Sie können Operatoren mit Ihrer Suchanfrage kombinieren, um diese noch präziser einzugrenzen. Klicken Sie auf den Suchoperator, um eine Erklärung seiner Funktionsweise anzuzeigen.
Findet Dokumente, in denen beide Begriffe in beliebiger Reihenfolge innerhalb von maximal n Worten zueinander stehen. Empfehlung: Wählen Sie zwischen 15 und 30 als maximale Wortanzahl (z.B. NEAR(hybrid, antrieb, 20)).
Findet Dokumente, in denen der Begriff in Wortvarianten vorkommt, wobei diese VOR, HINTER oder VOR und HINTER dem Suchbegriff anschließen können (z.B., leichtbau*, *leichtbau, *leichtbau*).
Diese Studie untersucht die Auswirkungen von Warmverformungen in mehreren Durchgängen auf einen mikrolegierten Stahl mit mittlerem Kohlenstoffanteil und konzentriert sich dabei auf Schlüsselthemen wie thermomechanische Verarbeitung, Mikrostrukturentwicklung und die Auswirkungen verschiedener Verarbeitungsvariablen. Die Forschung untersucht, wie Dehnung, Verformungstemperatur, Kühlraten, Einweichzeit und Übergangszeit die Entwicklung ultrafeinkörniger (UFG) und kugelförmiger Mikrostrukturen beeinflussen. Die Studie nutzt fortgeschrittene Techniken wie Elektronenrückstreuung (EBSD) und Feldmesskanonen-Rasterelektronenmikroskopie (FEG-REM), um mikrostrukturelle Veränderungen zu analysieren. Die Forschungsergebnisse zeigen, dass eine warme Deformation die dynamische Sphäroidisierung von Perlstein aktiviert und eine kontinuierliche dynamische Rekristallisation in Ferrit fördert. Die Kinetik der Sphäroidisierung wird anhand einer Avrami-Gleichung beschrieben, wobei eine geschätzte Aktivierungsenergie die Rolle von Leerständen bei der Beschleunigung des Prozesses anzeigt. Die mikrostrukturelle Evolution während der Kühl- und Einweichzeit zeigt, dass angesammelte Belastungsenergie statische metallurgische Phänomene antreibt und zu raffinierten Mikrostrukturen mit feinen, globulisierten Zementpartikeln führt. Die Studie kommt zu dem Schluss, dass eine Steuerung der Interpasszeit die kontinuierliche Rekristallisation optimieren und zugleich die Ostwald-Reifung minimieren kann. Mit einer flächengewichteten mittleren Korngröße von 0,8 um m bei 700 ° C werden hervorragende Ergebnisse erzielt. Diese umfassende Analyse liefert wertvolle Einblicke in die Optimierung industrieller Prozesse zur Entwicklung fortschrittlicher Stahlmikrostrukturen.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
The microstructure development of 55VNb1 microalloyed steel after warm deformation via multi-pass biaxial compression tests was studied, and the effect of thermomechanical conditions on spheroidisation of cementite lamellae and ferrite recrystallisation for a range of deformation temperatures (600–700 °C), cooling/soaking time (water quenching, air cooling, 10 and 30 min of soaking time) and interpass time (0–10 s) was analysed. During deformation, the spheroidisation of pearlite is dynamically accelerated mainly by boundary splitting mechanism together with the rapid dissolution of cementite, while ferrite softening is attributed to dynamic recovery and continuous dynamic recrystallisation. The strong microstructural evolution during cooling/soaking time indicates that deformation energy accumulated is sufficient to activate metallurgical phenomena in both phases also statically. Static spheroidisation is a diffusive process, with rate controlled by the diffusion of vacancies, as suggested by the estimated activation energy. Ferrite refinement is the result of the evolution of continuous recrystallisation and pinning effect exerted by fine, globulised and homogeneously dispersed cementite particles. Increasing temperature causes accelerated kinetics in metallurgical phenomena; therefore, cooling/soaking time becomes key parameters to achieve ultrafine grained and spheroidised microstructures. Interpass time favours spheroidisation and promotes continuous recrystallisation; however, it must be carefully controlled to find a balance between recrystallisation and Ostwald ripening to optimise microstructural development.
1 Introduction
In order to meet the increasing needs of the automotive sector, the research on new-generation steels with improved mechanical properties has become a worldwide issue. Different mechanisms enable an increase in yield strength, but among them, the refinement of grain size stands out as the only mechanism that allows an increase in strength without compromising toughness. For this reason, new thermomechanical processes have been investigated and developed in recent decades to produce steels with finer grain sizes. Grain size refinement below 10 µm is very effective as a hardening mechanism; however, when grain size is below 1 µm, work hardening is limited, and elongation deteriorates [1, 2]. For this reason, some studies have focused on developing steels with sizes between 2 and 5 µm, with the aim of achieving a good compromise between strength and ductility [3‐5]. However, when a further increase in strength is required, it becomes necessary to further refine the grain, and it is in this context that ultrafine grained (UFG) steels emerge [6‐10].
