Time-Dependent Evolution of Volume Fraction and Stability of Retained Austenite in a Hot-Rolled and Intercritically Annealed Al-Alloyed Medium-Mn Steel

ORIGINAL RESEARCH ARTICLE Time-Dependen E olu ion o Volume F ac ion and S abili y o Re ained Aus eni e in a Ho -Rolled and In e c i ically Annealed Al-Alloyed Medium-Mn S eel ADAM SKOWRONEK , ADAM GRAJCAR , CARLOS GARCIA-MATEO , JOSE ´A. JIME ´NEZ , and ROUMEN H. PETROV The de elopmen o supe io mechanical p ope ies in medium-Mn equi es he op imiza ion o mic os uc u al pa ame e s such as e ained aus eni e (RA) s abili y, olume ac ion, and mo phology. The p esen wo k explo es he possibili y o using a con inuous annealing app oach ins ead o con en ional ba ch annealing o pe o m an in e c i ical annealing (IA) ea men in a ho - olled s ip o an Al-alloyed 5Mn s eel. Dila ome ic s udies we e pe o med a a empe a u e o 680 ºC wi h soaking imes anging om 1 o 300 min o ollow he mic os uc u al changes as a unc ion o ime. The mic os uc u es hus ob ained we e ho oughly cha ac e ized by means o X- ay diff ac ion, SEM and TEM, TEM-EDS mic oanalysis and EBSD phase and o ien a ion maps. I was obse ed ha wi h inc easing soaking imes, he olume ac ion o e ained aus eni e g adually inc eases, albei a he cos o i s s abili y. The compa ison o ma ensi e s a empe a u es (M s ) based on he chemical composi ion o aus eni e a 680 ºC wi h ha expe imen ally ob ained a highe p ocess empe a u e e ealed he effec o he g ain size on he educ ion o RA s abili y o longe p ocess imes. Acco dingly, mechanical es s esul s showed ha he yield s ess, ensile s eng h and ha dness dec ease wi h an inc ease in he IA soaking ime. h ps://doi.o g/10.1007/s11661-023-06998-x The Au ho (s) 2023 I. INTRODUCTION STEELS con aining e ained aus eni e (RA) a e becoming inc easingly a ac i e o he au omo i e sec o , as new ehicle sa e y and uel economy s anda ds manda e he use o alloys ha combine high duc ili y and high s eng h. [1] Such s eng h-duc ili y combina- ion may be ob ained by s ain-induced ma ensi ic ans o ma ion (SIMT) o e ained aus eni e, which enhances s ain ha dening. [2,3] The use o mul iphase s eels con aining RA allows dec easing ehicle mass since complica ed pa s cha ac e ized by simila s eng h and s iffness may be p oduced om hinne bu s onge me al shee s. Mo eo e , i some ac ion o RA emains a e he o ming ope a ions, i may be used o any subsequen e en p o iding be e ene gy abso p ion p ope ies. [4,5] Re ained aus eni e s abili y is a key ac o con olling he kine ics o SIMT and, he e o e, he mechanical beha io o mul iphase s eels. I depends on nume ous ma e ial and s uc u al ac o s as: chemical composi ion, [6] mo phology, [7] g ain size, [8] s eng h o su ounding phases, [9] and e en c ys allog aphic ex- u e. [10] The mos ecen g oup o alloys specifically designed o ha e high ac ions o s able RA in he mic os uc u e a e he medium-Mn s eels, which a e he nex s ep de elopmen a e he high manganese s eels. The la e ones, despi e con aining ully aus eni ic mic os uc u e which p o ided he excellen combina- ion o p ope ies, u ned ou o be oo expensi e o be used in he indus y. Medium-Mn s eels ypically con ain 3 o 12 w pc o Mn wi h a ca bon con en o up o 0.2 w pc . [7] The addi ion o Mn is in ended o imp o e he s abili y o he aus eni e by educing he M s empe a u e a a lowe ma e ial cos . ADAM SKOWRONEK and ADAM GRAJCAR a e wi h he Depa men o Enginee ing Ma e ials and Bioma e ials, Facul y o Mechanical Enginee ing, Silesian Uni e si y o Technology, Kona skiego 18a S ., Gliwice 44-100, Poland. Con ac e-mail: [email p o ec ed] CARLOS GARCIA-MATEO and JOSE ´A. JIME ´NEZ a e wi h he Na ional Cen e o Me allu gical Resea ch CENIM-CSIC, A . De G ego io del Amo 8, Mad id 28040, Spain. ROUMEN H. PETROV is wi h he Resea ch G oup Ma e ials Science and Technology, Depa men o Elec omechanical, Sys ems and Me al Enginee ing, Ghen UE ´ni e si y, Tech Lane Ghen Science pa k-Campus A Zwijnaa de, Technologiepa k 46, Ghen 9052, Belgium and also wi h he Depa men o Ma e ials Science and Enginee ing, Del Uni e si y o Technology, 3mE-TUDel , Building 