Astm stp 1159 1992

Foreword In 1988, the ASTM Committee E-9 on Fatigue approved the formation of a Task Group on Fretting Fatigue Testing to develop standards for the fretting fatigue test methods and equi

Trang 2

STP 1159

Standardization of Fretting

Fatigue Test Methods and

Equipment

M Helmi Attia and R B Waterhouse, editors

ASTM Publication Code Number (PCN)

04-011590-30

As M

1916 Race Street Philadelphia, PA 19103

Trang 3

Library of Congress Cataloging-in-Publication Data

Standardization of fretting fatigue test methods and equipment / M

Helmi Attia and R B Waterhouse, editors

(STP ; 1159)

Proceedings from a symposium held in San Antonio, Tex., Nov 12-13, 1990

"ASTM publication code number (PCN) 04-011590-30."

Includes bibliographical references and index

ISBN 0-8031-1448-6

1 Materials Fatigue Testing Standards Congresses 2 Fatigue

testing machines Standards Congresses I Attia, M Helmi

(Mahmoud Helmi) II Waterhouse, R B (Robert Barry), 1922-

III Title: Fretting fatigue test methods and equipment

IV Series: ASTM special technical publication ; 1159

TA418.38.$68 1992

CIP Copyright | 1992 AMERICAN SOCIETY FOR TESTING AND MATERIALS, Philadelphia, PA All rights reserved This material may not be reproduced or copied, in whole or in part, in any printed, mechanical, electronic, film, or other distribution and storage media, without the written consent of the publisher

Photocopy Rights

Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by the AMERICAN SOCIETY FOR TESTING AND MATERIALS for users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided that the base fee i

of $2.50 per copy, plus $0.50 per page is paid directly to CCC, 27 Congress St., Salem, MA 01970; (508) 744-3350 For those organizations that have been granted a photocopy license by CCC, a separate system of payment has been arranged The fee code for users of the Transactional Reporting Service is 0-8031-1448-6/92 $2.50 + 50

Peer Review Policy

Each paper published in this volume was evaluated by three peer reviewers The authors addressed all

of the reviewers' comments to the satisfaction of both the technical editor(s) and the ASTM Committee

on Publications

The quality of the papers in this publication reflects not only the obvious efforts of the authors and the technical editor(s), but also the work of these peer reviewers The ASTM Committee on Publications acknowledges with appreciation their dedication and contribution to time and effort on behalf of ASTM

Printed in Baltimore, MD July 1992

Downloaded/printed by

Trang 4

Foreword

In 1988, the ASTM Committee E-9 on Fatigue approved the formation of a Task Group on

Fretting Fatigue Testing to develop standards for the fretting fatigue test methods and equip-

ment This task group, chaired by one of the editors of this special publication (M H Attia)

has recognized the gravity of its responsibility and realized the need for an international coop-

erative effort to achieve its objective As a first step towards this goal, the idea of organizing a

symposium on this subject matter was born

This publication, Standardization of Fretting Fatigue Methods and Equipment, contains

papers presented at the Symposium of the same name in San Antonio, TX on 12-13 Novem-

ber 1990 The symposium was sponsored by ASTM Committee E-9 on Fatigue Dr M Helmi

Attia, of Ontario Hydro Research Division, Toronto, Ontario, Canada and Dr R B Water-

house, of the University of Nottingham, Nottingham, UK, presided as symposium chairmen

and are the editors of the resulting publication

The Cover

The photoelastic picture on the cover depicts the change in the stress field and the contact

pressure distribution at the fatigue specimen/fretting pad interface as a result of the change in

the height of the pad The latter is usually chosen arbitrarily and as such, the variability in the

test results is not unexpected It is hoped that the picture will capture the attention of those

involved with fretting fatigue testing to the necessity of standardizing the test specimens con-

figuration, methods, and equipment

The picture was obtained from the Fretting Laboratory, Mechanical Research Department,

Ontario Hydro Research Division

Trang 5

Contents

Overview M H ATTIA AND R B WATERHOUSE

A Historical Introduction to Fretting Fatigue R B WATERHOUSE

OPENING PAPER

FUNDAMENTAL ASPECTS OF FRETTING FATIGUE TESTING CONCEPTUAL FRAMEWORK

Mechanisms of Fretting Fatigue and Their Impact on Test Methods

Testing Methods in Fretting Fatigue: A Critical Appraisal L VINCENT,

Fretting and Contact Fatigue Studied with the Aid of Fretting M a p s - -

Variables of Fretting Process: Are There 50 of T h e m ? - - J DOBROMIRSKI 60

FUNDAMENTAL ASPECTS OF FRETTING FATIGUE TESTING MECHANICS OF CONTACT

The Development of a Fretting Fatigue Experiment with Well-Defined

Determination and Control of Contact Pressure Distribution in Fretting F a t i g u e - -

Fretting Fatigue Analysis of Strength Improvement Models with Grooving or

Knurling on a Contact Surface T HATTORI, M NAKAMURA,

Effect of Contact Pressure on Fretting Fatigue of High Strength Steel and

Titanium Alloy K NAKAZAWA, M SUMITA, AND N MARUYAMA l l 5

C o p y r i g h t b y A S T M I n t ' l ( a l l r i g h t s r e s e r v e d ) ; W e d D e c 2 3 1 9 : 0 1 : 5 5 E S T 2 0 1 5

D o w n l o a d e d / p r i n t e d b y

Trang 6

A Critical Review of Fretting Fatigue Investigations at the Royal Aerospace

Establishment D B RAYAPROLU AND R COOK

Fretting Fatigue in the Power Generation Industry: Experiments, Analysis, and

Integrity Assessment T C LINDLEY AND K J NIX

Techniques for the Characterization of Fretting Fatigue Damage c RUIZ,

Z P WANG, AND P H WEBB

The Influence of Fretting Corrosion on Fatigue Strength of Nodular Cast Iron and

Steel under Constant Amplitude and Load Spectrum Tests G FISCHER,

V GRUBISIC, AND O BUXBAUM

Adaptation of a Servohydraulic Testing Machine to Investigate the Life of

Machine Components Operating under Fretting Conditions J LABEDZ

Steam Turbine Steel Y MUTOH, T SATOH, AND E TSUNODA

The Fretting Fatigue Properties of a Blade Steel in Air and Vapor Environments

D YUNSHU, Z BAOYU, AND L WEILI

The Application of Electrochemical Techniques to Evaluate the Role of Corrosion

in Fretting Fatigue of a High Strength Low Alloy Steel s PRICE AND

L CLOUTIER, M ST-LOUIS, AND A LEBLOND

Fretting Fatigue of Carbon Fiber-Reinforced Epoxy Laminates o JACOBS,

K SCHULTE~ AND K FRIEDRICH

231

243

CLOSING PAPER Fretting Fatigue Testing: Current Practice and Future Prospects for

Trang 7

of test data do, however, exist with proper and comprehensive understanding of the sources of uncertainties

Objectives

The main objectives of this symposium/publication are as follows:

1 Review the present state of knowledge and the current fretting fatigue testing practice

2 Identify the areas of uncertainties in conducting fretting fatigue testing, including the design of the test specimens, as well as the measurement and control aspects

3 Identify the measures that should be taken to improve the repeatability of test results and

to minimize their dependence on the design of the test equipment

4 Examine the future prospects for standardization, and identify the areas that warrant further research

This book will be useful to tribologists, physicists, and mechanical engineers who are involved with fretting fatigue testing and those who are concerned with contact problems, particularly where fatigue and vibration are concerned, for example, in turbines, generators, aircraft, structures involving steel ropes, and so on The paper presented by Hattori et al., for example, shows how problems have been overcome in the design of steam turbines Vincent

et al and Vingsbo discussed the use of fretting maps for controlling the fretting fatigue damage

in practice Other papers show the effectiveness of certain preventative measures such as surface treatment and cathodic protection in marine environments The papers presented in this publication cover the response of common-place materials, such as steel and aluminum, as well as the less conventional materials such as composites

Overview of the Papers of the Symposium

This special technical publication contains 20 papers written by renowned authorities in this field The opening keynote paper, presented by R B Waterhouse, provides a global overview

of the problems of fretting fatigue testing and presents the author's perspective and views on the main issues that should be addressed in any attempt to standardize fretting fatigue testing

In addition, a total of four invited keynote papers were also presented by Vingsbo, Hoeppner,

1 Copyright 9 1992by ASTM International www.astm.org

Trang 8

2 FRETTING FATIGUE TEST METHODS AND EQUIPMENT

Hills, and Vincent to simulate and set the stage for focused and fruitful discussion during the

symposium The closing paper by Attia, the Chairman of the ASTM Task Group E9.08.02 on

Fretting Fatigue Testing, examines the future prospects for standardization in relation to the

current practice The paper presents also the results of a survey in which the input was solicited from 65 active researchers in various parts of the world

This special technical publication reflects the trends and testing philosophy in ten different

countries and is, therefore, characterized by its international flavor Apart from the opening and closing position papers, the papers of this symposium are grouped in five sections:

FundamentaI Aspects of Fretting Fatigue Testing Conceptual Framework

This section includes four papers that provide a conceptual framework for the mechanical

and physical interactions associated with the fretting fatigue process and testing Following a

brief presentation of the historical evolution of the fretting fatigue testing, Hoeppner reviewed

the mechanism of fretting fatigue and the contributions that have been made in understanding

the crack nucleation and in characterizing the fretting fatigue damage He underlined those

parameters that can be considered as mechanism controlling and presented the recent devel-

opments in micromechanical modeling The paper concluded with the recommendation for

standards development and the identification of some areas that warrant further research

