Microstructural evolution and mechanical properties in steels treated by quenching & partitioning with the partitioning stage in teh intercritical range.

Ph.D. Thesis MICROSTRUCTURAL EVOLUTION AND MECHANICAL PROPERTIES IN STEELS TREATED BY QUENCHING & PARTITIONING WITH THE PARTITIONING STAGE IN THE INTERCRITICAL RANGE Author: Eider Del Molino Duran Bilbao, May 2022 Advisors: Javier Jesús González Martínez Maribel Arribas Tellería (cc)2022 EIDER DEL MOLINO DURAN (cc by 4.0) Ph.D. Thesis MICROSTRUCTURAL EVOLUTION AND MECHANICAL PROPERTIES IN STEELS TREATED BY QUENCHING & PARTITIONING WITH THE PARTITIONING STAGE IN THE INTERCRITICAL RANGE Author: Eider Del Molino Duran Bilbao, May 2022 Advisors: Javier Jesús González Martínez Maribel Arribas Tellería Science makes people reach selflessly for truth and objectivity; it teaches people to accept reality, with wonder and admiration, not to mention the deep awe and joy that the natural order of things brings to the true scientist. Lise Meitner ACKNOWLEDGMENTS Tesia idazten hasi nintzen momentuan ez nuen inondik inora espero atalik zailenetakoa (zailena ez bada) hau izango zenik. Asko izan baitzarete bide luze honetan, animoak eman eta nigan sinistu izan duzuenok eta ez nuke inor bidean utzi nahi. En primer lugar, quiero agradecer a Tecnalia el haberme dado la oportunidad de realizar la tesis. En especial, a mi directora de tesis de Tecnalia, Maribel Arribas, por haber puesto todo tu empeño en hacer que este viaje me fuera lo más sencillo y enriquecedor posible. Soy incapaz de imaginarme una persona más idónea para realizar ese desempeño. Por supuesto, gracias a el resto de los compañeros y compañeras de trabajo, comidas, cafés, etc., por haberme acogido desde el primer día como una más y haber hecho que estos años se pasen volando. Se me hace imposible mencionaros a todos y todas individualmente, pero sí que me gustaría darles las gracias en particular a Iñaki e Igor, por todo el tiempo dedicado a apoyarme en el desarrollo de gran parte experimental de la tesis. También quiero agradecer a la UPV/EHU la oportunidad de realizar esta tesis. A mi director, el profesor Javier Jesús González, por tu asesoramiento en todos los momentos en los que lo he necesitado y por haberme aportado tanta experiencia. Tampoco me gustaría dejar de lado al profesor José Tomás San José, por haberme ofrecido tu ayuda de una forma totalmente desinteresada desde que comencé mis estudios de Máster y hasta el día de hoy. ¡Muchas gracias a los dos! I would like to extend my gratitude to the Advanced Steel Processing and Products Research Center (ASPPRC) at Colorado School of Mines. Thanks to Casey Gilliams for the time sacrificed, without which part of this thesis would not have been possible. Thanks also to Prof. John G. Speer and Prof. Emmanuel De Moor because, although the pandemic situation has not allowed me to meet you in person, it has been a true honor to discuss the results of my thesis with experts like you. I would also like to thank Dr. Artem Arlazarov, from ArcelorMittal Maizières, for the supply of the medium Mn steels employed in this thesis and for sharing your knowledge with me. Por otro lado, ya que esta tesis se ha desarrollado en el marco de varios proyectos de investigación, quería alargar mis agradecimientos a la financiación recibida en el programa RFCS, de la Comisión Europea, para la ejecución del proyecto HIGHQP (grant number: 709855) y el programa HAZITEK, del Gobierno Vasco, para la ejecución del proyecto CHALET (expediente: ZL-2020/00434). Eskerrik asko nire kuadrilako lagunei, nigatik arduratzeagatik eta nire ausentzia (batez ere azkenengo hilabete hauetakoa) ulertzeagatik. Zuek gabe bide hau ez litzateke berdina izango. Mila esker unibertsitateko lagunei ere, zeren, nahiz eta sakrifikatutako denborak zuekin are gutxiago egoteko aukera eman didan, frogatu didazuenez, urte gutxi batzuetan eraikitako adiskidetasuna denboran mugarik gabe luzatu daitekeelako (txip-txip). Muchas gracias a mi familia. A mi amama, porque aunque no la llamo, ni visito tanto como debería (casi) siempre tiene buenas palabras para mí. A mi aita, por esa manera tan peculiar que ha tenido siempre de motivarme, que con los años he ido “entendiendo”, aunque más de una vez me haya sacado de quicio. A mi ama, por ser un pilar fundamental en mi vida y un espejo de superación en el que mirarme. Azkenengo hitzak, urte hauetan zehar gehien aguantatu nauen pertsonarentzat. Jon Ander, eskerrik asko nire oreka puntua izateagatik. Momentu onetan eta, batez ere, txarretan, irribarre bat ateratzen jakiteagatik. Indarrak transmititzeagatik. Bizi izandako une guztiengatik eta biziko ditugunengatik. Guztiei, eskerrik asko bihotz bihotzez. Muchas gracias. Thank you. LABURPENA Erresistentzia handiagoko altzairuak sortzeko eskariak, beste propietate batzuk mantentzen diren bitartean, hala nola konformagarritasun plastikoa edo soldagarritasuna, Erresistentzia Handiko Altzairu Aurreratuak garatzea eskatu du (AHHS, ingeleseko sigletan). Tesi honetan ikertutako Q&P (Quenching and Partitioning) altzairuak AHSSen hirugarren belaunaldiaren barruan daude. Altzairu horien ezaugarri nagusia austenita atxikitu balio altu samarrak eta matrize martensitikoa dituztela da. Horrek erresistentzia eta konformagarritasun handia ematen die eta, beraz, oso interesgarriak dira automobilak fabrikatzeko. Q&P altzairuen prozesatzea bi etapako tratamendu termiko baten datza. Lehenengoan, altzairua Ms-Mf (hau da, martensita hasierako tenperatura - martensita amaierako tenperatura) tarteko tenperatura lehenetsi batera hozten da, mikroegitura partzialki martensitikoa eta austenitikoa sortzeko. Bigarrenak, partizio etapa izenekoak (altzairu martensitikoen iraoketa klasikoa gogorarazten duena), austenitaren karbono edukia aberastea du helburu, martensitaren karbono edukia (partzialki) murriztuz. Horrela, karbonoz egonkortutako austenita mikroegituran atxikitzea lortzen da, giro tenperaturara hoztu ondoren. Tesi honetan, Q&P tratamendua Mn edukia duten altzairuetan aplikatu da, eta berritasuna partizio etapa tarte interkritikoari dagokion tenperatura batean egin dela da, martensitatik austenitarako transformazioaren fenomenoa eraginez. Horrela, atxikitako austenita eduki handiagoak lortu nahi izan dira eta, horretaz gain, austenita karbonoak ez ezik, manganesoak ere egonkortu du. Altzairuaren konposizio kimikoak (manganeso eta nikel eduki desberdineko altzairutzat ikertuta) eta tratamendu termikoak (tenplaketa tenperatura eta partizio denbora) bilakaera mikroegituralean, fase austenitikoaren egonkortzean eta trakzioarekiko propietateetan duten eragina ikertu da. Tratatutako altzairuen mikroegituraren karakterizazioari esker, manganeso gehien duten altzairuetan, atxikitutako austenita eduki handiak baieztatu ahal izan dira, baita fase austenitikoko manganeso aberastea ere. Gainera, partizio etaparen aurreko mikroegiturak (tenplaketa tenperaturak zehaztua) eragin handia duela ikusi da, eta baldintzetako batzuetan mikroegitura finagoa, austenita eduki handiagoak eta aparteko trakzioarekiko propietateak lortu izan dira, trakzioarekiko erresistentziaren eta elongazio totalaren arteko produktuan 30 GPa% gaindituz. Nikela gehitzea onuragarria izan da austenita atxikitu gehiago eta produktu horren balio handiagoak lortzeko. Trakzioarekiko emaitzak martensita 3.1 Materials and processing ...................................................................................... 39 3.1.1 Materials ........................................................................................................ 39 3.1.2 Design of the Q&P cycles based on dilatometry ........................................... 43 3.1.3 Application of the Q&P heat treatments: Furnaces and salt baths ............... 51 3.2 Microstructural characterization ............................................................................ 53 3.2.1 X-ray diffraction (XRD) .................................................................................. 53 3.2.2 Scanning Electron Microscope (SEM and FE-SEM) ..................................... 54 3.2.3 Electron Backscatter Diffraction (EBSD) ....................................................... 55 3.2.4 Transmission Electron Microscopy (TEM) .................................................... 56 3.3 Mechanical behavior ............................................................................................. 59 3.3.1 Tensile test .................................................................................................... 59 3.3.2 Interrupted tensile test ................................................................................... 60 3.3.3 Hardness ....................................................................................................... 60 3.3.4 Tribology: pin-on-disk (PoD) .......................................................................... 61 3.3.5 Toughness ..................................................................................................... 63 Chapter 4 Results and discussion .................................................................................... 67 4.1 Q&P applied to medium Mn and Ni steels ............................................................ 67 4.1.1 Dilatometry curves of the Q&P heat treatments ............................................ 67 4.1.2 Microstructure after the Q&P heat treatments............................................... 70 4.1.3 Tensile properties .......................................................................................... 77 4.1.4 Interrupted tensile test results ....................................................................... 81 4.1.5 Analysis of the influence of the heat treatment parameters on microstructure and austenite stabilization ............................................................................................ 82 4.1.6 Analysis of the influence of Ni on microstructure and austenite stabilization 86 4.1.7 Theoretical analysis of austenite stabilization using DICTRA ....................... 87 4.1.8 Relationship between microstructure and tensile properties ........................ 92 4.2 Q&P applied to medium and high C steels ........................................................... 96 4.2.1 Dilatometry study to select the steels and Q&P cycle to further study ......... 96 4.2.2 Dilatometry curves of the Q&P cycle applied in furnaces and salt baths ..... 97 4.2.3 Microstructure after Q&P treatment ............................................................... 99 4.2.4 Hardness ..................................................................................................... 102 4.2.5 Wear behavior ............................................................................................. 103 4.2.6 Retained austenite stability ......................................................................... 105 4.2.7 Toughness ................................................................................................... 106 Chapter 5 Conclusions ..................................................................................................... 111 5.1 Q&P applied to medium Mn and Ni steels .......................................................... 111 5.2 Q&P applied to medium and high C steels ......................................................... 112 Chapter 6 Future work ...................................................................................................... 117 6.1 Q&P applied to medium Mn and Ni steels .......................................................... 117 6.2 Q&P applied to medium and high C steels ......................................................... 118 Chapter 7 Bibliography ..................................................................................................... 121 Publications related to the doctoral thesis .................................................................... 137 Appendix A: Dilatometry .................................................................................................. 139 Appendix B: Route followed in the furnaces and salt baths ........................................ 151 Appendix C: Tensile test specimens ............................................................................... 152 Appendix D: Hardness measurements ........................................................................... 155 Appendix E: Charpy specimens ...................................................................................... 156 List of Figures Figure 1. Global Formability Diagram (2021) comparing strength and elongation of current and emerging steel grades. Courtesy of WorldAutoSteel [2]. ................................................ 4 Figure 2. Schematic of a typical Q&P cycle representing the different microstructures and C content of each phase within the cycle, where α’ is martensite, α’sec is secondary martensite and γ is austenite. ................................................................................................................... 6 Figure 3. Total elongation versus tensile strength diagrams after the application of Q&P heat treatments showing the effect of C and Mn additions [49]. .................................................. 11 Figure 4. Total elongation versus ultimate tensile strength for Q&P treated 0.3C-Mn-1.6Si steels containing various levels of manganese [86]. ........................................................... 13 Figure 5. Volume fraction of retained austenite and corresponding carbon content after full austenization, quenching to different temperatures and partitioning at 400 °C for 100 s [108]. .............................................................................................................................................. 16 Figure 6. Volume fraction of retained austenite after Q&P treatment at different partitioning temperatures and times [8]. .................................................................................................. 19 Figure 7. Microstructural evolution that undergoes during Q&P cycles showing the different phases and carbon contents of each phase expected at austenization, quenching temperature, at the end of partitioning, and final microstructure. In this figure, it is assumed that neither austenite decomposition nor carbide precipitation occurred. ........................... 20 Figure 8. SEM micrographs of a typical Q&P microstructure containing retained austenite (RA), tempered martensite (M1), secondary martensite (α’sec), and martensite-austenite islands (MA). Adapted from [117]. ........................................................................................ 20 Figure 9. Schematic of the typical processing route of medium-Mn steels, where α’ is martensite, α is ferrite and γ is austenite. ............................................................................ 25 Figure 10. Comparison of Charpy impact energies at different temperatures of Q&T and Q&P treated specimens [194]. ...................................................................................................... 30 Figure 11. Light optical metallography (LOM) images and hardness of the hot rolled microstructures. The 2Mn (a), 4Mn (b), 6Mn (c) and 6Mn2Ni (d) steels after hot rolling, and the 4Mn (e), 6Mn (f) and 6Mn2Ni (g) steels after soft annealing. ........................................ 41 Figure 12. Theoretical CCT curves of 1.2990 (a), 1.2344 (b), and 300M (c) steels obtained with JMatPro. ........................................................................................................................ 42 Figure 13. Global view on L78 RITA dilatometer. ................................................................ 43 Figure 14. Dilatometry curves representing the determination of Ac1 and Ac3 temperatures of the 2Mn (a) and 1.2344 (b) steels. In (a) Ac1 was determined as the temperature with maximum change in length just before transformation started and Ac3 was determined by the tangent method; in (b) both temperatures were obtained by the tangent method. ............. 44 Figure 15. Dilatometry curve of the 2Mn steel for the determination of Ms, Mf and martensite transformation curve (a) and the corresponding martensitic transformation curve showing the determination of the temperature corresponding to a transformation of 75% martensite (b). .............................................................................................................................................. 46 Figure 16. Dilatometry curve showing the obtention of TART of the 6Mn2Ni steel reheating from QT25. ............................................................................................................................ 48 Figure 17. Schematic of the Q&P cycles carried out in this work for medium Mn steels. ... 49 Figure 18. Schematic of the Q&P cycles carried out in this work for medium and high C steels. .............................................................................................................................................. 50 Figure 19. Sheet of the 2Mn steel prepared to carry out Q&P thermal treatment in furnaces and salt baths with one thermocouple welded in the center and other in the edge. The dimensions of the showed sheet were 250 mm length, 50 mm wide, and 1.5 mm thick. ... 51 Figure 20. Cylinder of the 300M steel employed to carry out the Q&P thermal treatment in furnaces and salt baths representing the position of the thermocouples. ........................... 52 Figure 21. Schematic representation of the route followed to carried out Q&P thermal cycles in the furnaces and salt baths. Blue and red labels and arrows represents cooling and heating stages, respectively. ............................................................................................................. 52 Figure 22. Schematic representation of Bragg diffraction of crystallographic planes. ........ 53 Figure 23. Funnel where samples were glued to be polished. ............................................ 57 Figure 24. Hand drill employed to take out 3 mm diameter discs from the samples thinned down to 100 μm. ................................................................................................................... 57 Figure 25. Sample holder for the electrolytic bath and electropolishing equipment. ........... 58 Figure 26. Intensity-Voltage curve. ...................................................................................... 58 Figure 27. Dimensions of the tensile specimens. All measurements are expressed in mm. .............................................................................................................................................. 59 Figure 28. Samples cut for XRD measurements from tensile specimens after the application of different strains. ................................................................................................................ 60 Figure 29. Vickers hardness test principle. .......................................................................... 61 Figure 30. Pin-on-disk system. ............................................................................................. 61 Figure 31. Cross-sectional areas of the wear track employed for the measurement of wear rate. ....................................................................................................................................... 62 Figure 32. Standard subsize specimen dimensions (ASTM A370). .................................... 63 Figure 33. Dilatometry curves of the applied Q&P cycles: 2Mn steel (a), 4Mn steel (b), 6Mn steel (c), and 6Mn2Ni steel (d). ............................................................................................ 68 Figure 34. Relative change in length measured during the partitioning stage for the ref cycle. .............................................................................................................................................. 69 Figure 35. Relative change in length measured during the partitioning stage for the QT10Pt1000 and the QT25-Pt1000 cycles: 6Mn (a) and 6Mn2Ni (b) steels. Evolution of the austenite formed during the partitioning (reverted γ) in the QT10-Pt1000 and the QT25Pt1000 cycles in the 6Mn (c) and 6Mn2Ni (d) steels. .......................................................... 69 Figure 36. Retained austenite contents as a function of partitioning time measured after the application of all Q&P cycles. ............................................................................................... 71 Figure 37. FE-SEM micrographs corresponding to the QT25-Pt1000 cycle for the 6Mn (a), and 6Mn2Ni (b) steels; the QT10-Pt1000 cycle for the 6Mn (c), and 6Mn2Ni (d) steels; and the QT10-Pt3600 cycle for the 6Mn2Ni steel (e). ................................................................. 73 Figure 38. EBSD scans corresponding to the QT25-Pt1000 cycle for the 6Mn (a), and 6Mn2Ni (b) steels; and the QT10-Pt1000 cycle for the 6Mn (c), and 6Mn2Ni (d) steels. ................. 74 Figure 39. TEM analysis of 6Mn2Ni steel: Micrographs and SAD patterns for the QT25 (a) and the QT10 (b) conditions; EDS analysis representing the distribution of Mn concentration for the QT25 (c) and QT10 (d) conditions, and Ni concentration for the QT25 (e) and QT10 (f) conditions; concentration profiles of Mn and Ni line scans performed within the areas marked in (a) and (b) for the QT10 (g) and QT25 (h) conditions. ........................................ 76 Figure 40. TEM micrographs of the 6Mn steel (a) and the 6Mn2Ni steel (c) after QT10-Pt1000 cycle, TEM-EDS spectrum of a globular carbide marked in (a) (b), and compositional mapping of the (c) micrograph (d). In the micrograph (c), acicular and globular carbides are visible; in the compositional map (d) a higher concentration of Mn is visible in the globular and acicular carbides............................................................................................................ 77 Figure 41. Engineering stress-strain curves obtained after Q&P treatments for the 2Mn steel. .............................................................................................................................................. 78 Figure 42. Engineering stress-strain curves obtained after Q&P treatments for the 4Mn steel. .............................................................................................................................................. 78 Figure 43. Engineering stress-strain curves obtained after Q&P treatments for the 6Mn steel. .............................................................................................................................................. 78 Figure 44. Engineering stress-strain curves obtained after Q&P treatments for the 6Mn2Ni steel. ..................................................................................................................................... 79 Figure 45. Relationship between total elongation (%) and tensile strength (MPa) obtained after Q&P cycles in 2Mn, 4Mn, 6Mn and 6Mn2Ni steels. .................................................... 81 Figure 46. Engineering stress-strain curves obtained for the QT10-Pt1000 condition and RA measured at different strains for the 6Mn (a) and 6Mn2Ni (b) steels. ................................. 82 Figure 47. TEM micrograph of the 6Mn2Ni steel after the QT10-Pt1000 condition (a); and line scans of Mn and Ni weight concentration (b), corresponding to the line shown in (a). ....... 87 Figure 48. DICTRA simulations for 6Mn2Ni steel at 640 ºC: schematic of the initial conditions of the simulations for set up (1) (a); results for set up (1), showing a comparison between the growth of austenite in QT10 and QT25 with Pt1000 and diffusion of Mn (b) and Ni (d) (the initial interface was set in the same position for QT10 and QT25), and showing the comparison between the growth of austenite with Pt300, Pt1000, and Pt3600 in QT10 and diffusion of Mn (c) and Ni (e). ............................................................................................... 