34, Mekelweg 2, Del 2628 CD, The Ne he lands. Manusc ip submi ed Oc obe 27, 2022; accep ed Janua y 31, 2023. A icle published online Feb ua y 20, 2023 METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 54A, APRIL 2023—1443 Cu en ly, he IA hea ea men allows ob aining he la ges ac ion o s able RA in he mic os uc u e o he s eel. [7] This enables o imp o e hei plas ic p ope ies significan ly in compa ison o he momechanically p o- cessed s eels. [11,12] IA consis s o a single aus eni iza ion in he a+c egion, whe e a wide ange o phase a io and chemical elemen s edis ibu ion may be ob ained. The wo mos impo an pa ame e s o he p ocess a e aus eni iza ion empe a u e and soaking ime. Bo h pa ame e s con ol he aus eni e ac ion and he diffusion ange o C and Mn, which a e a he same ime he wo mos impo an elemen s in defining he s abili y o he RA. [13] In gene al, an inc ease o he aus eni iza ion empe a u e inc eases he aus eni e ac- ion and accele a es i s o ma ion. [14,15] Howe e , he e comes a poin when he o e all C and Mn concen a ion in s eel and he diffusion a e canno keep up wi h he apid inc ease in he RA ac ion, esul ing in a poo s abili y, and a o ing ma ensi ic ans o ma ion du - ing cooling, which de e io a es duc ili y. The e o e, many o he IA op imiza ion a emp s ha e ocused on long- ime ba ch annealing a mode a e empe a u es be ween 580 C and 650 C, depending on he chemical composi ion. [16–19] This hea ea men equi es a lo o ime and ene gy, and he e o e, effo s should be made o op imize he use o in e c i ically ea ed medium-Mn s eels in he mo e ene gy efficien con inuous annealing line. Fo his eason, i is necessa y o in es iga e he mic os uc u al e olu ion o medium-Mn s eels o e a wide ange o sho p ocessing imes, bu a ele a ed empe a u es, o de e mine whe he such a combina ion o p ocessing T/ allows good mechanical p ope ies o be main ained. Fu he mo e, he use o highe Al con en in medium-Mn-con aining s eels no only widens bu also inc eases he in e c i ical in e al owa ds highe em- pe a u es, [20] hus p omo ing an inc ease o he ans- o ma ion and diffusion a e. [14] This no el app oach, combining he inc ease o he p ocessing empe a u e wi h an app op ia e chemical design o he s eel, can u he sho en he du a ion o he IA ea men . Addi ionally, an inc eased con en o Al in med- ium-Mn s eels widens he in e c i ical ange owa ds highe empe a u es, [20] p omo ing inc eased ans o - ma ion a e and as e diffusion. [14] Such no el app oaches combining he inc eased p ocess empe a- u e and p ope chemical design o he s eel may u he sho en he IA p ocess. The e o e, he aim o his wo k is o in es iga e he changes and dependencies in RA ac ion, i s s abili y, and he co esponding mechanical p ope ies o Al- ich 5 pc Mn s eel as a unc ion o he soaking ime. II. EXPERIMENTAL PROCEDURE A. Ma e ial and Hea T ea men The chemical composi ion o he Al-alloyed med- ium-Mn s eel used in his in es iga ion is gi en in Table I. Chemical composi ion was de e mined by op ical emission spec oscopy (OES) a e ho olling and quenching a 2/3 o he shee hickness, 4.5 mm, as he mos ep esen a i e a ea, no influenced by possible mic oseg ega ion. The addi ion o 1.6 w . pc Al, whose main ole in he s eel is o p e en ca bide p ecipi a ion, addi ionally causes a widening o he in e c i ical empe a u e ange om 648 C(A C1 ) up o 924 C (A C3 ). [14] Fu he de ails in he alloy design and p o- duc ion p ocess s eps a e a ailable in a p e ious wo k. [14] The fi s s age o in es iga ions conce ning he influence o soaking empe a u e showed ha he op imal mic os uc u e condi ion con aining 35 pc o s able RA we e ob ained a 680 C o 60 min o p ocess du a ion. [14] This empe a u e was he e o e adop ed in he p esen s udy, whe e he a en ion is ocused on he effec o he soaking ime be ween 1 o 300 min in o de o ully desc ibe he changes and dependencies aking place in he ma e ial. The kine ics o aus eni e o ma ion du ing IA, and ma ensi ic ans o ma ion upon subsequen cooling o oom empe a u e, we e s udied by dila ome y using cylind ical samples o 4 mm diame e and 10 mm leng h in a high esolu ion dila