Recrystallization and grain growth of a nano/ultrafine structured austenitic stainless steel during annealing under high hydrostatic pressure
The aim of this study was to investigate the effect of high hydrostatic pressure applied during annealing on the processes of recrystallization and grain growth in nanostructured austenitic stainless steel 316LVM. The nanostructures were obtained by profile rolling to a total strain of 3.4 and by high-pressure torsion to a total strain of 79. These processes resulted in microstructures consisting of nanotwins and nanograins, respectively. The deformed samples were annealed at 900 °C for 10 min under atmospheric or hydrostatic pressures of 2 and 6 GPa. The resulting microstructures were examined using transmission and scanning electron microscopy techniques. The mechanical properties were evaluated in microhardness measurements. It was established that annealing under high hydrostatic pressure retards recrystallization and grain growth, both in profile-rolled and high-pressure torsion-processed samples. The magnitude of retardation depends on the character of the grain boundaries. The non-equilibrium high-angle grain boundaries present in the high-pressure torsion-processed sample show higher mobility under pressure than the nanotwinned and low-angle boundaries in the profile-rolled sample.
Recovery, recrystallization and grain growth are some of the most important processes that affect the properties of crystalline materials. For this reason, many papers have been devoted to understanding these phenomena in microcrystalline [1, 2], and recently in nanocrystalline, materials [3, 4, 5]. During the annealing of nanocrystalline materials produced by severe plastic deformation (SPD) techniques, one can observe typical processes of recovery, recrystallization and grain growth, but they follow a different course than in materials deformed by conventional techniques [6, 7, 8]. In general, nanomaterials are less thermally stable than their microcrystalline counterparts, and recrystallization begins below their usual recrystallization temperatures. However, it has also been shown that, by creating uniform nanostructures having a uniform grain size distribution and a high content of high-angle grain boundaries, thermal stability can be significantly improved [9, 10, 11]. The enhanced thermal stability of a nanostructured austenitic stainless steel can be obtained by adding yttrium powders during mechanical milling . The addition of 1.5 wt% stabilizes the microstructure around 116 nm grain size after 3 h annealing at 1100 °C. Moreover, the creation of deformation-induced nanotwins makes the microstructure stable up to 800 °C . The processes of recrystallization and grain growth in nanostructured, and especially single-phased, materials have been quite well described. Nevertheless, one issue which remains almost completely neglected is how such materials behave during annealing under high hydrostatic pressure. This subject has been raised only in the author’s previous work [14, 15]. In those studies, annealing under high hydrostatic pressure was applied to a hydrostatically extruded (HE) austenitic stainless steel to optimize its mechanical properties, particularly its strength to ductility balance. It was possible to achieve a controlled slowing of recrystallization processes to produce a nanostructured austenitic stainless steel possessing a good combination of strength and ductility (an ultimate tensile strength of 1247 MPa and a total elongation of 24.4%). The only experiments that have been performed under high pressure recently refer to the Ge precipitation rate in Ge+ ion-implanted SiO2 films , a study on the effect of annealing under pressure on the material properties of Cu2ZnSn(S,Se)4 thin films  and an enhancement of magnetic refrigeration performance in metamagnetic MnCoSi alloy by high-pressure annealing .
Furthermore, it was found during annealing of cold-rolled copper under a pressure of 4.2 GPa that the pressure retarded recrystallization . The general conclusion that can be drawn from this work regarding grain growth is that the higher the hydrostatic pressure applied, the more significant the decrease in grain boundary mobility. Moreover, the impact of high pressure on the <100>, <110> and <111> tilt boundaries was analyzed [18, 22]. The results showed that the movement of <110> tilt boundaries depends on the activation energy, as this movement is effected by the cooperative motion of several atoms. The movement of the <100> and <111> tilt boundaries is not related to the activation energy and is effected by a single atom mechanism. It was also noticed during experiments on normal grain growth in aluminum under high hydrostatic pressure that the migration of low-angle grain boundaries was slowed down even more, nearly frozen in comparison with the case of general grain boundaries . This is because low-angle grain boundaries can migrate by vacancy grain boundary migration, which is highly limited under high pressure.
