Effect of Intercritical Annealing Temperature on Austenite Formation in Medium-Mn Steels: A Thermodynamic and Experimental Study

Effect of Intercritical Annealing Temperature on Austenite Formation in Medium-Mn Steels: A Thermodynamic and Experimental Study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Effect of Intercritical Annealing Temperature on Austenite Formation in Medium-Mn Steels: A Thermodynamic and Experimental Study Mateusz MORAWIEC, Jarosław OPARA, Aleksandra KOZŁOWSKA, Adam GRAJCAR This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6319163/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 8 You are reading this latest preprint version Abstract The effects of intercritical annealing (IA) temperature and manganese content on austenite formation and stability in medium-Mn steels were investigated using combined computational and experimental approaches. Three steels containing 3, 4, and 5 wt.% Mn were annealed at 700–760 °C for 60 min, and their microstructures were analysed to assess the influence of Mn content on phase transformation behaviour. Thermodynamic (JMatPro, Thermo-Calc) and kinetic (DICTRA) simulations were used to model phase stability, austenite growth, and elemental partitioning under both equilibrium and non-equilibrium conditions. The modelling results were validated using dilatometry, scanning electron microscopy (SEM), and energy-dispersive X-ray spectroscopy (EDS). The results showed that increasing Mn content promotes higher austenite fractions during IA but reduces its carbon enrichment, which adversely affects thermal stability due to higher martensite start (Ms) temperatures. DICTRA simulations also revealed that Mn and Al develop distinct concentration gradients across the ferrite/austenite interface, especially in higher-Mn steels, and that these gradients correlate with the measured EDS values in retained austenite laths. The revealed microstructures were composed of ferrite, retained austenite, and small amounts of cementite and martensite-austenite constituents. Overall, the study demonstrates that Mn content strongly affects both the amount and stability of retained austenite, and that IA parameters must be carefully optimized to tailor microstructure and mechanical behaviour in medium-Mn advanced high-strength steels. medium-Mn steel thermodynamic modelling kinetic modelling intercritical annealing retained austenite thermodynamic stability element partitioning Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 1. Introduction Typically, medium-Mn automotive sheet steels subjected to intercritical annealing (IA) heat treatment develop a multiphase microstructure predominantly composed of ferrite and austenite [1-3], but may also contain martensitic-austenitic (M–A) constituents or small amounts of other phases depending on heat treatment parameters. In the intercritical range between A c1 and A c3 temperatures, ferrite and austenite form from either an initial ferrite-pearlite or a martensitic microstructure [4]. The nature of this initial microstructure significantly affects both the kinetics of austenite and ferrite formation and the morphology of the resulting phases. A duplex ferrite–austenite microstructure that develops from lath martensite tends to inherit a layered morphology, whereas starting with a ferrite-pearlite microstructure leads to more globular morphologies [5]. Moreover, the final intercritical microstructure is significantly influenced by the prior thermomechanical history. Hot rolling typically yields a lath-type starting microstructure [5], which, upon intercritical annealing, retains a layered morphology; in contrast, cold-rolled material that undergoes recrystallization during heating leads to a more globular morphology [4]. In general, multiphase steels with a lath-type microstructure exhibit a superior combination of high strength and ductility [7]. Morawiec et al. [8] reported that increasing the Mn content from 3 to 5 wt.% fosters the formation of such lath-type microstructures due to enhanced hardenability. In addition, the higher dislocation density in lath-type microstructures suppresses Lüders elongation and flow curve serrations in stress–strain curves [9, 10]. An essential structural phenomenon during the IA of multiphase steels is the partitioning of key alloying elements, particularly carbon and manganese, from ferrite to austenite [11]. As a result, a high austenite fraction (20-40%) can be stabilized at room temperature (RT) [1-4]. Because carbon atoms have a small atomic radius, C diffusion occurs relatively quickly, which is crucial for maintaining austenite stability in automotive steel sheets produced in continuous annealing lines, with soaking times of 60– 180 s. Conversely, the slower diffusion of Mn necessitates significantly longer times. Therefore, manganese is used to enhance austenite stability mainly in batch annealed steels, requiring several hours of soaking [12]. Both elements effectively lower the martensite start ( M s ) temperature, and their combined concentrations commonly suppress M s temperature below RT. It should be also noted that higher Mn contents in steel reduce the degree of carbon enrichment in the austenitic phase [13]. Elements like Al and Si can also diffuse preferentially to ferrite; in particular, Al elevates M s temperature, diminishing the thermodynamic stability of austenite. Nevertheless, aluminium remains a popular addition in various Advanced High Strength Steels (AHSS) due to its lightweighting potential [14]. Mn is notably prone to microsegregation [15]. In addition, it can hinder the redistribution of C from ferrite to austenite [16]. As a result, the ferrite-to-austenite ratio strongly depends on the austenite’s chemical composition and time-temperature parameters applied during IA process [17, 18]. The amount and stability of retained austenite (RA) significantly influence the mechanical properties of medium-Mn steels, making the optimization of IA conditions a critical issue. Furthermore, recent studies have expanded the understanding of austenite formation and stability in medium-Mn steels. In particular, manganese partitioning has been shown to proceed preferentially along grain boundaries and dislocations, thereby accelerating austenite growth during intercritical annealing [19]. The formation of Mn-enriched cementite can also influence local compositional gradients and affect how austenite nucleates [20]. Moreover, room-temperature quenching and partitioning (Q&P) treatments applied to medium-Mn steels have been reported to preserve the austenite fraction while improving its mechanical stability [21]. In parallel, nanostructured bainitic steels (nanobainite) have also been developed to maximize strength and toughness via extremely fine austenite/ferrite lamellae, where the control of C partitioning plays a similarly critical role [22, 23]. These steels—although differing in processing route—share with medium-Mn steels a reliance on the stabilization of austenite and kinetic tailoring of phase transformations. The last findings highlight that, although carbon content crucially determines mechanical stability, annealing temperature plays a more decisive role in the thermal stability of austenite than soaking time [24]. These insights underscore the importance of carefully tuning both chemical composition and processing parameters to achieve an optimal balance of austenite fraction and stability – an aspect central to the development of advanced high-strength steels with enhanced performance. The mentioned correlations between IA parameters and chemical composition of steels can be analysed using various computational tools based on the CALculation of PHAse Diagrams (CALPHAD) methodology [25]. Common software packages for metallic alloys include JMatPro ® [26], Thermo-Calc ® and DICTRA™ [27]. While JMatPro and Thermo-Calc primarily model phase transformations under equilibrium conditions and allow predicting the thermodynamic behaviour of materials across a wide range of temperatures, pressures, and chemical compositions. However, the kinetics of element diffusion is equally critical for accurate predictions during intercritical annealing. In this regard, advanced modules (e.g., DICTRA) or integrated approaches are required to account for time-dependent diffusion and transformation mechanisms. For example, Dykas et al. [28] used JMatPro and Thermo-Calc to model phase diagrams and continuous cooling transformation (CCT) diagrams of medium-Mn steels, investigating how alloying elements such as Mn and C influence phase transition kinetics in steels with 0.1–0.2 wt.% C and 2–10 wt.% Mn. Morawiec et al. [8] used JMatPro to examine the influence of Mn content on the bainite fraction. These studies align with research on nanobainitic steels, where precise control over isothermal transformation temperatures and carbon distribution is key to achieving the desired microstructure [29]. JMatPro is also a valuable tool for simulating theoretical CCT and TTT (Time-Temperature-Transformation) diagrams [28]. Recently, Elaraby et al. [30] demonstrated how a CALPHAD-based approach using Thermo-Calc and JMatPro can be employed to design medium-Mn steels with enhanced hydrogen embrittlement resistance, highlighting the potential of computational methods to optimize alloy composition for improved austenite stability and mechanical performance. In turn, Tian et al. [31] applied DICTRA to investigate nano-scale Mn and Ni concentration gradients from the interface to the interior of retained austenite lath. Likewise, Hu et al. [32] explored how C and Mn distributions evolve at the ferrite/austenite interface during IA. Overall, such thermodynamic modelling tools are highly useful for designing and optimizing processing routes in metallic alloys. However, theoretical predictions must ultimately be corroborated through appropriate experimental validations. Since the amount and stability of retained austenite play a crucial role in the mechanical properties of medium-Mn steels, and both the intercritical annealing process and manganese content significantly influence these aspects, a thorough investigation of their combined effects on austenite formation is essential. The aim of this study was to analyse the relationships between IA temperature, steel Mn content, and the resulting C, Mn, and Al contents in austenite using both computational and experimental approaches. To capture the equilibrium perspective, we used JMatPro and Thermo-Calc to model phase stability at various temperatures and compositions, while kinetic simulations (DICTRA) provided insight into non-equilibrium diffusion and transformation behaviour relevant to the short times typical of industrial processes. Dilatometry, energy-dispersive X-ray spectroscopy (EDS), and scanning electron microscopy (SEM) were then employed to verify the extent of these transformations in practice, bridging the gap between idealized equilibrium predictions and actual non-equilibrium conditions. 2. Material and experimental procedure 2.1. Materials The analysed materials were medium-Mn steels containing 3 to 5 wt.% Mn. The chemical compositions are listed in Table 1. A relatively low carbon content (0.16-0.17 wt.%) is beneficial for steel weldability. Because manganese is an austenite stabilizer, various fractions of retained austenite can be obtained in steels containing 3–5 wt.% Mn. In addition, aluminium and silicon prevent cementite precipitation [13]. Molybdenum was added to increase the steel’s hardenability and improve its hot ductility [33]. Table 1. Chemical compositions of analysed materials in wt.%. Steel C Mn Al Si Mo Nb 3Mn 0.17 3.1 1.6 0.22 0.22 0.04 4Mn 0.18 3.6 1.7 0.23 0.20 0.04 5Mn 0.16 5.0 1.6 0.20 0.20 0.04 The investigated steels were produced via induction melting. The ingots were then hot forged to a thickness of 22 mm and subsequently hot rolled in the temperature range from 1200–900°C, reducing the thickness to 9 mm. The final thermomechanical rolling consisted of three passes at deformation temperatures of 1050 °C, 950 °C and 850 °C, ultimately reaching a thickness of 4.5 mm. The flat samples were air-cooled to RT. Owing to the high hardenability provided by manganese, the resulting microstructures were fully martensitic [6, 7]. 2.2. Thermodynamic and kinetic modelling In order to capture both full equilibrium and non-equilibrium aspects of austenite formation, we used JMatPro ® (Genereal Steels Module, database version 13) [34] and Thermo-Calc ® (TCFE12, MOBFE7) [27] for equilibrium-based phase stability and element partitioning, while kinetic simulations of diffusion during intercritical annealing were performed using the DICTRA™ module within the Thermo-Calc ® software package. This approach accounts for the short soaking times typical of industrial IA processes, providing insight into time-dependent redistribution of elements under conditions far from full equilibrium. Although element partitioning during IA is computed here under equilibrium assumptions, these data offer a reference state that reflects the maximum extent of redistribution for carbon, manganese, and aluminium. The chemical compositions used in the simulations are listed in Table 1. In practice, the slower diffusion rate of manganese often means that full equilibrium is not achieved during short IA times. Nevertheless, equilibrium-based results can help identify trends—such as the limiting fraction of austenite and the potential distribution of elements at higher temperatures—which guide the subsequent kinetic analysis. The M s temperature for each composition and IA temperature was estimated via JMatPro’s internal models, which combine thermodynamic data with empirically derived equations for M s . This approach uses the computed phase compositions (notably the C and Mn content in austenite) to determine M s . Although these calculations assume a (near-)equilibrium austenite composition, they serve as a baseline for understanding which conditions favour the retention of austenite upon cooling. To assess how closely the simulated M s corresponds to real, non-equilibrium conditions, we compared the computed M s values with those derived from dilatometric measurements of the investigated steels. Given that full equilibrium is rarely reached in the short times typical of IA, DICTRA simulations provided additional insight into how incomplete manganese diffusion affects the time-dependent growth of austenite. By simulating diffusion profiles across ferrite/austenite interfaces, we captured deviations from the equilibrium composition predicted by JMatPro and Thermo-Calc. This combined approach—equilibrium-based phase stability plus kinetic analysis—helps bridge the gap between idealized predictions and the actual partitioning behaviour. 2.3. Dilatometric investigations Cylindrical specimens, measuring 10 mm in length and 4 mm in diameter, were machined from the hot-rolled material with their long axes parallel to the rolling direction. The heat treatment was carried out using a high-resolution Bähr DIL805A/D dilatometer, where temperature was measured using a S-type thermocouple welded to the specimen’s centre. All experiments were conducted under vacuum, and helium served as the coolant. The dilatometric data were analysed in accordance with ASTM A1033-04 [35]. To ensure a uniform, fully martensitic starting microstructure—favourable for the austenite reversion transformation (ART) during IA [6]—the specimens were first austenitized (at 1100 °C for 5 min) and subsequently quenched to RT at approximately 60 °C/s. This rapid cooling rate was chosen to minimize any ferritic or bainitic transformations [8]. Next, the samples were heated at 3°C/s to selected IA temperatures between 700°C and 760°C and held for 60 min. The heating rate of 3 °C/s was chosen in accordance with standard dilatometric practice [6-8], while the temperature range (700–760 °C) aligns with the intercritical region reported as suitable for medium-Mn steels [7]. A 60 min soaking time was adopted to allow partial diffusion of Mn and C into austenite [6]. Finally, the specimens were cooled to RT at 60 °C/s, which is both feasible in the dilatometer setup and sufficient to suppress ferritic or bainitic transformations upon cooling [8]. The schematic illustration of the entire heat treatment cycle applied to the investigated steels is presented in Fig. 1. In order to determine the critical A c 1 and A c 3 temperatures, the samples were subjected to a specialized time–temperature cycle. First, they were heated from room temperature to 1100 °C at a rate of 2.5 °C/min. After reaching 1100 °C, the specimens were rapidly cooled down to room temperature at 10 °C/s. This procedure allowed precise identification of the A c 1 and A c 3 points based on the recorded temperature and dilatation curves, in accordance with [35]. 