E‘Sé ll“!!!WI“!IHIIUIIWINIUIWIIUNNHIIWHIIHHI tr ”'5 e “a?! on. m— .;=La'—'unn' Michigan $2.22: ‘5 ’ .“ o‘-‘ A"- .5!“ I?! U."Ii This is to certify that the thesis entitled Effect of changing load conditions on Dynamic Recrystallization. presented by Sanjeev Deshpande. has been accepted towards fulfillment of the requirements for D G.Gottstein Majorprofessor 5/16/1986 Date 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU RETURNING MATERIALS: Place in book drop to “33551155 remove this checkout from .—_—-—. your record. FINES will be charged if book is returned after the date stamped below. MN! 31%62692 EFFECT OF CHANGING LOAD CONDITIONS ON DYNAMIC RECRYSTALLIZATION by San jeev Deshpande A Dissertation Submitted to Michigan State University East Lansing, Michigan in partial fulfilment of the requirements for the degree of Master of Science Department of Metallurgy, Mechanics and Materials Science I 986 Eitectdduulmhndcaflitlunmueetdmicmummzetmwasstmm lmmwnstedmm.slmbmmlyuystelmmpolwystaleto.61m(imminent new). Itvethmtmyplaysvay Imprrtmt minim~ the InItIetImotO/nunic Wilhelm. In all "tandem millennium Weathtiusm by sidic mlinget (bimetim tunpu‘eiu‘e thedmunic rw'ystellizetim (DRX) stress (impart WING/tunic mystellizetim was twat in tin cyclic tests. cycling in tinelestic mini Smuhmfloiblea‘ my little Him! in the mystellizetion behavior, but wclino in the platic reoim in both directions melted in (Manic Wilhelm et minutely 25$ stress level Wwith multimic testing m "I WWI. All timidsumiirnmmic rmy emmwmmm iu‘mnic mystelliutim tor polvaystels also. simillr toeimlecrystels AWEELS lite thisupwtmitytoormmysim wetlhuhtomyemisw Prat. acutsteinmitlmt hislelpanrlmithnethisproiectmldmt imbecnem I mldelaoliketotlutmy colleagueMrk Waterbury tar his help In setting.» lmtrunentatkm, wantshellmyirmmtheMMM Wtwtnoheipulmemmmwiflcim their time aid etiofls without my mpleints this project is tuned by Departmmt Of Energy by want In DE—F002-BSEMS2OS Table of Contents LIST OF FIGURES TABLE I: TEST DATA (pm ll-l3) Chapter Title I Introduction 2 Brief Review of Previous Results 3 Experimental Conditions 3. I Specimen Preparation 5. I I Mechanical Testing Specimens 3. I 2 Optical Microscopy Specimens 3.13 Electron Microscopy Specimens 3.2 Mechanical Testing 4 Types of Experiments 5 Results and Discussions 5. I Effect of Static Recovery 5.2 Effect of Strain Path 5.3 The Critical Deformation Parameters 5.4 Welic Deformation 6 Conclusions References Paco 14 15 16 16 I9 20 Figure A Page No. 21 22 23 24 25 26 27 28 29 List of Figures Caption Mechanical Testing Protective Gas Chanber Arrangement. Tensile stress curves of Cu polycrysrtels deformed at 400 °C and a strain rate = 2.5 - l0'4s". 0- continously (bformed, I - deformation interrupted for I.3h at 80 MPa, v - deformation interrupted for 20h at 80 MPa. Strain hardening rate versus diagrams for the curves shown In figure I . Compression hardening curves of Cu polycrystels after a predeformetion in tension to 80 MPa. I - compression parallel to the tensile axis, v - compression perpendicular to the tensile axis. Strain hardening dimrams for curves shown in figure 3. Deformation geometry for the single crystal tests. 0 ( TD) denotes the tensile axis and I , A (001 , C02) represents the compressin axes. The code indicates the activated slip systems. Strain hardening diagrams for curves in figure 7. Resolved shear stress— strain curves for oriented Cu single crystals during compression. I - compression axis parallel to tensile axis, 0 - compression along (I I2> perpendicular to tensile axis, A - compression axis along (I i I) perpendicular to tensile axis. The ratio of recrystallizatim stress to the (extrmoiated 0 Id = 0) steady state flow stress for Cu single and polycrystals as function of the recrystallization stress. Figure 9 l0 ll 12 Page No. 30 31 32 32 33 34 35 Caption a) Tensile hardening curves of Cu polycrystals with ISO cycles in the tensile rang after predeformation to 80 MPa. I - stress amplitude 65 MPa, A - stress