Effects of Cooling Rate During Quenching and Tempering

materials

Article Eﬀects of Cooling Rate during Quenching and Tempering Conditions on Microstructures and Mechanical Properties of Carbon Steel Flange

Haeju Jo 1,2, Moonseok Kang 1, Geon-Woo Park 1 , Byung-Jun Kim 1, Chang Yong Choi 3, Hee Sang Park 3, Sunmi Shin 1, Wookjin Lee 1 , Yong-Sik Ahn 2,* and Jong Bae Jeon 1,* 1 Energy Materials and Components R&D Group, Korea Institute of Industrial Technology, Busan 46938, Korea; [email protected] (H.J.); [email protected] (M.K.); [email protected] (G.-W.P.); [email protected] (B.-J.K.); [email protected] (S.S.); [email protected] (W.L.) 2 Department of Materials Science and Engineering, Pukyong National University, Busan 48513, Korea 3 Research Center, Felix Technology Co., Ltd., Busan 46733, Korea; [email protected] (C.Y.C.); [email protected] (H.S.P.) * Correspondence: [email protected] (Y.-S.A.); [email protected] (J.B.J.); Tel.: +82-51-629-6361 (Y.-S.A.); +82-54-367-9407 (J.B.J.) Received: 30 July 2020; Accepted: 17 September 2020; Published: 21 September 2020

Abstract: This study investigated the mechanical properties of steel in flanges, with the goal of obtaining high strength and high toughness. Quenching was applied alone or in combination with tempering at one of nine combinations of three temperatures TTEM and durations tTEM. Cooling rates at various flange locations during quenching were first estimated using finite element method simulation, and the three locations were selected for mechanical testing in terms of cooling rate. Microstructures of specimens were observed at each condition. Tensile test and hardness test were performed at room temperature, and a Charpy impact test was performed at 46 C. All specimens − ◦ had a multiphase microstructure composed of matrix and secondary phases, which decomposed under the various tempering conditions. Decrease in cooling rate (CR) during quenching caused reduction in hardness and strength but did not affect low-temperature toughness significantly. After tempering, hardness and strength were reduced and low-temperature toughness was increased. Microstructures and mechanical properties under the various tempering conditions and CRs during quenching were discussed. This work was based on the properties directly obtained from flanges under industrial processes and is thus expected to be useful for practical applications.

Keywords: ﬂange; F70 steel; heat treatment; multiphase microstructure; mechanical property

1. Introduction The flange is a component used to interconnect pipelines and must provide a strong and reliable connection. Since pipelines are increasingly being developed for harsh environments such as high pressure, low temperature and corrosive atmospheres, flanges are thus also required to have excellent mechanical properties and reliability [1–5]. The mechanical properties of thick steel plates used for large structures such as flanges are controlled using post-heat treatment [6–9]. Quenching is used for the purpose of improving strength but it has a detrimental effect on low temperature toughness due to the formation of a hard secondary phase. Thus, tempering has been used to obtain a balance of strength and toughness by optimizing the microstructure [4,10]. In low carbon steels, it is well known that tempering after quenching induces decomposition of as-quenched martensite into tempered martensite and precipitates, static recovery of dislocation and release of residual stress [11–14]. Martensite after tempering loses strength but gains

Materials 2020, 13, 4186; doi:10.3390/ma13184186 www.mdpi.com/journal/materials Materials 2020, 13, 4186 2 of 16 toughness by the release of embedded carbons, which later form precipitates. Not only tempered martensite itself but also the precipitates affect mechanical properties as they vary by their type, morphology and size [11,12]. Generally appropriate control of various process parameters included tempering temperature TTEM and holding time tTEM can yield the desired mechanical properties. However, when a thick steel plate is subjected to a given TTEM and tTEM, different parts of it cool at different rates; this variation causes non-uniformity of microstructures and variation in final mechanical properties [8,9,15–17]. After hot-forging of the flange, a post-heat treatment process can also be used to control its mechanical properties, but the combination of high strength and high toughness is difficult to obtain due to the non-uniform microstructure and the complicated effects of the parameters of the tempering process. When dealing with bulk flanges of large dimensional size and complex structure, it should be taken into consideration that heat treatment would not be uniformly applied to each location of the products. However, most previous studies have tended to use standardized specimens [18–22] or sheet plates [9,16] of certain sizes to which uniform heat treatment can be applied within the specimens, so the results cannot be directly used to understand the bulk flanges. It is rather useful, especially from an industrial point of view, to investigate the effect of heat treatment on flanges by directly obtaining local properties of interest from the flanges under heat treatment processes. Therefore, this study considered 10-inch-diameter weld neck flanges at three locations selected according to cooling rates (CRs) estimated using finite element method (FEM) simulation. Microstructures were investigated at the corresponding locations on the untempered flange and on flanges that had been subjected to one of nine types of tempering conditions (3 T 3 t ). Specimens were cut from TEM × TEM each location of each flange and then subjected to tensile and hardness tests at room temperature and a Charpy impact test at 46 C. The results were used to determine the correlation between − ◦ microstructure and mechanical properties under various CRs and tempering conditions.

2. Background Low-carbon steel sheets have various microstructures and therefore various characteristics. Microstructures are distinguished into matrix and secondary phase. According to the morphological characteristics, the matrix is classified as granular bainitic ferrite (GBF), lath bainitic ferrite (LBF), acicular ferrite (AF) and quasi-polygonal ferrite (QPF), and the secondary phases are classified as martensite (M), martensite–austenite constituent (MA), degenerated pearlite/pearlite (DP/P) and carbide (C). GBF and LBF are two different morphologies of bainitic structure, based on descriptions in the literature [23,24]. GBF consists of block ferrite and a granular M/A constituent. GBF is transformed during continuous cooling and can transform over a wide range of CRs. For these reasons, GBF tends to form a major structure in low carbon steels that have cooled at various rates [17,25–27]. After GBF transformation ends, LBF is formed in the region of retained austenite with rich carbon content, but not rich enough to form M/A constituent [11]. LBF has a parallel lath shape and has relatively higher strength than GBF due to the fine laths and rich carbon content [25–27]. The transformation temperature ranges of AF and GBF are similar; typically, AF is formed at a faster CR than GBF. AF nucleates within grains in austenite as lath-shaped needles that radiate irregularly in various directions to form basket weaves. These chaotically arranged AF laths impede crack propagation so the AF microstructure produces an excellent combination of strength and toughness [17,25,27]. QPF transforms at slow CRs at high temperatures. It has higher dislocation density than polygonal ferrite and has irregular grain boundaries. Secondary phases often form at QPF grain boundaries [25]. Compared to other matrices, OPF has the lowest strength but excellent ductility. M is formed by cooling austenite at a greater than critical CR, which prevents carbon from spreading. As a result, carbon forms a solid solution and M is an unstable microstructure with high Materials 2020, 13, 4186 3 of 16 dislocation density. M has a lath phase in low-carbon steel, so it has very high hardness and strength, but M is brittle, so its presence degrades toughness. MA consists of martensite and retained austenite. The distribution of martensite in MA depends on the stability of retained austenite and affects the mechanical properties of MA [4,17,28–30]. P is a eutectoid structure composed of ferrite and cementite with a lamellar structure and forms at highMaterials temperatures 2020, 13, x FOR with PEER slow REVIEW CR. The mechanical properties of P are between those of martensite3 of 17 and ferrite. DP has almost the same mechanical properties as P. DP is transformed by low carbon concentrationdislocation density. or increased M has CR a lath compared phase in to low that-carbon of P [31 steel,]. so it has very high hardness and strength, butC M is is usually brittle, observedso its presence on grain degrades boundaries toughness. and MA impurities consists of form martensite and aggregate and retained at the austenite end of. phaseThe transformation.distribution of martensite C is also observedin MA depends inside GBFon the and stabili mayty be of finely retained precipitated austenite byand tempering. affects the Themechanical influence ofproperties C on the of mechanical MA [4,17,28 properties–30]. depends on the shape and size [32]. P is a eutectoid structure composed of ferrite and cementite with a lamellar structure and forms 3. Materialsat high temperature and Methodss with slow CR. The mechanical properties of P are between those of martensite andIn ferrite. this study, DP has low-alloy almost steelthe same (Table mechanical1) was used properties to make as 10-inch-diameter P. DP is transformed weld neckby low flanges. carbon concentration or increased CR compared to that of P [31]. The raw steel was heated at 1250 ◦C for 12 h in a furnace. Hot-forging and punching were performed C is usually observed on grain boundaries and impurities form and aggregate at the end of phase at 850 to 1250 ◦C using a hydraulic press. The completed flanges were then subjected to post-heat transformation. C is also observed inside GBF and may be finely precipitated by tempering. The treatment of the entire flange in a furnace (Figure1). All flanges were heated at a heating rate of 80 ◦C/h influence of C on the mechanical properties depends on the shape and size [32]. to 940 ◦C and held at that temperature for 4 h and quenched in a water bath that had a temperature of 25 to 50 C. Afterward, flanges were reheated at 80 C/h to T = 450, 530 or 600 C, held at that 3. Materials◦ and Methods ◦ TEM ◦ temperature for tTEM = 1, 3 or 5 h, then air-cooled. Specimens were named according to their tempering conditionsIn th (Tableis study,2). low-alloy steel (Table 1) was used to make 10-inch-diameter weld neck flanges. TheCRs raw during steel was quenching heated byat 1 location250 °C for of flange12 h in werea furnace. estimated Hot-forging using FEM and simulation punching were (Forge performed NxT 3.0, Enginsoft,at 850 to Rome, 1250 °C Italy). using Thermal a hydraulic parameters press. The for thecompleted heat transfer flanges simulation were then during subjected annealing to post- andheat quenchingtreatment were of the set entire as those flange of a low-alloyin a furnace steel (Figure (C45 steel),1). All provided flanges were by the heated commercial at a heating FEM software.rate of 80 Metal°C/h to to metal 940 °Ccontact and heldand metal at that to temperature fluid quenching for 4 were h and applied quenched to the in heat a water transfer bathsimulation. that had a Threetemperature rectangular of 25 zones to 50(20 °C. mmAfterward,19 mm) flanges were were then reheated selected at (Figure 80 °C/h2 a):to T theyTEM = had450, fast530 CRor 600 (FC) °C, × 10.6held Cat/s, that middle temperature CR (MC) for 4.6tTEM C= /1,s or3 or slow 5 h, CRthen (SC) air-cooled.2.5 C /Specimenss (Figure2b). were The named sample according size was to − ◦ − ◦ − ◦ definedtheir tempering considering conditions the production (Table of2). test specimens.