2.4. Microstructural investigations A standard metallographic preparation procedure [36] was applied to samples prior to microscopic observation. The specimens were cut into half perpendicularly to their length, then mechanically ground using SiC paper up to 2000 grit, polished with a diamond paste (up to 1 μm), and etched in 4% nital for 5 seconds to reveal microstructural detailes. The microstructure was examined using a Zeiss Supra 25 scanning electron microscope (SEM) in secondary electron (SE) mode. Energy-dispersive X-ray spectroscopy (EDS) was used to measure changes in Mn and Al concentrations in RA laths. The calibration was conducted using a molybdenum (Mo) standard with a purity of 99.99%, ensuring high precision and stability in energy measurements essential for accurate chemical composition analysis. The energy resolution of the detector was set at 137 eV, enabling clear distinction of spectral lines and providing detailed results with minimal noise interference. The acquisition time was configured to 6400 ns, which allows for optimal signal quality and accurate spectral intensity representation while maintaining a good signal-to-noise ratio. These parameters were selected to ensure reliable and precise results, tailored to the characteristics of the point analyses conducted. 3. Results and discussion 3.1. Equilibrium and kinetic analysis of reversed austenite formation Thermodynamic and kinetic simulations were performed to investigate the influence of IA temperature and Mn content on austenite fraction and elemental partitioning in medium-Mn steels. The calculations focused on phase stability and diffusion behaviour relevant to industrial annealing conditions. Figures 2.a–2.c present the theoretical fractions of austenite (FCC), ferrite (BCC), and cementite (CEM) as a function of intercritical annealing temperature, predicted for 3MnNb, 4MnNb, and 5MnNb steels, respectively. The simulations were carried out under the assumption of full thermodynamic equilibrium. Two thermodynamic software packages—JMatPro and Thermo-Calc—were used to enable a comparative assessment of phase stability in the intercritical region. In all compositions, an increase in the IA temperature led to a higher fraction of austenite, consistent with the expected transformation behaviour in medium-Mn steels. Additionally, a higher nominal manganese content resulted in a greater amount of austenite at the same temperature due to Mn’s strong austenite-stabilizing effect. At 700 °C, Thermo-Calc simulations predicted the austenite (FCC) phase fraction to be approximately 24.5 % for the 3MnNb steel, 28.1 % for the 4MnNb steel, and 37.9 % for the 5MnNb steel, indicating a clear increase in austenite stability with increasing Mn content. This trend was also observed at higher temperatures. For example, at 760 °C, the predicted FCC phase fractions rose to 35.5 % for 3MnNb, 40.3 % for 4MnNb, and 54.2 % for 5MnNb. These results confirm the strong stabilizing effect of manganese on the austenite phase during intercritical annealing and highlight the sensitivity of austenite fraction to both chemical composition and annealing temperature in medium-Mn steels. JMatPro simulations yielded slightly lower austenite fractions—typically by 2–6 % across the analysed temperature range. Notably, this discrepancy became more pronounced with increasing Mn content and temperature. The observed differences likely result from variations in the thermodynamic databases and interaction parameters used by each software package. The equilibrium phase diagrams (Fe-Fe 3 C system) for the investigated steels, along with the corresponding austenite transformation start ( A e 1 ) and finish ( A e 3 ) temperatures, were calculated using Thermo-Calc ® and are illustrated in Figures 2.d–2.f. These diagrams provide complementary insight into the thermodynamic stability of BCC, FCC, and CEM phases within the intercritical temperature range, supporting the interpretation of austenite formation trends under equilibrium conditions. With increasing Mn content, both the equilibrium austenite formation start temperature and finish temperature shift to lower values, confirming the strong austenite-stabilizing effect of manganese. Specifically, the A e 3 temperatures are 940 °C, 924 °C, and 867 °C for 3MnNb, 4MnNb, and 5MnNb steels, respectively. The corresponding A e 1 temperatures are 654 °C, 631 °C, and 469 °C for the 3MnNb, 4MnNb, and 5MnNb steels, respectively. These temperatures mark the onset of austenite nucleation at the expense of cementite, that is, the beginning of cementite dissolution [6, 37]. It should be noted, however, that in the investigated steels with an initial martensitic microstructure, cementite dissolution does not occur due to the absence of pre-existing due to the absence of pre-existing CEM, as will be discussed in a later section in the structure, as will be discussed in a later section. Moreover, a progressive expansion of the three-phase (BCC + FCC + CEM) coexistence region is observed with increasing Mn concentration, indicating an extended temperature range in which ferrite, austenite, and cementite are thermodynamically stable. One can observe in equilibrium phase diagrams that in the lower part of the intercritical range (e.g., below 660 °C for 5MnNb and 680 °C for 3MnNb and 4MnNb), a sharp increase in the carbon concentration in FCC was observed, reflecting the progressive dissolution of cementite into the growing austenite. This process becomes more gradual at higher Mn contents, where the difference between minimum and maximum carbon concentrations within FCC is more pronounced. These findings are consistent with the literature data for Mn-rich steels [6,13,37], which report that manganese decreases the chemical potential of carbon in cementite, thereby moderating the driving force for carbon partitioning into austenite. As a result, austenite enrichment in carbon was significantly lower in the 5MnNb steel compared to the 3MnNb and 4MnNb grades, especially at elevated IA temperatures. Since the carbon content in austenite strongly influences the M s temperature, the lower C enrichment in 5MnNb steel suggests that its retained austenite may exhibit reduced thermal stability relative to steels with lower Mn contents. In addition to the equilibrium phase boundaries, Figures 2.d–2.f also present the results of DICTRA simulations, showing the average carbon concentration in retained austenite laths predicted under non-equilibrium diffusion conditions at selected intercritical annealing temperatures (700 °C, 720 °C, and 760 °C) after 60 minutes of isothermal holding. These values are superimposed onto the equilibrium phase diagrams for direct comparison. The horizontal error bars associated with these data points do not represent statistical uncertainty but instead reflect the minimum and maximum carbon concentrations across the austenite lath obtained from DICTRA simulations. This variation is a consequence of carbon segregation driven by interface-controlled diffusion kinetics during intercritical annealing. The significance of this carbon heterogeneity within retained austenite and its correlation with local phase stability will be discussed in detail in a subsequent section. The equilibrium behaviour of manganese in the Fe–Mn system was analysed for the 3MnNb, 4MnNb, and 5MnNb steels using Thermo-Calc ® , as illustrated by the phase diagrams in Figures 3.a–3.c. These diagrams provide a comprehensive view of the phase equilibria, including the phase boundaries between ferrite, austenite, and cementite, as well as the temperature ranges in which these phases can coexist. As the temperature increases, the Mn concentration in the FCC phase decreases, particularly above 700 °C, primarily because of reduced manganese partitioning from ferrite under equilibrium conditions. Steels with higher nominal Mn content, especially 5MnNb, exhibit greater enrichment of manganese in the FCC phase across the intercritical region. These results are consistent with previous findings by Liu et al. [38], who reported that higher annealing temperatures reduce the Mn and C content in austenite, thereby lowering its thermal stability and promoting chemical inhomogeneity within austenite laths. Furthermore, the three-phase region (BCC + FCC + CEM) is most extensive in the 3MnNb steel and progressively narrows as the Mn content increases, ultimately giving way to a two-phase (BCC + CEM) field in the 5MnNb alloy. This shift in phase stability is a direct result of changes in chemical potential and the distribution of substitutional elements. A similar analysis was performed for aluminium in the Fe–Al system, with the corresponding equilibrium phase diagrams shown in Figures 3.d–3.f. These diagrams illustrate not only the solubility limits of Al in austenite, but also the phase boundaries for ferrite and cementite, as well as the two-phase coexistence regions. As the temperature increases, the aluminium content in the FCC phase rises gradually. Below 900 °C, this increase is nearly linear, and the differences among the steel grades are relatively small. However, above 900 °C, a sharp, non-linear rise in Al solubility in austenite is observed, and the differences in FCC solubility boundary Al content among the alloys become more pronounced. Notably, the extent of austenite enrichment in Al remains significantly lower than that observed for carbon (Fig. 3a) and manganese (Fig. 3.b), reflecting both the strong ferrite-stabilizing effect of aluminium and its limited solubility in the FCC phase under equilibrium conditions. To complement the equilibrium analysis, DICTRA simulations were performed to evaluate the non-equilibrium partitioning of manganese and aluminium across the ferrite/austenite interface. The Fe–Mn and Fe–Al diagrams shown in Figures 3.a–3.f include these simulation results, displaying the average concentrations of Mn and Al in retained austenite (FCC) after 60 minutes of isothermal holding at selected intercritical annealing temperatures (700 °C, 720 °C, and 760 °C). These values are superimposed on the equilibrium phase diagrams to enable direct comparison between thermodynamic predictions and kinetically influenced elemental distributions. The horizontal error bars associated with the DICTRA data points do not reflect statistical variation but rather represent the range of local concentrations (minimum–maximum) of Mn and Al within the FCC phase, derived from the simulated concentration profiles. The effects of local Mn and Al concentration gradients in retained austenite—particularly their influence on phase stability and the initiation of martensitic transformation—will be further examined in the following sections. To analyze the kinetics of austenite formation and the concurrent partitioning of the main alloying elements (C, Mn, Al) during intercritical annealing (IA), one-dimensional diffusion simulations were performed using the DICTRA module of Thermo-Calc ® . The geometry of the simulation domains was defined based on quantitative measurements of the martensite and austenite laths observed in the microstructure after IA at 700 °C for 60 minutes, the results of which are shown in Fig. 4.a. The IA temperature at 700 °C was selected as a reference point because it lies near the lower end of the intercritical range, where diffusion distances and kinetic limitations are expected to exert the strongest influence on partitioning behaviour. The width of martensite (w α′ ) and austenite (w γ ) laths was measured from high-resolution SEM micrographs, and the mean distance from center to center was used to estimate the effective diffusion path. In the simulations, this spacing was taken to represent the martensite-side domain thickness, as indicated schematically in Fig. 4.b. The initial microstructure, consisting of low-carbon martensite (α′), was treated as ferrite (α) due to its minimal tetragonality after IA, which renders it effectively indistinguishable from ferrite in X-ray diffraction analysis [1]. Specifically, three diffusion domains were defined in DICTRA for the 3MnNb, 4MnNb, and 5MnNb steels, with total sizes of 325 nm, 260 nm, and 455 nm, respectively. In each case, the model consisted of a supersaturated ferrite (α′) region and a 1 nm-thick austenite (γ) nucleus positioned at the left boundary to initiate phase reversion. A schematic of the simulation geometry and a high-magnification SEM micrograph used for dimensional calibration are shown in Fig. 4.b. Figures 5.a–5.i illustrate the simulated evolution of reversed austenite during the ART process, along with the corresponding concentration profiles of carbon, manganese, and aluminium in the 3MnNb, 4MnNb, and 5MnNb steels after 60 minutes of isothermal annealing at selected intercritical temperatures: 700 °C, 720 °C, and 760 °C. For reference, the early-stage elemental distributions after 1 second of simulation are also shown to highlight the initial partitioning behaviour and the steep concentration gradients developing at the α/γ interface in the initial phase of the IA process. Carbon, due to its high diffusivity in both phases, exhibited rapid enrichment in the growing austenite lath regardless of steel composition. The simulated profiles showed steep concentration gradients across the α′/γ interface at lower temperatures, which gradually became more uniform as IA temperature increased. At 700 °C, sharp C peaks were observed within the γ phase, especially for 3MnNb and 4MnNb, with maximum values reaching ~0.56 wt.% for 3MnNb. These values decreased with increasing temperature due to accelerated homogenization and volumetric expansion of the FCC phase, leading to lower average and maximum C concentrations. Despite these differences, all three alloys showed an increasing γ width and transformed volume fraction with temperature, indicating that even with lower C concentrations, austenite was stabilized by cumulative contributions from other elements and longer exposure time. Notably, carbon partitioning was most efficient in 3MnNb, where the diffusion distance was shortest, leading to higher localized enrichment near the interface. Unlike carbon, manganese exhibited sluggish diffusion kinetics, resulting in strong chemical gradients across the α′/γ interface, particularly at lower IA temperatures. The simulations revealed that Mn redistribution was incomplete even after 60 minutes at 700 °C, especially in 5MnNb, where diffusion paths were longest due to the coarser initial martensitic structure. At 700 °C, the Mn concentration ranged from ~5.3 wt.% (min) to ~9.1 wt.% (max) in the γ phase of 5MnNb, indicating pronounced inhomogeneity. With increasing temperature, Mn partitioning became more efficient, yet a strong discrepancy between max and min Mn values persisted in all alloys. The 3MnNb steel showed relatively narrower segregation ranges, likely due to both shorter diffusion distances and lower nominal Mn content. These results highlight the critical role of Mn kinetics in determining the local austenite composition and suggest that Mn inhomogeneity can induce local variations in M s temperature and mechanical stability of retained austenite. Aluminium, although generally considered a slow diffuser in BCC and FCC iron, showed measurable partitioning behaviour in the DICTRA simulations. Similar to Mn, Al exhibited asymmetric profiles across the α′/γ boundary, with higher concentrations near the ferritic side and lower solubility within the austenite. At lower temperatures (700 °C), a steep Al gradient was observed within the γ phase, particularly in 3MnNb, where the max and min values differed by over 0.44 wt.%.The segregation range (max–min Al) decreased with increasing IA temperature, confirming that higher temperatures facilitated more uniform redistribution, albeit still kinetically limited. Among the three steels, 5MnNb exhibited the highest retained austenite fraction and largest lath width, yet its Al content remained lower within the FCC phase, especially at 700 °C. This effect is attributed to the low solubility of Al in austenite and its tendency to accumulate in ferrite, where it contributes to stabilizing the BCC phase and potentially increases M s locally. Furthermore, the average concentrations of C, Mn, and Al in retained austenite, calculated from the DICTRA simulations and superimposed on the equilibrium phase diagrams (Figures 2 and 3), demonstrate that the partitioning behaviour deviates significantly from equilibrium thermodynamic predictions. Not only do the mean compositions fall well below the respective solubility limits defined by the equilibrium boundaries, but even the maximum local concentrations at the ferrite/austenite interface remain far from the equilibrium tie-lines. This disparity highlights the non-equilibrium nature of the intercritical annealing process, especially under industrially relevant timescales, and underscores the critical role of kinetic constraints in shaping the compositional gradients and stability of austenite in medium-Mn steels. The DICTRA simulations confirm that all three main alloying elements (C, Mn, and Al) in the investigated steels undergo complex, temperature-dependent redistribution during intercritical annealing, collectively governing the growth and stability of austenite. While carbon partitions rapidly and uniformly into the FCC phase, Mn and Al exhibit pronounced concentration gradients due to their lower diffusivity. These gradients arise from the limited mobility of substitutional elements and reflect the non-uniform, kinetically controlled redistribution under non-equilibrium conditions [19]. Manganese, in particular, shows stronger segregation compared to aluminium, consistent with its lower diffusion coefficient and stronger interaction with carbon [13]. The extent of elemental partitioning is further influenced by initial microstructural parameters (domain size), annealing temperature, and nominal alloy composition. Importantly, the simulations suggest that local heterogeneities in Mn and Al concentration within retained austenite may introduce significant variations in local stability (i.e., variations in M s temperature), which could manifest as sequential martensitic transformation upon cooling [38]. To further validate the predicted partitioning behaviour of substitutional elements during IA, the manganese and aluminium concentrations in RA laths were experimentally measured using EDS and compared with the average values obtained from DICTRA simulations. Figure 6 presents this comparison for the 3MnNb, 4MnNb, and 5MnNb steels after 60 minutes of annealing at 700 °C. The EDS measurements, performed on individual austenite laths identified in SEM micrographs, provide insight into the local chemical composition resulting from element partitioning during IA. In the simulations, it was assumed that the reverted austenite remained thermally stable and did not transform into martensite upon cooling, thereby persisting as retained austenite. This assumption justifies the direct comparison between DICTRA-predicted average concentrations in the austenite phase and experimentally measured EDS data. The results for manganese, summarized in Figure 6.a, show a good overall agreement between DICTRA simulations and EDS measurements. For the 3Mn and 4Mn steels, the simulated Mn contents in RA (4.69 wt.