Study of dynamic behaviour via Taylor anvil test and structure observation of M300 maraging steel fabricated by the selective laser melting method

Jou nal o Manu ac u ing P ocesses 125 (2024) 283–294 A ailable online 20 July 2024 1526-6125/© 2024 The Au ho s. Published by Else ie L d on behal o The Socie y o Manu ac u ing Enginee s. This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/). S udy o dynamic beha iou ia Taylo an il es and s uc u e obse a ion o M300 ma aging s eel ab ica ed by he selec i e lase mel ing me hod S epan Kolomy , Mi osla Jopek * , Jose Sedlak, Ma ek Benc, Jan Zouha B no Uni e si y o Technology, Facul y o Mechanical Enginee ing, B no, Czech Republic ARTICLE INFO Keywo ds: Ma aging s eel Taylo an il es Selec i e lase mel ing Johnson–cook High s ain a e Mic oha dness S uc u e ABSTRACT This pape deals wi h he M300 high s eng h ma aging s eel ab ica ed ia selec i e lase mel ing me hod. Mechanical p ope ies especially mic oha dness and comp essi e yield s eng h o ma aging s eel in as-p in ed s a e we e obse ed. The acqui ed da a was implemen ed in Johnson-Cook cons i u i e equa ion used o nume ic simula ion, which showed he sa is ac o y co ela ion wi h he obse ed expe imen . Dynamic beha - iou unde high s ain a e (impac eloci y eached 185 m.s −1 ) was in es iga ed ia Taylo An il Tes . The expe imen e ealed s uc u e and geome ical changes accompanied wi h he c ea ion o cha ac e is ic unnel- like and cylind ical a eas on he de o med sample. Impac ed sample o ehead ea u ed he inc ease o mic o- ha dness (465 HV) accompanied wi h s uc u e changes. The s uc u e in he unnel-like a ea exhibi ed he dec ease o a e age g ain size, which eached he minimum (3.1 μ m) in he icini y o he impac ed o ehead. Close analysis e ealed ha he high s ain a e caused he inc ease o high ac ion high-angle g ain bounda ies (50,8 %) and highe geome ically necessa y disloca ion densi y (52.77 1.m −2 ) in he unnel-like a ea. 1. In oduc ion Addi i e manu ac u ing (AM) is he p ocess o c ea ing a pa based on a digi al model c ea ed ia Compu e Aided Design (CAD) so wa e. In AM, ma e ial is g adually added (laye by laye ), while in machining, ma e ial is g adually emo ed. The pa is inally o med by sin e ing indi idual laye s o a de ined hickness, which a e s acked on op o each o he [1]. This echnology enables p oduc ion o complex shapes ha would no be possible o p oduce by con en ional me hods [2,3]. AM is used and ep esen ed in se e al sec o s such as he au omo i e and ae ospace indus ies. I could also ind i s applica ion in heal hca e and ashion [4–6]. AM enables o manu ac u e a ious ypes o ma e ials, such as ool s eel [7] s ainless s eel 15-5PH [8] and 316 L s eel [9–11], Inconel 718 [12–14], Ni alloy [15], Ti alloy [16,17] o Al alloy [18,19], whose s uc u e and mechanical p ope ies can di e depending on he p in ing pa ame e s, s a egies, hea ea men , e c. [20,21]. Gene ally, he beha iou o mos ma e ials unde dynamic loading is di e en han unde s a ic/quasi-s a ic loading [22,23]. Machine com- ponen s such as gea boxes, c anksha s e c. a e subjec ed o dynamic loading dec easing hei li e. The use o dynamic es s p o ides a ma- e ial assessmen , which helps o design new componen s. Dynamic es s will ob ain inpu s in o he ma e ial models, which can be used o nume ic simula ion. The simula ion p o ides eal esul s, which can be used in he design o machine componen s. The me hods, which can be used o e alua e dynamic p ope ies o he ma e ial a e he Spli Hop- kinson P essu e Ba (SHPB) me hod and he Taylo An il Tes (TAT) o en pe o med o gain ha dening beha iou s o nume ous alloys a ul a-high s ain [24]. TAT can also be used in e e se mode o e alua e he esponse o he ma e ial o dynamic loading o can be used in bal- lis ics es s [25]. Using he esul s o he dynamic p ope ies in p ac ice could lead o a be e , mo e accu a e and as e design o a machine componen [26]. Many s udies and a icles ocused on he in es iga ion o he dy- namic p ope ies o ma e ials. Rua e al. [27] s udied composi es wi h epoxy esin polyme ma ix. The mic os uc u e o he es samples we e in es iga ed using scanning elec on mic oscopy. Dynamic es s o he ma e ial we e pe o med using he SHPB es , which p o ided s ess- s ain cu es o di e en s ain a es. Lu and Li [28] in es iga ed he dynamic beha iou o polyme s a high s ain a es using he SHPB. The expe imen ally measu ed dependence o s ess on s ain o di e en s ain a es is subsequen ly implemen ed in o he dynamic cons i u i e Johnson-Cook equa ion including he e ec o s ain a e and empe - a u e. Gilbe son [29] in es iga ed he dynamic p ope ies o wood using he SHPB. The de o ma ion o he sample was measu ed by s ain * Co esponding au ho . E-mail add ess: [email p o ec ed] (M. Jopek). Con en s lis s a ailable a ScienceDi ec Jou nal o Manu ac u ing P ocesses jou nal homepage: www.else ie .com/loca e/manp o h ps://doi.o