89 Figure 49. DICTRA simulations for 6Mn2Ni steel at 640 ºC: schematic of the initial conditions of the simulations for set up (2) (a), and results for set up (2) configuration, where austenite nucleated at θ/α interface (b). .............................................................................................. 91 Figure 50. Tensile properties of the Q&P treated samples: yield strength (a); tensile strength (b); total elongation (c); and TEL x TS product (d). .............................................................. 93 Figure 51. Graphical representation of the relationship between the product of tensile strength and total elongation and RA for each Q&P cycle of the 6Mn and 6Mn2Ni steels. 93 Figure 52. Volume percent of each phase in the 6Mn and 6Mn2Ni steels after the application of Q&P cycles calculated based on dilatometry curves and RA measured by XRD. .......... 94 Figure 53. Evolution of normalized RA during tensile tests stopped at different strains for 6Mn and 6Mn2Ni steels after QT10-Pt1000 cycle. ...................................................................... 95 Figure 54. Dilatometry curves, hardness and RA content for the 1.2990 steel after the PT300 (a) PT400 (b) cycles; for the 1.2344 steel after the PT300 (c) and PT400 (d) cycles; and for the 300M steel after the PT300 (e) and PT400 (f) cycles. ................................................... 97 Figure 55. Dilatometry curves of the thermal Q&P cycle applied in the furnaces and salt baths for the 1.2990 and 300M steels. ........................................................................................... 98 Figure 56. Relative change in length measured during the partitioning stage for the Q&P cycle applied to the 1.2990 and 300M steels. ...................................................................... 99 Figure 57. SEM micrograph of the 300M steel after the Q&P treatment. ............................ 99 Figure 58. SEM micrographs (a) and (b), and EDS analysis of the matrix (c), globular carbide (d), and coarse carbides (e) and (f) of the 1.2990 steel after the Q&P treatment. ............ 100 Figure 59. Phase fraction as a function of temperature of (a) 1.2990 and (b) 300M steels, and composition of 1.2990 steel (c) M7C3 and (d) M23C6 phases obtained by Thermo-Calc calculations. ........................................................................................................................ 101 Figure 60. Hardness values measured in the 1.2990 and 300M steels after the application of Q&T or Q&P heat treatments. ............................................................................................ 103 Figure 61. Wear rate (mm3/N.m) calculated from the PoD tests. Comparison between Q&P and Q&T treated 1.2990 (a) and 300M (b) steels; and comparison between Q&P treated 1.2990 and 300M steels tested with the same linear velocity (c). ..................................... 104 Figure 62. RA measured in the surface and wear tracks formed during PoD characterization of the 1.2990 and 300M steels. .......................................................................................... 105 List of Tables Table 1. Short review of typical steels with Fe-C-Mn-Si/Al compositions employed in Q&P process. .................................................................................................................................. 9 Table 2. Short review of steels with Fe-C-Mn-Si compositions and Ni and/or Cr additions employed in Q&P process. ................................................................................................... 10 Table 3. Short review of compositions employed in Q&P steels with Mo and/or V additions. .............................................................................................................................................. 10 Table 4. Tensile properties and volume % of retained austenite obtained by Q&P. ........... 24 Table 5. Typical mechanical property ranges for industrially produced QP980 and QP1180. Adapted from [151]. .............................................................................................................. 27 Table 6. Replacing DP590 with QP980 allows for downgauging. Adapted from [157]. ...... 27 Table 7. Composition (wt.%) of the medium Mn steels employed in this work. .................. 40 Table 8. Composition (wt.%) of the medium and high C steels employed in this work. ...... 43 Table 9. Phase transformation temperatures (Ac1, Ac3, Ms and Mf), critical cooling rates (CCR) and the austenite measured in the quenching state of all the steels employed in this work. .............................................................................................................................................. 45 Table 10. QT25, QT10 and TART values employed in the Q&P cycles corresponding for each steel. ..................................................................................................................................... 48 Table 11. Selection of the QT, PT and Pt parameters employed in each Q&P cycle. ........ 49 Table 12. Summary of conditions carried out in each steel. ................................................ 49 Table 13. Q&P cycle parameters (QT, PT and Pt) employed in each condition and steel.. 51 Table 14. Parameters used for the pin-on-disk tests. .......................................................... 62 Table 15. Comparison between retained austenite content (RA %) measured by XRD and EBSD. Average austenite grain size (RA nm) measured by EBSD. ................................... 74 Table 16. Summary of tensile properties for the 2Mn, 4Mn, 6Mn and 6Mn2Ni steels after each Q&P treatment. ............................................................................................................ 79 Table 17. Summary of the austenite and secondary martensite values estimated and measured at the end of the partitioning and in the final microstructure. .............................. 84 6 Figure 2. Schematic of a typical Q&P cycle representing the different microstructures and C content of each phase within the cycle, where α’ is martensite, α’sec is secondary martensite and γ is austenite. Development of Q&P treatment has so far been focused to sheet processing directed to automotive industry. However, development of Q&P type microstructures in thicker sections is of interest to heavy industry, including mining and mineral processing. For such applications, the thicknesses of the steel sheets are too low and therefore steel plates are required. In that aspect, only a few studies considered the possibility of applying Q&P in steel plates [9–11]. 1.2.2 Carbon partitioning Carbon supersaturation in martensite is ordinarily eliminated by mechanisms like carbide precipitation during tempering [12,13]. In addition, the temperature which is used normally in steel treatment processes is too low for substantial amounts of carbon diffusion to occur [12]. Consequently, although carbon-enriched retained austenite in martensitic steels was not unknown [14], the thermodynamics of carbon partitioning between martensite and retained austenite has been scarcely considered. Recently, a model has been developed in order to address carbon partitioning from as-quenched martensite into austenite. This model predicts the endpoint of partitioning when martensite is in metastable equilibrium with austenite. It assumes that the diffusion of α’sec Temperature (ºC) Time (s) γ γ γ γ Cγ= Ci/ Cα’ = Ci Cγ= Ci Cγ> Ci/ Cα’ < Ci α' γ γ γ γ γ γ γ γ γ γ γγ Ac1 Ac3 Ms QT PT 7 substitutional atoms is restricted, the martensite/austenite interface is immobile, and carbide formation and carbon segregation to dislocations in martensite are assumed to be suppressed [6]. In particular, it has been recently suggested that the carbon partitioning from martensite into austenite is controlled by the constrained carbon equilibrium (CCE) criterion [12,13,15]. According to this, to reach the metastable martensite/austenite equilibrium by the completion of carbon partitioning, both equal chemical potential of carbon in martensite and austenite and conservation of all substitutional atoms in each phase are required. The results of the CCE model served to successfully propose the novel Q&P process. However, the grain size of the austenite and its mechanical stability due to the appearance of dislocations during cooling can affect both the Ms temperature [16] and the predictions of the phase fractions [17]. The morphology of the austenite and the characteristics of the martensite, such as carbon supersaturation and dislocation density [18,19], can also influence carbon diffusion, along with the coupling of austenite with neighboring phases. A later analysis [20] about the character of the interface in the partitioning stage included the possibility of a mobile interface based on experimental evidence of interface migration. In recent studies [21,22], it was found out that the interface migrates during the partitioning step. This migration was evident since they measured a significant increase in austenite fraction after partitioning as compared to the as-quenched condition. Zhong et al. [23] found that in Q&P steels with high amount of C the interface probably migrates to increase the austenite fraction, whereas in low C Q&P steels it probably migrates to reduce it. 1.2.3 Competing reactions during Q&P heat treatment In addition to carbon partitioning into austenite, other processes could occur during the partitioning step. Including decomposition of austenite to other phases, such as bainite, formation of cementite and/or transitional carbides, and carbon trapping in martensite interfaces and dislocations. 1.2.3.1 Carbides precipitation Controlling the precipitation of carbides is of vital importance to obtain the desired microstructures in a Q&P process [12]. Since the stabilization of the austenite is based on its carbon-enrichment, it is crucial that carbon is not lost in competitive reactions. Thus, it is necessary to understand and control any carbide precipitation that might occur during the Q&P cycle. The untransformed austenitic region during the first quench can vary depending on its morphology, which can be interlath-lamellar or blocky, or the composition of the steel 8 [13]. Furthermore, the exact distribution of the austenitic regions at quenching temperature is not known. Therefore, it can be concluded that the carbon escape route may vary for each case. In the martensite, fine transition carbides generally are not considered detrimental, whereas cementite can be more concerning. Therefore, the greatest effort has been made to understand when transition carbides are replaced by cementite [24,25], rather than the initiation of transition carbides precipitation itself. However, any transition carbide precipitation decreases austenite carbon-enrichment potential in Q&P processing and, hence, it is necessary to develop a better understanding of the initiation of transition carbide formation, including composition and processing effects [26,27]. Carbon chemical potential is quite higher in as-quenched martensite than in retained austenite, so the carbide nucleation is more likely to occur in BCC ferrite than in austenite [27,28]. The α/γ interface is also a propitious site for carbide formation [28]. Carbon trapping at crystallographic defects, and austenite decomposition to ferrite and cementite as a result of martensite tempering during Q&P is consistent with the lower experimental austenite fractions often achieved compared with predictions [18,19,29–31]. 1.2.3.2 Other phase transformations Apart from carbide precipitation, austenite decomposition can also take place during the partitioning. The normally employed range of partitioning temperatures (350 ºC – 450 ºC) enables the bainite-ferrite formation from austenite [32]. Some authors demonstrated that bainitic transformation in austenite–martensite mixture is faster than that starting from singlephase austenite [33,34]. The authors suggested that the boundary between martensite and austenite serves as a nucleation site for bainitic ferrite. Bainite formation decreases the amount of austenite available for carbon enrichment and therefore is often undesirable [35]. Some researchers assume that small amounts of bainite are formed mostly in the beginning of partitioning [36]. Further enrichment of remaining austenite with carbon lowers the Bs temperature which should significantly suppress the bainitic reaction so that eventually Bs can drop below the partitioning temperature and bainitic reaction can stop completely. In some recent studies [37–39], in which the employed partitioning temperature was higher than usual, decomposition of untransformed austenite into perlite was observed. The rapid carbon enrichment of austenite and the higher partitioning temperature promote pearlite formation at prior austenite grain boundaries and cementite precipitation within austenite 9 films [38]. This transformation is obviously undesired, as consumes carbon for austenite enrichment and decreases the retained austenite content. The last competing reaction is the formation of secondary martensite during the final cooling. If competitive reactions occurred during the partitioning stage, part of the austenite might not be sufficiently carbon-enriched to be stable at room temperature and, consequently, it might transform into secondary martensite at the last cooling [13,38,40,41], which was considered as detrimental for ductility [40,42]. 1.2.4 Alloying elements employed in Q&P steels In this section the influence of different alloying elements on Q&P process and resultant properties is analyzed. The role of alloying elements in Q&P steels could be described in this way: • Prevent any competing reaction such as carbide precipitation or decomposition of austenite into ferrite, bainite or pearlite during partitioning [6,13,43]; the addition of silicon as alloying element is the key to obtain the desired results. • Improve the hardenability and final mechanical properties of the steel [43,44]; the addition of manganese is the simplest key in this question, existing other possibilities. Alloying elements also influence on phase transformation temperatures (Ac1, Ac3, Ms, and Mf), which are important to design Q&P cycles. From the industrial point of view, the lower these temperatures, the more reduction in the energy consumption and costs. In Table 1, Table 2 and Table 3 a short review of typical compositions containing the elements discussed in the next sections is shown. The typical Q&P steels contain Fe, C, Mn and Si, while other elements could be also added depending on the purpose. Table 1. Short review of typical steels with Fe-C-Mn-Si/Al compositions employed in Q&P process. Studies with Fe-C-Mn-Si/Al compositions Ref. C Mn Si Al [45] 0.25 3.00 1.50 - [46] 0.29 3.00 1.40 - [47] 0.25 3.00 1.50 0.02 [48] 0.20 3.50 1.54 - 0.20 3.50 0.45 0.22 [49] 0.2 3.00 1.60 0.06 0.29 2.95 1.59 0.06 0.28 4.95 1.64 0.06 [50] 0.29 1.49 1.47 0.26 [51] 0.19 1.61 0.35 1.10 10 Table 2. Short review of steels with Fe-C-Mn-Si compositions and Ni and/or Cr additions employed in Q&P process. Studies with Fe-C-Mn-Si compositions and Ni and/or Cr additions Ref. C Mn Si Ni Cr [52] 0.19 6.00 - - - 0.19 5.80 1.40 - - 0.19 6.00 - 2.10 - 0.19 5.70 1.40 1.60 - [53] 0.24 1.38 1.39 0.03 - 0.21 1.44 1.44 1.01 - 0.28 1.41 1.46 1.99 - [54] 0.2 1.54 1.30 0.07 1.48 0.19 1.52 1.32 1.53 0.01 [55] 0.21 4.00 1.60 - 1.00 [56] 0.22 1.30 0.25 - 0.20 0.27 1.50 1.61 - - 0.28 1.46 1.58 - 0.97 [57] 0.12 0.87 0.26 - 12.00 Table 3. Short review of compositions employed in Q&P steels with Mo and/or V additions. Studies with Mo and/or V additions Ref. C Mn Si Ni Cr Mo V [58] 0.43 0.59 2.03 0.07 1.33 0.03 - 0.43 0.59 2.03 0.07 0.03 - [59] 0.37 2.27 2.45 1.47 0.80 0.58 - 0.22 2.42 2.49 1.39 0.72 0.49 - 0.39 2.39 2.64 4.83 0.78 0.50 - 0.28 2.37 3.00 5.30 0.71 0.60 - [60] 0.20 1.50 1.50 - - 0.13 - [61] 0.21 1.65 1.67 - 0.03 - 0.20 0.22 1.89 0.52 - 0.03 - - [62] 0.24 1.90 1.50 - - - - 0.24 1.88 1.47 - - - 0.03 0.24 1.83 1.52 - - - 0.16 [9] 0.10 1.51 1.48 - - 0.30 0.04 1.2.4.1 Carbon C is important in Q&P steels because it helps retaining austenite and significantly increases the strength of the martensite [63]. C also decreases Ms temperature and therefore, if sufficient austenite carbon-enrichment occurs, austenite retention at room temperature is enhanced. Morito et al. [64] reported that, increasing C content up to 0.6 wt.%, also increased dislocation densities of as-quenched martensitic microstructures. These results suggest that the strengthening from increasing C content may be strongly correlated with the resulting 11 dislocation density. However, the C content is generally kept at levels in which the carbon equivalent remains in the range of weldable steels. The effect of increasing C content from 0.2 wt.% to 0.3 wt.% on tensile properties of Q&P steels has been studied in [49] for a steel containing 3 wt.% Mn and 1.6 wt.% Si. In the fully austenitized condition, the 0.2 wt.% C steel exhibited ultimate strength levels of 1200-1450 MPa and total elongations of 9–15% and increasing it to 0.3 wt.% increased the ultimate tensile strength levels into a range of 1400–1700 MPa and also total elongation levels to 11–17%. The results obtained in the study are represented in Figure 3. Figure 3. Total elongation versus tensile strength diagrams after the application of Q&P heat treatments showing the effect of C and Mn additions [49]. 1.2.4.2 Silicon and Aluminum It is known that the formation of undesirable cementite can be inhibited by adding Si, Al, or their combination. Most of the Q&P studies employ Si-bearing grades with 1–2.5 wt.% Si (typically around 1.5 wt.% Si is added) [56,65]. Si plays an important role in carbon precipitation suppression during Q&P process [62,66]. Since the solubility of Si in cementite is negligible, for cementite precipitation to occur during tempering, Si must diffuse into the matrix from the carbide-matrix interface. Simultaneously, carbon needs to diffuse in the opposite direction. At low Si levels, this process is controlled by carbon diffusion, while at higher Si levels, Si diffusion begins to control cementite precipitation. Furthermore, Owen [67] reported that higher tempering temperatures are required for the decomposition of austenite with higher Si contents. Earlier researchers showed that Si retards or even eliminates bainite formation in the partitioning step due to the near-to-zero 12 solubility of Si in the cementite phase [68]. These studies suggest that between 1.25 and 2 wt.% Si is required to suppress bainite formation during reheating [69–72]. Si content may affect the final mechanical properties. However, increasing the Si amount to 1.5–2 wt.% does not seem to considerably affect hardness values. By contrast, other investigations reported a slight enhancement of hardness [69,73]. The problem associated with high Si contents is the formation of Si superficial oxide, together with the iron oxide, during the hot rolling. These oxides are difficult to remove by pickling and cause surface finish problems by reducing hot dip galvanizability [74,75]. A possible alternative is the use of Al which, like Si, is not soluble in cementite and hence prevents carbide formation without deterioration of galvanizability [35]. Aluminum alloying has been shown to result in significant retained austenite fractions, and to increase austenite carbon content in TRIP steels [35,76,77]. However, utilization of aluminum in steels for Q&P processing is limited because it has such a strong effect on the Ac3 temperature that only the intercritical initial heating is possible. The response of both Si and Al in Q&P steels has been investigated in multiple studies [48,50,51,78–80], as displayed in Table 1. In [80] the results show that increasing Al contents leads to a decrease in retained austenite fraction, strength and elongation levels, comparing with Si alloyed steel. Santofimia et al. [48] studied two Q&P steels with different Si and Al contents and they found that the partial substitution of Si by Al lead to lower retained austenite and higher cementite fractions. Furthermore, they reported a decrease in the strength levels. Ande and Sluiter [81] conducted a density functional theory (DFT) calculation and they showed that Si is almost twice more effective than Al suppressing carbides. Al has also been reported to delay strain-induced martensite transformation, enhancing TRIP effect [76,82,83]. 1.2.4.3 Manganese Austenite stabilization is mainly controlled by C enrichment although other austenite stabilizing elements, such as Mn, can be effective as well [66]. Mn is a common alloying element in this type of Q&P steels since it has effective austenite stabilization capacity, it increases hardenability by retarding austenite to ferrite, perlite or bainite transformation during initial fast cooling, it helps lowering phase transformation temperatures (strongly decreases the Ms temperature), and contribute to the overall strength of the steel by solid solution strengthening [84]. Mn contents added to Q&P steels are typically in the order of 1.5 wt.%, with the aim of avoiding diffusional transformations during the quenching step. 13 Furthermore, since Mn retards bainitic reaction can be intentionally added to avoid competing reactions also in the partitioning. Higher contents of Mn are considered to enhance austenite retention, employing the austenite stabilizing effects of Mn to adjust the relative amounts of martensite and austenite [85]. In recent Q&P studies [37,45–49,55], interesting tensile properties were obtained with steels containing increased Mn contents. For example, the use of Mn additions of 3 wt.% and 5 wt.% in Q&P steels has been successful in developing very high strength cold-formable sheet steels with tensile strengths exceeding 1500 MPa, in combination with total elongations above 20% (Figure 4) [86]. However, an earlier study of the same authors showed that, after a partitioning stage at 400 ºC, the 5 wt.% Mn containing steel showed a microstructure with untempered martensite, leading to very low ductility with most samples failing at strains lower than 2% [85]. Figure 4. Total elongation versus ultimate tensile strength for Q&P treated 0.3C-Mn-1.6Si steels containing various levels of manganese [86]. High Mn contents lead to the development of strong band structure, which might reduce TRIP effect and, consequently, the ductility of the steel [84]. As a medium carbide-forming element, Mn will decrease the diffusivity of C in austenite and thus, slow down the growth kinetics of bainitic ferrite [87]. 1.2.4.4 Chromium and Nickel Additions of Ni and Cr have both been reported to increase retained austenite amounts after Q&P [56,58]. Pierce et al. [54] also observed that a comparable fraction of austenite was stabilized in the alloy with Cr with respect to the alloy with Ni, being Cr rather cheaper. Cr additions up to ~7 wt.% are reported to be austenite stabilizers in Fe-Cr binary steels, while higher amounts starts to stabilize ferrite rather than austenite [90]. The reasons why Cr 14 is a good austenite stabilizer is because it lowers Ms temperature and considerably reduces the C diffusivity in austenite, increasing the resistance of martensite to tempering [88], slowing austenite decomposition kinetics [89] and, consequently, delaying bainite transformation [56]. E. J. Seo et al. [56] analyzed the role of Q&P treatment on a Si (1.6 wt.%) and Cr (1 wt.%) added medium Mn steel. They observed that the addition of Cr resulted in an increase in the retained austenite amount and, consequently, significantly improved plasticity without decreasing of the strength. On the other hand, Ni also reduces the Ms temperature but is a significantly stronger austenite stabilizer than Cr and enhances the kinetics of C diffusion in austenite [88]. Additionally, it can assist in controlling major competing reactions, such as precipitation of carbides inside austenite and pearlite formation, and in increasing thermal stability of austenite [52]. Increasing wt.