In the view of published work within the subject of annealing under the high hydrostatic pressure, it seems that annealing under high hydrostatic pressure can be an effective way to control the grain growth in nanostructured materials and produce fine grain materials which could not be obtained under atmospheric pressure due to fast and uncontrollable grain growth. However, this requires full understanding of processes taking place during annealing under high hydrostatic pressure. Therefore, the aim of the current study was to investigate the influence of high pressure on the recrystallization and grain growth of a nanostructured stainless steel 316LVM. To this end, specimens of an austenitic stainless steel were deformed by profile rolling (PR)—a conventional deformation technique, and by high-pressure torsion (HPT)—an SPD technique; the samples were then annealed under a pressure of 2 and 6 GPa and, for comparison, under atmospheric pressure. The HPT process was chosen because it is one of the most efficient SPD processes for grain refinement [28, 29, 30, 31, 32, 33].
Materials and methods
Chemical composition (wt%) of austenitic stainless steel 316LVM
The samples were conventionally deformed using profile rolling (PR) with a reduction in cross section of 23.8, which corresponds to a strain value of approximately 3.4. In this case, the strain was calculated according to the equation ɛ = 2ln (d1/d2), where d1 (ϕ = 12 mm) is the initial diameter and d2 (ϕ = 2.2 mm) the final diameter. It must be pointed out that this calculation is only an approximation, since the cross section is not actually a circle. Additionally, the material was cut into disks with a diameter of 10 mm and a thickness of 0.8 mm. The disks were processed at room temperature using an HPT device at a constant pressure of 6.0 GPa. The disks were torsionally strained to 5 revolutions. The strain was well defined as simple shear, γ, and was calculated according to the equation γ = 2π × r × n/t, where r, n and t are the distance from the torsion axes, the number of applied revolutions and the mean thickness of the sample, respectively. The equivalent strains ɛeq = γ/√3 calculated 3.5 mm from the central point of the sample after 5 revolutions were equal to 79. A phase analysis of the as-received, HPT- and PR-processed samples was performed on a Bruker D8 Advance diffractometer with nickel-filtered copper radiation (λ = 0.154056 nm). The data were collected in a range between 10° and 120° 2Θ, with a step width of Δ2Θ = 0.02° and a counting time 5 s. The energy of the emitter beam was 40 kV, and the current was 40 mA.
After PR, samples of 3 mm in height were cut from a 500 mm rod. After the HPT experiments, disks of 5 mm in diameter were cut in such a way that the radius of the sample after HPT became the diameter of the sample for the annealing experiments. The dimensions of the samples were limited by the dimensions of the toroidal high-pressure cell. The samples were annealed at 900 °C for 10 min, either at atmospheric pressure (0.1 MPa) or at a hydrostatic pressure of 2 or 6 GPa and. To verify whether annealing under high hydrostatic pressure leads to the creation of similar microstructures as conventional annealing for a shorter time, additional experiments of annealing under 0.1 MPa for 4 min were performed. A toroidal high-pressure cell [34, 35] was used for the high-pressure annealing, and a conventional furnace for annealing at atmospheric pressure. Further on in this text, these samples will be referred to as PR_10_0.1 MPa, PR_10_2GPa, PR_10_6GPa, PR_4_0.1 MPa, HPT_10_0.1 MPa, HPT_10_2GPa, HPT_10_6GPa and HPT_4_0.1 MPa.
The microstructure of the samples was investigated using a Hitachi SU 8000 scanning electron microscope working at 5 kV in the BSE mode and a JEOL JEM 1200 EX transmission electron microscope working at 120 kV. For the scanning electron microscopy examination, the samples were prepared by electropolishing using Struers electrolyte A3. The parameters for electropolishing were as follows: voltage − 15 V, time − 15 s. For the transmission electron microscopy examination, the samples were prepared by mechanical polishing to a disk thickness of about 100 μm. Further thinning to reach a thickness appropriate for electron transparency was carried out by electropolishing using Struers electrolyte A2. In the case of the PR samples, the microstructure was analyzed in the center of the samples. In the case of the HPT samples, for microstructure analysis of the cross sections, disks with a diameter of 3 mm were cut from the edge regions of each disk so that the areas of observation were 3.5 mm from the central point. Further thinning was carried out by means of a conventional procedure. Qualitative and quantitative studies as well as the percentage share of the recrystallized areas of the microstructures were conducted using stereological and image analysis methods [36, 37]. The grain size was determined as the equivalent diameter, d2, defined as the diameter of a circle having an area equal to the surface area of a given grain. The grain shape was described by grain elongation factor, defined as the ratio of the maximum to the equivalent diameter dmax/d2. To establish the variation of the size of individual grains, a variation coefficient, CV(d2), defined as the ratio of the standard deviation SD (d2) to the mean value, was determined.