Vincent, Berthier, and Godet applied their concept of "velocity accommodation" to the

fretting process and showed that the relative displacement and velocity difference between the

core of contacting solids are accommodated at different sites (the rubbing solids, their inter-

face, or the surface screens) and according to different modes (elastic, rupture, shear, and roll-

ing) Depending on the surface tensile stresses and whether adhesive welds break before crack

initiation, it was indicated that the material responds to fretting in three different ways: no

degradation, crack formation, and particle detachment Since different material responses can

be observed during a single test, the authors stressed the importance of constructing "fretting

maps" to identify the material response to specific running conditions To extend the velocity

accommodation and fretting maps concepts to fretting fatigue testing and to overcome the

classical problem of the dependence of the displacement amplitude on the body stress level,

the authors proposed a new "fretting-static fatigue" testing method This method, which is

based on applying a constant body stress and controlling the slip amplitude independently,

requires a set of fretting maps to be produced for different loads, slip amplitudes, and number

of cycles The authors proposed also a measure for the "severity" of the test, and outlined how

the design engineer can use these maps to identify and avoid fretting fatigue failures In this

paper, some fundamental questions were raised, regarding the contact mechanics parameters

that govern crack initiation/propagation, and the significance of the drop in the fatigue

strength measured in fretting fatigue test machines The latter issue was discussed in relation

to the formation/retainment of wear debris, and the effect of the machine stiffness

The subject of fretting maps, which define the effect of the process parameters on the extent

of the stick, partial- and gross-slip regimes, was also discussed by Vingsbo Using a simple

model of surface asperities in elastic contact with a' perfectly flat semi-infinite body under

cyclic loading, the author concluded that surface fatigue is promoted by fretting under mixed

stick-slip conditions, both in terms of cyclic stress concentrations and plastic deformation in

the contact zone The author's view on establishing fretting maps for a given tribo-system to

control the fretting fatigue damage in practice is readily applicable to the design of a controlled

fretting fatigue testing system

Perhaps the most difficult problem to be encountered in developing standards for a con-

trolled and well-defined fretting fatigue test is handling the large number of process variables

The popular list of variables, which was originally assembled by Collins in 1964, includes as

Trang 9

OVERVIEW 3

many as 50 variables! In reviewing the stress models, which were successfully used to predict the fretting fatigue strength, Dobromirski argued that the vast majority of these variables,

which are not explicitly included in the stress models, can be treated as "secondary" variables

that influence the process through their effect on the "primary" variables The latter is a short list of three variables, namely, the coefficient of friction, the displacement amplitude, and the

contact pressure The coefficient of friction was further singled out and identified as the main

primary variable By re-examining a large sample of available fretting wear/fatigue data, from this perspective, the author was able to use the coefficient of friction as a c o m m o n denomi- nator to explain the effect of various process parameters on the fretting fatigue test results

Beyond the obvious benefit of reducing the list of variables to a manageable and practical num-

ber, Dobromirski's analysis should be taken one step further to alert all of us that the time has come to treat the coefficient of friction as one of the parameters that should be measured in fretting fatigue testing It will be noted throughout this book that the emphasis on the critical

role of friction force is echoed by many others

Fundamental Aspects of Fretting Fatigue Testing Mechanics of Contact

This section includes four papers that deal with the theoretical aspects of the mechanics of contact, and the application of numerical techniques; for example, finite-element and bound-

ary-element methods to calculate the contact stresses Experimental verification, using the

caustics method, is also presented The authors maintained their focus on the main objectives

of this symposium and presented their analysis in terms of two important issues: the design of the fretting pad/fatigue specimen and the method of applying the normal contact load

The paper presented by Hills and Nowell is centered around the idea that specimen/pad

geometry should be amenable to a well-defined stress field and fracture mechanics analysis

They highlighted the drawbacks associated with the flat-ended fretting pad; for example, the

singularities in the contact stress distributions and the difficulty in defining the slip-stick zones

They recommended the adoption of a "cylindrical bridges against flat tensile specimens" configuration, since it allows changing the contact size while keeping constant normal load, as well

as controlling the normal and tangential contact forces independently The paper deals with

some points of interest to those involved with the task of developing standards for fretting

fatigue tests, namely, the contact size threshold phenomenon and the nature of the distribution

of the coefficient of friction over the contact area

Using the boundary element method, Sato studied the effects of clamping position (central

versus edge clamping) as well as the bridge height on the magnitude and the distribution of the

contact pressure at the specimen/bridge interface The results of the plane-stress analysis of the

bending fatigue problem were validated experimentally, using the method of caustics The concept of"equivalent stress amplitude," as defined by Tresca's yield criterion, was proposed

by the author for estimating the fretting fatigue strength From the S-N fretting fatigue test

results, it was established that the bridge height affects the fatigue life only under central clamp-

ing conditions (negative effect) The author was successful in interpreting these results in rela-

tion to the contact pressure amplitude, defined as half the difference between the compressive

and tensile contact pressures at the outer edge of the contact area The paper was concluded with the recommendation to use either central clamping when the bridge height-to-contact length H / L ratio is unity, or to use edge clamping for fretting fatigue tests with other H / L ratios

To improve the fretting fatigue strength, the author demonstrated a way of reducing the contact pressure amplitude through the machining of grooves in the fatigue specimen near the

end of the bridge

The application of the boundary element method for calculating the contact pressure dis-

tribution and the concept of controlling it through grooving and surface knurling were also

Trang 10

discussed by Hattori, Nakamura, and Ishizuka In this paper, the fretting fatigue limit was pre-

dicted using the fracture mechanics approach These predictions were also verified experi-

mentally The paper addresses some interesting points in relation to the measurement and

modeling of the effective stiffness of the contact interface The example given in the paper for

improving the fretting fatigue strength through optimization of the groove geometry (to cour~-

teract the negative notch effect with the positive effect associated with the rise in the threshold

stress intensity range factor) provides a methodology for designing the configuration of fretting

fatigue test specimens

The effect of the average contact pressure on the fretting fatigue strength was further invesL

tigated experimentally by Nakazawa, Sumita, and Maruyama The test results indicated that

the relationship between the fretting fatigue life and the contact pressure is influenced by the

stress amplitude At low-stress amplitude ( < 2 0 % of the 0.2% PS of high strength steel), this

relationship is nonmonotonic and passes through a minimum and then a maximum before

reaching a constant level At high-stress amplitude ( > 4 0 % of the 0.2% PS), the increase in the

contact pressure leads to a continuous drop in the fretting fatigue life The authors reported

also the increase in the frictional stress amplitude with the increase in the contact pressure For

the steel used, it has been indicated that the crack initiation sites shift from the middle portion

of the contact area to the outer edge as the contact pressure is increased This observation is of

a particular importance to fracture mechanics analysts who usually assume that cracks initiate

at the contact edge

Fretting Fatigue Testing Methods and Equipment

In this section which includes five papers, the present state of the art in fretting fatigue testing

is reviewed, and the relative merits of various test methods are evaluated A few recommen-

dations were made regarding the adoption of commercial equipment, proven techniques and

experimental test rigs, as a starting point for standards development Some interesting con-

cepts and observations were also made, providing guidelines for conducting proper simulative

tests

The fretting fatigue testing and research activity at the Royal Aerospace Establishment

(RAE) the U.K was critically reviewed by Rayaprolu and Cook Over the last 15 years, the

test methods and test variables at RAE were progressively changing to satisfy specific require-

ments and objectives Four stages or test series were identified by the authors to reflect such a

change The conventional fretting fatigue setup with a proving ring was used in the first test

series to investigate the effects of the pad span, contact load and body loading type on the

fatigue endurance The second test series was motivated by the need for knowing the local

stresses induced by fretting in order to apply fracture mechanics models Here, the frictional

force measurement was introduced In the third stage, the experimental research effort was

directed towards identifying the separate effects of the contact, frictional, and body loads on

the fatigue process Using a biaxial fatigue machine with phase linked actuators, a fourth series

of tests is being currently undertaken to examine the effect of cyclic load variations on the

cyclic frictional load, as well as crack initiation and propagation The paper summarizes also

the work related to fracture mechanics modeling at RAE Recommendations for standard test

setup, procedures, and future work were presented in the last two sections of the paper To

improve the fracture mechanics prediction capability, the effect of the contact parameters on

crack initiation and growth, particularly with reference to initiation sites and angular and short

crack growth, was identified as an important area for further research It is worth noting that

this recommendation is well founded by the observations made by Nakazawa et al

The paper given by Lindley and Nix described the two fretting fatigue test methods used at

the National Power Technology and Environmental Centre in the United Kingdom These

Trang 11

OVERVIEW 5

methods are similar to those used and recommended in the previous paper by Rayaprolu and