g/10.1016/j.jmap o.2024.07.057 Recei ed 21 Ma ch 2023; Recei ed in e ised o m 15 Ap il 2024; Accep ed 10 July 2024 Jou nal o Manu ac u ing P ocesses 125 (2024) 283–294 284 gauge and digi al image analysis. SHPB has been ound o be a use ul ool o assessing he ma e ial p ope ies o wood subjec ed o high s ain a es o sho loading imes. Chen e al. [30] in es iga ed he dynamic p ope ies o specimens made om conc e e. Tes ing was pe o med using The SHPB o di e en loading angles and di e en impac eloci ies. Fo lowe impac eloci ies, he s ess-s ain cu e ea u ed a simila cou se o quasi-s a ic loading, in which he ini ial c ack occu s in he cen e o he specimen and p opaga es ou wa d. Impac eloci y has a signi ican ole in e ms o local s ess dis ibu ion in he sample. Kang e al. [31] compa ed he dynamic p ope ies o shale and g ani e. Dynamic p ope ies we e e alua ed using SHPB. They ound ha dynamic loading educed he impac s eng h. Renliang e al. [32] in es iga ed he dynamic beha iou o sands one ia he SHPB. The empe a u e o he samples was main ained a −15 ◦C du ing he expe imen . The es esul s e ealed ha when he a e age s ain a e was low, he sample s eng h changed g adually bu when i was high, i s s eng h changed apidly. Li, Xu, Chen e al. [33] in es iga ed he dynamic p ope ies o coal samples subjec ed o di e en impac e- loci ies. They obse ed an ob ious s ain a e e ec on he dynamic s eng h o specimens, which exhibi ed c acks unde de ined impac ed angle, a ec ing s ess concen a ion in specimens and changed hei dynamic beha iou . High demands on he high-se ies p oduc ion o such componen s, which a e ypically manu ac u ed om Fe ma e ial by au oma ic sequen ial machines, which can p oduce componen s a s ain a es up o 10 3 s −1 . To ensu e success ul indus ial p oduc ion he mapping o dynamic beha iou o he used ma e ial is c i ical. Liu e al. [34] used SHPB o he in es iga ion o U71Mn s eel (used o ails) he dynamic p ope ies o e a wide ange o s ain a es and empe a u es. Johnson- Cook models we e de e mined ega ding his ma e ial based on he expe imen al da a. Jopek [35] in es iga ed he dynamic beha iou o TRISTAL low ca bon s eel using he TAT. The expe imen was pe o med on a de ice in he labo a o y o high de o ma ion a es a he Facul y o Mechanical Enginee ing in B no. The LS DYNA p og am was used o simula e he mechanical beha iou o TRISTAL s eel unde dynamic loading. Kuncicka e al. [36] showed he esul s o he impo an pa- ame e s implemen ed in Johnson-Cook cons i u i e equa ion o TRIS- TAL s eel. Nume ic simula ion con i med a sa is ac o y co ela ion wi h expe imen ally obse ed da a. The s udy o dynamic beha iou o ma e ials p oduced by he 3D p in ing me hod ha e been add essed by se e al s udies. Zhang e al. [37] in es iga ed he dynamic p ope ies o 3D-p in ed ce amics. Due o di e en s uc u es o AM pa compa ed o adi ionally manu ac u ed he di e ences in dynamic beha iou was obse ed. Sha ma e al. [38] in es iga ed he e ec o p in ing pa ame e s on he dynamic beha iou o polyamide samples p oduced by he selec i e lase mel ing (SLM) me hod. S o age modulus inc eases as he lase powe , bed empe a u e and scan leng h inc eased while i dec eased wi h he inc ease in he scan spacing. Mansou e al. [39] in es iga ed he dynamic p ope ies o PLA ma e ial samples p oduced ia FDM echnology. A ew samples we e ein o ced wi h g aphene and subsequen ly subjec ed o dynamic es s. The esul s showed ha he inco po a ion o g aphene imp o ed he elas ic modulus, s eng h, and ha dness o he 3D p in ed samples. S udies [40,41] in es iga ed he dynamic beha iou o Ti-6Al-4 V samples ab ica ed by AM. The esul s ob ained h ough nume ic simula ion showed a linea ela ionship wi h expe imen ally ob ained esul s. Dynamic-mechanical p ope ies o AM AlSi10Mg aluminium alloy we e in es iga ed ia SHPB o gain s ess-s ain cu es. Di e ences in he dynamic p ope ies o his alloy p oduced wi h di e en ma e ial o ien a ions accompanied wi h aniso opic p ope ies we e obse ed [42–44]. A se ies o TATs we e pe o med on addi i ely (by SLM) and con en ionally manu ac u ed 316 L s ainless s eel samples. A di e ence in dynamic p ope ies was obse ed be ween e ical, ho izon al (SLM) and con en ionally manu ac u ed specimens [45,46]. Resea che s [47,48] es ed AM as-buil ma aging s eel specimens, which we e sub- jec ed o dynamic loadings o in es iga e he dynamic beha iou , de o ma ion mechanisms, mic os uc u al and ex u e e olu ion ia SHPB. Resul s showed, as-buil samples exhibi ed ac u e a he s ain a e o 3500 s −1 . Ma aging s eel p oduced by AM is no pe ec ly known in ega d o he dynamic beha iou . No many exis ing s udies we e ocused on dynamic in es iga ion o M300 ma aging s eel ab ica ed ia SLM. The aim o his s udy is o analyse he dynamic beha iou o M300 ma aging s eel ia Taylo an il es . Mechanical p ope ies especially comp essi e yield s eng h and mic oha dness in as-p in ed s