep bearings wrapped in dry-oiled paper when not in use (even for short periods of storage).
  • Clean the inside of bearing housing and shaft surfaces thoroughly before replacing bearings.
  • Before mounting a bearing, including press fit types, apply a light coat of oil to the shaft and the bore of the inner race; removal at a later date will be much easier.
  • Make sure it fits. Improper fits can expand the inner race and take up all the bearing clearance. This will cause the bearing to run hot and fail prematurely.
  • When assembling the bearing in the pillow block or mounting, care must be taken so as not to crush the outer race. Excessive interference fit will cause the outer race to be crushed, causing the bearing clearance to be reduced and prematurely fail.


  • Do not use dirty, brittle, chipped or improvised tools.
  • Do not work on wooden bench tops, without suitable covering, or use wooden mallets.
  • Do not use tetra-ethyl lead gasoline as solvents, use safety approved petroleum solvents.
  • Do not spin uncleaned bearings.
  • Do not spin a bearing with compressed air; many air systems in steel mills contain considerable water. In addition, high-speed rotation without grease or thorough oil is injurious to a bearing. Water in a bearing will cause rust.
  • Do not use cotton waste or dirty cloths to wipe or handle bearings during maintenance operations, inspection, and reinstallation. Possible contamination from foreign materials may result.
  • Do not handle a new or reconditioned bearing with bare hands; cover and transport with a clean rag or use clean gloves.
  • Do not nick or scratch any polished bearing surfaces. Avoid dropping any bearing.
  • Do not remove the bearing's protective wrapping and coatings until the bearing is to be used.
  • Do not overheat a bearing. If a bearing is heated much over 250 degrees Fahrenheit, damage can be done to the hardness of the bearing. Under no circumstances should a bearing ever be heated with open flame. A bearing that has been heated with a flame is noticeable by its discoloration.
  • Do not tighten retainers too tight on taper bearings. To do so on bearings without spacers will eliminate the bearing's diametrical clearance, causing the bearings to run hot and fail prematurely.
  • Never use a hammer on the bearing. This will cause the balls or rollers to indent and damage the races.
  • Do not use a puller on the outer races to pull the bearing off the shaft. By doing this, damage can be done to the rollers or races.

Bearing Failure Analysis

Since the bearings of a machine are among its most vital components, the ability to learn as much as possible from bearing failures is of utmost importance.

In designing the bearing mounting, the first step is to decide which type and size of bearing should be used. The choice is usually based on a certain desired life for the bearing. The next step is to design the application with allowance for the prevailing service conditions. Unfortunately, too many of the ball and roller bearings installed never attain their calculated life expectancy because of something done or left undone in handling, installation, and maintenance.

The calculated life expectancy of any bearing is based on four assumptions:

  1. Good lubrication in proper quantity will always be available to that bearing.
  2. The bearing will be mounted without damage.
  3. Dimensions of parts related to the bearing will be correct.
  4. There are no defects inherent in the bearing.


However, even when properly applied and maintained, the bearing will be subjected to one cause of failure: fatigue of the bearing material. Fatigue is the result of shear stresses cyclically applied immediately below the load carrying surfaces and is observed as spalling away of surface metal.

Although spalling can be readily observed, it is necessary to discern between spalling produced at the normal end of a bearing’s useful life and that triggered by causes found in three major classifications of premature spalling. These classifications are lubrication, mechanical damage, and material defects. Most bearing failures can be attributed to one or more of the following causes:

The actual beginning of spalling is invisible because origin is usually below the surface. The first visible sign is a small crack, which cannot be seen nor can its effects be heard while the machine operates. Figure 63 through Figure 65 illustrate the progression of spalling. The spot on the inner ring in Figure 63 will gradually spread to the condition seen in the ring of Figure 65 where spalling extends around the ring. By the time spalling reaches proportions shown in Figure 64, the condition should make itself known by noise.

Figure 63: Incipient Fatigue Spalling

Figure 64: More Advanced Spalling

Figure 65: Greatly Advanced Spalling

If the surrounding noise level is too great, a bearing’s condition can be evaluated by using a monitoring device, such as a stethoscope or a shock pulse analyzer.