Microhardness measurements were conducted on polished cross sections of the PR and HPT samples. These measurements were made using a Zwick microhardness tester under a load of 200 g. The values of the Vickers microhardness, Hv, were recorded along a diameter of the PR and HPT samples with a separation of 0.1 mm.
X-ray diffraction analysis
Microstructure observation after plastic deformation
Microstructure observations after annealing
Applied annealing pressure at 900 °C hinders recrystallization and grain growth in the PR-processed samples and grain growth in the HPT-processed samples. In the case of the PR samples, the equivalent diameter reaches 0.42 μm under 2 GPa and only 0.087 μm under 6 GPa. Between the recrystallized areas, one can notice areas where there has been no recrystallization. In the case of the HPT samples, the equivalent diameter reaches 1.35 μm under 2 GPa and 0.58 μm under 6 GPa, which is greater than in the PR samples. Moreover, in contrast to the PR samples, there are no non-recrystallized areas between the recrystallized grains. As well as investigating the effect of high hydrostatic pressure on the equivalent diameter of the PR and HPT samples, attention was paid to the impact of high hydrostatic pressure on the elongation parameter and coefficient of variation. It seems that the high hydrostatic pressure applied had no impact on the elongation parameter, which was approximately 1.3 for all annealing conditions for both the PR and HPT samples. Nevertheless, the pressure applied during annealing had a significant impact on the coefficient of variation. There are two tendencies visible. In the case of the PR samples, the coefficient of variation decreases with an increase of pressure from 0.79 under 0.1 MPa to 0.46 under 6 GPa, whereas in the case of the HPT samples it increases from 0.47 under 0.1 MPa to 0.64 under 6 GPa. It means that the abnormal grain growth, represented by a high coefficient of variation, is favorized in the PR sample under the atmospheric pressure and in the HPT sample under the increased pressure.
Comparison of the thermal stability of HPT- and PR-processed samples under 0.1 MPa
Even though the TEM observations confirmed the refinement of the microstructures of the HPT- and PR-processed samples, their behavior during annealing under 0.1 MPa was different. The differences are visible after annealing for 4 and 10 min. After 4 min of annealing, it can be noticed that in the PR-processed samples discontinuous recrystallization occurred, whereas in the HPT-processed samples the recrystallization was continuous. This discontinuous recrystallization (frequently called primary recrystallization) is the result of an inhomogeneous microstructure produced during deformation. If the microstructure after deformation is inhomogeneous, which is true for low-to-moderate strains (in the case of the PR-processed samples − ε = 3.4) during deformation, it means that there are preferred sites for the formation of nuclei having high local misorientations, e.g., highly misoriented regions within deformation bands . The process of heterogeneous nucleation leads, over a longer annealing time, to a microstructure that is fully recrystallized but of diversified grain size. In the case of the austenitic stainless steel, continuous recrystallization was reported for a total strain of 6.4 reached during multi-axial compression . During annealing of the HPT-processed sample that was deformed to a high strain of 79, the strain-induced high-angle grain boundaries change to conventional ones. As a result, homogeneous nucleation and, for longer times, normal grain growth were perceived. This kind of behavior has been previously observed in austenitic stainless steels deformed by HPT and subsequently annealed .
