LEE:FATIGUETESTINGANDANALYSIS FinalProof 21.6.2004 6:19pm page i Fatigue Testing and Analysis (Theory and Practice) This page intentionally left blank LEE:FATIGUETESTINGANDANALYSIS FinalProof 21.6.2004 6:19pm page iii Fatigue Testing and Analysis (Theory and Practice) Yung-Li Lee DaimlerChrysler Jwo Pan Universityof Michigan Richard B. Hathaway WesternMichigan University Mark E. Barkey University of Alabama . . . . AMSTERD.AM BO.STON HEIDE.LBERG LONDON. NEW YORK OXFORD PARIS SAN DIEGO . SAN FRANCISCO SINGAPORE SYDNEY TOKYO LEE:FATIGUETESTINGANDANALYSIS FinalProof 21.6.2004 6:19pm page iv ElsevierButterworth–Heinemann 200WheelerRoad,Burlington,MA01803,USA LinacreHouse,JordanHill,OxfordOX28DP,UK Copyright(cid:1)2005,ElsevierInc.Allrightsreserved. Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,ortransmittedin anyformorbyanymeans,electronic,mechanical,photocopying,recording,orotherwise, withoutthepriorwrittenpermissionofthepublisher. PermissionsmaybesoughtdirectlyfromElsevier’sScience&TechnologyRights DepartmentinOxford,UK:phone:(þ44)1865843830,fax:(þ44)1865853333, e-mail:[email protected] lineviatheElsevierhomepage(http://www.elsevier.com),byselecting‘‘Customer Support’’andthen‘‘ObtainingPermissions.’’ Recognizingtheimportanceofpreservingwhathasbeenwritten,Elsevierprintsitsbookson acid-freepaperwheneverpossible. LibraryofCongressCataloging-in-PublicationData ISBN0-7506-7719-8 BritishLibraryCataloguing-in-PublicationData AcataloguerecordforthisbookisavailablefromtheBritishLibrary. (Applicationsubmitted) ForinformationonallButterworth–Heinemannpublications visitourWebsiteatwww.bh.com 04 05 06 07 08 09 10 10 9 8 7 6 5 4 3 2 1 PrintedintheUnitedStatesofAmerica LEE:FATIGUETESTINGANDANALYSIS FinalProof 21.6.2004 6:19pm page v Thank our families for their support and patience. To my parents and my wife Pai-Jen. - Yung-Li Lee To my Mom and my wife Michelle. - Jwo Pan To my wife Barbara. - Richard B. Hathaway To my wife, Tammy, and our daughters Lauren and Anna. - Mark E. Barkey This page intentionally left blank LEE:FATIGUETESTINGANDANALYSIS FinalProof 21.6.2004 6:19pm page vii Preface Over the past 20 years there has been a heightened interest in improving quality,productivity,andreliabilityofmanufacturedproductsintheground vehicleindustryduetoglobalcompetitionandhighercustomerdemandsfor safety, durability and reliability of the products. As a result, these products mustbedesignedandtestedforsufficientfatigueresistanceoveralargerange of product populations so that the scatters of the product strength and loadinghave to be quantified foranyreliability analysis. Therehavebeencontinuingeffortsindevelopingtheanalysistechniques for those who are responsible for product reliability and product design. The purpose of this book is to present the latest, proven techniques for fatigue dataacquisition,dataanalysis,testplanningandpractice.Morespecifically, itcoversthemostcomprehensivemethodstocapturethecomponentload,to characterizethescatterofproductfatigueresistanceandloading,toperform the fatigue damage assessment of a product, and to develop an accelerated life test plan forreliability target demonstration. Theauthorshavedesignedthisbooktobeausefulguidelineandreferenceto thepracticingprofessionalengineersaswellastostudentsinuniversitieswho areworkinginfatiguetestinganddesignprojects.Wehaveplacedaprimary focus on an extensive coverage of statistical data analyses, concepts, methods,practices, and interpretation. The material in this book is based on our interaction with engineers and statisticiansintheindustryaswellasbasedonthecoursesonfatiguetesting vii LEE:FATIGUETESTINGANDANALYSIS FinalProof 21.6.2004 6:19pm page viii viii Preface andanalysisthatweretaughtatOaklandUniversity,UniversityofMichigan, Western Michigan University and University of Alabama. Five major con- tributorsfromseveralcompaniesanduniversitieswerealsoinvitedtohelpus enhance the completeness of this book. The name and affiliation of the authors areidentified atthe beginning ofeachchapter. There are ten chapters in this book. A brief description of these chapters is given in the following. Chapter 1 (Transducers and Data Acquisition) is first presented to address the importance of sufficient knowledge of service loads/stresses and how to measuretheseloads/stresses.Theserviceloadshavesignificanteffectsonthe resultsoffatigueanalysesandthereforeaccuratemeasurementsoftheactual service loads are necessary. A large portion of the chapter is focused on the