The time between incipient and advanced spalling varies with speed and load, but in any event, it is not a sudden condition that will cause destructive failure within a matter of hours. Complete bearing failure and consequent damage to machine parts is usually avoided because of the noise the bearing will produce and the erratic performance of the shaft carried by the bearing.

Patterns of Load Paths and their Meaning in Bearing Damage

There are many ways bearings can be damaged before and during mounting and in service. The pattern or load zone produced on the internal surfaces of the bearing, by the action of the applied load and the rolling elements, is a clue to the cause of failure. To benefit from a study of load zones, one must be able to differentiate between normal and abnormal patterns. Figure 66 illustrates how an applied load of constant direction is distributed among the rolling elements of a bearing.

Figure 66: Load Distribution Within a Bearing

The large arrow indicates the applied load and the series of small arrows show the share of this load that is supported by each ball or roller in the bearing. The rotating ring will have a continuous 360 degree zone while the stationary ring will show a pattern of approximately 150 degrees. Figure 67 illustrates the load zone found inside a ball bearing when the inner ring rotates and the load has a constant direction.

Figure 67: Normal Load Zone - Inner

Figure 68 illustrates the load zone resulting if the outer ring rotates relative to a load of constant direction, or where the inner ring rotates and the load also rotates in phase with the shaft.

Figure 68: Normal Load Zone Outer Ring Rotating Relative to Load

Figure 69 illustrates the pattern we find in a deep groove ball bearing carrying an axial load and Figure 70 shows pattern from excessive axial load. This is the one condition where the load paths are the full 360 degrees of both rings. Combined thrust and radial load will produce a pattern somewhere between the two, as shown in Figure 71.

Figure 69: Normal Load Zone - Axial Load

Figure 70: Load Zone When Thrust Loads Are Excessive

Figure 71: Normal Load Zone Combined Thrust and Radial Load

With combined load, the loaded area of the inner ring is slightly off-center and the length in the outer is greater than that produced by radial load, but not necessarily 360 degrees. In a two-row bearing, a combined load will produce zones of unequal length in the two rows of rolling elements. If the thrust is of sufficient magnitude, one row of rolling elements can be completely unloaded.

When an interference fit is required, it must be sufficient to prevent the inner ring from slipping on the shaft but not so great as to remove the internal clearance of the bearing. If the fit is too tight, the bearing can be internally preloaded by squeezing the rolling elements between the two rings. In this case, the load zones observed in the bearing indicate that this is not a normal life failure as Figure 72 shows.

Figure 72: Load Zone From Internally Preloaded Bearing Supporting Radial Load

Both rings are loaded through 360 degrees, but the pattern will usually be wider in the stationary ring where the applied load is superimposed most on the internal preload.

Distorted or out-of-round housing bores can radially pinch an outer ring. Figure 73 illustrates the load zone found in a bearing where the housing bore was initially out-of-round or became out-of-round by bolting the housing to a concave or convex surface. In this case, the outer ring will show two or more load zones depending on the type of distortion. This is actually a form of internal preload.

Figure 73: Load Zones Produced by Out-of-Round Housing Pinching Housing Outer Ring

Figure 74 is a picture of a bearing mounted in an out-of-round housing that pinched the stationary outer ring. This is a mirror view and shows both sides of the outer ring raceway.

Figure 74: Mirror View of Outer Ring Distorted by Housing

Certain types of roller bearings can tolerate only very limited amounts of misalignment. When misaligned, a deep groove ball bearing produces load zones not parallel to the ball groove on one or both rings, depending on which ring is misaligned. Figure 75 shows the load zone when the outer ring is misaligned relative to the shaft. Figure 76 shows the patterns that appear when the inner ring is misaligned relative to the housing.

Figure 75: Load Zone Produced When Outer Ring is Misaligned Relative to Shaft

Figure 76: Load Zones When Inner Ring is Misaligned Relative to Housing

Cylindrical roller bearings and angular contact ball bearings are also sensitive to misalignments but it is more difficult to detect this condition from the load zones. With this background of failure patterns, the following failure descriptions should be meaningful.