% Ni also lead to grain refinement and increases the retained austenite fraction in volume. However, due to the higher cost, Ni content should be kept to a minimum [91]. Rizzo et al. [59] observed that a high Ni content steel, apparently having larger amounts of retained austenite relative to the other experimental alloys, had less desirable strength/ductility combinations. On the contrary, other researchers reported that lower additions offer better mechanical properties because of a large fraction of retained austenite in volume [56,92,93]. Finally, Kibum et al. [53] observed that increasing Ni content the amount and stability of retained austenite increased. The Ni addition also led to grain refinement. The strain-induced transformation kinetics were retarded and they also reported solid-solution strengthening effect. Consequently, an improvement in tensile strength and elongation, and hardness was achieved. 1.2.4.5 Molybdenum and Vanadium Normally, the composition of the Q&P steel does not contain any carbide forming elements, such as Nb, V, Ti, Mo, etc. This means that the Q&P steel excludes the advantage of precipitation strengthening of carbides and fine-grain strengthening [94]. However, there are several studies which investigate the influence of these elements in Q&P process [65,86– 95]. The addition of Mo and/or V to the compositions of Q&P steels pursues the strengthening produced by carbide precipitation. Some studies carried out an additional tempering after the quenching and partitioning process to further promote this [10,98,99]. Mo is also reported to improve the hardenability of the steel [9]. Another property of both Mo and V is that they act as austenite grain refinement [9,60,97,98]. Regarding mechanical properties V addition 15 improves strength and work hardening [60] and also wear resistance [95] of the Q&P treated steels. Zhang et al. [60] also reported that V addition improved the ductility by refining the grain size and enhancing the austenite stability. 1.2.5 Q&P cycle parameters The original Q&P process (Figure 2) can be summarized in four main steps: • Austenization: In this step the steel is fully austenitized (or, sometimes, partially, in order to introduce ferrite into the microstructure with the aim of decreasing yield strength and raising elongation [100]). The initial microstructure [101,102], soaking temperature (ST) [103,104] and soaking time (St) [103] influence on the characteristics of the austenite formed in this step. • Quenching: The steel is quenched from the ST to a temperature between Ms and Mf, called quenching temperature (QT), with the aim of obtaining a partially martensitic and partially austenitic microstructure (and partially ferritic in the case of the partially austenitized steels). The rate of cooling during quenching is important because it is necessary to prevent the formation of lower bainite in case of cooling after full austenization. • Partitioning: During this step the carbon migrates from martensite into austenite. In one-step Q&P cycles the partitioning temperature (PT) is the same than QT, whereas in two-step Q&P cycles it is higher, and usually above Ms. In the time that steel remains in PT, called partitioning time (Pt), competitive reaction such as carbide precipitation or austenite decomposition might occur. • Final cooling: The steel is cooled to room temperature. If austenite is sufficiently carbon-enriched, secondary martensite formation is supposed to be avoided. The control of processing parameters such as quenching temperature, partitioning temperature and partitioning time is crucial to obtain the desired microstructures, and therefore mechanical properties. Hence, numerous studies have been carried out to investigate the influence of these parameters. 1.2.5.1 Quenching temperature (QT) Quenching temperature determines the austenite and martensite fractions for the following partitioning step. If QT is too low, near to Mf, a small amount of austenite will be available for carbon-enrichment and final retention. On the contrary, a high QT, slightly below Ms, left little martensite to provide carbon to the austenite and, consequently, austenite is unlikely to be 22 is stable. On the other hand, Si helps to stabilize austenite by suppressing cementite precipitation, thus leaving more C available to enrich the austenite [122]. Therefore, increasing the addition of austenite stabilizer elements can lead to a higher retained austenite amount. Carbon content in austenite depends on partitioning parameters and, typically, was found to reach ~1 wt.%. However, as mentioned above, C gradients may appear within either bulk of steel or across a single austenite grain and consequently the less carbon-enriched areas may transform into secondary martensite during final cooling. Stabilization of austenite during partitioning step compromised two important stages: carbon diffusion from martensite into austenite, which in general is a quick stage (it can take place in less than 1 s), and homogenization of carbon within the austenite, which was found to need much longer time [123]. DICTRA simulation can be of great help when studying these two stages [38,124]. Hidalgo et al. [38] simulated carbon partitioning at the partitioning stage in a 0.3C-4.5Mn-1.5Si steel and they found that, at 400 ºC, less than 1 s was needed for carbon partitioning to occur. However, 50 s were necessary for its distribution to become homogenized in a 100 nm thickness austenite lath. In addition, some recent studies reported that the stability of retained austenite also depends on its morphology and therefore on the locations where it was formed during microstructure evolution. Numerous studies proved the coexistence of blocky and film-like retained austenite morphologies, being reported the film-like austenite as the most stable. Sun et al. found three types of retained austenite in 0.2C–1.5Si–1.9Mn steel, including austenite films located between the martensite laths with a width of about 100 nm, blocky austenite incorporated into ferritic matrix, and ultrafine austenite films, of about 20–30 nm thick, between the plates of ferritic bainite [106]. On the other hand, Arlazarov et al. [125] investigated the role the grain size plays in thermal stability of austenite in a Medium-Mn steel and they found that a grain size smaller than 0.5 μm clearly lowered the Ms, increasing austenite stability. 1.3.2 Mechanical stability of the austenite The mechanical stability of retained austenite is the stability against the formation of martensite under the influence of deformation or transformation-induced plasticity (TRIP) effect. The TRIP effect increases the work hardening rate and delays necking, thereby improving uniform elongation (UE). The studies performed with Q&P steels often indicate no direct correlation between volume fraction of retained austenite and ductility of steel, which 23 means that the role of stability of the austenite looks more important than that of its volume fraction. If the austenite into martensite phase transformation occurs at small strains it is unable to retard the necking of the material. On the contrary, if it is to stable (i.e., no phase transformations at large strains) then it will not contribute to the TRIP-effect [126,127]. The stability of RA in steels that exhibit the TRIP effect is affected by chemical composition [126] (mainly C content), grain size [128], morphology [129], and surrounding microstructure [130]. Therefore, a better understanding of RA stability can aid in designing a microstructure that will provide the desired combination of strength and formability. Jacques et al. [126] performed experiments on steels with low and high silicon contents and similar amounts of RA. The difference between the RA in the two steels consisted in the C content and the amount of the surrounding phases, being the grain sizes similar in both steels. The high silicon steel’s austenite had the higher C content, due to the carbide precipitation occurred in the low Si steel, and the studies demonstrated that higher C content decreased the austenite transformation rate during straining. Therefore, they concluded that the high silicon steels had a slower RA transformation rate during straining due to the more carbon-enriched RA required a higher amount of strain to transform into martensite. Xiong’s et al. [129] researches revealed that although blocky austenite had higher carbon content (1.14 wt.%) compared to that in film-like austenite (0.64 wt.%) located between martensite laths, “high C” blocky RA appeared to be less resistant to martensite transformation. The blocky austenite began to transform into martensite at tensile strain of 2% and by 12% strain all blocky austenite had transformed. Nevertheless, numerous film-like austenite grains were still present at that strain, indicating that carbon content is not always a dominating factor. This could be explained by de differences in yield strength of the surrounding phases. In the work [129] the Q&P cycle was started with a partial austenization in the intercritical range, and blocky austenite was surrounded by ferrite, while film-like austenite was surrounded by higher yield strength martensite laths and the martensite transformation requiring volume expansion could be suppressed. The other factor is grain size effect. Generally, it was reported that coarse austenite grains are less stable than the finer ones. In [83] it was studied the influence of grain size on austenite stability of Q&P steels and they concluded that the optimal austenite grain size was between 0.01 and 1 μm. RA grain sizes above 1 μm transformed into austenite at quite low strains, whereas a size below 0.01 μm led to a too stable austenite that did not transform even at high strain levels. 24 Finally, De Knijf et al. [131] reported that austenite stability was negatively affected by the appearance of secondary martensite. As secondary martensite is formed, the distribution of strains in the microstructure can be significantly affected, since the amount of strain that can be accommodated in tempered martensite is drastically reduced. Consequently, the transformation stability of austenite decreases. 1.4 Relationship between microstructure and mechanical properties Q&P steels are of significant interest to generate desirable combinations of strength and formability. The key of these properties is the microstructure combining a martensitic matrix with significant fractions of carbon-enriched retained austenite [17]. However, the correlation between mechanical properties with the parameters of microstructure and its constituents is still a subject of study. Table 4 shows some examples of the tensile properties obtained in Q&P treated steels and the retained austenite measured after the treatment. Table 4. Tensile properties and volume % of retained austenite obtained by Q&P. Steel UTS (MPa) TEL (%) RA (%) Ref. 0.2C1.96Mn1.49Si0.25Mo 1280-1510 4-15 11-16 [80] 0.3C3Mn1.6Si 1500 17 1-15 [49] 0.2C3.5Mn1.5Si 1415-1630 15-23 8-19 [48] 0.2C1.57Mn1.55Si 1220-1340 12-16 6-14 [132] 0.5C2.01Mn1.03Si1.58Cr0.34Mo0.11Ti 1741-1931 2-12 12-28 [133] 0.28C4.08Mn1.42Si 1399-1548 13-16 12-19 [134] Several studies on Q&P steels reported the trend that the increase in RA is accompanied by increasing total elongation (TEL) but not so much by uniform elongation (UE), which reflects the contribution of TRIP effect. As mentioned in the previous section, it is the stability of the retained austenite, rather than its amount, which determines the final mechanical properties. Seo et al. [135] concluded that, in Q&P processed medium Mn steels, the key factors determining the mechanical properties were the kinetics of the mechanically-induced austenite to martensite transformation and the amount of secondary martensite. One characteristic of Q&P steels is the elevated YS/TS ratio due to a considerable amount of tempered martensite and, therefore, the strain hardening rate of these steels is relatively low. This results favorable for hole expansion, but the presence of secondary martensite significantly reduce YS and, consequently, the YS/TS ratio [8]. 25 1.5 Medium Mn steels (MMnS) Medium manganese steels (MMnS), with Mn contents between 4 wt.% and 10 wt.% [52], are considered another promising third generation AHSS due to their combinations of properties. These steels are based on a concept proposed by Miller [136] in 1972. In this case, cold rolled martensitic microstructures are intercritically annealed forming austenite by the so-called Austenite Reverse Transformation (ART) phenomenon [137]. Austenite stabilization can be achieved through diffusion of substitutional elements, such as Mn, (and also C) into austenite during high temperature processing [138,139], resulting in rather high contents of retained austenite (20–40%) [140–142]. A schematic of this process is shown in Figure 9. The stability of retained austenite and ultra-fine microstructure in MMnS, seem to play a key role in the impressive strength/ductility balance of these steels [143–145]. Figure 9. Schematic of the typical processing route of medium-Mn steels, where α’ is martensite, α is ferrite and γ is austenite. Intercritical annealing stage carried out in MMnS implies much longer times (typically hours) than the partitioning stage in Q&P process (typically ranging from a few seconds to few minutes). On the other hand, having both a tempered martensite matrix, the ultrafine-grained ferrite in medium Mn steels is softer. Therefore, the strength of Q&P steels is generally higher than that of medium Mn steels, but Q&P steels have a much lower elongation due to a lower RA fraction [146]. Although last years advanced, the understanding of medium manganese steels is still under development. The microstructure evolution, microstructure-properties relationship, the effect of alloying elements and contents, effects of processing and post-manufacturing, etc. are Ac3 RT Temperature (ºC) Time (s) α' Hot rolling Ac1 Ms Cold rolling α' α' α' γ γ γγ ART annealing γ γ γγ α α 26 some of the aspects which need to be approached. Moreover, although it is easier than for second-generation advanced high strength steels, such as TWIP steels, processing of MMnS in industrial environments is still challenging due to their relatively high alloy contents [147]. Even though the phase transformation behavior given in medium Mn steels during its processing is relatively understood, the precipitation and dissolution of the carbides and how does it affect austenite reversion required further study. Specifically, the understanding of how the heating rate and austenite reversion temperature affects the evolution of carbides would offer new awareness for the novel microstructure designs in MMnS [146]. 1.6 Q&P treatment for automotive industry The requirements of the automotive industry for weight reduction, improved fuel efficiency, and CO2 mitigation along with crash safety standards imposed by governments make the development of third generation AHSS a priority to the steel industry, as they provide an opportunity for the development of cost-effective and light-weight parts with improved safety and optimized environmental performance [148–150]. The ultra-high strength and excellent ductility, or formability, of Q&P steels make them well suited for weight reduction in car bodies, while increasing occupant safety. Q&P steels have a higher stretch-forming capability than conventional high strength steels (HHS) due to the significantly higher rate of work hardening. Furthermore, compared to most HSS with the similar tensile strength, Q&P steels have substantially higher formability; therefore, they are especially suitable for structural and safety parts of automobiles, such as cross members, longitudinal beams, B-pillar reinforcements, sills, and bumper reinforcements, which are difficult to cold form with conventional HSS with the same strength [151]. Q&P steels are still at an early stage of industrial implementation and might be developed for both high-strength components in automotive and other applications. 1.6.1 Commercialized Q&P steels To date, two levels of Q&P steels are in global production, with 980 MPa and 1180 MPa in tensile strength. The enhanced properties of Q&P steels offer benefits over similar strength steels of other microstructures. For example, compared against Dual Phase steel with similar yield and tensile strength, Baosteel produce a Q&P steel which shows higher uniform elongation, total elongation, work hardening index, and higher lowest point on the forming 27 limit curve (FLC0) [152]. ArcelorMittal reports similar strength and elongation properties of the QP980 steel, with a targeted 23% hole expansion ratio [153]. QP980 is undergoing increasing use in automotive production. In 2016 General Motors presented the first application in the Chevrolet Sail from SAIC-GM [154]. Later, the 2021 Ford Bronco used hot dip galvanized QP980 in five components of the front and rear floor assemblies [155]. 60% of the body structure of the 2021 Jeep Grand Cherokee L is made from AHSS, with some parts stamped from 3rd generation AHSS [156]. Table 5 contains typical mechanical property ranges for industrially produced QP980 and QP1180 [151]. Table 5. Typical mechanical property ranges for industrially produced QP980 and QP1180. Adapted from [151]. Material Yield strength (MPa) Tensile strength (MPa) Total elongation (%) QP980 650-800 980-1050 17-22 QP1180 950-1150 1180-1300 8-14 A recent conference highlighted several applications (Table 6) where thinner gauge QP980 replaced DP590 in General Motors vehicles [157]. The same presentation showed the example of QP980 replacing press hardening steels in B-pillar reinforcements and door anti-intrusion beams in a First Auto Works vehicle. Table 6. Replacing DP590 with QP980 allows for downgauging. Adapted from [157]. Part Material Part model Part prototype Kick Down Lower DP590 1.8 → QP980 1.6 Weight saving: 0.38 kg A Pillar Inner Lower DP590 1.2 → QP980 1.0 Weight saving: 0.20 kg Hinge Pillar Inner DP590 1.2 → QP980 1.0 Weight saving: 0.58 kg A Pillar Inner upper DP590 1.2 → QP980 1.0 Weight saving: 0.30 kg 1.6.2 Weldability of Q&P steels In addition to low weight and good mechanical behavior, high levels of weldability are important to have a competitive material to the automotive industry since high-quality welded joints are required to ensure the vehicle body's safety [158,159]. However, numerous studies 28 demonstrated that, due to the thermal effect of welding, apparent material softening can occur during different types of welding methods, such as laser welding (LW) [160–162], resistance spot welding (RSW) [163–165], gas metal arc welding (GMAW) [160,166] or friction stir welding (FSW) [167–170], in AHSSs like DP steel [160,165–168], TRIP steel [161,169,171], and martensitic steel [162,166,171]. Usually, property losses are related with the tempering of the material, which produces heterogeneities in the microstructure of the heat-affected zone where local stress or strain concentration is preferentially developed during deformation. In Q&P steels the martensite is tempered during partitioning, while the austenite is stabilized. Therefore, the internal structures of Q&P steels are more stable. Recent studies verified the extraordinary welding performance of QP980 steel during LW [172–174] and FSW [175,176], and equal strength joints to the steel were successfully fabricated with negligible heat-affected zone softening. Wang et al. [177] demonstrated the viability of achieving high strength joints in the QP1180 steel, while in TRIP [169] and DP [170] steels with similar strength levels are difficult to achieve such joint strength levels. The performance of the joint is determined by the welding method since, under various heat source energies, welding speeds, and cooling rates, the dimension and degree of softening of the heat-affected zone can vary considerably [160,178,179]. Although several works showed that Q&P steels have high welding performance [172,173,175,177,180], further research is required to determine the influence on the microstructure and mechanical properties, especially in ultrahigh strength steels (>1 GPa). 1.7 Q&P treatment for applications requiring wear resistance It is complex to estimate the wear resistance of a material, as it is defined by the conjunction of many properties of the material and varies with the wear condition, environment, and mechanism. Typically, wear resistance is directly related with hardness, although many other factors must be considered. Increased hardness is generally obtained by increasing the carbon content of the steel. However, the higher carbon content typically leads to a decrease in other properties of interest, such as toughness, bendability, and weldability [181]. Therefore, other methods are being investigated to increase the wear resistance of steel without drastic loss of other important properties. 29 Although quenched and tempered (Q&T) steels show good abrasive wear performance, some studies demonstrated that microstructures containing RA, despite having lower hardness, presented better wear resistance. The benefits of RA were observed in block on ring testing of austempered ductile iron [182], pin on-disk testing of both mottled cast iron [183] and a D-2 tool steel [184] and a non-standard impact-abrasion test of a 13.8 wt.% Cr, 2.7 wt.% Mn containing steel [185]. More recently, Q&P steels have been studied in a wear context and have shown to be of interest for wear applications. All the Q&P studies related to wear are very recent and most of them correspond to the mining industry. In general, it is concluded that the retained austenite is very beneficial for the tribological behavior (wear and friction). Under mechanical loading, a higher amount of retained austenite helps to obtain a thicker hardened layer due to transformation into martensite. In Q&P steels, apart from the direct relationship between hardness and wear resistance, the latter also depends on the microstructure and, therefore, on the Q&P process conditions (QT, PT, Pt) [186]. Generally, higher austenite content translates into better abrasive wear performance. A clear example of this phenomenon is shown in the work of Wasiak et al. [187], who observed that wear is reduced by practically 50% in the case of Q&P compared to Q&T in a 35CrSiMn5-5-4 steel, so that the improvement obtained is substantial. The employed Q&P cycle consisted of a full austenization followed by an oil quenching to 235 ºC and a partitioning at 260 ºC, obtaining 20% of RA. P. Wolfram et al. [188] calculated the volume loss of Q&P treated 9260 steel with normalized samples with respect to the performance of AR400F samples during dry sand and rubber wheel (DSRW) wear tests, and here also a beneficial effect of austenite was confirmed. In a subsequent study, the material was examined in a simulated wear track after controlled scratch tests for different microstructural conditions: after fast quenching and after quenching and tempering (Q&T) and quenching and partitioning (Q&P). The rapidly cooled material fractured out of the wear track in response to the indentation force. In Q&T, the wear track edges showed a chipping mechanism, while much less fragmentation was observed with the austenite-containing Q&P condition, which helps explaining the improved wear resistance of the Q&P microstructures. The same team also confirmed that the high Si Q&P processed steel exhibited a better combination of hardness and wear resistance than the same Q&T processed steel. Another recent study [189] studied the wear resistance of direct quenched and partitioned (DQ&P) steels subjected to impeller-tumbler impact-abrasive wear testing [189]. This study 30 found that wear performance depended solely on hardness, regardless of microstructure, and did not see improvements with increasing RA. However, it must be considered that the wear resistance measured at laboratory level depends on the tests carried out, and these are very sensitive to the test parameters. Generally, wear applications require thicker materials than those used in the automotive industry. Therefore, there is a need to understand the thermal gradients that can arise when applying Q&P to thicker samples. Stewart et al. [190] carried out some simulations to reproduce the thermal history that would occur in a 300M plate steel of 18 mm thick at different thicknesses, observing different quenching temperatures and non-isothermal partitioning depending on the thickness. Then, they evaluated the microstructures arising from each thermal cycle, discovering that the retained austenite varied between 5% and 27% between the surface and the center of the sample. As appreciable amounts of austenite were retained through the simulated plate thickness regardless of initial quench, they concluded that the insensitivity to quenching conditions might suggest behavior that would be considered beneficial in the framework of robust industrial implementation. 