amplitude 79 MPa. b) Strain hardening rate of curves in Figure 9(a). Cumulative stress-strain curve of Cu polycrystal with constant I .53 strain amplitude. (Nola frequency 0.3/min. (T=4OO °C). Cumulative stress-strain curve of Ni polycrystal with constant I 3 strain amplituib. (Note frequency OAS/min. (T=600 °C). Optical micrographs of Ni specimen cycled at 600 00. Note the bouandry curvature and newly recrystallized grains at the orignal grain boundary. TEM micrographs of the specimens cycled at 0.5 Tm. a) dislocation arrangement in the vicinity of a grain boundary in Ni. b) dislocation arrangement in the grain interior of Cu. Note the crystallographically sharp bands. CHARTER ONE We It is estdllished that a veriety of ice. materials m (Manic mystallizatim I.e. recrystallize while being inform at temperatwes wove 0.4 Tm, aid fit the sane material dagiventemperaturemdgiven strain ratethedynunic reuystallianimstraaaismnie Mini 1] This mty isvay importalt in all the hot wakingmmwell for components which we used in high temperature implications It is father eetdllidlui that once the critical stressor critical strain (which is mimately I28 for static recrystallization aid mimaiely 22! for Music recrystallizdilll while correspmding values for strws as critical stress value) is crmsul airing uwmm eventually the material will unrbrgl recrystallizatiul. This critical strain is minutely 0.8 (1 ill: (Minnie rerrystellialtion strain chem as till pain simian] Eventlum "wwmsrekmwnmeprmlem whicllrelneimis, ilcllllltle predictedluwaduminthemfa‘mation conditions dfectstheouxralceofwmnic recl‘yslallizditll. Baicallytwomaclmismscmbeaaumadtocontrol flieocclmoiwnllllic myatallizatilll. II Mleaiim 2] Will of ruystallized rains. In single crystals tlere is ammo evirhice that nucleation controls the process [3] but in polyuystatsd/mnic reuystallizotimmaybewowthcontmlledbeceuaethednmic recrystallizatim strms is much Iowa. Sdiai mm havesImn that dynamic recrystallization stress keeps on increasing with m cycle of mystallization, when sitting out with single crystal aid informing ttroudl multiple recrystallizatim wcles The divious rmson fit this Will‘ is that win Mics in polycrystalline materials diact the (Nannie recrystallization such that they with more Iavordlla nucleation Sim inmlyuystalswhichcm 08$“de mmmls. This Billullivelalt W, with the exclusion that dynanic recrystallialtim occlrs wowth cmtrolled in polyuwtals In this investigtion the problem was utt‘essedby invaaiglting theefiect of change in thestrainpothondmunicrecrystallization. Thestrainpathofintuumfmae, isthet prior tothecritical strainer the effectofstreaeeswhichn less that theirItical stress for Municor static mystallzation Materials selected for the tests were polycrwtalline Ni aid Cu, aid single crystals of Cu. Moatofthetestscolmctedsofro‘eollm,becauswweismuchmreferancemtaavailmle fir this matrialald It iseaeiil~ tomsinglecrystalsofm tlmofNi. Alsotheibfa‘mationmd Manic reuystallization behavior ofOu Isvelysimiia' toNiiSl this isbecolaetheratioof stating fault energy to sham moMus aid Burqpr's vector isvely much compudlle, which cll'itrols (Mimic recovery all my the initidion d Manic reuystallizdm CHABIELM MW The mermaid/tunic reaystalllzation tiring high temperdtremlileiion is now well whitewater many metals Malorltyoftha invasilmtionswerecnnllded lodynonlc tests, pa'ticulu‘ly in torsion, tension and comprossiml .5, 6] Tiedlslwationmnsityof ameterialcreataiorlngtnwu'klngclliberelmedby rmywreuysiallizatim.80thtinseprmessesreobservedtoocclr notallylnllrstdic condition, i.e. tiring allieeIingof cold worked couples, butalsounmr (Manic colldltlllls, I.e. dringhigl temperatlraihformdion Once itwasestdlllshedthatdmanic recrystallizatimihmouxlr Ithmbeen ilivestilpted extensively In polyuystalsmdtheresultshmeprovedto beertremly useful In wietyof mpliceiions, pvticulvly in min ref inunentll’l] But then the physical processes which control the (Manic recrystallization were not withstood plmerly; mainlybecausaof