Failure Due to Defective Bearing Seats on Shafts and in Housings

The calculated life expectancy of a roller bearing presupposes that its comparatively thin rings will be fitted on shafts or in housings that are as geometrically true as modern machine shop techniques can produce. Unfortunately, there are factors that produce shaft seats and housing bores that are oversized or undersized, tapered, or oval. Figure 77 shows another mirror view of a spherical roller bearing outer ring that has been seated in an out-of-round housing. Notice how the widest portions of the roller paths are directly opposite each other.

Figure 77: Mirror View of Spherical Roller Bearing Outer Ring Pinched by Housing

The same condition can be produced by seating the bearing in a housing with a correctly made bore, but where the housing is distorted when it is secured to the machine frame. An example is a pillow block bolted to a pedestal that is not flat.

Figure 78 shows the condition resulting from a not fully supported bearing outer ring. The impression made on the bearing OD by a turning chip left in the housing when the bearing was installed is seen in the left-hand view. This outer ring was subsequently supported by the chip alone with the result that the entire load was borne by a small portion of the roller path. The heavy specific load imposed on that part of the ring immediately over the turning chip produced the premature spalling seen in the figure.

Figure 78: Fatigue From Chip in Housing Bore

On both sides of the spalled area there is a condition called fragment denting, which occurred when fragments from the flaked surface are trapped between the rollers and raceway. When the contact between a bearing and its seat is not perfect, small movements produce a condition called fretting corrosion (Figure 79 and Figure 80). Fretting started the crack which, in turn, triggered the spalling.

Figure 79: Wear Due to Fretting Corrosion

Figure 80: Advanced Wear and Cracking Due to Fretting Corrosion

Fretting corrosion can also be found in applications such as railroad journal boxes, where machining of the seats is accurate but because of service conditions, the seats deform under load. This type of fretting corrosion on the outer ring does not, as a rule, detrimentally affect the life of the bearing. Shaft seats or journals as well as housing bores can yield and produce fretting corrosion. Figure 81 illustrates damage by movement on a shaft.

Figure 81: Fretting Caused by Yield in the Shaft Journal

The fretting corrosion covers a large portion of the surface of both the inner ring bore and the journal. The axial crack through the inner ring started from surface damage caused by the fretting. Bearing seats that are concave, convex, or tapered also causes bearing damage. On such a seat, a bearing ring cannot make contact throughout its width. The ring therefore deflects under the loads and fatigue cracks commonly appear axially along the raceway. Cracks caused by faulty contact between a ring and its housing is shown in Figure 82.

Figure 82: Cracks Caused by Faulty Housing Fit


Misalignment is a common source of premature spalling. It occurs when an inner ring is seated against a shaft shoulder that is not square with the journal, or where a housing shoulder is out-of-square with the housing bore. Misalignment arises when two housings are not on the same centerline. A bearing ring can be misaligned even though it is mounted on a tight fit, but is not pressed against its shoulder and is left cocked on its seat. Bearing outer rings in slip-fitted housings can be left cocked across their opposite corners.

Using self-aligning bearings does not cure some of the foregoing misalignment faults. When the inner ring of a self-aligning bearing is not square with its shaft seat, the inner ring is required to wobble as it rotates. This results in smearing and early fatigue. Where an outer ring is cocked in its housing across corners, a normally floating outer ring can become axially held as well as radially pinched in its housing. The effect of a pinched outer ring was shown earlier.

Ball thrust bearings suffer early fatigue when mounted on supports that are not perpendicular to the shaft axis, because one short load zone of the stationary ring carries the entire load. When the rotating ring of the ball thrust bearing is mounted on an out-of-square shaft shoulder, the ring wobbles as it rotates. The wobbling rotating ring loads only a small portion of the stationary ring and causes early failure.

Figure 83 shows smearing within a ball thrust bearing when either one of two conditions occurs. First, the two rings may not be parallel to each other during operation, and secondly, the load may not be sufficient at the operating speed to hold the bearing in its designed operational attitude.

