ASM Metals HandBook Vol. 8 - Mechanical Testing and Evaluation

Подождите немного. Документ загружается.

42. R.E. Petersen, Stress Concentration Factors, John Wiley & Sons, New York, 1973

43. R.E. Petersen, Trans. ASME (Appl. Mech. Sect.), Vol 58, 1936, p A-149

44. R. Stickler and B. Weiss, in Ultrasonic Fatigue, J.M. Wells, O. Buck, L.D. Roth, and J.K. Tien, Ed.,

TMS-AIME, Warrendale, PA, 1982, p 135

45. S. Purushothoman, J.P. Wallace, and J.K. Tien, Ultrasonics International 1973 Conference

Proceedings, IPC Science and Technology Publications, Guildford, Surrey, U.K., 1973, p 244

46. W. Hessler, H. Mullner, and B. Weiss, Met. Sci., May 1981, p 225

47. B. Weiss, R. Stickler, J. Fembock, and K. Pffafinger, Fatigue Eng. Mater. Struct., Vol 2, 1979, p 73

48. W. Hoffelner and P. Gudmundson, Eng. Fract. Mech., Vol 16, 1982, p 365

49. W. Hoffelner, J. Phys. E, Sci. Instrum., Vol 13, 1980, p 617

50. N. Dowling, Cyclic Stress-Strain and Plastic Deformation Aspects of Fatigue Growth, STP 637, ASTM,

Philadelphia, 1976, p 97

51. B. Weiss, Determination of The Threshold Stress Intensity Value of Mo and Mo-Alloys using a 20 kHz

method (German), Metall, Vol 34, 1980, p 636

52. S. Purushothoman and J.K. Tien, Metall. Trans. A, Vol 9, 1975, p 367

53. J. Babouk, K. Kromp, W. Kromp, and P. Bajons, in Ultrasonic Fatigue, J.M. Wells, O. Buck, L.D.

Roth, and J.K. Tien, Ed., TMS-AIME, Warrendale, PA, 1982, p 51

54. C. Laird and P. Charlsey, in Ultrasonic Fatigue, J.M. Wells, O. Buck, L.D. Roth, and J.K. Tien, Ed.,

TMS-AIME, Warrendale, PA, 1982, p 183

55. J.K. Tien, in Ultrasonic Fatigue, J.M. Wells, O. Buck, L.D. Roth, and J.K. Tien, Ed., TMS-AIME,

Warrendale, PA, 1982, p 1

56. L.E. Coffin, in Ultrasonic Fatigue, J.M. Wells, O. Buck, L.D. Roth, and J.K. Tien, Ed., TMS-AIME,

Warrendale, PA, 1982, p 423

Fretting Fatigue Testing

S.J. Shaffer and W.A. Glaeser, Battelle Memorial Institute

Introduction

FRETTING is a special wear process that occurs at the contact area between two materials under load and

subject to slight relative movement by vibration or some other force. Damage begins with local adhesion

between mating surfaces and progresses when adhered particles are removed from a surface. When adhered

particles are removed from the surface, they may react with air or other corrosive environments. Affected

surfaces show pits or grooves with surrounding corrosion products. On ferrous metals, corrosion product is

usually a very fine, reddish iron oxide; on aluminum, it is usually black. The debris from fretting of noble

metals does not oxidize.

Under fretting conditions, fatigue strength or endurance limits can be reduced by as much as 50 to 70% during

fatigue testing (e.g., see Fig. 1a). During fretting fatigue, cracks can initiate at very low stresses, well below the

fatigue limit of nonfretted specimens. In fatigue without fretting, the initiation of small cracks can represent

90% of the total component life. The wear mode known as fretting can cause surface microcrack initiation

within the first several thousand cycles, significantly reducing the component life. Additionally, cracks due to

fretting are usually hidden by the contacting components and are not easily detected. If conditions are favorable

for continued propagation of cracks initiated by fretting, catastrophic failure can occur (Fig. 1b). As such,

prevention of fretting fatigue is essential in the design process by eliminating or reducing slip between mated

surfaces.