tliemplex intoractionsof disllmtloi taigles, mains, higimglewainbolmmsmtc Mefocuswsthangivuitom‘stmdmmismstlmm testson single crystals in ordr to simplify the matters byelimineting tlieg~ain ileum intermitona mummleclystalmilmtsmmtainet alconelpwiththeresultsthatthe mumhreuystallizdimdnslummmworimtiemimm i’lra allsimt Uu‘metim pain, in acne mm, «Willi, strain rdeuuitunpudmfinmic rurystallidim taut oft Willi/at attinitevolmoi flow strm In tinir minute theyobsa‘vedthevaluesofahelrstrainvrledmawimrmbut thesiressremaimlcum. So it was mluthd that dynamic recrystallization is set off at acritical value of flow stress rather than at a critical strain; this was also observed wring present investith In the sane paler they cane up with results that the actual value of the mystallizatim stress mpenm on the material Hid m the (bfwmatim Wilma With increasing temperature ind increasing strain rate the critical flow stress or the mystallizetion stress is m to W. One more important otmrvation airing the investigtion was dynunic recrystallization in single crystals wm found to be trigpred by the Gimmatim illtllced illstdlillties of the main structure The factors which lrise this inddfility can be the corltmlling factors for dyntlnic recrystallization. This kindof study was counted by eottstein old Kockson single crystals of Cu and HITS] Fran the previous statics in Al single crystals it w diservai that it never mystallizes but always recovers, thisstrongrecovery isolatothehigi ratioofstackingfaultemrwtosher modulus and Burger‘s vector. This particular behavior had led to a hypothesis that d/nanic recovery anddyrlllnic recrystallization me two competing processes. but from Unuwiments on CilmdNieottsteinoidKodtscanewwithaconclusion tllatdynonicrecoveryrether theta competing process. is a precondition tooccurenceof dynenic recrystallization. Dynonic recovery of dislocations leads to mmt of cell wallson local scale giving rise amid 71 which being more mobile cm form fluctuatims in the homowneom information structtre old when subg‘ains of Wtial tree reformation trimr dymnic recrystallization. This cllriclusion also helped in Wag lower sires values at hidier temperatures. where formation of subbounthries is easier old they hm hither mobility. OtMr that temperature we more fair which is found to be inillnncing (Manic rwystalllzetion is thestrain rate it wesWthatat tinsunetanpa‘attre lithe deformation is conflicted at higher strain rates. rau'ystallization stress increased i l in "lemma: they havaalsosl'lown thatbydmgingdeformation pathevenwith same strain rate we get different value of recrystallizdion stress Tdting into amount all these experimental results we found it of pa‘ticulfl' interest to find out effect of chums in the information path on (Manic reuystaiiization QflAE! EB THREE Ex crime t Co ' ‘ons 3. ll Specimen Praprltion i‘iatar‘ials - a) Cu Polycrystalline aid single crystal 99.992 pure b) Ni Polycrystalline 99.99! pure 3.l l] HachuIIicai Tasting Specimens All sunples for polycrystalline Cu aid Ni were minhirlei Tran I/2' diunetar rats. Thespecimen Wh was 3" with cylihti‘icai mewtionof l" Wit 8!! 3/l6' Cusampleswereannealedat temperatureof600°Cfor6tmrthenchemically etched to remove mprcximately 0. i mm. surface layer with 502 HMO} Finally they were chemically polished in a solution of emal parts of H3P04. Cii5w0ii aid HN05 to remove any surfwa irregularities Ni sampleswere stress relievedat a temperatureof680 “C, than chemically polished to remove mimateiy 0. 1 mm layer of tin surface with a miutim of 708 i003 mom i 20gns c1304. 