Figure 83: Smearing in a Ball Thrust Bearing

If the condition arises from non-parallelism of the rings, the smearing seen in Figure 83 occurs when the balls pass from the loaded into the unloaded zone. Secondly, if the rings are parallel to each other but the speed is too high in relation to the load, centrifugal force causes each ball to spin instead of roll at its contact with the raceway; resulting in smearing. Smearing from misalignment will be localized in one zone of the stationary ring whereas smearing from gyral forces will be general around both rings.

Where two housings supporting the same shaft do not have a common centerline, only self-aligning ball or roller bearings will be able to function without inducing bending moments. Cylindrical and tapered roller bearings, although crowned, can accommodate only very small misalignments. Appreciable misalignment results in edge-loading, causing a source of premature fatigue. Edge-loading from misalignment was responsible for the spalling in the bearing ring shown in Figure 84.

Figure 84: Fatigue Caused by Edge-Loading

Advanced spalling due to the same cause can be seen on the inner ring and a roller of a tapered roller bearing in Figure 85.

Figure 85: Advanced Spalling Caused by Loading

Faulty Mounting Practice

Premature fatigue and other failures are often due to abuse and neglect before and during mounting. Prominent among causes of early fatigue is the presence of foreign matter in the bearing and its housing during operation. The effect of trapping a chip between the OD of the bearing and the bore of the housing was shown in Figure 78. Figure 86 shows the inner ring of a bearing where foreign matter has been trapped between the raceway and the rollers causing brinelled depressions. This condition is called fragment denting. Each of these small dents is the potential start of premature fatigue. Small particles of foreign matter cause wear and, when the original, internal geometry is changed, the calculated life expectancy cannot be achieved.

Figure 86: Fragment Denting

Impact damage during handling or mounting results in brinelled depressions this becomes the start of premature fatigue. An example of this is shown in Figure 87, where the spacing of flaked areas corresponds to the distance between the balls. The bearing has obviously suffered impact and, if installed, the fault should be apparent by noise of vibration during operation.

Figure 87: Fatigue Caused by Impact Damage During Handling or Mounting

Cylindrical roller bearings are easily damaged in mounting, especially when the rotating part with the inner ring mounted on it is assembled into a stationary part with its outer ring and roller set assembled. Figure 88 shows the inner ring of a cylindrical roller bearing that was damaged because the rollers had to slide forcibly across the inner ring during assembly.

Figure 88: Smearing Caused by Excessive Force in Mounting

Here again the spacing of the damage marks on the inner ring is the same as the distance between rollers. The smeared streak in Figure 88 is shown enlarged eight times in Figure 89.

Figure 89: Smearing Enlarged 8x

If a bearing is subjected to loads greater than those calculated to arrive at the life expectancy, premature fatigue results. Unanticipated or parasitic loads can arise from faulty mounting practice. An example of parasitic load can be found in the procedure of mounting the front wheel of an automobile. If the locknut is not backed off after the specific torque to seat the bearing is applied, parasitic load may result. Another example would be any application where a bearing should be free in its housing, but because of pinching or cocking, it cannot move with thermal expansion. Figure 90 shows the effect of a parasitic thrust load.

Figure 90: Spalling from Excessive Thrust

The damaged area is not in the center of the ball groove as it should be, but is high on the shoulder of the groove. The ring shown in Figure 91 is a self-aligning ball bearing subjected to an abnormally heavy thrust load. Usually in such cases, evidence of axial restraint will appear either as the imprint of a housing shoulder on the outer ring face, or as areas of fretting on the O.D. of the bearing.

Figure 91: Spalling from Parasitic Thrust

Interference between rotating and stationary parts can result in destructive cracks in the rotating bearing ring. The roller bearing inner ring in Figure 92 shows the effect of contact with an end cover while the bearing ring rotated.