ome e , Bah 805A. This de ice enables o s udy phase ans o ma ion by moni o ing he ela i e change in leng h (RCL) o he sample using a Linea Va iable Diffe en ial T ansduce (LVDT) in con ac wi h used silica push ods. The dila ome e uses an induc ion coil and helium blown sys em o hea and cool he sample. The empe a u e is con olled h ough- ou he expe imen wi h an S- ype he mocouple welded a mid leng h o he sample su ace. The dila ome ic da a we e analyzed acco ding o ASTM A1033-04. The ini ial ma ensi ic mic os uc u e p esen in he ho - olled pla e, was subjec ed o he mal ea men s consis ing o hea ing a 3 C/s up o an IA empe a u e o 680 C, hen holding o 1, 5, 15, 60, 120, 300 min and finally cooling o oom empe a u e a 60 C/s. The e olu ion o e i e and aus eni e ac ion a he IA empe a u e as a unc ion o he soaking ime was calcula ed om dila ome ic esul s using a me hod p oposed in p e ious s udy. [14] B. Mic os uc u al Cha ac e iza ion The mic os uc u es ob ained a e he diffe en p o- posed IA ea men s we e e alua ed by means o scanning elec on mic oscopy (SEM), ansmission elec- on mic oscopy (TEM), and X- ay diff ac ion (XRD). Specimens o SEM and XRD we e p epa ed ollowing s anda d me allog aphic p ocedu es ha included mechanical g inding wi h SiC pape up o 2000 g id and polishing wi h 3 and 1 lm g ain size diamond pas e. A final polish using a 35 nm colloidal silica suspension Table I. Chemical Composi ion o he S eel Ho -Rolled Band Used o This In es iga ion Chemical elemen , w pc C Mn Al Si Mo Fe-balance 0.16 4.7 1.6 0.20 0.20 93.12 1444—VOLUME 54A, APRIL 2023 METALLURGICAL AND MATERIALS TRANSACTIONS A (OPS) was included o ensu e a fla su ace wi h minimal de o ma ion. Fo SEM obse a ions a 3 pc Ni al solu ion was used o e ch he polished specimens o ~4s a oom empe a u e o de elop he mic os uc u al ea u es. SEM obse a ions we e pe o med wi h a FEI Quan a FEG 450 SEM equipped wi h an ene gy-dispe - si e X- ay spec ome e (EDS) and an elec on backsca - e diff ac ion (EBSD) de ec o . As he s uc u e p esen in he ma e ial exhibi s la h-like mo phology, he la h hickness was measu ed ins ead o he g ain diame e o desc ibe he mic os uc u e and o unde s and ma e ial p ope ies such as ha dness and s ain–s ess ela ionship. RA la h hickness was de e mined by a leas 200 linea in e cep s on SEM mic og aphs o ob ain eliable s a is- ics. EBSD measu emen s we e pe o med wi h SEM ope a ing a 20 kV and a p obe cu en o ~2.5 nA. All samples we e il ed a 70º and obse ed a a wo king dis ance o 15 mm o maximize backsca e ed elec ons. EBSD mapping has been ca ied ou in a hexagonal scan g id on an a ea o abou 40 940 lm 2 a s ep sizes o 70 nm. The OIM-TSLda a Analysis so wa e .7.31 has been used o da a p ocessing. A wo-s ep clean-up p ocedu e was pe o med o emo e he bad da a poin . In a fi s s ep, i was es ablished a g ain ole ance angle o 5and a minimum g ain size g ea e han 2 pixels. Finally, in he second s ep, all poin s wi h CI (confidence index) <0.1 we e emo ed om he measu emen as no being obus . The olume ac ion o RA was de e mined om he XRD pa e ns eco ded wi h Co adia ion (kKa 1 = 0.1789 nm) in a Panaly ical X’Pe P o MPD diff ac- ome e ) equipped wi h a PIXcel 3D-de ec o in he diffe en hea ea ed ma e ials and in he uni o m de o ma ion a ea o he samples es ed in ension o ailu e. Con en ional h o 2hscans we e pe o med o e a2h ange o 45 o 105 deg wi h a s ep size o 0.026 deg and scan speed o 0.6 s/s ep. The collec ed diff ac ion pa e ns we e analyzed using he e sion 3.0 o he High Sco e Plus so wa e (Panaly ical). The RA ac ion as well as i s la ice pa ame e we e de e mined using he Rie eld efinemen me hod and he c ys allog aphic in o ma ion o e i e and aus eni e. [21] The C concen- a ion in RA was calcula ed om he aus eni e la ice pa ame e using he ollowing exp ession. [22] ac¼3:556 þ0:0453xCþ0:00095xMn þ0:0056xAl;½1 whe e acis in A ˚,xC,xMn and xAl a e elemen s concen a ion in aus eni e (w pc ). The samples o XRD measu emen s o he RA ac ion a e uniaxial ensile de o ma ion we e cu om uni o m de o ma ion a eas o he ensile samples. TEM obse a ions we e pe o