1.7.1 Toughness in Q&P steels for wear applications In Q&P steels for wear applications, toughness is usually a desired property. Reported literature show that generally this parameter is improved in Q&P steels with respect to those treated by a conventional Q&T treatment [191–193]. This improvement is likely due to the lower hardness values measured in Q&P steels. Figure 10 shows a comparison between the Charpy impact energies obtained for a Fe-0.19C-1.5Si-1.46Mn steel treated by two different Q&P and Q&T cycles [194]. As it can be seen, the improvement achieved by the Q&P treatment is substantial, specially at high temperatures. Figure 10. Comparison of Charpy impact energies at different temperatures of Q&T and Q&P treated specimens [194]. 31 Chapter 2 Motivation and objectives 38 39 Chapter 3 Materials and experimental procedures In this chapter the materials and techniques employed in this work are described. The reasons for selecting the chemical compositions, as well as the procedures employed to carry out the Q&P cycles are presented. Finally, the methods and experimental techniques employed to perform both microstructural and mechanical characterization are also described. 3.1 Materials and processing 3.1.1 Materials The materials employed in this work were divided in two groups: • Low C medium Mn steels with different Mn and Ni additions to be treated by high partitioning temperature Q&P treatments and study of their microstructure and strength-ductility behavior. These materials were produced in the laboratory in the form of cold rolled sheets and were studied as potential candidates for the automotive sector. • Medium and high C steels employed to investigate the benefits of Q&P heat treatment in relation with wear applications. Several commercial grades were Q&P treated and characterized in terms of microstructure and hardness, wear resistance and toughness properties. In this case, classical Q&P cycles were applied, and the application of high partitioning temperatures was planned as part of a future work. 3.1.1.1 Medium Mn steels Four steels, referred to as 2Mn, 4Mn, 6Mn and 6Mn2Ni, and which composition is shown in Table 7, were studied. The 2Mn steel, which has a typical composition of a Q&P steel, was established as the reference steel. The other steels were medium Mn steels and 6Mn2Ni 40 steel also contained Ni. Alloying elements and their amounts were selected in order to study the influence of Mn additions (the 4Mn and 6Mn steels with respect to the 2Mn steel), and Ni additions (the 6Mn2Ni steel with respect to the 6Mn steel). The addition of both Mn and Ni was made to promote the partitioning of these elements in order to increase the stability of the austenite. The addition of Ni was also motivated by its lower segregation in comparison with Mn. Silicon was added to supress carbides formation. Table 7. Composition (wt.%) of the medium Mn steels employed in this work. Steel ref. C Mn Ni Si P S N 2Mn 0.2 2.0 - 1.5 0.013 0.0015 0.0031 4Mn 0.19 3.8 - 1.4 0.012 0.0017 0.0035 6Mn 0.19 5.8 - 1.4 0.011 0.0013 0.003 6Mn2Ni 0.19 5.7 1.6 1.4 0.014 0.0019 0.0041 Ingots of four steel grades were produced in the laboratory (ArcelorMittal, France, in the frame of the HIGHQP project, grant No 709855) using a vacuum induction melting furnace. The ingots were reheated to ~1250 ºC and a roughing stage was performed to decrease the thickness from 60 mm to 30 mm. The obtained slabs were then cut into five small ingots which were hot rolled after being reheated to ~1250 ºC. The hot rolling consisted of five hot rolling passes (Tend of rolling approximately above 900 ºC) and coiling at 500 ºC. The hot rolled sheets of about 4 mm thickness were ground on both surfaces to decrease the thickness to 2.8 mm. A softening anneal was applied before cold rolling which consisted of holding at 600 °C for 1 h followed by water quench. Then, the sheets were cold rolled with approximately 50% reduction to obtain a final thickness of 1.5 mm. Hot rolled materials were characterized before and after softening annealing by microstructure observation with a light optical microscope (LOM) after polishing down to 1 μm and etching with sodium metabisulfite (ArcelorMittal, France, in the frame of the HIGHQP project, grant No 709855). Vickers microhardness tests with 9.8 N load were also performed with the hot rolled samples. The results of the hardness measurements of hot rolled materials at different steps, together with some examples of the observed microstructures, are presented in Figure 11. The hardness of the 2Mn steel was 224 HV, which was acceptable for further cold rolling and therefore was not heat treated. In contrast, the obtained hardness of other steels was quite high (420 HV–490 HV). This means that direct cold rolling of these hot rolled sheets was difficult, and therefore specific softening treatments were applied as explained in the previous paragraph. After the softening treatments, the obtained hardness was in range of 280 HV–320 HV, which allows for laboratory cold rolling procedures. 41 Figure 11. Light optical metallography (LOM) images and hardness of the hot rolled microstructures. The 2Mn (a), 4Mn (b), 6Mn (c) and 6Mn2Ni (d) steels after hot rolling, and the 4Mn (e), 6Mn (f) and 6Mn2Ni (g) steels after soft annealing. As can be seen in the micrographs of Figure 11, the 2Mn steel mainly presented a ferrite-pearlite structure with some minor fractions of bainite and martensite. The microstructure in the remaining steels consisted of a mixture of bainite–martensite. In the case of the 4Mn steel, the major phase seemed to be bainite and in the 6Mn and 6Mn2Ni steels, it was martensite. These microstructure observations were in agreement with the measured microhardness levels of the different hot rolled sheets. The presence of hard phases such as bainite and martensite was a consequence of the higher Mn and Ni contents added to the steels. Softening annealing performed on the 4Mn, 6Mn, and 6Mn2Ni steels resulted in the recovery of the structure (decrease of dislocation density), which was helpful in the reduction of steel hardness and, in some cases, led to the precipitation of carbides. 3.1.1.2 Medium and high C steels For this part of the work, three commercial medium and high C steels were selected (Table 8). The carbon content of the selected steels was high enough to guarantee a minimum hardness, allowing to achieve wear resistance values similar to those currently achieved in fields where wear resistance is required. In addition, it was ensured that the steels to be studied contained a significant amount of silicon in their chemical composition in order to avoid the formation of cementite in the partitioning stage of the Q&P treatment. On the other hand, it must be considered that Q&P treatment has been typically studied in low thickness products (such as cold rolled sheets). However, the products employed in wear applications may show higher thickness. As hardenability is a constraint if Q&P is applied in high thickness products, the steels were intended to contain a series of chemical elements that ensure (a) 2Mn, 224 HV1 (b) 4Mn, 427 HV1 (c) 6Mn, 484 HV1 (d) 6Mn2Ni, 481 HV1 Hot Rolled (e) 4Mn, 281 HV1 (f) 6Mn, 310 HV1 (g) 6Mn2Ni, 315 HV1 Soft-Annealed 42 hardenability, as Mn, Cr, etc. At the same time, these elements were intended to retard the formation of bainite in the partitioning stage and harden the martensitic matrix. The selection of the 3 commercial steel grades was done supported by JMatPro (v10.2) thermodynamic software. Using this software critical cooling rates to avoid diffusion-based phase transformations (CCR) and hardness in the as-quenched state were calculated for several commercial steel grades. Based on these results, 1.2990, 1.2344, and 300M steels were selected (Table 8) with the criteria that critical cooling rates were minimized, and the hardness values were maximized, ensuring a suitable content of Si. The continuous cooling transformation (CCT) diagrams calculated by JMatPro employed for the predictions of CCR and hardness for the selected steels are shown in Figure 12. The first two steel (1.2990 and 1.2344) are tool steels which are typically heat treated by quenching and tempering. These steels contain high amounts of carbides forming elements (Cr, Mo, V), which give rise to the secondary hardening phenomenon in the tempering. The main difference among both steels is the C content. The third steel, 300M, was selected mainly due to the high Si content. Figure 12. Theoretical CCT curves of 1.2990 (a), 1.2344 (b), and 300M (c) steels obtained with JMatPro. (a) 1.2990 (b) 1.2344 (c) 300 M 43 Table 8. Composition (wt.%) of the medium and high C steels employed in this work. Steel ref. C Si Mn Cr Mo Ni V 1.2990 0.92 0.96 0.36 7.9 1.0 - 1.5 1.2344 0.39 0.97 0.37 5.1 1.2 - 0.92 300M 0.41 1.6 0.83 0.78 0.38 1.8 0.06 3.1.2 Design of the Q&P cycles based on dilatometry A LINSEIS L78 RITA dilatometer (Tecnalia, Spain) was used to obtain the phase transformation temperatures, critical cooling rates , Q&P cycle parameters, and then physically simulate the whole Q&P cycles to understand microstructure evolution by the analysis of the expansion/contraction behavior. A picture of L78 RITA dilatometer is presented in Figure 13. The dilatometer follows the length variations of the sample occurring during the imposed heat treatment. The samples were heated and maintained at temperature by induction heating. Temperature control was done by Type K thermocouples. The samples were maintained by vertical hollow quartz rods, being the upper rod mobile. Hence, when length variations occurred, one rod moved, and the linear displacement was captured by an LVDT (Linear Variable Differential Transducer) sensor. Figure 13. Global view on L78 RITA dilatometer. In the present work, cylindrical samples, of 3 mm diameter and 10 mm length, were used. After machining, the samples were cleaned by immersion in ethanol within an ultrasound bath (to remove dirt and oils from machining) and just before the test, they were cleaned again with ethanol. Type K thermocouple was welded in the middle of the clean surface of the sample. To avoid oxidation during treatment vacuum was done in the experimental chamber, then, a small amount of helium (He) was injected. 44 3.1.2.1 Obtention of phase transformation temperatures and critical cooling rates First, Ac1 and Ac3 phase transformation temperatures were determined by dilatometry. For that, samples were heated up to 1000 °C (in the case of Medium Mn steels) and 1100 ºC (in the case of medium and high C steels) at 5 °C/s (medium Mn steels) and 10 ºC/s (medium and high C steels) and then, after 120 s of maintenance, cooled down to room temperature. Two tests were carried out for each medium Mn steel, whereas only one test was carried out for each medium and high C steel. Once the dilatometry curve was obtained, Ac1 and Ac3 temperatures were determined by the application of the tangent method. As in medium Mn steels a deviation was observed just before austenite transformation curve, Ac1 was considered the temperature with maximum relative change in length just before transformation started. As reference of each Ac1 obtention methods, in Figure 14 the curves of the 2Mn and 1.2344 steels showing Ac1 and Ac3 temperatures are presented. The dilatometry curves for all steels are shown in Appendix A. The results for all steels are shown in Table 9. It can be seen that, in medium Mn steels, Mn additions led to lower austenization temperatures and Ni addition resulted in a further decrease. Since the medium and high C steels have different amounts of different alloying elements it is not possible to stablish a direct correlation between the austenization temperatures. Figure 14. Dilatometry curves representing the determination of Ac1 and Ac3 temperatures of the 2Mn (a) and 1.2344 (b) steels. In (a) Ac1 was determined as the temperature with maximum change in length just before transformation started and Ac3 was determined by the tangent method; in (b) both temperatures were obtained by the tangent method. Once Ac3 temperatures were determined, critical cooling rates to avoid any undesirable transformation during quenching were experimentally determined by dilatometry. The 0.006 0.008 0.010 0.012 0.014 0.016 0.018 0.020 600 700 800 900 1000 Relative change in length (ΔL/L0) Temperature (ºC) Ac1 Ac3 (a) 2Mn steel 0.012 0.014 0.016 0.018 0.02 0.022 0.024 800 900 1000 1100 Relative change in length (ΔL/L0) Temperature (ºC) Ac1 Ac3 (b) 1.2344 steel 45 procedure to obtain these values was as follows: first, samples were heated up to a soaking temperature 50 ºC above Ac3 and after a soaking time of 120 s were cooled down to room temperature employing different cooling rates. The first cooling rate selected for each steel was based on theoretical results obtained by a thermodynamic software. If that cooling rate resulted in any further transformation apart from martensitic a higher one was tested. The dilatometry curves for all steels are shown in Appendix A and all CCR are presented in Table 9. It can be seen that increasing Mn in medium Mn steels led to a considerable decrease of this parameter and the alloying contents of the medium and high C steels were also sufficient to achieve low CCRs. Table 9. Phase transformation temperatures (Ac1, Ac3, Ms and Mf), critical cooling rates (CCR) and the austenite measured in the quenching state of all the steels employed in this work. Steel ref. Ac1 (ºC) Ac3 (ºC) Ms (ºC) Mf (ºC) CCR (ºC/s) As-quenched RA (%) Medium Mn steels 2Mn 760 920 363 215 60 1 4Mn 730 873 313 101 0.5 – 2 3 6Mn 695 841 252 <RT <0.1 7 6Mn2Ni 690 793 219 <RT <0.1 7 Medium and high C steels 1.2990 881 987 342 <RT 0.5 – 1 <5 1.2344 866 993 317 ~RT <1 - 300M 758 875 270 ~RT <2 - 3.1.2.2 Obtention of martensitic transformation curve There are different ways to obtain the martensitic transformation curve. The first method, purely theoretical, consists in calculating the Ms temperature by means of the Andrews equation (1) [195] and, then, calculate the austenite fraction corresponding to each temperature by means of the Koistinen-Marburger equation (2) [196]: 𝑀𝑠=539 −423𝐶 − 30.4𝑀𝑛 −17.7𝑁𝑖 −11.0𝑆𝑖 −12.1𝐶𝑟 − 7.5𝑀𝑜 (1) where the alloying elements are expressed in weight percent. 𝑓 𝛾= exp(−𝛼𝑚×(𝑀𝑠− 𝑇)) (2) where fγ is austenite volume fraction; T is current temperature, ºC; and αm is a constant coefficient with a value equal to 0.011 K-1. Other method, the one employed in this work, is to obtain the transformation curve experimentally by dilatometry. Thus, the samples were heated up to Ac3 + 50 ºC at 5 ºC/s 46 (medium Mn steels) and 10 ºC/s (medium and high C steels) and, after 120 s, were cooled down to room temperature. The applied cooling rate was 100 ºC/s in the 2Mn steel, 20 ºC/s in the medium Mn steels, and 45 ºC/s in the medium and high C steels. Then, the samples were reheated up to 500 ºC at 5 ºC/s. Applying the lever rule between the expansion of the untransformed austenite curve and that from the reheating, martensite transformation curves were calculated. The final change in length was associated with the amount of martensite in the as-quenched state, which was obtained from the following dilatometry: a sample was quenched to room temperature from Ac3 + 50 ºC after austenitizing it for 120 s, considering the same cooling rate as before. Then, austenite content of the sample was measured by magnetization in medium Mn steels (TU Delft, the Netherlands, in the frame of the HIGHQP project, grant No 709855), and X-ray diffraction in medium and high C steels (SGIker, UPV/EHU, Spain). The measured values are collected in the last column of Table 9. From the martensitic transformation curve, Ms and Mf temperatures were obtained as the temperatures in which 5% and 95% of the martensite transformation occurred, respectively. As an example, in Figure 15a the dilatometry curve of the 2Mn steel is shown and the Ms and Mf temperatures for all the steels are presented in Table 9. The dilatometry curves for all steels used to obtain the martensitic curve are shown in Appendix A. The QTs selected for each Q&P cycle were obtained from the corresponding martensitic transformation curve and are detailed in the next section. In Figure 15b the procedure to obtain a QT corresponding to 25% of untransformed austenite (QT25) is represented by red arrows. Figure 15. Dilatometry curve of the 2Mn steel for the determination of Ms, Mf and martensite transformation curve (a) and the corresponding martensitic transformation curve showing the determination of the temperature corresponding to a transformation of 75% martensite (b). 0.000 0.005 0.010 0.015 0.020 0.025 0 100 200 300 400 500 Relative change in length (ΔL/L0) Temperature (ºC) B A C Lever rule: α’ = AB/AC (a) 0 10 20 30 40 50 60 70 80 90 100 0 100 200 300 400 500 Martensite (vol.%) Temperature (ºC) (b) 47 3.1.2.3 Design of the heat treatments (QT, PT and Pt) The selection of the thermal cycle parameters followed a different strategy depending on the required final properties. Therefore, a series of treatments was designed to medium Mn steels and another to medium and high C steels. Design of cycle parameters in medium Mn steels: In all cases, full austenization was performed by heating at Ac3 + 50 ºC. Then, after 120 s of soaking, samples were cooled down to the selected QTs at a cooling rate above CCR. After 5 s at the QT, samples were heated up to the established PT. Finally, once the partitioning time was completed, samples were cooled down to room temperature. In a first stage, three Q&P cycles were designed: A reference cycle with a typical partitioning temperature of 400 ºC and a quenching temperature corresponding to 25% of untransformed austenite (QT25), and two high partitioning temperature cycles with different quenching temperatures, one of which coincided with that used in the reference cycle and other corresponding to 10% of untransformed austenite (QT10). In this way, both the influence of PT and QT on the microstructure could be investigated. High partitioning temperature was selected as the temperature in which austenite reverse transformation starts (TART). The selection of TART as partitioning temperature was motivated from previous work, in which it was observed that temperatures beyond the intercritical range lead to undesirable decomposition of austenite into pearlite [39]. Higher partitioning temperatures might also lead to excessive austenite which would be difficult to stabilize. Nevertheless, the investigation of the effect of partitioning temperatures around TART would be interesting as part of a future work. The consideration of two QT conditions (QT10 and QT25) followed by a partitioning stage at TART, aimed at investigating the influence of different contents of pre-existing austenite on the formation of reverted austenite. The pre-existing austenite contents were defined based on a previous work [112] where it was reported that ductility strongly decreased when pre-existing austenite was higher than 30% due to the presence of secondary martensite. The formation of secondary martensite was reported to be the consequence of insufficient austenite stabilization during the partitioning stage. Therefore, in this thesis, pre-existing austenite contents were kept below 30%. The temperature for the start of austenite reverse transformation (TART) upon heating from the QT, was determined by dilatometry for each steel. For this, samples were heated to 54 Bragg’s law relates the angle θ, the wavelength of the beam (λ) and the spacing between the planes of atoms in the material (d), and n is an integer representing the order of the diffraction peak: 𝑛𝜆 = 2𝑑 sin 𝜃 (3) Additionally, the volume of material that has planes with the same space and orientation to the diffraction and the intensity of a diffracted beam in a particular direction are proportional. Thus, from the relative intensities of the diffracted beams, the relative volume of a polycrystalline material phases can be estimated [197]. The samples studied in this work are assumed to be consisted of ferrite (which represent martensite) and austenite. According to ASTM E975-13, the volume fraction of austenite (Vγ) can be calculated as follows: 𝑉 𝛾=𝐼𝛾𝑅𝛾 ⁄ 𝐼𝛼𝑅𝛼 ⁄+ 𝐼𝛾𝑅𝛾 ⁄ (4) where Iγ and Iα are the measured intensities of a particular (hkl) peak above the background of austenite and ferrite, respectively; and Rγ and Rα are the calculated intensities of a particular (hkl) austenite and ferrite peak (standard value), respectively. In this work, the employed (hkl) peak for both austenite and ferrite was (200), being the corresponding R values 34.78 and 20.73, respectively. Thus, in this work, for the measurement of RA in the Q&P samples, X-ray powder diffraction patterns were collected by using a Bruker D8 Discover diffractometer (SGIker, UPV/EHU, Spain) equipped with a Cr Twist tube, V filter (λ = 2.291 Å), PolyCapTM (1μ single crystal cylinders) system for parallel been generation (divergence of 0.25°), and a 1-D LynxEye detector (active length in 2θ 2.7º). The samples were mounted on a Eulerian Cradle with automatic controlled X-Y-Z stage. Data were collected from 50 to 120° 2θ (step size = 0.04 and time per step = 1 s). Peak area intensity was evaluated using the peak-fit option of the WinPLOTR software. 3.2.2 Scanning Electron Microscope (SEM and FE-SEM) Scanning electron microscopy (SEM) is a microstructural characterization technique in which a high-energy electron beam is focused onto a fine probe that scatters inelastically when it strikes the surface of a solid sample. The electrons are inelastically scattered, thus generating various signals from the sample that are collected and amplified. Scanning the 55 probe beam across the sample surface forms a digitized image displaying one or more of the collected signals on a monitor that has the same time base as the probe scan. Various signals are generated as a result of the impact of the incident electrons, being the secondary electrons the most commonly used, although characteristic X-rays, high energy backscattered electrons, visible cathodoluminescence and the net specimen current have all been used to acquire microstructural information from samples examined in the scanning electron microscope [198]. In this work a field emission source for the electron beam was used in some of the medium Mn steel samples to improve the performance of SEM (FE-SEM). Samples selected for FE-SEM (JEOL® JSM7000F) (Colorado School of Mines, EEUU) were prepared by conventional polishing to 1 µm, followed by a final polishing with colloidal silica. Then, there were etched with 2% Nital for 30 s. Characterization conditions were a beam voltage of 15 KV, a medium probe current, and a working distance of 10 mm. SEM (QUANTA 200 FEI) (Tecnalia, Spain) characterization was performed to the Q&P samples of the 1.2990 and 300M steels. In this case, samples were prepared by conventional polishing to 1 µm and etched with 2% Nital for 1 min. A beam voltage of 25 KV, and a working distance of 10 mm were employed. In addition, energy-dispersive X-ray spectroscopy (EDS) analysis was carried out in order to identify the composition of some of the phases presented in the 1.2990 steel. 