% and 5.37 wt.%, respectively) are slightly higher than the corresponding EDS values (4.55 wt.% and 4.78 wt.%), yet the differences fall within the experimental and simulation standard deviations. For the 5Mn steel, the EDS data indicate a higher average Mn content (7.02 wt.%) compared to the simulation (6.15 wt.%), although both values exhibit relatively large standard deviations (1.22 and 1.03, respectively). This suggests a greater compositional heterogeneity in the RA laths of the 5Mn steel, consistent with prior reports indicating that increased nominal Mn content enhances segregation effects during austenite reversion [19,21,]. Xu et al. [39] also observed that austenite nucleating at metastable carbides tends to exhibit strong Mn concentration gradients, affecting local phase stability. The convergence between simulated and experimental mean values, alongside comparable deviations, demonstrates the ability of DICTRA to reasonably capture Mn partitioning trends in medium-Mn steels subjected to IA, especially under short annealing durations representative of industrial conditions. A similar comparison was carried out for aluminium, as shown in Figure 6.b. The DICTRA-predicted aluminium contents in the austenite phase after IA at 700 °C for 60 minutes show slightly lower values than those measured experimentally by EDS, but the trends across the three steel compositions remain consistent. For the 3Mn, 4Mn, and 5Mn steels, the simulated Al concentrations in RA were 1.43 wt.%, 1.52 wt.%, and 1.48 wt.%, respectively, while the EDS measurements yielded slightly higher averages of 1.73 wt.%, 1.62 wt.%, and 1.72 wt.%. The differences are within a reasonable range (≤0.25 wt.%) and can be partially attributed to the limited spatial resolution and averaging effects inherent to EDS analysis at sub-micrometer scales, as well as assumptions in the DICTRA model regarding the initial interface geometry and mobility databases. Moreover, the standard deviations from both DICTRA and EDS are relatively low, indicating that aluminium partitioning into the austenite phase is more uniform compared to manganese. This is in line with the higher diffusivity of Al and its weaker thermodynamic interaction with carbon compared to Mn, which results in smoother concentration gradients and more homogeneous elemental distribution, as previously observed in experimental and modelling studies of medium-Mn steels [19]. Taken together, the combined comparison of experimental and simulated results for both Mn and Al confirms the reliability of DICTRA simulations in capturing the essential features of element partitioning during ART, supporting their use as a predictive tool for designing microstructures in medium-Mn steels. 3.2. Determination of characteristic temperatures Dilatometric experiments were performed to determine the key phase transformation temperatures (A c 1 , A c 3 , and M s ) for the investigated medium-Mn steels. It is important to note that each IA cycle was initiated from a fully martensitic microstructure, which ensured that any transformation observed during reheating was associated with austenite reversion or new phase formation, rather than the dissolution of pre-existing phases such as pearlite or cementite. Figure 7 presents the dilatometric results for the 3MnNb, 4MnNb, and 5MnNb steels, illustrating the methodology used to determine critical transformation temperatures. In Figures 7.a–7.c, the dilatation curves—presented in terms of relative change of length (RCL)—and their first derivatives are shown, enabling the identification of the A c 1 and A c 3 points based on the inflection and slope changes in thermal expansion behaviour. These characteristic temperatures were evaluated from heating curves, where the tangents were applied to the transition regions between the linear segments to define the onset and completion of austenite formation. It should be emphasized that the interpretation of A c 1 in these medium-Mn steels with an initial martensitic structure poses certain challenges. Unlike traditional ferrite-pearlite microstructures, where A c 1 corresponds to the start of cementite dissolution into austenite, the martensitic structure promotes the precipitation of fine cementite particles at sub-A c 1 temperatures during reheating [6]. This precipitation process introduces dilatometric artifacts that interfere with the typical linear thermal expansion trend, complicating the accurate determination of A c 1 from dilatation curves. The critical A c 1 temperatures were determined to be 626 °C for 3MnNb, 656 °C for 4MnNb, and 631 °C for 5MnNb. The corresponding A c 3 temperatures were 931 °C, 956 °C, and 934 °C, respectively. These values vary systematically with the manganese content, consistent with the well-known austenite-stabilizing effect of Mn, which promotes γ-phase formation at lower temperatures. A lower A c 1 temperature also broadens the intercritical processing window, which is advantageous for industrial control of IA parameters. Figure 7.d presents the cooling segments of the dilatometric curves used to determine the M s temperatures for each steel composition. These measurements were conducted on samples austenitized at 1100 °C for 5 min and subsequently quenched to room temperature at approximately 60 °C/s to produce a fully martensitic microstructure. The M s temperatures were identified as the points corresponding to the onset of rapid contraction, associated with the γ → α′ martensitic transformation. The measured M s values were 458 °C for 3MnNb, 436 °C for 4MnNb, and 354 °C for 5MnNb. This trend reflects the increasing thermal stability of austenite with higher Mn content, consistent with the austenite-stabilizing effect of manganese. Figure 8 presents dilatometric curves recorded during cooling of the investigated steels (3MnNb, 4MnNb, and 5MnNb) after intercritical annealing for 60 minutes at different temperatures. In Figure 8.a, which corresponds to IA at 700 °C, no transformation-related deviations were observed. The cooling curves are nearly linear, without inflection points typically associated with phase transformations, indicating that the microstructures formed during IA remained stable down to room temperature. This suggests that the applied IA parameters were sufficient to ensure effective partitioning of carbon and manganese from ferrite to austenite, stabilizing the retained austenite and suppressing the γ → α′ martensitic transformation [40, 41]. By contrast, Figure 8.b presents the dilatometric curves obtained during cooling of the same steels after intercritical annealing at 720 °C and 760 °C. In these cases, deviations from the linear contraction trend were observed, corresponding to the onset of martensitic transformation. This transformation was clearly detected in the 5Mn steel annealed at both 720 °C and 760 °C, and in the 3Mn steel after annealing at 760 °C. The transformation onset was identified by a characteristic elongation of the sample, reflecting the volume expansion associated with the γ → α′ transformation. Interestingly, despite its higher nominal Mn content, the 5Mn steel exhibited a more pronounced martensitic transformation after IA at 760 °C compared to the 3Mn steel. This behavior can be attributed to the increased volume fraction of austenite formed at elevated IA temperatures, which was not sufficiently enriched in stabilizing elements such as C and Mn. As a result, the austenite remained metastable and transformed partially into martensite during cooling. These findings are consistent with previous studies by Skowronek et al. [6], who demonstrated that intercritical annealing within the range of 680–700 °C is sufficient to stabilize austenite in 5 wt.% Mn steels, thereby suppressing martensitic transformation upon cooling to room temperature. Figure 9 illustrates the variation of M s temperature as a function of IA temperature for the investigated steels. Contrary to the initial assumption, the M s temperature increases with increasing IA temperature. This trend reflects a reduction in the chemical stability of austenite formed at higher annealing temperatures, which is attributed to a lower degree of enrichment in carbon and manganese due to faster diffusion dynamics and a higher volume fraction of austenite to be stabilized. The observed behaviour is consistent with the findings of Liu et al. [38] and Zhang et al. [5], who reported that the thermal stability of reversed austenite decreases at elevated IA temperatures due to dilution effects and heterogeneous partitioning. Notably, the 5Mn steel exhibits lower M s temperatures compared to 3Mn and 4Mn steels across the entire IA temperature range, which confirms the strong stabilizing effect of manganese on austenite. This is further supported by prior thermodynamic analyses indicating that Mn reduces the chemical potential of carbon in austenite and cementite [31, 37], thereby promoting austenite stability. According to theoretical calculations, the IA temperature required to suppress M s below room temperature was approximately 680 °C for 5Mn steel, and close to 700 °C for 3Mn and 4Mn steels. 3.3. Microstructure observations after intercritical annealing The microstructures of the investigated steels after intercritical annealing at 700 °C for 60 minutes are shown in Figure 10. They were primarily composed of lath-like ferrite and retained austenite (RA), with the morphology and relative phase fractions varying depending on the manganese content. Note that, in the present study, the lath regions denoted as ferrite correspond to regions that were originally fully martensitic prior to IA treatment. Following the 60-minute annealing, significant carbon partitioning into austenite occurred, leading to a marked reduction in tetragonality and rendering the former martensite structurally and chemically indistinguishable from ferrite. This interpretation is consistent with the assumption adopted in the DICTRA simulations, where the low-carbon martensite was modelled as ferrite due to its negligible tetragonality after partitioning. While a detailed quantitative phase analysis was beyond the scope of the present work, qualitative evaluation of the SEM images suggests that the steel containing 3 wt.% Mn exhibits a higher proportion of ferrite (Fig. 10.a), whereas the 5Mn steel reveals a more extensive presence of retained austenite. This trend aligns with the thermodynamic predictions shown in Fig. 2, where the increased Mn content promotes austenite stability and fraction under intercritical annealing conditions. Moreover, a visual comparison indicates that the thickness of RA laths increases with higher Mn content; this observation is in qualitative agreement with DICTRA simulation results (Fig. 5). However, it should be noted that no direct measurements of RA thickness were performed, and a more detailed quantitative analysis (e.g., via image analysis or EBSD) would be necessary to validate this tendency. The presence of cementite was also confirmed, particularly in the 3Mn and 4Mn steels (Fig. 10.b), consistent with previous studies. Mueller et al. [37] reported similar findings in Fe–0.19C–4.39Mn steel, where manganese partitioning from ferrite to cementite reduced the chemical potential of carbon in cementite. This, in turn, decreased the driving force for carbon to diffuse toward austenite, thereby reducing the extent of carbon enrichment in RA and stabilizing cementite during intercritical annealing. Conclusion The effect of Mn content in a range of 3-5 wt.% on the phase transformation kinetics during intercritical annealing of multiphase steels was analysed using thermodynamic and kinetic modelling and experimental approaches. The main findings of the study are: An increase in the Mn content in steel leads to a higher equilibrium and simulated fraction of austenite formed during intercritical annealing. Modelling at 700 °C predicted ~24 % of FCC in 3MnNb steel and ~38% in 5MnNb steel. This trend was confirmed by SEM observations and is consistent with the austenite-stabilizing effect of Mn. A higher Mn content in steel reduces the carbon content in austenite, thereby increasing its martensite start temperature and reducing thermal stability. This behaviour was evident from both DICTRA simulations and experimental M s determinations via dilatometry. The IA temperature significantly affects both austenite fraction and chemical composition. At higher annealing temperatures, the increased amount of austenite leads to a dilution effect, resulting in lower concentrations of C and Mn per unit volume of austenite and thus a higher M s temperature. The microstructures of all investigated steels were primarily composed of lath-like ferrite and retained austenite. In the 3MnNb steel, only ferrite and RA were identified, whereas the 4MnNb steel additionally exhibited the presence of fine carbides. In contrast, the 5MnNb steel revealed small MA constituents, indicating increased heterogeneity of the microstructure with higher Mn content. Qualitative observations suggest that the thickness of RA laths increases with increasing Mn content, which is consistent with trends predicted by DICTRA simulations. DICTRA simulations revealed strong concentration gradients of Mn and Al across the austenite/ferrite interface, particularly at lower IA temperatures. These non-equilibrium conditions were confirmed by some discrepancy between simulated and equilibrium phase boundaries. EDS measurements confirmed the general trends predicted by DICTRA. In particular, the standard deviation of Mn content within retained austenite laths increased with bulk Mn concentration, reflecting incomplete homogenization due to limited diffusion during intercritical annealing. Al partitioning was more uniform, but still followed the modelled trend. Declarations Funding: This research was funded in whole or in part by National Science Center through the grant no. 2021/41/N/ST8/03371. For the purpose of Open Access, the author has applied a CC-BY public copyright license to any Author Accepted Manuscript (AAM) version arising from this submission. Data Availability Statement: Not applicable. Conflicts of Interest: The authors declare no conflict of interest. References Chandan AK, Bansal GK, Kundu J, Chakraborty J, Chowdhury SG. Effect of prior austenite grain size on the evolution of microstructure and mechanical properties of an intercritically annealed medium manganese steel. Sci. Eng. A 2019;768:138458, doi:10.1016/j.msea.2019.138458. Tjahjanto DD, Turteltaub S, Suiker ASJ. Crystallographically based model for transformation-induced plasticity in multiphase carbon steels. Continuum Mech Thermodyn 2008;19:399-422, doi:10.1007/s00161-007-0061-x. Kaczmarczyk J, Grajcar A, Kozłowska A, Cruz Banuelos JS, Opara J, Tatke N. Modelling of strain-induced martensite formation in advanced medium-Mn automotive sheet steel. Continuum Mech Thermodyn 2023;35:1841-1858, doi:10.1007/s00161-023-01202-2. Kang S, Hofer RW, Speer JG, Krizan D, Matlock DK, De Moor E. Intercritical annealing response of medium manganese steels having different carbon concentrations. Steel Res. Int. 2018;89:1800069, doi:10.1002/srin.201800069. Zhang BG, Zhang XM, Liu HT. Microstructural evolution and mechanical properties of Ni-containing light-weight medium-Mn TRIP steel processed by intercritical annealing. Mater. Sci. Eng. A 2020;793:139289, doi:10.1016/j.msea.2020.139289. Skowronek A, Grajcar A, Kozłowska A, Janik A, Morawiec M, Petrov RH. Temperature-dependent microstructural evolution of Al-rich medium-Mn steel during intercritical annealing. Metall. Mater. Trans. A 2022;53A:3012–3021, doi:10.1007/s11661-022-06721-2. Skowronek A, Grajcar A, Garcia-Mateo C, Jiménez JA Petrov RH. Time-dependent evolution of volume fraction and stability of retained austenite in a hot-rolled and intercritically annealed Al-alloyed medium-Mn steel. Metall. Mater. Trans. A 2023;54:1443–1454, doi:10.1007/s11661-023-06998-x. Morawiec M, Opara J, Garcia-Mateo C, Jimenez JA, Grajcar A. Effect of Mn on the chemical driving force and bainite transformation kinetics in medium-manganese alloys. Therm. Anal. Calorim. 