med in an S/TEM TITAN 80–300 mic oscope ope a ing a 200 kV in lamellas o app ox. 8 98lm 2 cu along he olling di ec ion wi h ocused ion beam (FIB). Semiquan i a i e mic oanalyses by ene gy-dispe si e X- ay spec oscopy (EDS) using TEM we e pe o med o e alua e changes in Mn and Al concen a ions in RA as a unc ion o he soaking ime. Fo his goal, a leas wen y measu e- men s we e done in andom selec ed a eas o diffe en g ains. F om hese esul s he heo e ical M s empe a u es we e de e mined using he ollowing equa ions. [23–25] Ms¼561 474 C33 Mn 21 Mo ½2 Ms¼545601:2ð1Exp 0:868CðÞÞ34:4Mn 13:7Si15:4Mo1:4Al ½3 Ms¼273 545:8e1:365C34:4Mn 7:5Si þ30 AlÞ ½4 C. Mechanical P ope ies The dependence o he mechanical p ope ies wi h soaking ime was e alua ed a oom empe a u e by means o ha dness and ensile es s. Vicke s ha dness measu emen s we e pe o med wi h a mic oha dness es e FB-700 a a load o 1 kg . Ten measu emen s o each sample we e aken and he mean ha dness de e - mined a e disca ding he wo ex eme alues. Sub-sized fla ensile samples wi h a gauge leng h o 40, 6 mm wide and 2.5 mm hick we e used. They we e machined om he cen al pa o ho - olled shee along he olling di ec ion. These samples we e subjec ed o IA ea - men s a 680 ºC in a Nabe he m LT15/12/P330 muffle u nace unde a gon a mosphe e wi h he same holding ime used in dila ome y es s. The ensile p ope ies we e de e mined acco ding o ASTM E8 s anda d om s a ic ensile es pe o med a a nominal s ain a e o 10 –3 s 1 on a uni e sal es ing machine fi ed wi h an ex ensome e . III. RESULTS AND DISCUSSION A. Fo ma ion and Chemical S abili y o RA The ela i e change in leng h (RCL) measu ed by dila ome y du ing he iso he mal IA ea men a 680º C es s clea ly show a p og essi e con ac ion o he sample wi h inc easing holding ime, as obse ed in Figu e 1(a), which is mainly due o he aus eni e- e- e ed- ans o ma ion (ART) o ma ensi e, and o a lesse ex en o he dissolu ion o cemen i e pa icles in o his phase. Medium-Mn s eels a e some imes suscep ible o seg ega ion o his elemen du ing solidifica ion and consequen ly he RCL measu ed du ing he IA may change om sample o sample. [26] As he a e age kine ics o aus eni e- e e ed- ans o ma ion measu ed in se e al dila ome ic samples shows high consis ency, i was deduced ha he ma e ial used o his s udy does no p esen s ong mic oseg ega ion o alloying ele- men s. F om same Figu e 1(a), i is possible o deduce ha RCL dec eased p og essi ely and 50 and 90 pc o he e-aus eni iza ion ans o ma ion was eached a e ~4 min and ~50 min, espec i ely. This figu e also METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 54A, APRIL 2023—1445 shows a pla eau in he RCL cu e a e ~160 min associa ed wi h he comple ion o he ans o ma ion, which is mo e han 3 imes longe ha he ime needed o an IA a 720 C as epo ed in a p e ious s udy. [14] This esul shows he g ea impo ance o annealing empe a u e and ime in he kine ics o aus eni e o ma ion du ing IA ea men s. The aus eni e ac ion should inc ease as he anneal- ing empe a u e and/o ime inc ease, and his phase mus be e ained a oom empe a u e o ensu e an excellen combina ion o s eng h and duc ili y. As shown in Figu e 1(b) i is obse ed ha ega dless o he holding ime used, when cooling o oom empe a u e, he dila ome ic cu es emain linea and no ma ensi ic ans o ma ion is de ec ed in any case. The de ec able diffe ences in he slopes o he abo e men- ioned cu es a e explained by he diffe ences in he he mal expansion coefficien s o he diffe en mic os uc u es ob ained (diffe en aus eni e/ e i e a ios) du ing he IA ea men s. Acco ding o he me hod p oposed in a p e ious wo k, [14] he heo e ical ac ion o high empe a u e aus eni e has been calcu- la ed om he dila ome ic cu es (Figu e 2(a)) and compa ed o he expe imen al da a (XRD). A good ag eemen be ween he esul s by bo h me hods is obse ed. This figu e shows he e en ion o a consid- e able amoun o aus eni e (abou 16 pc ) in he mic os uc u e a e jus 1 min o annealing a 680 ºC. The a e o ans o ma ion o aus eni e inc eases apidly du ing he fi s 15 min (33 pc ac ion), and hen dec eases d as ically, needing ~160 min o each a final ac ion o ~41 pc . As an icipa ed, he s abili y o RA is closely ela ed o i s ca bon and manganese con en , which has been de e mined om he aus eni e la ice pa ame e using Eq. [1] and by dispe si e ene gy spec ome y (EDS) measu emen s in TEM, espec i ely. The measu ed change in concen a ion o bo h elemen s as a unc ion o he soaking ime is shown in Figu e 2(b). The ca bon concen a ion in RA (Figu e 2(b)) is he highes a e Fig. 1—Dila ome ic esul s showing: (a) he e olu ion o he ela i e change in leng h (RCL) as a unc ion o he soaking ime; g een, ed, and iole do s ma k he soaking ime equi ed o comple e 50, 90 and 100 pc , espec i ely, o he ans o ma ion and (b) dila ome ic cu e, RCL, ob ained on cooling o RT a e he diffe en IA ea men s (Colo figu e online). Fig. 2—Influence o he soaking ime on he e olu ion o : (a) heo e ical and expe imen al aus eni e ac ion calcula ed om dila ome ic cu es and XRD da a, espec i ely; (b) Mn and C con en s in aus eni e de e mined om EDS spec a and aus eni e la ice pa ame e , espec i ely; and (c)M s empe a u e acco ding o Eqs. [2–4] (Colo figu e online). 1446—VOLUME 54A, APRIL 2023 METALLURGICAL AND MATERIALS TRANSACTIONS A 1 min o annealing (0.74 w pc ) and dec eases o 0.39 w pc a e 5 h, ollowing he aus eni e o ma ion eco ded by dila ome y. The C concen a ion in he aus eni e dec eases apidly in he fi s 15 min o iso he mal soaking as he g ea es amoun o aus eni e o ms apidly up o ha poin . In case o he Mn concen a ion (Figu e 2(b)), he e is a sha p inc ease in he ea ly s age o IA om 4.8 o 6.5 pc a e jus 1 min. Such in ense en ichmen in a sho soaking ime was measu ed also by o he esea che s. [27,28] As he soaking ime inc eases, he de ec ed en ichmen a e becomes mo e sluggish, eaching a Mn concen a ion o ~8 w pc a e 5 h o IA ea men . The diffe en beha io o C and Mn pa i ioning can be explained by he diffe ences in hei diffusion a es, which a 680 ºC o he o me is much as e (~10 –10 m 2 s 1 in e i e and ~10 –13 m 2 s 1 in aus eni e) [29] han in he la e (~10 –16 m 2 s 1 in e i e and ~10 –19 m 2 s 1 in aus eni e), [29] he e o e, enabling he C concen a ion o ollow closely he RA ac ion changes, [30] while changes in Mn a e mo e sluggish. [31] Mn concen a ion in aus eni e inc eases sligh ly e en a e 5 h o annealing a 680 C, while ca bon achie es a maximum alue and s a s o d op a e only 1 min. The de ec ed a ia ion o C and Mn in e ained aus eni e mus eflec in i s s abili y agains ma ensi ic ans o ma ion. Figu e 2(c) shows he e olu ion o he M s calcula ed using Eqs. [2–4] and using RA composi- ion gi en in Figu e 2(b). I is obse ed ha , al hough he M s alues diffe depending on he o mula used, he o e all endency is almos iden ical. Despi e he sho soaking ime, a e only 1 min o soaking ime aus eni e exhibi s he highes chemical s abili y, lowes M s em- pe a u e, due o he e y as en ichmen in ca bon (~0.74 pc ) and manganese (~6.5 pc ) o i s ela i ely low ac ion (16 pc ). Du ing he fi s 15 min o annealing he M s empe a u e inc eases significan ly which is expec ed o cause a d op o he he mal s abili y o he e ained aus eni e. The eason is ha despi e he appa en inc ease in manganese con en o 7.4 pc , he s abili y canno be main ained because o as dec easing C concen a ion in an inc easing RA ac ion (32 pc ). A e p olonged soaking imes he men ioned in e play o as diffusing C and slow diffusing Mn allows o main ain e y simila s abili y o he e ained aus eni e, M s empe a u e, up o 5 h. The M s empe a u e changes p esen ed in Figu e 2(c) we e calcula ed only on he basis o changes in he chemical composi ion o aus eni e. I should be emem- be ed ha RA s abili y and he M s empe a u e a e also influenced by o he ac o s besides chemical composi ion which we e men ioned in he in oduc ion sec ion. Fo his eason, ano he IA a ian wi h iden ical p ocess pa ame e s bu a highe IA empe a u e, 760 C, was es ed using dila ome y o eco d ac ual changes in M s empe a u e. The calcula ed aus eni e ac ion a 760 C inc eases o 38 and 69 pc a e 2 s and 1 h o soaking, espec i ely, a e which i emains almos cons an . A ele a ed empe