In the case of medium and high C steels, the lamellar morphology of the pearlite, which forms after conventional hot rolling, limits the formability and ductility of the material. For this reason, it has traditionally been necessary to subject them to costly soft annealing treatments (also known as spheroidisation treatments) in which the lamellar structure of the pearlite decomposes into globulised carbides [11, 12]. Thus, after treatment, microstructure is comprised by a continuous matrix of ductile ferrite with uniformly dispersed cementite particles which altogether offer a lower resistance to deformation [13, 14]. However, on an industrial scale, these treatments consume large amounts of energy and time and make the product irreversibly more expensive. For these reasons, new strategies with the aim of reducing or even eliminating these costly treatments have been proposed [15‐17].
Anzeige
Therefore, for medium and high C steels, those technologies that simultaneously promote grain refinement and pearlite spheroidisation appear to be of particular interest. There are currently different methods that promote these microstructural changes which can be classified into severe plastic deformation (SPD) and advanced thermomechanical processes (ATP) [18]. It should be noted, however, that SPD methods generally require the accumulation of very large plastic deformations as well as complex designs in the deformation sequences [8, 19, 20], making their application to large-scale production complex and demanding large investments. ATP, on the other hand, employ lower strain accumulations (in the range of 1.0–3.6) to produce UFG steels, and although less effective than SPD methods from a grain refinement point of view, they might be more adaptable to current facilities. Within this family of processes, which includes deformation induced transformation (DIT) [9, 21] or intercritical rolling [21], warm rolling [7, 22, 23] stands out for its potential industrial application by means of multi-pass warm caliber rolling (WCR) [24]. This technology is characterised by applying relatively severe deformations below Ae1 temperature, in the ferritic–pearlitic range, through multiple passes with moderate levels of deformation.
The effect of warm deformation on low C steels has been studied by different scholars [6, 7, 24‐26]. However, there is a limited number of studies in the case of medium or high C steels [27‐30]. In this context, the present study evaluates the effect of multi-pass warm deformation on a medium C microalloyed steel, with the aim of understanding the effect of thermomechanical variables such as strain, deformation temperature, cooling rate, soaking time and interpass time on microstructural development. The ultimate goal is to identify the deformation conditions suitable for producing UFG and spheroidised microstructures. This identification will enable the definition of appropriate rolling conditions to be applied in future WCR trials.
2 Experimental procedure
A hot-rolled 55VNb1 steel bar with a diameter of 90 mm was industrially produced with the following composition (wt.%): C 0.53, Mn 0.84, Si 0.32, V 0.13 and Nb 0.02. Compression specimens were machined out from the centre of this bar. Multi-pass biaxial compression tests were carried out on a Gleeble HDS-V40 thermal mechanical simulator testing system. The samples were heated to 20 °C above the test temperature (indicated as Ttest + 20 °C in Fig. 1) at a rate of 1 °C/s and held for 180 s. Subsequently, the specimens were cooled down to the test temperature (600–700 °C) at a rate of 1 °C/s and held for 60 s. Then, the specimens were subjected to 14 compression passes, applying a strain per pass εp of 0.2 and rotating the samples 90° after each pass, applying a total strain εT of 2.8. The selected strain rate (\(\dot{\varepsilon }\)) was 1 s–1. After the deformation process, the samples were cooled down using different cooling rates such as water quenching (WQ) or air cooling (AC), or different soaking time (10 or 30 min) was applied prior to air cooling. Additionally, 10 s of interpass time (IT) between each pass has been included in some cycles. A scheme of the thermomechanical schedule along with the different parameters used is shown in Fig. 1.
Biaxial compression tests lead to a non-uniform distribution of the resulting deformation across the section of the sample. Smaller deformation accumulates below the area of contact with the tools, while the central region may accumulate significantly higher local strain levels than the nominal strain applied [31].
Anzeige
Tested samples were sectioned along their longitudinal axis for microstructural analysis. Optical microscopy and field emission gun scanning electron microscopy (FEG-SEM) were applied after conventional polishing and etching with 2% Nital. The ASTM E-562–89 standard was applied to calculate the volume fraction of ferrite. The procedure is based on a manual point-count method, where the volume fraction of ferrite is estimated as the average percentage of points in a regular gird identified as ferrite in optical micrographs. Additionally, the ASTM E-112–96 standard was used to determine the mean interlamellar spacing in FEG-SEM micrographs as the reciprocal of NL, i.e., the number of intersections of cementite lamellae with the circumference of a circular network. Particle area and spheroidisation degree have been analysed by image analysis of randomly selected FEG-SEM micrographs. For each condition, between 5 and 8 random view fields were selected to ensure a representative sample, analysing more than 103 particles per condition. Following the O’Brien and Hosford criteria [32], particles with an aspect ratio lower than 3 have been considered to be spheroidised. Thus, the degree of spheroidisation has been calculated as the ratio between the spheroidised cementite area and the total cementite area (spheroidised and non-spheroidised).