3.2.3 Electron Backscatter Diffraction (EBSD) Electron backscatter diffraction (EBSD) is a scanning electron microscope (SEM) based technique. It allows the measurement of microtexture (texture on the scale of the microstructure) [199], microstructure quantification [200], grain and phase boundary characterization [201,202], phase identification [203] and strain determination [204] in crystalline multiphase materials of any crystal structure. By focusing an electron beam on a crystalline sample, diffracted patterns are obtained, which are subsequently evaluated and indexed. Generally, this is done automatically, and the data is output graphically and statistically. The most versatile and illuminating results are the OIM maps, which are a quantitative representation of an area of the analyzed microstructure in terms of its crystallographic constituents [205]. One such map is image quality (IQ), which is a metric that describes the quality of a diffracted pattern. An IQ map is constructed by mapping the measured IQ value for each diffraction pattern obtained during an OIM scan to 56 a gray or color scale. Both the "perfection" of the crystal lattice and the atoms present within the diffraction volume affect the IQ [206,207]. In this work, scans were performed on a field emission scanning electron microscope (FE-SEM, JEOL® JSM7000F) (Colorado School of Mines, EEUU) with a beam voltage of 20 KV and a medium probe current. The acquisition of EBSD scans was done using a step size of 0.06 μm. Selected samples were prepared by conventional polishing to 1 µm, followed by a final polishing with colloidal silica. The obtained results consisted on IQ maps and identification of average grain size of the retained austenite. 3.2.4 Transmission Electron Microscopy (TEM) The transmission electron microscope (TEM) can be comprehended as a tool engineered specifically for the analysis and visualization of samples at micrometric and nanometric scale. This kind of electron microscope has capability of reveal highly complex levels of detail which are inaccessible by a conventional light microscope. In the TEM high energy electrons are elastically scattered as they penetrate a thin specimen. The transmitted electrons are then focused by electromagnetic lenses to form a well-resolved image that can be viewed on a fluorescent screen or a charge-coupled device [208]. In this work microstructure characterization of some of the Q&P treated medium Mn steel samples was done by means of a TEM (Talos F200i field emission gun instrument equipped with a Brüker X-Flash100 XEDS spectrometer) (SGIker, UPV/EHU, Spain). Elemental maps were obtained by XEDS in the STEM mode under a high annular dark field (HAADF) detector for Z contrast imaging in STEM conditions (camera length of 160 mm) using a pixel size of 2 nm, a dwell time of 900 s and an image size of 512 x 512 pixels. Thin-foil specimens were prepared for the observation in the transmission electron microscope. A thin-foil is a 3 mm diameter disk with a central hole around which the TEM observation was performed. The thin-foil specimens must be less than 100 nm in thickness to minimize inelastic scattering of the transmitted beam as it passes through the specimen. Good specimen preparation is critical. The procedure employed in this work for the preparation of the foils was as follow: • First, sheets of Q&P treated material were cut into pieces of about 10 mm x 10 mm. • Once the sample was cut, it was glued to a resin block or a metallic funnel (Figure 23) with Loctite. 250 μm thick copper tape was glued to each side of the sample and the sample was sanded down to that thickness. 57 • Having thinned the sample to 250 μm, being the last pass with 1200 grit sandpaper, the sample was peel off and stick it back to polish the other side. This time, 100 μm tapes were placed and sample was thinned down to that thickness, finishing the polished with 1 μm cloth. Figure 23. Funnel where samples were glued to be polished. • In this way samples had approximately 100 μm thickness, and 3 mm diameter disks could be obtained. This is done lowering the red lever of the "hand drill" shown in Figure 24. As many disks as possible were obtained from each sample. Figure 24. Hand drill employed to take out 3 mm diameter discs from the samples thinned down to 100 μm. • Next, the side of the disk that was polished only down to 1200 grit sandpaper, was finished by polishing down to 1 μm cloth. 58 • Thus, disks were completely mirror polished on both sides. To make the central hole, around which the TEM observation was to be carried out, an electropolishing bath was used. The employed electrolyte was a mixture of 5% perchloric acid and 95% ethanol (absolute). The equipment has a sample holder to place the discs (Figure 25). Figure 25. Sample holder for the electrolytic bath and electropolishing equipment. • Once the first disc was placed, a scan was carried out at different temperatures to find the appropriate voltage and temperature to make the hole by electropolishing. Through these scans, the intensity-voltage curve was represented at the temperature of the bath. The curve must consist of 3 parts (Figure 26), a first ascending part in which the sample would be attacked, a second in which a plateau is found, where the sample would be polished, and a third, also ascending, in which the sample would be spoiled by chopping. Testing with temperatures of -10 °C and -15 °C was impossible to distinguish the plateau. At room temperature it was clearly seen that the plateau was at 29 V. Figure 26. Intensity-Voltage curve. mA V Suitable V Etching area Polishing area Chopping area 59 • Once the voltage and temperature to be used were established, it was advisable to carry out the electropolishing with a new disk, since when performing the scanner in a wide voltage range, the sample likely was already etched, polished and even chopped. • The equipment was therefore programmed to operate at 29 V. A sensitivity of 10% was set (this is measured by detecting the light that passes through the sample so that the test ends as soon as the hole has been made, around of 1 minute with these samples). • Once electropolished, the disk was quickly immersed in ethanol to remove any traces of acid and was passed from one ethanol bath to another by immersing the sample up to three different baths. • Finally, the disk was left to dry, and it was observed if the holes were performed by an optical microscope. 3.3 Mechanical behavior 3.3.1 Tensile test Tensile Testing is a form of tension testing whereby controlled tension is applied to a sample until it fully fails. This is one of the most common mechanical testing techniques. It is used to find out how strong a material is and how much it can be strained before it breaks. In this work, this test method was used to determine yield strength, tensile strength, and ductility of all the medium Mn steels Q&P treated in the furnaces and salt baths. Two tests were performed for each steel/Q&P condition. Tensile specimens were water cut from the treated sheets. The dimensions of the specimens corresponded to a standard geometry of 50 mm gauge length and are specified in Figure 27. Figure 27. Dimensions of the tensile specimens. All measurements are expressed in mm. 60 Tests were performed in a universal INSTRON tensile testing machine (Tecnalia, Spain), with a strain rate of 0.001 s-1 and a contact extensometer following the UNE-EN ISO 6892-1:2019. 3.3.2 Interrupted tensile test Interrupted tensile tests were performed to the selected steels and Q&P conditions in order to assess the evolution of the RA under mechanical loading. For that, once the total elongation was known for a given steel/Q&P condition, different strain levels were selected to stop the tensile tests. Then, the RA fraction was measured by X-ray diffraction in the way aforementioned in the section 3.2.1. In the tensile specimens with low strains the center section of the specimen was selected for the measurement of RA, whereas in specimens with higher strains the thinner area of the tensile specimen was selected. RA was measured in broken specimens as well. In Figure 28 the different sections employed for RA measurements are shown. Figure 28. Samples cut for XRD measurements from tensile specimens after the application of different strains. 3.3.3 Hardness Materials with wear resistance requirements must have a relatively high hardness. Thus, in medium and high C steels studied in this work hardness was an important parameter. Hardness measurements were carried out in a Vickers Hardness Tester FV-700 model (FUTURE-TECH) (Tecnalia, Spain) using a 10 kg load, applied for 10 s according to the ISO 6507-1. To minimize errors, 5 indentations per sample were performed. The Vickers test consists of an indenter in the form of a straight pyramid with a square base and with a specific angle between opposite faces at the vertex (α), which is introduced with Low strain High strain Fractured specimen 61 a certain force (F) on the surface of the specimen of the material to be tested. Subsequently, the diagonals of the imprint (d1 and d2) remaining on the surface are measured and, finally hardness value is obtained by means of (5), where “d” is the average value between d1 and d2. The Vickers hardness test principle is depicted in Figure 29. 𝐻𝑉 =0.189 × 𝐹 𝑑2 (5) Figure 29. Vickers hardness test principle. 3.3.4 Tribology: pin-on-disk (PoD) To characterize the wear resistance of medium and high C steels, pin-on-disk (PoD) tests were carried out in a MICROTEST tribometer (Tecnalia, Spain) to the Q&P and Q&T treated samples of 1.2990 and 300M steels. The PoD consists in placing two materials in contact, keeping one of them in motion and both subjected to constant pressure for a determined time or distance. As shown in Figure 30, this test reproduces the unidirectional sliding process between two different materials under certain conditions. Figure 30. Pin-on-disk system. In these tests, the disks (materials investigated in this work) are the ones in motion. The pin can be a cylinder with a flat surface, spherical or a ball. In this case, the pin was a 6 mm diameter ball. The dimension of the disk for the application of PoD were 4 mm thickness F α d1 d2 62 disks, with a diameter of 100 mm and 35 mm in the 1.2990 and 300M steels, respectively, and an inner hole of 5.5 mm diameter. The test parameters are summarized in Table 14. Two tests were performed for each steel/Q&P condition. Table 14. Parameters used for the pin-on-disk tests. Steel ref. 1.2990 300M Test radius (mm) 15 and 45 15 Lineal velocity (cm/s) 78.3 and 208.3 78.3 Load (N) 20 20 Temperature (ºC) RT RT Relative humidity (%) 50 50 Pin material Al2O3 Al2O3 Pin dimensions (mm) Ø6 Ø6 Distance (m) 15000 5638 After carrying out the test, the worn volume was evaluated on the disk, measuring a series of wear profiles. The evaluation consists of making 4 profiles on the track left in the test, thus obtaining 4 cross-sectional areas (Figure 31). This operation was carried out using a Dektak 150 Contact Profilometer. Then, the wear volume was calculated by the following geometrical relation: 𝑉 = 2𝜋𝑟 ∙ 𝐴 (6) where r is the test radius in mm, and A is the average wear area obtained by profilometer, expressed in mm2. Figure 31. Cross-sectional areas of the wear track employed for the measurement of wear rate. Finally, the specific wear rates (K) in mm3/N∙m of the investigated materials were obtained. This rate is defined as the volume worn divided by the normal load and the test distance using the following equation: K = 𝑉 𝐿 ∙ 𝑑 (7) Being V the volume of wear (mm3), L the normal load (N) and d the total distance of the test (m). 63 3.3.5 Toughness Toughness is the ability of a material to absorb energy and plastically deform without fracturing. The toughness of a material can be measured using a small specimen of that material. A typical testing machine is a Charpy V-notch impact test. This is a dynamic test in which a notched specimen is struck and broken by a single blow in a specially designed testing machine. In this work, a preliminary toughness study was realized with the 1.2990 steel to compare the energy absorption capacity of Q&P and Q&T treated steels. Due to material limitation, standard subsize specimens were used. Three test per heat treatment were performed. Figure 32 shows the dimensions of the standard subsize specimens according to ASTM A370. Figure 32. Standard subsize specimen dimensions (ASTM A370). The equipment used to carry out the impact tests was an AMSLER RKP 300 model pendulum from ROELL + KORTHAUS (University of Cantabria, Spain), with a capacity of 300 J and a hammer weight of 20.4 kg (200 N), with a travel angle of up to 150 ° and maximum impact speed of 5.42 m/s. 70 The dilatometry curves during final cooling to room temperature were analyzed to study the possible formation of secondary martensite (Figure 33). The curve was linear in all the QT10 conditions except for the 2Mn steel, which showed a secondary martensite transformation temperature (Ms2) of 167 ºC. This means that the formation of secondary martensite was not significant in the steels with at least 4 wt.% of Mn and, therefore, most of the austenite available at the end of partitioning was retained in the final microstructure. In this sense, the QT10 condition seemed to be beneficial for the stabilization of austenite. In contrast, for the QT25 condition, the dilation curves deviated from linear contraction, which denoted secondary martensite transformation during final cooling. The secondary Ms2 were determined to be 157 ºC, 133 ºC, 103 ºC and 57 ºC in the 2Mn, 4Mn, 6Mn and 6Mn2Ni steels, respectively. The lower Ms2 temperature observed with the increase of the alloying content in the steels indicated that austenite stabilization was enhanced, which was likely due to the known stabilization capacity of Mn [66] and Ni [37,52,53,210]. 4.1.2 Microstructure after the Q&P heat treatments The retained austenite contents were measured by XRD for all Q&P conditions and steels. Then, the most interesting conditions were selected for being analysed by means of FE-SEM, EBSD and TEM. The retained austenite contents measured by XRD are shown in Figure 36 as a function of partitioning time for all Q&P cycles and steels. Regarding ref cycle, the alloying content greatly influenced on the stabilization of austenite. In the 2Mn steel only 7% of the 25% austenite available at the QT was stabilized. As mentioned in the previous section, greater expansion was observed during partitioning in this steel which was related to bainite formation, thus, it is likely that a considerable amount of austenite transformed into bainite during partitioning resulting in a lower content of final austenite. With higher alloying contents the austenite was effectively retained at room temperature, achieving to stabilize all of it in the 6Mn2Ni steel. After the high partitioning temperature cycles with a partitioning time of 1000 s, almost no retained austenite was detected in the 2Mn steel regardless of the QT condition, showing that 2 wt.% Mn was not enough to stabilize austenite. In the rest of the steels, all the cycles resulted in a final retained austenite content greater than the austenite content existing at the QT with the exception of the QT25-Pt1000 cycle in the 4Mn steel, likely due to the secondary martensite transformation observed in the dilatometry curve (Figure 33b). In the case of the 71 QT25-Pt1000 cycle applied to the 6Mn steel, the increase in austenite content was small, but in the 6Mn2Ni steel was considerable. This is in good agreement with the Ms2 measured in the dilatometry curves (Figure 33), where this temperature decreased with the alloying content of the steel. In all QT10 conditions, a very substantial increase was observed. The maximum content of RA was 47%, which was obtained after the application of the QT10-Pt1000 cycle in the 6Mn2Ni steel. The Ni addition clearly resulted in a significant increase in RA content, regardless of the thermal treatment conditions. Furthermore, comparing the QT25-Pt1000 and the QT10-Pt1000 conditions, it can be seen that a lower austenite content at the QT clearly resulted in an increased content of austenite in the final microstructure of the 6Mn and 6Mn2Ni steels. As to the influence of partitioning time in the QT10 condition, increasing partitioning time from 300 to 1000 s also increased RA content. However, a further increase to 3600 s resulted in a slight decrease in RA content. The influence of partitioning time on RA content was very similar for the 6Mn and 6Mn2Ni steels. Figure 36. Retained austenite contents as a function of partitioning time measured after the application of all Q&P cycles. 0 5 10 15 20 25 30 35 40 45 50 55 0 500 1000 1500 2000 2500 3000 3500 4000 RA (%) Partitioning time (s) 6Mn –QT10 γ = 25 % γ = 10 % 6Mn2Ni –QT10 6Mn2Ni –QT25 6Mn –QT25 4Mn –QT25 4Mn –QT10 2Mn –QT25 2Mn –QT10 2Mn –ref 4Mn –ref 6Mn2Ni –ref 6Mn –ref 72 The 6Mn and 6Mn2Ni steels after QT10-Pt1000 and QT25-Pt1000 conditions were selected for being observed in the FE-SEM due to the high retained austenite contents. Additionally, the microstructure after QT10-Pt3600 cycle was observed in the 6Mn2Ni steel to analyse the influence of partitioning time. The microstructure expected in these samples can be described as follows: After heating to obtain a fully austenitic microstructure, the Q&P steels were quenched to the predetermined QT in order to form a partially austenitic and partially martensitic microstructure. Then, heating to a partitioning temperature, which corresponded to the austenite reversion transformation start temperature, the pre-existing austenite was expected to grow and be enriched by Mn and C provided by the martensite [13,49]. In the 6Mn2Ni steel, Ni enrichment could also occur [211]. Thus, after final quenching, the austenite which contained enough C and Mn would become stable and be retained at room temperature [13,49], whereas the less enriched austenite would transform into secondary martensite (α’sec). In this way, the expected final microstructure should consist of C and Mn depleted primary martensite (M1), RA laths and blocky RA, and martensite/austenite (MA) islands consisting of secondary martensite with fine RA. Carbide precipitation was also expected as a result of the tempering of martensite due to the high partitioning temperature employed. In the FE-SEM micrographs obtained in the present work (Figure 37), bright thin films were observed in all micrographs, which were likely retained austenite. The dark phase was interpreted to be primary martensite (M1) containing a considerable amount of carbides. The carbides exhibited both acicular/plate and globular morphologies. Mn retards cementite dissolution, so observed carbides could include some cementite that did not dissolve in prior steps, along with cementite formed as a consequence of tempering of the martensite. The regions with dark-grey center and white edges are generally recognized as MA islands. In the QT25-Pt1000 micrographs large areas of MA were evident. However, as shown later, the amount of secondary martensite determined from the dilatometry curves and XRD results was not significant. Therefore, the interpretation of these areas in the FE-SEM images was not entirely clear, but was presumably indicative of transformation during cooling near QT. In both steels, the microstructure observed after the QT25-Pt1000 cycle presented coarse and blocky constituents, whereas after the QT10-Pt1000 cycles the microstructure was thinner and exhibited a lath-type appearance. The microstructure features in the 6Mn2Ni steel after the QT10-Pt3600 cycle (Figure 37e) were thinner than after the QT25-Pt1000 cycle, but coarser than after the QT10-Pt1000 cycle. 73 Figure 37. FE-SEM micrographs corresponding to the QT25-Pt1000 cycle for the 6Mn (a), and 6Mn2Ni (b) steels; the QT10-Pt1000 cycle for the 6Mn (c), and 6Mn2Ni (d) steels; and the QT10-Pt3600 cycle for the 6Mn2Ni steel (e). EBSD phase maps were obtained for all the Pt1000 conditions of the 6Mn and 6Mn2Ni steels. EBSD phase maps are shown in Figure 38, where the red phase was identified as RA, green was martensite, and black indicates unidentified regions. From EBSD measurements, retained austenite content and size were obtained. In Table 15 a comparison between the RA content measured by XRD and EBSD is shown. The austenite content measured by EBSD was lower than that measured by XRD, which might be due to three reasons. First, the dark regions in the EBSD maps can be MA islands containing retained austenite. However, these regions are mostly recognized as unidentified regions. Second, detection of Acicular carbides Globular carbides M1 MA (a) 6Mn –QT25-Pt1000 (b) 6Mn2Ni –QT25-Pt1000 RA laths (c) 6Mn –QT10-Pt1000 (d) 6Mn2Ni –QT10-Pt1000 (f) 6Mn2Ni –QT10-Pt3600 74 nanometer sized film like retained austenite stabilized between the martensitic laths is difficult to identify with EBSD due to low resolution. Third, the area and depth of the measurements were different, playing an important role the step used in EBSD characterization. Most of the dark zones in the QT25 cycles (Figure 38a-b) likely corresponded to MA islands, whereas the thin dark zones between martensite areas in the QT10 condition (Figure 38c-d) likely included nanometer sized film-like retained austenite. In Table 15, average RA sizes estimated from EBSD measurements are also shown. In the 6Mn steel, the QT10-Pt1000 condition resulted in a coarser average austenite size in comparison with the QT25-Pt1000 condition. However, in the QT10-Pt1000 condition, the minimum austenite size was 110 nm and fine scale film-like retained austenite areas were mostly not detected in the measurement. Figure 38. EBSD scans corresponding to the QT25-Pt1000 cycle for the 6Mn (a), and 6Mn2Ni (b) steels; and the QT10-Pt1000 cycle for the 6Mn (c), and 6Mn2Ni (d) steels. Table 15. Comparison between retained austenite content (RA %) measured by XRD and EBSD. Average austenite grain size (RA nm) measured by EBSD. Q&P cycle 6Mn 6Mn2Ni XRD (RA %) EBSD (RA %) EBSD (RA nm) XRD (RA %) EBSD (RA %) EBSD (RA nm) QT25-Pt1000 26 17.4 168 38 25.2 220 QT10-Pt1000 43 20.4 237 48 17.6 161 (a) 6Mn –QT25-Pt1000 (b) 6Mn2Ni –QT25-Pt1000 (c) 6Mn –QT10-Pt1000 (d) 6Mn2Ni –QT10-Pt1000 75 In Figure 39, TEM characterization results are presented for the 6Mn2Ni steel. Figure 39a and Figure 39b show TEM micrographs after the application of the QT25-Pt1000 and QT10-Pt1000 cycles, respectively. In both cases, the microstructure consisted of lath-type constituents which were identified as martensite and austenite by means of Selected Area Diffraction (SAD). Carbides were also observed, and their characterization is described later. In general, the austenite and martensite laths appeared thinner in the QT10-Pt1000 condition. Moreover, a large number of laths were thinner than the detection limit employed in EBSD analysis (~100 nm). The laths identified as austenite in the QT25-Pt1000 condition had a width of 53 ± 5 nm, which would explain the lower RA contents measured by EBSD in comparison with XRD measurements. With the aim of understanding the partitioning behaviour of Mn and Ni, TEM – Energy Dispersive Spectroscopy (EDS) analysis was performed and the compositional maps (Figure 39c-f) and concentration profiles (Figure 39g-h) were obtained for Mn and Ni alloying elements. The compositional maps revealed that the Mn concentration was not homogeneous, with austenite films enriched in this element. On the contrary, Ni concentration was more homogeneous. The concentration profiles obtained for the QT25-Pt1000 condition (Figure 39g), revealed high Mn enrichment in the region identified as austenite by SAD, obtaining a maximum concentration of 15 wt.