a u e, all p ocesses aking place du ing soaking a e accele a ed and aus eni e ac ion is highe . [14] The changes in aus eni e ac ion a e simila as a 680 C, howe e , o sho e soaking ime. The M s empe a u e again is he lowes o he lowes RA ac ion despi e e y sho soaking ime (2 second) and inc eases wi h soaking du a ion. Wha is impo an he e, e en a e he e i e–aus eni e equilib ium is ob ained a e 60 min o soaking ea men , is ha he M s empe a u e ends o con inuously inc ease (Figu e 3) wi h soaking ime, which diffe s om he obse ed endency a 680 Co annealing empe a u e (Figu e 2(c)). As measu ed, he C concen a ion ollows closely changes in aus eni e ac- ion and e y quickly eaches equilib ium, whe eas Mn diffuses o aus eni e con inuously (Figu e 2(b)). The e- o e, o cons an ac ion o aus eni e in he mic os uc- u e, he M s empe a u e should also be cons an o g ow sligh ly i we conside only he influence o chemical composi ion o aus eni e. The g ow h endency o expe - imen al M s shows he influence o ac o s o he han chemical composi ion as g ain size o ma ix s eng h, on aus eni e s abili y. The local edis ibu ion o Mn in samples soaked a 680 C has also been de e mined om EDS da a ob ained using a spec ome e a ached o an SEM and TEM. Bo h, TEM elemen al line p ofiles and SEM–EDS mappings in Figu e 4show simila e olu ion o he Mn con en wi h soaking ime. Clea and significan changes in manganese concen a ions in e i e and aus eni e egions we e eco ded. Manganese and ca bon edis ibu ed o he aus eni e, while we e deple ed in e i e (Figu e 4). Aus eni e manganese en ichmen is ai ly uni o m in a e 1 min o soaking (Figu e 4(a)). By inc easing he soaking ime o 300 min, u he inc ease i s deg ee o en ichmen al hough i became mo e he e ogeneous, deno ed by peaks o he Mn concen a ion p ofiles close o he e i e/aus eni e in e aces, [32] see Figu e 4(b). This en- dency is e en s onge i he RA is adjacen o coa se e i ic g ains and is caused by significan ly as e Mn diffusion in e i e han in aus eni e. [33] A apid deple- ion o e en la ge e i e g ains in o manganese and ca bon occu s. Howe e , hese elemen s accumula e close o he aus eni e g ain bounda y, which is con- fi med by o he esea che s. [34] B. Mic os uc u e E olu ion The o ma ion o aus eni e du ing IA o ho - olled medium manganese s eels wi h ma ensi ic ini ial mic os uc u e occu s in an aus eni e e e se Fig. 3—The changes in aus eni e ac ion (A) and co esponding expe imen al M s a 760 C o diffe en soaking ime de e mined om he dila ome ic cu es (Colo figu e online). METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 54A, APRIL 2023—1447 ans o ma ion (ART) mode by he g ow h o in e la h aus eni e nuclea ed along ma ensi e la h bounda ies. [35] The newly o med phase he e o e inhe i s he la h-like mo phology p esen in he pa en ma ensi e (Figu e 5(a)) leading o a duplex mic os uc u e wi h aus eni e and e i e lamella (Figu e 5(b)-(d)). As desc ibed p e iously, he aus eni e has an ini ial apid g ow h and a e 1 min o annealing a ela i ely high olume ac ion o hin la h-like e ained aus eni e dis ibu ed be ween coa se empe ed ma ensi e la hs is ob ained, as shown in Figu e 5(b). On he o he hand, his mic os uc u e also included a significan amoun o ca bide pa icles p ecipi a ed in he cen al a ea and a he la h and g ain bounda ies o empe ed ma ensi e, whe e he C and Mn a oms end o seg ega e. [36] The supe sa u a ed C in ma ensi e in s a ing mic os uc u es pa i ioned in o he in e c i ical aus en- i e and also p ecipi a ed as ca bide pa icles du ing he fi s s age o he IA ea men (Figu es 5(b) and (c)). Figu e 6shows some ca bide pa icles a highe magnifica ion and he diff ac ion pa e n om an a ea con aining one o hese ca bide pa icles, which is compa ible wi h cemen i e. These cemen i e pa icles a e almos comple ely dissol ed a e 15 min o IA and he olume ac ion o RA inc eased up o 33 pc . Al hough he amoun o aus eni e has doubled as he soaking ime inc eases o 15 min, he ca bon con en o his phase has only been educed by 0.26 pc due o he dissolu ion o he cemen i e pa icles. The a e o aus eni e o ma ion o soaking imes longe han 60 min d ama ically dec eased. Fo soaking imes be ween 120 and 300 min, he RA con en emains Fig. 4—EDS esul s: TEM mic og aphs wi h Mn elemen al line ob ained in sample annealed o (a) 1 min and (b) 300 min, (c) SEM mic og aphs o a sampled annealed o 300 min oge he wi h (d) Mn and (e) C mappings ob ained in he same zone (Colo figu e online). 