Cook, namely, the proving ring and the biaxial test rigs The advantages of the latter system

were discussed in terms of controlling the contact load and the relative slip between the spec-

imen and the pad, as well as applying variable amplitude loading The paper describes also

alternative fretting pad geometries and emphasizes the requirements for frictional force mea-

surement during the test The two approaches of fretting fatigue analysis, the S-N curve and

the fracture mechanics modeling, were also reviewed

For a proper simulative fretting fatigue testing, Ruiz, Wang, and Webb introduced the

fatigue-fretting damage parameter (FFDP), as a measure of the severity of fretting fatigue dam-

age This parameter is a function of the tangential stress along the line of contact, the interface

shear stress, and the relative slip and, therefore, includes the variables that control the initia-

tion of fretting surface damage (wear) and the growth of the cracks The main thrust of the

paper is centered around the importance of getting the three components of the FFDP right

in any test designed to reproduce the conditions prevailing in a real structural joint The paper

discussed further the issue of controlling these variables in three types of tests: biaxial, tension/

compression, and 3-point bending tests The authors pointed out the proper choice of the test

method, depending on the ductility of the material tested

The paper presented by Fischer, Grubisic, and Buxbaum deals with a very important and

fundamental issue in fretting fatigue testing: the effect of load sequence The experimental

study carried out by the authors on the fretting fatigue behavior of nodular cast iron under

constant amplitude and load spectrum (random sequence) throws the light on a few important

findings First, the c o m m o n test practice of constant stress amplitude produces more reduction

in the fretting fatigue strength because of higher slip amplitude and higher degree of "embed-

ding." Second, the widely accepted notion of the negative effect of the contact pressure on the

fretting fatigue strength (under constant stress amplitude) cannot be extended to the case in

which the stress amplitude follows a random sequence Third, the significant improvement in

the fretting fatigue strength with residual compressive stresses, for example, due to shot peen-

ing, was not observed in plain fatigue testing under spectrum load Although these conclusions

cannot be generalized, at the moment, beyond the test conditions reported by the authors, they

demonstrate the importance of proper simulation of the loading conditions encountered in

practice and suggest the improved repeatability of the test results under random sequence

loading, even when the contact pressure and residual stresses are not precisely controlled and

defined

Labedz's paper deals with the adaptation of commercially available servo-hydraulic testing

machines and the use o f a univeral test rig for fretting testing The proposed test method is in

harmony with Dobromirski's concept of primary/secondary variables and considers only five

essential test variables The author brings to our attention two test parameters that are usually

ignored in fretting wear/fatigue testing: the contact temperature and the residual stresses The

effect of the latter was experimentally investigated to confirm its importance and to demon-

strate the proposed test method

Environmental and Surface Conditions

This section includes three papers that deal with the effect of surface residual stresses and

the environmental conditions (for example, temperature, vapor content, and corrosivity) on

the fretting fatigue test results These papers point out the importance of monitoring and dupli-

cating the environmental conditions and the state of stresses at the surface of the specimen

Some experimental techniques, for example, X-ray diffraction, scanning electron microscopy

(SEM), atomic emission spectroscopy (AES), Mossbauer spectrometry, and electrochemical

techniques were described

Trang 12

The effects of the compressive residual stresses and the environmental temperature on the

fretting fatigue test results were investigated by Mutoh, Satoh, and Tsumoda Some consid-

erations for testing and measuring frictional forces at elevated temperatures were discussed

The paper examines also the relationship between the coefficient of friction and the stress

amplitude It has been concluded that for the given test conditions, this relationship is unique

regardless of the temperature and the surface residual stresses This behavior was attributed to

the insensitivity of the following mechanisms to surface and environmental conditions: oxi-

dation (to temperature), and surface roughness and hardness (to shot peening)

The paper presented by Yunshu, Baoyu, and Weili focused on the effect of the environment

on the debris structure and its tribological properties Using surface analysis techniques, the

authors concluded that if the environmental conditions promote the wear debris to act as an

effective solid lubricant, the fretting fatigue strength will be partially restored, as in the case of

blade steel fretted in vapor They also concluded that the environmental effects become less

important in the presence of compressive stresses

The paper presented by Price and Taylor is concerned with two issues: the synergistic effect

of the mechanical and electrochemical components of the fretting fatigue process and the

application of electrochemical techniques to separate and evaluate the role of corrosion in tests

run in aqueous environment An experimental setup was developed to control the corrosivity

of the medium and to identify the electrochemical dissolution process through the use of

impressed cathodic protection For the test conditions specified in the paper, the authors

concluded that the electrochemical processes have the greatest influence on the fatigue life of

high-strength low-alloy steel The paper draws the attention to the requirement of assessing the

contribution of the corrosion action in fretting fatigue testing, and provides a method for

achieving that

Nonconventional Materials and Test Methods

This section includes two papers that deal with nonconventional test configuration and

materials The fretting fatigue testing system developed by Cardou, Cloutier, St Louis, and

Leblond to test overhead electrical conductors is based on exciting the conductor at the span

midpoint, with a controlled cyclic deflexion The concept of primary and secondary test vari-

ables was independently applied in this paper, and two test methods were followed, namely

the wire fracture time sequence and fracture location analysis

In the paper presented by Jacobs, Friedrich, and Schulte, a special test setup was developed

to study the mechanism of fretting fatigue of carbon fiber reinforced expoxy (CFRE) lami-

nates In contrast to the observation made by Lindley and Nix, the authors found that the

fretting fatigue life of CFRE is significantly affected by the fretting pad material This was con-

tributed to the mechanism of interaction between fretting wear damage and fatigue, which is

also sensitive to the contact pressure and the hardness of the fretting pad material The authors

established that the fretting fatigue mechanism of fiber reinforced polymers is characterized

by multiple matrix cracking along the fibers and, therefore, the available fracture mechanics

models are not applicable to these materials A theoretical model for the "fretting fatigue load

versus number of cycles to failure" and the "specific pseudo-wear rate" was developed and

verified experimentally

Acknowledgment

The editors are indeed grateful to the authors for their valuable and original contributions

The effort of the reviewers in streamlining and improving the clarity of the presentation is

highly appreciated Special thanks are due to Dr R Frishmuth, of the Vecto Gray Inc., Hous-

Trang 13

The editors would like to express their thanks to the officers and members of the ASTM Committee E-9 on Fatigue for their support and also to the publication staffofASTM for their patience and support that made this publication possible

This publication is only one aspect of the symposium The sessions and the discussions con- tribute greatly to the mission of the symposium The effort of the co-chairmen of the sessions

is acknowledged and appreciated The editors are thankful to the attendees of the symposium for the interesting points and useful comments they made during the discussions that followed the paper presentaion, and during the panel discussion session Their enthusiasm to follow up this symposium with similar conferences in the future is appreciated and well taken The editors hope that those concerned with the subject of fretting fatigue will find this publication useful and stimulating

M Helmi Attia

Ontario Hydro Research Division, Toronto, Ontario, Canada; symposium chairman and editor

R B Waterhouse

Department of Materials, Engineering and Materials Design, University of Nottingham- symposium chairman and editor

Trang 14

R B W a t e r h o u s e ~

A Historical Introduction to Fretting Fatigue

ization of Fretting Fatigue Test Methods and Equipment, ASTM STP ]15 9, M Helmi Attia and

R B Waterhouse, Eds., American Society for Testing and Materials, Philadelphia, 1992, pp 8-

9

KEY WORDS: fretting fatigue, fatigue properties, historical perspective, crack propagation

Fretting was first reported by Eden et al in 1911 [1 ] who found that brown oxide debris was

formed in the steel grips o f their fatigue machine in contact with a steel specimen It was not

until 1927 that Tomlinson [2] conducted the first investigation of the process and designed

two machines to produce small-amplitude rotational movement between two annuli in the

first case, and an annulus and a flat in the second The movement was controlled by a long

lever system Since the resultant debris on his steel specimens was the red iron oxide c~Fe20~,

which had arisen from chemical reaction with oxygen in the air, he coined the phrase "fretting

corrosion." He also established that the damage could be caused by movements with ampli-

tudes as small as a few millionths of an inch ( ~ 125 nm) and the important fact that relative

movement had to occur, which he termed "slip."

The effect that fretting could have on fatigue properties was first investigated by Warlow-

Davies [3] in 1941, who produced fretting damage on the gage length o f steel fatigue specimens

and found a subsequent reduction in fatigue strength caused by the pitting of the surface, of

between 13 and 17% This was to be expected, but later investigations, particularly by McDow-

ell [4] showed that the conjoint action of fretting and fatigue, which is the usual case in prac-

tice, was much more dangerous, producing strength reduction factors of 2 to 5 and even

greater F e n n e r and Field [5] in 1958 demonstrated that fretting greatly accelerated the crack

initiation process I published my first research paper in 1961 and showed that recrystallization

o f the ferrite occurred in the fretted region when a bright drawn mild steel was subjected to

fretting fatigue [6] The first major investigation was by Nishioka and Hirakawa who pub-

lished a series o f six detailed papers that were inspired by a problem encountered in the rolling

stock o f the Shinkansen [ 7] Subsequent experimental investigations have been based on their

valuable work They also were the first people, together with Liu et al [8], to attempt an anal-

ysis of fretting fatigue This is an area that has seen great developments in the succeeding years

and forms a major part o f this publication

References

[ 1] Eden, E M., Rose, W N., and Cunningham, F L., "Endurance of Metals," Proceedings of the Insti-

tute ( f Mechanical Engineers" Vol 4, 1911, pp 839-974

[2] Tomlinson, G A., "The Rusting of Steel Surfaces in Contact," Proceedings q/the Royal Society, A

Vol 115, 1927, pp 472-483

Department of Materials Engineering and Materials Design, University of Nottingham, University

Park, Nottingham NG7 2RD, England

8

Copyright 9 1992by ASTM International www.astm.org

Trang 15

WATERHOUSE ON A HISTORICAL INTRODUCTION 9

[3] Warlow-Davies, F J., "Fretting Corrosion and Fatigue Strength," Proceedings of the Institute on

Mechanical Engineers, Vol 146, 1941, p 32

[4] McDowell, J R., "Fretting Corrosion Tendencies of Several Combinations of Materials," Sympo-

sium on Fretting Corrosion, STP 144 American Society for Testing and Materials, Philadelphia,

1953, pp 24-39

[5] Fenner, A J and Field, J E., "La Fatigue Dans les Conditions de Frottement," Rev MOt., Vol 55,

