T'- AD-A140 812 MULTITECHNIQUE STUDIES ON FRETTING FATIGUE: INFLUENCE I/IOF SURFACE TREATMENTU) BUNDESANSTALT FUERMATERIALPRUEFUNG BERLIN (GERMANY F R) H SANDER JAN 84

UNCLASSIFIED DAJA4583-C-019 F/G 11/6 L

EL

11.25 =4 0E1.6 1

MICROCOPY RESOLUTION TEST CHARTI N~TWAL 6UREAU OF STANDARDS- 1963-AI

1 -

-Mil

AD R+D A~4~NIr

* Multitechnique Studies on Fretting Fatigue:

. Influence of Surface Treatment

0

Annual Technical Report

by

0 Harro SanderJanuary 1964

United States Army

EUROPEAN RESEARCH OFFICE OF THE U.S. ARMY

London England

Contract Number DAJA 45-83-C-0019

QTICLmjm MA 0 41984.J

* Bundesanstalt fOr Materialpr3fung Berlin (SAM)

"* Approved for Public Release: distribution unlimited

84 05 03 005

. ' , ,

SECURITY CLASSIFICATION4 OF THIS PAGE (Wm, Dol te)

REPORT DOCUMENTATION PAGE BEFORE COMPLETING FORM1. REPORT NUMBER .GVTACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER

4. TITLE (amld Subtle) S YEO EOTAPRO OEE

MULTITECHNIQUE STUDIES ON FRETTING Annual. Jan.-Dez.1983FATIGUE; INFLUENCE OF SURFACE____________

TREAMENT6. PERFORMING ORO. REPORT NUMBER

7. AUTHOR(.) U. CONTRACT OR GRANT NMNER(o)

Harro Sander DAJA 45-83-C-0019

3. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT. TASK

Bundesanstalt fUr Materialprdfung (BAM) AE&OKMTUURUnter den Eichen 87D - 1000 Berlin 45

11. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE

January 198413. NUMBER OF PAGES

1014. MONITORING AGE1NCY NAME ADDORESS(if different frm Cont,.ifbd Office) 15. SECURITY CLASS. (of *#a ftpwat

unclassifiedI5.. 0ECLASSIFICATION/DOWNGRAOING

SCHEDULE

14. DISTRIBUTION STATEMENT (of tirn Rbport)

17. DISTRIBUTION STATEMENT (of 1h. oabaroet emtmd to Block. It d11fferewt 1mw. Repott)

I*. SUPPLEMENTARY MOTES

19. KEY WORDS (Cmmitinuo m ,.ronr aif. If noeseary .E idmtit? by Nook swiabo)

fretting fatigue, surface treatment, surface roughness,steel, fatigue cracks, grinding, polishing, contact

*~~ ATURAC7 (Cb0104000601 RM Wfmmemp~~b OO*mOTfhe problem of fretting fatigue is becoming more acute as in-creasing demands are placed on materials. As the conditions of

* contacting surfaces are of great influence for the frettingfatigue process, a project has been initiated to investigate theinfluence of mechanical surface treatments on the fretting fa-

* tigue behaviour of steel specimens. One of the initial stepstaken has been the development of an experimental apparatus and -

DD P,.3 W3 ammO. 9No 411 i.Vs IS SSETIE unclassifiedSEGUm VT CLAM04PC ATMOP muI m ts MM A MINV

unclassifiedIMCUPY CLASSFICATION OF THIS PAV o n Ae

cont. of 20.a test technique that allows the fretting and fatigue process tooccur simultaneously. The experimental program and the apparatusneeded for investigating the effect of mechanical surface treat-ments on fretting fatigue is described

Aoo.ession For

NTIS GRA&IDTIC TAB

Unannounoed 00Justificat io

DistributionI

AvailabilitY CodesAvail and/or

iSt Special

unclassifiedginsvin" Ci. gPICATM OF TMi PAe60Wuga 000bu

: @-- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Abstract

The problem of fretting fatigue is becoming more acuteas increasing demands are placed on materials. As theconditions of contacting surfaces are of great influenceon the fretting fatigue process, a project has beeninitiated to investigate the influence of mechanicalsurface treatments on the fretting fatigue behaviour ofsteel specimens. One of the initial steps taken hasbeen the development of an experimental apparatus and atest technique that allows the fretting and fatigue pro-cess to occur simultaneously. The experimental programand the apparatus needed for investigating the effectof mechanical surface treatments on fretting fatigueis described.