3. i2] Optical flicroscapy Specimens Specimenswaracut from thetensileor compression sonplesbyveryslowspeed diunond wheel cutting along the looting direction and subsementiy mounted in cold setting resin before mechanical polishing WmlishingCu molesmetctn‘lihamiutimofwx "03 + 602 013mm Ni sempleewereetchedin 67ml ”0;, + 33ml Cit-5m + lml ill. 3. l3) Electron flicroscepy Specimens Thin slices of wpromimately 0.3 to 0.4 mm thickness were cut from tensile samples byslow spaaddimlondwheel cuttingsuch that the foil plate normal is perpendicula‘ tothe directionof loading ThesewerethinnedrbwntoQi mm with 508% + 508 H20forCumd l00rnl HN05 + ZOgnsCuSO4 for Ni. The final jet polishing was rhne in a Tenupol jet polishing device Electrolytes used wereDZ corrosivet") forCuandABt“) for Ni. (*) 02 andABnregistered trahmrizsof StruersSciantific instrumalts. Inc 3.2] mechanical Testing All tests were waistcoat a floor motel electro-medlanicai lnstron testing machine with a 800 itg terlsim- compression load celi. Pull rodsmdbutton headg‘ipsweredesigledandmechlnedoutofmo heat resistalt stainless steel. To avoid mtitbtion wring testing a cylintricai potective challbar was instilled as shown in the fig A A stainless steel ring with a clwmof 0.75 mm to the pull rod is weithdat theiopofthe tube.whiiethelowerringwhfch fitson immpull rodismarbofinvrrmdthe dimensionsrechosenstulthatat temperatures inexcessof200°00ti¢lt fitwiththeltwrer pull rodwasobtainedmeto thediffarence in thermal expansion between im ringald lower puii rod With thiscillliber MWapf‘otwtiveatlmedWS N2 + lOSszas maintained airing the test inside the charmer to minimize oxitbtim. F low rate of l2 litres/Mr at 5 PSI was found sufficient to avoid oxicbtion. Hither flow rates mavoitiad sincems starts burning at the top of the chamber, also it creates beckpressure which should be avoilbdt The this tnlisitim is carried out by AID conversion by processing the ice] cell simal into a computer [iBl‘i-XTl. This snarled to store all the loui—displwunent this from the testatm interval of LS seconds.withmaccurer.yofg~eater that 0.l$. Thisrhtawasthen further processed to at the relevant information like, stress—strain curves tlld withu‘thning coefficient ctrves. Similtrly. bythedigitai output from thecomputer themevemaltofthecrosdm could be controlled from the computer. For all the tests strain rate was maintained mimicry 2.5 x 10" sec". The temperature selected for all the tests wu approximately 0.5Tm, so for Gr it was 400 i 5°Cmini600+5°Q HAPTER FQUR - *‘0-00 types of Experiments l he experiments an be categorized into the following main amps. 2. Uniaxial tension and compression tests on polycrystalline m. a Tensile deformation to a stress below the critical stress for static or dynanic recrystallization and then holding the specimen for a specific time at the reformation temperature followed by uniexial tension till dynamic recrystallization. b. Simila‘ tests in compression. . Deformation of specimens in tensial to thesane stress vaiueas from 2. removing the specimen, which was then cut to prime: the compression sanpies for: 3. Compression in the same direction as tension. b. Compression at 90° to the tensile direction. . Similar tests as in 3 on Co single crystal. a. Tension followed by compression in the sane axis. b. Tension followed by compression at 90° to the tensile axis but where slip systems remain sane. c. Tension followed by compression at 90° to the talsilea‘ientation where aflitional slip wstems a‘e activated . chliCtbformation a CYcIicaily loading elastically only in tension at different loads followed by deformation in tension to dynanic recrystallization. crystal. i0 b. Loading cyclically in plastic region in tension and compression with fixed strain anpliturh till dynanic recrystallization. This experiment was ctlrltitlcted on polyu'ystaiiine Cu and Ni as well as Cu single TABLE t .n...-. I...-~.—¢._CI--I..I>.--ooo .- . - -ooo—-..-ae..-_ o... n.-.-....--...—-o.o-n -. .. canon—on-.. no“... -u--.- ~w-—.~o--~--- oc- TastNo. ortx Stress“ dDRX/dt‘aaturation Typeof Test CUTENT LPCU2 LPCUS LTPCUZ lTPCU? CPCU3 MP1! 1 56.00 l32.00 ”1.90 l3t.65 l 26.84 l 56.00 0.810 0.850 0.823 0.820 0.850 0.870 Uniaxial tension Loaf cycled between i6 Kg aid 76 Kg in tension. riax. stress 65 "Pa Load cycled betwwn i6 Kg aid 96 Kg in tension. i‘lax. stress 79 l'iPa Pretbformed in tension to 80 "Pa Annealed for 80 mimics at information temperature, reformed in tension till dynanic recrystallization. Pretbformed in tension to 80 ”Pa Annealed for 20 hrs. Deformed in tension till dynanic recrystallization tiniaxiai compression. t2 W n...oee‘u-..oooa.e.n--o..