1958, pp 475-485

[6] Waterhouse, R B., "Influence of Local Temperature Increases on the Fretting Corrosion of Mild

Steel," Journal of Iron and Steel Institute Vol 197, 1961, pp 301-305

[7] Nishioka, K and Hirakawa, K., "Fundamental Investigations of Fretting Fatigue," Bulletin of the

Japan Society of Mechanical Engineers, Vol 12, 1969, pp 180-187,397-407, 408-414, 692-697;

Vol 15, 1972, pp 135-142

[8] Liu, H W., Corten, H T., and Sinclair, G M., "Fretting Fatigue Strength of Titanium Alloy RC

130B," Proceedings ofASTM, Vol 57, 1957, pp 623-641

Trang 16

Opening Paper

Trang 17

R o b e r t B W a t e r h o u s e I

The Problems of Fretting Fatigue Testing

REFERENCE: Waterhouse, R B., "The Problems of Fretting Fatigue Testing," Standardiza- tion of Fretting Fatigue Test Methods and Equipment, ASTM STP 1159, M Helmi Attia and R

B Waterhouse, Eds., American Society for Testing and Materials, Philadelphia, 1992, pp 13-

19

ABSTRACT: Fretting fatigue testing usually arises as the result of some failure which it is imper- ative to overcome either by a modified design or application of some surface treatment In such cases the test rig is usually designed to replicate the actual situation as closely as possible (e.g., a press fit or riveted joint) with environmental conditions as near to those occurring in practice (e.g., high temperature or a marine atmosphere) In laboratory testing the purpose is much wider, and usually entails, for instance, the assessment of different materials for their susceptibility to fretting damage, or the effect of variables such as clamping load, amplitude of slip and frequency,

in particular environmental conditions The type of Fatigue test is very relevant (i.e., whether the response is the same in rotating-bending, push-pull or torsion, and whether in the latter two cases, a mean stress is applied) The nature of the contact is also material (i.e., whether it is flat- on-flat or cylinder-on-flat, or even crossed cylinder) How is the clamping stress to be applied? If

by a proving ring, then the pressure can change according to whether the debris is trapped or can escape; a dead weight method might be preferable Surface finish and residual stress are factors which must be taken into account All these matters must be considered in devising a recommended testing procedure

KEY WORDS: fretting fatigue, fatigue testing, fretting device, clamping pressure, slip amplitude, frequency

W i t h fretting being the small a m p l i t u d e oscillatory m o v e m e n t between two contacting surfaces, it is o b v i o u s that in m a c h i n e s a n d structures subjected to vibration its potential occurrence is to be c o m m o n l y expected T h e p r o d u c t i o n o f wear debris, although a nuisance and with possible long-term consequences, is n o t so serious as the initiation o f fatigue cracks and their subsequent propagation, where the m o v e m e n t arises f r o m the cyclic stressing o f o n e o f the c o m p o n e n t s or in the presence o f a static tensile stress S o m e o f the failures initiated by fretting fatigue have had tragic c o n s e q u e n c e s (e.g., the loss o f the C h i n o o k helicopter in the

N o r t h Sea in N o v e m b e r 1986) [I ]; others have had serious e c o n o m i c consequences [2], as in the case o f the failure o f a p o w e r station generator rotor M a n y failures reported in the literature have had less devastating consequences These have i n v o l v e d the o u t p u t shaft flange o f a helicopter [3], turbine disc failure in a gas turbine aero-engine [4], failure o f wire reinforce-

m e n t s in radial tires [5], steel ropes [6], the supporting j o i n t o f a railway line [7] and an arti- ficial hip j o i n t [8] to n a m e a few T h e investigation o f such incidents has provided useful infor-

m a t i o n for avoiding c o n t i n u i n g failures, b u t frequently m o r e i n f o r m a t i o n is required which can only be achieved by s o m e f o r m o f testing

Associate Reader, Department of Materials Engineering and Materials Design, University of Not- tingham, Nottingham, NG7 2RD, United Kingdom

Trang 18

The Object of Fretting Fatigue Testing

Where an immediate fretting problem has to be tackled, the usual experimental arrange-

m e n t is designed to replicate as closely as possible the actual failure situation (e.g., a fly wheel

on a shaft where the author constructed a scale model o f the system, a turbine blade/disc dovetail fixing [9], the rail failure mentioned above [10], or bridge suspension ropes [11]) One of the great difficulties here is to ascertain and then reproduce the mechanical conditions of the original joint, particularly the loading in the contact and the amplitude of movement Some- times these may be measured experimentally [ 12], or failing that, finite element analysis may provide the answer The environmental conditions should also be reproduced, e.g., temperature, humidity, seawater or salt spray, or industrial atmospheres The purpose of the testing is

to examine the effects o f such factors as modification to the original design, surface finishing and surface treatments, such as shot-peening, surface rolling, anti-fret coatings, or even a change of base material Some manufacturers will try several or all possibilities together, whereas, if time permits, it is more economical in the lbng run to pinpoint the o p t i m u m solu- tion and hence to identify the source o f the original problem Of the options available,

i m p r o v e m e n t in design is the most satisfactory, but this may be disruptive and expensive The more general objective o f fretting fatigue testing is the assessment of the susceptibility

of different materials to this type of failure In the case of metallic alloys this may involve consideration o f such factors as stacking fault energy heat-treatment (particularly in the case of steels whether tempering or normalizing is to be preferred) whether cold worked or annealed, forged or cast, and, o f course, hardness Further to this is the combination of different materials, since contacts of the same material have been thought inadvisable, as local welding is more likely Also important is the effect o f mechanical variables, such as contact pressure, amplitude o f slip, frequency, the influence o f random or sinusoidal loading, the effect of residual stresses and the nature of the surface finish Finally, a systematic study of environmental influences may be necessary, particularly the temperature, whether high or low, the compo- sition o f the gaseous or liquid environment and possible fluctuations thereof

Practical Considerations

Specimen

The type o f specimen is dictated to some extent by the choice of fretting contact, discussed below The most c o m m o n contact is flat-on-flat or cylinder-on-flat This means that the specimen generally has to have a gauge length with parallel flats This also means that the specimen must itself be machined from plate or sheet material, or if the specimen is of circular cross section, parallel flats must be machined on the gauge length Typical examples are shown in Fig 1 The existence o f corners means that there are stress concentrations present In practice,

it is usually found that the fretting is much more severe and cracks initiate there rather than

P O U S H LONGITUDINALLY

SECTION A - A

FIG l Design of specirnen with machined flats

Trang 19

WATERHOUSE ON PROBLEMS OF FRETTING FATIGUE TESTING 15

at the corners of the specimen Attention must be paid to the type of grips holding the specimen, since fretting is often experienced in them This can usually be overcome by using screw threads or applying an anti-fret lubricant or some form &insert

Fretting Contact

Fretting contact is usually provided by some form of bridge A pair of these is clamped on

to the gauge length by a proving ring, which is strain gauged to allow adjustment of the clamping pressure In a push-pull machine, one end of the bridge can be located on the grips In this case the amplitude of slip is determined unequivocally, whereas, in the isolated bridge, it is usually assumed that the slip is equally distributed between the two feet of the bridge Provided that loading is the same on each foot, experimental observation of the damage supports this assumption Figure 2 shows two possibilities If the specimen is vertical, as in many servohydraulic machines, a dead-loading arrangement may be possible to apply the normal load

In rotating-bending types, the proving ring must be carefully balanced to prevent vibration The feet of the bridge may be flat and sufficiently low in height to minimize elastic deformation and allow relative slip to occur They may also be chamfered or cylindrical Cylinder/flat con-

FIG 2 Arrangement of specimens and fretting bridges" (a) unlocated (b) located on grips

Trang 20

tacts allow stress distribution to be calculated initially, but as soon as any wear occurs, the contact situation is changed

The disadvantage of the proving ring is that if debris is formed and retained in the contact, since it is usually o f greater volume than the metal from which it arises, the pressure will increase On the other hand, if the debris can escape, there will be a decrease in pressure Thus, pressure needs to be checked regularly Choosing the material of the bridge the same as that of the specimen eliminates a complicating variable and possibly gives the most severe type of fretting damage The length dimension of the bridge will determine the range of amplitude slip, which is governed by extremes in the cyclic stressing of the specimen

Normal Load

The effect of normal load on the fretting fatigue strength is of the form shown in Fig 3 It would seem advisable to choose a load which is within the horizontal part of the curve There- fore, slight variations, due to the effects of debris, will not be significant Too heavy a load may result in digging in if the bridge has sharp corners A recent survey of the published literature

by the author indicated that in steel specimens and flat-footed bridges the range of clamping pressure was mainly between 20 and 1,90 MPa

Amplitude of Slip

In most of the investigations the amplitude of slip has not been constant, but has depended

on the range of alternating stress, and hence alternating strain, in the specimen This is the usual situation in practical cases of fretting fatigue The range of slip amplitude can be imposed

by suitable choice of the length dimen,;ion of the bridge related to the stress range under consideration An elaborate arrangement was developed by Nishioka and Hirakawa to allow the amplitude of slip to be kept constant whatever the stress in the specimen [14]

Most damaging range o f slip is between 18 [15] and 25u [16], although Lindley and Nix have suggested that coefficient of friction is a more important factor [17] A typical plot of

300

( 9

z i.iJ rr"