1448—VOLUME 54A, APRIL 2023 METALLURGICAL AND MATERIALS TRANSACTIONS A p ac ically cons an . Mo eo e , his longe holding ime esul s in a clea agmen a ion o hick e i ic la hs, which leads o a homogeneous wo-phase s uc u e obse ed in Figu e 5(d). This figu e shows ha inc eas- ing he soaking ime esul s in an inc ease o he mean la h hickness o RA and a widening o he la h hickness dis ibu ion (Figu e 5(e) and ( )). The mos significan inc ease in he la h hickness (abou 40 pc ) occu s wi hin he fi s 15 min o IA, wi h an a e age la h hickness inc easing om 107 ±30 o 150 ±55 nm, when he soaking ime is inc eased om 1 o a e 15 min, espec i ely. Figu e 4(e) shows an almos linea inc ease o he aus eni e la h hickness o longe soaking imes, eaching a alue o 240 ±90 nm a e 5 h. Wang and Van de Zwaag showed ha dec easing he RA size inc eases he numbe o ma ensi e nuclei needed o ob ain a gi en ma ensi e ac ion and hus inc eases i s s abili y. [37] I is caused by RA bounda ies ac ing as obs acles o ma ensi e g ow h. In gene al, all p oduced mic os uc u es show a e y fine mic os uc- u e, which should s abilize RA much mo e han sugges ed om he calcula ed alues o M s Fig. 5—SEM images o he mic os uc u e p esen in (a) ini ial sample, a e soaking a 680 ºC o (b) 1 min, (c) 15 min, (d) 300 min, and e olu ion wi h soaking ime o (e) a e age la h hickness and ( ) equency o hickness edis ibu ion; RA – e ained aus eni e; M – ma ensi e; M T – empe ed ma ensi e; F – e i e; h– ca bides (Colo figu e online). Fig. 6—TEM mic og aphs o he sample ea ed o 1 min (a) b igh field image; (b) diff ac ion pa e n om he a ea ma ked by a da k blue ci cle (Colo figu e online). METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 54A, APRIL 2023—1449 (Figu e 2(c)). The inc ease in RA la h hickness and ma ix empe ing wi h soaking p og ess, howe e , a e ac o s dec easing he ela i e RA s abili y sligh ly, which was e idenced by expe imen al M s da a measu ed o he ea men a 760 C (Figu e 3). The RA la h hickness e olu ion may also explain why he a e age en ichmen o RA in Mn is so high jus a e 1 min o IA. The Mn en iches as hin laye s, whe e he diffusion dis ance in RA is sho (Figu e 4(a)). Fu he en ich- men is inhibi ed by limi ed diffusion a e in hickening la hs (Figu e 5(e)), hence, he inc ease in he a e age Mn concen a ion slows down significan ly as soaking ime inc eases (Figu e 2(b)). I should be no ed ha all samples wi h up o 120 min o IA con ain some ac ion o RA la hs wi h a hickness o £50 nm (Figu e 5( )), while o he 300 min sample he smalles g ains a e >100 nm hick. This may p o e ha new aus eni e g ains nuclea e almos du ing he en i e applied soaking ime, which esul s in a educ ion and agmen a ion o coa se e i ic g ains as obse ed in he SEM images (Figu e 5). The RA la h hickness dis ibu ion widens wi h inc easing soaking ime, which should posi i ely affec he kine ics o SIMT and should allow an inc ease in he plas ici y o he ma e ial by means o a g adual ans o ma ion o RA since he la ge g ains, less mechanically s able, will p omp ly ans o m o ma en- si e a he beginning o plas ic de o ma ion, while he fine g ains, mo e mechanically s able, will ans o m as u he plas ic de o ma ion is applied. EBSD analysis (Figu e 7) indica es ha aus eni e g ains nuclea e fi s a bounda ies o ma ensi ic la hs. Wi h con inued soaking, he aus eni e g ows along he ma ensi ic la hs and s a s o su ound hem and ge s hicke as he RA ac ion inc eases (Figu es 7(b)-( )). EBSD maps o samples ea ed o 1 and 5 min show an unde es ima ed ac ion o RA since he e y hin <50 o 70 nm aus eni e la hs o med du ing he sho p ocess du a ion (Figu es 5(b) and ( )) a e no esol ed by EBSD, since he su ounding phase in e - e es wi h hei diff ac ion pa e ns leading