Warm deformation usually produces a substructure that is difficult to resolve by optical or electron microscopy; therefore, electron backscattered diffraction (EBSD) techniques were applied. The samples were prepared using conventional metallographic methods and a final soft polishing with colloidal silica. Areas of 50 µm × 50 µm were scanned using a 0.1 µm step size. AZtecCrystal software package was used to analyse the EBSD scan data acquired. Image quality maps, grain boundary maps with different tolerance angles and inverse pole figures were obtained. EBSD measures also allowed for quantitative studies. Thus, the boundary length density, defined as the boundary line length per unit area (in µm−1), was determined. Determining the grain size in warm deformed materials is often not straightforward because the recrystallisation process responsible for grain size refinement, usually identified as continuous recrystallisation, often leads to grains that are not completely surrounded by high angle grain boundaries (> 15°) [33]. Taking this into account, grain size has been defined by the misorientation criteria above 15°, but allowing grain completion when the misorientation is above 5°. Following this criterion, grain size distributions in terms of cumulative area fractions and in terms of area fractions for the ultrafine grains (< 2 µm) were determined. As shown in a previous work [34], warm deformation can lead to very heterogeneous microstructures, in which very fine grains develop in some regions, while in others, only a limited substructure is generated in the deformed grains. For this reason, the grain size was weighted by area, to properly consider the influence of the coarser grains [35, 36]. Thus, the grain size has been reported as area-weighted mean grain size, \(\overline{{D_\text{w} }}\), which can be defined as:
where Di is the equivalent diameter of grain i; Ai is the area of grain i; and N is the total number of grains. The equivalent diameter is defined as the diameter of a circle that has the same area as the actual grain.
3 Results and discussion
3.1 Starting material
The microstructure at the centre of the industrially produced bar is shown in Fig. 2a, b. Ferrite grains can be observed decorating prior austenite grain boundaries and also as idiomorphic ferrite, nucleated intragranularly on inclusions. Quantitative metallographic measurements indicate that ferrite volume fraction is approximately 13.5% ± 1.7%, while the rest of the microstructure comprises pearlite colonies. Measurements also indicate that the interlamellar spacing in the pearlite is around 0.22 ± 0.03 μm. The EBSD images (Fig. 3) show ferrite grains around pearlite colonies, with diverse orientations. The cementite lamellae in the pearlite are not well defined; however, they can be distinguished on the image quality (IQ) map in Fig. 3a. Both, ferrite and pearlite colonies, appear to be surrounded by high angle grain boundaries (HAGBs, misorientation > 15°), as shown in Fig. 3b, although in some grains, low angle grain boundaries (LAGBs, misorientation between 2° and 15°) can be observed, probably associated with the pearlite colonies. Figure 3c shows the orientation map, in which the misorientation pattern within the pearlite can be observed more clearly. As a result, evidence has been found that some LAGBs are already present in the pearlite microstructure prior to any deformation being applied.
Fig. 2
Optical image (a) and FEG-SEM image (b) of ferrite–pearlite microstructure at centre of hot rolled bar
EBSD images of microstructure at centre of hot rolled bar. a IQ map; b low (2°–15°, red) and high (> 15°, black) grain boundary map; c inverse pole figure
The application of multi-pass biaxial compression tests results in a heterogeneous strain distribution which is shown in the scheme in Fig. 4a and in the optical micrograph in Fig. 4b. Two different zones can be identified. The first one is located at the centre of the compressed specimen where accumulated strain is maximum (high strain region (HSR)). As shown in Fig. 4c, the microstructure in this region is elongated in the direction perpendicular to the applied stress. The second zone is located below the tools where strain accumulation is minimum (low strain regions (LSR)), and as a consequence, microstructure remains almost equiaxed (Fig. 4d). Microstructural characterization has been carried out in the HSR, where major microstructural development is expected in both ferritic and pearlitic zones.
Fig. 4
Scheme (a) and optical micrograph (b) showing strain distribution after multi-pass biaxial deformation and detail of corresponding HSR microstructure (c) and LSR microstructure (d) for sample 600-WQ
At the mesoscopic level, strain application causes the pearlite colonies and the ferrite grains to elongate in the direction perpendicular to the compression axis. Figure 5a, b shows the features developed in both phases using a combination of FEG-SEM images and grain boundary maps obtained by means of EBSD in the water quenched samples after deformation at 600 and 700 °C. As can be seen, both phases are affected by the severe strain imposed, and important changes take place, although microstructures are far from being homogeneous, and former pearlite colonies and ferrite regions are distinguishable in the material (see ferrite regions identified in the FEG-SEM images).