-e—— -— Ice-I An Test No. on; Stress 6 0M! o Saturation Type of Test Pa LCPCU2 l25.80 0.850 Pretkformad in compression to 80 i‘iPa Holding at thformdion temperature for 20 hrs Daform in compression to ORX TCPCUZ i42.40 0.830 Pramtl'mai in tmsim to 80 "Pa foliowod by deformation in compression to rmrystallization. TCPCU‘I i34.30 - Premformed in tension to 80 i'iPa, followed by deformation in compression perpendicular to the tensile axis to raz'ystallization. (MPSCLS i i i .00 — Strain cycling of Cu polycrystal in tension ald canpression with strain anplitulh of 1.58. l3 15W ’4---_ .v—e-ea .- 0“-”x. .-——-.~. ~ ..uma-“ a.-- “’oa-“o-— mo Test No. onx Stress «remix «saitiratim Type of Test MPa NiPSC4 72.00 -- Strain cycling of Ni polycrystal in tension and canpression with strain amplitumof LOX. TCSCUS 52.00 0.83 at single crystal with (i i0> orientation for tensile axis, tbformed to 25 "Pa shaa‘ stress in tension, followed by compression in perpendicular direction in (l 12> orientation. TCSCU6 55.00 0.86 Cu single crystal , pretbformed in tension followed by compression in the sane axis. TCSCU? 38.00 0.86 Cu single crystal, compression axis (i l i>, perpaidicuia' to the mnsiie axis. l4 CHAPTER H VF. Realms and Discussions 5. l] Effect of Static Recovery The harrbning curve of Cu polycrystal, (bformed in uniaxial tension until dynamic recrystallization, is shown in the iigure l. Dynamic recrystallization is indicated by the maximum stress value on the true stress-strain curve. It was shown before that recrystallization mrs snrnewhat prior to the peak stress value on the hardening curve, but the difference in the stress values is only marginal so this pad: stress will be refered to as recrystallization stress in the following lhis is the stress which is shown to be repromcible and the most simiiicmt quantity to indicate dynanic recrystallization. lhe results are shown in the figure I. lhe specimen were predefrrmod in tension to 80 MPa( approximately 0.5 an ), then the test interupted and ariruealed for 80 minutes aid 20 hours. lhen chiermation was continued until recrystallization occured From the y‘aphs it is clea‘ that static recovay which is the effect of annealing results in a slight loweringof the recrystallization stress; the 7.0 hour annealed sample simed l9: disease in the recrystallization stress. AnotMr interesting point to be noted is, the annealing has also affected the strain hrrbning behavior. this is indicated by the strain hmhning rate (dc/dew 0 ) curves in figure 2. For ainaaledsunples therate is lower for agiven valueofstress for aeontinually informed specimen. This indicates a chum in the subg‘ain structure. Such a behavior is not obwwed in polycrystalline Al. Cu or Ni single crystals aid is only reported of Al single crystals where it is attributed to subg‘ain ewssning which is not observed in Cu polycrystal at this ternprature. So it is concluded that the presence of grain bounmries affects the dynamic recovery l5 substantially, an important fact for the initation of dynamic recrystallization, which was shown to be recovery controlled Similrr results were obtained irorn the samples tested in compression tests. 5.2] The Effect of Strain Path in this set of experiments the deformation aid recrystallization behavior associated with the chance in the strain path were studied All specimens were predeformed to 80 MPa (mproximately 0.5 0R) in tension, followed by compression either along an axis parallel to tensile axis or perpendiwlar to the tensile axis. The resulting stress—strain curves are shown in the figure 3. it can be seen that the effmt of compression on dynamic reuystalllzatlm is very much Wt on the direction of compression axis. in the tests where cornpresion is conducted prallal to the tensile axis. there is only slidit wrease in the