a e we e e alua ed o gain he in o ma ion abou he ma e ial model, which is implemen ed in o he Johnson-Cook cons i u i e equa ion and subsequen ly used o nume ic simula ion. The Real expe imen is accompanied wi h he ollowing analysis e ealing a se e e s uc u e change. C ys allog aphic ex u e, a e age g ain size, high angle g ain bounda ies and geome i- cally necessa y disloca ion densi y a e desc ibed in de ail. 2. Expe imen al p ocedu e Me al powde o he M300 ma aging s eel was used o p oduce he samples. Fou a ious ypes o samples, which we e used o di e en ypes o analysis we e ab ica ed ( ypes o samples a e lis ed in Table 1). Fu he mo e, he chemical composi ion o as-p in ed sample ( ype 1) was de e mined using he Q4 TASMAN (B uke , Rudice, Czech Repub- lic) de ice. The gained chemical composi ion o his sample (p esen ed in Table 2) complied wi h he i gin chemical composi ion o powde p o ided by he p oduce . Cylind ical samples o es ing we e ab i- ca ed using he RENISHAW RenAM 500E (RENISHAW, New Mills, UK) 3D p in e . Meande Ha ching Pa en s a egy wi h 67◦ o a ion a e each laye was used o ab ica ion o he samples (see Fig. 1a). The s a egy p o ides su icien mel ing p ocess, which exhibi s low po osi y and o ma ion o a a ou able s uc u e [49]. Sample po osi y ( ype 2) was e alua ed by op ical mic oscope Olympus DSX500, which was used o scan he su ace o he sample. The building di ec ion (BD) was di ec ed o he Z axis ( e ical di ec ion). P ocess pa ame e s used o ab ica ion a e lis ed in Table 3. The eq. (1) [50,51] was used o he calcula ion o he ene gy densi y, which was se du ing he building p ocess. E=P h (1) whe e E is lase ene gy densi y (J.mm −3 ), P is lase powe (W), is scanning speed (mm.s −1 ), h is ha ch spacing (mm) and is laye hick- ness (mm). Mic oha dness o he samples ( ype 3) was measu ed in as-p in ed s a e. Mic oha dness measu emen was pe o med in wo a ious di- ec ions i.e. ho izon al and e ical, using load o 200 g and dwell ime 10s. To gain he comp essi e yield s eng h, necessa y machining ope a ion was ca ied ou . Bo h aces o he sample we e no pe ec ly la a e ab ica ion ( he e was a suppo on he i s ace and he second ace was no la ). Pos p ocess ace u ning b ough a ou able su ace quali y on bo h aces. Subsequen ly he samples we e placed in o a special ix u e, which was used o comp essi e es and we e p essed a he cons an speed 10 mm.min −1 a he oom empe a u e 22 ◦C±1. Machine wi h an EDC 60 (B no, Czech Republic) con ol uni was used o comp essi e loading. Fo da a acquisi ion he TIRA es .2.1 so - wa e was used. S ess–loga i hmic s ain cu es we e de e mined om acqui ed da a. Table 1 Types o di e en samples used o di e en ype o analysis. Type Analysis s a e amoun [−] 1 Chemical composi ion as- ab ica ed 1 2 Po osi y e alua ion as- ab ica ed 1 3 Tes ing o mechanical p ope ies as- ab ica ed 8 4 Dynamic es ing as- ab ica ed 1 S. Kolomy e al. Jou nal o Manu ac u ing P ocesses 125 (2024) 283–294 285 Tes o dynamic beha iou ia TAT ook place in he labo a o y o high s ain a es a he Facul y o Mechanical Enginee ing BUT in B no. The de ice used o es ing o he dynamic p ope ies is shown in Fig. 2a. The sample ( ype 4) oge he wi h he ca ie (see Fig. 2b) is placed in he ai essel om whe e he sample was i ed owa ds he ixed ba . Imminen ly he sample was sepa a ed om he ca ie , he impac e- loci y was measu ed. The impac eloci y eached max. Value o 185 m. s −1 in his expe imen . Fig. 2c shows ini ial dimensions o he sample used o dynamic es ing, whe e L 0 is he leng h o he sample and D 1 and D 2 a e diame e s on bo h sample aces. Fig. 2d ep esen s ma e ial model wi h he key pa ame e s (L = inal leng h, X =leng h o non-de o med a ea, L =leng h o de o med a ea, D1 =diame e o non-de o med cylind ical a ea, D max =max. Diame e a e de o ma ion) subse- quen ly in eg a ed in he nume ic simula ion. De ailed s uc u e obse a ion o sample a e TAT, which was p e- pa ed by puck cas ing and au oma ically polished and e ched by 2 % ni al solu ion was examined. Mic oha dness o de o med sample was measu ed in YZ plane (see Fig. 1b) along and pe pendicula o sample axis (see Fig. 1c). The s uc u e was obse ed in he YZ plane using he Olympus DSX500 (Olympus, P ag, Czech Republic) ligh mic oscope and he KEYENCE VHX-F (KEYENCE, Mechelen, Belgium) digi al mi- c oscope. Tescan Ly a 3 XMU FEG/SEMxFIB (Tescan, B no, Czech Re- public) wi h Symme y EBSD de ec o was used o de ailed s udy o he mic os uc u e. The scan s ep o he EBSD (elec on-backsca e ) scanning was 0.15 μ m. Nume ical simula ion o TAT was ealised using so wa e ANSYS Explici dynamic (Ansys Inc., Canonsbu g, Pennsyl ania, USA). Hexag- onal mesh was applied o e alua e he sample a e TAT ia ini e elemen me hod (FEM). Close o 270 housand elemen s wi h 308 housand nodes we e implemen ed o he de ail de ini ion, wi h no applied e-meshing. The s a ing empe a u e was se as 22 ◦C. The TAT nume ical simula ion was ca ied ou using Johnson-Cook (J–C) eq. (2). Each pa ame e om he equa ion was de e mined expe imen ally and compa ed o pa ame e s acqui ed by he nume ic simula ion. σ = ( σ 0+Bφn) • ⎛ ⎝1+Cln ˙ φ ˙ φ0 ⎞ ⎠• (1−T*m)(2) whe e σ 0 is s a ic low s ess, B is wo k ha dening coe icien , ˙ φis s ain a e, n is wo k ha dening exponen , C is s ain a e sensi i i y, m is he he mal so ening coe icien , and T* is homological empe a u e de ined in he ange 0 <T* <1 by eq. (3) T*=T−T0 Tm−T0 (3) whe e T 0 is he e e ence empe a u e (a s a ic low s ess σ 0 = σ ), and T m is he mel ing empe a u e. 3. Resul s and discussions The p ocess o 3D p in ing is based on he concep o placing each laye by laye un il he whole pa is ab ica ed. Di ec p in ed su ace is depic ed in Fig. 3.Fig. 3a shows he on iew o he as-p in ed sample whe e he ha ching pa hs a e clea ly isible. Dis ance be ween each pa h is app oxima ely equal o 0.095 mm (ha ch spacing). Fig. 3b shows he side iew o he p in ed sample. This su ace ea u ed a non-su icien su ace quali y a e p in ing. The co e laye o p in ed samples con- ains plen y o mic opo es and oids, which can nega i ely a ec dy- namic and mechanical p ope ies [52]. The e o e, i is needed o machine su aces, which a e conside ed o be unc ional. AM echnology especially SLM can p oduce nea ly ully dense and high-pe o mance pa s o a e y complex shape [53]. The combina ion o implemen ed p ocess pa ame e s could a ec he o ming s uc u e and he po osi y [54]. Po es, ca i ies, o oids could appea due o he p o ec ing gas, which is en apped du ing he sin e ing p ocess [55,56]. S uc u e, which epo s lowe po osi y would ea u e highe a igue esis ance and mechanical p ope ies [57]. Po osi y was examined on Table 2 Chemical composi ion o es ed sample. Elemen Ni Co Mo Ti Si Mn C P Fe Q4 TASMAN [W %] 18.38 8.84 4.67 0.71 0.01 0.04 0.02 0.02 66.40 Fig. 1. (a) Meande Ha ching Pa en 67 ◦ o a ion a e each laye and (b) C oss- sec ion plane depic ed in he sample subjec ed o TAT and (c) lines 1–3 used o mic oha dness measu emen and places I-III in YZ plane used o s uc u e obse a ion. Table 3 P ocess pa ame e s used o ab ica ion o he samples. Ene gy densi y [J/ mm 3 ] Lase powe [W] Scanning speed [mm/s] Lase hickness [mm] Ha ch spacing [mm] 133 410 810 0.04 0.095 Fig. 2. (a) De ice used o es ing o dynamic p ope ies ia TAT, (b) assemble con ains o sample and ca ie and ma e ial model (c) be o e de o ma ion (d) a e de o ma ion. S. Kolomy e al. Jou nal o Manu ac u ing P ocesses 125 (2024) 283–294 286 he sample ( ype 2 men ioned in Table 1), which was ab ica ed wi h iden ical p ocess pa ame e s. Po osi y was e alua ed in ho izon al (see Fig. 4a) and e ical di ec ion (see Fig. 4b). Th ee indi idual planes in each di ec ion (pe pendicula and pa allel o BD) we e chosen. The planes we e si ua ed on op, in middle heigh and in he bo om o he sample. These planes we e sepa a ely scanned on op ical mic oscope Olympus DSX500. Po osi y o he sample is calcula ed as a e age alue om each single e alua ed plane (see Table 4). Fig. 4 shows he mic o and mac o po es, which a e si ua ed dominan ly in co e laye o he sample. In case o emo al he co e laye (0.4 mm om su ace) he global po osi y would dec ease d ama ically. The po osi y was also e alua ed wi hou conside ing he co e laye , which b ough he sig- ni ican educ ion o po osi y. 3.1. Mechanical p ope ies The as-p in ed samples ( ype 3 co esponding o Table 1) we e used o he mic oha dness e alua ion. Fou indi idual samples we e measu ed o de e mine mic oha dness in ho izon al (pe pendicula o BD) and e ical (pa allel o BD) di ec ions, which is depic ed in Fig. 5. The middle heigh plane was used o measu e mic oha dness in ho izon al di ec ion due o he expec a ions o he highes po osi y connec ed wi h lowe mic oha dness [58]. The esul ing alues o mic oha dness a e gi en as he a e age alues calcula ed om 12 in- den a ions in each di ec ion. A e age mic oha dness in ho izon al and e ical di ec ion ea u ed o 411 HV and 410 HV espec i ely. I could be concluded; he mic oha dness is independen on i s measu ing di ec ion. Bo h aces o as-p in ed samples ( ype 3) we e no pe ec ly la a e ab ica ion. The i s ace o he sample ea u ed a building suppo , and he second ace did no exhibi wi h an app op ia e su ace quali y. Face u ning p o ided a a ou able su ace quali y, which ensu ed he su i- cien condi ions o comp essi e es ing. The comp essi e s ess –log- a i hmic s ain cu es o he examined samples a e shown in Fig. 6. The di e en comp ession s oke was applied, which a ied o each sample: sample 1 ea u ed he 6 mm s oke, sample 2 was exposed o 7 mm comp ession s oke, sample 3 and sample 4 we e exposed o 8 mm and 9 mm comp ession s oke. Due o he a ious comp ession s oke he imposed loga i hmic s ain was di e en o each sample. The a e age Fig. 3. Su ace quali y in as-p in ed s a e: (a) on iew on SLM piece and (b) side iew. Fig. 4. Ligh op ical mic oscopy (LOM) images show po osi y (c oss-sec ion in middle heigh ) in (a, b) ho izon al di ec ion and (c, d) e ical di ec ion, (a–d) co esponds o sample ype 2 acco ding o Table 1. Table 4 E alua ed po osi y o he sample in ho izon al a e ical di ec ion. SAMPLE Type 2 Po osi y in ho izon al di ec ion [%] Po osi y in e ical di ec ion [%] Sec ion 1 2 3 1 2 3 Unbounded a ea 0.161 0.180 0.170 0.160 0.150 0.190 Bounded a ea 0.022 0.025 0.031 0.025 0.023 0.029 Fig. 5. Mic oha dness o ou di e en samples in as-p in ed s a e. Mic o- ha dness is gi en in ho izon al and e ical di ec ion. S. Kolomy e al. Jou nal o Manu ac u ing P ocesses 125 (2024) 283–294 287 comp essi e yield s eng h calcula ed om all samples ea u ed 1177 ± 5 MPa. Addi i e manu ac u ed pa should wi hs and he highe dynamical load he highe mechanical p ope ies, he pa exhibi ed [59]. The inal dimensions ( inal heigh , maximal and minimal diam- e e ) o comp essed samples we e impu ed in o nume ic simula ion o pe o m dynamic beha iou . 3.2. Dynamic es ing Dynamic beha iou o he sample ( ype 4) was examined using he TAT de ice. Dimensions o es ed sample in unde o med s a e we e measu ed and subsequen ly used o inal e alua ion as well as he di- mensions a e de o ma ion. U ilisa ion o acqui ed pa ame e s led o e alua ion o dynamic beha iou ia FEM. Impac eloci y was 185 m. s −1 , which in lic ed uni o m plas ic de o ma ion depic ed in Fig. 7. De o med sample can be di ided in wo cha ac e is ic a eas i.e., unnel- like, and cylind ical a ea. The i s a ea was exposed o high s ain a e, which led in c ea ion o he unnel-like a ea, co esponding o se e e plas ic de o ma ion. The c ea ion o unnel-like a ea ins ead o ba el (in case o comp essi e es ) p o ed ha he sample was exposed o high s ain a e (dynamic loading) and was no in luenced by he ic ion. The second a ea was de ined as unde o med cylinde wi h he ini ial diam- e e . The ac , ha any mac o c acks we e obse ed, can be a ibu ed o low impac eloci y. Highe impac eloci y can cause des uc ion o he sample due o c acks ini ia ion [29]. The nex ac o , which can a ec de o ma ion could be impu i y, oids, po es e c., con ained in co e laye , which can lead o c ack ini ia ion and apid des uc ion o he sample [60]. The sample a e high s ain a e was obse ed ia scanning elec on mic oscopy (SEM). SEM was used o de ail scan o a eas con aining de ec s on he impac ed egion o he sample, which is depic ed in Fig. 8. The uni o m plas ic de o ma ion o he sample o ehead (depic ed in Fig. 8a) con ained a huge numbe o aces equally sp ead in whole a ea. Fig. 8b showing he de ail a ea con aining isible impu i ies ma ked in yellow ci cle wi hin his a ea. In addi ion, hey can se e as a s ess concen a o , om which c acks can ini ia e du ing he mechanical as well as dynamic loading. The de o med su ace con ained plen y o c acks o med a e he on al impac . Close de ail o c ack is p esen ed in Fig. 8c. Each c ack is o med in main di ec ion wi h se e al seconda y mic o sc a ches pe pendicula o main c ack. 3.3. Nume ical simula ion The choice o he cons i u i e equa ion was based on se e al e- qui emen s in ol ing he physical na u e o he e en . The J–C equa ion includes i e pa ame e s, whe e wo pa ame e s, i.e. he s a ic low s ess and he ha dening coe icien , we e ob ained om comp essi e es . The acqui ed esul s om comp essi e es showed ha he comp essi e yield s eng h a e in he ange 1150 MPa o 1204 MPa. The alue 1204 MPa was se in J–C cons i u i e equa ion, which was used o he TAT nume ic simula ion. Fig. 9a shows he dependences be ween he e o s (pe cen age di - e ences be ween he expe imen and he simula ion in he obse ed a eas) on he s ain a e sensi i i y. These cu es we e e alua ed om he sample, which ea u ed he 185 m.s −1 impac speed. The pe cen age di e ences we e obse ed in he a eas o maximum sample diame e D max and inal leng h L . The a e age e o eached a minimum alue a C=0.138. When he s ain a e coe icien inc ease/dec ease he de- ia ion would inc ease. A e he op imiza ion o he s ain a e sensi- i i y pa ame e C, he so ening exponen m was also op imized. TAT expe imen s we e also pe o med a ele a ed empe a u es, speci ically o a measu ed su ace impac empe a u e o 175 ◦C, whe e he hea ed sample was accele a ed owa ds he ba 's ace. A he ba 's ace, he su ace empe a u e o he sample was eco ded wi h a non-con ac empe a u e me e . By compa ing he con ou o he sample, namely he maximum sample diame e D max and he inal leng h L o he sample a e he expe imen and op imizing D max and L o he sample in he nume ical simula ion gi en he eco ded su ace impac empe a u e, he so ening exponen m was op imized. The esul s o he so ening exponen op imiza ion can be seen in Fig. 9b. The dependence o de o ma ion s ain on he loga i hmic de o ma- ion was ob ained (see Fig. 10) by subs i u ing indi