Figure 92: Cracks Caused by Contact With End Cover While Bearing Ring Rotated

Damage Due to Improper Fits

To decide if a bearing ring, either inner or outer, should be mounted with an interference or slip fit on its shaft or its housing, it must be determined whether the ring rotates or is stationary, with reference to the direction of the load. The degree of tightness or looseness is governed by the magnitude of the load and the speed. If a bearing ring rotates relative to the load direction, an interference fit is required. If the ring is stationary with reference to the load, it is fitted with some clearance and is called a slip fit. The degree of fit is governed by the concept that heavier loads require greater interference. The presence of shock or continuous vibration calls for a heavier interference fit of the ring that rotates relative to the load. Lightly loaded rings, or rings with considerable load that operate at extremely slow speeds that rotate relative to the load, may use a lighter fit or, in some cases, a slip fit.

Consider two examples. In an automobile front wheel, the direction of the load is constant - the pavement is always exerting an upward force on the wheel. In this case, the outer rings or cups are rotating and are press-fitted into the wheel hub while the inner rings or cones are stationary and are slip fitted on the spindle.

On the other hand, the bearings of a conventional gear drive have their outer rings stationary relative to the load and are slip fitted but the inner rings rotate relative to the load and are mounted with an interference fit. There are some cases where it appears necessary to mount both inner and outer rings of a bearing with interference fits due to a combination of stationary and rotating load or loads of undetermined amounts.

Such cases are designed with bearings that can allow axial expansion at the rollers rather than at a slip-fitted ring. Such a mounting would consist of a cylindrical roller bearing at one end of the shaft and a self-contained bearing (deep groove ball or spherical roller bearing) at the other end.

Some examples of the effects of incorrect fitting follow. Figure 93 shows the bore surface of an inner ring that is damaged by relative movement between it and its shaft while rotating under a constant direction load. This relative movement, called creep, can result in the scoring shown in Figure 93.

Figure 93: Scoring of Inner Ring Caused by "Creep"

If lubricant can penetrate the loose fit, the appearance of the bore as well as the shaft seat will be a brilliant polish, as in Figure 94.

Figure 94: Wear Due to Creep

When a normally tight fitted inner ring does creep, the damage is not confined to the bore surface but can have its effect on the faces of the ring. Contact with shaft shoulders or spacers can result in either wear or severe rubbing cracks (Figure 92 depending on the lubrication condition). Figure 94 also shows how a shaft shoulder wore into the face of a bearing inner ring when relative movement occurred. Wear between a press-fitted ring and its seat is an accumulative damage. The initial wear accelerates the creep, which, in turn, produces more wear. The ring loses adequate support, develops cracks, and the products of wear become foreign matter to fragment, dent, and internally wear the bearing.

Excessive fits also result in bearing damage by internally preloading the bearing as shown in Figure 82 or inducing dangerously high hoop stresses in the inner ring. Figure 95 illustrates an inner ring that cracked because of excessive interference fit.

Figure 95: Axial Cracks Caused By an Excessive Interference Fit

Housing fits that are unnecessarily loose allow the outer ring to fret, creep, or even spin. Examples of fretting were seen in Figure 89 and Figure 90. The lack of support to the outer ring results from excessive looseness as well as from faulty housing bore contact. A cracked outer ring was shown in Figure 90.

Inadequate or Unsuitable Lubrication

All bearings need lubricants for reliable operation. The curvature of the contact areas between rolling element and raceway in normal operation results in minute amounts of sliding motion in addition to the rolling. Also, the cage must be carried on either the rolling elements, some surface of the bearing rings, or a combination of these. In most types of roller bearings, there are roller end faces that slide against a flange or a cage.

For these reasons, adequate lubrication is even more important at all times. The term "lubrication failure" is too often taken to imply that there was no oil or grease in the bearing. While this does happen occasionally, the failure analysis is normally not that simple. Many cases require a thorough examination of the lubricant’s properties including the amount of lubricant applied to the bearing and the operating conditions. If any one of these factors does not meet requirements, the bearing can be said to have failed from inadequate lubrication.

Viscosity of the oil, either as oil itself or as the oil in grease, is the primary characteristic of adequate lubrication. The nature of the soap base of a grease and its consistency, along with the viscosity of the oil, are the main quality points when considering a grease. For the bearing itself, the quantity of a lubricant required at any one time is usually rather small, but the supply must be such that a sufficient quantity is constantly available. If the lubricant is also a heat removal medium, then a larger quantity is required.