CLAMPING PRESSURE, MPa

FIG 3 Fretting Jatigue stren~h versus clamping load

Trang 21

WATERHOUSE ON PROBLEMS OF FRETTING FATIGUE TESTING

0.8

17

0.6 _o

I.- 0.4

Z iii

~ 0.2

AMPLITUDE OF SLIP, p.m FIG 4 Co~,~cienl (~/i/riction versus amplitude of slip

coefficient of friction against amplitude of slip is shown in Fig 4 as the transition from partial

to total slip

Type of Testing

In a push-pull machine, a non-zero mean stress can be applied There is evidence that mean

stress has a somewhat greater effect in fretting fatigue than in plain fatigue [ 18] In plane bend-

ing it is usual for one end of the fretting device to be located in the grips Since a parallel sided

plane specimen will have a varying bending m o m e n t along its length, it will be important to

locate the fretting contact at exactly the same point in comparable tests A shaped specimen

would overcome this to some extent, provided that the apparent contact areas were kept rea-

sonably constant In addition, there has been little consideration of torsional fatigue testing in

fretting fatigue

The question o f frequency of the alternating stress should also be considered Higher fre-

quencies allow testing time to be cut down However, possible heating of the specimen and

fretting contact have to be borne in mind, particularly as chemical interaction with the envi-

r o n m e n t is an important feature of fretting

Conclusion

In devising a standard test for fretting fatigue it would seem advisable to stipulate a particular

type of specimen and fatigue testing facility An agreed type of fretting bridge is also necessary

Since the coefficient of friction is an important factor, the fretting bridge should, if possible,

be strain gauged to allow for its measurement The normal load applied should be related to

the yield pressure o f the material The amplitude o f slip should be limited, since at higher val-

ues the wear process predominates and the effect on fatigue is reduced, as illustrated in Fig 5

Surface finish and surface residual stress need to be stipulated or, at least, recorded For testing

in normal atmosphere, the temperature and humidity should be recorded and, if possible, held

within certain limits

The question remains whether the information gathered from such testing can be applied

to a particular practical situation A recent paper suggests that results of the fatigue limit of

D o w n l o a d e d / p r i n t e d b y

Trang 22

[1] Department of Transport Aircraft Accident Report 2/88 H.M Stationery Office 1989

[2] Lindley, T C., McIntyre, P., Snow, D J., iand Wilson, J D., "Fatigue and Environmental Cracking

in Turbo-generators," Proceedings, Sixth Thermal Generation Specialists Meeting, Madrid, 1981

[3] Eckert, J and Richter, R., "Fretting Corrosion of an Output Shaft Flange," Prakt Met., Vol 2 l,

1984, pp 140-143

[4] Lindblom, T and Burman, G., "Fatigue Failure under Fretting Conditions," Proceedings, Confer-

ence on High Temperature Alloys for Gas Turbines, Liege Belgium 4-6 Oct 1982, Riedel Publishing Co., Dordrecht, 1982, pp 673-684

[5] Seitz, N and Schmid, R., "Korrosion bei PKW-Stahlgurtelreifen," Kautschuk + Gurnrni-Kunstoffe,

Vol 40, 1987, pp 20-27

[6] Hobbs, R E and Ghavani, K., "The Fatigue of Structural Wire Strands," Int J Fatigue, 1982, pp

69-72

[7] Okazaki, A., Urashima, C., Sugino, K., Matsumoto, H., and Hattori, M., "Upper Fillet Crack in

Bolted Joint of Rails and its Causes," Transactions t?fl.S.1.J., Vol 23, 1983, pp B22

[8] Smethurst, E and Waterhouse, R B., "Causes of Failure in Total Hip Prostheses," J Mat Sci., Vol

12, 1977, pp 1781-1792

[9[ Ruiz, C and Chen, K C., "Life Assessment of Dovetail Joints between Blades and Discs in Aero-

Engines," Fatigue of Engineering Materials and Structures, I.Mech.Eng., London, 1986, pp 187-

194

[10] Urashima, C., Sugino, K., Nishida, S-I., and Matsumoto, H., "Factors on the Upper Fillet Crack

Initiation and its Preventive Measures," Transactions ofl.S.I.Z, Vol 23, 1983, pp B23

[11] Blakeborough, A and Cullimore, M S G., "Fretting in the Fatigue of Wire Rope," Advances in

Fracture Research, 6th Int Conf on Fracture, New Delhi, India, 4-10 Dec 1984, Pergamon, New

York, Vol 3, 1984, pp 2133-2141

[12] Fisher, N J and Ingham, B., "Measurement of Tube-to-Support Dynamic Forces in Fretting-Wear

Rigs," Journal of Pressure Vessel Technology, Transactions of ASME, Vol 111, 1989, pp 385-393

Trang 23

WATERHOUSE ON PROBLEMS OF FRETTING FATIGUE TESTING 19

[13] Waterhouse, R B., "The Role of Adhesions and Delaminations in the Fretting Wear of Metallic

Materials," Wear, Vol 45, 1977, pp 355-364

[14] Nishioka, K and Hirakawa, K., "Fundamental Investigations of Fretting Fatigue Pt.2 Fretting

Fatigue Testing Machine and Some Test Results," Bulletin ofJ.S.M.E., Vol 12, 1969, pp 180-187

[15] Fenner, A J and Field, J E., "La Fatigue dans les Conditions de Frottement," Rev M~t., Vol 55,

1958, pp 475-485

[16] Gaul, D J and Duquette, D J., "The Effect of Fretting and Environment on Fatigue Crack Initia-

tions and Early Propagations in a Quenched and Tempered 4130 Steel," Met Trans A., Vol 11A,

1980, pp 1555-1561

[ 17] Nix, K J and Lindley, T C., "The Influence of Relative Slip Range and Contact Materials on the

Fretting Fatigue Properties of 3.5 NiCrMoV Rotor Steel," Wear, Vol 125, 1988, pp 147-162

[18] Fenner, A J and Field, J E., "A Study of the Onset of Fatigue Damage Due to Fretting," Trans-

actions, N.E Coast Instn Engrs and Shipbldrs, Vol 76, 1960, pp 184-228

[19] Gotoh, Y and Ohuchida, H., "Effect of Corrosive Environment on Fretting Fatigue Under Plane

Bending," Journal of the Society of Materials Science, Japan, Vol 38, 1989, pp 816-822

Trang 24

Fundamental Aspects of Fretting Fatigue Testing Conceptual Framework

Trang 25

D a v i d W H o e p p n e r I

Mechanisms of Fretting-Fatigue and Their

Impact on Test Methods Development

REFERENCE: Hoeppner, D W., "Mechanisms of Fretting Fatigue and Their Impact on Test

and Materials, Philadelphia, 1992, pp 23-31

ABSTRACT: At the beginning of this century very little information was available related to the

phenomenon of fretting and fretting-fatigue However, significant progress has been made in

developing an understanding of the mechanisms of fretting-fatigue in this century Progress

toward developing a holistic view of the process of fretting fatigue is presented

Contributions that have been made in understanding nucleation of fretting-fatigue damage are

reviewed As well, characterization of damage is discussed related to improving our understand-

ing Those parameters that are viewed as mechanism controlling, thus important to test methods

development, are reviewed The progress made in micromechanical modeling is also reviewed

The paper concludes with recommendations for the standards development group to consider

in relation to test methods and some suggestions for future research and development

KEY WORDS: fretting-fatigue, fretting corrosion, fretting wear, mechanisms, modeling,

standards

Fretting-fatigue has been a nemesis to designers, manufacturers, and operators of equipment of all types since h u m a n s first used machines to their advantage It has been in the 20th century that our progress at identifying, characterizing, and designing for prevention, allevi- ation, and controlling fretting fatigue has become more formalized Furthermore, we have characterized both the mechanisms and parameters that are involved in the fretting process However, no standardized procedures, at least within ASTM auspices, have been developed

to aid engineers at either the prospective or retrospective design stage This continues to be a major technological challenge This symposium, as well as previous ASTM symposia concerned with this subject [ 1-3], attempts to focus attention on this problem As a small portion

of this focus, this paper discusses some aspects of mechanisms that have been revealed during the recent decades Subsequently, the role of various parameters in influencing these mechanisms is discussed Finally, some conclusions and suggestions for future research are presented

Occurrence of Fretting Fatigue

Numerous publications have documented the occurrence of fretting or potential occurrence

of fretting in any mechanically fastened joint or in surfaces in contact under "small" relative motion If one or both of the contacting surfaces are under cyclic load in addition, then one

1 Professor and Chair, Mechanical Engineering Department, University of Utah, Salt Lake City, UT

84112,

Trang 26

or b o t h o f the m e m b e r s m a y experience fretting fatigue Reference 4 d o c u m e n t s m a n y specific

c o m p o n e n t s where fretting fatigue occurs S o m e are in the following systems:

W Barrois, a leader in design in the aircraft industry a n d N o r t h Atlantic Treaty Organiza-

t i o n - A d v i s o r y G r o u p for A e r o s p a c e Research and D e v e l o p m e n t - S t r u c t u r e s and Materials

Panel ( N A T O - A G A R D - S M P ) , wrote the following in 1970 [5]:

Until about 1940 fatigue investigations included: some basic research into the physics of metals in

an effort to discover its mechanism; systematic rotating-bending tests on smooth conical or waisted

specimens in order to qualify each metal by identifying its fatigue limit; finally, check tests on actual

parts

Between 1940 and 1945 it became obvious to most specialists that the rotating-bending test on

smooth specimens was not representative of the behaviour of actual components and that the fatigue

limit was not the only characteristic of interest One improvement consisted in performing axial ten-

sion-compression tests on cylindrical specimens having a V-groove with a rounded root, the smooth

specimen being replaced by a waisted one, with a very large radius of curvature, so that the heat

induced by internal damping should no longer increase the temperature of the specimen and falsify

the test results Furthermore, it was recognized that the purpose of fatigue testing is not to specify the

fatigue limit, which is often hypothetical, but to provide the entire stress versus number of loadings

curve from static strength, corresponding to one load application, up to a large number of cycles in

laboratory tests of short duration This number being large in comparison with the cycles sustained

by the structure during its service life, testing had to be speeded up as much as is practicable without

distorting the results

However, a factor o f great importance in many service fatigue incidents was left out in the notched-

specimen test: the contact alteration by friction due to very small relative displacements o f the various

parts o f an assembly during the loading and unloading cycles This phenomenon, known as "fretting,"

consists in the welding o f asperities on the surfaces in contact and in the tearing-up o f these micro-

welds; it is responsible for the initiation offatigue cracks in assemblies and, in cases like the fitting o f

wheels on shafts or the bearing o f bolts in lugs, it may reduce theJatigue life to a tenth ~ f what it would

otherwise be 2

After 1970, n u m e r o u s publications a n d conference proceedings emerged that provided

additional focus on the challenge o f fretting-fatigue prevention, control, and estimation [6-

2 Italics mine

Trang 27

HOEPPNER ON MECHANISMS OF FRETTING FATIGUE 25

19] The seminal book of Waterhouse [7] has provided much insight into fretting-fatigue Another book, edited by Waterhouse [i8], focused attention on the engineering challenge related to anticipation of fretting-fatigue in a prospective design as well as the prevention, prediction, and control o f fretting-fatigue Although some of the mechanistic understanding of fretting-fatigue has been applied to the repair of such damage, this is still a very great challenge The role of modeling the fretting damage is vital here and fracture mechanics has been coupled with fretting-fatigue and studied for many years This aspect will be briefly discussed in a later section The intensive efforts of the past 20 years have made the mechanisms by which fretting- fatigue proceeds much clearer It has been recognized for many years that a systems view is needed to study and design for fretting fatigue prevention since there are so many parameters involved References [20-30] have presented the need for a systems view in studying fretting fatigue The work by Czichos reported in [15] and expanded in [30] is a significant help to all engineers and scientists embarking on the journey to help develop fretting-fatigue standards The first step in the systems view is to understand why we do fretting-fatigue experiments Figure 1 presents a simplified view of the reasons Basically, the designer is interested in either the determination of a fretting-fatigue life reduction factor (sometimes part of a "joint design allowable") or in evaluating a fretting protection system Additional experimentation often is done to verify a fretting-fatigue prevention scheme

The systems view of fretting-fatigue is presented in Figs 2 and 3 The efforts of numerous investigators have indicated that designing to resist fretting fatigue is an extremely difficult

ON FRETTING STUDIES

i Purpose

I

J

i

Understanding ~ Test Methods Life ~Tlesting / of Fundamentals

FIG 1 Reasons.[or conducting frening-]atigue experimental studies

Trang 28

9 constant load amplitude

9 variable load amplitude

~ Chemical or L , electrochemical factors

f Potential Current density Passivity Oxide Time Pitting Dissolution Embrittlement Film formation

Basic Material ~ Mechanical Deformation ]

i

[ Response to environment, 1 combined contact and

cyclic mechanical deformation

- ~ " t Cyclic Loading ] -

I Stress range

I Stress amplitude ]Frequency

I Sequence of loading (spectrum)

t, Time/Waveform effects Product form, thickness, geometry, inspectability

i

FIG 3 A systems framework previously proposedjbr planning experimental studies q[[?etting-[atigue

Trang 29

task Part of this difficulty results from a lack of understanding of the mechanisms [7,9,10] and the lack of standardized fretting fatigue apparatus, definitions, and test methods Thus, this paper will focus additional attention on this significant challenge To assist with this, a clearer view of the mechanisms will be helpful

M e c h a n i s m s of Fretting-Fatigue

The National Materials Advisory Board (NMAB) report [4] provides an extensive review

of fretting fatigue mechanisms and is recommended to all who have an interest Numerous other studies have been reported in the literature Over the years, the understanding of the mechanisms of fretting fatigue have been reduced, by the author and others, to the following phases of fretting fatigue Aspects of this approach are summarized in Ref31 as well as numerous other writings

Surface or Crack Formation/ Crack Propagation Instability

Near Surface ~ Nucleation ~

Damage

Furthermore, intensive studies by m a n y investigators have shown that the mechanisms in Phase I, undoubtedly the most criticalphase offretting-fatigue, and the most difficult to study, result in "damage" that is summarized in Table 1 It has also been found that numerous parameters influence the "damage" formation The role of numerous parameters on fretting- fatigue mechanisms is below summarized by M a n n [32]:

Consequently, all parameters which may affect the generation of the service induced surface

"damage" must be reproduced in the laboratory as closely as possible, if there is to be any hope of developing transfer functions relating the experiments to "reality."

These parameters are discussed below beginning with the loading related parameters already mentioned Along with each are comments indicating the effect of the variable on the fretting fatigue mechanisms and the component life under conditions meeting the requirements for fretting fatigue Cyclic Load (Stress) Provides energy input for both crack nucleation and subsequent propagation; generates slip between the faying surfaces with the amplitude depending on the level of cyclic stress; increases the cyclic stress, reducing life; and gives a lower frequency which can reduce life by providing more time per cycle for corrosion (fretting interactions)

TABLE 1 "Damage"produced underJkettingz/btigue conditions

9 Metal Transfer (More General Material)

9 Cracks of Various Geometries at Various Angles to the Surface, Including Parallel

9 Fretting Craters

Trang 30

Normal Load (Stress) Generates contact stresses in body of fatigue specimen; provides energy

input for production of most surface damage mechanisms; and increases normal stress, usually low-

ering life, although, if sufficient stress is applied such that slip amplitude is reduced or eliminated,

can increase life

Slip Amplitude Results from applied load; its presence undoubtedly plays a role in surface dam-

age mechanisms; and it usually increases the slip amplitude, reducing fatigue life, but in some cases

higher slip amplitudes are associated with increased life due to the wearing away of nucleated

cracks

Number of Fretting Cycles As the number of fretting cycles increases, the amount of surface

damage produced increases; for some conditions a fretting fatigue damage threshold exists in terms

of the number of fretting cycles; continued fretting beyond the damage threshold produces no further

reduction of fatigue life; and this clearly relates to the energy threshold for crack nucleation alluded

to earlier and will definitely depend on the type of damage generated

Geometry of Mating Components This in part controls state of stress in components and, con-

sequently, affects crack growth thresholds and crack growth rates; extremely important in retaining

debris between contacting surfaces, which in turn affects surface damage mechanisms (detrimental

to life by generating pitting and enhancing Mode I crack growth rates); and retained debris affects

friction forces and stress sta~e

Elasticity This affects both stress state and slip amplitudes

Hardness Generally, harder surfaces resist fretting damage, however, they also reduce the tough-

ness of the surface layer which lowers the threshold energy requirement to nucleate fretting fatigue

cracks; this may or may not offset the advantage of the harder surface

Microstructure This can control near surface crack growth and crack nucleation via grain size,

hardness, etc

Combinations Similar metal contacts promote welding of asperities, enhancing debris produc-

tion and reducing fatigue life; galvanic cell considerations are important; and friction and thus tan-

gential surface stresses in fatigue component can be dramatically affected

Surface Roughness "Rough" surfaces can provide escape routes for debris (increasing life) or

also encourage gouging, scratching, and debris production which is detrimental to fatigue life

Environment Consideration must be given to temperature, humidity, and corrosive atmo-

sphere; and fretting fatigue life normally longer in vacuum due to absence of oxygen (which otherwise

can form oxides harder than parent material)

It should be noted that under each parameter, e.g., cyclic stress, normal stress, slip amplitude,

materials, and environment, there are offsetting or opposing phenomena as the magnitudes or con-

ditions of the controlling parameters vary This clearly is due to the interactions and synergisms of

these variables and therefore the complexity of the subject Consequently, it becomes of paramount

importance to view fretting fatigue deliberations from a systems context

F r o m M a n n ' s analysis, a n d the m a n y works on m e c h a n i s m s , this need for a systems frame-

w o r k is essential Thus, to develop standards in fretting-fatigue, a systems framework and even

greater c o o r d i n a t i o n o f activity will be required than in d e v e l o p m e n t o f previous fatigue

related standards

T h e recognition that a fretting-fatigue d a m a g e threshold exists has had an influence on the

e v o l u t i o n o f b o t h m e c h a n i s t i c and engineering understanding In addition, this d e v e l o p m e n t

has had a bearing on the research d o n e to characterize the m a n y parameters that are involved

H o w e v e r , a great deal o f research still needs to be d o n e to fully characterize the influence o f

the m a n y p a r a m e t e r s in the pre-threshold d o m a i n

T h e discovery o f the threshold also has had an influence on the d e v e l o p m e n t o f fracture

m e c h a n i c s based m o d e l s for fretting-fatigue, as briely discussed in the next section Even

t h o u g h significant progress has been m a d e since the d a m a g e threshold, m u c h controversy o v e r

the c o n t r o l l i n g p a r a m e t e r s still exists This is d r a m a t i z e d by m a n y o f the ideas in other papers

o f this s y m p o s i u m

Trang 31

H O E P P N E R O N M E C H A N I S M S O F F R E T T I N G F A T I G U E 29

Modeling

Subsequent to the presentation of the conceptual framework for a fretting fatigue damage threshold in [9,10], numerous publications [33,34] focus on the utilization of fracture mechanics based modeling These have all focused around the development of a methodology

as shown in Fig 4 The phases of this process that deal with crack propagation can be dealt with to an extent However, the phases on the left, i.e., fretting "damage" production and formation of cracks, and structurally dependent crack propagation are still very difficult to deal with because &limited knowledge in those areas Part of the difficulty in studying these areas

is the lack of standards concerned with fretting-fatigue

Future Needs

Based upon the extensive progress made in understanding the mechanisms of fatigue and characterizing the parameters that influence fretting fatigue, certain needs become evident First, clear definitions of all terminology are essential Even though we have developed definitions of fretting (Terminology Relating to Erosion and Wear, A S T M G 40), fretting corrosion (Definition of Terms Relating to Corrosion and Corrosion Testing, ASTM G 15), fretting wear (ASTM G 40), and fatigue (Definition of Terms Relating to Fatigue, ASTM E 1150),

further unification o f these definitions focusing on fretting fatigue is essential In addition, definitions of fretting fatigue, damage, induced cracks, damage threshold, etc are essential to improving the uniform reporting of information on fretting fatigue