2023;148:1567-1576, doi.org/10.1007/s10973-022-11664-2. Bhowmik N, Ghosh S, Mandal S, Haldar A, Chattopadhyay PP. Genesis of plasticity-induced serrated metal flow in medium-Mn steel. Calphad 2022;77:102425, doi:10.1016/j.calphad.2022.102425. Kozłowska A, Grzegorczyk B, Morawiec M, Grajcar A. Explanation of the PLC effect in advanced high-strength medium-Mn steels. A review. Materials 2019;12:4175, doi: 10.3390/ma12244175. Morawiec M, Grajcar A, Zalecki W, Garcia-Mateo C, Opiela M. Dilatometric study of phase transformations in 5Mn steel subjected to different heat treatments. Materials 2020;13:958, doi:10.3390/ma13040958. Kwok TWJ, Dye D. A review of the processing, microstructure and property relationships in medium Mn steels. Mater. Rev. 2023;68:1098-1134, doi:10.1080/09506608.2023.2199617. Pereloma EV, Gazder AA, Timokhina I. Retained austenite: Transformation-Induced Plasticity. Encyclopedia of Iron, Steel, and Their Alloys. Taylor and Francis, New York 2016, 3088–3103. Kaar S, Krizan D, Schwabe J, Hofmann H, Hebesberger T, Commenda C, Samek L. Influence of the Al and Mn content on the structure-property relationship in density reduced TRIP-assisted sheet steels. Mater. Sci. Eng. A 2018;735:475-486, doi:10.1016/j.msea.2018.08.066. Hidalgo J, Celada-Casero C, Santofimia MJ. Fracture mechanisms and microstructure in a medium Mn quenching and partitioning steel exhibiting macrosegregation. Mater. Sci. Eng. A 2019;754:766-777, doi:10.1016/j.msea.2019.03.055. Kalsar R, Sanamar S, Schell N, Brokmeier HG, Saha R, Ghosh P, Bhagat AN, Suwas S. Elemental partitioning in medium Mn steel during short-time annealing: An in-situ study using synchrotron X-rays. Materialia, 2020;9:100594, doi:10.1016/j.mtla.2020.100594. Kang S, Speer JG, Krizan D, Matlock DK, De Moor E. Prediction of tensile properties of intercritically annealed Al-containing 0.19C–4.5Mn (wt%) TRIP steels. Mater. Des. 2016;97:138–146, doi:10.1016/j.matdes.2016.02.058. Srinivas R, Varanasi BG, Ponge D. Effect of Nb micro-alloying on austenite nucleation and growth in a medium manganese steel during intercritical annealing. Acta Mater. 2022;229:117786, doi:10.1016/j.actamat.2022.117786. Varanasi RS, Lipińska-Chwałek M, Mayer J, Gault B, Ponge D. Mechanisms of austenite growth during intercritical annealing in medium manganese steels. Scripta Mater. 2022;206:114228. doi:10.1016/j.scriptamat.2021.114228. Mueller, J., Matlock, D., Speer, J., & De Moor, E. (2023). Effect of Mn-Enriched Cementite on Austenite Formation During Intercritical Annealing. International Symposium on New Developments in Advanced High-Strength Sheet Steels, AHSS 2023 , 215-223. http://doi.org/10.33313/298/023. He BB, Pan S, Huang MX. Extra work hardening in room-temperature quenching and partitioning medium Mn steel enabled by intercritical annealing. Mater. Sci. Eng. A 2020;797:140106, doi:10.1016/j.msea.2020.140106. K.D.H. Bhadeshia. Nanostructured bainite. Proceedings of the Royal Society A: Mathematical, 466 (2010): 3–18. doi:10.1098/rspa.2009.0407. Caballero FG, Miller MK, Garcia-Mateo C. Tracking solute atoms during bainite reaction in a nanocrystalline steel. Mater Sci Technol (2010) 26 889-898, doi:10.1179/026708310X12635619987943. Zhou TP, Wang CY, Wang C, Cao WQ, Chen ZJ. Austenite stability and deformation-induced transformation mechanism in cold-rolled medium-Mn steel. Mater. Sci. Eng. A 2020;798:140147, doi:10.1016/j.msea.2020.140147. Lukas HL, Fries SG, Sundman B. Computational Thermodynamics: The Calphad Method. Cambridge: Cambridge University Press; 2007. Saunders N, Guo Z, Li X, Miodownik AP, Schillé J-Ph. Using JMatPro to model materials properties and behavior. JOM 2003;55:60-65, doi:10.1007/s11837-003-0013-2. J-O Andersson, Thomas Helander, Lars Höglund, Pingfang Shi, Bo Sundman, Thermo-Calc & DICTRA, computational tools for materials science, CALPHAD 26 (2002) 273-312, https://doi.org/10.1016/S0364-5916(02)00037-8. Dykas J, Samek L, Grajcar A, Kozłowska A. Modelling of phase diagrams and continuous cooling transformation diagrams of medium manganese steels. Symmetry 2023;15:381, doi:10.3390/sym15020381. Morawiec M, Opara J, Grajcar A. Nanobainite formation in high-Al medium-Mn steels: thermodynamic approach. J Therm Anal Calorim (2024) 10511-10520, doi:10.1007/s10973-024-13441-9. Elaraby M, Ali M, Eissa M, Kömi J, Karjalainen P, Javaheri V. Computational designing approach for medium manganese steels with potential better hydrogen embrittlement resistance. In: Proceedings of SIMS EUROSIM (2024) 228-235. doi:10.3384/ecp212.032. Tian C, Guo H, Hu B, Enomoto M, Shang C. Influence of nano-scale concentration gradient of alloying elements on the ductility in an intercritically annealed and tempered medium Mn steel. Mater. Sci. Eng. A 2021;810:141009, doi:10.1016/j.msea.2021.141009. .Hu B, Zheng Q, Lu Y, Ji C, Liang T, Zheng C, Li D. Stabilizing austenite via intercritical Mn partitioning in a medium Mn steel. Scripta Mater. 2023;225:115162, doi:10.1016/j.scriptamat.2022.115162. Gramlich A, Lange R, Zitz U, Büßenschütt K. Air-hardening die-forged con-rods—achievable mechanical properties of bainitic and martensitic concepts. Metals 2022;12:97, doi:10.3390/met12010097. Sente software Ltd. (2005) ‘A collection of free downloadable papers on the development and application of JMatPro’, http://www.sentesoftware.co.uk/biblio.html. ASTM A1033-04, Standard practice for quantitative measurement and reporting of hypoeutectoid carbon and low-alloy steel phase transformations; ASTM International: west Conshohocken. ASTM International. ASTM E3-11(2017), Standard Guide for Preparation of Metallographic Specimens. West Conshohocken, PA: ASTM International; 2017. Mueller JJ, Hu X, Sun X, Ren Y, Choi K, Barker E, Speer JG, Matlock DK, De Moor E. Austenite formation and cementite dissolution during intercritical annealing of a medium-manganese steel from a martensitic condition. Mat. Des. 2021;203:109598, doi:10.1016/j.matdes.2021.109598. Liu G, Li B, Xu S, Tong S, Wang X, Liang X, Sun X. Effect of intercritical annealing temperature on multiphase microstructure evolution in ultra-low carbon medium manganese steel. Mater. Char. 2021;173: 110920, doi:10.1016/j.matchar.2021.110920. Xu Y, Li W, Du H, Jiao H, Liu B, Wu Y, Ding W, Luo Y, Ni Y, Min N, Liu W, Jin X. Tailoring the metastable reversed austenite from metastable Mn-rich carbides. Acta. Mater. 2021;214:116986, doi:10.1016/j.actamat.2021.116986. Kamoutsi H, Gioti E, Haidemenopoulos GN. Kinetics of solute partitioning during intercritical annealing of a medium-Mn steel. Metall. Mater. Trans. A 2015;46:4841–4846, doi:10.1007/s11661-015-3118-7. Kalsar R, Sanamar S, Schell N, Brokmeier HG, Saha R, Ghosh P, Bhagat AN, Suwas S. Elemental partitioning in medium Mn steel during short-time annealing: An in-situ study using synchrotron X-rays. Materialia, 2020;9:100594, doi:10.1016/j.mtla.2020.100594. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 05 Sep, 2025 Reviews received at journal 04 May, 2025 Reviewers agreed at journal 24 Apr, 2025 Reviewers agreed at journal 22 Apr, 2025 Reviewers invited by journal 01 Apr, 2025 Editor assigned by journal 01 Apr, 2025 Submission checks completed at journal 01 Apr, 2025 First submitted to journal 27 Mar, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6319163","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":445835702,"identity":"c520763d-f57e-4dcc-8351-8335fa7eaf40","order_by":0,"name":"Mateusz MORAWIEC","email":"","orcid":"","institution":"Silesian University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Mateusz","middleName":"","lastName":"MORAWIEC","suffix":""},{"id":445835703,"identity":"c40fe5a1-fa88-413f-99ea-55a350c8c4f8","order_by":1,"name":"Jarosław OPARA","email":"data:image/png;base64,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","orcid":"","institution":"Łukasiewicz Research Network – Upper Silesian Institute of Technology","correspondingAuthor":true,"prefix":"","firstName":"Jarosław","middleName":"","lastName":"OPARA","suffix":""},{"id":445835704,"identity":"df5d5039-443c-4db7-8077-288eb31e6ef4","order_by":2,"name":"Aleksandra KOZŁOWSKA","email":"","orcid":"","institution":"Silesian University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Aleksandra","middleName":"","lastName":"KOZŁOWSKA","suffix":""},{"id":445835705,"identity":"17ec0a92-006d-42c1-b239-75d785feca26","order_by":3,"name":"Adam GRAJCAR","email":"","orcid":"","institution":"Silesian University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Adam","middleName":"","lastName":"GRAJCAR","suffix":""}],"badges":[],"createdAt":"2025-03-27 09:38:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6319163/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6319163/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":81089970,"identity":"d91d8b0d-5350-4a0f-827c-8a4c9de6abac","added_by":"auto","created_at":"2025-04-22 06:57:47","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":295769,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic illustration of heat treatment process applied for investigated steels\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/2ec4071be70ff80024fdc4b3.png"},{"id":81089972,"identity":"6357bf47-50d6-4e65-a73a-73297d28e412","added_by":"auto","created_at":"2025-04-22 06:57:47","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":657333,"visible":true,"origin":"","legend":"\u003cp\u003ePhase fractions calculated using JMatPro\u003csup\u003e®\u003c/sup\u003e and Thermo-Calc\u003csup\u003e®\u003c/sup\u003e (a–c) and Fe–Fe₃C equilibrium phase diagrams (d–f) for the 3MnNb, 4MnNb, and 5MnNb steels. In (a–c), the equilibrium fractions of BCC, FCC, and Fe₃C are plotted as functions of temperature. In (d–f), the Thermo-Calc\u003csup\u003e®\u003c/sup\u003e-derived Fe–Fe₃C phase diagrams are shown\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/2cc94a5306247ddb452c4e8c.png"},{"id":81089971,"identity":"5e107347-e16c-4c94-9f82-f4e4459a0197","added_by":"auto","created_at":"2025-04-22 06:57:47","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":488199,"visible":true,"origin":"","legend":"\u003cp\u003eCalculated equilibrium phase diagrams for the Fe–Mn system (a–c) and the Fe–Al system (d–f) in the 3MnNb, 4MnNb, and 5MnNb steels\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/bf98456574af55b8968b9827.png"},{"id":81090420,"identity":"66a37645-be7c-4d7c-ab57-0e7ae294fb5f","added_by":"auto","created_at":"2025-04-22 07:05:47","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":582320,"visible":true,"origin":"","legend":"\u003cp\u003eQuantitative measurements of martensite and austenite lath widths in the microstructure after intercritical annealing at 700 °C for 60 minutes (a); schematic representation of the simulation domain used in DICTRA modelling, where w\u003csub\u003eα′\u003c/sub\u003e denotes the martensite lath width and w\u003csub\u003eγ\u003c/sub\u003e represents the initial austenite lath width prior to partitioning (b).\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/fc04fbc345d395dd827c4ee2.png"},{"id":81091579,"identity":"e3d5ada0-04bb-41a7-8a5c-8538821f05c8","added_by":"auto","created_at":"2025-04-22 07:13:47","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":510099,"visible":true,"origin":"","legend":"\u003cp\u003eDICTRA-simulated diffusion profiles of carbon (a, d, g), manganese (b, e, h), and aluminium (c, f, i) in reversed austenite for 3MnNb, 4MnNb, and 5MnNb steels. Profiles shown after 60 min of isothermal annealing at 700 °C, 720 °C, and 760 °C\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/528b800d46d3339bf9c7e0bc.png"},{"id":81089979,"identity":"5430ee46-58e1-4b8f-b47b-d08d573914c8","added_by":"auto","created_at":"2025-04-22 06:57:47","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":161987,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of manganese and aluminium concentrations in retained austenite laths in 3MnNb, 4MnNb, and 5MnNb steels, based on EDS measurements and DICTRA-predicted average values after 60 min of intercritical annealing at selected temperatures\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/cb7a19d94fdc7933ee30b148.png"},{"id":81089986,"identity":"20211ed8-5f76-46c8-bf0a-b770fedc2015","added_by":"auto","created_at":"2025-04-22 06:57:47","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":457995,"visible":true,"origin":"","legend":"\u003cp\u003eDilatometric analysis of the investigated medium-Mn steels: (a–c) heating dilatation and differential curves used for the determination of A\u003csub\u003ec1\u003c/sub\u003e and A\u003csub\u003ec3\u003c/sub\u003e temperatures for 3MnNb, 4MnNb, and 5MnNb steels, respectively; (d) cooling dilatograms used to determine the M\u003csub\u003es\u003c/sub\u003e temperature for each composition, following austenitization at 1100 °C and quenching to room temperature\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/557854bfd8a75d154ca99cf3.png"},{"id":81090418,"identity":"becefe62-af79-4c84-a608-48d2c3b7ed5b","added_by":"auto","created_at":"2025-04-22 07:05:47","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":228938,"visible":true,"origin":"","legend":"\u003cp\u003eDilatometric curves recorded during cooling of the investigated steels (3MnNb, 4MnNb, and 5MnNb) after intercritical annealing for 60 minutes: (a) at 700 °C, showing no detectable martensitic transformation upon cooling; (b) at higher IA temperatures (720 °C and 760 °C), where inflection points indicate partial martensitic transformation in selected alloys\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/7427476459f9fd975798d0f2.png"},{"id":81089982,"identity":"1b491c59-57d4-40c0-a61e-f30feddf28b6","added_by":"auto","created_at":"2025-04-22 06:57:47","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":55127,"visible":true,"origin":"","legend":"\u003cp\u003eJMatPro\u003csup\u003e®\u003c/sup\u003e-based predictions of M\u003csub\u003es\u003c/sub\u003e temperature and carbon concentration in austenite as a function of IA temperature for steels with varying Mn contents\u003c/p\u003e","description":"","filename":"image9.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/af00a244f057ab2bf3657488.png"},{"id":81089983,"identity":"04b7c415-7699-485a-b6da-9a59954db2ed","added_by":"auto","created_at":"2025-04-22 06:57:47","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":1443979,"visible":true,"origin":"","legend":"\u003cp\u003eSEM images of the steels with different Mn contents after IA at 700°C for 60 minutes: (a) 3MnNb; (b) 4MnNb; (c) 5MnNb. RA – retaned austenite; F – ferrite; MA – martensitic-austenitic islands\u003c/p\u003e","description":"","filename":"image10.png","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/ed105f3e7a1203ff10d9d8ba.png"},{"id":81092173,"identity":"3b98becc-729d-4c36-9c0e-5d2e409e117c","added_by":"auto","created_at":"2025-04-22 07:21:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4852180,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6319163/v1/8567f32c-1b6c-48cf-901f-0e4f7c851504.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Effect of Intercritical Annealing Temperature on Austenite Formation in Medium-Mn Steels: A Thermodynamic and Experimental Study","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eTypically, medium-Mn automotive sheet steels subjected to intercritical annealing (IA) heat treatment develop a multiphase microstructure predominantly composed of ferrite and austenite [1-3], but may also contain martensitic-austenitic (M–A) constituents or small amounts of other phases depending on heat treatment parameters. In the intercritical range between \u003cem\u003eA\u003csub\u003ec1\u003c/sub\u003e\u003c/em\u003e and \u003cem\u003eA\u003csub\u003ec3\u003c/sub\u003e\u003c/em\u003e temperatures, ferrite and austenite form from either an initial ferrite-pearlite or a martensitic microstructure [4]. The nature of this initial microstructure significantly affects both the kinetics of austenite and ferrite formation and the morphology of the resulting phases. A duplex ferrite–austenite microstructure that develops from lath martensite tends to inherit a layered morphology, whereas starting with a ferrite-pearlite microstructure leads to more globular morphologies [5]. Moreover, the final intercritical microstructure is significantly influenced by the prior thermomechanical history. Hot rolling typically yields a lath-type starting microstructure [5], which, upon intercritical annealing, retains a layered morphology; in contrast, cold-rolled material that undergoes recrystallization during heating leads to a more globular morphology [4]. In general, \u0026nbsp; multiphase steels with a lath-type microstructure exhibit a superior combination of high strength and ductility [7]. Morawiec et al. [8] reported that increasing the Mn content from 3 to 5 wt.