TableTable 1. Chemical1. Chemical composition composition (mass (mass %) %) of of the the present present steel steel sheet. sheet. Element Fe C Si Mn P S Cr Ni As Element Fe C Si Mn P S Cr Ni As Content Bal. 0.1605 0.246 1.266 0.0147 0.0045 0.191 0.014 0.004 Content Bal. 0.1605 0.246 1.266 0.0147 0.0045 0.191 0.014 0.004 B Ca Cu Mo N Nb Sn Ti V Al 0.0002B 0.0002 Ca Cu0.027 Mo0.082 0.0037 N Nb0.0019 Sn0.002 Ti0.0016 V0.054 Al 0.026 0.0002 0.0002 0.027 0.082 0.0037 0.0019 0.002 0.0016 0.054 0.026

Figure 1. Schematic illustration of post-heat treatment process. Figure 1. Schematic illustration of post-heat treatment process.

Table 2. Post-heat treatment conditions and symbols.

Specimen Tempering Tempering Name Temperature (°C) Time (h) Q × × 450_1 450 1 450_3 450 3 450_5 450 5 MaterialsMaterials2020 2020, 13,, 418613, x FOR PEER REVIEW 4 of4 of1716

530_1 530 1 Table 2. Post-heat treatment conditions and symbols. 530_3 530 3 530_5 Specimen Tempering530 Tempering 5 600_1 Name Temperature600 (◦C) Time (h) 1 600_3 Q 600 3 600_5 600× × 5 450_1 450 1 450_3 450 3 CRs during quenching450_5 by location of flange 450 were estimated using 5FEM simulation (Forge NxT 3.0, Enginsoft, Rome, Italy). Thermal parameters for the heat transfer simulation during annealing 530_1 530 1 and quenching were set as those of a low-alloy steel (C45 steel), provided by the commercial FEM 530_3 530 3 software. Metal to metal530_5 contact and metal to 530 fluid quenching were applied5 to the heat transfer simulation. Three rectangular zones (20 mm × 19 mm) were then selected (Figure 2a): they had fast 600_1 600 1 CR (FC) −10.6 °C/s, middle CR (MC) −4.6 °C/s or slow CR (SC) −2.5 °C/s (Figure 2b). The sample size 600_3 600 3 was defined considering600_5 the production of test spec 600imens. 5

Figure 2. (a) Schematic illustration of each specimen location according to CR. Dimensions: mm. (b) CoolingFigure 2. curve (a) Schematic of each specimenillustration obtained of each byspecimen FEM simulation location according (Forge NxT to CR. 3.0). Dimensions: Colors in (a mm.) location (b) Cooling curve of each specimen obtained by FEM simulation (Forge NxT 3.0). Colors in (a) location and (b) cooling curves indicate cooling rates; red, green and blue mean FC ( 10.6 ◦C/s), MC ( 4.6 ◦C/s) and (b) cooling curves indicate cooling rates; red, green and blue mean FC (−−10.6 °C/s), MC (−−4.6 °C/s) and SC ( 2.5 ◦C/s) respectively. and SC− (−2.5 °C/s) respectively. Microstructure observation, hardness test, tensile test and Charpy impact test were conducted using specimensMicrostructure taken observation, at three locations hardness (Figure test, tensile3a,b). test To observeand Charpy microstructure, impact test were 5 conducted10 3 mm using specimens taken at three locations (Figure 3a,b). To observe microstructure, 5 ×× 10 ×× 3 mm specimens were obtained along the vertical direction of the flange and then polished physically specimens were obtained along the vertical direction of the flange and then polished physically using using SiC paper and diamond paste. Then, they were etched using 4% Nital solution (4% NHO3 + SiC paper and diamond paste. Then, they were etched using 4% Nital solution (4% NHO3 + ethanol), ethanol), and then the matrix was observed using an optical microscope (Olympus corporation, GX51F, and then the matrix was observed using an optical microscope (Olympus corporation, GX51F, Okayama, Japan) and the secondary phase was identified using a scanning electron microscope (JEOL, Okayama, Japan) and the secondary phase was identified using a scanning electron microscope 7200F, Tyoko, Japan) and energy dispersive spectrometer (EDS) (Oxford Instruments, Nanoanalysis, (JEOL, 7200F, Tyoko, Japan) and energy dispersive spectrometer (EDS) (Oxford Instruments, Oxford,Nanoanalysis England)., Ox Usingford, these England microphotographs,). Using these micro microstructurephotographs types, microstructure were classified types according were to theirclassified morphological according featuresto their morphological and their area features fractions and were their calculated. area fractions To confirm were calculated.the presence To of MAconfirm constituents the presence in themicrostructure, of MA constituents the specimens in the microstructure, were etched the using specimens Klemm’s were I solution etched (50using mL saturatedKlemm’s aqueous I solution Na2 (S520O mL3 + saturated1 g K2S2O aqueous5) for 40 Na to 502S2O s3 and + 1 theng K2S observed2O5) for 40 using to 50 ans and optical then microscope. observed To understandusing an optical phase microscope. evolution To during understand tempering, phase evolution equilibrium during phases tempering, were predictedequilibrium based phases on themodynamicswere predicted calculation based on using themodynamics Thermocalc calculation with TCFE using 9 database Thermocalc (Thermocalc, with TCFE ver. 2020b). 9 database (Thermocalc,Hardness tests ver. were2020b). performed using a Rockwell hardness (HRC) tester (Hanbando commerce, HBD-DBEHardness SW2259, tests Seoul, were Korea)performed on specimens using a Rockwell with dimensions hardness (HRC) of 13 tester13 (Hanbando5 mm taken commerce, along the × × verticalHBD- directionDBE SW2259, of the Seoul, flange. Korea) Hardness on specimens was measured with dimensions using a C-typeof 13 × 13 scale. × 5 mm Nine taken measurements along the werevertical taken direction from each of partsthe flange. and thenHardness the maximum was measured and minimumusing a C-type values scale. were Nine discarded measurements and the average was calculated for the rest. Materials 2020, 13, x FOR PEER REVIEW 5 of 17

were taken from each parts and then the maximum and minimum values were discarded and the average was calculated for the rest. Tensile test specimens were round according to sub-size standard of ASTM-E8 and were prepared along the horizontal direction of the flange. The tests were conducted using a tensile testing machine (MTS criterio, model 45, Eden Prairie, MN, USA) and were gauged once at a crosshead speed of 1.5 mm/min at room temperature. The Charpy impact test was performed on specimens that had been machined according to ASTM-E23 standard; these samples were taken along the horizontal direction of the flange. Charpy Materialsimpact2020 ,tests13, 4186 were conducted once at −46 °C using an impact tester (ZwickRoell, PSW750+TZE, Ulm,5 of 16 Germany).

