metals Article Effect of Ausforming Temperature on the Microstructure of G91 Steel Javier Vivas *, Carlos Capdevila, José Antonio Jimenez, Miguel Benito-Alfonso and David San-Martin Centro Nacional de Investigaciones Metalúrgicas (CENIM), Consejo Superior Investigaciones Científicas (CSIC), Avda Gregorio del Amo, 8, E 28040 Madrid, Spain; [email protected] (C.C.); [email protected] (J.A.J.); [email protected] (M.B.-A.); [email protected] (D.S.-M.) * Correspondence: [email protected]; Tel.: +34-91-553-89-00; Fax: +34-91-534-74-25 Received: 18 May 2017; Accepted: 23 June 2017; Published: 27 June 2017 Abstract: The development of thermomechanical treatments (TMT) has a high potential for improving creep-strength in 9Cr-1Mo ferritic/martensitic steel (ASTM T/P91) to operate at temperatures beyond 600 ◦C. To maximize the number of nanoscale MX precipitates, an ausforming procedure has been used to increase the number of nucleation sites for precipitation inside the martensite lath. Relative to standard heat treatments (consisting of austenitization at about 1040 ◦C followed by tempering at about 730 ◦C) this processing concept has enabled achieving a microstructure containing approximately three orders of magnitude higher number density of MX precipitates having a size around four times smaller in ASTM T/P91 steel. On the other hand; this TMT has little effect on the size and number density of M23C6 particles. The optimized microstructure produced by this TMT route proposed is expected to improve the creep strength of this steel. Keywords: creep resistant steels; carbonitrides precipitation; martensite; tempering; thermomechanical treatment; ferritic/martensitic steel; MX nanoprecipitates 1. Introduction The 9-12Cr Ferritic/martensitic (FM) steels are widely used in power generating and petrochemical plants for operating temperatures up to 620 ◦C because of their excellent combination of creep strength, thermal properties, and oxidation resistance as well as acceptable room temperature mechanical properties and also good weldability [1]. In the designs of advanced power plant components for future fission and fusion power plants, which will operate at temperatures up to 650 ◦C, the primary emphasis was placed on the development of new steel grades with superior long-term creep and thermal fatigue properties [2,3]. Several promising ferritic steels have been developed by the addition of very expensive alloying elements such as W and Co [4–7]. However, there are economic and technical advantages for using the same basic composition of ASTM T/P 91 (here after named G91) and considerably raising creep-strength of this material by the development of thermomechanical treatments (TMT) [8–11]. The creep resistance of a G91 steel results from the combination of several types of barriers to dislocation motion: a high density of martensitic lath boundaries, solid solution strengthening with elements with much larger atomic size than iron, and fine dispersion of second phase particles [12,13]. Although the overall creep strain rate is the result of the combined effects of these mechanisms, it has been reported in previous works that the creep strength is mainly increased by the precipitation of fine MX carbonitrides in the matrix [14,15], which are very effective barriers to pin dislocations, and very stable during long-term aging [16–18]. The goal of this paper is to develop a processing route to improve the creep strength of a conventional G91 steel. Prior to the optimization of TMT, thermodynamic calculations using ThermoCalc® will be carried out to foresee the most promising microstructures. In order to maximize Metals 2017, 7, 236; doi:10.3390/met7070236 www.mdpi.com/journal/metals Metals 2017, 7, 236 2 of 11 The goal of this paper is to develop a processing route to improve the creep strength of a Metals 2017, 7, 236 2 of 11 conventional G91 steel. Prior to the optimization of TMT, thermodynamic calculations using ThermoCalc® will be carried out to foresee the most promising microstructures. In order to maximizethe number the of number nanoscale of nanoscale MX precipitates, MX precipitates, an ausforming an ausforming procedure procedure will be appliedwill be toapplied increase to increasethe number the ofnumber nucleation of nucleation sites for precipitation sites for prec insideipitation the martensite inside the lath martensite relative tolath standard relative heat to standardtreatments. heat Previous treatments. works Previous have investigated works have the inve effectstigated of applying the effect a TMT of applying instead a of TMT a conventional instead of aheat conventional treatment demonstratingheat treatment thedemonstrating improvement the in creepimprovement strength in achieved creep strength with a finer