Fig. 5
FEG-SEM micrographs and grain boundary maps obtained in HSR for samples 600-WQ (a) and 700-WQ (b)
The application of warm deformation affects the morphology of the cementite lamellae due to the high strain accumulated in the region. This strain is sufficient to activate the dynamic spheroidisation of the cementite lamellae irrespectively of the deformation being applied at 700 or 600 °C, although to a different degree and extent depending on the observed region. The heterogeneities developed have been related to the different alignment of the colonies with respect to the applied strain direction, which affects the localised deformation and provides different driving force for the spheroidisation [27, 30]. Therefore, in some regions, the lamellae lose their straight laminar morphology and become kinked and partially or completely fragmented, whereas in others, dynamic spheroidisation progresses to a greater extent, resulting a microstructure comprised of fine and globulised particles.
The spheroidisation of pearlite is a direct consequence of the topologically unstable character of the cementite lamellae. The process is driven by the Gibbs-Thompson effect, which is responsible for the continuous mass transport between the high curvature and flat areas of the lamellae. In non-deformed samples, these sources of curvature are related to the presence of crystalline defects and imperfections in the lamellae after the phase transformation. Defects such as termination of lamellae or holes provide sites of high curvature where spheroidisation is initiated, until lamellar splitting occurs, following the fault migration theory [37]. However, these processes are slow and consequently conventional spheroidisation treatments consume significant amounts of energy and time [32].
The application of warm deformation promotes both lamellae fragmentation and development of kinks and shears in the lamellae, which constitute additional high curvature points for the dynamic spheroidisation process to take place. Figure 6a shows a region where several deformation bands are formed contributing to the break-up of the lamellae [38]. A detail of the spheroidisation process taking place in these very distorted platelets is shown in Fig. 6b. These highly kinked lamellae are more prone to spheroidisation as kinks provide high curvature areas for the start of the mass transport process responsible of spheroidisation, and some authors suggested that dynamic spheroidisation occurs as a result of the shearing of lamellae followed by penetration of the ferrite, so as to complete fragmentation and subsequent rounding [39].
Fig. 6
Examples of deformation bands (a) and detail of spheroidisation process (b) taking place in very kinked and distorted platelets
Figure 7a shows an area of lamellar pearlite almost perpendicular to the observation plane with clear indices of the first stages of globulisation. At certain points, some grooves appear on the cementite plates, leading to the formation of necks (see arrows in the image) prior to the separation of cementite as globular particles. As a result of the splitting process, isolated particles can also be observed in the same micrograph. Additionally, Fig. 7b shows some fragmented plates inclined with respect to the observation plane. It can be observed that plates are transformed into a complete pattern of subgrains, with sizes ranging from about 50 up to 200 nm. Thus, the recovery process proceeds dynamically along the lamellae, resulting in the lamellae being transformed into a complete pattern of equiaxed crystallites. At triple points between cementite sub-boundaries and ferrite, the establishment of the local equilibrium of the surface tension involves the formation of grooves along the subgrains, leading to the necking. The continuous mass flux away from the grooves leads finally to the final lamella splitting. Thus, sub-boundary formation greatly accelerates lamellae fragmentation, which is consistent with the thermal groove theory [40], suggesting that this is the main spheroidisation mechanism in warm deformed samples. These sub-boundaries are introduced into the lamellae by deformation and posterior recovery. The process initiates at the intersection between sub-boundaries and ferrite matrix by forming grooves along the sub-boundaries. Owing to the curvature difference, a continuous mass flux from the triple points to the adjacent areas in the lamella takes place, which finally provokes cementite splitting along the channels. In the model proposed by Kampe et al. [40], this process results in elongated cementite shapes (essentially constituted by cylindrical shapes) which are unstable against Rayleigh instability phenomenon [37], so finally split into globular shapes.
Fig. 7
Examples of grooved lamellae in lamellar pearlite almost perpendicular to observation plane (a) and fragmented plates inclined with respect to observation plane (b)
Additionally, dark contrast crystallites can be observed at some points in the plates (Fig. 7b), indicating that the dissolution of cementite is an important mechanism leading to globulisation. FEG-SEM images suggest that cementite dissolution takes place dynamically, while deformation is applied. The rapid dynamic spheroidisation is related to the excess of dislocations generated during plastic deformation, and as a result, the effective diffusion coefficient of carbon might be increased through dislocation core [39]. Transmission electron microscopy (TEM) work on hot deformed pearlite has revealed the presence of dislocations in ferrite and at the cementite/ferrite interphase [11]. This increase in the dislocation density coupled with the relatively high temperature applied during warm deformation enhances cementite dissolution in the matrix.