FigureFigure 3. Schematic 3. Schematic illustration illustration ofof testing location location in inflange ﬂange for samples for samples.. (a) Horizontal (a) Horizontal direction, direction, (b– (b–f)f) vertical vertical direction direction,, (c) (microstructure,c) microstructure, (d) hardness (d) hardness test (marked test as (marked orange), ( ase) tensile orange), test, ((ef) Charpy tensile test, (f) Charpyimpact impacttest. Dimensions: test. Dimensions: mm. mm.

Tensile4. Results test and specimens Discussion were round according to sub-size standard of ASTM-E8 and were prepared along theQ horizontal-condition specimens direction ofhad the a matrix flange. that The included tests were LBF, conductedAF, GBF and using QPF and a tensile secondary testing phases machine (MTSM, criterio, MA, DP/P model and C 45, (Figure Eden 4). Prairie, These microstructure MN, USA) andtypes were were gaugedclassified once by morphological at a crosshead features speed of 1.5 mm(Table/min 3) at [23 room–25,33,34 temperature.]. The area fraction of phases (Figure 5) was measured three times in each and Theaveraged Charpy using impact an image test was analyzer performed [35]. C on microstructures specimens that were had beensmaller machined than other according microstructures to ASTM-E23 standard;and very these rare, samples so C were was taken excluded along due the to horizontal difficulties direction in identification of the flange. and measurement Charpy impact of tests area were fraction. conducted once at 46 C using an impact tester (ZwickRoell, PSW750+TZE, Ulm, Germany). GBF was a− major◦ microstructure at all CRs (Table 3). As CR increased, the fractions of LBF, AF 4. Resultsand M andincreased Discussion but the fractions of QPF and DP/P decreased. Q-condition specimens had a matrix that included LBF, AF, GBF and QPF and secondary phases M, MA, DP/P and C (Figure4). These microstructure types were classified by morphological features (Table3)[ 23–25,33,34]. The area fraction of phases (Figure5) was measured three times in each and averaged using an image analyzer [35]. C microstructures were smaller than other microstructures and very rare, so C was excluded due to difficulties in identification and measurement of area fraction. GBF was a major microstructure at all CRs (Table3). As CR increased, the fractions of LBF, AF and M increased but the fractions of QPF and DP/P decreased. In FC-Q (Figure4a,b), the matrix was mainly composed of GBF (47 4.1% area fraction), followed ± by AF (16 4.0%) and LBF (13 1.0%), and the secondary phase was mostly M (17 2.0%) with small ± ± ± amounts of DP/P (5 1.5%) and MA (2 0.2%). Because of the fast CR, FC-Q had fine grains; coarse ± ± grains were not observed. Materials 2020, 13, x FOR PEER REVIEW 6 of 17

In FC-Q (Figure 4a,b), the matrix was mainly composed of GBF (47 ± 4.1% area fraction), followed by AF (16 ± 4.0%) and LBF (13 ± 1.0%), and the secondary phase was mostly M (17 ± 2.0%) with small amounts of DP/P (5 ± 1.5%) and MA (2 ± 0.2%). Because of the fast CR, FC-Q had fine Materialsgrains;2020, 13coarse, 4186 grains were not observed. 6 of 16 In MC-Q (Figure 4c,d), the matrix was mainly composed of GBF (51 ± 4.7% area fraction) but included QPF (6 ± 3.4%), and the secondary phases were M (11 ± 3.1%) and DP/P (10 ± 0.7%); In MC-Q (Figure4c,d), the matrix was mainly composed of GBF (51 4.7% area fraction) but compared to FC-Q, this was a lower area fraction of Am and a higher area fraction± of DP/P. The area included QPF (6 3.4%), and the secondary phases were M (11 3.1%) and DP/P (10 0.7%); compared fraction of MA± was still low (2 ± 0.1%), and the grains were coarser± than in FC-Q. ± to FC-Q, thisIn SC was-Q (Fig a lowerure 4 areae,f), the fraction matrix ofwas Am composed and a higher of similar area area fraction fractions of DP of /OPFP. The (39 area± 7.9% fraction area of MA wasfraction still) lowand (2GBF 0.1%),(37 ± 5.9 and%); theAF and grains LBF were were coarsernot observed. than inThe FC-Q. main secondary phase was DP/P ± In(15 SC-Q ± 1.4%); (Figure M (74 ±e,f), 1.3%) the was matrix relatively was uncommon. composed ofThe similar area fraction area fractions of MA was of the OPF same (39 as in7.9% FC- area ± fraction)Q and and MC GBF-Q. (37Grains5.9%); in SC- AFQ were and coarser LBF were than not in observed.FC-Q and MC The-Q, main because secondary in SC-Q, phase the CR was was DP /P ± (15 1.4%);slow so M the (7 specimen1.3%) wasstayed relatively at a high uncommon.temperature for The a arearelatively fraction long of time. MA was the same as in FC-Q ± ± and MC-Q.The Grains area infractions SC-Q wereof MA coarser were not than affected in FC-Q by andCR, MC-Q,unlike the because other microstructure in SC-Q, the CR types was. The slow so effect of CR on the formation of MA has not been established. Previous studies have reported that the specimen stayed at a high temperature for a relatively long time. the area fraction of MA increased [26,35] or decreased [36,37] as CR increased.

FigureFigure 4. (a– 4.c) ( Opticala–c) Optical and and (d– (fd)– scanningf) scanning electron electron micrographsmicrographs of of Q Q specimens specimens etched etched by 4% by Nital. 4% Nital. (a,d) FC (−10.6 °C/s), (b,e) MC (−4.6 °C/s), (c,f) SC (−2.5 °C/s). (LBF: lath bainitic ferrite, AF: acicular (a,d) FC ( 10.6 ◦C/s), (b,e) MC ( 4.6 ◦C/s), (c,f) SC ( 2.5 ◦C/s). (LBF: lath bainitic ferrite, AF: acicular ferrite,− GBF: granular bainitic− ferrite, QPF: quasi-polygonal− ferrite, M: martensite, DP: degenerated ferrite, GBF: granular bainitic ferrite, QPF: quasi-polygonal ferrite, M: martensite, DP: degenerated pearlite, MA: martensite–austenite constituent). pearlite, MA: martensite–austenite constituent).

Table 3. Microstructure and area fraction (%) of experimental steel according to the CR in the quenching-only treatment, Q.

Matrix Secondary Phase Name LBF AF GBF QPF M DP/P MA FC-Q 13 1.0 16 4.0 47 4.1 - 17 2.0 5 1.5 2 0.2 ± ± ± ± ± ± MC-Q 8 3.0 12 4.4 51 4.7 6 3.4 11 3.1 10 0.7 2 0.1 ± ± ± ± ± ± ± SC-Q - - 37 7.9 39 5.9 7 1.3 15 1.4 2 0.1 ± ± ± ± ± Materials 2020, 13, x FOR PEER REVIEW 7 of 17

Table 3. Microstructure and area fraction (%) of experimental steel according to the CR in the quenching-only treatment, Q.

Matrix Secondary Phase Name LBF AF GBF QPF M DP/P MA FC-Q 13 ± 1.0 16 ± 4.0 47 ± 4.1 - 17 ± 2.0 5 ± 1.5 2 ± 0.2 Materials 2020,MC13, 4186-Q 8 ± 3.0 12 ± 4.4 51 ± 4.7 6 ± 3.4 11 ± 3.1 10 ± 0.7 2 ± 0.1 7 of 16 SC-Q - - 37 ± 7.9 39 ± 5.9 7 ± 1.3 15 ± 1.4 2 ± 0.1

Figure 5. Optical micrographs of microstructure etched by Klemm’s I to distinguish MA (colored as Figure 5. Optical micrographs of microstructure etched by Klemm’s I to distinguish MA (colored as white) according to various CRs and heat treatment conditions: (a,d,g,j,m) FC (−10.6 °C/s), (d,e,h,k,n) white) according to various CRs and heat treatment conditions: (a,d,g,j,m) FC ( 10.6 C/s), (b,e,h,k,n) MC MC (−4.6 °C/s), (c,f,i,l,o) SC (−2.5 °C/s), (a–c) Q, (d–f) 450_1, (g–i) 450_5, (j–l) 600_1,− (m◦–o) 600_5. ( 4.6 C/s), (c,f,i,l,o) SC ( 2.5 C/s), (a–c) Q, (d–f) 450_1, (g–i) 450_5, (j–l) 600_1, (m–o) 600_5. − ◦ − ◦ The area fractions of MA were not affected by CR, unlike the other microstructure types. The effect of CR on the formation of MA has not been established. Previous studies have reported that the area fraction of MA increased [26,35] or decreased [36,37] as CR increased. The area fraction of MA was significantly reduced after tempering [38,39]. Q specimens (Figure5a,b,c) had numerous small MA inclusions. In 450_1 specimens, some of the relatively coarse MA remained, but the rest disappeared (Figure5d,e,f). In 450_5 specimens (Figure5g,h,i) with increased tempering time and in 600_1 specimens (Figure5j,k,l) with increased TTEM, almost all of the MA disappeared. In 600_5 (Figure5m,n,o) with increases in both TTEM and tTEM, MA was not observed. From the literature [4,10,29,30], the disappearance could be explained by decomposition of MA to ferrite matrix and cementite. Decomposition of MA could be expected to reduce hardness and strength and improve toughness. However, the area fraction of MA was 2% in all specimens before tempering (Table3) and so should not affect their mechanical properties. Materials 2020, 13, 4186 8 of 16