An insufficient quantity of lubricant at medium to high speeds generates a temperature rise and usually a whistling sound. An excessive amount of lubricant produces a sharp temperature rise due to churning in all but exceptionally slow speed bearings. Conditions inducing abnormally high temperatures can render a normally adequate lubricant inadequate.

When lubrication is inadequate, surface damage will result. This damage progresses rapidly to failures that are often difficult to differentiate from a primary fatigue failure. Spalling will occur and often destroy the evidence of inadequate lubrication. However, if caught soon enough, indications that pinpoint the real cause of the short bearing life can be found. One form of surface damage is shown in stages in Figure 96.

Figure 96: Progressive Stages of Spalling Caused by Inadequate Lubrication

The first visible indication of trouble is usually a fine roughening or waviness on the surface. Later, fine cracks develop, followed by spalling. If there is insufficient heat removal, the temperature may rise high enough to cause discoloration and softening of the hardened bearing steel. This happened to the bearing shown in Figure 97. In some cases, inadequate lubrication initially appears as a highly glazed or glossy surface which, as damage progresses, takes on a frosty appearance and eventually spalls. The highly glazed surface is shown on the roller of Figure 98.

Figure 97: Discoloration and Softening of Metal Caused by Inadequate Lubrication and Excessive Heat

Figure 98: Glazing Caused by Inadequate Lubrication

In the frosty stage, it is sometimes possible to feel the "nap" of fine slivers of metal pulled from the bearing raceway by the rolling element. The frosted area will feel smooth in one direction, but have a distinct roughness in the other. As metal is pulled from the surface, pits appear and frosting advances to pulling. An example of pulling is shown in Figure 99.

Figure 99: Effect of Rollers Pulling Metal from the Bearing Raceway

Another form of surface damage is called smearing. It appears when two surfaces slide and the lubricant cannot prevent adhesion of the surfaces. Minute pieces of one surface are torn away and rewelded to either surface. Examples are shown in Figure 100 through Figure 103.

Figure 100: Smearing on Spherical Roller End

Figure 101: Smearing on Cylindrical Rollers Caused by Ineffective Lubrication

Figure 102: Smearing on Cage Pockets Caused by Ineffective Lubrication

Figure 103: Smearing on Cylindrical Outer Raceway

A peculiar type of smearing occurs when rolling elements slide as they pass from the unloaded to the loaded zone. Figure 104 illustrates the patches of skid smearing, one in each row. A lubricant that is too stiff also causes this type of damage. This is particularly likely to happen if the bearing is large.

Figure 104: Skid Smearing on Spherical Outer Raceway

Wear of the bearing as a whole also results from inadequate lubrication. The areas subject to sliding friction such as locating flanges and the ends of rollers in a roller bearing are the first parts to be affected. Figure 105 and Figure 106 illustrate the damage done and the extent of the wear.

Figure 105: Rollers Welded to Rib Because of Ineffective Lubrication

Figure 106: Grooves Caused by Wear Due to Inadequate Lubrication

Where high speeds are involved, inertial forces become important and the best lubrication is demanded. Figure 107 shows an advanced case of damage from high speed with inadequate lubrication. Inertia forces acting on the rolling elements at high speed and with sudden starting or stopping can result in high forces between rolling elements and the cage.

Figure 107: Grooves Caused by Wear Due to Inadequate Lubrication

Figure 108 shows a large bore tapered roller bearing failure due to an insufficient amount of lubricant resulting from a low flow rate in a circulating oil system.

Figure 108: Broken Cage Caused by Ineffective Lubrication

The area between the guide flange and the large end of the roller is subject to sliding motion. This area is more difficult to lubricate than those areas of rolling motion, accounting for the discoloration starting at the flange contact area. The heat generated at the flange caused the discoloration of the bearing and resulted in some of the rollers being welded to the guide flange.