Keywords: fretting fatigue, surface treatment, surfaceroughness, steel, fatigue cracks, grinding, poli-shing, contact resistance, magnetic induction.

Table of Contents page

1. Introduction 12. Experimental objectives 23. Experimental approach 34. Specimens 45. Test apparatus 56. Control of experiments and data 8

acquisition" 7. Conclusion 9

List of Illustrations

Fig. 1 - Schematic diagram of fretting 3Ffatigue test setupFig. 2 - Test specimen 4Fig. 3 - Fretting pad 5Fig. 4 - Schematic diagram of test 7

apparatus

q

. . .... " ... .. . .. lllir.. ... .. .lir' i i'r ... ., il-1 -.. . .. ..• = ",-.n-. W .... .

1. Introduction

Fretting occurs where there is repeated movement betweenmetallic surfaces in contact. Its effect on fatigue isto speed up the initiation of fatigue cracks and it canbe extremely damaging. This phenomenon is called fret-ting fatigue. Failures due to fretting fatigue have be-come increasingly numerous. Fretting fatigue in techni-cal systems is recognized as a major source of failurein establishing and maintaining structural integrity. Itfrequently leads to high maintenance and inspectioncosts, increasing loss time and excessive retrofit re-quirements. Thus, durability and safety considerationsrelate to the presence of fretting fatigue. The reasonsfor the increasing incidence of such failures are thegreater demands placed on components, closer tolerances,and the scaling up of existing plant to a point wheremovement at mating surfaces exceeds the critical ampli-tude for fretting.

Though intensive research activity will certainly clearup many problems concerning fretting fatigue in the nearfuture, one practical question which will still be askedby engineers, remains: 'Can fretting fatigue be entirelyeliminated by careful design?'The opinion of Waterhouse /l/ is, that like fatigue itwill be impossible to avoid fretting fatigue but itseffects can be minimized.

As the conditions of the contacting surfaces has beenrecognized as an important factor in the frettingfatigue process, a research program was started to pro-vide information regarding the effects of surface con-ditions caused by mechanical surface treatments andfretting fatigue behaviour. The primary purpose of thisresearch is to determine the effect of different surface

* treatments of steel specimens on their fatigue and fret-ting fatigue behaviour.

The initial phase of the research investigation was con-sumed in the development of an experimental apparatusfor the study of fretting fatigue. The following sec-tions describe the projected test method and the main

* components and functioning of the experimental apparatus.

/I/ R.B. Waterhouse, Fretting Fatigue, MaterialsScience and Engineering, 25 (1976) 201-296

2. Experimental objectives

The behaviour of two surfaces in contact under load, i.e.whether they behave elastically or plastically, dependsto some extent on the surface roughness and stress state.These surface conditions are changed by surface treat-ments. As mating surfaces of structural parts are trea-ted in a certain way, it is important for service lifewhether the surface treatment decreases or increases theresistance against fretting fatigue initiated damage.In the projected investigation,specimens of steel shouldbe tested in a comparative manner under both uniaxialfatigue loading and uniaxial fatigue loading with super-imposed fretting. Particular attention should be payedto those processes which lead to a nucleation of fatiguecracks, whereas the process of crack propagation isconsidered to be of minor interest because the stateof surface is of less influence on crack propagation.Through variation of parameters of the mechanical surfacetreatment (grinding, polishing) information should begained which are of high interest for the field of prac-tice as they indicate steps to reduce damage by frettingfatigue.