Tempering decomposed the secondary phases. As TTEM and tTEM increased, cementite became spheroidized and coarsened. Spheroidization of cementite is thermodynamically driven by a decrease in ferrite/cementite interfacial energy. Small cementite belongs to large cementite by Ostwald ripening. These cementites are coarsened and the total surface area to volume ratio is decreased [40,41]. M content decreased as TTEM and tTEM increased. Comparison to the Q specimen (Figure6a) indicated that the decomposition of M started in the 450_1 specimen (Figure6d); pre-existing M broke up into ferrite and cementite, and then they mixed with ﬁne cementite and precipitated after tempering. In the 450_5 specimen (Figure6g) and the 600_1 specimen (Figure6j), M was decomposed completely into ﬁne ferrite and tempered martensite. Cementite became coarsened and nearly spherical in the 600_5 specimen (Figure6m). Materials 2020, 13, x FOR PEER REVIEW 9 of 17

Figure 6. Scanning electron micrographs of (a,d,g,j,m) martensite, (b,e,h,k,n) degenerated Figure 6. Scanning electron micrographs of (a,d,g,j,m) martensite, (b,e,h,k,n) degenerated pearlite/pearlite, (c,f,i,l,o) cementite according to tempering conditions in the MC (−4.6 °C/s) location: pearlite/pearlite, (c,f,i,l,o) cementite according to tempering conditions in the MC ( 4.6 C/s) location: (a–c) Q, (d–f) 450_1, (g–h) 450_5, (j–l) 600_1, (m–o) 600_5. − ◦ (a–c) Q, (d–f) 450_1, (g–h) 450_5, (j–l) 600_1, (m–o) 600_5.

Materials 2020, 13, 4186 9 of 16

DP/P content was affected by TTEM and tTEM. DP/P content did not differ noticeably between Q (Figure6b) and 450_1 specimens (Figure6e). In 450_5 (Figure6h) and 600_1 specimens (Figure6k), pre-existing degenerated cementite of DP was spheroidized, and the edges of lamellar cementite of P became rounded. In the 600_5 specimen (Figure6n), spheroidized cementite was coarsened and non-degenerated lamellar cementite was still present. C content was not affected by TTEM or tTEM and did not differ noticeably between Q specimens (Figure6c) and 450_1 specimens (Figure6f). This indicates the occurrence of the decomposition and coarsening of C in 450_5 (Figure6i) and 600_1 specimens (Figure6l). In the 600_5 specimen (Figure6o), C was spheroidized in comparison to the specimens of 450_5 and 600_1. It was attempted to identify carbides in decomposed M during tempering and the results are depicted in Figure7. As can be seen in Figure7a,b, EDS analyses of carbides in 450_1 and 600_5 specimens clearly revealed that carbide formation was promoted with increasing tempering time and temperature. The stoichiometry of the carbides was close to M3C, cementite carbides. These results were well matched with the stable phase predicted by equilibrium thermodynamics calculation, as shown in Figure7c. Within the present tempering temperatures, cementite was predicted to be stable. This is also in good agreement with previous reports in which decomposition of M into M3C cementite carbides was observedMaterials [ 102020,42, 13]., x FOR PEER REVIEW 10 of 17

Figure 7. EDS analysis of carbides in decomposed martensite in (a) 450_1 and (b) 600_5 specimens. Figure 7. EDS analysis of carbides in decomposed martensite in (a) 450_1 and (b) 600_5 specimens. (c) (c) Calculated equilibrium phase diagram of the present alloy. Calculated equilibrium phase diagram of the present alloy.

The mechanical properties of the steels were affected by the decomposition of the secondary phase and the decomposition velocity of each type according to tempering conditions. M decomposed faster than DP/P under the same tempering conditions; this phenomenon has been reported previously [43,44]. M is unstable at high temperatures and has a pronounced tendency to decompose to ferrite and cementite, which are thermodynamically stable microstructures. Furthermore, carbon’s activation energy for diffusion is low, so carbon has a strong tendency to escape the microstructure. In contrast, DP/P is stable and has markedly lower spheroidizing velocity than that of unstable structures [40]. Moreover, tempering drives various phenomena such as the release of residual stress, decline in dislocation density and appearance of fine precipitates. These phenomena combine with the transformation of MA and decomposition of the secondary phase identified to control the mechanical properties of the steels [4,10,19,30,45]. Materials 2020, 13, x FOR PEER REVIEW 11 of 17

As the CR decreased, hardness under the same tempering condition decreased (Figure 8). Hardness decreased as a result of the change from LBF and AF to QPF in the matrix and from M to DP/P in the secondary phase. FC-Q, MC-Q and SC-Q had hardness values of 19, 15.7 and 11.7 HRC, Materials 2020, 13, 4186 10 of 16 respectively. Grain size in each specimen was not quantified, so the contribution of the grain size could not be calculated. However, as CR increased, the grain size decreased and the hardness increased,The mechanical so we infer properties that the of grain the steels refinement were a ffdependingected by the on decomposition the CR also influenced of the secondary the increased phase andhardness the decomposition [18,20,46]. velocity of each type according to tempering conditions. M decomposed faster than DPThe/P underdegree the of samehardness tempering decrease conditions;d after tempering this phenomenon and the has range been of reported the decrease previously increased [43,44 ].as MT isTEM unstable and tTEM at increased. high temperatures The range and of has reduction a pronounced was decreased tendency toas decompose the CR during to ferrite quenching and cementite, slowed. whichThe hardness are thermodynamically of FC specimens stable decreased microstructures. by 5.2 HRC Furthermore, from 19 (FC- carbon’sQ) to 14.8 activation HRC (FC energy-650_5); for the dihardnessffusion is of low, MC sospecimens carbon has decreased a strong by tendency 4.9 HRC tofrom escape 15.7 the(MC microstructure.-Q) to 10.8 HRC In (MC contrast,-650_5), DP and/P the is stablehardness and hasof SC markedly specimens lower decreased spheroidizing by 3.8 velocityHRC from than 11.7 that (SC of-Q) unstable to 7.9 structuresHRC (SC-650_5). [40]. Moreover, Previous temperingstudies [2,47 drives–49] varioushave reported phenomena that TTEM such changes as the the release hardness of residual of M, LBF, stress, P and decline ferrite in dislocationand that the densitydecreasing and appearancerange of hardness of fine precipitates.is in the order These of M, phenomena LBF and P combineand is highly with thedepend transformationent on TTEM. of In MAcontrast, and decomposition the hardness of of ferrite the secondary was not phasemuch affected identified by to tempering. control the mechanical properties of the steels [Decomposition4,10,19,30,45]. of the secondary phase increased as TTEM and tTEM increased (Figures 5 and 6). HardnessAs the CRwas decreased, reduced by hardness decomposition under the and same softening tempering of hard condition secondary decreased phases (Figure that were8). Hardness enriched decreasedin carbon. as Tempering a result of theof LBF change decreased from LBF the andhardness AF to due QPF to in the the diffusion matrix and of fromcarbon M toin DPLBF/P and in the the secondaryreductionphase. in the dislocation FC-Q, MC-Q density. and SC-Q The hadmajor hardness reasons values for the of change 19, 15.7 in and the 11.7hardness HRC, of respectively. ferrite were Grainreduced size residual in each specimen stress and was reduced not quantified, dislocation so density the contribution [47,49]. Tempering of the grain reduced size could hardness not be by calculated.releasing residual However, stress, as CR decreasing increased, the the dislocation grain size decreased density and and causi theng hardness precipitation increased, of solute so we carbon infer thatin the the microstructure grain refinement. depending on the CR also influenced the increased hardness [18,20,46].