recrystallization stress because some slip systems are activated as per the Taylor's theory, while when the compression is in the perpendicular direction there is substantial decrease in the recrystallization stress. This has to be attributed to the different (bformation mostly did so a different development of dislocation substructure. A chum in the substructure by operating initially inactive slip systems results in a charm in the recovery rate which is indicated by the strain winning rate curves in fig.” 4 and tiereforeadiffarent reu'ystallization behavior. in orrbr to substantiate the results from these tests, similar set of tests was corimcted on Cu single crystals. The single crystal geometry is shown in the figire 5, where tensile axis was parallel to [ 1 i0]. Compression was cornictad in the axis parallel to tensile axis If) andalong two perpendicular axis ll l l] de l i2] . after predatorming the single crystal in tension. The [ l l l] and [ l l2] axes are known to remain stdile (hiring compression of f.c.c. crystals. The deformation mommy (figure 5) is such that the compression along [ l l2] mtivated the same slip planes as tension along [ l l0] . while compression along [ i i l] activated a slip plane not operated under tension along [ l 10]. As can be sesn from the figure 6 activation of adiitional slip plane results in a very strong Wnamic recovery old which is reflected by substmtial decrease in the recrystallization stress as shown in the figure 7. 5.3] The Critical Deformation Parameters From previous single crystal experiments [5]. it was conclurhd that the onset of dynanic recrystallization is controlled by the dynamic recovery rate which is reflected by a constant ratio of the resolved racrystalization stress 'R to the (hypothetical) steady state flow stress. rR/ 15:083. Since randadiffer onlybytheSchmidfactor,also UR/ as hastobe (Install. This ratio is plotted in fimre 8 as function of mystallization stress. It is seen that irrapactive whether single or polycrystal or whatever the strain history of the specimen, the ratio is imiably 0.84 4 0.03. Hence, the present experiments confirm that dynamic recrystallization is controlled by dynamic recovery in single aid polycrystals. Enhmced recovery by shaming the strain path, therefore, favors the onset of dynunic mystallization, i.e. lowers the mystallization stress. 5.4] Wells Deformation The experiments ascribed in the previous sectims confirm that even a proiormd whaling treatment has only a mild effect on the onset of Wmlc recrystallizatim, hates the repeated unloading periods during cyclic rhiormation should not consimrdny effect the dynamic recrystallization behavior. This was also confirmed by a set of tests where cyclic mformation was conducted in tension after monotonic prerbformation to 80 r‘IPa (approximately 0.5 0 R). (Nelle deformation was performed with constant stress amplitude of 65 MPa, i.e. only in the elastic regime, and 79 MPa, i.e. upto the transient to plastic reformation. Wcling was stopped after ISO cycles. Subserpently, deformation was continued by a monotonic tension until dynanic recrystallization occurad (figure 9). The effect of cycling in the elastic rm is cornpu'able to static wheeling for the same period of time without cyclic deformation Wells tests into the plastic regime were cormcted on Ni aid Cu polycrystals at 0.5 Tm (i.e. 400 °C for Cu aid 600 °C for Ni) with constant strain anpliturb of 1.53 (for Cu) aid 1% (for Ni). The strain rate was some as in the monotonic tests (wproximately l0"4s‘ ‘) corresponding to a cycle frequemy of mproximately of 0.3 or (0.45) min‘ '. The respective stress-strain curves (figures to, l l ) reveal an initial period of strain humping follmed by a steady decrm of flow stress. The cumulative strain (or number of cycles) to the maximum stress (hpanrls on the strain ampliturh For l5}! strain ampliturb (Co) the peat stresswas reached alrarn/ after 4 cycles, while for is ampliturh strain (Ni) the pad: stress is reached only after lOl