Fig. 1 Effects of fretting. (a) Comparison of fatigue life for 4130 steel under fretting and nonfretting

conditions. Specimens were water quenched from 900 °C (1650 °F), tempered 1 h at 450 °C (840 °F), and

tested in tension-tension fatigue. Normal stress was 48.3 MPa (7 ksi); slip amplitude was 30–40 μm. (b)

Example of catastrophic fatigue due to fretting of a flanged joint

The initiation of fatigue cracks in fretted regions depends mainly on the state of stress in the surface,

particularly stresses caused by high friction. The direction of growth of the fatigue cracks is associated with the

direction of contact stresses and takes place in a direction perpendicular to the maximum principal stress in the

fretting area. After formation due to fretting, cracks propagate initially under shear (mode II) conditions under

the influence of the near-surface shear-stress field due to friction of fretting. Beyond that, tensile (mode I) crack

propagation under bulk cyclic stresses controls further propagation.

The topics covered in this article are:

• Mechanisms of fretting and fretting fatigue

• Typical occurrences of fretting fatigue

• Fretting fatigue testing

• Prevention methods

Many investigators have contributed to the theoretical and practical research in the field of fretting and fretting

fatigue, and the information in this section is derived from their work. Several general texts are available (Ref

1, 2, 3, 4). Reference 5 is another key source for information and illustrations of fretting fatigue failures. In

addition, more current discussions and background are provided in the book Fretting Fatigue edited by R.B.

Waterhouse and T.C. Lindley. The article “On Fretting Maps” by O. Vingsbo and S. Söderberg is another

useful general reference on fretting. (See the Selected References for complete bibliographic information.)

As yet, general techniques or models permitting prediction of crack initiation due to fretting are limited.

However, an understanding of the factors contributing to fretting fatigue can help minimize the risk and extent

of damage. The examples presented in this article from case studies, theoretical work, and laboratory

investigations are intended to assist the reader in recognizing the potential for fretting fatigue in design and

materials selection. General principles and practical methods for the abatement or elimination of fretting fatigue

are summarized in Table 1. More recent information on fretting fatigue testing can be found in ASTM STP

1367 (listed under “Selected References” at the end of this article).

Table 1 Reduction or elimination of fretting fatigue

Principle of abatement ormitigation

Practical method

Reduction in surface shear forces

• Reduction in surface normal forces

• Reduction in coefficient of friction with coating or

lubricants

Reduction/elimination of stress

concentrations

• Large radii

• Material removal (grooving)

• Compliant spacers

Introduction of surface compressive stress

• Shot or bead blasting

• Interference fit

• Nitriding/heat treatment

Elimination of relative motion

• Increase in surface normal load

• Increase in coefficient of friction

Separation of surfaces

• Rigid spacers

• Coatings

• Compliant spacers

Elimination of fretting condition

• Drive oscillatory bearing

• Remove material from fretting contact (pin joints)

• Separation of surfaces (compliant spacers)

Improved wear resistance

• Surface hardening

• Ion implantation

• Soft coatings

• Slippery coatings

Reduction of corrosion • Anaerobic sealants

• Soft or anodic coatings

References cited in this section

1. R.B. Waterhouse, Ed., Fretting Fatigue, Applied Science, 1981

2. M.H. Attia and R.B. Waterhouse, Ed., Standardization of Fretting Fatigue Test Methods and

Equipment, STP 1159, ASTM, 1992

3. D.A. Hills and D. Nowell, Mechanics of Fretting Fatigue, Kluwer Academic Publishers, 1994

4. R.B. Waterhouse, Fretting Corrosion, Fretting Fatigue, Pergamon Press, 1972

5. Proc. Specialists Meeting on Fretting in Aircraft Systems, AGARD-CP-161, Advisory Group for

Aerospace Research and Development, 1974

Fretting Fatigue Testing

S.J. Shaffer and W.A. Glaeser, Battelle Memorial Institute

Fretting and Fretting Fatigue Mechanisms

In general, fretting occurs between two tight-fitting surfaces that are subjected to a cyclic, relative motion of

extremely small amplitude. Although certain aspects of the mechanism of fretting are still not thoroughly

understood, the fretting process is generally divided into the following three parts: initial conditions of surface

adhesion, oscillation accompanied by the generation of debris, and fatigue and wear in the region of contact.