Second, a compilation of terminology specifically concerned with fretting fatigue would aid our efforts a great deal It also would be of help to develop a uniform format (protocol) for reporting fretting fatigue experimental results

Crack Know Formotion

FIG 4 Fatigue life estimation JbrJkeuing-fatigue

Trang 32

Third, more coordinated efforts, such as could be stimulated within ASTM Committee E 9

on Fatigue would be helpful A round-robin activity could be developed ifa sufficient number

of participants become available

Fourth, the development of a fretting fatigue experimental practice guide similar to A Ten- tative Guide for Fatigue Testing and the Statistical A nalysis of Fatigue Data, A S T M STP 91A; Supplement to Manual on Fatigue Testing, A S T M - S T P 91; and Handbook of Fatigue Testing,

A S T M - S T P 566 would be extremely helpful This effort is already underway in laboratories

at the University of Utah It has been found that the diversity of fretting-fatigue apparati is astounding In addition, many fretting-fatigue investigators, even though well intentioned, violate fatigue testing practices as elucidated in Practice for Conducting Constant Amplitude Axial Fatigue Tests of Metallic Materials, A S T M E 466; Practice for Verification of Constant Amplitude Dynamic Loads in an Axial Load Fatigue Testing Machine, A S T M E 467; Rec- ommended Practice for Constant-Amplitude Low-Cycle Fatigue Testing, A S T M E 606; and

A S T M E 1150; etc This is a situation which could be improved upon in the future with such

a fretting fatigue experimental practice guide

Statistically planned round robin activities will eventually be needed, but it is imagined that

a great deal must be done before this can take place Nonetheless, the necessity to agree on an experimental protocol, including apparatus, and conduct statistically planned fretting fatigue experiments is essential to our further progress

Finally, continued interaction under ASTM auspices to accelerate activity is vital The interest and activity in fretting fatigue has obviously increased as attested to by the other papers in this volume Collectively, it is desirable that all researchers and engineers working in the field keep this m o m e n t u m going forward

Conclusions

From the information presented, the following conclusions can be made:

1 Mechanisms studied have resulted in identification of various forms of damage that occurs under fretting fatigue conditions

2 The principal parameters involved in fretting fatigue have been identified qualitatively Additional research is necessary to characterize quantitative influences

Acknowledgments

I wish to acknowledge the support of all my students over the past sixteen years In addition,

my colleagues at Battelle Memorial Institute and Lockheed Aircraft Corporation have been a great inspiration to all my efforts on fretting The past support of the Office of Naval Research and the Natural Sciences and Engineering Research Council of Canada is gratefully acknowledged Bob Jeal, now technical director of Hawker de Havilland in Australia, has continued

to interact with me and question my ideas on fretting-fatigue as well as numerous other areas

I am grateful to him for his interest and critical assessment of ideas Paula Jorgensen typed the manuscript and aided with the preparation of my presentation I am indebted to her for her efforts My thanks also to Mark Thomsen who aided with the graphics

References

[1] Symposium on Fretting Corrosion, STP 144, ASTM, Philadelphia, 1953

[2] Materials Evaluation Under Fretting Conditions, STP 780, ASTM, Philadelphia, 1981

[3] Selection and Use of Wear Tests for Metals, STP 615, ASTM, Philadelphia, 1977

Trang 33

[7] Waterhouse, R B., Fretting Corrosion, Pergamon Press, New York, 1972

[8] Devereaux, O F., McEvily, A J., Staehle, R W (editors), Corrosion Fatigue: Chemistry, Mechan-

neers), Houston, 1972

[9] Hoeppner, D W., Uhlig, H H., "Fretting, Cavitation, and Rolling Contact Fatigue Critical Intro- duction," Corrosion Fatigue: Chemistry, Mechanic's, and Microstructure, NACE-2 Conference, NACE (National Association of Corrosion Engineers), Houston, 1972, p 607

[10] Waterhouse, R B., "The Effect of Fretting Corrosion in Fatigue Crack Initiation," Corrosion

ciation of Corrosion Engineers), Houston, 1972, pp 608-616

[11] Hoeppner, D W., Goss, G G., "Research on the Mechanism of Fretting Fatigue," Corrosion

[12] Salkind, M J., Lucas, J J., "Fretting Fatigue in Titanium Helicopter Components," Corrosion

[ 13] Lum, D W., Crosby, J J., "Fretting Resistant Coatings for Titanium Alloys," Corrosion Fatigue:

Corrosion Engineers), Houston, 1972, pp 631-641

[ 14] Starkey, W L., "A New Fretting Fatigue Testing Machine," Corrosion Fatigue: Chemistry, Mechan-

neers), Houston, 1972, pp 642-645

[ 15] Fretting in Aircraft Systems, papers presented at the 39th meeting of the Structures and Materials Panel, CP-161, NATO, AGARD Symposium, 1975

[16] Barrois, W G., Manual on the Fatigue of Structures H Causes and Prevention of Structural Dam-

No 9, Nov 1975

[17] Barrois, W G., Manual on the Fatigue of Structures H Causes and Prevention of Damage 7

[18] Waterhouse, R B., editor, Fretting Fatigue, Applied Science Publishers, Ltd., Essex, England,

1981

[20] Waterhouse, R B., "Fretting Wear," Wear, Vol 100, 1984, pp 107-118

[21 ] Hoeppner, D W., "Fretting of Aircraft Control Surfaces," Specialists Meeting on Fretting in Aircraft Systems, published in AGARD Conference Proceedings No 161, AGARD, 7 Rue Ancelle 92200 Neuilly Sur Seine, France, 1974, pp 9-13

[22] Goss, G L., Hoeppner, D W., "Characterization of Fretting Fatigue Damage by SEM Analysis,"

Wear, Vol 24, 1973, pp 77-95

[23] Goss, G L., Hoeppner, D W., "Normal Load Effects in Fretting Fatigue of Titanium and Alumi- num Alloys," Wear, Vol 27, 1974, pp 153-159

[24] Hoeppner, D W., "Comments on 'Initiation and Propagation of Fretting Fatigue Cracks'" (letter

to the editor), Wear, Vol 43, 1977, pp 267-270

[25] Hoeppner, D W., "Environmental Effects in Fretting Fatigue," Fretting Fatigue, R B Waterhouse, ed., Applied Science Publishers, Ltd., Essex, England, 1981, pp 143-158

[26] Hoeppner, D W., "Material/Structure Degradation Due to Fretting and Fretting-lnitiated Fatigue,"

Canadian Aeronautics and Space Journal, (Third Quarter, 1981 ), Vol 27, No 3, pp 213-221

[27] Hoeppner, D W., Gates, F L., "Fretting Fatigue Considerations in Engineering Design," Wear,

Trang 34

[31] Hoeppner, D W., "Parameters that lnput to Application of Damage Tolerance Concepts to Critical Engine Components," in NATO-AGARD CP-393, Conference on Damage Tolerance Concepts for Critical Engine Components, NATO-AGARD, 1985

Based Fretting Fatigue Studies with Eleetrohydraulic Closed Loop Servo-Control of Axial Load, Normal Load and Slip Amplitude," masters thesis, University of Toronto, Department of Mechan- ical Engineering, Toronto, Ontario, Canada, 1982

Fatigue Process," Specialist Meeting on Corrosion Fatigue NATO-AGARD, 52nd Meeting of the Structures and Materials Panel, 04:05-10, Cesme, Turkey, 1981

[34] Edwards• P R.• ``The App•icati•n •f Fracture Mechanics t• Predicting Fretting Fatigue••• in Fretting-

Trang 35

Leo Vincent, 1 Yves Berthier, 2 a n d Maurice Godet 2

Testing Methods in Fretting Fatigue"

A Critical Appraisal

A Critical Appraisal," Standardization of Fretting Fatigue Test Methods" and Equipment, ASTM

rials, Philadelphia, 1992, pp 33-48

ABSTRACI': Fatigue damage in fretting after crack formation is only initiated under very spe-

cific conditions of both amplitude and load Experience shows, that for a given load the velocity

difference between the core of the rubbing specimen and the friction pads is accommodated as

amplitude increases through the three following mechanisms:

1 Elastic displacements in homogeneous rubbing solids (or first bodies) is a non-destructive

process which will not alter specimen life

2 Elastic displacements in cracked rubbing solids (or first bodies) can lead to fatigue failure

This mechanism is dangerous as it can lead to failure and thus to life reduction

3 Shear in debris beds (third-bodies), formed by particles detached from first-bodies In most

instances, these beds protect the rubbing surfaces and prevent or at least retard crack for-

mation and thus fatigue

The significant drop in life observed experimentally in modified fatigue machines is noted only

because the conditions which lead to Mechanism 2 above are often favored in this type of device