Moreover, after 10 min of annealing, the HPT-processed samples showed smaller grain size than the PR-processed samples. The visibly enhanced thermal stability here may be explained by the fact that, during heating, the recovery of the non-equilibrium grain boundary structure (definitely present in a higher volume in the HPT-processed samples than in the PR-processed samples) proceeds quite rapidly due to their high diffusivity . Non-equilibrium grain boundaries are specific grain boundaries that possess an increased free energy density, increased width, a high density of dislocations associated with the near-boundary region, and correspondingly large residual microstrains . This leads to a rapid decrease in the driving force of the grain growth. Another important factor explaining the enhanced thermal stability of a HPT sample is the more uniform microstructure after deformation in the case of HPT than in a PR-processed sample. The impact of the uniform microstructure on the thermal stability was proven in the experiment on molybdenum . The thermal stability of molybdenum processed by HPT exceeded significantly the respective one for multi-step forging despite the lower deformation degree in the latter case, which was related to the more homogeneous grain size distribution in the former case. As a result, the mobility of grain boundaries decreases, which results in their enhanced thermostability.
Impact of high hydrostatic pressure on recrystallization and grain growth in nanostructured austenitic steel 316LVM
The above results indicate that hydrostatic pressures of 2 and 6 GPa have a considerable impact on the process of recrystallization and grain growth in the PR samples and mainly on grain growth in the HPT samples. The fact that increasing the annealing pressure retards recrystallization and grain growth has been investigated in the past in numerous studies [20, 21, 22, 23, 24, 25, 26, 27]. For example, in one such early study, it was found that a pressure of 4.2 GPa applied while annealing polycrystalline copper cold-rolled to 98% retarded both the initiation and the rate of recrystallization . In another work, it was proved that that grain growth decreased by a factor of 1.3 under a high pressure of 1.2 GPa in aluminum rolled to a 90% reduction . A similar effect of recrystallization retardation was observed during annealing at 300 °C of an Al-2%Mg alloy under an applied stress of 10 MPa .
Firstly, in the PR-processed samples, most of the microstructure is occupied by nanotwins and elongated deformation bands consisting of subgrains, whereas in the HPT-processed samples there is a preponderance of nanograins. In previous studies on micrograined materials, it was explained that low-angle grain boundaries, present in the PR samples, move by the vacancy diffusion mechanism . However, the vacancy concentration in the material decreases with increasing pressure during annealing . For this reason, the movement of low-angle grain boundaries is strongly slowed down during annealing under high hydrostatic pressure. This explains the fact that there are non-recrystallized areas in the PR-processed samples after annealing under high hydrostatic pressure. Moreover, in the PR samples there is a high density of nanotwin boundaries viewed as 60° <111> twist boundaries or 70.5° <110> tilt boundaries. It is known that <110> tilt boundaries move by some cooperative motion of several atoms, while <100> and <111> tilt boundaries move by a single atom mechanism, which can be achieved more easily under high hydrostatic pressure [21, 22]. This fact also explains the retardation of recrystallization under high hydrostatic pressure in the PR-processed samples. A strong retardation of grain growth was also observed in a nanotwinned austenitic stainless steel refined by HE after annealing for 10 min under a pressure of 6 GPa (d2 = 133 nm). The behavior of HE-processed and PR-processed samples during annealing under high hydrostatic pressure confirms that the migration of twins boundaries is hampered under high pressure.
Secondly, these samples differ in the equivalent strain applied during deformation, which was 3.4 and 79 for the PR- and HPT-processed samples, respectively. According to previous studies, the higher the plastic deformation and hydrostatic pressure applied during deformation, the higher the density of vacancies in the material [46, 47]. The excess vacancy concentration in pure Cu and Ni samples processed by HPT can achieve values of (0.9–20)*10^(−4) [46, 47], whereas in α’-martensite processed by HPT the equivalent value are (5.2 ± 3.6)*10^(−4) . Since austenitic stainless steel is a material having a low stacking fault energy and a high melting temperature, one can predict that a high density of agglomerated vacancies will appear in the HPT-processed samples, as opposed to the PR-processed samples. This means that even though the vacancy concentration decreases with increasing pressure during annealing, in the HPT-processed samples it must be much higher than in the PR-processed samples during annealing under high hydrostatic pressure, enabling dislocation climbing and the migration of grain boundaries even under a pressure of 6 GPa.