Fretting wear occurs from repeated shear stresses that are generated by friction during small-amplitude

oscillatory motion or sliding between two surfaces pressed together in intimate contact. Surface cracks initiate

in the fretting wear region. The relative slip amplitude is typically less than 50 μm (0.002 in.), and

displacements as small as 10

-4

μm have produced fretting. Generation of fine wear debris that usually oxidizes

is an indication of fretting wear (Fig. 2). The following factors are known to influence the severity of fretting:

• Contact load. As long as fretting amplitude is not reduced, fretting wear will increase linearly with

increasing load.

• Amplitude. There appears to be no measurable amplitude below which fretting does not occur. However

if the contact conditions are such that deflection is only elastic, it is not likely that fretting damage will

occur. Fretting wear loss increases with amplitude. The effect of amplitude can be linear, or there can be

a threshold amplitude above which a rapid increase in wear occurs (Ref 4). The transition is not well

established and probably depends on the geometry of the contact.

• Frequency. When fretting is measured in volume of material removed per unit sliding distance, there

does not appear to be a frequency effect.

• Number of cycles. An incubation period occurs during which fretting wear is negligible. After the

incubation period, a steady-state wear rate is observed, and a more general surface roughening occurs as

fretting continues.

• Relative humidity. For materials that rust in air, fretting wear is higher in dry air than in saturated air.

• Temperature. The effect of elevated temperature on fretting depends on the oxidation characteristics of

the material.

Fig. 2 Fretting wear scars. (a) On steel (arrows indicate fatigue crack). Courtesy of R.B. Waterhouse,

University of Nottingham. (b) On high-purity nickel. Courtesy of R.C. Bill, NASA Lewis Research

Center

In terms of fatigue, the following three primary variables contribute to shear stresses at the surface and, hence,

are important for crack initiation and initial propagation of fretting fatigue cracks:

• Normal load (e.g., contact pressure)

• Relative displacement (slip amplitude)

• Coefficient of friction

The other primary variable is the bulk tensile stresses that control crack propagation beyond the limit of the

surface-induced stress field. Secondary factors, including surface roughness, surface contaminants, contact size,

debris accumulation, and environment affect fretting fatigue through their influence on the primary variables.

Effective lubrication will reduce friction stresses and wear-particle accumulation.

Fretting Modes and Contact Conditions. The oscillatory motion responsible for fretting can be induced by

system vibrations or by cyclic loading of one of the components. The relative displacement can be either

amplitude controlled or load controlled, or a combination of both. Methods to control fretting fatigue depend on

which of these two modes dominates the contact conditions.

Stress Conditions. The nominal macroscopic normal stress between the two surfaces is defined by the normal

force divided by the nominal area of contact. Subsurface stress distributions can be computed using Hertzian

calculations and the macroscopic contact geometry. The normal stress is also influenced by geometric stress

concentrations. The real area of contact is limited to the contacting tips of the microscopic asperities on each of

the surfaces, and the local (microscopic) normal stress is dictated by the yield strength of the softer of the two

materials. Superimposed on the local normal stresses are shear stresses resulting from the friction of relative

displacement of the two contacting members. The magnitude of the shear stresses induced by asperity contact

depends on the coefficient of friction (due to adhesive forces between, and interpenetration of, asperities), the

local load, asperity geometry, the elastic moduli of the two surfaces, and the amplitude of relative displacement.

Strain Conditions. If the amplitude of oscillation is small, the shear strains are elastic, and the contact condition

is one of sticking or no slip. Even under microelastic displacements, fatigue cracks can form due to reverse

bending at the bases of the contacting asperities. If the amplitude of oscillation is large, depending on the

strength and ductility of the asperities in contact, and on the adhesive forces acting between them, the asperities

will be forced to pass over one another and slip occurs.