FigureFigure 8. 8.Hardness Hardness of of ( a(a)) FC FC ( (−10.610.6 °C/sC/s),), ((bb)) MCMC ((−4.6 °C/sC/s),), ( (cc)) SC SC (−2 ( 2.5.5 °C/sC/s)) specimens specimens according according to − ◦ − ◦ − ◦ tothe the tempering tempering conditions. conditions. Tempering Tempering conditions conditions represented represented in black, red, red, green green and and blue blue mean mean

quenching,quenching, 450, 450 530, 530 and and 600 600◦C, °C respectively., respectively.

TheTensile degree strength of hardness and yield decreased strength after decreased tempering with anddecreasing the range CR ofunder the decreasethe same increasedtempering asconditionTTEM ands. ThetTEM tensileincreased. strength The range and yield of reduction strength was were decreased highest as in the FC- CRQ (730 during and quenching 578 MPa, slowed.respectively) The hardness (Figure of9a FC), followed specimens by decreased MC-Q (668, by 5.2492 HRC MPa) from (Figure 19 (FC-Q) 9b) and to 14.8SC- HRCQ (638, (FC-650_5); 455 MPa) the(Figure hardness 9c). ofAll MC specimens specimens had decreased high strength by 4.9 due HRC the from high 15.7 content (MC-Q) of GBF to 10.8 in the HRC microstructure. (MC-650_5), andMicrostructure the hardness types of SC and specimens their fractions decreased were by affected 3.8 HRC by from CR 11.7 and (SC-Q) affected to the 7.9 samples’ HRC (SC-650_5). strengths Previous(Table 3, studies Figure [4).2,47 –49] have reported that TTEM changes the hardness of M, LBF, P and ferrite and that theStrengths decreasing were range decreased of hardness after istempering in the order (Fig ofure M, LBF9), but and not P andas quickly is highly as dependenthardness (Fig on TureTEM 8.). InDecomposition contrast, the hardness of the ofsecondary ferrite was phase not much was a causedﬀected by by tempering. precipitation of solute carbon during temperingDecomposition (Figure 6). of Carbon the secondary is in solid phase solution increased and has as T aTEM strongand strengtheningtTEM increased effect; (Figures therefore5 and6 ).the Hardnessstrength wasdecreased reduced significantly by decomposition as the carbon and softening content ofin hardthe secondary secondary phase phases decreased that were due enriched to the indiffusion carbon. of Tempering carbon atoms of LBF into decreased M3C cementite the hardness carbides due during to the tempering diﬀusion of[10 carbon]. The formation in LBF and of the the reductioncarbides inwas the indeed dislocation observed density. in tempered The major specimens reasons, for as theshown change in Figure in the 7 hardness. of ferrite were reducedThe residual reduction stress ranges and ofreduced strength dislocation after tempering density under [47,49 ]. CRs Tempering were defined reduced as the hardness difference by releasingbetween residual the Q specimen stress, decreasing and the tempered the dislocation specimen density that andhad causing the low precipitationest strength. In of soluteFC specimens, carbon inthe the reduction microstructure. ranges were 80 MPa for tensile strength and 55 MPa for yield strength; in both MC and SC specimens,Tensile strength the reduction and yield ranges strength were decreased 60 MPa for with tensile decreasing strength CR and under 40 MPa the for same yield tempering strength. conditions. The tensile strength and yield strength were highest in FC-Q (730 and 578 MPa, respectively) (Figure9a), followed by MC-Q (668, 492 MPa) (Figure9b) and SC-Q (638, 455 MPa) (Figure9c). Materials 2020, 13, x FOR PEER REVIEW 12 of 17

Materials 2020, 13, 4186 11 of 16 Tempering reduced strength by decomposing hard phases, and the reduction ranges increased with the increase in the area fraction of the hard phase. The similarity of the reduction ranges of strength Allin specimens MC and SC had may high be strengtha result of due complex the high phenomena content of that GBF include in the microstructure.not only the fraction Microstructure of the hard typesphase and but their also fractions grain size were and a ﬀfineected precipitates. by CR and aﬀected the samples’ strengths (Table3, Figure4).

FigureFigure 9. 9.Tensile Tensile properties properties of of (a )(a FC) FC ( 10.6(−10.6C °C/s/s), ()b, )(b MC) MC ( 4.6(−4.6C °C/s/s),) (,c )(c SC) SC ( (−22.5.5 C°C/s/s) specimens) specimens − ◦ − ◦ − ◦ accordingaccording to to tempering tempering conditions. conditions. The The solid solid line line and and dotted dotted line line show show tensile tensile strength strength and and yield yield strength.strength. Tempering Tempering conditions conditions represented represented in in black, black, red, red, green green and and blue blue mean mean quenching, quenching, 450, 450 530, 530 andand 600 600◦C, °C respectively., respectively.

StrengthsStrain–stress were curves decreased of the after entire tempering specimen (Figure showed9), dome but not-shaped as quickly, continuous as hardness yielding (Figure behavior8). Decomposition(Figure 10). This of the phenomenon secondary phase could was be related caused to by high precipitation mobile dislocation of solute carbon density during without tempering pinning (Figureeffect6 ). of Carbon carbons is or in solidprecipitates solution [50 and–53 has]. On a strong the contrary, strengthening some previous e ffect; therefore studies the have strength shown decreaseddiscontinuous significantly yielding as the after carbon tempering content or in ageing the secondary [50–56]. phaseIt is well decreased accepted due ttohat the discontinuous diffusion of carbonyielding atoms is caused into M 3byC cementiteunpinning carbides of dislocation during from tempering the Cottrell [10]. The atmosphere, formation which of the formed carbides by was the indeedinteraction observed between in tempered dislocation specimens, and interstitial as shown atoms in Figure such7 .as carbon and nitrogen [52,54,55]. Heat treatmentThe reduction could induce ranges diffusion of strength of inter afterstitial tempering atoms into under the dislocation CRs were core defined so dislocation as the di ffiserence pinned betweenduring the deformation Q specimen and and able the to tempered escape from specimen the atmosphere that had the by lowest reaching strength. the upper In FC specimens, yield point. theMoreover, reduction this ranges discontinuous were 80 MPa yielding for tensile can strength be caused and by 55 the MPa pinning for yield effect strength; of fine in bothprecipitates MC and of SCcritical specimens, size on the dislocation reduction [ ranges53]. For were breaking 60 MPa away for tensilefrom pinning strength precipitates, and 40 MPa higher for yield stress strength. will be Temperingrequired to reduced drive dislocation. strength by Therefore, decomposing it can hard be postulated phases, and that the both reduction chemical ranges composition increased and with the thetempering increase in conditions the area fraction of this study of the may hard have phase. been The insufficient similarity of to the form reduction the Cottrell ranges atmosphere of strength or in to MCgrow and precipitates SC may be a to result the ofcritical complex size phenomena that is necessary that include for the not development only the fraction of discontinuous of the hard phase yield butbehavior. also grain size and fine precipitates. Strain–stress curves of the entire specimen showed dome-shaped, continuous yielding behavior (Figure 10). This phenomenon could be related to high mobile dislocation density without pinning effect of carbons or precipitates [50–53]. On the contrary, some previous studies have shown discontinuous yielding after tempering or ageing [50–56]. It is well accepted that discontinuous yielding is caused by unpinning of dislocation from the Cottrell atmosphere, which formed by the interaction between dislocation and interstitial atoms such as carbon and nitrogen [52,54,55]. Heat treatment could induce diffusion of interstitial atoms into the dislocation core so dislocation is pinned during deformation and able to escape from the atmosphere by reaching the upper yield point. Moreover, this discontinuous yielding can be caused by the pinning effect of fine precipitates of critical size on dislocation [53]. For breaking away from pinning precipitates, higher stress will be required to drive dislocation. Therefore, it can be postulated that both chemical composition and the tempering conditions of this study mayFigure have 10. beenEngineering insuffi cientstress– tostrain form curves the Cottrell of (a) FC atmosphere (−10.6 °C/s), or(b) to MC grow (−4.6 precipitates °C/s), (c) SC to(−2 the.5 °C/s critical) size thatspecimens is necessary according for the to development the tempering of conditions. discontinuous Line colors yield and behavior. patterns distinguish tempering temperature: black is quenching, red is 450 °C and blue is 600 °C, respectively. Solid lines are 1 h, dotted lines are 5 h (excluded quenching).

Chary V-notch (CVN) impact energy was not significantly affected by CR (Figure 11). The experimental temperature (−46 °C) was assumed to be in the ductile-to-brittle transition temperature Materials 2020, 13, x FOR PEER REVIEW 12 of 17

Tempering reduced strength by decomposing hard phases, and the reduction ranges increased with the increase in the area fraction of the hard phase. The similarity of the reduction ranges of strength in MC and SC may be a result of complex phenomena that include not only the fraction of the hard phase but also grain size and fine precipitates.