In addition to the standard methods to measure normaland tangential forces, stresses, displacements etc.the effects should be studied by electric, magnetic andthermal measuring technics. It is believed that addi-tional insight could be achieved with regard to thefretting fatigue process from these special measurements.The magnetic and electric measuring methods should beutilized for the nondestructive and continuous detectionof the damaging processes. Thus the given possibility ofthe early detection of a stress induced damage of ma-

* * terial should permit the experimental detection of theelementary processes which leads to fretting fatigue da-mage by further methods of investigation. As the beha-viour of two surfaces in contact under load depends tosome extent on the surface topography or the surfaceroughness, the surface topography and its change causedby cyclic stressing should be investigated by means oflight and scanning electron microscope and roughnessmeasurements. Determinations of surface hardness accord-ing to different numbers of loading cycles informabout the state of strain hardening.

-- 2 -

3. Experimental approach

To investigate the influence of surface treatments onfretting fatigue, the fretting fatigue conditions occur-ringin practice were simulated in laboratory testing bymodel tests. This method is a more economical way to besuccessful than the method of cost and time expensivedirect tests (test of practice). The model test is con-centrated on the phenomenon 'fretting fatigue'. Thusthe elementary damaging processes should be directlyaccessible by experiment.

normal load

fretting pad fretted interface

test/ oscillating

specimen bearing fatigue loadsupport

Fig. 1 - Schematic diagram of fretting fatigue testsetup

Figure 1 shows diagramatically a typical fretting fa-tigue situation, in this case the test setup used forthe experiments. This configuration simulates a struc-tural joint in which members represented by the testspecimen and the fretting pad are attached by mechani-cal fasteners. The normal load gives a contact pressureat the fretted interface representing what wouldbe created by the fasteners. In such a situation thefretting accellerates fatigue crack initiation and alsothe early stages of crack growth.

-- 3 -

4. Specimens

A basic objective of the experimental work is to usematerials and test conditions similar to those found inpractice. Thus ferromagnetic steel should be selectedfor both the axially loaded test specimens and the fret-ting pads which touch the flat specimen on one side.The practical system which should be simulated by modeltests are the systems gear flange/shaft and bearingrace/shaft. Thus it is chosen as material for specimenthe shaft steel German Standard W.No.1.6582 which hasno AISI equivalent. The gear steel German StandardW.No.1.5752 (AISI 9314) and the bearing steel GermanStandard W.No.1.3505 (AISI 52 100) are selected as ma-terial for fretting pads.

The heat treatment of the materials should be the sameas in practice. The contact surfaces of specimens andpads should be polished so that pads and specimens willhave the same surface roughness.

, , 150 v

Fig. 2 - Test specimen. 5 mm thick. (dimensions in mm)

The temporary specimen geometry is shown in Fig. 2, thefretting pad in Fig. 3. The cylindrical fretting padtouches the specimen with its shell along a line contact.Comparative tests without fretting should be also con-ducted on specimens of the same design.

- 4 -

fI

B 5

A3

C'.

contact slipsurface direction

Fig. 3 - Fretting pad (dimensions in mm)

5. Test apparatus

The fretting fatigue machine was developed to meet the

following requirements:

1. The fatigue load would be applied axially.

2. The mean load, alternating load, normal load,relative surface displacements, friction load,and frequency would be controllable and moni-tored throughout each test.

-- 5 -

It was decided that the basic fatigue unit should be an* electrohydraulic servocontrolled unit to provide the

greatest versatility and sensitivity. Consequently, aservo hydraulic testing machine under closed loop con-trol is used to apply axial loads to the specimen. Thepeak load of the testing machine is 40 kN. The frequen-cy of the push pull loading ranges from 0 Hz to 150 Hz.The load unit assembly consists of A two-column loadframe, a servo-hydraulic actuator, and a strain gageload cell. The strain gage load cell is mounted on thecrosshead to sense load and produces anelectrical sig-nal for readout and switchable feedback in the servocontrol loop. Hydraulically actuated grips provideself-aligning of the specimens with rapid specimen in-sertion. Chucks with insulating layers allow to isolatethe gripped specimen from the testing machine. This is

4, advantageous for the measurement of the electricai con-tact resistance of the fretting system.