% fosters the formation of such lath-type microstructures due to enhanced hardenability.\u0026nbsp;In addition, the higher dislocation density in lath-type microstructures suppresses Lüders elongation and flow curve serrations in stress–strain curves [9,\u0026nbsp;10]. An essential structural phenomenon during the IA of multiphase steels is the partitioning of key alloying elements, particularly carbon and manganese, from ferrite to austenite [11]. As a result, a high austenite fraction (20-40%) can be stabilized at room temperature (RT) [1-4]. Because carbon atoms have a small atomic radius, C diffusion occurs relatively quickly, which is crucial for maintaining austenite stability in automotive steel sheets produced in continuous annealing lines, with soaking times of 60– 180 s. Conversely, the slower diffusion\u0026nbsp;of Mn necessitates significantly longer times. Therefore, manganese is used to enhance austenite stability mainly in batch annealed steels, requiring several hours of soaking [12]. Both elements effectively lower the martensite start (\u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e) temperature, and their combined concentrations commonly suppress\u0026nbsp;\u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e temperature below RT. It should be also noted that higher Mn contents in steel reduce the degree of carbon enrichment in the austenitic phase [13]. Elements like Al and Si can also diffuse preferentially to ferrite; in particular, Al elevates\u0026nbsp;\u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e temperature, diminishing the thermodynamic stability of austenite. Nevertheless, aluminium remains a popular addition in various Advanced High Strength Steels (AHSS) due to its lightweighting potential [14]. Mn is notably prone to microsegregation [15]. In addition, it can hinder the redistribution of C from ferrite to austenite [16]. As a result, the ferrite-to-austenite ratio strongly depends on the austenite’s chemical composition and time-temperature parameters applied during IA process [17,\u0026nbsp;18]. The amount and stability of retained austenite (RA) significantly influence the mechanical properties of medium-Mn steels, making the optimization of IA conditions a critical issue. Furthermore, recent studies have expanded the understanding of austenite formation and stability in medium-Mn steels. In particular, manganese partitioning has been shown to proceed preferentially along grain boundaries and dislocations, thereby accelerating austenite growth during intercritical annealing [19]. The formation of Mn-enriched cementite can also influence local compositional gradients and affect how austenite nucleates [20]. Moreover, room-temperature quenching and partitioning (Q\u0026amp;P) treatments applied to medium-Mn steels have been reported to preserve the austenite fraction while improving its mechanical stability [21]. In parallel, nanostructured bainitic steels (nanobainite) have also been developed to maximize strength and toughness via extremely fine austenite/ferrite lamellae, where the control of C partitioning plays a similarly critical role [22,\u0026nbsp;23]. These steels—although differing in processing route—share with medium-Mn steels a reliance on the stabilization of austenite and kinetic tailoring of phase transformations. The last findings highlight that, although carbon content crucially determines mechanical stability, annealing temperature plays a more decisive role in the thermal stability of austenite than soaking time [24]. These insights underscore the importance of carefully tuning both chemical composition and processing parameters to achieve an optimal balance of austenite fraction and stability – an aspect central to the development of advanced high-strength steels with enhanced performance. The mentioned correlations between IA parameters and chemical composition of steels can be analysed using various computational tools based on the CALculation of PHAse Diagrams (CALPHAD) methodology [25]. Common software packages for metallic alloys include JMatPro\u003csup\u003e®\u003c/sup\u003e [26], Thermo-Calc\u003csup\u003e®\u003c/sup\u003e and DICTRA™\u0026nbsp;[27]. While JMatPro and Thermo-Calc primarily model phase transformations under equilibrium conditions\u0026nbsp;and allow predicting the thermodynamic behaviour of materials across a wide range of temperatures, pressures, and chemical compositions. However, the kinetics of element diffusion is equally critical for accurate predictions during intercritical annealing. In this regard, advanced modules (e.g., DICTRA) or integrated approaches are required to account for time-dependent diffusion and transformation mechanisms. For example, Dykas et al. [28] used JMatPro and Thermo-Calc to model phase diagrams and continuous cooling transformation (CCT) diagrams of medium-Mn steels, investigating how alloying elements such as Mn and C influence phase transition kinetics in steels with 0.1–0.2 wt.% C and 2–10 wt.% Mn. Morawiec et al. [8] used JMatPro to examine the influence of Mn content on the bainite fraction. These studies align with research on nanobainitic steels, where precise control over isothermal transformation temperatures and carbon distribution is key to achieving the desired microstructure [29]. JMatPro is also a valuable tool for simulating theoretical CCT and TTT (Time-Temperature-Transformation) diagrams [28]. Recently, Elaraby et al. [30] demonstrated how a CALPHAD-based approach using Thermo-Calc and JMatPro can be employed to design medium-Mn steels with enhanced hydrogen embrittlement resistance, highlighting the potential of computational methods to optimize alloy composition for improved austenite stability and mechanical performance. In turn, Tian et al. [31] applied DICTRA to investigate nano-scale Mn and Ni concentration gradients from the interface to the interior of retained austenite lath. Likewise, Hu et al. [32] explored how C and Mn distributions evolve at the ferrite/austenite interface during IA. Overall, such thermodynamic modelling tools are highly useful for designing and optimizing processing routes in metallic alloys. However, theoretical predictions must ultimately be corroborated through appropriate experimental validations.\u003c/p\u003e\n\u003cp\u003eSince the amount and stability of retained austenite play a crucial role in the mechanical properties of medium-Mn steels, and both the intercritical annealing process and manganese content significantly influence these aspects, a thorough investigation of their combined effects on austenite formation is essential. The aim of this study was to analyse the relationships between IA temperature, steel Mn content, and the resulting C, Mn, and Al contents in austenite using both computational and experimental approaches. To capture the equilibrium perspective, we used JMatPro and Thermo-Calc to model phase stability at various temperatures and compositions, while kinetic simulations (DICTRA) provided insight into non-equilibrium diffusion and transformation behaviour relevant to the short times typical of industrial processes. Dilatometry, energy-dispersive X-ray spectroscopy (EDS), and scanning electron microscopy (SEM) were then employed to verify the extent of these transformations in practice, bridging the gap between idealized equilibrium predictions and actual non-equilibrium conditions.\u003c/p\u003e"},{"header":"2. Material and experimental procedure","content":"\u003cp\u003e\u003cem\u003e2.1. Materials\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe analysed materials were medium-Mn steels containing 3 to 5 wt.% Mn. The chemical compositions are listed in Table 1. A relatively low carbon content (0.16-0.17 wt.%) is beneficial for steel weldability. Because manganese is an austenite stabilizer, various fractions of retained austenite can be obtained in steels containing 3\u0026ndash;5 wt.% Mn. In addition, aluminium and silicon prevent cementite precipitation [13]. Molybdenum was added to increase the steel\u0026rsquo;s hardenability and improve its hot ductility [33].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1.\u0026nbsp;\u003c/strong\u003eChemical compositions of analysed materials in wt.%.\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"425\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSteel\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMn\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAl\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSi\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 47px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMo\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNb\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003e3Mn\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e3.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e1.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 47px;\"\u003e\n \u003cp\u003e0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e0.04\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003e4Mn\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e3.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e1.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 47px;\"\u003e\n \u003cp\u003e0.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e0.04\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003e5Mn\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e5.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e1.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e0.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 47px;\"\u003e\n \u003cp\u003e0.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e0.04\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe investigated steels were produced via induction melting. The ingots were then hot forged to a thickness of 22 mm and subsequently hot rolled in the temperature range from 1200\u0026ndash;900\u0026deg;C, reducing the thickness to 9 mm. The final thermomechanical rolling consisted of three passes at deformation temperatures of 1050 \u0026deg;C, 950 \u0026deg;C and 850 \u0026deg;C, ultimately reaching a thickness of 4.5 mm. The flat samples were air-cooled to RT. Owing to the high hardenability provided by manganese, the resulting microstructures were fully martensitic [6, 7].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.2. Thermodynamic and kinetic modelling\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eIn order to capture both full equilibrium and non-equilibrium aspects of austenite formation, we used JMatPro\u003csup\u003e\u0026reg;\u003c/sup\u003e (Genereal Steels Module, database version 13) [34] and Thermo-Calc\u003csup\u003e\u0026reg;\u003c/sup\u003e (TCFE12, MOBFE7) [27] for equilibrium-based phase stability and element partitioning, while kinetic simulations of diffusion during intercritical annealing were performed using the DICTRA\u0026trade; module within the Thermo-Calc\u003csup\u003e\u0026reg;\u003c/sup\u003e software package. This approach accounts for the short soaking times typical of industrial IA processes, providing insight into time-dependent redistribution of elements under conditions far from full equilibrium. Although element partitioning during IA is computed here under equilibrium assumptions, these data offer a reference state that reflects the maximum extent of redistribution for carbon, manganese, and aluminium. The chemical compositions used in the simulations are listed in Table 1. In practice, the slower diffusion rate of manganese often means that full equilibrium is not achieved during short IA times. Nevertheless, equilibrium-based results can help identify trends\u0026mdash;such as the limiting fraction of austenite and the potential distribution of elements at higher temperatures\u0026mdash;which guide the subsequent kinetic analysis.\u003c/p\u003e\n\u003cp\u003eThe \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e temperature for each composition and IA temperature was estimated via JMatPro\u0026rsquo;s internal models, which combine thermodynamic data with empirically derived equations for \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e. This approach uses the computed phase compositions (notably the C and Mn content in austenite) to determine \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e. Although these calculations assume a (near-)equilibrium austenite composition, they serve as a baseline for understanding which conditions favour the retention of austenite upon cooling. To assess how closely the simulated \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e corresponds to real, non-equilibrium conditions, we compared the computed \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e values with those derived from dilatometric measurements of the investigated steels.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eGiven that full equilibrium is rarely reached in the short times typical of IA, DICTRA simulations provided additional insight into how incomplete manganese diffusion affects the time-dependent growth of austenite. By simulating diffusion profiles across ferrite/austenite interfaces, we captured deviations from the equilibrium composition predicted by JMatPro and Thermo-Calc. This combined approach\u0026mdash;equilibrium-based phase stability plus kinetic analysis\u0026mdash;helps bridge the gap between idealized predictions and the actual partitioning behaviour.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.3. Dilatometric investigations\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eCylindrical specimens, measuring 10 mm in length and 4 mm in diameter, were machined from the hot-rolled material with their long axes parallel to the rolling direction. The heat treatment was carried out using a high-resolution B\u0026auml;hr DIL805A/D dilatometer, where temperature was measured using a S-type thermocouple welded to the specimen\u0026rsquo;s centre. All experiments were conducted under vacuum, and helium served as the coolant. The dilatometric data were analysed in accordance with ASTM A1033-04 [35]. To ensure a uniform, fully martensitic starting microstructure\u0026mdash;favourable for the austenite reversion transformation (ART) during IA [6]\u0026mdash;the specimens were first austenitized (at 1100 \u0026deg;C for 5 min) and subsequently quenched to RT at approximately 60 \u0026deg;C/s. This rapid cooling rate was chosen to minimize any ferritic or bainitic transformations [8]. Next, the samples were heated at 3\u0026deg;C/s to selected IA temperatures between 700\u0026deg;C and 760\u0026deg;C and held for 60 min. The heating rate of 3 \u0026deg;C/s was chosen in accordance with standard dilatometric practice [6-8], while the temperature range (700\u0026ndash;760 \u0026deg;C) aligns with the intercritical region reported as suitable for medium-Mn steels [7]. A 60 min soaking time was adopted to allow partial diffusion of Mn and C into austenite [6]. Finally, the specimens were cooled to RT at 60 \u0026deg;C/s, which is both feasible in the dilatometer setup and sufficient to suppress ferritic or bainitic transformations upon cooling [8]. The schematic illustration of the entire heat treatment cycle applied to the investigated steels is presented in Fig. 1. In order to determine the critical \u003cem\u003eA\u003csub\u003ec\u003c/sub\u003e\u003c/em\u003e\u003csub\u003e1\u003c/sub\u003e and \u003cem\u003eA\u003csub\u003ec\u003c/sub\u003e\u003c/em\u003e\u003csub\u003e3\u003c/sub\u003e temperatures, the samples were subjected to a specialized time\u0026ndash;temperature cycle. First, they were heated from room temperature to 1100 \u0026deg;C at a rate of 2.5 \u0026deg;C/min. After reaching 1100 \u0026deg;C, the specimens were rapidly cooled down to room temperature at 10 \u0026deg;C/s. This procedure allowed precise identification of the \u003cem\u003eA\u003csub\u003ec\u003c/sub\u003e\u003c/em\u003e\u003csub\u003e1\u003c/sub\u003e and \u003cem\u003eA\u003csub\u003ec\u003c/sub\u003e\u003c/em\u003e\u003csub\u003e3\u003c/sub\u003e points based on the recorded temperature and dilatation curves, in accordance with [35].\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e2.4. Microstructural investigations\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eA standard metallographic preparation procedure [36] was applied to samples prior to microscopic observation. The specimens were cut into half perpendicularly to their length, then mechanically ground using SiC paper up to 2000 grit, polished with a diamond paste (up to 1 \u0026mu;m), and etched in 4% nital for 5 seconds to reveal microstructural detailes. The microstructure was examined using a Zeiss Supra 25 scanning electron microscope (SEM) in secondary electron (SE) mode. Energy-dispersive X-ray spectroscopy (EDS) was used to measure changes in Mn and Al concentrations in RA laths. The calibration was conducted using a molybdenum (Mo) standard with a purity of 99.99%, ensuring high precision and stability in energy measurements essential for accurate chemical composition analysis. The energy resolution of the detector was set at 137 eV, enabling clear distinction of spectral lines and providing detailed results with minimal noise interference. The acquisition time was configured to 6400 ns, which allows for optimal signal quality and accurate spectral intensity representation while maintaining a good signal-to-noise ratio. These parameters were selected to ensure reliable and precise results, tailored to the characteristics of the point analyses conducted.