cycles, but followed by a more (rastic stress morease in subsament wcles The occurence of dynan ic recrystallization was confirmed by metalloa‘whic investimtions (figlras l2, l3). Newly recrystallized gains are observed on the grain boundries of the wormed material, typical for d/nunic recrystallization in polycrystals [ 9). Thecurvatureofthebourmies in "are 12 also indicatethatstrain irxlmdwain bounty migration (SlBl'l) had occurad die to an imbalance of dislocation tensities in uljacent trains. Since the volume of newly formed gains is very small, it may be hypothesized, whether SIB" is actually the major softening process unrbr these conditions. Preliminary TEl'l observations, however, could not confirm this hypothesis. I8 The dislocation structure in these specimen is a well defined cell structure (figures l4. i5). (implied to monotonic tests up to dynunic reclystallization at the same temperature and strain rate, the structure appears more recovered. The average cell size ( linear interwpt) is in theorder of S m, which compass to 1.5 m in monotonic testsat R- This is reasonable when consilbring a stress ratio 0 man! u cyc = 300/74 and the commonly obmrved relationship a. d = constant. it is remarkable thoug'l, that Manic recrystallization mured in a microstructure with a consitbrably less stored energy than in monotmically (Minna! spmimens at the onset of (Manic rmrystallizatim. This result substmtiates that not the stress but dynamic recovery controls the initation of (Manic recrystallization. The relationship between cyclic may, dmamic recovery and microstructure lbvelopment is presently unmr investigation by systematically va'ying the mfcrmation conditions. i] 2] 3] 19 Cl PT‘ X Conclusions The influence of static recovery and the effect of strain path chaos on dynanic recrystallization have been studied. Both conditions are relevant to the interpretation of data in cyclic test. it was found that both static recrystallization ald lcai inversion from tension to compression facilitates dynanlc recrystallization but only sllmtly compasd to changes to strain paths which change the slip mometry. Low fremency cyclic (bfcrmation in tension within the elastic regime has alout the sane effect on dynanic recrystallization as static recovery for the sane period. The initation of dynanic recrystallization is dynamic recovery controlled irrespective of the deformation conditions the ratio a R/ a S is approximately constant. Wclic information with a i 2 strain ampliturb llnlces (Mlanic recrystallization after 100 cycles (T - 0.8 Tm, 0.5 cycles/min.) at 25! of the recrystallization stress for monotonic tests under the same deformation conditions. Recrystallization progresses very slowly, maybe assisted by Strain induced Bounty filtration The dislocation mlcrostructurecorresponds in scale tomcnotonictcststothesanestress. but the structure appears more recovered ll 2) 3) 4) 5) 6) 7) 8) 9) 20 Bflfl'fllfififi ll. i‘lechingand 0. Gottstein: Recrystallizationof Metallic Materials ppi9S i978 RA Petkcvic, r1. 0. Lutonand J.J.Jonas. Actaflet. v27 ppi633-48 i979 0.00ttstein, D.Zabardjaiiald H.l‘1ecking l‘ietalSclenceviS pp223--27 1979 T. Sakai and J.J.Jonas. Acta l'let. v26 i978 6.00ttsteinand V. i“. Kocks. Actaitet v3l ppl75-88 l983 H. J. l‘tcOussnald J. J.Jonas "RecoveryaldRecrystallization wring Him Temperature Deformation" in Treatisecni'laterial Sciermaldtechrlolmy, v6 pp393 l975 T. Hasegawaand V. F. Kooks. Acta metal v27 ppl705 i979 T.Sakaiand J.J.Jonas Actametal v32 pp i89 i984 L. C. Limand R. Raj: Acta Metal v32 pp1183 I984 2] vi UPPER PULL O . O STAiNLCSS STEEL 8:1 :33? o 0 ° W 0 ° ‘ 2 Mill “6 “’8 e. 8 ° areas 0 TN 0 o o o a o o o . a 0 Al 0 C) vv 7 C) O O 8 r ' -« o ' Mt ° l E ~8—mv O ‘ 0 0 I O ’/r': I > LOWER. , PULL GAS Figure A 22 H meamwa 3... 2.45m mac» 3 I hack»... A «30...: I PZNPDO O on omp (VdW) $83818 3081 73 owp :55: mmwmem m3: sneak; A «Dock; I Fzmhao o om m mcamwa _ .......,-_.....l L... ‘9 (vac) VLBHL 24 on u‘ nauaoh A «Donoh I cacao 0 1x; z_