In order to induce fretting in addition to the axialfatigue loads applied by the hydraulic actuator afretting arrangement was designed which is shown inFig. 4.This arrangement is based on the bridge principlewith one foot of the bridge anchored on the upper spe-cimen grip so that all the movement occurs at the otherfoot. The unfixed lower end of the bridge carries thecylindrical fretting pad, which touches the midsectionof the specimens frontside. The midsection of the spe-cimen is supported from the backside by a PTFE block toprevent bending and fretting on the backside of the fa-tigue specimen. The clamping load is applied to the

r bridge at its end by a proving ring (not described inFig. 4). The friction parameters - normal and tangentialforce and the wear rate -were measured by the straingauge technique. The slip amplitude is changeable byvariation of the length of the fretting bridge by meansof distance rings mounted between the ring fastener andthe grip on the fixed side of the bridge. A directmeasurement of slip amplitude is provided for by capaci-tance gauge technique.

It is possible to insert a coil system into the frettingdevice for the nondestructive detection of damage(cracks) in an early stage by means of magnetic induction.

- 6 -

X view

2 -5 4

Fig. 4 - Schematic diagram of test apparatus1 specimen, 2 fretting pad, 3 support bearing,4 guide springs with strain gauges for measure-ment of linear abrasion, 5 guide springs withstrain gauges for measurement of tangentialforce, 6 hydraulically actuated grips of theservo hydraulic testing machine, 7 ringfastener for the fretting device

-- 7 -

6. Control of experiments and data acquisition

The process of damage due to fretting fatigue is influen-ced by a large number of variables. To investigate thedamaging processes by experiments numerous adjustments oftest parameters and single measurements of variousvariables are necessary. This task of measurement isperformed most effectively by a largely automatic con-trol of the experimental arrangements by computer. Someof the benefits that can be provided by computerizedtesting are:

- Improved accuracyThe repeatability of the tests isincreased because the test is con-trolled by a computer program andnot by an operator, thus eliminatingmany human errors.

- Improved efficiencyLess time is required for test moni-toring and data reduction. Fewer testsare required, for they are run rightthe first time and the significantdata are properly recorded.

- More versatilityInstead of separate equipment beingrequired for specific tasks the compu-terized system can be used. Whentesting requirements change, only thecomputer program is to change, and notthe hardware.

- Increased capabilityWith a computer in the system, theability is given to complete teststhat may be impossible to performwith reasonable accuracy in any otherway.

Considering these facts the control of experiments anddata acquisition is mastered by an automatic data acqui-sition system. The instrumentation side of the system

1~ provides a multiplexing scanner. Analog inputs switchedby the scanner are connected to a voltmeter for measure-ment and conversion to a digital signal. Two voltmeters

- 8 -

are provided for in the system to cover a wide range ofmeasurement, speed, accuracy, and resolution. The centralcontrol unit is the programmable desk top computerHP 9825 T. This computer controls the experiments and

measurements being made, processes the acquired data,and outputs the results in a suitable format. A thermalIprinter and a graphics plotter belong to the peripheralside of the data acquisition system.

7. Conclusion

The first year of the projected investigations was con-sumed to design an experimental arrangement suitable forthe study of fretting fatigue. A tribological loadingsystem with measuring device to determine the mechanicalstandard parameters was developed. Computer programswere made for fretting fatigue tests to measure the me-chanical parameters characterizing the pad-specimensystem after predetermined numbers of loading cycles.Preliminary tests run satisfactory. The next steps areto insert the possibility of crack indication by the mag-netic induction method and the measurement of the elec-trical resistance into the experimental program. To getinformation about the thermal conditions of the inter-face region the application of infrared microscope andthermocouple elements should be proved.

Acknowledgments

The author acknowledges the continuing support of theEuropean Research Office of the U.S. Army (ERO) and ex-presses appreciation to Dr. Robert C. Bill of the LewisResearch Center for his interest and advices. Thanks arealso given to Milka Javor for her work due to solve ma-terial problems and Lothar Scherpinski for his assistancein constructing the entire system.

-9-