idual pa ame e s in o he cons i u i e equa ion acco ding o J–C. The cu es we e plo ed o di e en s ain a es, i.e., 0.02, 1200, 3000 s −1 . The cu e acqui ed by comp essi e es ( ep esen ed by he yellow line) was also included. The obse ed de ia ion be ween yellow and blue lines (comp essi e es and J–C equa ion) ea u ed he minimal de ia ion. The pa ame e s Fig. 6. Comp essi e s ess –loga i hmic s ain cu es o all he examined samples in as-buil condi ions, he sample placed in ix u e and is also included. Fig. 7. Sample subjec ed o TAT: (a) on iew o he impac ed o ehead o he sample, (b) side iew p esen ing unnel-like and cylind ical a ea o he sample. S. Kolomy e al. Jou nal o Manu ac u ing P ocesses 125 (2024) 283–294 288 co esponding o J–C equa ion, which we e inally ob ained by nume ic simula ion a e shown in Table 5. The esul s o he e ec i e s ain, he ac ual s ain in indi idual axis is depic ed in Fig. 11a-d. Fig. 11a shows he dis ibu ion o XY s ain componen s h oughou he es ed sample a e TAT, which exhibi ed he highes de o ma ion a impac ed o e- head. Nume ically p edic ed dis ibu ions o he s ess is depic ed in Fig. 8. SEM images o he de o med TAT sample o ehead imposed o high s ain a e: (a) on iew o he sample, (b) a ea o in e es , (c) de ail o mic o c ack. Fig. 9. Dependence o he de ia ion on he s ain a e sensi i i y (a) and dependence o he de ia ion on he so ening exponen m (b). Fig. 10. Dependence o de o ma ion s ain on he loga i hmic de o ma ion. Table 5 Nume ically ob ained pa ame e s o J–C cons i u i e equa ion. σ 0 [MPa] B [−] n [−] C [−] m [−] 1104 1099 0.449 0.138 0.46 Fig. 11. Nume ically p edic ed dis ibu ions o s ain in XY (a), dis ibu ion o s ess (b), YY s ain (c) and XX s ain (d). S. Kolomy e al. Jou nal o Manu ac u ing P ocesses 125 (2024) 283–294 289 Fig. 11b whe eas Fig. 11c p edic s he dis ibu ion o YY s ain wi hin es ed sample, which eached maximum in icini y o he cen e. Fig. 11d p edic s dis ibu ion o XX s ain in he TAT es ed sample. 3.4. Mic oha dness e alua ion Mic oha dness in de o med s a e a e TAT was e alua ed in e ical (YZ plane is depic ed Fig. 1b) di ec ion. The mic oha dness (acco ding o Vicke s HV 0.2) was e alua ed in h ee indi idual measu emen s (lines 1–3 co esponding o Fig. 1c) in longi udinal and ans e se di- ec ion. Mic oha dness in line 1 (depic ed in Fig. 12a) was measu ed along he sample axis. A o al o 193 indi idual inden ions in 0.1 mm we e pe o med in his line. The cou se o he mic oha dness co e- sponds wi h he con ou (depic ed in Fig. 12a) o de o med sample. The mic oha dness copies he cou se o he con ou , which eached he highes obse ed alue (465 HV) in he icini y o he impac ed sample o ehead due o se e e plas ic de o ma ion no iced in his egion. This is an inc ease o 13,1 % wi hin his a ea, compa ed o he sample in as- p in ed s a e. The cou se shows ha he mic oha dness dec eased wi h he inc easing dis ance, which is p obably caused due o ac o s co e- sponding wi h a e age g ain size, high angle g ain bounda ies and disloca ion densi y. These ac o s we e a ec ed by se e e plas ic de o ma ion (desc ibed in Sec ion 3.5). The e alua ed mic oha dness oscilla ed wi hin cylind ical a ea (inc eased and dec eased) acco ding o he place, whe e he mic oha dness was measu ed. The s uc u e (close iew is desc ibed in ollowing chap e ) o he as-buil samples ea u ed mel ing pool s uc u e wi h wo dis inc egions, which co esponds o di e en e ching con as (see Fig. 12A2). Da ke a eas p esen ing a empe ed ma ensi e egion, which al e na e wi h b igh e a eas co e- sponding o less ex ensi e empe ing [61]. SEM image (see Fig. 12A1) e ealed a p esence o ine seconda y ca bides wi hin empe ed a ea. The mic oha dness measu emen s con i med he lowe mic oha dness o he da ke a eas (T =Tempe ed =393 ±5 HV 0.2), hen b igh e a eas o less empe ed ma ensi e (Q =Quench =420 ±5 HV 0.2), which can be conside ed as quickly quenched egion. O he cu en wo k [62] discussed he ac ha as each subsequen laye is deposi ed, he empe a u e inc eases due o hea ans e om he solidi ying laye owa ds o he p e iously deposi ed laye . The inc ease o empe a u e may cause he empe ing p ocess in neighbou ing laye s (i.e., in insic empe ing du ing he SLM p ocess). Based on acqui ed da a, he cylin- d ical a ea o as-de o med sample can be conside ed o ha e he same mic oha dness as he sample in as-p in ed s a e. In o he wo ds, he s ain a e in luenced he mic oha dness only in impac ed unnel-like a ea o he sample, as well as he cylind ical a ea wi hin he sample exhibi ed no signi ican changes. Mic oha dness in line 2 and 3 (see Fig. 12b) was measu ed pe pen- dicula o he sample axis. To al o 33 and 26 indi idual inden ions in dis ance o 0.2 mm we e pe o med in hese lines, whe e line 2 p esen s highe a e age alue (459 ±8 HV) o mic oha dness han line 3 (409 ± 7 HV). This is an inc ease o 12.2 % co esponding wi h geome ical