\u003c/p\u003e"},{"header":"3. Results and discussion","content":"\u003cp\u003e\u003cem\u003e3.1. Equilibrium and kinetic analysis of reversed austenite formation\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThermodynamic and kinetic simulations were performed to investigate the influence of IA temperature and Mn content on austenite fraction and elemental partitioning in medium-Mn steels. The calculations focused on phase stability and diffusion behaviour relevant to industrial annealing conditions.\u003c/p\u003e\n\u003cp\u003eFigures 2.a\u0026ndash;2.c present the theoretical fractions of austenite (FCC), ferrite (BCC), and cementite (CEM) as a function of intercritical annealing temperature, predicted for 3MnNb, 4MnNb, and 5MnNb steels, respectively. The simulations were carried out under the assumption of full thermodynamic equilibrium. Two thermodynamic software packages\u0026mdash;JMatPro and Thermo-Calc\u0026mdash;were used to enable a comparative assessment of phase stability in the intercritical region. In all compositions, an increase in the IA temperature led to a higher fraction of austenite, consistent with the expected transformation behaviour in medium-Mn steels. Additionally, a higher nominal manganese content resulted in a greater amount of austenite at the same temperature due to Mn\u0026rsquo;s strong austenite-stabilizing effect. At 700 \u0026deg;C, Thermo-Calc simulations predicted the austenite (FCC) phase fraction to be approximately 24.5 % for the 3MnNb steel, 28.1 % for the 4MnNb steel, and 37.9 % for the 5MnNb steel, indicating a clear increase in austenite stability with increasing Mn content. This trend was also observed at higher temperatures. For example, at 760 \u0026deg;C, the predicted FCC phase fractions rose to 35.5 % for 3MnNb, 40.3 % for 4MnNb, and 54.2 % for 5MnNb. These results confirm the strong stabilizing effect of manganese on the austenite phase during intercritical annealing and highlight the sensitivity of austenite fraction to both chemical composition and annealing temperature in medium-Mn steels. JMatPro simulations yielded slightly lower austenite fractions\u0026mdash;typically by 2\u0026ndash;6 % across the analysed temperature range. Notably, this discrepancy became more pronounced with increasing Mn content and temperature. The observed differences likely result from variations in the thermodynamic databases and interaction parameters used by each software package.\u003c/p\u003e\n\u003cp\u003eThe equilibrium phase diagrams (Fe-Fe\u003csub\u003e3\u003c/sub\u003eC system) for the investigated steels, along with the corresponding austenite transformation start (\u003cem\u003eA\u003csub\u003ee\u003c/sub\u003e\u003c/em\u003e\u003csub\u003e1\u003c/sub\u003e) and finish (\u003cem\u003eA\u003csub\u003ee\u003c/sub\u003e\u003c/em\u003e\u003csub\u003e3\u003c/sub\u003e) temperatures, were calculated using Thermo-Calc\u003csup\u003e\u0026reg;\u003c/sup\u003e and are illustrated in Figures 2.d\u0026ndash;2.f. These diagrams provide complementary insight into the thermodynamic stability of BCC, FCC, and CEM phases within the intercritical temperature range, supporting the interpretation of austenite formation trends under equilibrium conditions. With increasing Mn content, both the equilibrium austenite formation start temperature and finish temperature shift to lower values, confirming the strong austenite-stabilizing effect of manganese. Specifically, the \u003cem\u003eA\u003csub\u003ee\u003c/sub\u003e\u003c/em\u003e\u003csub\u003e3\u003c/sub\u003e temperatures are 940 \u0026deg;C, 924 \u0026deg;C, and 867 \u0026deg;C for 3MnNb, 4MnNb, and 5MnNb steels, respectively. The corresponding \u003cem\u003eA\u003csub\u003ee\u003c/sub\u003e\u003c/em\u003e\u003csub\u003e1\u003c/sub\u003e temperatures are 654 \u0026deg;C, 631 \u0026deg;C, and 469 \u0026deg;C for the 3MnNb, 4MnNb, and 5MnNb steels, respectively. These temperatures mark the onset of austenite nucleation at the expense of cementite, that is, the beginning of cementite dissolution [6, 37]. It should be noted, however, that in the investigated steels with an initial martensitic microstructure, cementite dissolution does not occur due to the absence of pre-existing due to the absence of pre-existing CEM, as will be discussed in a later section in the structure, as will be discussed in a later section.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMoreover, a progressive expansion of the three-phase (BCC + FCC + CEM) coexistence region is observed with increasing Mn concentration, indicating an extended temperature range in which ferrite, austenite, and cementite are thermodynamically stable. One can observe in equilibrium phase diagrams that in the lower part of the intercritical range (e.g., below 660 \u0026deg;C for 5MnNb and 680 \u0026deg;C for 3MnNb and 4MnNb), a sharp increase in the carbon concentration in FCC was observed, reflecting the progressive dissolution of cementite into the growing austenite. This process becomes more gradual at higher Mn contents, where the difference between minimum and maximum carbon concentrations within FCC is more pronounced. These findings are consistent with the literature data for Mn-rich steels [6,13,37], which report that manganese decreases the chemical potential of carbon in cementite, thereby moderating the driving force for carbon partitioning into austenite. As a result, austenite enrichment in carbon was significantly lower in the 5MnNb steel compared to the 3MnNb and 4MnNb grades, especially at elevated IA temperatures. Since the carbon content in austenite strongly influences the \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e temperature, the lower C enrichment in 5MnNb steel suggests that its retained austenite may exhibit reduced thermal stability relative to steels with lower Mn contents.\u003c/p\u003e\n\u003cp\u003eIn addition to the equilibrium phase boundaries, Figures 2.d\u0026ndash;2.f also present the results of DICTRA simulations, showing the average carbon concentration in retained austenite laths predicted under non-equilibrium diffusion conditions at selected intercritical annealing temperatures (700 \u0026deg;C, 720 \u0026deg;C, and 760 \u0026deg;C) after 60 minutes of isothermal holding. These values are superimposed onto the equilibrium phase diagrams for direct comparison. The horizontal error bars associated with these data points do not represent statistical uncertainty but instead reflect the minimum and maximum carbon concentrations across the austenite lath obtained from DICTRA simulations. This variation is a consequence of carbon segregation driven by interface-controlled diffusion kinetics during intercritical annealing. The significance of this carbon heterogeneity within retained austenite and its correlation with local phase stability will be discussed in detail in a subsequent section.\u003c/p\u003e\n\u003cp\u003eThe equilibrium behaviour of manganese in the Fe\u0026ndash;Mn system was analysed for the 3MnNb, 4MnNb, and 5MnNb steels using Thermo-Calc\u003csup\u003e\u0026reg;\u003c/sup\u003e, as illustrated by the phase diagrams in Figures 3.a\u0026ndash;3.c. These diagrams provide a comprehensive view of the phase equilibria, including the phase boundaries between ferrite, austenite, and cementite, as well as the temperature ranges in which these phases can coexist. As the temperature increases, the Mn concentration in the FCC phase decreases, particularly above 700 \u0026deg;C, primarily because of reduced manganese partitioning from ferrite under equilibrium conditions. Steels with higher nominal Mn content, especially 5MnNb, exhibit greater enrichment of manganese in the FCC phase across the intercritical region. These results are consistent with previous findings by Liu et al. [38], who reported that higher annealing temperatures reduce the Mn and C content in austenite, thereby lowering its thermal stability and promoting chemical inhomogeneity within austenite laths. Furthermore, the three-phase region (BCC + FCC + CEM) is most extensive in the 3MnNb steel and progressively narrows as the Mn content increases, ultimately giving way to a two-phase (BCC + CEM) field in the 5MnNb alloy. This shift in phase stability is a direct result of changes in chemical potential and the distribution of substitutional elements.\u003c/p\u003e\n\u003cp\u003eA similar analysis was performed for aluminium in the Fe\u0026ndash;Al system, with the corresponding equilibrium phase diagrams shown in Figures 3.d\u0026ndash;3.f. These diagrams illustrate not only the solubility limits of Al in austenite, but also the phase boundaries for ferrite and cementite, as well as the two-phase coexistence regions. As the temperature increases, the aluminium content in the FCC phase rises gradually. Below 900 \u0026deg;C, this increase is nearly linear, and the differences among the steel grades are relatively small. However, above 900 \u0026deg;C, a sharp, non-linear rise in Al solubility in austenite is observed, and the differences in FCC solubility boundary Al content among the alloys become more pronounced. Notably, the extent of austenite enrichment in Al remains significantly lower than that observed for carbon (Fig. 3a) and manganese (Fig. 3.b), reflecting both the strong ferrite-stabilizing effect of aluminium and its limited solubility in the FCC phase under equilibrium conditions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo complement the equilibrium analysis, DICTRA simulations were performed to evaluate the non-equilibrium partitioning of manganese and aluminium across the ferrite/austenite interface. The Fe\u0026ndash;Mn and Fe\u0026ndash;Al diagrams shown in Figures 3.a\u0026ndash;3.f include these simulation results, displaying the average concentrations of Mn and Al in retained austenite (FCC) after 60 minutes of isothermal holding at selected intercritical annealing temperatures (700 \u0026deg;C, 720 \u0026deg;C, and 760 \u0026deg;C). These values are superimposed on the equilibrium phase diagrams to enable direct comparison between thermodynamic predictions and kinetically influenced elemental distributions. The horizontal error bars associated with the DICTRA data points do not reflect statistical variation but rather represent the range of local concentrations (minimum\u0026ndash;maximum) of Mn and Al within the FCC phase, derived from the simulated concentration profiles. The effects of local Mn and Al concentration gradients in retained austenite\u0026mdash;particularly their influence on phase stability and the initiation of martensitic transformation\u0026mdash;will be further examined in the following sections.\u003c/p\u003e\n\u003cp\u003eTo analyze the kinetics of austenite formation and the concurrent partitioning of the main alloying elements (C, Mn, Al) during intercritical annealing (IA), one-dimensional diffusion simulations were performed using the DICTRA module of Thermo-Calc\u003csup\u003e\u0026reg;\u003c/sup\u003e. The geometry of the simulation domains was defined based on quantitative measurements of the martensite and austenite laths observed in the microstructure after IA at 700 \u0026deg;C for 60 minutes, the results of which are shown in Fig. 4.a. The IA temperature at 700 \u0026deg;C was selected as a reference point because it lies near the lower end of the intercritical range, where diffusion distances and kinetic limitations are expected to exert the strongest influence on partitioning behaviour. The width of martensite (w\u003csub\u003e\u0026alpha;\u0026prime;\u003c/sub\u003e) and austenite (w\u003csub\u003e\u0026gamma;\u003c/sub\u003e) laths was measured from high-resolution SEM micrographs, and the mean distance from center to center was used to estimate the effective diffusion path. In the simulations, this spacing was taken to represent the martensite-side domain thickness, as indicated schematically in Fig. 4.b. The initial microstructure, consisting of low-carbon martensite (\u0026alpha;\u0026prime;), was treated as ferrite (\u0026alpha;) due to its minimal tetragonality after IA, which renders it effectively indistinguishable from ferrite in X-ray diffraction analysis [1]. Specifically, three diffusion domains were defined in DICTRA for the 3MnNb, 4MnNb, and 5MnNb steels, with total sizes of 325 nm, 260 nm, and 455 nm, respectively. In each case, the model consisted of a supersaturated ferrite (\u0026alpha;\u0026prime;) region and a 1 nm-thick austenite (\u0026gamma;) nucleus positioned at the left boundary to initiate phase reversion. A schematic of the simulation geometry and a high-magnification SEM micrograph used for dimensional calibration are shown in Fig. 4.b.\u003c/p\u003e\n\u003cp\u003eFigures 5.a\u0026ndash;5.i illustrate the simulated evolution of reversed austenite during the ART process, along with the corresponding concentration profiles of carbon, manganese, and aluminium in the 3MnNb, 4MnNb, and 5MnNb steels after 60 minutes of isothermal annealing at selected intercritical temperatures: 700 \u0026deg;C, 720 \u0026deg;C, and 760 \u0026deg;C. For reference, the early-stage elemental distributions after 1 second of simulation are also shown to highlight the initial partitioning behaviour and the steep concentration gradients developing at the \u0026alpha;/\u0026gamma; interface in the initial phase of the IA process.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCarbon, due to its high diffusivity in both phases, exhibited rapid enrichment in the growing austenite lath regardless of steel composition. The simulated profiles showed steep concentration gradients across the \u0026alpha;\u0026prime;/\u0026gamma; interface at lower temperatures, which gradually became more uniform as IA temperature increased. At 700 \u0026deg;C, sharp C peaks were observed within the \u0026gamma; phase, especially for 3MnNb and 4MnNb, with maximum values reaching ~0.56 wt.% for 3MnNb. These values decreased with increasing temperature due to accelerated homogenization and volumetric expansion of the FCC phase, leading to lower average and maximum C concentrations. Despite these differences, all three alloys showed an increasing \u0026gamma; width and transformed volume fraction with temperature, indicating that even with lower C concentrations, austenite was stabilized by cumulative contributions from other elements and longer exposure time. Notably, carbon partitioning was most efficient in 3MnNb, where the diffusion distance was shortest, leading to higher localized enrichment near the interface.\u003c/p\u003e\n\u003cp\u003eUnlike carbon, manganese exhibited sluggish diffusion kinetics, resulting in strong chemical gradients across the \u0026alpha;\u0026prime;/\u0026gamma; interface, particularly at lower IA temperatures. The simulations revealed that Mn redistribution was incomplete even after 60 minutes at 700 \u0026deg;C, especially in 5MnNb, where diffusion paths were longest due to the coarser initial martensitic structure. At 700 \u0026deg;C, the Mn concentration ranged from ~5.3 wt.% (min) to ~9.1 wt.% (max) in the \u0026gamma; phase of 5MnNb, indicating pronounced inhomogeneity. With increasing temperature, Mn partitioning became more efficient, yet a strong discrepancy between max and min Mn values persisted in all alloys. The 3MnNb steel showed relatively narrower segregation ranges, likely due to both shorter diffusion distances and lower nominal Mn content. These results highlight the critical role of Mn kinetics in determining the local austenite composition and suggest that Mn inhomogeneity can induce local variations in \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e temperature and mechanical stability of retained austenite.\u003c/p\u003e\n\u003cp\u003eAluminium, although generally considered a slow diffuser in BCC and FCC iron, showed measurable partitioning behaviour in the DICTRA simulations. Similar to Mn, Al exhibited asymmetric profiles across the \u0026alpha;\u0026prime;/\u0026gamma; boundary, with higher concentrations near the ferritic side and lower solubility within the austenite. At lower temperatures (700 \u0026deg;C), a steep Al gradient was observed within the \u0026gamma; phase, particularly in 3MnNb, where the max and min values differed by over 0.44 wt.%.The segregation range (max\u0026ndash;min Al) decreased with increasing IA temperature, confirming that higher temperatures facilitated more uniform redistribution, albeit still kinetically limited. Among the three steels, 5MnNb exhibited the highest retained austenite fraction and largest lath width, yet its Al content remained lower within the FCC phase, especially at 700 \u0026deg;C. This effect is attributed to the low solubility of Al in austenite and its tendency to accumulate in ferrite, where it contributes to stabilizing the BCC phase and potentially increases \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e locally.