%, and a lower Mn concentration in the laths identified as martensite. In the latter, Mn content was around the nominal value (grey laths) or above this value (white laths), and consequently, the laths were deduced to be primary and secondary martensite, respectively. Likely, initial austenite laths existing at QT25 were too large to be completely enriched in Mn and, as a consequence, a Mn gradient was observed from the boundary to the center of the initial austenite lath, existing a higher Mn concentration near the boundary. Hence, secondary martensite transformed from the interior of the austenite laths remaining from cooling to the QT, originating MA islands in the final microstructure [112,114]. On the contrary, in the QT10 condition (Figure 39h) a higher percentage of the lath was enriched in Mn and the concentration profile was more homogeneous. Furthermore, the Mn concentration in the enriched laths was significantly higher than in the QT25 condition. Therefore, it can be concluded that in the high partitioning temperature Q&P cycles applied in this work, Mn partitioning occurred and the lower QT condition was beneficial for the Mn enrichment of austenite laths, which likely contributed to austenite stabilization. Ni concentration profiles did not show such clear concentration differences between laths as 76 those of Mn, although in the QT10-Pt1000 cycle Ni concentration profile showed a trend similar to that shown by Mn. The effect of Ni is further discussed in section 4.1.6. Figure 39. TEM analysis of 6Mn2Ni steel: Micrographs and SAD patterns for the QT25 (a) and the QT10 (b) conditions; EDS analysis representing the distribution of Mn concentration for the QT25 (c) and QT10 (d) conditions, and Ni concentration for the QT25 (e) and QT10 (f) conditions; concentration profiles of Mn and Ni line scans performed within the areas marked in (a) and (b) for the QT10 (g) and QT25 (h) conditions. 6Mn2Ni –QT10-Pt1000 6Mn2Ni –QT25-Pt1000 0 5 10 15 20 0.0 0.2 0.4 0.6 0.8 wt.% Distance (μm) 0 5 10 15 20 0.0 0.2 0.4 0.6 0.8 wt.% Distance (μm) α’ (h) Ni 200 nm Ni 200 nm Mn 200 nm Mn 200 nm (c) (g) Mn Ni Mn Ni RA RA α’sec RA RA RA RA α’ α’ 200 nm (b) 200 nm (a) 01-1 10-1 [111] α’ [101] -202 020 -202 020 [101] 020 200 [001] α’ γ α’ (d) (e) (f) α’ α’ γα’ 77 The microstructure characterization was finished by performing a TEM-EDS analysis of the carbides observed in the microstructure. As presented before, globular and acicular/plate carbides were present in the microstructure. The TEM-EDS analysis of the carbides revealed that the carbides contained Mn. As an example, the EDS spectrum of a globular carbide observed in the 6Mn steel after the QT10-Pt1000 cycle is shown in Figure 40b. Compositional mapping in an area examined in the 6Mn2Ni steel after the QT10-Pt1000 cycle, which contained both globular and acicular carbides, revealed the enrichment of both types of carbides with Mn (Figure 40d). In the 6Mn2Ni steel, the TEM-EDS analysis did not show the presence of Ni in the carbides. The formation of Mn enriched carbides resulted in less C and Mn available for austenite stabilization. Figure 40. TEM micrographs of the 6Mn steel (a) and the 6Mn2Ni steel (c) after QT10-Pt1000 cycle, TEM-EDS spectrum of a globular carbide marked in (a) (b), and compositional mapping of the (c) micrograph (d). In the micrograph (c), acicular and globular carbides are visible; in the compositional map (d) a higher concentration of Mn is visible in the globular and acicular carbides. 4.1.3 Tensile properties Figure 41, Figure 42, Figure 43 and Figure 44 show the stress-strain curves for all Q&P cycles of the 2Mn, 4Mn, 6Mn and 6Mn2Ni steels, respectively. All data obtained from the tensile tests are summarized in Table 16: Yield strength (YS), tensile strength (TS), YS/TS ration, total elongation (TEL), and TS x TEL product. These data correspond to the average values obtained from two tests. In Appendix C the photographs of one fractured tensile specimen per steel and condition is shown. Tensile properties are related to the microstructure in section 4.1.8. 200 nm (a) 6Mn –QT10-Pt1000 0 100 200 300 400 500 600 700 800 0 2 4 6 8 Intensity (Counts) Energy (keV) C (b) EDS spectrum Mn-L Fe-L Si Mn-Kα Fe-Kα Mn-Kβ Fe-Kβ 200 nm 200 nm Acicular carbides Globular carbides Acicular carbides Globular carbides (c) 6Mn2Ni –QT10-Pt1000 (d) 6Mn2Ni –QT10-Pt1000 Mn 78 Figure 41. Engineering stress-strain curves obtained after Q&P treatments for the 2Mn steel. Figure 42. Engineering stress-strain curves obtained after Q&P treatments for the 4Mn steel. Figure 43. Engineering stress-strain curves obtained after Q&P treatments for the 6Mn steel. 0 200 400 600 800 1000 1200 1400 0 5 10 15 20 25 30 Engineering Stress (MPa) Engineering Strain (%) QT10-Pt1000 QT25-Pt1000 ref 2Mn 0 200 400 600 800 1000 1200 1400 0 5 10 15 20 25 30 Engineering Stress (MPa) Engineering Strain (%) QT10-Pt1000 QT25-Pt1000 ref 4Mn 0 200 400 600 800 1000 1200 1400 0 5 10 15 20 25 30 Engineering Stress (MPa) Engineering Strain (%) QT10-Pt300 QT10-Pt1000 QT10-Pt3600 QT25-Pt1000 ref 6Mn 79 Figure 44. Engineering stress-strain curves obtained after Q&P treatments for the 6Mn2Ni steel. Table 16. Summary of tensile properties for the 2Mn, 4Mn, 6Mn and 6Mn2Ni steels after each Q&P treatment. Steel ref. Q&P cycle YS (0.2% offset) (MPa) TS (MPa) YS/TS ratio TEL (%) TS x TEL (GPa%) 2Mn ref 974 1273 0.76 9.9 12.5 QT25-Pt1000 512 767 0.67 18.3 14.0 QT10-Pt1000 523 759 0.69 20.1 15.2 4Mn ref 833 1259 0.66 3.7 4.7 QT25-Pt1000 581 1182 0.49 13.6 16.1 QT10-Pt1000 524 1064 0.49 20.3 21.5 6Mn ref 735 993 0.74 0.6 0.5 QT25-Pt1000 585 1251 0.47 6.9 8.6 QT10-Pt300 644 1158 0.56 24.6 28.5 QT10-Pt1000 651 1199 0.54 24.9 29.9 QT10-Pt3600 584 1204 0.48 17.4 20.9 6Mn2Ni ref 896 1170 0.77 1.6 1.9 QT25-Pt1000 580 1329 0.44 11.1 14.7 QT10-Pt300 727 1182 0.61 23.7 28.0 QT10-Pt1000 694 1207 0.57 28.1 33.8 QT10-Pt3600 705 1222 0.58 28.8 35.2 It can be seen that in the Q&P cycles with a partitioning temperature of 400 °C (ref cycle), the increase in Mn resulted in a noticeable decrease in elongation, and, also, in tensile strength. The further addition of Ni in 6Mn2Ni did not result in an improvement of tensile properties, obtaining, also in this case, a low elongation. 0 200 400 600 800 1000 1200 1400 0 5 10 15 20 25 30 Engineering Stress (MPa) Engineering Strain (%) QT10-Pt300 QT10-Pt3600 QT10-Pt1000 QT25-Pt1000 ref 6Mn2Ni 86 austenite reversion kinetics would be in the PLE-1 stage. During this stage the austenite fraction was still growing, but the reverted austenite amount formed so far led to a relatively high amount of RA. By 1000 s of partitioning the growth of austenite was slowed and it is likely that the kinetics were at the end of the PLE-1 stage or the beginning of the PLE-2 stage. Reverted austenite was not far from its maximum and thus outstanding values of RA were obtained in this condition. Finally, at 3600 s of partitioning, the controlling kinetics would already be in the PLE-2 stage and no increase of austenite fraction occurred, leading to the slight decrease of RA measured by XRD. From a chemical point of view, a large difference in austenite stability was not expected between the 1000 s and 3600 s partitioning time conditions. However, in FE-SEM microstructures (Figure 37d and Figure 37f) a coarsening of austenite laths was observed. This behavior was hypothesized to occur due to the coalescence of austenite grains during growth [218]. As mentioned above, grain size might influence the thermal stability of austenite [125], which can explain the slight decrease in RA measured after 3600 s partitioning time cycles in comparison with 1000 s cycles, where thinner austenite was observed. 4.1.6 Analysis of the influence of Ni on microstructure and austenite stabilization The addition of Ni in the 6Mn2Ni steel was beneficial for increasing RA content in the final microstructure in all heat treatment conditions. The Ni addition did not seem to affect significantly the austenite formation kinetics (as denoted by the similar contraction observed in the partitioning stage, Figure 33 and Figure 35). However, Ni was effective in stabilizing a higher content of austenite. The latter could be related to an enrichment of austenite not only with Mn, but also with Ni. The behaviour of Ni partitioning was analyzed based on compositional maps and concentration profiles shown in Figure 39. In the QT10 condition, some redistribution of Ni can be appreciated in the compositional map (Figure 39f), although it was not as distinctive as in the Mn map. However, the concentration profile was very similar to that shown by Mn (Figure 39h), with areas enriched in Ni (apparently austenite) and areas depleted in Ni (identified as martensite), in the same positions as in the Mn profile. Thereby, the QT10 condition also seemed favourable for Ni partitioning. In the case of the QT25 condition, the Ni partitioning behaviour was less clear and would require further investigation. On the other hand, as presented in section “4.1.2 Microstructure characterization”, the TEM-EDS analysis revealed that the carbides contained Mn. Besides, in the 6Mn2Ni steel, the analysis seemed to indicate that Ni was not present in the carbides. The composition of 87 carbides in Ni containing steels was studied in recent work. Pierce et al. [54] observed plate-like and globular carbides in TEM after partitioning at 450 ºC for 300 s in a 0.2C-1.5Mn-1.3Si steel containing 1.5% Ni. Atom probe tomography (APT) revealed Mn enrichment of the carbides. They did not observe significant partitioning of Ni between carbide and matrix. Thus, both the morphology and chemical composition of the carbides observed in the present work were in agreement with that observed by Pierce et al. Similarly, Clarke et al. [219] reported no substantial partitioning of Ni between the matrix and the carbide in 4340 steel during tempering at 450 ºC for 2 h. However, Clarke et al. showed via APT that significant rejection of Ni from carbides occurred during tempering at 575 ºC for 2 h, with Ni enrichment persisting near the carbide/ferrite interface. With the aim of further investigating the possible Ni enrichment in the carbides, a line scan was performed in TEM across two globular carbides in the QT10-Pt1000 condition for 6Mn2Ni steel (Figure 47b). Both carbides contained a high amount of Mn, but no Ni enrichment was evident in the carbides or near the carbide/matrix interfaces. Figure 47. TEM micrograph of the 6Mn2Ni steel after the QT10-Pt1000 condition (a); and line scans of Mn and Ni weight concentration (b), corresponding to the line shown in (a). 4.1.7 Theoretical analysis of austenite stabilization using DICTRA As shown earlier, the ART occurred in the high partitioning temperature stage and Mn partitioning from martensite into austenite was observed in TEM-EDS scans. The Mn concentration profiles were different depending on the QT condition. Furthermore, a substantial carbide fraction was observed, which contained a significant content of Mn. According to recent studies [220–223], austenite is likely to nucleate at carbide interfaces. Therefore, austenite formation could be the result of the growth of pre-existing austenite along with formation of new austenite at carbide interfaces. With the aim of better 0 5 10 15 20 25 30 0 0.05 0.1 0.15 0.2 wt.% Distance (μm) 200 nm Mn Ni (a) 6Mn2Ni –QT10-Pt1000 (b) 6Mn2Ni –QT10-Pt1000 88 understanding these possible phenomena, DICTRA simulations were performed for the 6Mn2Ni steel considering the partitioning conditions employed in the present work. In the simulations two different set ups were considered. In the first one, the growth of pre-existing austenite in the austenite/martensite microstructures was simulated. In the second one, austenite nucleation at carbide interfaces was simulated. In both set ups, a single cell planar geometry was used [224]. Because of symmetry, only the half thickness was considered in the DICTRA calculation domain. Ferrite was considered instead of martensite, since martensite is not included in the thermodynamic and kinetic database. The results of these simulations were the Mn concentration profile in the considered systems and the growth of austenite for different partitioning times. The growth of austenite can be observed by the displacement of the interface. In set up (1), 10 and 25 volume percent of austenite (QT10 and QT25 conditions, respectively) were considered in the initial austenite/martensite microstructure. The dimension of the martensite was set to 0.200 μm. It was based on the TEM observations of Krauss and colleagues, indicating that most martensitic lath widths range from approximately 0.150 to 0.200 μm [226,227], which seems consistent with the TEM analysis presented in this work. Corresponding austenite dimensions were obtained using the “constant ferrite width approach” [225]; i.e., the width of the austenite lath corresponded to a percentage of the total width (martensite + austenite) equal to the volume percent of austenite fixed in the QT. Thus, austenite initial lath widths equal to 0.22 or 0.66 μm were fixed. In the case of the QT10 condition, different Pt-s were considered. A schematic illustration of these initial conditions is shown in Figure 48a, and the results for the simulation comparing different QT and Pt conditions are shown in Figure 48b and Figure 48d, and Figure 48c and Figure 48e, respectively. In set up (2), the average carbide size and Mn composition measured from TEM analysis were considered, 0.020 μm and 24.6 wt.%, respectively. The partitioning dimension in martensite was set to 0.200 μm, as in set up (1). The initial composition of α phase was considered from the equilibrium composition for the 6Mn2Ni steel at 640 ºC obtained by Thermo-Calc (i.e., 0.003 C wt.% and 1.998 Mn wt.%). In addition, in order to simplify the simulation, the nucleation process was ignored, adding a 0.001 μm length austenite phase, with the same composition as the martensite, between the cementite and martensite phases. The initial conditions for this set up are shown in Figure 49a and the obtained results are shown in Figure 49b. 89 Figure 48. DICTRA simulations for 6Mn2Ni steel at 640 ºC: schematic of the initial conditions of the simulations for set up (1) (a); results for set up (1), showing a comparison between the growth of austenite in QT10 and QT25 with Pt1000 and diffusion of Mn (b) and Ni (d) (the initial interface was set in the same position for QT10 and QT25), and showing the comparison between the growth of austenite with Pt300, Pt1000, and Pt3600 in QT10 and diffusion of Mn (c) and Ni (e). In set up (1), austenite grew into the initial martensite lath and was enriched considerably in Mn, while the martensite was considerably depleted, almost reaching the equilibrium value (Figure 48b). With regard to Ni, it can be seen that, even the austenite was partially enriched, and the martensite was partially depleted, a 1000 s partitioning was not enough time to reach the equilibrium composition. Furthermore, the distribution of the Ni in the austenite lath was 0 1 2 3 4 5 6 7 8 9 10 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Mn wt.% Distance (μm) QT10 QT25 Initial γ/αinterface (b) 6Mn2Ni, set up (1), Pt1000 conditions Eq. Mn in FCC Eq. Mn in BCC γ α 0 1 2 3 4 5 6 7 8 9 10 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Mn wt.% Distance (μm) Initial γ/αinterface Pt300 Pt1000 Pt3600 (c) 6Mn2Ni, set up (1), QT10 conditions Eq. Mn in FCC Eq. Mn in BCC γ α γα 0.033 μm 0.100 μm γα 0.100 μm0.011 μm set up for the QT25 condition set up for the QT10 condition (a) Initial conditions for set up (1) 0 1 2 3 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Ni wt.% Distance (μm) QT10 QT25 Initial γ/αinterface Eq. Mn in FCC Eq. Mn in BCC γ α (d) 6Mn2Ni, set up (1), Pt1000 conditions 0 1 2 3 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Ni wt.% Distance (μm) Pt300 Pt1000 Pt3600 Initial γ/αinterface Eq. Mn in FCC Eq. Mn in BCC γ α (e) 6Mn2Ni, set up (1), QT10 conditions 90 quite heterogeneous (Figure 48d). Regarding the influence of QT, the DICTRA calculations appear to be in good agreement with the dilatometry results, where it was seen that the QT10 condition led to a greater amount of austenite formation (Figure 35). Thus, according to Figure 48b and Figure 48d, after 1000 s at the PT, the austenite lath size grew more in the QT10 condition, obtaining a final length 7.5 times the initial one, whereas for the QT25 condition the size was only 3.0 times the initial. With regards to Mn and Ni diffusion distance, the enriched width was quite similar for both QT conditions. However, the percentage of the lath size enriched in Mn was greater for the QT10 than the QT25 condition, specifically 78% and 60% austenite was enriched in the QT10 and the QT25, respectively. These percentages are somewhat lower regarding Ni diffusion distance, specifically 61% and 50% in the QT10 and the QT25, respectively. These observations were consistent with TEM-EDS line scans (Figure 39g and Figure 39h), and would imply that the quantity of secondary martensite formed in the QT25 condition during final cooling would be higher. Calculations for different Pt-s, employing QT10, are shown in Figure 48c and Figure 48e. As expected, the growth of the austenite was greater with increased partitioning time, as the interface was moved to longer distances. The growth increment difference with increasing partitioning time from 300 to 1000 s was substantially larger than increasing it from 1000 s to 3600 s, although in these DICTRA simulations the aforementioned coalescence phenomenon was not considered. At Pt300, there was a clear distinction between the lath region enriched in Mn and the non-enriched region, which might lead to some secondary martensite formation. In addition, the martensite lath was not completely Mn depleted. At Pt1000, Mn was more homogeneously distributed across the lath, but there was still a small not enriched length, while at Pt3600 all the length was somewhat enriched, although a compositional difference within the lath was still visible and, thereby, the less enriched part of the lath would potentially transform into secondary martensite. Regarding Ni diffusion, the austenite length without enrichment was the same for the three Pt conditions. However, the Ni distribution within the lath was different. At Pt300, the Ni enrichment/depletion in the austenite and martensite, respectively, was poor and very localized nearby the interface. At Pt1000, the Ni enrichment in the austenite continued being located near the interface, but the maximum peak was over the equilibrium composition. The Ni depletion in the martensite led to a homogeneous distribution, although the content was still over the equilibrium value. Finally, at Pt3600, the behavior was like at Pt1000, obtaining a slightly higher Ni enrichment in the austenite near the interface. 91 Figure 49. DICTRA simulations for 6Mn2Ni steel at 640 ºC: schematic of the initial conditions of the simulations for set up (2) (a), and results for set up (2) configuration, where austenite nucleated at θ/α interface (b). In set up (2), different austenite sizes and Mn profiles were obtained when comparing with set up (1). First, the Mn content across the austenite lath was almost constant and very close to the equilibrium content (8.4 wt.%). Second, the austenite lath size was significantly smaller than the size obtained in set up (1), specifically, the total size of the lath after 3600 s of partitioning did not reach 0.008 μm, with even lower values after 1000 and 300 s, 0.005 and 0.003 μm, respectively. In addition, according to DICTRA calculations, the austenite grew into the martensite rather than dissolving the cementite. Yan et al. [228] also observed that the lath size of austenite was much smaller when it nucleated at the interface between cementite and martensite than when it nucleated at the boundary of the martensite. However, comparing with the set up (1) on this work, they did not have pre-existing austenite in their simulations and, comparing with the set up (2). they found that carbides were totally dissolved after holding for 2000 s at the intercritical temperature. Additionally, Luo et al. [229] suggested that the dissolution of cementite was the main reason for the high enrichment of austenite with Mn, and the limited diffusion distance of Mn resulted in a refined austenite. In general, Mn contents measured experimentally in austenite (Figure 39g-h) were significantly higher than those obtained by DICTRA. Furthermore, in DICTRA simulations, the austenite width enriched in Mn was less than observed by TEM-EDS scans. The regions θ α 0.010 μm 0.100 μm (a) Initial conditions for set up (2) 0 5 10 15 20 25 0.008 0.01 0.012 0.014 0.016 0.018 Mn wt.% Distance (μm) Pt300 Pt1000 Pt3600 Initial θ/αinterface Eq. Mn in FCC Eq. Mn in BCC 0 5 10 15 20 25 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 Mn wt.% Distance (μm) (b) 6Mn2Ni, set up (2), γnucleated at θ/αinterface Zoom γ/α θ/γ 92 not enriched in Mn (or less enriched) would lead to a quite higher extent of secondary martensite transformation than that observed by dilatometry. This means that DICTRA calculations could underestimate Mn diffusion through the austenite. As said before, in the simulations ferrite was considered instead of martensite, since martensite is not included in the thermodynamic and kinetic database. The thermodynamic properties of martensite are very much the same as those of ferrite, however, the kinetic parameters may deviate between the two phases due to larger numbers of lattice defects, particularly dislocations, in martensite. Thus, mobilities of all alloying elements in martensite are increased compared to ferrite [124], which could explain the differences between the simulations and experimental data. 4.1.8 Relationship between microstructure and tensile properties As presented in “4.1.3 Tensile properties” section, tensile properties strongly decreased by the addition of Mn in the ref cycle. The addition of Ni did not result in an improvement of these properties. Although the retained austenite content measured by XRD increased by the addition of Mn and Ni, in the present case, this increase did not result in better tensile properties. By rising the partitioning temperature to the start of ART temperature, it was shown that tensile strength decreased, and elongation increased in the 2Mn and 4Mn steels, which was mainly related to the tempering of the martensite phase. From now on, the relationship between microstructure and tensile properties is analyzed for the 6Mn and 6Mn2Ni steels heat treated by high partitioning temperature cycles. First, with the aim of better analyzing the effect of the Q&P cycle conditions and the addition of Ni on the tensile properties, the tensile results are shown as a function of partitioning time in Figure 50. In general, the better tensile properties were obtained with the 6Mn2Ni steel in the QT10 conditions, except for TS, which was higher after the application of QT25-Pt1000 cycle. For the 6Mn2Ni steel, the properties did not deteriorate with the increase of the Pt, whereas in the 6Mn steel YS and TEL decreased for the Pt of 3600 s. The product of tensile strength and total elongation is a parameter often applied to evaluate and compare steels for the automotive sector. In Figure 51, TS x TEL is represented as a function of RA. Except for QT10-Pt300 condition in the 6Mn steel, it can be said that the product TS x TEL increased with the austenite content. Thus, the presence of a higher content of RA in Ni containing steel and after the QT10 condition, likely contributed to the outstanding TS x TEL values obtained in this case. Moreover, the finer austenite laths for the 93 QT10 condition likely played an important role in improving the TS x TEL product [230,231]. On the contrary, the greater amount of secondary martensite reduced the ductility for the QT25 condition. Figure 50. Tensile properties of the Q&P treated samples: yield strength (a); tensile strength (b); total elongation (c); and TEL x TS product (d). Figure 51. Graphical representation of the relationship between the product of tensile strength and total elongation and RA for each Q&P cycle of the 6Mn and 6Mn2Ni steels. 