Figure 9. Tensile properties of (a) FC (−10.6 °C/s), (b) MC (−4.6 °C/s), (c) SC (−2.5 °C/s) specimens according to tempering conditions. The solid line and dotted line show tensile strength and yield strength. Tempering conditions represented in black, red, green and blue mean quenching, 450, 530 and 600 °C, respectively.

Strain–stress curves of the entire specimen showed dome-shaped, continuous yielding behavior (Figure 10). This phenomenon could be related to high mobile dislocation density without pinning effect of carbons or precipitates [50–53]. On the contrary, some previous studies have shown discontinuous yielding after tempering or ageing [50–56]. It is well accepted that discontinuous yielding is caused by unpinning of dislocation from the Cottrell atmosphere, which formed by the interaction between dislocation and interstitial atoms such as carbon and nitrogen [52,54,55]. Heat treatment could induce diffusion of interstitial atoms into the dislocation core so dislocation is pinned during deformation and able to escape from the atmosphere by reaching the upper yield point. Moreover, this discontinuous yielding can be caused by the pinning effect of fine precipitates of critical size on dislocation [53]. For breaking away from pinning precipitates, higher stress will be required to drive dislocation. Therefore, it can be postulated that both chemical composition and the tempering conditions of this study may have been insufficient to form the Cottrell atmosphere or to grow precipitates to the critical size that is necessary for the development of discontinuous yield Materials 2020, 13, 4186 12 of 16 behavior.

Materials 2020, 13, x FOR PEER REVIEW 13 of 17

increase in CVN impact energy could be possibly explained with the tempering effect which induced FigureFigure 10. 10.Engineering Engineering stress–strain stress–strain curves curves of of (a )(a FC) FC ( (10.6−10.6C °C/s/s),) (,b ()b MC) MC ( (4.6−4.6 C°C/s/s),) (,c ()c SC) SC ( (−22.5.5 C°C/s/s) ) martensite to lose its brittleness by rejecting carbons− ◦and annihilating− ◦ dislocation.− Moreover◦ , as specimensspecimens according according to to the the tempering tempering conditions. conditions. Line Line colors colors and and patterns patterns distinguish distinguish tempering tempering evidenced in Figure 6, the rejected carbons formed cementite carbides in tempered martensite and temperature:temperature: black black is is quenching, quenching, red red is is 450 450 C°C and and blue blue is is 600 600 C,°C, respectively. respectively.Solid Solidlines lines are are 1 1 h, h, those carbides were spheroidized with increasing◦ tempering◦ time and temperature. It can be dotteddotted lines lines are are 5 5 h h (excluded (excluded quenching). quenching). postulated that the carbides with high aspect ratio could cause local concentrated stress, which can lead to fracture, but spheroidized cementite relieves local concentrated stress, and thereby inhibit CharyChary V-notch V-notch (CVN)(CVN) impact energy energy was was not not signifi signiﬁcantlycantly affected aﬀected by by CR CR (Figure (Figure 11). 11 The). initiation and propagation of cracks and increases toughness (Figure 12c) [13]. Tempering also Theexperimental experimental temperature temperature (− (4646 °C)◦C) was was assumed assumed to to be be in in the the d ductile-to-brittleuctile-to-brittle transitiontransition temperature temperature − (DBTT)reduces range. dislocation Steels thatdensity had aand similar residual composition stress, and to thatthese of changes the steels also in thiscontribute experiment to the had increase DBTTs in intoughness the range of[19,6025–62C]. to 100 C, as in earlier investigations [9,16,30,57–59]. − ◦ − ◦

FigureFigure 11. 11.CVN CVN impact impact energy energy of of (a )(a FC) FC ( 10.6(−10.6C °C/s/s), ()b, ()b MC) MC ( (4.6−4.6C °C/s/s), ()c, )(c SC) SC ( (−22.5.5C °C/s/s) specimens) specimens − ◦ − ◦ − ◦ accordingaccording to to the the tempering tempering conditions. conditions. Colors Colors indicate indicate tempering tempering conditions: conditions: black, black, red, red, green green and and blueblue mean mean quenching, quenching, 450, 450 530, 530 and and 600 600◦C, °C respectively., respectively.

After tempering, all specimens under various CRs had increased CVN impact energy (Figure 10). It increased with both TTEM and tTEM. By observing the fracture surface (Figure 12a,b), it was found that tempering increased the area fraction of the dimpled surface (red lines in Figure 12a), indicating that tempering promoted ductile fracture. The size distribution of the dimple did not show any feasible diﬀerence after tempering at 450 ◦C but became irregular after tempering at 600 ◦C. This increase in CVN impact energy could be possibly explained with the tempering eﬀect which induced martensite to lose its brittleness by rejecting carbons and annihilating dislocation. Moreover, as evidenced in Figure6, the rejected carbons formed cementite carbides in tempered martensite and those carbides were spheroidized with increasing tempering time and temperature. It can be postulated that the carbides with high aspect ratio could cause local concentrated stress, which can lead to fracture, but spheroidized cementite relieves local concentrated stress, and thereby inhibit initiation and propagation of cracks and increases toughness (Figure 12c) [13]. Tempering also reduces dislocation density and residual stress, and these changes also contribute to the increase in toughness [19,60–62].

Figure 12. (a) Increased ductile-fractured surface (inside red lines) with tempering, (b) dimpled surface, (c) schematic illustration of stress relief on spheroidized carbides. Materials 2020, 13, x FOR PEER REVIEW 13 of 17

(DBTT) range. Steels that had a similar composition to that of the steels in this experiment had DBTTs in the range of −25 °C to −100 °C, as in earlier investigations [9,16,30,57–59]. After tempering, all specimens under various CRs had increased CVN impact energy (Figure 10). It increased with both TTEM and tTEM. By observing the fracture surface (Figure 12a,b), it was found that tempering increased the area fraction of the dimpled surface (red lines in Figure 12a), indicating that tempering promoted ductile fracture. The size distribution of the dimple did not show any feasible difference after tempering at 450 °C but became irregular after tempering at 600 °C. This increase in CVN impact energy could be possibly explained with the tempering effect which induced martensite to lose its brittleness by rejecting carbons and annihilating dislocation. Moreover, as evidenced in Figure 6, the rejected carbons formed cementite carbides in tempered martensite and those carbides were spheroidized with increasing tempering time and temperature. It can be postulated that the carbides with high aspect ratio could cause local concentrated stress, which can lead to fracture, but spheroidized cementite relieves local concentrated stress, and thereby inhibit initiation and propagation of cracks and increases toughness (Figure 12c) [13]. Tempering also reduces dislocation density and residual stress, and these changes also contribute to the increase in toughness [19,60–62].

Figure 11. CVN impact energy of (a) FC (−10.6 °C/s), (b) MC (−4.6 °C/s), (c) SC (−2.5 °C/s) specimens Materialsaccording2020, 13, 4186to the tempering conditions. Colors indicate tempering conditions: black, red, green and13 of 16 blue mean quenching, 450, 530 and 600 °C, respectively.

FigureFigure 12. 12.( a()a Increased) Increased ductile-fractured ductile-fractured surface surface (inside (inside red lines) red lines)with tempering, with tempering (b) dimpled, (b) dimpled surface, (surfacec) schematic, (c) schematic illustration illustration of stress reliefof stress on spheroidizedrelief on spheroidized carbides. carbides.

5. Conclusions The present work investigated the effect of heat treatment on microstructure and mechanical properties of flanges by directly obtaining local properties of interest from the flanges under the heat treatment process. Conclusions are as follows:

(1) During quenching, cooling rates varied among locations due to non-uniform cooling in ﬂanges. Microstructure was strongly aﬀected by cooling rates in a way that area fraction of either hard or soft constituent phases was determined by cooling rate. FC-Q has the highest area fraction of hard phase such as LBF, AF and M while SC-Q showed the area volume fraction of softer phases like QPF and DP/P. (2) Both strength and hardness were dependent on cooling rates; faster cooling rates induced hard phases so that hardness and strength resultantly increased. CVN impact energy at 46 C, − ◦ however, did not show clear dependence on cooling rates. (3) Tempering evidently changed microstructure by decomposing of secondary phases such as M and P and spheroidizing cementite carbide. Accordingly, hardness, strength and CVN impact energy were improved.