With slip, the possibility of wear exists from either

adhesion, abrasion, or delamination. All of these material-removal mechanisms lead to a roughening of the

surface, the creation of sites for crack initiation, and the generation of wear debris. Cracks can also be initiated

by pitting that occurs during fretting.

Conditions for Slip. The local contact conditions may be predominantly displacement controlled or load

controlled. In displacement-controlled fretting contacts undergoing full slip, the amplitude of motion is

controlled by the external displacements, an example being the relative displacement imposed on adjoining

strands of a wire rope passing over a pulley. For force-controlled fretting contacts, the displacement depends on

the macroscopic shear force, the normal force, and the coefficient of friction; for example, the mating forces of

a bolted flange or a hub/shaft press fit interface. No slip occurs until the shear stress exceeds the product of the

normal force and local coefficient of friction. The condition for slip is met when

τ > μσ

(Eq 1)

where τ is the local shear stress, μ is the coefficient of friction between contacting surfaces, and σ

is the local

normal stress.

Regions of both slip and no-slip can occur at an interface of contacting solids. This is most easily seen for

convex contacts using the elastic stress analysis of a sphere pressed into a plate, as shown in Fig. 3 (Ref 6). The

normal (Hertzian) stress field is an elliptical distribution with the maximum stress occurring under the center of

the contact. The shear stresses, which promote relative slip, are a maximum at the edges of the contact (limited

by yielding) and a minimum in the center. The forces resisting sliding due to the shear stress are given by μN.

The condition pictured is known as partial slip. As μN is increased, the region of sticking expands, and vice

versa. When the shear forces of fretting are superimposed on this stress field, the result is a smaller “stick” area.

This analysis can also apply to a cylindrical contact or can be adapted to microscopic asperity contacts.

Fig. 3 Stress distribution for hemispherical contact pressed into flat plate. Source: Ref 6

For contact between nominally flat surfaces, the stress state is different, though the slip condition is still defined

by Eq 1. The macroscopic stress concentrations for well-defined geometries can be computed using finite-

element modeling (FEM) analysis, although wear will change the assumed contact geometry. A general

treatment of the subject can be found in Ref 3.

Fatigue-Crack Nucleation from Fretting. Crack nucleation due to fretting must involve a stress concentration or

discontinuity. At the microscopic level, examples include: microcracks formed at the base of asperities due to

reverse-bending fatigue of the asperities, stress concentrations in the pits left by sheared adhering “cold-

welded” asperity junctions, corrosion pits that form due to removal of protective oxides by fretting, grooves due

to abrasion, or delamination of a thin surface region whose work-hardening capacity has been exhausted. At the

macroscopic level, cracks are proposed to form solely as a result of geometric stress concentrations, usually at

the edges of the fretting contact region, where shear stresses are predicted to be highest, at some microscopic

inhomogeneity. The two views are not significantly different. The second view is more amenable to modeling

and FEM analysis.

The location of crack nucleation depends on the contact conditions. Under full-slip conditions, in the absence of

stress concentrations at the edge of the contact, cracks can nucleate anywhere in the contact region. The number

of asperity interactions per cycle depends on the asperity distribution (surface roughness) and the amplitude of

relative motion. Several cracks may be formed. Their stress fields can interact and lead to a decrease in the

stress field associated with a single crack. This may explain why multiple nonpropagating cracks are often

found in association with fretting. Under partial-slip conditions, the cracks always form at the border between

the slip and the no-slip regions. In this case, multiple cracks are proposed to result from the movement of the

slip/no-slip boundary due to the generation of debris (Ref 7). Though less likely, cracks can also form in the

region of no slip (full sticking) due to reciprocating subsurface shear stresses associated with reversing elastic

deformation of the contacting asperities and stress concentrations at their bases leading to local microplastic

deformation and fatigue.

Fatigue-Crack Propagation during Fretting Fatigue. Crack propagation is initially driven by the stress state

dominated by the surface shearing. As such, when viewed in cross section, the crack direction initially appears

at an angle to the surface of between 35 and 55°. Mode II crack propagation dominates this region. The mode II

propagation may depend on material parameters such as grain size, texture, and phase morphology. Because the

surface shear stresses fall off rapidly with depth, the crack will either arrest, or, if static or alternating tensile

stresses exist in the bulk material, will change direction and run perpendicular to the surface as the driving

forces come under the control of the bulk tensile forces (Fig. 4). The depth at which this occurs depends on the

magnitude of the surface shear stresses, which depend on the coefficient of friction and normal contact stress.