To avoid lubrication-related surface failures, be aware of the following:

  • Sufficient elastohydrodynamic film prevents surface distress (glazing, pitting).
  • Good boundary lubrication guards against smearing and sliding surface wear.
  • Clean lubricants prevent significant wear of rolling surfaces.
  • Sufficient lubricant flow keeps bearing from overheating.

As long as an elastohydrodynamic oil film can separate the surfaces of rolling element and raceway in rolling contact, surface distress is avoided. The continuous presence of the film depends on contact area, the load it carries, the speed, operating temperature, the surface finish, and the oil viscosity.

When the elastohydrodynamic oil film proves suitable in the rolling contacts, experience shows it is generally satisfactory for the sliding contacts at cages and guide flanges. When, in unusual applications, viscosity selection must be governed by the sliding areas, experience has proven that the viscosity chosen is capable of maintaining the necessary elastohydrodynamic film in the rolling contacts.

Ineffective Sealing

Although foreign matter can enter a bearing during mounting, its most direct and sustained area of entry can be the housing seals. The result of gross change in bearing internal geometry has been pointed out. Bearing manufacturers realize the damaging effect of dirt and take extreme precautions to deliver clean bearings. Not only assembled bearings, but also parts in process, are washed and cleaned. Freedom from abrasive matter is so important that some bearings are assembled in air-conditioned white rooms. Dramatic examples of combined abrasive particle and corrosive wear, both due to the defective sealing, are shown in Figure 109 and Figure 110.

Figure 109: Advanced Abrasive Wear

Figure 110: Advanced Abrasive Wear

Figure 111 shows a deep groove ball bearing, which has operated with an abrasive in it. The balls have worn to such an extent that they no longer support the cage and the latter has been rubbing on the lands of both rings.

Figure 111: Advanced Abrasive Wear

In addition to abrasive matter, corrosive agents should be excluded from bearings. Water, acid, and those agents that deteriorate lubricants result in corrosion. Figure 112 illustrates how moisture in the lubricant can rust the end of a roller. The corroded areas on the rollers of Figure 113 occurred while the bearing was not rotating. Acids forming in the lubricant with water present etches the surface as shown in Figure 114.

Figure 112: Rust on End of Roller Caused by Moisture in Lubricant

Figure 113: Corrosion on Roller Surfaces Caused by Water in Lubricant While Bearing Was Standing Still

Figure 114: Corrosion of Roller Surface Caused by Formation of Acids in Lubricant with Some Moisture Present

The lines of corrosion seen in Figure 115 are caused by water in the lubricant as the bearing rotates.


Rolling bearings exposed to vibration while the shafts are not rotating are subject to damage called false brinelling. The evidence can be either bright polished depressions or the characteristic red-brown stain of fretting. The oxidation rate at the point of contact determines the appearance. Variation in the vibrating load causes minute sliding in the area of contact between rolling elements and raceways. Small particles of material are set free from the contact surfaces and may or may not be immediately oxidized. The debris thus formed acts as a lapping agent and accelerates the wear.

Another identification of damage of this type is the spacing of the marks on the raceway. The spacing of false brinelling will be equal to the distance between the rolling elements, just as it is in some types of true brinelling. If the bearing has rotated slightly between periods of vibration, more than one pattern of false brinelling damage may be seen. A type of false brinelling with abrasive present is seen in Figure 116.

Figure 116: False Brinelling Caused by Vibration with Bearing Stationary

There was no rotation between the two rings of the bearing for considerable periods of time, but while they were static they were subject to severe vibration. False brinelling developed with a production of iron oxide, which in turn acted as a lapping compound.

A combination of vibration and abrasion in a rotating bearing is seen in the wavy pattern shown in Figure 117. When these waves are more closely spaced, the pattern is called fluting and appears similar to cases shown in the section titled Passage of Electric Current. Metallurgical examination is often necessary to distinguish between fluting caused solely by abrasion and vibration, or by vibration and passage of electric current.

Figure 117: False Brinelling Caused by Vibration in Presence of Abrasive Dirt While Bearing was Rotating

Since false brinelling is a true wear condition, such damage can be observed even though the forces applied during vibration are much smaller than those corresponding to the static carrying capacity of the bearing. However, the damage is more extensive as the contact load on the rolling elements increases.