o low IQ alues a hese loca ions. The g ain miso ien a ion (Figu e 8(a)) dis ibu ion o he ini ial sample co e- sponds o ha o a ypical ma ensi ic mic os uc- u e. [38,39] The e a e isible peaks o low angle (5 deg >) and high angle (>55 deg) bounda ies ela ed o inc eased disloca ion densi y [40] and mu ual o ien a ion o ma ensi ic la hs, espec i ely. [38,39] Wi h an inc ease in aus eniza ion ime, he a o emen ioned peaks ade ou (Figu e 8(a)), while a as inc ease in he numbe o g ains wi h a miso ien a ion o ~45 deg is isible (Figu e 8(b)), which ep esen he a/co ien a ion ela- ionship. [38] Based on he EBSD phase map (Figu e 7) and miso ien a ion esul s (Figu e 8(b)) i can be assumed ha each single e i ic g ain is su ounded by RA wi h an a e age o 60 pc a e 5 min o soaking and almos 80 pc a e 5 h, ensu ing a e y homoge- neous mic os uc u e. This is a diffe en ela ionship han in he case o IA a e cold olling, whe e bo h he e i e and aus eni e g ains a e o en loca ed nex o each o he in a andom manne . [7,36] This endency may influence he diffe en mechanical beha io o bo h ypes o medium-Mn s eels. Fig. 7—EBSD phase maps o samples soaked o diffe en imes (Colo figu e online). 1450—VOLUME 54A, APRIL 2023 METALLURGICAL AND MATERIALS TRANSACTIONS A C. Mechanical P ope ies In a s eel, each phase p esen in he mic os uc u e con ibu es o he o e all mechanical p ope ies o he agg ega e. Thus, hei ha dness and ensile p ope ies a e de e mined by con ibu ion o he ma ix ( e i e, baini e, ma ensi e o a combina ion o hem) and he olume ac ion, shape, and mechanical s abili y o aus eni e. All hese mic os uc u al pa ame e s can be modified in a ho - olled band pla e by adap ing he ime and/o empe a u e used o he IA ea men , making possible he ailo ing o he mechanical p ope ies o hese ma e ials. The ha dness o hea ea ed samples (Figu e 9) dec eases wi h p olonged aus eni iza ion ime om 318 ±10 HV1 a e 1 min o 279 ±8 HV1 a e 5 h, which is ela ed o ma ensi e empe ing and inc ease o he RA ac ion. The same endency is also obse ed in he epo ed YS alues, which dec ease om 859 ±16 MPa o 591 ±11 MPa o he same ime ange as abo e (Figu e 10). Since he ha dness and yield s ess o he empe ed ma ensi e is highe han ha o aus eni e, he a e age ha dness alues o he s eel will dec ease as he olume ac ion o RA inc eases. The dec ease o he UTS is much smalle han o he YS ( om 938 ±18 o 890 ±20 MPa) due o in ense SIMT in he samples soaked o mo e han 15 min. Such a esul was an icipa ed as he p olonged soaking causes an inc ease in he ac ion o RA wi h lowe s abili y, whose ans o ma ion p omo es an enhanced s ain ha dening. [2] None o he samples exhibi discon inuous yielding due o he specific la h-like mo phology. Bo h uni o m elonga ion (UEl) and o al elonga ion (TEl) (Figu e 10) o he samples ini ially inc ease wi h inc eas- ing IA ime demons a ing he highes UEl alue o 19.5 ±0.5 pc ollowing a 15 min ea men and he highes TEl o 28.3 ±0.6 pc ollowing a 60 min ea men . Thei duc ili y dec eases sligh ly o final alues o 16 ±0.4 pc o UEl and 26.5 ±0.8 pc o TEl a e a ea men o 5 h. Fo samples ea ed wi hin 1 and 5 min he o e all s ain ha dening is low due o ema kably high s abili y and co esponding low M s empe a u es, and hin la hs (<110 ±30 nm) o RA. Also, he RA in hose samples is su ounded by empe ed ma ensi e, while in he es o he samples he ma ix is composed o e i e. The high s eng h o he empe ed ma ensi e ma ix u he enhances RA s abili y. [9] The e o e he influence o his ac o should be he s onges in he sample ea ed o only 1 min. High RA s abili y in his sample is con- fi med by he p esence o 12 pc o his phase in he UEl egion a e ensile es (Figu e 11). I means ha RA exhibi ing high s abili y does no ans o m e en unde Fig. 10—Tensile es esul s o samples annealed o diffe en imes (Colo figu e online). Fig. 9—Ha dness (HV1) esul s o samples annealed o diffe en imes. Fig. 8—EBSD esul s, (a) dis ibu ion o g ain bounda ies miso ien a ion; and (b) ac ion o g ain bounda ies wi h miso ien a ion o ~45 (Colo figu e online). METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 54A, APRIL 2023—1451