The effect of deformation on the ferrite microstructure is clearly illustrated when EBSD images of the starting material in Fig. 3 are compared with those in Fig. 5a, b corresponding to the HSR of the water quenched samples deformed at 600 and 700 °C, respectively. Grain boundary maps show that the application of warm deformation leads to a significant increase in LAGBs and HAGBs, although to a lesser extent in the latter case. Ferrite has a high stacking fault energy; therefore, dislocation rearrangement takes place readily all over in this phase, and, as a consequence, dynamic recovery dominates the restoration process during deformation application. As a result, LAGBs extensively forms throughout the material, forming a very fine substructure. A significant increase in HAGBs can also be observed, although their development is more heterogeneous and less extensive. Consequently, in some regions, coarse grains can be observed, highly elongated in the direction perpendicular to the applied stress, whereas in others, very fine grains are developed. Results obtained in a previous work on the same material for samples deformed by single-pass uniaxial compression tests (ɛT = 1) [34] showed that the evolution of the ferrite was highly dependent on the region under study (ferrite or pearlite) and the degree of spheroidisation of the cementite lamellae. Moreover, plastic strain partitioning is expected to take place between the different phases (ferrite and pearlite), and between ferrite and cementite lamellae inside pearlite colonies.
3.3 Microstructure evolution during cooling and soaking time
The microstructure evolution when applying different cooling rates and soaking time after warm deformation process at 600 °C is shown in Fig. 8. The evolution of the ferrite matrix and cementite spheroidisation has been examined using a combination of FEG-SEM images and grain boundary maps obtained by means of EBSD.
Fig. 8
Microstructure evolution after application of air cooling (600-AC) (a), 10 min of soaking time (600-10min-AC) (b), 30 min of soaking time (600-30min-AC) (c) and 10 s of interpass time (600-IT10s-AC) (d) in samples deformed at 600 °C
FEG-SEM micrographs show that the application of slower cooling rates and soaking time allows the spheroidisation process to evolve in the material; as a result, after 10 min of soaking time, spheroidisation seems to be completed. This indicates that spheroidisation takes place also statically after the deformation process. Increasing soaking time to 30 min favours the homogenisation of the microstructure, so that the differences between pearlitic and ferritic regions are no longer evident. Once the spheroidisation process is completed, the soaking time at this temperature allows the finer particles to dissolve and the carbon in the matrix to diffuse into the ferritic zones with a lower carbon concentration where it reprecipitates. Dissolved carbon tends to diffuse towards the ferrite driven by the compositional gradient (carbon concentration is higher in the former colonies compared to the proeutectoid ferrite) and assisted by diffusion through the boundaries and sub-boundaries developed as a consequence of the softening process taking place in the ferrite, so that finally carbon precipitates at ferrite [27].
Grain boundary maps show that softening processes also evolve in the ferrite during cooling and soaking time. The fraction of new ultrafine grains (< 2 µm) tends to increase as slower cooling rates and soaking time are applied, so that a very fine and rather homogeneous microstructure is developed after 30 min of soaking. This indicates that continuous recrystallisation takes place statically. Its evolution is intrinsically related to the spheroidisation process of pearlite, which also progresses in the material [41], as a consequence of the Zenner pinning effect exerted by finer particles. These particles generate a significant dragging force that impedes the migration of grain boundaries and sub-boundaries, thereby promoting the development of an ultrafine microstructure.
Microstructural evolution suggests that the strain energy accumulated during multi-pass biaxial compression tests in form of defects in both ferrite and cementite lamellae is sufficient to activate not only the spheroidisation and recovery/recrystallisation processes dynamically, but also statically. This indicates that under these conditions, at 600 °C, the soaking time is a key parameter to obtain ultrafine microstructures, with fully spheroidised, fine and homogeneously dispersed cementite particles.
Additionally, microstructure evolution when warm deformation process is conducted at 700 °C is also shown in Fig. 9.
Fig. 9
Microstructure evolution after application of air cooling (700-AC) (a), 10 min of soaking time (700-10min-AC) (b) and 10 s of interpass time (700-IT10s-AC) (c) in samples deformed at 700 °C
At 700 °C, images show the application of air cooling is enough for the spheroidisation process to complete and differences between pearlitic and ferritic regions are no longer evident in the material. The grain boundary map also shows that ferrite evolves very significantly under these conditions, developing a high density of HAGBs. As a result, after air cooling, the microstructure comprises very fine and homogeneously dispersed particles together with a homogeneous and very fine structure of grains and subgrains. When subjected to 10 min of soaking at 700 °C, a highly significant evolution also occurs in the microstructure. FEG-SEM image reveals that the spheroidised cementite particles coarsen driven by the Ostwald ripening process, and both FEG-SEM image and grain boundary map show recrystallisation and grain growth have taken place in the material. Thus, as Ostwald ripening process advances, the pinning effect exerted by the fine cementites decreases and the grain boundaries after recrystallisation are able to migrate. The strong microstructural evolution that takes place at 700 °C indicates that the higher applied temperature together with the elevated stored deformation energy during the deformation process causes accelerated kinetics in metallurgical phenomena.