500 550 600 650 700 750 800 0 1000 2000 3000 4000 Yield Stregth (Mpa) Partitioning time (s) (a) 6Mn2Ni –QT10 6Mn2Ni –QT25 6Mn –QT25 6Mn –QT10 1100 1150 1200 1250 1300 1350 1400 0 1000 2000 3000 4000 Tensile Stregth (Mpa) Partitioning time (s) (b) 6Mn2Ni –QT10 6Mn2Ni –QT25 6Mn –QT25 6Mn –QT10 0 5 10 15 20 25 30 35 0 1000 2000 3000 4000 Total Elongation (%) Partitioning time (s) (c) 6Mn2Ni –QT10 6Mn –QT25 6Mn2Ni –QT25 6Mn –QT10 0 5 10 15 20 25 30 35 40 0 1000 2000 3000 4000 TEL x TS (GPa%) Partitioning time (s) (d) 6Mn2Ni –QT10 6Mn –QT25 6Mn2Ni –QT25 6Mn –QT10 0 5 10 15 20 25 30 35 40 20 25 30 35 40 45 50 TS x TEL (GPa%) Retained Austenite (%) 6Mn 6Mn2Ni QT25-Pt1000 QT10-Pt300 QT10-Pt1000 QT10-Pt3600 94 The presence of secondary martensite can be considered further, as it can be critical for the deterioration of tensile properties [112]. Even though secondary martensite transformation was difficult to detect by dilatometry for some of the conditions, a deeper analysis allows an approximation of the amount of each constituent in the final microstructure. Thus, the volume fraction of the phases existing in the final microstructure after heat treatments was estimated employing RA measurements (XRD) and dilatometry curves (Figure 52). Secondary martensite was obtained by comparing the change in length at the final cooling with the change in length given by the martensite transformation curve on the directly quenched sample, and primary martensite was calculated by balance [37]. From the graph, it can be confirmed that the greatest content of secondary martensite was obtained after the QT25-Pt1000 cycle, particularly in the 6Mn steel. The presence of a hard and brittle secondary martensite in this condition can explain the lower total elongation and higher tensile strength obtained after the QT25-Pt1000 cycle (Figure 50). Figure 52. Volume percent of each phase in the 6Mn and 6Mn2Ni steels after the application of Q&P cycles calculated based on dilatometry curves and RA measured by XRD. On the other hand, it is widely accepted that the transformation of austenite into martensite induced by deformation is also a critical factor that determines tensile properties. From the results obtained in the interrupted tensile tests, the evolution of normalized RA with strain in QT10-Pt1000 cycles for the 6Mn and 6Mn2Ni steels was represented (Figure 53). In both steels, little austenite transformation occurred up to 10% strain, where only about 15% of the austenite had transformed. Afterwards, the behavior was rather different in each steel. In the 6Mn steel the austenite content gradually decreased with strain, whereas in the 6Mn2Ni steel 56 55 70 59 52 50 56 52 31 38 25 37 43 48 39 44 13 7545254 0 20 40 60 80 100 Phase volume (%) primary martensite retained austenite secondary martensite 6Mn 6Mn2Ni6Mn 6Mn2Ni6Mn 6Mn2Ni 6Mn 6Mn2Ni QT25-Pt1000 QT10-Pt300 QT10-Pt1000 QT10-Pt3600 95 the entirety of the remaining RA at 20% strain (77% of the initial RA) transformed close to the uniform strain, likely resulting in a high fraction of very hard martensite and thus provoking early fracture and the absence of post-uniform elongation. Although the 6Mn steel showed a more gradual decrease in the austenite fraction, it is likely that the amount of martensite transformed at strains close to the uniform elongation was still high and also provoked the absence of post-uniform elongation. From the evolution of RA with strain, it can be deduced that adding Ni increased the mechanical stability of retained austenite and retarded the kinetics of the strain-induced martensite transformation. It can be explained by considering the effect of Ni on the strength of martensite. First, according to Hidalgo et al. [92] the strength of martensite can affect the mechanical stability of austenite. They concluded that austenite surrounded by a stronger martensitic matrix was mechanically more stable than that surrounded by a weaker martensite. Second, in the present work, Ni decreased the Ms temperature, so the martensite formed during the first cooling to QT should contain more dislocations and, thus, it was likely harder and contributed to the higher mechanical stability of austenite in this steel. Figure 53. Evolution of normalized RA during tensile tests stopped at different strains for 6Mn and 6Mn2Ni steels after QT10-Pt1000 cycle. Stress-strain curves in Figure 43 and Figure 44 indicated discontinuous yielding in QT10-Pt1000 and Pt3600 cycles, but not so in the shorter-time Pt300 condition. Raabe et al. [231] recently reviewed different mechanisms contributing to the occurrence of discontinuous yielding. They observed that ultra-refining the grain sizes, materials which normally exhibit continuous yielding, such as pure Al, austenitic steels and interstitial-free (IF) steels, yield discontinuously. In the present work, a very fine martensite/austenite microstructure was observed by TEM characterization after the QT10-Pt1000 cycle in the 6Mn and 6Mn2Ni 0 10 20 30 40 50 60 70 80 90 100 0 5 10 15 20 25 30 Normalized RA (%) Engineering strain (%) 6Mn2Ni –QT10-PT1000 6Mn –QT10-Pt1000 102 As it can be seen in Figure 59a, the main precipitated phase in 300M steel was cementite, therefore, the carbides observed inside the tempered martensite in Figure 57 were likely this phase. In the 1.2990 steel up to around 10% of M7C3 and M23C6 phases can be expected in the microstructure (Figure 59b) according to Thermo-Calc, being M = Fe, Cr, V or Mo. The composition of each phase as a function of temperature is shown in Figure 59c-d. Both phases contain a high content of Cr, but only M7C3 phase contain a relatively high V content. Besides, M23C6 phase shows a considerable Mo content, which was barely detected by EDS analysis. Therefore, it can be said that the phases analyzed in Figure 58a-b (points x2, x3 and x4) might be M7C3, although a further research would be needed to better determine the type of carbide. According to Thermo-Calc, the dissolution temperature of M7C3 phase (~1220 ºC) is significantly higher than the austenization temperature employed in the Q&P treatments (1037 ºC), which means that likely these carbides were eutectic carbides present in the initial microstructure or precipitated during austenization and were not significantly affected by the heat treatment [234,235]. In cast irons with elevated contents of Cr, eutectic M7C3 carbides tend to transform into M23C6 carbides. However, when Cr content was below 10-25 wt.% M7C3 carbides do not appear to undergo any structural changes [235]. On the other hand, the austenization and subsequent quenching might led to a formation of secondary spheroid M7C3 particles (point x2 in Figure 58a) within the primary austenitic phase [236]. 4.2.4 Hardness Hardness measurements were performed in the Q&T and Q&P treated samples in both steels. The average values obtained after five indentations are shown in Figure 60 and the values obtained after each indentation are shown in Appendix D. The Q&P treatment resulted in a considerably lower hardness, especially in the 1.2990 steel. The 1.2990 steel contains relatively high contents of carbides forming elements like Cr (7.9%), V (1.5 wt.%) and Mo (1 wt.%). Thus, in the tempering, apart from eliminating the stresses generated during the martensitic transformation and soften the structure restoring the toughness of the material, secondary precipitation hardening can be achieved [10,98,99]. As it was seen in Figure 58, Cr, Mo and V containing carbides were visible in the microstructure of the Q&P treated 1.2990 steel. However, the total amount of carbides was likely lower than in the Q&T steel. Furthermore, the Q&P steel contained carbon-enriched retained austenite which likely lowered the overall hardness value [237]. 103 In 300M steel, Q&P treatment resulted in lower hardness as well, although the decrease was less significant in comparison with the 1.2990 steel. Unlike in the 1.2990 steel, precipitation hardening does not occur in this steel, only little cementite precipitation inside the tempered martensite was observed (Figure 57). Thus, tempering treatment is only employed to eliminate residual stresses produced during the quenching and restore the toughness [8,238]. As in the 1.2990 steel, the presence of carbon-enriched austenite left less carbon available for the tempered martensite, thus decreasing the overall hardness of the Q&P steel. Figure 60. Hardness values measured in the 1.2990 and 300M steels after the application of Q&T or Q&P heat treatments. 4.2.5 Wear behavior Wear rates were calculated from the pin-on-disk tests for both steels and both Q&P and Q&T treatments. Figure 61 shows comparisons between the values obtained with each steel and heat treatment. When comparing the influence of the heat treatment (Figure 61a-b), the difference in the wear rate was insignificant both in the 1.2990 and the 300M steels. However, it was shown before that the difference in hardness between the Q&P and Q&T treated steels was considerably high, especially in the 1.2990 steel. Thus, Q&P treated samples showed a significantly lower hardness but the same wear resistance than Q&T samples. This behavior demonstrates that hardness was not the only parameter affecting the wear resistance and that the retained austenite may play a key role [186]. On the other hand, when comparing the Q&P treated steels at the same linear velocity conditions (Figure 61c), as expected, the 1.2990 steel clearly showed a better wear resistance. Both steels had a similar RA content, hence, in this case, the hardness and the 62.7 56.5 54.5 51.2 0 10 20 30 40 50 60 70 1.2990 300M Hardness (HRC) Q&T Q&P 104 composition of the steel were decisive when determining the wear response of each steel. The 1.2990 steel has a higher C content and hardness and additionally, it is very likely that the carbides observed in the microstructure of this steel were M7C3 and M23C6, as predicted by Thermo-Calc Figure 59a), which resulted beneficial for wear resistance [239]. Figure 61. Wear rate (mm3/N.m) calculated from the PoD tests. Comparison between Q&P and Q&T treated 1.2990 (a) and 300M (b) steels; and comparison between Q&P treated 1.2990 and 300M steels tested with the same linear velocity (c). It is difficult to compare the results obtained in this work with those reported in the literature. On the one hand, there is very little literature studying the wear resistance applied to Q&P steels, and even less studying the wear behavior through unidirectional sliding tests. On the other hand, the results obtained from PoD tests strongly depend on the testing conditions, therefore, the obtained wear rate results are specific for the employed conditions. However, most of the results reported in the literature show an improvement in the wear resistance by the application of Q&P heat treatments. Wasiak et al. [187], for example, observed that, after being Q&P treated, the 35CrSiMn5-5-4 steel exhibited twice better wear resistance under the dry sliding friction than after being Q&T treated. Wang et al. [240] found that by applying a Q&P treatment to a ductile cast iron wear resistance was improved and that the employed partitioning time affected the obtained wear rate, first increasing with the increase of the partitioning time and then decreasing with a further time increasing. In [192], the authors concluded that compared with Q&T treatment, wear resistance of a high C steel was improved by applying a series of Q&PT (quenching & partitioning-tempering) treatments, which was associated with the formation of film-like and blocky austenite during partitioning stage, but an increase in the partitioning temperature from 250 ºC to 400 ºC was found to be disadvantageous due to an increase in cementite carbide precipitation. 8.55 8.59 0 2 4 6 8 10 12 14 Q&P Q&T K (x10-6 mm3N-1 m-1) 11.00 10.60 Q&P Q&T 8.78 11.00 1.2990 300M (a) 1.2990 (208.3 cm/s) (b) 300M (78.3 cm/s) (c) Q&P (78.3 cm/s) 105 In the present work, wear resistance was not improved by the substitution of the Q&T by Q&P treatment. However, only one Q&P condition was studied, and it is possible that the optimization of the Q&P cycle parameters could result in an improvement. Nevertheless, the results showed similar wear rates after both treatments with significantly lower hardness after Q&P, which might be advantageous for other properties typically required in this type of steels, such as toughness. 4.2.6 Retained austenite stability The austenite transformation into martensite it is thought to be responsible for enhancing the wear behavior in Q&P steels [186]. Therefore, austenite stability was studied measuring the RA content in the wear track of the PoD tested disk in order to examinate the transformation given during the tests. In Figure 62, the RA contents measured by XRD in the wear track are shown for both steels and different linear velocity conditions employed (the latter only for 1.2990 steel) and these values are compared with RA contents measured before in the surface of the disks. Figure 62. RA measured in the surface and wear tracks formed during PoD characterization of the 1.2990 and 300M steels. In the 1.2990 steel and the lower linear velocity condition, austenite transformation was not observed, therefore, material did not benefit from transformation hardening. The higher linear velocity condition resulted in the transformation of less than 20% of the initial RA content. Therefore, it can be said that the mechanical stability of austenite against transformation in this steel, and, in the test conditions employed in this work, was too high. It is known that the C content is the main factor affecting RA stability [129,241,242]. In the 1.2990 steel the C 22 20 23 12 18 0 4 8 12 16 20 24 28 1.2990 300M RA (%) Surface Wear track (78.3 cm/s) Wear track (208.3 cm/s) 106 content was high (0.92%), and it is likely that, in spite of the high fraction of carbides present in the microstructure, the C content in the RA after the Q&P treatment was too high to allow martensite transformation. Possibly, with a higher load or linear velocity the improvement through the application of Q&P with respect to Q&T would be clearer, although further study is required. In the 300M steel, due to material limitations, only the test with lower linear velocity was performed. In this case, a higher amount of the austenite transformed during the tests, specifically 40% of the initial RA. This steel contains a lower C content (0.41%), which might result in a lower C content in RA after Q&P treatment and consequently in a higher transformation into austenite. However, this transformation seemed to be still low to be beneficial for wear resistance. Same as in the 1.2990 steel, it is likely that employing a higher load or lineal velocity conditions the results would show a clearer improvement after the application of the Q&P treatment. Apart from the C content, it is known that other factors can affect the stability of austenite. It has been reported that film RA can be mechanically more stable in martensitic microstructures than blocky RA, suggesting that this stability may be due to surrounding martensite laths suppressing the transformation of film RA [129,243,244]. In the SEM image of the 300M steel (Figure 57) some areas with blocky austenite were observed, whereas in the 1.2990 steel (Figure 58) austenite laths predominated. Furthermore, in the case of the 1.2990 steel, a stronger matrix could lead to an increase in the mechanical stability of RA and sometimes this makes the transformation of RA to martensite not to occur even under the actions of stress or strain [191]. In this work, only a preliminary study of the wear behavior of Q&P steels was performed, and future tests and characterization would be needed to better elucidate the factors determining the stability of austenite in both steels and the influence of the microstructure in wear behavior. 4.2.7 Toughness As described before, 1.2990 steel showed important hardness differences between the Q&P and Q&T treated samples whereas the wear rates were similar. The lower hardness values measured in Q&P samples were expected to result in a better toughness [191–193], which would be a clear advantage, as those applications requiring wear resistance generally also require toughness. Thus, Charpy V-notch impact test were performed in both the Q&P and Q&T samples employing subsize standard specimens. 107 The toughness values obtained are summarized in Table 18 and the photographs of each specimen are shown in Appendix E. As it can be seen, the obtained values were too low to make an accurate comparison between the Q&P and Q&T treated steels. Particularly, the toughness of the Q&P steel was quite lower than expected based on literature results. Lai et al. [192,193] reported impact toughness values in the range of 17-22 J/cm2 for a steel containing 1.2 wt.% C and RA and hardness values similar to those obtained in this work, 18% RA and 600 HV (~54,5 HRC), respectively. In addition, Liu et al. [191] measured an impact toughness of 19.4 J/cm2 in a 0.95 wt.% C steel with 20% of RA and an approximate hardness of 600 HV, although the steel had a bainitic matrix rather than martensitic. An explanation for the low toughness can be related with the machining operation of the Charpy specimens. Due to the complexity of machining the dimensions of the V-notch, it was done by wire cutting. There is no way to prove that through this cutting operation the austenite around the notch did not transform into martensite, which could be a possible reason for obtaining such low values in the impact tests, although further investigations would be required. Table 18. Toughness values (J/cm2) obtained in Charpy V-notch impact test of the Q&P and Q&T treated 1.2990 steel. Toughness (J/cm2) Test 1 Test 2 Test 3 Average Q&T 5.7 4.9 3.8 4.8 Q&P 4.5 3.7 3.7 4.0 108 109 Chapter 5 Conclusions 110 111 Chapter 5 Conclusions In this thesis four medium Mn and Ni steel, and three medium and high C steels treated by different Q&P treatments were investigated. In this chapter the conclusions drawn from all the obtained results are presented. 5.1 Q&P applied to medium Mn and Ni steels Conventional Q&P cycles and high partitioning temperature Q&P cycles in which the austenite reverse transformation phenomenon occurred, were investigated in four medium Mn steel with different Mn and Ni amounts. The following conclusions were drawn: 1. The addition of 2 wt.% Mn was not sufficient to stabilize high amounts of austenite. In the ref cycle competing reactions occurred during partitioning, likely decomposition of austenite into bainite. In high partitioning temperature cycles secondary martensite transformation occurred during final cooling, resulting in a final microstructure with less than 5 wt.% of RA. 2. In the steels with four or more wt.% of Mn, both austenite content and tensile properties were improved with the application of the high partitioning temperature cycles, independently of the QT and Pt conditions of the cycle. 3. In the 6Mn and 6Mn2Ni steels, treated with high PT cycles, large amounts of retained austenite were obtained in the final microstructure, which were comparable to those obtained after intercritical annealing of medium Mn steels. The addition of Ni further increased the content of retained austenite. The presence of pre-existing austenite (due to interrupted quenching) before the partitioning stage considerably reduced the partitioning time needed for austenite stabilization in comparison with some conventional intercritical annealing heat treatments of medium Mn steel. 4. In the 6Mn and 6Mn2Ni steels, a lower quenching temperature resulted in faster austenite formation kinetics in the high temperature partitioning stage, less formation of secondary martensite in the final cooling, and a final microstructure with a higher content of RA, which showed a lath-type microstructure with finer constituents. 118 6.2 Q&P applied to medium and high C steels In this thesis, only a preliminary study of the benefits of Q&P treatment for wear resistance requiring applications was performed. There is still a lot of background work to be done in the study of Q&P steels in relation to wear applications and toughness. • The influence of different microstructures on the wear performance of the steels should be studied. Particularly, the effect that the amount of each phase as retained austenite, secondary martensite or bainite has in wear resistance should be tested. • Besides austenite amount, how its carbon content and morphology influence on its stability against wear tests must be study. For that purpose, microstructures with different RA contents, carbon in RA and morphology should be created and studied through wear tests employing different conditions. • Once the effect of the microstructure is understood, an optimization of Q&P cycle parameters is necessary to achieve the desired microstructures. • In this work only pin-on-disk test were carried out. The behavior of Q&P steels in other types of wear tests, such as abrasive test, could be studied. • Regarding toughness, in this work an improvement after the application of the Q&P treatment was not verified, despite the lower hardness. There is uncertainty as to whether the retained austenite transformed during notch machining, therefore Charpy tests could be performed on unnotched specimens to check whether an improvement in toughness is indeed seen after application of the Q&P cycle. • Since thermal gradients between the surface and the center of the disks were not observed during the heat treatments, the feasibility of applying the heat treatment into thicker specimens without obtaining heterogeneities in the microstructure could be studied. • A study about energetical and economic advantages of replacing the conventional Q&T treatment by the Q&P treatment should be performed. • In this part of the work, high partitioning temperature Q&P cycles were not applied, as further study is still required to better understand the influence of low partitioning temperature treatments on wear and toughness behavior. The application of high partitioning temperatures could also be interesting as part of future work. 119 Chapter 7 Bibliography 120 121 Chapter 7 Bibliography [1] T. Nanda, V. Singh, V. Singh, A. Chakraborty, S. Sharma, Third generation of advanced high-strength steels: Processing routes and properties, Proceedings of the Institution of Mechanical Engineers, Part L: Journal of Materials: Design and Applications. 233 (2019) 209–238. doi:10.1177/1464420716664198. [2] WorldAutoSteel, Advanced High-Strength Steels Guidelines Version 7.0, WorldAutoSteel. (2021). https://www.worldautosteel.org/steel-basics/automotive-advanced-high-strengthsteel-ahss-definitions/. [3] E. De Moor, P.J. Gibbs, J.G. Speer, D.K. Matlock, J.G. Schroth, Strategies for thirdgeneration advanced high-strength steel development, Iron and Steel Technology. 7 (2010) 133–144. [4] J. Zhao, Z. Jiang, Thermomechanical processing of advanced high strength steels, Progress in Materials Science. 94 (2018) 174–242. doi:10.1016/j.pmatsci.2018.01.006. [5] E. De Moor, Advanced High-Strength Sheet Steels for Automotive Applications BT - HighPerformance Ferrous Alloys, in: R. Rana (Ed.), Springer International Publishing, Cham, 2021: pp. 113–151. doi:10.1007/978-3-030-53825-5_4. [6] J. Speer, D.K. Matlock, B.C. De Cooman, J.G. Schroth, Carbon partitioning into austenite after martensite transformation, Acta Materialia. 51 (2003) 2611–2622. doi:10.1016/S1359-6454(03)00059-4. [7] A.M. Streicher, J.G. Speer, D.K. Matlock, B.C. De Cooman, Quenching and partitioning response of a Si-added TRIP sheet steel, International Conference on Advanced High Strength Sheet Steels for Automotive Applications - Proceedings. (2004) 51–62. [8] N. Fonstein, Advanced High Strength Sheet Steels, 2015. doi:10.1007/978-3-319-191652. [9] S.C. Hong, Mechanical properties of high-Si plate steel produced by the quenching and partitioning process, Metals and Materials International. 13 (2007) 439–445. doi:10.3365/mmi.2007.12.439. [10] S. Zhou, K. Zhang, N. Chen, J. Gu, Y. Rong, Investigation on High Strength Hot-rolled Plates by Quenching-partitioning-tempering Process Suitable for Engineering, ISIJ International. 51 (2011) 1688–1695. doi:10.2355/isijinternational.51.1688. [11] M.C. Somani, L.P. Karjalainen, D.A. Porter, D.K. Misra, Evaluation of the behaviour and properties of a high-Si steel processed using direct quenching and partitioning, Materials Science Forum. 