\u003c/p\u003e\n\u003cp\u003eFurthermore, the average concentrations of C, Mn, and Al in retained austenite, calculated from the DICTRA simulations and superimposed on the equilibrium phase diagrams (Figures 2 and 3), demonstrate that the partitioning behaviour deviates significantly from equilibrium thermodynamic predictions. Not only do the mean compositions fall well below the respective solubility limits defined by the equilibrium boundaries, but even the maximum local concentrations at the ferrite/austenite interface remain far from the equilibrium tie-lines. This disparity highlights the non-equilibrium nature of the intercritical annealing process, especially under industrially relevant timescales, and underscores the critical role of kinetic constraints in shaping the compositional gradients and stability of austenite in medium-Mn steels.\u003c/p\u003e\n\u003cp\u003eThe DICTRA simulations confirm that all three main alloying elements (C, Mn, and Al) in the investigated steels undergo complex, temperature-dependent redistribution during intercritical annealing, collectively governing the growth and stability of austenite. While carbon partitions rapidly and uniformly into the FCC phase, Mn and Al exhibit pronounced concentration gradients due to their lower diffusivity. These gradients arise from the limited mobility of substitutional elements and reflect the non-uniform, kinetically controlled redistribution under non-equilibrium conditions [19]. Manganese, in particular, shows stronger segregation compared to aluminium, consistent with its lower diffusion coefficient and stronger interaction with carbon [13]. The extent of elemental partitioning is further influenced by initial microstructural parameters (domain size), annealing temperature, and nominal alloy composition. Importantly, the simulations suggest that local heterogeneities in Mn and Al concentration within retained austenite may introduce significant variations in local stability (i.e., variations in \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e temperature), which could manifest as sequential martensitic transformation upon cooling [38].\u003c/p\u003e\n\u003cp\u003eTo further validate the predicted partitioning behaviour of substitutional elements during IA, the manganese and aluminium concentrations in RA laths were experimentally measured using EDS and compared with the average values obtained from DICTRA simulations. Figure 6 presents this comparison for the 3MnNb, 4MnNb, and 5MnNb steels after 60 minutes of annealing at 700 \u0026deg;C. The EDS measurements, performed on individual austenite laths identified in SEM micrographs, provide insight into the local chemical composition resulting from element partitioning during IA. In the simulations, it was assumed that the reverted austenite remained thermally stable and did not transform into martensite upon cooling, thereby persisting as retained austenite. This assumption justifies the direct comparison between DICTRA-predicted average concentrations in the austenite phase and experimentally measured EDS data. The results for manganese, summarized in Figure 6.a, show a good overall agreement between DICTRA simulations and EDS measurements. For the 3Mn and 4Mn steels, the simulated Mn contents in RA (4.69 wt.% and 5.37 wt.%, respectively) are slightly higher than the corresponding EDS values (4.55 wt.% and 4.78 wt.%), yet the differences fall within the experimental and simulation standard deviations. For the 5Mn steel, the EDS data indicate a higher average Mn content (7.02 wt.%) compared to the simulation (6.15 wt.%), although both values exhibit relatively large standard deviations (1.22 and 1.03, respectively). This suggests a greater compositional heterogeneity in the RA laths of the 5Mn steel, consistent with prior reports indicating that increased nominal Mn content enhances segregation effects during austenite reversion [19,21,]. Xu et al. [39] also observed that austenite nucleating at metastable carbides tends to exhibit strong Mn concentration gradients, affecting local phase stability. The convergence between simulated and experimental mean values, alongside comparable deviations, demonstrates the ability of DICTRA to reasonably capture Mn partitioning trends in medium-Mn steels subjected to IA, especially under short annealing durations representative of industrial conditions.\u003c/p\u003e\n\u003cp\u003eA similar comparison was carried out for aluminium, as shown in Figure 6.b. The DICTRA-predicted aluminium contents in the austenite phase after IA at 700 \u0026deg;C for 60 minutes show slightly lower values than those measured experimentally by EDS, but the trends across the three steel compositions remain consistent. For the 3Mn, 4Mn, and 5Mn steels, the simulated Al concentrations in RA were 1.43 wt.%, 1.52 wt.%, and 1.48 wt.%, respectively, while the EDS measurements yielded slightly higher averages of 1.73 wt.%, 1.62 wt.%, and 1.72 wt.%. The differences are within a reasonable range (\u0026le;0.25 wt.%) and can be partially attributed to the limited spatial resolution and averaging effects inherent to EDS analysis at sub-micrometer scales, as well as assumptions in the DICTRA model regarding the initial interface geometry and mobility databases. Moreover, the standard deviations from both DICTRA and EDS are relatively low, indicating that aluminium partitioning into the austenite phase is more uniform compared to manganese. This is in line with the higher diffusivity of Al and its weaker thermodynamic interaction with carbon compared to Mn, which results in smoother concentration gradients and more homogeneous elemental distribution, as previously observed in experimental and modelling studies of medium-Mn steels [19].\u003c/p\u003e\n\u003cp\u003eTaken together, the combined comparison of experimental and simulated results for both Mn and Al confirms the reliability of DICTRA simulations in capturing the essential features of element partitioning during ART, supporting their use as a predictive tool for designing microstructures in medium-Mn steels.\u003c/p\u003e\n\u003cp\u003e3.2. Determination of characteristic temperatures\u003c/p\u003e\n\u003cp\u003eDilatometric experiments were performed to determine the key phase transformation temperatures (A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e1\u003c/sub\u003e,\u0026nbsp;A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e3\u003c/sub\u003e, and\u0026nbsp;M\u003csub\u003es\u003c/sub\u003e) for the investigated medium-Mn steels. It is important to note that each IA cycle was initiated from a fully martensitic microstructure, which ensured that any transformation observed during reheating was associated with austenite reversion or new phase formation, rather than the dissolution of pre-existing phases such as pearlite or cementite.\u003c/p\u003e\n\u003cp\u003eFigure 7 presents the dilatometric results for the 3MnNb, 4MnNb, and 5MnNb steels, illustrating the methodology used to determine critical transformation temperatures. In Figures 7.a\u0026ndash;7.c, the dilatation curves\u0026mdash;presented in terms of relative change of length (RCL)\u0026mdash;and their first derivatives are shown, enabling the identification of the\u0026nbsp;A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e1\u003c/sub\u003e and A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e3\u003c/sub\u003e points based on the inflection and slope changes in thermal expansion behaviour. These characteristic temperatures were evaluated from heating curves, where the tangents were applied to the transition regions between the linear segments to define the onset and completion of austenite formation. It should be emphasized that the interpretation of A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e1\u003c/sub\u003e in these medium-Mn steels with an initial martensitic structure poses certain challenges. Unlike traditional ferrite-pearlite microstructures, where A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e1\u003c/sub\u003e corresponds to the start of cementite dissolution into austenite, the martensitic structure promotes the precipitation of fine cementite particles at sub-A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e1\u003c/sub\u003e temperatures during reheating [6]. This precipitation process introduces dilatometric artifacts that interfere with the typical linear thermal expansion trend, complicating the accurate determination of A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e1\u003c/sub\u003e from dilatation curves. The critical A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e1\u003c/sub\u003e temperatures were determined to be 626 \u0026deg;C for 3MnNb, 656 \u0026deg;C for 4MnNb, and 631 \u0026deg;C for 5MnNb. The corresponding A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e3\u003c/sub\u003e temperatures were 931 \u0026deg;C, 956 \u0026deg;C, and 934 \u0026deg;C, respectively. These values vary systematically with the manganese content, consistent with the well-known austenite-stabilizing effect of Mn, which promotes \u0026gamma;-phase formation at lower temperatures. A lower A\u003csub\u003ec\u003c/sub\u003e\u003csub\u003e1\u003c/sub\u003e temperature also broadens the intercritical processing window, which is advantageous for industrial control of IA parameters.\u003c/p\u003e\n\u003cp\u003eFigure 7.d presents the cooling segments of the dilatometric curves used to determine the\u0026nbsp;M\u003csub\u003es\u003c/sub\u003e temperatures for each steel composition. These measurements were conducted on samples austenitized at 1100 \u0026deg;C for 5 min and subsequently quenched to room temperature at approximately 60 \u0026deg;C/s to produce a fully martensitic microstructure. The M\u003csub\u003es\u003c/sub\u003e temperatures were identified as the points corresponding to the onset of rapid contraction, associated with the \u0026gamma; \u0026rarr; \u0026alpha;\u0026prime; martensitic transformation. The measured M\u003csub\u003es\u003c/sub\u003e values were 458 \u0026deg;C for 3MnNb, 436 \u0026deg;C for 4MnNb, and 354 \u0026deg;C for 5MnNb. This trend reflects the increasing thermal stability of austenite with higher Mn content, consistent with the austenite-stabilizing effect of manganese.\u003c/p\u003e\n\u003cp\u003eFigure 8 presents dilatometric curves recorded during cooling of the investigated steels (3MnNb, 4MnNb, and 5MnNb) after intercritical annealing for 60 minutes at different temperatures. In Figure 8.a, which corresponds to IA at 700 \u0026deg;C, no transformation-related deviations were observed. The cooling curves are nearly linear, without inflection points typically associated with phase transformations, indicating that the microstructures formed during IA remained stable down to room temperature. This suggests that the applied IA parameters were sufficient to ensure effective partitioning of carbon and manganese from ferrite to austenite, stabilizing the retained austenite and suppressing the \u0026gamma; \u0026rarr; \u0026alpha;\u0026prime; martensitic transformation [40, 41].\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eBy contrast, Figure 8.b presents the dilatometric curves obtained during cooling of the same steels after intercritical annealing at 720 \u0026deg;C and 760 \u0026deg;C. In these cases, deviations from the linear contraction trend were observed, corresponding to the onset of martensitic transformation. This transformation was clearly detected in the 5Mn steel annealed at both 720 \u0026deg;C and 760 \u0026deg;C, and in the 3Mn steel after annealing at 760 \u0026deg;C. The transformation onset was identified by a characteristic elongation of the sample, reflecting the volume expansion associated with the \u0026gamma; \u0026rarr; \u0026alpha;\u0026prime; transformation. Interestingly, despite its higher nominal Mn content, the 5Mn steel exhibited a more pronounced martensitic transformation after IA at 760 \u0026deg;C compared to the 3Mn steel. This behavior can be attributed to the increased volume fraction of austenite formed at elevated IA temperatures, which was not sufficiently enriched in stabilizing elements such as C and Mn. As a result, the austenite remained metastable and transformed partially into martensite during cooling. These findings are consistent with previous studies by Skowronek et al. [6], who demonstrated that intercritical annealing within the range of 680\u0026ndash;700 \u0026deg;C is sufficient to stabilize austenite in 5 wt.% Mn steels, thereby suppressing martensitic transformation upon cooling to room temperature.\u003c/p\u003e\n\u003cp\u003eFigure 9 illustrates the variation of\u0026nbsp;M\u003csub\u003es\u003c/sub\u003e temperature as a function of IA temperature for the investigated steels. Contrary to the initial assumption, the M\u003csub\u003es\u003c/sub\u003e temperature increases with increasing IA temperature. This trend reflects a reduction in the chemical stability of austenite formed at higher annealing temperatures, which is attributed to a lower degree of enrichment in carbon and manganese due to faster diffusion dynamics and a higher volume fraction of austenite to be stabilized. The observed behaviour is consistent with the findings of Liu et al. [38] and Zhang et al. [5], who reported that the thermal stability of reversed austenite decreases at elevated IA temperatures due to dilution effects and heterogeneous partitioning. Notably, the 5Mn steel exhibits lower M\u003csub\u003es\u003c/sub\u003e temperatures compared to 3Mn and 4Mn steels across the entire IA temperature range, which confirms the strong stabilizing effect of manganese on austenite. This is further supported by prior thermodynamic analyses indicating that Mn reduces the chemical potential of carbon in austenite and cementite [31, 37], thereby promoting austenite stability. According to theoretical calculations, the IA temperature required to suppress M\u003csub\u003es\u003c/sub\u003e below room temperature was approximately 680 \u0026deg;C for 5Mn steel, and close to 700 \u0026deg;C for 3Mn and 4Mn steels.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e3.3. Microstructure observations after intercritical annealing\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe microstructures of the investigated steels after intercritical annealing at 700 \u0026deg;C for 60 minutes are shown in Figure 10. They were primarily composed of lath-like ferrite and retained austenite (RA), with the morphology and relative phase fractions varying depending on the manganese content. Note that, in the present study, the lath regions denoted as ferrite correspond to regions that were originally fully martensitic prior to IA treatment. Following the 60-minute annealing, significant carbon partitioning into austenite occurred, leading to a marked reduction in tetragonality and rendering the former martensite structurally and chemically indistinguishable from ferrite. This interpretation is consistent with the assumption adopted in the DICTRA simulations, where the low-carbon martensite was modelled as ferrite due to its negligible tetragonality after partitioning. While a detailed quantitative phase analysis was beyond the scope of the present work, qualitative evaluation of the SEM images suggests that the steel containing 3 wt.% Mn exhibits a higher proportion of ferrite (Fig. 10.a), whereas the 5Mn steel reveals a more extensive presence of retained austenite. This trend aligns with the thermodynamic predictions shown in Fig. 2, where the increased Mn content promotes austenite stability and fraction under intercritical annealing conditions. Moreover, a visual comparison indicates that the thickness of RA laths increases with higher Mn content; this observation is in qualitative agreement with DICTRA simulation results (Fig. 5). However, it should be noted that no direct measurements of RA thickness were performed, and a more detailed quantitative analysis (e.g., via image analysis or EBSD) would be necessary to validate this tendency. The presence of cementite was also confirmed, particularly in the 3Mn and 4Mn steels (Fig. 10.b), consistent with previous studies. Mueller et al. [37] reported similar findings in Fe\u0026ndash;0.19C\u0026ndash;4.39Mn steel, where manganese partitioning from ferrite to cementite reduced the chemical potential of carbon in cementite. This, in turn, decreased the driving force for carbon to diffuse toward austenite, thereby reducing the extent of carbon enrichment in RA and stabilizing cementite during intercritical annealing.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThe effect of Mn content in a range of 3-5 wt.