strain gage as a transducer of the accurate measurement of the strain/stress, which is the most significant predictor of fatigue life analyses. A variety of methodstoidentifythehighstressareasandhencethestraingageplacement inthetestpartarealsopresented.Measurementfortemperature,numberof temperaturecyclesperunittime,andrateoftemperatureriseisincluded.The inclusion is to draw the attention to the fact that fatigue life prediction is based on both the number of cycles at a given stress level during the service life and the service environments. The basic data acquisition and analysis techniques arealso presented. InChapter2(FatigueDamageTheories),wedescribephysicalfatiguemech- anisms of products under cyclic mechanical loading conditions, models to describe the mechanical fatigue damages, and postulations and practical implementations of these commonly used damage rules. The relations of crack initiation and crack propagation to final fracture are discussed in this chapter. In Chapter 3 (Cycle Counting Techniques), we cover various cycle counting methods used to reduce a complicated loading time history into a series of simple constant amplitude loads that can be associated with fatigue damage. Moreover the technique to reconstruct a load time history with the equivalent damage from a given cycle counting matrix is introduced in this chapter. InChapter4(Stress-BasedFatigueAnalysisandDesign),wereviewmethods of determining statistical fatigue properties and methods of estimating the fatigue resistance curve based on the definition of nominal stress amplitude. These methods have been widely used in the high cycle fatigue regime for decades and have shown their applicability in predicting fatigue life of notched shafts and tubular components. The emphasis of this chapter is on LEE:FATIGUETESTINGANDANALYSIS FinalProof 21.6.2004 6:19pm page ix Preface ix the applications of these methods to the fatigue design processes in the ground vehicle industry. In Chapter 5 (Strain-Based Fatigue Analysis and Design), we introduce the deterministic and statistical methods for determining the fatigue resistance parametersbasedonadefinitionoflocalstrainamplitude.Otheraccompan- iedtechniquessuchasthelocalstress-strainsimulationandnotchanalysisare also covered. This method has been recommended by the SAE Fatigue Design&EvaluationCommitteeforthelasttwodecadesforitsapplicability in the low and high cycle fatigue regimes. It appears of great value in the application ofnotched plate components. In Chapter 6 (Fracture Mechanics and Fatigue Crack Propagation), the text iswritten inamanner toemphasizethebasic concepts ofstress concen- trationfactor,stressintensityfactorandasymptoticcrack-tipfieldforlinear elastic materials. Stress intensity factor solutions for practical cracked geometries under simple loading conditions are given. Plastic zones and requirements of linear elastic fracture mechanicsarethen discussed. Finally, fatiguecrackpropagationlawsbasedonlinearelasticfracturemechanicsare presented. InChapter7(FatigueofSpotWelds),weaddresssourcesofvariabilityinthe fatiguelifeofspotweldedstructuresandtodescribetechniquesforcalculat- ing the fatigue life of spot-welded structures. The load-life approach, struc- tural stress approach, and fracture mechanics approach are discussed in details. In Chapter 8 (Development of Accelerated Life Test Criteria), we provide methods to account for the scatter of loading spectra for fatigue design and testing. Obtaining the actual long term loading histories via real time measurements appears difficult due to technical and economical reasons. As a consequence, it is important that the field data contain all possible loading events and the results of measurement be properly extrapolated. Rainflow cycle counting matrices have been recently, predominately used for assessing loading variability and cycle extrapolation. The following three main features are covered: (1) cycle extrapolation from short term measurement to longer time spans, (2) quantile cycle extrapolation from median loading spectra to extreme loading, and (3) applications of the extrapolation techniques toaccelerated lifetest criteria. In Chapter 9 (Reliability Demonstration Testing), we present various statis- tical-based test plans for meeting reliability target requirements in the accel- erated life test laboratories. A few fatigue tests under the test load spectra shouldbecarriedouttoensurethattheproductwouldpasslifetestcriteria.