The evolution of the mean particle area and the degree of spheroidisation as a function of the thermomechanical conditions and cooling rates/soaking time applied are shown in Fig. 10a, b.
Fig. 10
Evolution of mean particle area (a) and degree of spheroidisation (b) as a function of thermomechanical conditions and cooling rates/soaking time applied
Same trends are found at 600 and 700 °C for the evolution of both mean particle area and degree of spheroidisation. After air cooling, particle area tends to decrease when comparing with water quenching condition. This behaviour is related to the degree of spheroidisation, since a significant evolution in the spheroidisation process takes place during air cooling (from 26% to 78% at 600 °C and from 42% to 96% at 700 °C). The decrease in particle size is therefore related to the rapid spheroidisation process, which results in smaller particles as a consequence of the fragmentation and spheroidisation of the lamellae. During soaking time, spheroidisation process is completed (after 10 min of soaking, the degree of spheroidisation is 96% at 600 °C and 98% at 700 °C), and particle area tends to increase, slightly at 600 °C and significantly at 700 °C, as a consequence of Ostwald ripening process. Spheroidisation is considered a diffusion-controlled process, so the application of higher temperatures accelerates it, as observed in the figures.
Nijhof [42] proposed an Avrami equation to describe the kinetics of the spheroidisation phenomenon:
where α is the degree of spheroidisation; t is the time; and n, k are equation parameters. Based on this type of model, other authors have successfully described spheroidisation kinetics [43].
A linear relationship between \({\text{ln}(\text{ln}(1-\alpha )}^{-1})\) and \(\text{ln}t\) is found both at 600 and 700 °C, which indicates that the Eq. (2) is satisfied by the experimental data, so the parameters n and k have been determined by the least square method. Results are presented in Table 1.
Table 1
Results of parameters n and k
T/°C
600
700
n
0.351
0.314
k
0.295
0.558
Parameter k can be considered as a constant rate of spheroidisation and can be described by an Arrhenius-type equation:
$$k = A {\text{exp}}\left( { - \frac{Q}{RT}} \right)$$
(3)
where A is a pre-exponential factor; Q is the activation energy; R is the Boltzmann constant; and T is the absolute temperature. Hence, the activation energy Q has been calculated from the temperature dependence of the constant k using Eq. (3), resulting in Q = 45.0 kJ/mol, which is in the order of the interaction energy between carbon atoms and vacancies [44]. According to Chattopadhyay and Sellars [11], the major cause of accelerated spheroidisation during deformation is the development of an excess of vacancies, which directly increases the self-diffusion rate of iron and indirectly increases the flux of carbon to form particles by the formation and diffusion of carbon–vacancy complexes.
Figure 11a, b shows the quantification of LAGB and HAGB densities after different cooling strategies and soaking time applied at 600 and 700 °C. The quantification should always be analysed together with the microstructure images for a correct interpretation of the results. After water quenching, the LAGB and HAGB densities increase significantly at both temperatures, especially at 600 °C. At both temperatures, a higher fraction of LAGBs (accounting for 62% of the total boundary density) is obtained, suggesting that dynamic recovery is the main softening mechanism taking place during deformation. The increase in HAGB density together with the presence of very fine grains observed in the images has been related to the local activation of continuous dynamic recrystallisation [33]. This process enables the progressive transformation of LAGBs formed in the original deformed grains into HAGBs by the accumulation of dislocations at the LAGBs, either by the merging of lower angle boundaries during subgrain coalescence or by the subgrain growth with migration of low angle boundaries [27].
Fig. 11
Evolution of low and high angle grain boundary densities together with percentage of total they represent for samples tested at 600 °C (a) and 700 °C (b)
The analysis of the evolution of LAGB and HAGB densities under slower cooling rates or soaking time at 600 °C (Fig. 11a) shows that LAGB density tends to decrease (both in absolute value and as a percentage of the total boundary density), whereas HAGB density tends to increase. This behaviour is attributed to the activation of continuous static recrystallisation, which facilitates the transformation of LAGBs into HAGBs, while spheroidised particles control grain growth. As a result, an ultrafine microstructure with 60% HAGBs is developed after 30 min of soaking time.
At 700 °C (see Fig. 11b), metallurgical phenomena are accelerated, and as a consequence, after air cooling, an ultrafine microstructure with 63% HAGBs is developed. After 10 min of soaking, a significant decrease in the absolute value of both LAGB and HAGB densities is observed as a result of the grain growth that has taken place in the material after recrystallisation.