and s uc u e changes a ec ed by se e e plas ic de o ma ion, lowe a e age g ain size and high ac ion o high-angle g ain bounda ies (HAGB, desc ibed in ollowing chap e in de ail). Conside ing he cou se o he alues in each line, acqui ed da a esul ed highe dispe sion in line 3 co esponding o di e en measu ed egion ( empe ed and quench). The dispe sion in line 2 was diminished due he o e whelming e ec o lowe a e age g ain size, high ac ion HAGB and disloca ion densi y. 3.5. S uc u e obse a ion Fig. 13 shows a ligh op ical mic oscopy image aken om he sample subjec ed o dynamic es ing ia TAT. The images ep esen he longi udinal c oss-sec ion in YZ plane. The s uc u e in he cylind ical a ea o he sample shows (see Fig. 13a) isible scanning pa hs and cha ac e is ic mel pool s uc u e ypically c ea ed by he sin e ing p ocess. The s uc u e ea u es semi-ellip ical g ains c ys allizing in he solidi ying laye , which ha e hei longe axis pe pendicula o he lase beam. The deposi ed laye is co e ed wi h he ollowing scanning pa h c ea ing he o e lapped a eas. This ype o obse ed s uc u e is ypical o c oss-sec ion made pa allel o he building di ec ion. SLM p ocess causes a nonuni o mal cooling a e, which is slowe inside he mel pool han a he bounda y esul ing in a di e en c ys alliza ion a es and undesi able seg ega ion inside he mel ing pools [63]. This seg ega ion can a ec mechanical p ope ies o pieces p oduced ia SLM. Due o lowe se laye hickness (40 μ m), i can be p edic ed ha indi idual laye s could be emel ed se e al imes. The emel ing p ocess can cause pa ial annealing o he al eady deposi ed laye esul ing in nonhomo- geneous mechanical p ope ies. S uc u e in unnel-like a ea o he sample is depic ed in Fig. 13b. De o ma ion unde s ain a e en ailed in a change o s uc u e. The shape o a mel pools exhibi ed longe main axis and sho e seconda y axis compa ed o pools in cylind ical a ea esul ing in change o mechanical p ope ies. EBSD O ien a ion image maps (OIMs) (depic ed in Fig. 14) we e cons uc ed on images cap u ed a 1000×magni ica ion ep esen lon- gi udinal (YZ plane) c oss-sec ion o he sample subjec ed o dynamic es ia TAT. The g ains a e shown in he maps, using colou s ( ed, g een and blue) ela ing o hei indi idual o ien a ions along he h ee p in- cipal c ys al di ec ions i.e., 〈001〉,〈101〉, and 〈111〉, acco ding o Mille indices ( he legend iangle is depic ed in each igu e). These o ien a ion maps a e p esen ed on each a ea I-III co esponding o Fig. 1c. The g ain o ien a ions in each a ea we e obse ed pa allel o he building di ec- ion i.e., he longi udinal images (YZ plane) we e e alua ed pa allel o he x axis (|| x). G ain s uc u e in he cylind ical a ea (depic ed in Fig. 14a) ea u ed mos ly an i egula shape wi hin his a ea I and exhibi ed no endency o o m in ce ain o ien a ion. G ain s uc u e depic ed in Fig. 14d and g p esen clus e s o ine equiaxed g ain and ew elonga ed g ains, which can be seen legibly. Conside ing he OIM and o ien a ion oge he , i appea s ha he elonga ed g ains o m pe pendicula o he sample axis. Fig. 12. Dependence o Vicke s mic oha dness on dis ance in YZ plane: (a) measu ed mic oha dness in line 1, (A1) p esence o ine seconda y ca bides, (A2) wo di e en egions i.e., empe ed, and quenched s uc u e, (b) measu ed mic oha dness in line 2 a 3. S. Kolomy e al. Jou nal o Manu ac u ing P ocesses 125 (2024) 283–294 290 Fig. 14d desc ibes elonga ed g ains in he cen e o he sample, which ea u e a endency o o m he <101>||z p e e en ial o ien a ion as well as he a ea a he bounda y o he sample (see Fig. 14g) exhibi s he same endency. Such clus e s o ine equiaxed g ains appea in icini y o he impac ed o ehead, whe e he applied s ess eached he maximum alue. The lowe impac ed s ess (co ela e wi h nume ic simula ion) on he edge o de o med egion o he sample esul ed in a lowe numbe o such clus e s o ine g ains. P asad e al. [64] obse ed c ea ion o he clus e o ine g ains in a eas which we e subjec ed o he emel ing p ocess o successi e laye s. The a e age g ain size in all a eas was examined (examina ion was ca ied ou in YZ plane conside ing he a eas I-III, see OIMs images depic ed in Fig. 14a, d, g) using g ain e e diame e (de ined as a maximum dis ance be ween wo poin s wi hin a single g ain). In each a ea I-III he a e age g ain size sligh ly a ied, bu he highes obse ed di e ence was among a eas I and II co esponding o cylind ical and unnel-like egion, espec i ely. A e age max. G ain e e diame e in a ea I (cylind ical shape o he sample, which was no signi ican ly a ec ed wi h s ain a e) was e alua ed as 10.2 μ m ( he examined g ain size in a ea I is depic ed in Fig. 14b). A ea II and III exhibi ed a s uc u e change compa ed o a ea I. Due o s ain a e a signi ican g ain e inemen was obse ed wi hin a eas II and III (each a ea II and III was chosen in unnel-like a ea o he sample). The a e age g ain size dec eased o 