Author Contributions: Conceptualization, J.B.J., B.-J.K., Y.-S.A.; methodology, G.-W.P., S.S., W.L.; validation, J.B.J., B.-J.K., Y.-S.A.; materials preparation, C.Y.C., H.S.P.; investigation and analysis, H.J., M.K., J.B.J., B.-J.K.; draft preparation, H.J.; review and editing, J.B.J., Y.-S.A. All authors have read and agreed to the published version of the manuscript. Funding: This study was supported by the R&D Program of the “World Class 300 Project (R&D) (S2641289, Development of high value-added forged flanges for extreme environments based on intelligent forging systems) of the SMBA (Korea)” and also partly supported by the R&D Program of the Korea Evaluation Institute of Industrial Technology (KEIT) as “Development of Sub-merged-Arc-Welded Pipe and Extruded Pipe Made of High Manganese Steel for Cryogenic Usage in LNG Ship and Oshore Plant (10080728)”. Conflicts of Interest: The authors declare no conflict of interest. Materials 2020, 13, 4186 14 of 16

References

1. Kim, J.; Suh, Y.; Hwang, S.; Shin, S.Y. Effect of Rolling Conditions on Microstructure and Mechanical Properties of Thick Steel Plates for Offshore Platforms. Korean J. Mater. Res. 2018, 28, 478–488. [CrossRef] 2. Sayed, A.A.; Kheirandish, S. Affect of the Tempering Temperature on the Microstructure and Mechanical Properties of Dual Phase Steels. Mater. Sci. Eng. A 2012, 532, 21–25. [CrossRef] 3. Radionova, L.V.; Ivanov, V.A.; Strugov, S.S. Study of Force and Kinematic Parameters of the Low-Waste, Zero-Draft Hot Forging of Welding Neck Flanges on the Basis of a Combined Expansion-Extrusion Deforming Procedure. Mater. Sci. Forum 2019, 946, 823–831. [CrossRef] 4. Zhou, Y.; Jia, T.; Zhang, X.; Liu, Z.; Misra, R. Investigation on Tempering of Granular Bainite in an Offshore Platform Steel. Mater. Sci. Eng. A 2015, 626, 352–361. [CrossRef] 5. Jung, I. Current Status of Offshore Plant Equipment Industry. Korean Soc. Mech. Eng. 2012, 52, 36–40. 6. Filho, A.I.; Da Silva, R.V.; De Oliveira, P.G.B.; Martins, J.B.R.; Strangwood, M. Influence of Niobium and Molybdenum on Mechanical Strength and Wear Resistance of Microalloyed Steels. Mater. Res. 2017, 20, 1029–1034. [CrossRef] 7. Chen, J.; Li, C.; Ren, J.; Tu, X.; Chen, L. Strength and Toughness of Fe-1.2Mn-0.3Cr-1.4Ni-0.4Mo-C Tempered Steel Plate in Three Cooling Processes. Mater. Sci. Eng. A 2019, 754, 178–189. [CrossRef] 8. Lima, A.P.D.S.; De Faria, G.L.; Filho, V.B.D.T.; Cândido, L.C. Effect of the Chemical Homogeneity of a Quenched and Tempered C-Mn Steel Pipe on the Mechanical Properties and Phase Transformations. Mater. Res. 2019, 22, 1–11. [CrossRef] 9. Liu, H.; Zhang, H.-Q.; Li, J. Thickness Dependence of Toughness in Ultra-Heavy Low-Alloyed Steel Plate after Quenching and Tempering. Metals 2018, 8, 628. [CrossRef] 10. Xie, Z.; Ma, X.; Shang, C.; Wang, X.; Subramanian, S. Nano-Sized Precipitation and Properties of a Low Carbon Niobium Micro-Alloyed Bainitic Steel. Mater. Sci. Eng. A 2015, 641, 37–44. [CrossRef] 11. Jiang, Z.; Wang, P.; Li, D.; Li, Y. Effects of Rare Earth on Microstructure and Impact Toughness of Low Alloy Cr-Mo-V Steels for Hydrogenation Reactor Vessels. J. Mater. Sci. Technol. 2020, 45, 1–14. [CrossRef] 12. Pan, F.; Chen, H.-L.; Su, Y.-H.; Su, Y.-H.; Hwang, W.-S. Inclusions Properties at 1673 K and Room Temperature with Ce Addition in SS400 Steel. Sci. Rep. 2017, 7, 1–8. [CrossRef][PubMed] 13. Ahn, Y.S. Change of Microstructures and Mechanical Properties in Mn-Mo-Ni Low Alloy Steel with Respect to Intercritical Heat Treatment Conditions. J. Korean Inst. Met. Mater. 2000, 38, 1309–1316. 14. Ju, Y.; Goodall, A.; Strangwood, M.; Davis, C. Characterisation of Precipitation and Carbide Coarsening in Low Carbon Low Alloy Q&T Steels during the Early Stages of Tempering. Mater. Sci. Eng. A 2018, 738, 174–189. [CrossRef] 15. Hong, S.; Lee, C.; Kim, M.-C.; Lee, B.-S. Effects of Microstructure Variation on Tensile and Charpy Impact Properties in Heavy-Section SA508 Gr.3 Low Alloy Steels for Commercial Reactor Pressure Vessel. J. Korean Inst. Met. Mater. 2017, 55, 752–759. [CrossRef] 16. Cheng, T.C.; Yu, C.; Yang, T.C.; Huang, C.-Y.; Lin, H.-C.; Shiue, R.-K. Microstructure and Impact Toughness of Offshore Steel. Arch. Metall. Mater. 2018, 63, 167–172. [CrossRef] 17. Liu, W.; Zang, M.; Zheng, Z.; Wang, Q. Influence of Cooling Process on Microstructure and Property of Low-Carbon V–N Micro-Alloyed Medium and Heavy High-Strength Steel Plate. Metallogr. Microstruct. Anal. 2018, 7, 434–442. [CrossRef] 18. Jia, S.-J.; Li, B.; Liu, Q.-Y.; Ren, Y.; Zhang, S.; Gao, H. Effects of Continuous Cooling Rate on Morphology of Granular Bainite in Pipeline Steels. J. Iron Steel Res. Int. 2020, 27, 681–690. [CrossRef] 19. Saastamoinen, A.; Kaijalainen, A.; Heikkala, J.; Porter, D.; Suikkanen, P. The Effect of Tempering Temperature on Microstructure, Mechanical Properties and Bendability of Direct-Quenched Low-Alloy Strip Steel. Mater. Sci. Eng. A 2018, 730, 284–294. [CrossRef] 20. Karmakar, A.; Sahu, P.; Neogy, S.; Chakrabarti, D.; Mitra, R.; Mukherjee, S.; Kundu, S. Effect of Cooling Rate and Chemical Composition on Microstructure and Properties of Naturally Cooled Vanadium-Microalloyed Steels. Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 2017, 48, 1581–1595. [CrossRef] 21. Chen, Z.; Nash, P.; Zhang, Y. Correlation of Cooling Rate, Microstructure and Hardness of S34MnV Steel. Metall. Mater. Trans. B Process Metall. Mater. Process. Sci. 2019, 50, 1718–1728. [CrossRef] 22. Yen, H.-W.; Chiang, M.-H.; Lin, Y.-C.; Chen, D.; Huang, C.-Y.; Lin, H. High-Temperature Tempered Martensite Embrittlement in Quenched-and-Tempered Offshore Steels. Metals 2017, 7, 253. [CrossRef] Materials 2020, 13, 4186 15 of 16