For Hertzian stresses of convex contacts, this depth is on the order of the half-width of the contact area. Beyond

this depth, mode I crack propagation analysis can be used to predict the growth rate under the bulk stress state.

Fig. 4 Example of fretting fatigue crack viewed in cross section. Courtesy of R.B. Waterhouse,

University of Nottingham

A phenomenon peculiar to fretting is that some of the fatigue cracks do not propagate because the effect of

contact stress extends only to a very shallow depth below the fretted surface. At this point, favorable

compressive residual stresses retard or completely halt crack propagation. Under full-slip conditions, the wear

rate caused by fretting occasionally outpaces the growth rate of surface-initiated fatigue cracks. In this situation,

fretting wear preempts fretting fatigue.

Footnote

In fretting, the term slip is used to denote small-amplitude surface displacements. In contrast to sliding,

which

denotes macroscopic displacements. Additionally, in this article slip

does not refer to the mechanism of

fatigue resulting as a consequence of dislocation motion.

References cited in this section

3. D.A. Hills and D. Nowell, Mechanics of Fretting Fatigue, Kluwer Academic Publishers, 1994

4. R.B. Waterhouse, Fretting Corrosion, Fretting Fatigue, Pergamon Press, 1972

6. R.D. Mindlin, Compliance of Elastic Bodies in Contact, J. Appl. Mech., Vol 16, 1949, p 259–268

7. R.B. Waterhouse, Theories of Fretting Processes, Fretting Fatigue, Applied Science, 1981, p 203–220

Fretting Fatigue Testing

S.J. Shaffer and W.A. Glaeser, Battelle Memorial Institute

Typical Systems and Specific Remedies

Fretting generally occurs at contacting surfaces that are intended to be fixed in relation to each other but that

actually undergo minute alternating relative motion that is usually produced by vibration. Fretting generally

does not occur on contacting surfaces in continuous motion, such as ball or sleeve bearings. There are

exceptions, however, such as contact between balls and raceways in bearings and between mating surfaces in

oscillating bearings and flexible couplings. Common sites for fretting are in joints that are bolted, keyed,

pinned, press fitted, or riveted; in oscillating bearings, splines, couplings, clutches, spindles, and seals; in press

fits on shafts; and in universal joints, baseplates, shackles, and orthopedic implants.

Three general geometries and loading conditions for fretting fatigue are considered in this section:

• Parallel surfaces clamped together with some type of fastener, such as a bolted flange or riveted lap joint

• Parallel surfaces loaded by means of a press or interference fit, such as a gear or wheel on a shaft

• Convex contacts, as found beneath a convex washer, between crossed cylinders such as wire rope

strands, or a sphere or a cylinder in a bearing race

Specific remedies to reduce fretting are given for these common examples. When frettting occurs, it often

cannot be eliminated but can be reduced in severity.

Parallel Contact with External Loading (Fastened Joints). Bolted flanges in pipe systems are common locations

for fretting fatigue. Cracks can occur in the plate either under a bolt head or washer (load controlled), on the

inside diameter of the bolt through-hole (displacement controlled), or on the surface of one plate at the point of

contact with the end of the other plate (load controlled) (Fig. 5a). Lap joints are found in both heavy plates and

thin sheets such as aircraft skins. Fretting can occur in the joint or under the head of a countersunk screw, bolt,

or rivet (Fig. 5b).