3.1 μ m and 3.6 μ m (see Fig. 14e and h) in a ea II and III espec i ely. This is an almos h ee imes ine s uc u e compa ed o a ea I ( he cylind ical pa o he sample). The c ys allog aphic ex u e in a eas I, II, and III was s udied conside ing he in e se pole igu es (IPFs) o YZ plane. IPFs cap u ed in a ea I (see Fig. 14c) ea u ed a dominan o ien a ion, which is o med o 〈111〉|| z. The c ys allog aphic ex u e o he as-buil sample can be a ec ed by he ene gy densi y deposi ed du ing sin e ing p ocess [65–67]. Fig. 14 and i showing cha ac e is ic IPFs o a eas II and III si ua ed in he unnel-like a ea o impac ed o ehead. Conside ing he IPFs o de o med egion in a ea II, i is e iden ha he p e alen c ys al o ien a ion along he build di ec ion was o med o <001>|| x o ien- a ion as well as in a ea III. The c ys allog aphic ex u e in hese a eas is ela ed a he o he impac de o ma ion han o he sin e ing p ocess. The HAGB ac ion (analysis based on OIM images, h eshold o a high- angle g ain bounda y conside ed o be 15) in cylind ical a ea I exhibi ed 30.6 %, which was he lowes obse ed ac ion. The obse ed HAGB exhibi ed no signi ican di e ence in a eas II and III (examined HAGB in bo h a eas we e 50.8 % and 48.2 % espec i ely). I is e iden ha a e se e e plas ic de o ma ion he inc ease o HAGB in unnel-like a ea o he sample was obse ed. Howe e , he highes HAGB ac ion was obse ed in he cen e o he impac ed o ehead. Fig. 15 shows geome ically necessa y disloca ion (GND) densi y maps es ima ed om he EBSD o ien a ion da a co esponding o a ea I- III. GNDs ha e been mapped in ield 100 ×100 μ m a ound g ain bounda ies a ec ed by s ain a e. Black pixels in each map a e co e- sponding o non-indexed egions. Colou -scale, which is p esen ed in each map illus a es he es ima ed amoun o disloca ion densi y (numbe o disloca ions/m 2 ) con ained wi hin each measu ed pixel on he map. Fig. 15a p esen s acqui ed GND in a ea I, which co esponding o he lowes alue o 11.25 p obably due o he insu icien mac oscopic s ess le el. The minimum GND densi y is in he cen e o he g ains wi h high GND accumula ion in he icini y o he g ain bounda ies. G ains, which showed a s ong endency o o m in <101>||z p e e en ial o ien a ion exhibi ed a low GND densi y on hese slip sys ems. The GND shows he highes le el 52.77 in he cen e o unnel-like a ea II (see Fig. 15b) co esponding o he highes numbe o high-angle g ain bounda ies, he lowes a e age g ain size and o he highes obse ed mic oha dness. Fig. 15c depic s a ea III, whe e he GND dec eased as well as a e age g ain size inc eased compa ed o a ea II. The se e e plas ic de o ma ion ea u ed a dominan e ec in he cen e o he unnel-like a ea o he de o med sample due o c ea ion o s uc u e wi h he lowes a e age g ain size, highes ac ion o HAGB and GND densi y. Based on ob ained da a, he mic os uc u e and geome ical changes we e in luenced by high impac eloci y accompanied wi h high s ain a e. Mos a ec ed egion (conside ing s uc u e and geome ical shape) was he unnel-like a ea in he icini y o impac ed o ehead o he sample. Howe e , he cylind ical a ea I ea u ed e y simila mic o- ha dness compa ed o as-p in ed samples. The inc ease o mic oha d- ness in he unnel-like a ea was p obably in luenced by se e al ac o s: a e age g ain size, numbe o disloca ions and HAGB ac ion. The applied impac eloci y esul ed in se e e plas ic de o ma ion, which mainly a ec ed he egion o he de o med o ehead. The a e age g ain size dec eased due o a c ea ion o he clus e o ine g ains in he i- cini y o he cen e sample axis. The c ea ion o such clus e s o ine g ains was accompanied wi h o ming o disloca ions p esen ed domi- nan ly on g ain bounda ies. The a ea II close o sample axis exhibi ed he highes HAGB ac ion, which dec ease wi h inc easing dis ance om he sample axis. These ac o s pe ec ly co ela e wi h nume ic simula- ion, which con i med he highes obse ed s ess in he cen e a ea II o he sample dec easing owa ds a ea III. Mapping o he dynamic beha iou will ob ain a signi ican da a, which can be used in nume ic simula ions. The acqui ed da a om comp essi e es ing we e imple- men ed in Johnson-Cook equa ion used o nume ic simula ion. Thanks o he J-C equa ion new dynamically loaded componen s can be designed as e and mo e accu a e (based on nume ic simula ion). App op ia e pa su ace quali y is expec ed o inc ease esis ance o dynamic cyclic loading as well as he inal po osi y o he componen . Fig. 13. Ligh op ical mic oscopy (LOM) images o a eas (a) I, (b) II p esen ed in Fig. 1c. S. Kolomy e al. Jou nal o Manu ac u ing P ocesses 125 (2024) 283–294 291 Fig. 14. OIMs o he sample subjec ed o dynamic es ia TAT in YZ plane (a) a ea I, (b) max. Fe e diame e in a ea I, (c) IPFs in a ea I, (d) a ea II, (e) max. Fe e diame e in a ea II, ( ) IPFs in a ea II, (g) OIMs in a ea III, (h) max. Fe e diame e in a ea III, (i) IPFs in a ea III. S. Kolomy e al.