706–709 (2012) 2824–2829. doi:10.4028/www.scientific.net/MSF.706709.2824. [12] J.G. Speer, F.C. Rizzo Assunção, D.K. Matlock, D. V. Edmonds, The “quenching and partitioning” process: Background and recent progress, Materials Research. 8 (2005) 417–423. doi:10.1590/S1516-14392005000400010. [13] D. V. Edmonds, K. He, F.C. Rizzo, B.C. De Cooman, D.K. Matlock, J.G. Speer, Quenching and partitioning martensite—A novel steel heat treatment, Materials Science and Engineering: A. 438–440 (2006) 25–34. doi:10.1016/J.MSEA.2006.02.133. [14] S.G. Sarikaya M, Thomas G, Steeds JW, Barnard SJ, Solute element partitioning and austenite stabilization in steels, Proceedings of an International Conference on Solid to Solid Phase Transformations. (1982) 1421–1425. 122 [15] Y. Toji, G. Miyamoto, D. Raabe, Carbon partitioning during quenching and partitioning heat treatment accompanied by carbide precipitation, Acta Materialia. 86 (2015) 137–147. doi:10.1016/J.ACTAMAT.2014.11.049. [16] H.S. Yang, D.W. Suh, H.K.D.H. Bhadeshia, More complete theory for the calculation of the martensite-start temperature in steels, ISIJ International. 52 (2012) 164–166. doi:10.2355/isijinternational.52.164. [17] J.G. Speer, E. De Moor, A.J. Clarke, Critical assessment 7: Quenching and partitioning, Materials Science and Technology (United Kingdom). 31 (2015) 3–9. doi:10.1179/1743284714Y.0000000628. [18] H.K.D.H. Bhadeshia, Carbon in cubic and tetragonal ferrite, Philosophical Magazine. 93 (2013) 3714–3725. doi:10.1080/14786435.2013.775518. [19] J.H. Jang, H.K.D.H. Bhadeshia, D.W. Suh, Solubility of carbon in tetragonal ferrite in equilibrium with austenite, Scripta Materialia. 68 (2013) 195–198. doi:10.1016/j.scriptamat.2012.10.017. [20] J.G. Speer, R.E. Hackenberg, B.C. Decooman, D.K. Matlock, Influence of interface migration during annealing of martensite/austenite mixtures, Philosophical Magazine Letters. 87 (2007) 379–382. doi:10.1080/09500830701194173. [21] G.A. Thomas, J.G. Speer, Interface migration during partitioning of Q&P Steel, Materials Science and Technology (United Kingdom). 30 (2014) 998–1007. doi:10.1179/1743284714Y.0000000546. [22] X. Zhanga, G. Shenb, C. Lic, J. Gud, Analysis of interface migration and isothermal martensite formation for quenching and partitioning process in a low-carbon steel by phase-field modelling, Modelling and Simulation in Materials Science and Engineering. 27 (2019) 075011. doi:https://doi.org/10.1088/1361-651X/ab3608. [23] N. Zhong, X. Wang, Y. Rong, L. Wang, Interface migration between martensite and austenite during quenching and partitioning (Q&P) process, Journal of Materials Science and Technology. 22 (2006) 751–754. [24] H.C. Lee, G. Krauss, Intralath carbide transitions in martensitic mediumcarbon steels tempered between 200 and 300C, in: Fundamentals of Aging and Tempering in Bainitic and Martensitic Steel Products, 1992: pp. 39–43. [25] G. Krauss, Tempering and structural change in ferrous martensitic structures, in: International Conference on Phase Transformations in Ferrous Alloys, 1984: pp. 101–123. [26] D.C.B. Streicher AM, Speer JG, Matlock DK, Quenching and partitioning response of a Siadded TRIP sheet steel, in: Proceedings of the International Conference on Advanced High-Strength Sheet Steels for Automotive Applications, 2004: pp. 51–62. [27] G. Speer, J.; Streicher, A.; Matlock, D.; Rizzo, F.; Krauss, Quenching and partitioning: a fundamentally new process to create high strength TRIP sheet microstructures, in: Austenite Formation and Decomposition, 2003: pp. 505–522. [28] L. Barbé, K. Conlon, B.C. De Cooman, Characterization of the metastable austenite in low-alloy FeCMnSi TRIP-aided steel by neutron diffraction, Zeitschrift Fuer Metallkunde/Materials Research and Advanced Techniques. 93 (2002) 1217–1225. doi:10.3139/146.021217. [29] M. Hillert, J. Ågren, On the definitions of paraequilibrium and orthoequilibrium, Scripta Materialia. 50 (2004) 697–699. doi:10.1016/J.SCRIPTAMAT.2003.11.020. [30] J.G. Speer, D.K. Matlock, B.C. DeCooman, J.G. Schroth, Comments on “On the definitions of paraequilibrium and orthoequilibrium” by M. Hillert and J. Ågren, Scripta Materialia, 50, 697–9 (2004), Scripta Materialia. 52 (2005) 83–85. doi:10.1016/J.SCRIPTAMAT.2004.08.029. [31] M. Hillert, J. Ågren, Reply to comments on “On the definition of paraequilibrium and orthoequilibrium,” Scripta Materialia. 52 (2005) 87–88. doi:10.1016/J.SCRIPTAMAT.2004.08.026. 123 [32] A.S. Nishikawa, M.J. Santofimia, J. Sietsma, H. Goldenstein, Influence of bainite reaction on the kinetics of carbon redistribution during the Quenching and Partitioning process, Acta Materialia. 142 (2018) 142–151. doi:10.1016/j.actamat.2017.09.048. [33] H. Kawata, K. Hayashi, N. Sugiura, N. Yoshinaga, M. Takahashi, Effect of martensite in initial structure on bainite transformation, Materials Science Forum. 638–642 (2010) 3307–3312. doi:10.4028/www.scientific.net/MSF.638-642.3307. [34] K. Seto, H. Matsuda, Application of nanoengineering to research and development and production of high strength steel sheets, Materials Science and Technology (United Kingdom). 29 (2013) 1158–1165. doi:10.1179/1743284713Y.0000000226. [35] J. Mahieu, J. Maki, B.C. De Cooman, S. Claessens, Phase Transformation and Mechanical Properties of Si-Free CMnAl Transformation-Induced Plasticity-Aided Steel, Metallurgical and Materials Transactions A. 33A (2002) 2573–2580. [36] J. Sun, H. Yu, S. Wang, Y. Fan, Study of microstructural evolution, microstructuremechanical properties correlation and collaborative deformation-transformation behavior of quenching and partitioning (Q&P) steel, Materials Science and Engineering: A. 596 (2014) 89–97. doi:10.1016/J.MSEA.2013.12.054. [37] M. Arribas, T. Gutiérrez, E. Del Molino, A. Arlazarov, I. De Diego-Calderón, D. Martin, D. De Caro, S. Ayenampudi, M.J. Santofimia, Austenite Reverse Transformation in a Q&P Route of Mn and Ni Added Steels, Metals. 10 (2020) 862. doi:10.3390/met10070862. [38] J. Hidalgo, C. Celada-Casero, M.J. Santofimia, Fracture mechanisms and microstructure in a medium Mn quenching and partitioning steel exhibiting macrosegregation, Materials Science and Engineering: A. 754 (2019) 766–777. doi:10.1016/J.MSEA.2019.03.055. [39] S. Ayenampudi, C. Celada-Casero, J. Sietsma, M.J. Santofimia, Microstructure evolution during high-temperature partitioning of a medium-Mn Quenching and Partitioning steel, Materialia. (2019). doi:10.1016/j.mtla.2019.100492. [40] F. HajyAkbary, J. Sietsma, G. Miyamoto, T. Furuhara, M.J. Santofimia, Interaction of carbon partitioning, carbide precipitation and bainite formation during the Q&P process in a low C steel, Acta Materialia. 104 (2016) 72–83. doi:10.1016/J.ACTAMAT.2015.11.032. [41] D.V. Edmonds, K. He, M.K. Miller, F.C. Rizzo, A. Clarke, D.K. Matlock, J.G. Speer, Microstructural Features of “Quenching and Partitioning”: A New Martensitic Steel Heat Treatment, Materials Science Forum. 539–543 (2007) 4819–4825. doi:10.4028/www.scientific.net/msf.539-543.4819. [42] F. HajyAkbary, M.J. Santofimia, J. Sietsma, Optimizing mechanical properties of a 0.3C1.5Si-3.5MnQuenched and partitioned steel, Advanced Materials Research. 829 (2014) 100–104. doi:10.4028/www.scientific.net/AMR.829.100. [43] M.J. Santofimia, L. Zhao, J. Sietsma, Overview of mechanisms involved during the quenching and partitioning process in steels, Metallurgical and Materials Transactions A: Physical Metallurgy and Materials Science. 42 (2011) 3620–3626. doi:10.1007/s11661011-0706-z. [44] H.K.D.H. Bhadeshia, S.R. Honeycombe, The Effects of Alloying Elements on Iron-Carbon Alloys, Steels. (2006) 71–93. doi:10.1016/B978-075068084-4/50006-6. [45] D. De Knijf, E.P. Da Silva, C. Föjer, R. Petrov, Study of heat treatment parameters and kinetics of quenching and partitioning cycles, Materials Science and Technology (United Kingdom). 31 (2015) 817–828. doi:10.1179/1743284714Y.0000000710. [46] A. Arlazarov, O. Bouaziz, J.P. Masse, F. Kegel, Characterization and modeling of mechanical behavior of quenching and partitioning steels, Materials Science and Engineering A. 620 (2014) 293–300. doi:10.1016/j.msea.2014.10.034. [47] I. De Diego-Calderón, D. De Knijf, J.M. Molina-Aldareguia, I. Sabirov, C. Föjer, R. Petrov, Effect of Q&P parameters on microstructure development and mechanical behaviour of Q&P steels, Revista de Metalurgia. 51 (2015). doi:10.3989/revmetalm.035. [48] M.J. Santofimia, T. Nguyen-Minh, L. Zhao, R. Petrov, I. Sabirov, J. Sietsma, New low 124 carbon Q&P steels containing film-like intercritical ferrite, Materials Science and Engineering: A. 527 (2010) 6429–6439. doi:10.1016/J.MSEA.2010.06.083. [49] E. De Moor, J.G. Speer, D.K. Matlock, J.H. Kwak, S.B. Lee, Effect of carbon and manganese on the quenching and partitioning response of CMnSi steels, ISIJ International. 51 (2011) 137–144. doi:10.2355/isijinternational.51.137. [50] K. Ren, Y. Kang, S. Zhu, Effect of partitioning time on the microstructure and mechanicalproperties of Q&P steels, Materials Science Forum. 749 (2013) 303–307. doi:10.4028/www.scientific.net/MSF.749.303. [51] I. De Diego-Calderón, M.J. Santofimia, J.M. Molina-Aldareguia, M.A. Monclús, I. Sabirov, Deformation behavior of a high strength multiphase steel at macroand micro-scales, Materials Science and Engineering: A. 611 (2014) 201–211. doi:10.1016/J.MSEA.2014.05.068. [52] S. Ayenampudi, Z. Arechabaleta, M. Arribas, Microstructural Impact of Si and Ni During High Temperature Quenching and Partitioning Process in Medium-Mn Steels, Metallurgical and Materials Transactions A. 52 (2021) 5–11. doi:10.1007/s11661-02106144-5. [53] K. Kim, S.J. Lee, Effect of Ni addition on the mechanical behavior of quenching and partitioning (Q&P) steel, Materials Science and Engineering A. 698 (2017) 183–190. doi:10.1016/j.msea.2017.05.030. [54] D.T. Pierce, D.R. Coughlin, K.D. Clarke, E. De Moor, J. Poplawsky, D.L. Williamson, B. Mazumder, J.G. Speer, A. Hood, A.J. Clarke, Microstructural evolution during quenching and partitioning of 0.2C-1.5Mn-1.3Si steels with Cr or Ni additions, Acta Materialia. 151 (2018) 454–469. doi:10.1016/j.actamat.2018.03.007. [55] E.J. Seo, L. Cho, B.C. De Cooman, Application of Quenching and Partitioning Processing to Medium Mn Steel, Metallurgical and Materials Transactions A: Physical Metallurgy and Materials Science. 46 (2015) 27–31. doi:10.1007/s11661-014-2657-7. [56] E.J. Seo, L. Cho, B.C. De Cooman, Application of Quenching and Partitioning (Q&P) Processing to Press Hardening Steel, Metallurgical and Materials Transactions A. 45A (2014) 4022–4037. doi:10.1007/s11661-014-2316-z. [57] X. Tan, Y. Xu, X. Yang, D. Wu, Microstructure–properties relationship in a one-step quenched and partitioned steel, Materials Science and Engineering: A. 589 (2014) 101– 111. doi:10.1016/J.MSEA.2013.09.063. [58] H. Jirková, L. Kučerová, B. Mašek, The Effect of Chromium on Microstructure Development During Q-P Process, Materials Today: Proceedings. 2 (2015) S627–S630. doi:10.1016/J.MATPR.2015.07.362. [59] F.C. Rizzo, A.R. Martins, J.G. Speer, D.K. Matlock, A. Clarke, B.C. De Cooman, Quenching and Partitioning of Ni-Added High Strength Steels, Materials Science Forum. 539–543 (2007) 4476–4481. doi:10.4028/www.scientific.net/msf.539-543.4476. [60] G.T. Zhang, N.Q. Zhu, B.W. Sun, Z.Z. Zhao, Z.W. Zheng, D. Tang, L. Li, Effect of v addition on microstructure and mechanical properties in c–mn–si steels after quenching and partitioning processes, Metals. 11 (2021). doi:10.3390/met11081306. [61] X.D. Tan, X.L. Yang, Y.B. Xu, Z.P. Hu, F. Peng, H. Zhang, Y.M. Yu, D. Wu, Effect of ferrite status on mechanical properties of hot-rolled directly quenched and partitioned steel, Materials Science Forum. 816 (2015) 736–742. doi:10.4028/www.scientific.net/MSF.816.736. [62] J. Tobata, K.L. Ngo-Huynh, N. Nakada, T. Tsuchiyama, S. Takaki, Role of silicon in quenching and partitioning treatment of lowcarbon martensitic stainless steel, ISIJ International. 52 (2012) 1377–1382. doi:10.2355/isijinternational.52.1377. [63] G. Krauss, Steels - Processing, Structure, and Performance, 2nd ed., ASM International, 2012. [64] J.N.T.M. S. Morito, Dislocation Density within Lath Martensite in Fe–C and Fe–Ni Alloys, ISIJ International. 43 (2003) 1475–1477. 125 [65] M. Cheng, H.W. Song, X. Li, S.H. Zhang, M. Cheng, T. Lin, Effects of quenching and partitioning process on mechanical properties of a hot-stamping steel, Materials Science Forum. 788 (2014) 340–345. doi:10.4028/www.scientific.net/MSF.788.340. [66] J.G. Speer, E. De Moor, K.O. Findley, D.K. Matlock, B.C. De Cooman, D. V Edmonds, Analysis of Microstructure Evolution in Quenching and Partitioning Automotive Sheet Steel, Metallurgical and Materials Transactions A. 42 (2011) 3591–3601. doi:10.1007/s11661-011-0869-7. [67] W.S. Owen, The Effect of Silicon on the Kinetics of Tempering, Transactions of the ASM. 46 (1954) 812–829. [68] S.S. Nayak, R. Anumolu, R.D.K. Misra, K.H. Kim, D.L. Lee, Microstructure–hardness relationship in quenched and partitioned medium-carbon and high-carbon steels containing silicon, Materials Science and Engineering: A. 498 (2008) 442–456. doi:10.1016/J.MSEA.2008.08.028. [69] T. Sourmail, F.G. Caballero, F. Moudian, D. De Castro, M. Benito, High hardness and retained austenite stability in Si-bearing hypereutectoid steel through new heat treatment design principles, Materials & Design. 142 (2018) 279–287. doi:10.1016/J.MATDES.2018.01.035. [70] C. Garcia-Mateo, F.G. Caballero, T. Sourmail, J. Cornide, V. Smanio, R. Elvira, Composition design of nanocrystalline bainitic steels by diffusionless solid reaction, Metals and Materials International. 20 (2014) 405–415. doi:10.1007/s12540-014-3002-9. [71] T. Sourmail, F.G. Caballero, C. Garcia-Mateo, V. Smanio, C. Ziegler, M. Kuntz, R. Elvira, A. Leiro, E. Vuorinen, T. Teeri, Evaluation of potential of high Si high C steel nanostructured bainite for wear and fatigue applications, Materials Science and Technology (United Kingdom). 29 (2013) 1166–1173. doi:10.1179/1743284713Y.0000000242. [72] T. Sourmail, V. Smanio, Low temperature kinetics of bainite formation in high carbon steels, Acta Materialia. 61 (2013) 2639–2648. doi:10.1016/J.ACTAMAT.2013.01.044. [73] V. Ollilainen, W. Kasprzak, L. Holappa, The effect of silicon, vanadium and nitrogen on the microstructure and hardness of air cooled medium carbon low alloy steels, Journal of Materials Processing Technology. 134 (2003) 405–412. doi:10.1016/S09240136(02)01131-7. [74] W. Zhanshan, Z. Li, W. Lv, S. Zhenyao, The influence of the substitution of Si by Al on the properties of hot rolled C-MN-Si TRIP steel, Materials Science Forum. 896 MSF (2017) 198–201. doi:10.4028/www.scientific.net/MSF.896.198. [75] J. Mahieu, B.C. De Cooman, S. Claessens, Galvanizability of high-strength steels for automotive applications, Metallurgical and Materials Transactions A: Physical Metallurgy and Materials Science. 32 (2001) 2905–2908. doi:10.1007/s11661-001-1042-5. [76] and K.E.Y.W. D. Vanderschueren, M. De Meyer, B. C. De Cooman, The Influence of the Substitution of Si by Al on the Properties of Cold Rolled C-Mn-Si TRIP Steel, ISIJ International. 39 (1999) 813–822. doi:10.4028/www.scientific.net/MSF.896.198. [77] E. Girault, A. Mertens, P. Jacques, Y. Houbaert, B. Verlinden, J. Van Humbeeck, Comparison of the effects of silicon and aluminium on the tensile behaviour of multiphase TRIP-assisted steels, Scripta Materialia. 44 (2001) 885–892. doi:10.1016/S13596462(00)00697-7. [78] M.C. Somani, D.A. Porter, L.P. Karjalainen, P.P. Suikkanen, D.K. Misra, Innovation and processing of novel tough ductile ultra-high strength steels through TMR-DQP processing route, Materials Science Forum. 783–786 (2014) 1009–1014. doi:10.4028/www.scientific.net/msf.783-786.1009. [79] H.L. Yi, P. Chen, Z.Y. Hou, N. Hong, H.L. Cai, Y.B. Xu, D. Wu, G.D. Wang, A novel design: Partitioning achieved by quenching and tempering (Q–T & P) in an aluminium-added lowdensity steel, Scripta Materialia. 68 (2013) 370–374. doi:10.1016/J.SCRIPTAMAT.2012.10.018. [80] J. De Moor, E.; Speer, J. G.; Matlock, D. K.; Föjer, C.; Penning, Effect of Si, Al and Mo 126 Alloying on Tensile Properties Obtained by Quenching and Partitioning, in: Materials Science and Technology, 2009: pp. 1554–1563. [81] C. Krishna Ande, M.H. Sluiter, First-Principles Calculations on Stabilization of Iron Carbides (Fe3C, Fe5C2, and η-Fe2C) in Steels by Common Alloying Elements, Metallurgical and Materials Transactions A. 43 (2012) 4436–4444. doi:10.1007/s11661012-1229-y. [82] P.J. Jacques, E. Girault, A. Mertens, B. Verlinden, J. Van Humbeeck, F. Delannay, The developments of cold-rolled TRIP-assisted multiphase steels. Al-alloyed TRIP-assisted multiphase steels, ISIJ International. 41 (2001) 1068–1074. doi:10.2355/isijinternational.41.1068. [83] L. Samek, E. De Moor, J. Penning, B.C. De Cooman, Influence of Alloying Elements on the Kinetics of Strain-Induced Martensitic Nucleation in Low-Alloy, Multiphase HighStrength Steels, Metallurgical and Materials Transactions A. 37A (2006) 109–124. [84] S. Kim, C. Gil Lee, I. Choi, S. Lee, Effects of Heat Treatment and Alloying Elements on the Microstructures and Mechanical Properties of 0.15 Wt Pct C Transformation-Induced Plasticity-Aided Cold-Rolled Steel Sheets, Metallurgical and Materials Transactions A. 32A (2001) 505–514. doi:https://doi.org/10.1007/s11661-001-0067-0. [85] E. De Moor, J.G. Speer, D.K. Matlock, J.H. Kwak, S.B. Lee, Quenching and partitioning of CMnSi steels containing elevated manganese levels, Steel Research International. 83 (2012) 322–327. doi:10.1002/srin.201100318. [86] E. De Moor, J.G. Speer, Bainitic and quenching and partitioning steels, Elsevier Ltd, 2017. doi:10.1016/B978-0-08-100638-2.00010-9. [87] J. Wang, P.J. Van Der Wolk, S. Van Der Zwaag, On the influence of alloying elements on the bainite reaction in low alloy steels during continuous cooling, Journal of Materials Science. 35 (2000) 4393–4404. [88] S.J. Lee, D.K. Matlock, C.J. Van Tyne, An empirical model for carbon diffusion in austenite incorporating alloying element effects, ISIJ International. 51 (2011) 1903–1911. doi:10.2355/isijinternational.51.1903. [89] J.G.S. E. De Moor, C. Föjer, A.J. Clarke, J. Penning, Quench & partitioning response of a Mo-alloyed CMnSi steel, in: Conf. New Dev. Metall. Appl. High Strength Steels, Minerals, Metals, & Materials, 2008: pp. 721–729. [90] G. Kirchner, T. Nishizawa, B. Uhrenius, The distribution of chromium between ferrite and austenite and the thermodynamics of the α/γ equilibrium in the Fe-Cr and Fe-Mn Systems, Metallurgical Transactions. 4 (1973) 167–174. doi:10.1007/BF02649616. [91] A.K. Patra, C.N. Athreya, S. Mandal, K.C. Hari Kumar, V. Subramanya Sarma, High strength-high ductility medium Mn steel obtained through CALPHAD based alloy design and thermomechanical processing, Materials Science and Engineering: A. 810 (2021) 140756. doi:10.1016/J.MSEA.2021.140756. [92] J. Hidalgo, R.M. Huizenga, K.O. Findley, M.J. Santofimia, Interplay between metastable phases controls strength and ductility in steels, Materials Science and Engineering: A. 745 (2019) 185–194. doi:10.1016/J.MSEA.2018.12.096. [93] S. Mishra, R.P. Dalai, Effect of quenching and partitioning treatment on low carbon medium manganese alloyed steels-A short review, Materials Today: Proceedings. 43 (2021) 593–596. doi:10.1016/J.MATPR.2020.12.107. [94] P. Xu, C. Li, W. Li, M. Zhu, K. Zhang, Effect of microstructure on hydrogen embrittlement susceptibility in quenching-partitioning-tempering steel, Materials Science and Engineering: A. 831 (2022) 142046. doi:10.1016/J.MSEA.2021.142046. [95] P. hu Chen, Y. bo Li, R. qing Li, R. peng Jiang, S. sheng Zeng, X. qian Li, Microstructure, mechanical properties, and wear resistance of VCp-reinforced Fe-matrix composites treated by Q&P process, International Journal of Minerals, Metallurgy and Materials. 25 (2018) 1060–1069. doi:10.1007/s12613-018-1657-9. [96] Y. Xu, X. Tan, X. Yang, Z. Hu, F. Peng, D. Wu, G. Wang, Microstructure evolution and 127 mechanical properties of a hot-rolled directly quenched and partitioned steel containing proeutectoid ferrite, Materials Science and Engineering: A. 607 (2014) 460–475. doi:10.1016/J.MSEA.2014.04.030. [97] N. Zhong, Y. Wang, K. Zhang, Y.H. Rong, Microstructual evolution of a Nb-microalloyed advanced high strength steel treated by quenching-partitioning-tempering process, Steel Research International. 82 (2011) 1332–1337. doi:10.1002/srin.201100125. [98] N. Zhong, X.D. Wang, L. Wang, Y.H. Rong, Enhancement of the mechanical properties of a Nb-microalloyed advanced high-strength steel treated by quenching–partitioning– tempering process, Materials Science and Engineering: A. 506 (2009) 111–116. doi:10.1016/J.MSEA.2008.11.014. [99] S. Zhou, K. Zhang, Y. Wang, J.F. Gu, Y.H. Rong, High strength-elongation product of Nbmicroalloyed low-carbon steel by a novel quenching-partitioning-tempering process, Materials Science and Engineering A. 528 (2011) 8006–8012. doi:10.1016/j.msea.2011.07.008. [100] C. Kickinger, C. Suppan, T. Hebesberger, R. Schnitzer, C. Hofer, Microstructure and mechanical properties of partially ferritic Q&P steels, Materials Science and Engineering: A. 815 (2021) 141296. doi:10.1016/J.MSEA.2021.141296. [101] R. Ding, D. Tang, A. Zhao, R. Dong, J. Cheng, X. Zhang, A new type of quenching and partitioning processing developed from martensitic pre-microstructure, Materials and Manufacturing Processes. 29 (2014) 704–709. doi:10.1080/10426914.2014.912304. [102] R. Ding, D. Tang, A. Zhao, R. Dong, J. Cheng, X. Meng, Effect of intercritical temperature on quenching and partitioning steels originated from martensitic pre-microstructure, Journal of Materials Research. 29 (2014) 2525–2533. doi:10.1557/jmr.2014.285. [103] G. Liu, S. Zhang, J. Li, J. Wang, Q. Meng, Fast-heating for intercritical annealing of coldrolled quenching and partitioning steel, Materials Science and Engineering: A. 669 (2016) 387–395. doi:10.1016/J.MSEA.2016.05.106. [104] D. De Knijf, A. Puype, C. Föjer, R. Petrov, The influence of ultra-fast annealing prior to quenching and partitioning on the microstructure and mechanical properties, Materials Science and Engineering: A. 627 (2015) 182–190. doi:10.1016/J.MSEA.2014.12.118. [105] M.J. Santofimia, L. Zhao, J. Sietsma, Model for the interaction between interface migration and carbon diffusion during annealing of martensite–austenite microstructures in steels, Scripta Materialia. 59 (2008) 159–162. doi:10.1016/J.SCRIPTAMAT.2008.02.045. [106] J. Sun, H. Yu, Microstructure development and mechanical properties of quenching and partitioning (Q&P) steel and an incorporation of hot-dipping galvanization during Q&P process, Materials Science and Engineering: A. 586 (2013) 100–107. doi:10.1016/J.MSEA.2013.08.021. [107] A.J. Clarke, J.G. Speer, M.K. Miller, R.E. Hackenberg, D. V. Edmonds, D.K. Matlock, F.C. Rizzo, K.D. Clarke, E. De Moor, Carbon partitioning to austenite from martensite or bainite during the quench and partition (Q&P) process: A critical assessment, Acta Materialia. 56 (2008) 16–22. doi:10.1016/J.ACTAMAT.2007.08.051. [108] M.J. Santofimia, L. Zhao, R. Petrov, C. Kwakernaak, W.G. Sloof, J. Sietsma, Microstructural development during the quenching and partitioning process in a newly designed low-carbon steel, Acta Materialia. 59 (2011) 6059–6068. doi:10.1016/J.ACTAMAT.2011.06.014. [109] E.A. Ariza-Echeverri, M. Masoumi, A.S. Nishikawa, D.H. Mesa, A.E. Marquez-Rossy, A.P. Tschiptschin, Development of a new generation of quench and partitioning steels: Influence of processing parameters on texture, nanoindentation, and mechanical properties, Materials & Design. 186 (2019) 108329. doi:10.1016/J.MATDES.2019.108329. [110] M.J. Santofimia, L. Zhao, R. Petrov, J. Sietsma, Characterization of the microstructure obtained by the quenching and partitioning process in a low-carbon steel, Materials Characterization. 59 (2008) 1758–1764. doi:10.1016/J.MATCHAR.2008.04.004. [111] E.J. Seo, L. Cho, B.C. De Cooman, Kinetics of the partitioning of carbon and substitutional alloying elements during quenching and partitioning (Q&P) processing of medium Mn