The distributions of ferrite grain size in terms of cumulative area fraction for samples deformed at 600 °C are shown in Fig. 12a. This type of plot represents the fraction of area (y-axis) covered by grains with a diameter smaller than a certain value (x-axis). It can be observed that curves shift to the left as slower cooling rates and soaking time are applied, indicating that larger area fractions are covered by finer grains as the microstructure evolves. Refinement is the result of the evolution of the continuous static recrystallisation and static spheroidisation processes. Thus, LAGBs are transformed into HAGBs, while fine and homogeneously dispersed cementite particles pin the boundaries and sub-boundaries, resulting in a microstructure with an area-weighted mean size of 2.1 µm, where 64% of the surface is covered by grains smaller than 2 µm after 30 min of soaking time (see Table 2). For comparison, results obtained in a previous work on the same material [34] for samples deformed by single-pass uniaxial compression tests (ɛT = 1) are also shown in Fig. 12. It is noteworthy that results obtained after the deformation process on water quenched samples show similar results after single-pass (ɛT = 1) and multi-pass (ɛT = 2.8) compression tests. However, microstructural evolution strongly depends on the accumulated deformation, so that no clear change is observed in the grain size distribution even when 30 min of soaking time is applied after the application of moderate deformation (ɛT = 1), while accumulated area distributions move to the left (towards finer grain sizes) during the application of cooling and soaking time after multi-pass deformation process (ɛT = 2.8). This indicates that the accumulated deformation energy is the driving force for metallurgical phenomena to take place [7].
Fig. 12
Grain size distributions in terms of cumulative area fraction for samples tested at 600 °C (a) and 700 °C (c) and detail of grain size distributions in terms of area fraction for ultrafine grains (< 2 µm) for samples tested at 600 °C (b) and 700 °C (d). Misorientation criteria > 15° (closing open grains with misorientation > 5°)
Grain size distributions in terms of area fraction for ultrafine grains (< 2 µm) are also shown in Fig. 12b. Area fraction of ultrafine grains tends to increase while the grain size remains stable, indicating that the pinning effect exerted by the finer cementite particles at 600 °C prevents boundaries and sub-boundaries to migrate. Only a slight growth in size is observed after 30 min of soaking time. For comparison, the results obtained after single-pass compression tests (ɛT = 1) show a lower development of area fraction of ultrafine grains which does not increase even after 30 min of soaking time, confirming that continuous static recrystallisation is not extensively activated after single-pass compression tests (ɛT = 1) in this material.
At 700 °C, Fig. 12c shows that a significant refinement takes place after air cooling. The same metallurgical phenomena occur as at 600 °C, but at a faster rate due to the higher temperature, resulting in increased kinetics. As a result, 71% of the surface is covered by grains smaller than 2 µm with an area-weighted mean size of 2.2 µm (see Table 2). Figure 12d also indicates that the increase in the ultrafine grain area fraction is accompanied by a refinement of the ultrafine grains as a consequence of the extensive activation of the continuous static recrystallisation and the pinning effect exerted by the cementite particles under these conditions. During soaking for 10 min, the large amount of deformation energy accumulated in the form of defects together with the enhanced diffusion accelerates particle coarsening via Ostwald ripening, which, in turn, facilitates boundary migration. As a result, ultrafine grains tend to coarsen, obtaining an area-weighted mean size of 3.8 µm where only 13% of the surface is covered by grains smaller than 2 µm. When comparing the results of the multi-pass compression tests with those of the single-pass compression tests, the same trends are obtained as at 600 °C. This indicates that the accumulated deformation energy is a key variable for microstructural evolution to take place, while temperature and time are controlling parameters for ultrafine grain size to be developed.
Table 2 shows a summary of the most relevant results obtained in the EBSD measurements. These results comprise the fraction of area covered by ultrafine grains (< 2 µm), the average grain size of ultrafine grains, the area-weighted mean size (Eq. (1)), and the HAGB and LAGB densities.
3.4 Effect of interpass time on microstructure evolution
To evaluate the effect of the interpass time, the thermomechanical cycle was modified to include 10 s between each of the 14 passes, after which the samples were air cooled.
The effect of interpass time on the microstructure evolution when deformation cycle is applied at 600 °C can be observed by comparing Fig. 8a, d. FEG-SEM images show that including 10 s between each pass favours the spheroidisation of pearlite. Figure 10b confirms that the spheroidisation process is completed under these deformation conditions (600-IT10s-AC), reaching a 98% degree of spheroidisation while the mean particle area slightly coarsens when compared with the sample air cooled without interpass time, as can be seen in Fig. 10a. Moreover, grain boundary maps (Fig. 8a, d) show that the addition of the interpass time has a significant effect also on the evolution of the ferrite, extensively activating continuous recrystallisation in the material. As a result, Fig. 11a shows an important increase in HAGB density, which reaches 64%. This allows ultrafine grains to be formed extensively in the material, as can be observed in Fig. 12a, b. The accumulated area distribution moves to the left (towards finer grain sizes), resulting in a microstructure with an area-weighted mean size of 1.5 µm, where 76% of the surface is covered by grains smaller than 2 µm (see Table 2).