An original static fretting fatigue approach is proposed

ment, cracking, velocity accommodation mechanism, fretting maps

Fretting fatigue is c o m m o n l y encountered in quasi-static loaded assemblies and has been studied widely It is one of the forms of surface damage which significantly limits machine element's life Waterhouse [ 1,2] distinguishes between fretting wear and fretting fatigue Rela- tions between these two types of damage are best approached through "interface" or "third- body" tribology [3,4] Many parameters govern fretting [5] and laws drawn from non-intrinsic test methods invoke corrosion effects, contact temperatures, etc [4] Extensive bibliographic surveys have already been published and only general approaches are discussed here

Stress analyses are discussed in the classical works quoted in reference [6] Chivers and Gor- delier [ 7] described fretting fatigue as the superposition of surface stresses due to the contact and to the bulk fatigue created by the external loadings Similarly, Nishioka and Hirakawa [8] described stress fields in which the friction coefficient varies during the first cycles These two analyses do not consider the slip induced material degradations except through the modification of friction coefficient values

Failure through fatigue crack propagation only occurs if one of the surface cracks reaches a Professor, Drpartement Matrriaux-M~canique Physique, CNRS URA 447, Ecole Centrale de Lyon, Collongue, BP 163, 69131 ECULLY Cedex, France

2 Charg6 de Recherche and Professor, respectively, Laboratoire de M~canique des Contacts, CNRS URA 856, 69621 VILLEURBANNE Cedex, France

Trang 36

critical length related to the contact area [9-12] Two propagation stages were identified First,

crack propagation is influenced by both externally applied load and stresses due to the contact,

then by the externally applied loads only when the crack has reached a given length Propa-

gation rates were measured as a function of displacement, contact loads [13-16] Short cracks

propagated under high velocity rates, even for stress intensity factors inferior to the threshold

value, and closing effects were used to describe specific behaviors and possible increases in

fatigue limits for high contact pressure levels [13]

Few studies listed surface degradations which act as initiation sites Fretting fatigue tests

were run on aluminum or titanium alloys to analyze the degraded area from which cracks

nucleate [ 17,18] Debris formed in these areas explained the lack of decrease in fretting fatigue

limits [ 19], brought about by fretting

Fretting maps serve today to determine the actual fretting regime and to identify contact

kinematics conditions (adhesion, partial slip, gross slip) [20, 21] Partial slip clearly appears as

the most detrimental mode for crack initiation Critical values of the slip amplitude were

found for which fretting fatigue lifetime is minimal [14] Designers need data on both fatigue

limit decrease, due to fretting, and scatter in life, needed in "safe crack growth" approaches,

to give estimations for time between revisions The methodology proposed here is different

from that used in classical fretting fatigue testers It was developed around work performed

near Lyon (France) during the last ten years by the laboratoire de "Mrcanique des Contacts"

de l'Institut National des Sciences Appliqures (INSA) and by the Laboratoire "Matrriaux-

Mrcanique Physique" de l'Ecole Centrale de Lyon (ECL)

Fretting Wear

Fretting Wear Tests

The results discussed were obtained in tests run on several materials including iron, titanium

or aluminum based alloys with different microstructures Tests were run on a modified fatigue

tester described elsewhere [22] For each cycle, the tangential force (F) is recorded as a func-

tion o f the imposed displacement (D) Both are plotted versus the number of cycles (N) in a

3D "tangential force F~ displacement D~ number of cycles N" graph named "friction log."

Here, the maximum displacement D is naturally twice the fretting amplitude (a) Third-body

action or crack formation are illustrated using these friction logs Friction logs, which are char-

acteristic of debris formation and third-body action (Fig 1), are divided in four:

1 Elimination of the natural pollution surface screens

2 Increase in specimen (or first body) interaction accompanied by an increase in friction

and by the corresponding first body modifications [23]

3 Metallic particle detachment or debris bed formation and gradual transition from a two

to a three-body contact

4 Three-body contacts are characterized by: (a)/t continuous formation and ejection of

debris (steady-state conditions prevail); and (b) the change in morphology and compo-

sition of the debris during their dwell time in the contact Metallic debris oxidize and

form the well known red powder when steels are tested

The transition between two- and three-body contacts, which can be dangerous from the

crack formation point of view, is discussed elsewhere [3] Figure 2 gives a classical friction log

for tests in which the cracking mode prevails Three parts are noted:

1 Elimination of natural screens, as described above

2 Increase in the tangential load which corresponds to no or partial slip conditions FD

Trang 37

VINCENT ET AL ON TESTING METHODS 35

FIG 2 Representative friction log when cracks are present

cycles are closed or very slightly open (elliptic) During this stage, cracks initiate mainly

at the edges of the contact

3 FD cycles open further This corresponds to an accommodation of the displacement by

crack opening and localized slip

Note that the friction log is made out of individual FD cycles which can take on different

shapes (Fig 3):

1 Closed (cc) F D cycle, characteristic of a non-dissipative process associated with purely

elastic accommodation found in the "stick" zone Machine and tangential contact stiff-

ness is given, by the slope of the FD line

2 Elliptic (ec) F D cycle, characteristic of a slightly dissipative process and generated in con-

tacts in which partial slip is found or in cracked configuration with interfacial crack fric-

tion Depending on operating conditions, the area of the ellipse varies in size

3 Trapezoidal (tc) cycle, characteristic of gross slip, the near horizontal segments are dis-

sipative, the near vertical segments correspond to the elastic displacements noted above

which are always present

Trang 38

FIG 3 Friction~displacement loops in fretting

Velocity Accommodation Sites and Modes

Figure 4 shows that a three-body contact [24] can be broken down into five basic elements: the two rubbing solids or first-bodies (FB 1 and FB2); the interface or third-body bulk (TBb); and the two screens (TBs 1 and TBs2), that separate that bulk from the first-bodies The screens and interface bulk form the third-body These five basic elements are known as sites and are numbered S~ to $5 The difference in velocity between points A and B is thus accommodated along line AB However, the velocity distribution between A and B is unknown outside of thick film or hydrodynamic lubrication

If, as seen above, the velocity is accommodated at different sites, it is also accommodated according to different modes Visualization has shown that accommodation can take place within any o f the five sites and, accordingly, to any o f the four modes labeled, respectively, M~

to M4, and which correspond to the elastic, rupture, shear, and rolling modes

Velocity Accommodation Mechanisms

All velocity a c c o m m o d a t i o n mechanisms (VAM) combine a site and a mode, and are labeled S, Mj In hydrodynamic lubrication, for instance, accommodation occurs across the bulk o f the third-body or site $3 and the fluid is sheared according to mode M3 The velocity

a c c o m m o d a t i o n mechanism is unique and identified as $3 M3

The combination o f five sites and four modes leads to 20 VAMs which are presented briefly below As first-bodies S~ and $5, and screens $2 and $4 can be inverted, only 12 mechanisms need be illustrated However, in real situations the two first-bodies and screens can be different, and the full 20 mechanisms must be examined Visualization techniques are used to identify both velocity a c c o m m o d a t i o n sites and modes

FIG 4 - - Velocity accommodation sites and modes

Trang 39

VINCENT ET AL ON TESTING METHODS 37

Visualization tests are run with either transparent (glass or sapphire) or standard specimens Observations are performed through the specimens when transparent and along the side Microscopes are used where necessary and video films are taken A very rigid test device was built for the purpose The magnitude of the displacements which can be accommodated varies with the effective velocity accommodation mechanism

Sites S1 and $5

Machine elements, test specimens and supports deform elastically (St M~); this VAM is commonly found in fretting, particularly under small to medium (20 to 50 ~m) amplitudes Elastic deformations accommodate larger amplitudes when, due to high contact stresses, cracks (S~ M2) are formed in the first-bodies (Fig 5) This is also fairly c o m m o n and was observed with metals [25] ceramics [26] polymers [27] and other materials Significant displacements are commonly accommodated through plastic shear of superficial layers (S~ M3) (Fig 6) and exceptionally [28] small displacements are accommodated (Fig 7) through first-body bulk roll formation (St M4) Thus, all four modes of velocity accommodation are observed in sites S~ and $5

Sites $2 and $4

Screens are so thin (10 -9 m) that it is difficult to visualize the accommodation (Fig 8) which occurs within them However, such screens [29] are elastic ($2 Mr), tear [30] ($2 M2), are believed to shear (Sz M3) [31], and rolls ($2 M4) have been observed [32] All four modes exist

in sites $2 and S,

FIG 5 Sj M2 accommodation (through cracks)

Trang 40

FIG 6 S~ Ms accommodation (through first-body shear)

Site $3

Third-bodies are thicker than screens (10 6 m) and their behavior is easier to observe [33]

They are elastic ($3 M0, tear ($3 M2) (Fig 9a), shear ($3 M3) (Fig 9b), and form rolls ($3 M,) (Fig 9c) Here, roll formation is c o m m o n and has been observed with ceramics (Fig 9c), polymers, elastomers, solid lubricants, etc The roll formation process is always the same The nat-

FIG 7 $1 M4 accommodation (through first-body roll formation)

Tiêu đề	Standardization of fretting fatigue test methods and equipment
Tác giả	M. Helmi Attia, R. B. Waterhouse
Trường học	University of Washington
Chuyên ngành	Materials Science
Thể loại	Special Technical Publication
Năm xuất bản	1992
Thành phố	Philadelphia

Định dạng
Số trang	275
Dung lượng	5,87 MB

Astm stp 1159 1992

106 FRETTING FATIGUE TEST METHODS AND EQUIPMENT

112 FRETTING FATIGUE TEST METHODS AND EQUIPMENT