False brinelling occurs most frequently during transportation of assembled machines. Vibration fed through a foundation can generate false brinelling of a shaft that is not rotating. False brinelling during transportation can always be minimized and usually eliminated by temporary structures that will prevent any rotation or axial movement of the shaft.

It is necessary to distinguish between false and true brinelling. Figure 118 and Figure 119 are 100X photomicrographs of true and false brinelling in a raceway, respectively.

In Figure 118 (true brinelling), there is a dent produced by plastic flow of the raceway material. The grinding marks are not noticeably disturbed and can be seen over the whole dented area. However, false brinelling (Figure 119) does not involve flow of metal but rather a removal of surface metal by attrition.

Figure 118: Example of True Brinelling (100x)

Figure 119: Example of False Brinelling (100x)

Notice that the grinding marks are removed. To further understand false brinelling, which is very similar to fretting corrosion, one should remember that a rolling element squeezes the lubricant out of its contact with the raceway, and the angular motion from vibration is so small that the lubricant is not replenished at the point of contact. Metal to metal contact becomes inevitable, resulting in sub-microscopic particles being torn from the high points. If protection by lubricant is absent, these minute particles oxidize and account for the red brown color usually associated with fretting. If there is a slower oxidation rate, the false brinelling depression can remain bright, thereby adding to the difficulty in distinguishing true from false brinelling.

Passage of Electric Current through the Bearing

In certain electrical machinery applications, there is the possibility that electric current will pass through a bearing. Current that seeks ground through the bearing can be generated from stray magnetic fields in the machinery. Welding on some part of the machine with the ground attached so that the circuit is required to pass through the bearing can also cause current to pass through a bearing.

An electric current can be generated by static electricity, emanating from charged belts or from manufacturing processes involving leather, paper, cloth, or rubber. This current can pass through the shaft to the bearing and then to ground. When the current is broken at the contact surfaces between rolling elements and raceways, arcing results. This produces very localized high temperature and consequent damage. The overall damage to the bearing is in proportion to the number and size of individual damage points. Figure 120 and the enlarged view (Figure 121) show a series of electrical pits in a roller and in a raceway of a spherical roller bearing.

Figure 120: Electric Pitting on Surface of Spherical Roller Caused by Passage of Relatively Large Current

Figure 121: Electric Pitting on Surface of Spherical Outer Raceway Caused by Passage of Relatively Large Current

The pit was formed each time the current broke in its passage between the raceway and roller. The bearing from which this roller was removed was not damaged generally to the same degree shown on this roller. In fact, this specific bearing was returned to service and operated successfully for several additional years. Thus, moderate amounts of electrical pitting do not necessarily result in failure.

Another type of electrical damage occurs when current passes during prolonged periods and the number of individual pits accumulate drastically. The result is fluting, shown in Figure 122 through Figure 126.

Figure 122: Fluting on Surface of Spherical Roller Caused by Prolonged Passage of Electric Current

Figure 123: Fluting on Inner Raceway of Cylindrical Roller Bearing Caused by Prolonged Passage of Electric Current

Figure 124: Fluting on Outer Raceway of Spherical Raceway of Spherical Roller Bearing

Figure 125: Fluting on Inner Raceway of Spherical Raceway

Figure 126: Fluting on Outer Raceway of Self-aligning Ball Bearing

This condition can occur in ball or roller bearings. Flutes can develop considerable depth, producing noise and vibration during operation and eventual fatigue from local overstressing. The formation of flutes rather than a homogeneous dispersion of pits cannot be clearly explained. It is possible that it is related to initial synchronization of shocks or vibrations and the breaking of the current. Once the fluting has started, it is probably a self-perpetuating phenomenon.

Individual electric marks, pits, and fluting have been produced in test bearings. Both alternating and direct current can cause the damage. Amperage rather than voltage governs the amount of damage. When a bearing is under radial load, greater internal looseness in the bearing appears to result in greater electrical damage for the same current. In a double-row bearing loaded in thrust, little, if any, damage results in the thrust-carrying row, although the opposite row may be damaged.