Fig. 5 Typical location of fretting fatigue cracks in (a) a bolted flange, and (b) a lap joint

For both these geometries, the reduction or abatement of fretting severity depends on whether the motion is

load controlled or displacement controlled. If it is displacement controlled, then reducing the contact stress and

minimizing the coefficient of friction at the interface is recommended. For lap joints, however, a reduction in

the coefficient of friction may result in insufficient load transmitted by the interface, transferring the load to the

fasteners and leading to their failure. For load-controlled motions, it may be possible to increase either the

clamping force or the coefficient of friction to completely eliminate relative motion between the two contacting

members. While adhesives can be used to eliminate the relative motion, their use complicates future

disassembly. If motion cannot be completely eliminated, then minimizing the coefficient of friction may help,

although this will likely lead to an increased slip amplitude. Alternatively, a thin compliant layer, such as

rubber or other polymer, may be able to absorb the deflection and prevent contact between the two members.

For pin joints, fretting can occur on diagonally opposite sides of the pin at the points of contact with the hole

due to vibrations or reversing loads. In these cases, White (Ref 8) showed that an increase in fatigue strength

can be achieved by machining flats on the sides of the pins to prevent contact at the position of maximum

stress, thus removing the region where fretting occurs (Fig. 6).

Fig. 6 Pin joint. (a) Fretting locations. (b) Material removal to eliminate region of highest stress

Parallel Surfaces without External Loading. Hubs, flywheels, gears, and other types of press-fit wheels, pulleys,

or disks on shafts are subject to fretting fatigue caused by reverse bending strains compounded by the stress

concentration where the shaft meets the disk (Fig. 7a). The introduction of lubricant in the interface can make

matters worse by increasing the relative slip. In this case, it is best to attempt a strong interference fit. This can

be achieved through cooling the shaft and heating the bore of the hole during assembly in order to produce

sufficiently high normal stresses to completely eliminate slip within the interface. After assembly, both surfaces

will also be in a state of compressive stress, providing further resistance to fatigue-crack propagation. Finally, if

possible, a stress-relieving groove or large radius on the shaft (Fig. 7b) should be incorporated into the design.

Fig. 7 Wheel on shaft. (a) Location of fretting fatigue cracks. (b) Stress-reduction grooves

Gas-turbine rotor-blade roots and other dovetail joints are potential locations of fretting fatigue failures (Fig.

8a). In this case, the loading conditions are variable and depend on the rotational speed. For these situations,

stress-relieving grooves can be incorporated into the design (Fig. 8b). Coatings to reduce the coefficient of

friction (and hence the surface shear forces) can also help. Experiments by Ruiz and Chen on simulated

blade/disk dovetail joints at 600 °C (1112 °F) indicated that shot peening followed by electroplating with a 10

μm (394 μin.) thick Co/C surface layer was effective (Ref 9). Another example is provided in the article by

Johnson in Ref 5.

Fig. 8 Dovetail joint. (a) Location of fretting fatigue cracks. (b) Stress-reduction grooves

Convex Surfaces. Fretting fatigue in control cables or wire ropes is caused by small relative displacements

between the individual strands as the cable flexes in passing over pulleys, or by varying stresses from wind or

water currents (Fig. 9). Stainless steel control cables are particularly susceptible because of the high friction and

galling propensity between the strands. Fatigue fractures of the inner strands of the cables make detection by

visual inspection virtually impossible until the ends of the fractured strands pop out through the outer strands.

Wire rope fretting fatigue in control cables is an example of displacement-controlled contact. As such, large

pulley diameters as a function of cable cross section can be specified in the design in order to decrease the

displacement and minimize fretting fatigue. Incorporation of lubricant in the cable will reduce strand-to-strand

shear forces, but only as long as the lubricant is contained within the rope interior by the outer strands.

Takeuchi and Waterhouse report that electrodeposited zinc coatings helped prevent fretting fatigue of wire rope

in sea water (Ref 10). The zinc provides both a reduction in the effects of corrosion and a low-shear-strength

surface film that reduces friction.

Fig. 9 Examples of fretting on inner strands of drag-line wire rope

In rolling-element bearings, high Hertzian normal stresses occur beneath the contact of bearing balls in their

races. Control-system or oscillatory-pivot bearings are often subjected to a low-amplitude, but high-frequency,

dithering motion leading to fretting or false brinelling between the balls and the race. (The difference between

false brinelling and fretting is discussed in the article “Fretting Wear” in Friction, Lubrication, and Wear