At 700 °C, interpass time effect can be studied by comparing Fig. 9a, c. FEG-SEM images show that spheroidisation process is completed in both cases while the mean particle area slightly coarsens when applying the extra 10 s between each pass. The quantification in Fig. 10a shows that mean particle area increases from 0.014 to 0.018 µm2, while the degree of spheroidisation is above 95% in both cases (Fig. 10b). Besides, grain boundary maps (Fig. 9a, c) show that the time between passes allows continuous recrystallisation to take place more extensively in the material, transforming LAGBs into HAGBs. As a result, Fig. 11b shows an important increase in HAGB density, which reaches 84%. In addition, spheroidised cementite particles still exert a significant pinning effect, controlling grain size and facilitating the achievement of an ultrafine microstructure, as can be observed in Fig. 12c, d. The accumulated area distribution move to the left (towards finer grain sizes), resulting in a microstructure with an area-weighted mean size of 0.8 µm, where all the surface is covered by grains smaller than 2 µm (see Table 2).
The results obtained indicate that the application of 10 s between the 14 passes not only increases the time the samples are heated by 130 s, but also provides time for the diffusive phenomena to occur favoured by the defects introduced after each pass. As a result, spheroidisation occurs faster and more extensively in the material. The interval also affects recovery and continuous recrystallisation, because the extra time between passes allows dislocations generated during deformation to rearrange. Thus, during the first few passes, the substructure is formed due to the rearrangement of dislocations, and as the deformation increases with subsequent passes, the extra time facilitates the dislocations introduced into the material to become trapped in the sub-boundaries already formed in the previous deformation stages, increasing the level of misorientation, until the LAGBs are transformed into HAGBs, following continuous recrystallisation. Li et al. [45] showed that increasing interpass time intensified the effect of strain on the HAGBs formation; however, this effect saturated when interpass time exceeded 100 s.
At 700 °C, results have shown that 10 s of interpass time is enough for spheroidisation and recrystallisation to be completed, inhibiting grain growth due to the pinning effect of fine cementite particles which do not coarsen excessively during the additional time. At 600 °C, however, the application of longer interpass time could promote the formation of a higher fraction of HAGBs, while Ostwald ripening is not expected to evolve significantly due to the lower temperature applied, which could ultimately help to optimise the ultrafine microstructure under these conditions.
4 Conclusions
1.
The application of warm multi-pass biaxial compression tests activates the dynamic spheroidisation of pearlite, irrespectively of the deformation being applied at 700 or 600 °C. Grooved lamellae have been found in some regions as a consequence of the dynamic recovery that takes place during the deformation process, in agreement with the thermal groove theory. Additionally, the excess of dislocations generated during plastic deformation increases carbon diffusion, favouring cementite dissolution, and altogether accelerates globulisation.
2.
Deformation leads to a significant rise in LAGBs in ferrite, indicating dynamic recovery is the main softening mechanism during the deformation application. HAGBs also increase, although in a lesser extent, suggesting that continuous dynamic recrystallisation is locally activated. The ferrite evolution is strongly influenced by the degree of dynamic spheroidisation of cementite lamellae achieved.
3.
Spheroidisation is controlled by diffusive processes; thus, the kinetics can be correctly described by an Avrami equation. The empirically estimated activation energy of about 45.0 kJ/mol indicates that the rate of spheroidisation of pearlite is determined by the development of an excess of vacancies, which directly increases the self-diffusion rate of iron and indirectly increases the flux of carbon to form particles by the formation and diffusion of carbon–vacancy complexes.
4.
Microstructural evolution during cooling and soaking time shows strain energy accumulated during deformation in both ferrite and pearlite is sufficient to activate metallurgical phenomena statically. Thus, refinement is the result of the evolution of continuous recrystallisation and pinning effect exerted by fine, globulised and homogeneously dispersed cementite particles. Increasing temperature causes accelerated kinetics in metallurgical phenomena; therefore, cooling/soaking time becomes key parameters to achieve UFG and spheroidised microstructures. By controlling them, microstructures with an area-weighted mean size of 2.1 and 2.2 µm have been obtained for 600-30min-AC and 700-AC conditions, respectively.
5.
Interpass time application influences the microstructural evolution as it favours spheroidisation and promotes continuous recrystallisation. The time between passes provides more time for diffusive phenomena to occur favoured by the defects introduced after each pass and for dislocations generated during deformation to rearrange, facilitating LAGBs transformation into HAGBs. Thus, it is a key parameter that can be controlled to optimise continuous recrystallisation while keeping Ostwald ripening to a minimum. In particular, excellent results have been obtained by adding 10 s between each pass at 700 °C, achieving a microstructure with an area-weighted mean size of 0.8 µm.
Acknowledgements
This work was financially supported by the European Coal and Steel Community (RFCS-2015. No. 709828). The authors would like to acknowledge the work of Stefan Meiler at Technische Universität Bergakademie Freiberg for his contribution to the multi-pass biaxial compression tests.
Declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationship that could have appeared influence the work reported in this paper.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
