_STP_91-A-1963.pdf

FftOFERTY OF1017AftTES 8C1ENTIA VBRITASTheuniversity ofMichiganLiabraries A GUIDE FORFATIGUE TESTING AND THESTATISTICAL ANALYSIS OFFATIGUE DATAP repared byCOMMITTEE-9 ON FATIGUEAMER ICAN S O CIETY FOR T ES T ING AND M ATERIALSRe g.U.S.Pat.Off.ASTM Speci al Techni cal No.91-A*(Second Edi ti onPrice:$5.00;to Member s:$4.00Pu blish ed by t h eAMER ICAN SO C I E T Y FOR T EST I NG AN D MATERIALS1916 Ra ce St reet,Ph iladelphia 3,Pa.1963 BY A MERICA N SO C I ETY FOR T EST I NG AND MATERIALS 1963Library of C o ngre s s C atal o g Card Nu m ber:63-163 31Pr i nt ed in Bal t imore,Md.Febr ua ry,19 64 FOREWORDThe Fi rst Edition of t his Guide was the composi te w ork of m a n ypeopl e w h o con tributed a great deal of time to the di sc uss i o n a nd wr it-i ng of the tex t un der the g ui dan ce of T ask G rou p Lea der,F.B.Stul en.A m a jor portion of t h e statistica l s ecti on wa s wr it t en by M i ss Mary N.Torrey.Georg e R.G oh n not only contribu ted to the dis c uss i o n and plan-nin g,but al so edi ted a n d a rra n ged for th e printing of th e advance copi esof the tex t.The coordi nati on of con tribution s and di scus si ons was done byH.N.Cummin g s.A ppreciable contribu tion s to th e stati stical parts of th eGuide w ere al s o m a de by D.H.Shaffer.In addi ti on to th e above,R.E.Peterson,H.F.Dodg e,D.P.Ga v er,R.H ooke,W.T.La n k for d,R.B.Murphy,W.C.Schulte,P.R.T oolin,and M.B.Wi l k con tributed to th edi scussi ons at vari ous conferen ces.The ori g i nal T ask Gr ou p was org ani zed und er the l eadershi p of J.T.Ra n som,and a fi rst rough dr aft was prepared i n 1954 and re vis e d i n 195 5.Ot h er contribu t ors to these d rafts w ere E.W.E l l i s,W.T.Lank ford,F.A.McCl i ntock,R.E.Pe te rs o n,E.H.Schuette,F.B.Stule n,and E.J.W ard.I n 1956,F.B.Stul en became L eader of the T as k G roup and the G ui de wascompl eted und er h is direction.Upon the for m a t ion of Subcommittee V I on the Stati sti cal Aspects ofFati g ue,th is subcommi ttee w as asked to revi ew the First E di ti on an d tomake a n y revi si ons necessary to brin g th e Guide up to d at e.A s a resultof t his st u dy,ex tensiv e revi si ons have been made i n various se c ti ons aspr in t ed i n this Se cond E dition.They in clu de:(1)rev i si ons i n the defin ition s(Secti on II)and their sepa ra te publicati on as ASTM Tentativ e Defin itionsE 2 0 6,1(2)an expansi on of Secti on I V on the n u m ber of test speci mens,(3)chang es i n Section V on tests of sig nifica n ce,and (4)the prepa ra tionof a new secti on,A ppendi x IV,on the use of the Wei bul l dist ribu tion func-tion for fat igue M e.T hi s w ork w as carri e d out by four T ask Groups headed by S.M.Ma rco,H.E.Fra nkel,M i ss M.N.Torrey,a nd C.A.Moy er,respectiv el y.Oth ersw h o assi sted in th e preparation of th e Second Edition w ere W.N.Fi ndl ey,R.A.H ell er,J.H.K.Ka o,H.N.Cummings,W.S.Hyler,B.Ruley,andG.R.Gohn,Cha irm a n of Subcommi ttee VI.1 Defin ition s of Terms Rela tin g t o Fa t igu eg a nd th e Stat ist ica l A nalysis of Fa t igu e Dat a(E 206),1962 Suppl ement to Boo k of AST M Standards,Par t 3.iii NO TE.T he So cie ty i s not resp onsible,as a bo dy,for the statem en tsand op inio ns advanc ed in thi s p ub li catio n.CONTENTSPAGEI.Purposes of Fa tigu e T e sti ng 1II.Definition s,Symbol s,and Abbrev ia tion s 2III.Test Procedures 8I V.Minimum Nu mber o f Test Sp eci mens and Their Se le c tio n 16V.Analysi s of Fatig ue Da ta 22A ppendicesM is ce ll aneous Referen ce T ab les 55A dditional T e c hni que for Dist ribu tion Shap e Not As sumed 68Anal ys is o f Correlation Between Two Vari ab les 69The Weibull Distribu tion Fu n ct ion for Fatigue Life 71References 78I ndex 81LIST OF TABLESTABL E PA GE1.Allo catio n o f Te s t Spe c ime ns fo r Probit Method o f Te s t 112.Minim um Number of Sp e c ime ns Nee ded for Determ in ing 95 Per Cent C o n-fidence In t erv a ls of Stated W idt h for a Population Mean,p 193.Min im um Num ber of Sp ecime ns N eeded for Determ inin g 95 Per Cen t C on-fidence In terva ls of Stated W idth for a Populati on Standard Deviation,a 194.Min im u m Nu mber of Speci me ns N eeded to Detect if the Stan dard Dev ia tionof a Population I s a Stated Percentag e of a Fi xe d Val ue 205.M inim um Number of Spec i me ns Nee de d in Each Samp l e to Detect if a Stand-ard Dev ia tion of One Popu lation I s a Stated M ultiple of the Sta ndard De vi a-ti o n o f Anothe r P o p ul atio n 2 06.M inim um Nu mber of Spec i me ns Ne e de d to Detect a Stated Differen ce Bet weena M ean an d a Fi xe d Va lu e 217.Minim u m Num ber of Sp e ci me ns Ne e ded to Detect a Stated Differen ce Betw eenthe Means of T w o Popula tions 218.M edi an P erc entage of Su rv iv ors for the Population 249.Con fidence Interv als for the M edian 2610.Approxi mate Con fiden ce Interv a ls for th e M e an 2711.Confi dence In t erv a ls for Pe rcentages 2812.Fati g ue Test Da ta 2913.F atigue T e st Da t a 3114.Pe rcentages Su rv iv ing 108 C