% on the phase transformation kinetics during intercritical annealing of multiphase steels was analysed using thermodynamic and kinetic modelling and experimental approaches. The main findings of the study are:\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eAn increase in the Mn content in steel leads to a higher equilibrium and simulated fraction of austenite formed during intercritical annealing. Modelling at 700 °C predicted ~24 % of FCC in 3MnNb steel and ~38% in 5MnNb steel. This trend was confirmed by SEM observations and is consistent with the austenite-stabilizing effect of Mn. \u0026nbsp;\u003c/li\u003e\n \u003cli\u003eA higher Mn content in steel reduces the carbon content in austenite, thereby increasing its martensite start temperature and reducing thermal stability. This behaviour was evident from both DICTRA simulations and experimental\u0026nbsp;\u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e determinations via dilatometry.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eThe IA temperature significantly affects both austenite fraction and chemical composition. At higher annealing temperatures, the increased amount of austenite leads to a dilution effect, resulting in lower concentrations of C and Mn per unit volume of austenite and thus a higher \u003cem\u003eM\u003csub\u003es\u003c/sub\u003e\u003c/em\u003e temperature.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eThe microstructures of all investigated steels were primarily composed of lath-like ferrite and retained austenite. In the 3MnNb steel, only ferrite and RA were identified, whereas the 4MnNb steel additionally exhibited the presence of fine carbides. In contrast, the 5MnNb steel revealed small MA constituents, indicating increased heterogeneity of the microstructure with higher Mn content. Qualitative observations suggest that the thickness of RA laths increases with increasing Mn content, which is consistent with trends predicted by DICTRA simulations.\u003c/li\u003e\n \u003cli\u003eDICTRA simulations revealed strong concentration gradients of Mn and Al across the austenite/ferrite interface, particularly at lower IA temperatures. These non-equilibrium conditions were confirmed by some discrepancy between simulated and equilibrium phase boundaries.\u003c/li\u003e\n \u003cli\u003eEDS measurements confirmed the general trends predicted by DICTRA. In particular, the standard deviation of Mn content within retained austenite laths increased with bulk Mn concentration, reflecting incomplete homogenization due to limited diffusion during intercritical annealing. Al partitioning was more uniform, but still followed the modelled trend.\u0026nbsp;\u003c/li\u003e\n\u003c/ul\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e This research was funded in whole or in part by National Science Center through the grant no. 2021/41/N/ST8/03371. For the purpose of Open Access, the author has applied a CC-BY public copyright license to any Author Accepted Manuscript (AAM) version arising from this submission.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability Statement:\u0026nbsp;\u003c/strong\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of Interest:\u003c/strong\u003e The authors declare no conflict of interest.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eChandan AK, Bansal GK, Kundu J, Chakraborty J, Chowdhury SG. Effect of prior austenite grain size on the evolution of microstructure and mechanical properties of an intercritically annealed medium manganese steel. Sci. Eng. A 2019;768:138458, doi:10.1016/j.msea.2019.138458.\u003c/li\u003e\n\u003cli\u003eTjahjanto DD, Turteltaub S, Suiker ASJ. Crystallographically based model for transformation-induced plasticity in multiphase carbon steels. Continuum Mech Thermodyn 2008;19:399-422, doi:10.1007/s00161-007-0061-x.\u003c/li\u003e\n\u003cli\u003eKaczmarczyk J, Grajcar A, Kozłowska A, Cruz Banuelos JS, Opara J, Tatke N. Modelling of strain-induced martensite formation in advanced medium-Mn automotive sheet steel. Continuum Mech Thermodyn 2023;35:1841-1858, doi:10.1007/s00161-023-01202-2.\u003c/li\u003e\n\u003cli\u003eKang S, Hofer RW, Speer JG, Krizan D, Matlock DK, De Moor E. Intercritical annealing response of medium manganese steels having different carbon concentrations. Steel Res. Int. 2018;89:1800069, doi:10.1002/srin.201800069.\u003c/li\u003e\n\u003cli\u003eZhang BG, Zhang XM, Liu HT. Microstructural evolution and mechanical properties of Ni-containing light-weight medium-Mn TRIP steel processed by intercritical annealing. Mater. Sci. Eng. A 2020;793:139289, doi:10.1016/j.msea.2020.139289.\u003c/li\u003e\n\u003cli\u003eSkowronek A, Grajcar A, Kozłowska A, Janik A, Morawiec M, Petrov RH. Temperature-dependent microstructural evolution of Al-rich medium-Mn steel during intercritical annealing. Metall. Mater. Trans. A 2022;53A:3012\u0026ndash;3021, doi:10.1007/s11661-022-06721-2.\u003c/li\u003e\n\u003cli\u003eSkowronek A, Grajcar A, Garcia-Mateo C, Jim\u0026eacute;nez JA Petrov RH. Time-dependent evolution of volume fraction and stability of retained austenite in a hot-rolled and intercritically annealed Al-alloyed medium-Mn steel. Metall. Mater. Trans. A 2023;54:1443\u0026ndash;1454, doi:10.1007/s11661-023-06998-x.\u003c/li\u003e\n\u003cli\u003eMorawiec M, Opara J, Garcia-Mateo C, Jimenez JA, Grajcar A. Effect of Mn on the chemical driving force and bainite transformation kinetics in medium-manganese alloys. Therm. Anal. Calorim. 2023;148:1567-1576, doi.org/10.1007/s10973-022-11664-2.\u003c/li\u003e\n\u003cli\u003eBhowmik N, Ghosh S, Mandal S, Haldar A, Chattopadhyay PP. Genesis of plasticity-induced serrated metal flow in medium-Mn steel. Calphad 2022;77:102425, doi:10.1016/j.calphad.2022.102425.\u003c/li\u003e\n\u003cli\u003eKozłowska A, Grzegorczyk B, Morawiec M, Grajcar A. Explanation of the PLC effect in advanced high-strength medium-Mn steels. A review. Materials 2019;12:4175, doi: 10.3390/ma12244175.\u003c/li\u003e\n\u003cli\u003eMorawiec M, Grajcar A, Zalecki W, Garcia-Mateo C, Opiela M. Dilatometric study of phase transformations in 5Mn steel subjected to different heat treatments. Materials 2020;13:958, doi:10.3390/ma13040958.\u003c/li\u003e\n\u003cli\u003eKwok TWJ, Dye D. A review of the processing, microstructure and property relationships in medium Mn steels. Mater. Rev. 2023;68:1098-1134, doi:10.1080/09506608.2023.2199617.\u003c/li\u003e\n\u003cli\u003ePereloma EV, Gazder AA, Timokhina I. Retained austenite: Transformation-Induced Plasticity. Encyclopedia of Iron, Steel, and Their Alloys. Taylor and Francis, New York 2016, 3088\u0026ndash;3103.\u003c/li\u003e\n\u003cli\u003eKaar S, Krizan D, Schwabe J, Hofmann H, Hebesberger T, Commenda C, Samek L. Influence of the Al and Mn content on the structure-property relationship in density reduced TRIP-assisted sheet steels. Mater. Sci. Eng. A 2018;735:475-486, doi:10.1016/j.msea.2018.08.066.\u003c/li\u003e\n\u003cli\u003eHidalgo J, Celada-Casero C, Santofimia MJ. Fracture mechanisms and microstructure in a medium Mn quenching and partitioning steel exhibiting macrosegregation. Mater. Sci. Eng. A 2019;754:766-777, doi:10.1016/j.msea.2019.03.055.\u003c/li\u003e\n\u003cli\u003eKalsar R, Sanamar S, Schell N, Brokmeier HG, Saha R, Ghosh P, Bhagat AN, Suwas S. Elemental partitioning in medium Mn steel during short-time annealing: An in-situ study using synchrotron X-rays. Materialia, 2020;9:100594, doi:10.1016/j.mtla.2020.100594.\u003c/li\u003e\n\u003cli\u003eKang S, Speer JG, Krizan D, Matlock DK, De Moor E. Prediction of tensile properties of intercritically annealed Al-containing 0.19C\u0026ndash;4.5Mn (wt%) TRIP steels. Mater. Des. 2016;97:138\u0026ndash;146, doi:10.1016/j.matdes.2016.02.058.\u003c/li\u003e\n\u003cli\u003eSrinivas R, Varanasi BG, Ponge D. Effect of Nb micro-alloying on austenite nucleation and growth in a medium manganese steel during intercritical annealing. Acta Mater. 2022;229:117786, doi:10.1016/j.actamat.2022.117786.\u003c/li\u003e\n\u003cli\u003eVaranasi RS, Lipińska-Chwałek M, Mayer J, Gault B, Ponge D. Mechanisms of austenite growth during intercritical annealing in medium manganese steels. Scripta Mater. 2022;206:114228. doi:10.1016/j.scriptamat.2021.114228.\u003c/li\u003e\n\u003cli\u003eMueller, J., Matlock, D., Speer, J., \u0026amp; De Moor, E. (2023). Effect of Mn-Enriched Cementite on Austenite Formation During Intercritical Annealing. \u003cem\u003eInternational Symposium on New Developments in Advanced High-Strength Sheet Steels, AHSS 2023\u003c/em\u003e, 215-223. http://doi.org/10.33313/298/023.\u003c/li\u003e\n\u003cli\u003eHe BB, Pan S, Huang MX. Extra work hardening in room-temperature quenching and partitioning medium Mn steel enabled by intercritical annealing. Mater. Sci. Eng. A 2020;797:140106, doi:10.1016/j.msea.2020.140106.\u003c/li\u003e\n\u003cli\u003eK.D.H. Bhadeshia. Nanostructured bainite. Proceedings of the Royal Society A: Mathematical, 466 (2010): 3\u0026ndash;18. doi:10.1098/rspa.2009.0407.\u003c/li\u003e\n\u003cli\u003eCaballero FG, Miller MK, Garcia-Mateo C. Tracking solute atoms during bainite reaction in a nanocrystalline steel. Mater Sci Technol (2010) 26 889-898, doi:10.1179/026708310X12635619987943.\u003c/li\u003e\n\u003cli\u003eZhou TP, Wang CY, Wang C, Cao WQ, Chen ZJ. Austenite stability and deformation-induced transformation mechanism in cold-rolled medium-Mn steel. Mater. Sci. Eng. A 2020;798:140147, doi:10.1016/j.msea.2020.140147.\u003c/li\u003e\n\u003cli\u003eLukas HL, Fries SG, Sundman B. Computational Thermodynamics: The Calphad Method. Cambridge: Cambridge University Press; 2007.\u003c/li\u003e\n\u003cli\u003eSaunders N, Guo Z, Li X, Miodownik AP, Schill\u0026eacute; J-Ph. Using JMatPro to model materials properties and behavior. JOM 2003;55:60-65, doi:10.1007/s11837-003-0013-2.\u003c/li\u003e\n\u003cli\u003eJ-O Andersson, Thomas Helander, Lars H\u0026ouml;glund, Pingfang Shi, Bo Sundman, Thermo-Calc \u0026amp; DICTRA, computational tools for materials science, CALPHAD 26 (2002) 273-312, https://doi.org/10.1016/S0364-5916(02)00037-8.\u003c/li\u003e\n\u003cli\u003eDykas J, Samek L, Grajcar A, Kozłowska A. Modelling of phase diagrams and continuous cooling transformation diagrams of medium manganese steels. Symmetry 2023;15:381, doi:10.3390/sym15020381.\u003c/li\u003e\n\u003cli\u003eMorawiec M, Opara J, Grajcar A. Nanobainite formation in high-Al medium-Mn steels: thermodynamic approach. J Therm Anal Calorim (2024) 10511-10520, doi:10.1007/s10973-024-13441-9.\u003c/li\u003e\n\u003cli\u003eElaraby M, Ali M, Eissa M, K\u0026ouml;mi J, Karjalainen P, Javaheri V. Computational designing approach for medium manganese steels with potential better hydrogen embrittlement resistance. In: Proceedings of SIMS EUROSIM (2024) 228-235. doi:10.3384/ecp212.032.\u003c/li\u003e\n\u003cli\u003eTian C, Guo H, Hu B, Enomoto M, Shang C. Influence of nano-scale concentration gradient of alloying elements on the ductility in an intercritically annealed and tempered medium Mn steel. Mater. Sci. Eng. A 2021;810:141009, doi:10.1016/j.msea.2021.141009.\u003c/li\u003e\n\u003cli\u003e.Hu B, Zheng Q, Lu Y, Ji C, Liang T, Zheng C, Li D. Stabilizing austenite via intercritical Mn partitioning in a medium Mn steel. Scripta Mater. 2023;225:115162, doi:10.1016/j.scriptamat.2022.115162.\u003c/li\u003e\n\u003cli\u003eGramlich A, Lange R, Zitz U, B\u0026uuml;\u0026szlig;ensch\u0026uuml;tt K. Air-hardening die-forged con-rods\u0026mdash;achievable mechanical properties of bainitic and martensitic concepts. Metals 2022;12:97, doi:10.3390/met12010097.\u003c/li\u003e\n\u003cli\u003eSente software Ltd. (2005) \u0026lsquo;A collection of free downloadable papers on the development and application of JMatPro\u0026rsquo;, http://www.sentesoftware.co.uk/biblio.html.\u003c/li\u003e\n\u003cli\u003eASTM A1033-04, Standard practice for quantitative measurement and reporting of hypoeutectoid carbon and low-alloy steel phase transformations; ASTM International: west Conshohocken.\u003c/li\u003e\n\u003cli\u003eASTM International. ASTM E3-11(2017), Standard Guide for Preparation of Metallographic Specimens. West Conshohocken, PA: ASTM International; 2017.\u003c/li\u003e\n\u003cli\u003eMueller JJ, Hu X, Sun X, Ren Y, Choi K, Barker E, Speer JG, Matlock DK, De Moor E. Austenite formation and cementite dissolution during intercritical annealing of a medium-manganese steel from a martensitic condition. Mat. Des. 2021;203:109598, doi:10.1016/j.matdes.2021.109598.\u003c/li\u003e\n\u003cli\u003eLiu G, Li B, Xu S, Tong S, Wang X, Liang X, Sun X. Effect of intercritical annealing temperature on multiphase microstructure evolution in ultra-low carbon medium manganese steel. Mater. Char. 2021;173: 110920, doi:10.1016/j.matchar.2021.110920.\u003c/li\u003e\n\u003cli\u003eXu Y, Li W, Du H, Jiao H, Liu B, Wu Y, Ding W, Luo Y, Ni Y, Min N, Liu W, Jin X. Tailoring the metastable reversed austenite from metastable Mn-rich carbides. Acta. Mater. 2021;214:116986, doi:10.1016/j.actamat.2021.116986.\u003c/li\u003e\n\u003cli\u003eKamoutsi H, Gioti E, Haidemenopoulos GN. Kinetics of solute partitioning during intercritical annealing of a medium-Mn steel. Metall. Mater. Trans. A 2015;46:4841\u0026ndash;4846, doi:10.1007/s11661-015-3118-7.\u003c/li\u003e\n\u003cli\u003eKalsar R, Sanamar S, Schell N, Brokmeier HG, Saha R, Ghosh P, Bhagat AN, Suwas S. Elemental partitioning in medium Mn steel during short-time annealing: An in-situ study using synchrotron X-rays. Materialia, 2020;9:100594, doi:10.1016/j.mtla.2020.100594.\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"continuum-mechanics-and-thermodynamics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"cmat","sideBox":"Learn more about [Continuum Mechanics and Thermodynamics](http://link.springer.com/journal/161)","snPcode":"161","submissionUrl":"https://submission.nature.com/new-submission/161/3","title":"Continuum Mechanics and Thermodynamics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"medium-Mn steel, thermodynamic modelling, kinetic modelling, intercritical annealing, retained austenite, thermodynamic stability, element partitioning","lastPublishedDoi":"10.21203/rs.3.rs-6319163/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6319163/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The effects of intercritical annealing (IA) temperature and manganese content on austenite formation and stability in medium-Mn steels were investigated using combined computational and experimental approaches. Three steels containing 3, 4, and 5 wt.% Mn were annealed at 700–760 °C for 60 min, and their microstructures were analysed to assess the influence of Mn content on phase transformation behaviour. Thermodynamic (JMatPro, Thermo-Calc) and kinetic (DICTRA) simulations were used to model phase stability, austenite growth, and elemental partitioning under both equilibrium and non-equilibrium conditions. The modelling results were validated using dilatometry, scanning electron microscopy (SEM), and energy-dispersive X-ray spectroscopy (EDS). The results showed that increasing Mn content promotes higher austenite fractions during IA but reduces its carbon enrichment, which adversely affects thermal stability due to higher martensite start (Ms) temperatures. DICTRA simulations also revealed that Mn and Al develop distinct concentration gradients across the ferrite/austenite interface, especially in higher-Mn steels, and that these gradients correlate with the measured EDS values in retained austenite laths. The revealed microstructures were composed of ferrite, retained austenite, and small amounts of cementite and martensite-austenite constituents. Overall, the study demonstrates that Mn content strongly affects both the amount and stability of retained austenite, and that IA parameters must be carefully optimized to tailor microstructure and mechanical behaviour in medium-Mn advanced high-strength steels.","manuscriptTitle":"Effect of Intercritical Annealing Temperature on Austenite Formation in Medium-Mn Steels: A Thermodynamic and Experimental Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-22 06:57:42","doi":"10.21203/rs.3.rs-6319163/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-05T09:50:25+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-04T12:20:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"23132494054663753296947047222272027285","date":"2025-04-24T13:14:38+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"294706982681127080465379360971958053127","date":"2025-04-23T01:04:44+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-04-01T11:53:31+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-04-01T11:39:40+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-04-01T11:36:02+00:00","index":"","fulltext":""},{"type":"submitted","content":"Continuum Mechanics and Thermodynamics","date":"2025-03-27T09:35:24+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"continuum-mechanics-and-thermodynamics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"cmat","sideBox":"Learn more about [Continuum Mechanics and Thermodynamics](http://link.springer.com/journal/161)","snPcode":"161","submissionUrl":"https://submission.nature.com/new-submission/161/3","title":"Continuum Mechanics and Thermodynamics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"d7c70824-f21d-49f5-8e78-19a1165ab221","owner":[],"postedDate":"April 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-03-21T20:38:53+00:00","versionOfRecord":[],"versionCreatedAt":"2025-04-22 06:57:42","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6319163","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6319163","identity":"rs-6319163","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}