achieved dispersion with ofa finerMX [ 8dispersion,9,11]. This of work MX tries[8,9,11]. to deepen This work our understanding tries to deepen of theour effect understanding of the ausforming of the effect temperature of the ausformingon the tempered temperature martensitic on microstructurethe tempered inmartensi order totic quantify microstructure the importance in order of ausformingto quantify in the the importanceTMT to optimize of ausforming the creep in strength. the TMT The to optimize conclusions the achievedcreep strength. in this The work conclusions will allow usachieved to clarify in thiswhat work processing will allow parameters us to clarify affect what the heterogeneousprocessing parameters formation affect of thermally the heterogeneous stable precipitates. formation of thermally stable precipitates. 2. Materials and Methods 2. MaterialsA commercial and Methods G91 FM steel supplied by CIEMAT (Madrid, Spain) in the form of a plate was usedA in commercial this work. TheG91 nominalFM steel chemical supplied composition by CIEMAT of (Madrid, this steel Spain) is given in inthe Table form1 .of This a plate material was ◦ usedwas receivedin this work. after The a heat nominal treatment chemical consisting composition of normalization of this steel at is 1040 givenC in for Table 30 min 1. This followed material by ◦ wastempering received at 730after Ca forheat 1 h,treatment both with consisting air cooling of tonormalization room temperature. at 1040 °C for 30 min followed by tempering at 730 °C for 1 h, both with air cooling to room temperature. Table 1. Nominal composition in wt % of Grade 91 ferritic/martensitic steel. Table 1. Nominal composition in wt % of Grade 91 ferritic/martensitic steel. Elements C Si Mn Cr Mo V Nb N Fe Elements C Si Mn Cr Mo V Nb N Fe Wt % 0.1 0.4 0.4 9.0 1.0 0.2 0.07 0.038 balance Wt %. 0.1 0.4 0.4 9.0 1.0 0.2 0.07 0.038 balance ThermodynamicThermodynamic calculations, calculations, used used as as guidel guidelinesines to to select select the the temperatures temperatures for for the the ® Thermomechanical Treatment Treatment (TMT), (TMT), were were performed performed with with the the Thermo-Calc Thermo-Calc® (Solna,(Solna, Sweden Sweden) ) softwaresoftware based based on on the CALPHAD (Computer (Computer Coupling Coupling of of Phase Phase Diagrams Diagrams and and Thermo-chemistry) Thermo-chemistry) usingusing the the database database TCFE8. TCFE8. The The new new TMT TMT approach approach proposed proposed in in this this work work is is shown shown schematically schematically in in FigureFigure 11.. CylindricalCylindrical samplessamples withwith 1010 mmmm lengthlength × 55 mm mm diameter diameter were were given given 20% 20% deformation at at −1 0.10.1 s s−1 inin a aBähr Bähr DIL DIL 805A/D 805A/D plastodilatometer plastodilatometer (TA (TA Instruments, Instruments, New New Castle, Castle, DE, DE, USA). USA). The The samples samples ◦ ◦ were heated at 5 °C/sC/s and and cooled cooled at at 50 50 °C/s.C/s. Figure 1. Thermomechanical Treatment (TMT) scheme. Figure 1. Thermomechanical Treatment (TMT) scheme. The microstructure of as-received material and material after TMT simulation treatments was analyzedThe microstructureusing optical microscopy of as-received, field material emission and gun material scanning after el TMTectron simulation microscopy treatments (FEG-SEM), was orientationanalyzed using maps optical obtained microscopy, by electron field emissionbackscatter gun diffraction scanning electron(EBSD), microscopytransmission (FEG-SEM), electron microcopyorientation (TEM), maps obtained and X-ray by electrondiffraction backscatter (XRD). St diffractionandard grinding (EBSD), and transmission polishing electron procedures microcopy were used(TEM), for and sample X-ray preparation, diffraction (XRD). which Standard included grinding a final polish and polishing with 1 micron procedures diamond were usedpaste. for Polished sample preparation, which included a final polish with 1 micron diamond paste. Polished specimens were Metals 2017, 7, 236 3 of 11 etched with a solution of 5 mL hydrochloric acid, 1 g of picric acid and 100 mL of ethyl alcohol (Vilella’s reagent) to develop the microstructural features. EBSD measurements were performed with a JEOL JSM 6500 FEG-SEM (JEOL Ltd., Tokyo, Japan) operating at 20 kV equipped with a fully automatic EBSD attachment from Oxford Instruments HKL (Abingdon, UK). Residual damage in EBSD samples from diamond polishing was removed through an additional polishing stage with colloidal (40 nm). EBSD mapping has been carried out on an area of about 900 × 400 µm2 at step sizes of 0.4 µm. The HKL Channel 5 software (Oxford Instruments, Abingdon, UK) has been used for data processing in order to obtain a graphical representation