23. Bramfitt, B.L.; Speer, J.G. A Perspective on the Morphology of Bainite. Metall. Trans. A 1990, 21, 817–829. [CrossRef] 24. Bhadeshia, H.K.D.H. Bainite in Steels, 2nd ed.; Taylor & Francis Group: Boca Raton, FL, USA, 2001; Volume21A, pp. 1–454. [CrossRef] 25. Krauss, G.; Thompson, S.W. Ferritic Microstructures in Continuously Cooled Low- and Ultralow-Carbon Steels. ISIJ Int. 1995, 35, 937–945. [CrossRef] 26. Kim, Y.M.; Shin, S.Y.; Lee, H.; Hwang, B.; Lee, S.; Kim, N.J. Effects of Molybdenum and Vanadium Addition on Tensile and Charpy Impact Properties of API X70 Linepipe Steels. Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 2007, 38, 1731–1742. [CrossRef] 27. Zhao, H.; Wynne, B.P.; Palmiere, E.J. Conditions for the Occurrence of Acicular Ferrite Transformation in HSLA Steels. J. Mater. Sci. 2018, 53, 3785–3804. [CrossRef] 28. Bhadeshia, H.; Honeycombe, R. Steels: Microstructure and Properties, 3rd ed.; Elsevier: Oxford, UK, 2006. 29. Josefsson, B.; Andrén, H.O. Microstructure of Granular Bainite. J. Phys. 1988, 49, 293–298. [CrossRef] 30. Jiang, Z.; Wang, P.; Li, D.; Li, Y. The Evolutions of Microstructure and Mechanical Properties of 2.25Cr-1Mo-0.25V Steel with Different Initial Microstructures during Tempering. Mater. Sci. Eng. A 2017, 699, 165–175. [CrossRef] 31. Samuel, F.H. A Crystallographic Study of Pearlite Growth in Steels. Trans. Iron Steel Inst. Jpn. 1983, 23, 403–409. [CrossRef] 32. Johnson, D.R.; Becker, W.T. Toughness of Tempered Upper and Lower Bainitic Microstructures in a 4150 Steel. J. Mater. Eng. Perform. 1993, 2, 255–263. [CrossRef] 33. Zajac, S.; Schwinnand, V.;Tacke,K.H. Characterisation and Quantification of Complex Bainitic Microstructures in High and Ultra-High Strength Linepipe Steels. Mater. Sci. Forum 2005, 500–501, 387–394. [CrossRef] 34. Bhadeshia, H.K.D.H. Developments in Martensitic and Bainitic Steels: Role of the Shape Deformation. Mater. Sci. Eng. A 2004, 378, 34–39. [CrossRef] 35. Han, S.Y.; Shin, S.Y.; Lee, S.; Kim, N.J.; Bae, J.-H.; Kim, K. Effects of Cooling Conditions on Tensile and Charpy Impact Properties of API X80 Linepipe Steels. Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 2010, 41, 329–340. [CrossRef] 36. Wang, S.C.; Yang, J.R. Effects of Chemical Composition, Rolling and Cooling Conditions on the Amount of Martensite/Austenite (M/A) Constituent Formation in Low Carbon Bainitic Steels. Mater. Sci. Eng. A 1992, 154, 43–49. [CrossRef] 37. Wang, C.; Wu, X.; Liu, J.; Xu, N. Transmission Electron Microscopy of Martensite/Austenite Islands in Pipeline Steel X70. Mater. Sci. Eng. A 2006, 438–440, 267–271. [CrossRef] 38. Reisinger, S.; Ressel, G.; Eck, S.; Marsoner, S. Differentiation of Grain Orientation with Corrosive and Colour Etching on a Granular Bainitic Steel. Micron 2017, 99, 67–73. [CrossRef] 39. Huda, N.; Midawi, A.R.; Gianetto, J.; Lazor, R.; Gerlich, A.P. Influence of Martensite-Austenite (MA) on Impact Toughness of X80 Line Pipe Steels. Mater. Sci. Eng. A 2016, 662, 481–491. [CrossRef] 40. Cree, A.M.; Faulkner, R.G.; Lyne, A.T. Cementite Particle Coarsening during Spheroidisation of Bearing Steel SAE 52100. Mater. Sci. Technol. (UK) 1995, 11, 566–571. [CrossRef] 41. Krauss, G. Steels Processing, Structure, and Performance, 2nd ed.; ASM International: Materials Park, OH, USA, 2005. 42. Gao, W.; Leng, Y.; Fu, D.; Teng, J. Effects of Niobium and Heat Treatment on Microstructure and Mechanical Properties of Low Carbon Cast Steels. Mater. Des. 2016, 105, 114–123. [CrossRef] 43. Karadeniz, E. Influence of Different Initial Microstructure on the Process of Spheroidization in Cold Forging. Mater. Des. 2008, 29, 251–256. [CrossRef] 44. Kamyabi-Gol, A.; Sheikh-Amiri, M. Spheroidizing Kinetics and Optimization of Heat Treatment Parameters in CK60 Steel Using Taguchi Robust Design. J. Iron Steel Res. Int. 2010, 17, 45–52. [CrossRef] 45. Garcia-Mateo, C.; Peet, M.; Caballero, F.G.; Bhadeshia, H.K.D.H. Tempering of Hard Mixture of Bainitic Ferrite and Austenite. Mater. Sci. Technol. 2004, 20, 814–818. [CrossRef] 46. Sahoo, R.; Kabir, A.S. Effect of Heat Treatments on Microstructures and Mechanical Properties of Ti-5553 Alloy. In TMS 2020 149th Annual Meeting & Exhibition Supplemental Proceedings; Springer: Cham, Switzerland, 2020; Volume 18, pp. 1799–1807. [CrossRef] Materials 2020, 13, 4186 16 of 16

47. Hayashi, K.; Miyata, K.; Katsuki, F.; Ishimoto, T.; Nakano, T. Individual Mechanical Properties of Ferrite and Martensite in Fe-0.16 Mass% C-1.0 Mass% Si-1.5 Mass% Mn Steel. J. Alloys Compd. 2013, 577, 593–596. [CrossRef] 48. Baltazar-Hernández, V.H.; Panda, S.; Kuntz, M.; Zhou, Y. Nanoindentation and Microstructure Analysis of Resistance Spot Welded Dual Phase Steel. Mater. Lett. 2010, 64, 207–210. [CrossRef] 49. Azuma, M.; Goutianos, S.; Hansen, N.; Winther, G.; Huang, X. Effect of Hardness of Martensite and Ferrite on Void Formation in Dual Phase Steel. Mater. Sci. Technol. (UK) 2012, 28, 1092–1100. [CrossRef] 50. Sirinakorn, T.; Uthaisangsuk, V. Effects of The Tempering Temperature on Mechanical Properties. J. Met. Mater. Miner. 2014, 24, 13–20. [CrossRef] 51. Yan, W.; Zhu, L.; Sha, W.; Shan, Y.; Yang, K. Change of Tensile Behavior of a High-Strength Low-Alloy Steel with Tempering Temperature. Mater. Sci. Eng. A 2009, 517, 369–374. [CrossRef] 52. Shen, Z.; Wang, B.; Liang, G.; Zhang, Y.; Han, K.; Song, C. Grain Boundary Pop-in, Yield Point Phenomenon and Carbon Segregation in Aged Low Carbon Steel. ISIJ Int. 2018, 58, 373–375. [CrossRef] 53. Li, Z.-M.; Li, X.; Yang, L.; Shen, Z.-Y.; Wang, B.-L.; Zhao, S.-L.; Liang, G.-F.; Song, C.-J. Effect of Coiling and Annealing Temperatures on Yield Point Behavior of Low-Carbon Steel. J. Iron Steel Res. Int. 2020, 27, 325–333. [CrossRef] 54. Han, K.; Toplosky, V.J.; Min, N.; Xin, Y.; Walsh, R.; Lu, J. Yielding and Strain-Hardening of Reinforcement Materials. IEEE Trans. Appl. Supercond. 2019, 29, 1–5. [CrossRef] 55. Ahn, T.-H.; Oh, C.-S.; Lee, K.; George, E.P.; Han, H.N. Relationship between Yield Point Phenomena and the Nanoindentation Pop-in Behavior of Steel. J. Mater. Res. 2012, 27, 39–44. [CrossRef] 56. Cottrell, A.H.; Bilby, B.A. Dislocation Theory of Yielding and Strain Ageing of Iron. Proc. Phys. Soc. Sect. A 1949, 62, 49–62. [CrossRef] 57. Hanamura, T.; Yin, F.; Nagai, K. Ductile-Brittle Transition Temperature of Ultrafine Ferrite/Cementite Microstructure in a Low Carbon Steel Controlled by Effective Grain Size. ISIJ Int. 2004, 44, 610–617. [CrossRef] 58. Xing, R.; Yu, D.; Xie, G.; Yang, Z.; Wang, X.; Chen, X. Effect of Thermal Aging on Mechanical Properties of a Bainitic Forging Steel for Reactor Pressure Vessel. Mater. Sci. Eng. A 2018, 720, 169–175. [CrossRef] 59. Choi, S.W.; Lee, W.I. Study on Fracture Behavior of Mild Steel Under Cryogenic Condition. J. Korean Inst. Gas. 2015, 19, 62–66. [CrossRef] 60. Ahn, Y.-S.; Kim, H.-D.; Byun, T.-S.; Oh, Y.-J.; Kim, G.-M.; Hong, J.-H. Application of Intercritical Heat Treatment to Improve Toughness of SA508 Cl.3 Reactor Pressure Vessel Steel. Nucl. Eng. Des. 1999, 194, 161–177. [CrossRef] 61. Yan, P.; Liu, Z.; Bao, H.; Weng, Y.; Liu, W. Effect of Tempering Temperature on the Toughness of 9Cr-3W-3Co Martensitic Heat Resistant Steel. Mater. Des. 2014, 54, 874–879. [CrossRef] 62. Davis, C.L.; King, J.E. Cleavage Initiation in the Intercritically Reheated Coarse-Grained Heat Affected Zone: Part II. Failure Criteria and Statistical Effects. Metall. Mater. Trans. A Phys. Metall. Mater. Sci. A 1996, 27, 3019–3029. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).