Thirdly, in the HPT-processed samples one can expect to find a high density of non-equilibrium grain boundaries. The specific structure of grain boundaries affects diffusivity, which is much higher than in the case of general high-angle grain boundaries and highly defected twin boundaries. It is also possible that there is a certain fraction of non-equilibrium grain boundaries present in the PR-processed samples, as it is highly deformed. Nevertheless, that value is much lower than in the HPT-processed samples. During annealing under 0.1 MPa, the rapid recovery of such boundaries may decrease their mobility. In contrast, under high pressure, such a rapid recovery may be suppressed, and the enhanced diffusivity and excess energy of the non-equilibrium boundaries may act as a driving force for grain growth.
Does annealing under high hydrostatic pressure result in the same microstructure as under atmospheric pressure?
In order to answer this question, experiments in which PR- and HPT-processed samples were annealed for 4 min under 0.1 MPa were performed. The results were compared with the microstructures obtained for annealing under 2 and 6 GPa for 10 min. It was discovered that, in the case of the PR-processed samples, the microhardness and microstructure after 4 min of annealing under 0.1 MPa are comparable to the values obtained after 10 min under 2GPa. However, they differ slightly in the percentage of the recrystallized area, which is greater for the PR_10_2GPa—approximately 80% than the PR_4_0.1 MPa—approximately 70%. The HPT-processed samples annealed for 4 min under 0.1 MPa were fully recrystallized and can be compared with the samples annealed for 10 min under 6 GPa. However, they differ slightly in the value of the coefficient of variation, which is greater for the samples annealed under high hydrostatic pressure. This might result from the fact that under high pressure applied during annealing the migration of vacancies is highly reduced and various grain boundaries tend to migrate by various mechanisms and consequently at various rates .
HPT processing allows obtaining the nanograined austenitic stainless steel 316LVM, while ultrafine grained one (with a mixture of nanotwins and elongated deformation bands) can be obtained by a more conventional processing of PR.
Annealing under atmospheric pressure revealed that HPT-processed samples are more thermally stable when compared to PR-processed one. This was attributed to the more uniform microstructure of a HPT samples and their higher content of non-equilibrium grain boundaries, which have the tendency to a rapid recovery during heating drastically reducing the driving force for grain growth.
Annealing under high hydrostatic pressures of 2 and 6 GPa retards the processes of recrystallization and grain growth in samples processed by both methods; however, the retardation is much more pronounced for a PR-processed samples. This was related to such microstructural features of HPT samples as a higher vacancy concentration, lower frequency of low-angle grain boundaries and higher fraction of non-equilibrium grain boundaries, which all enhance the mobility of grain boundaries under high pressure.
The microstructure reorganization during annealing under high pressure proceeds in the same way as during annealing under atmospheric pressure; however, for the latter the same grain size is obtained after shorter annealing time.
This work was supported by Polish NSC project SONATA No. UMO-2014/15/D/ST8/00532.
- 15.Gierlotka S, Grzanka E, Krawczynska A et al (2011) Strain relaxation and grain growth in 316LVM stainless steel annealed under pressure. Zeitschrift für Krist Proc 1:281–286Google Scholar
- 17.Ranjbar S, Menon MRR, Fernandes PA, Da Cunha AF (2015) Effect of selenization conditions on the growth and properties of Cu2ZnSn(S,Se)4thin films. In: Thin solid films. pp 188–192Google Scholar
- 25.Lojkowski W (1988) On the effect of high pressures on the mobility of atoms in grain boundaries. J Phys-Paris 49:545–549Google Scholar
- 31.Valiev RZ, Estrin Y, Horita Z et al (2006) Producing bulk ultrafine-grained materials by severe plastic deformation. J Microsc 58:33–39Google Scholar
- 34.Khvostantsev LG, Vereshchagin LF, Novikov AP (1997) A verkh-niz (up-down) toroid device for generation of high pressure. High Temp-High Press 9:637Google Scholar
- 41.Humphreys FJ, Hatherly M (1996) Recrystallization and related annealing phenomena. Pergamon Press, OxfordGoogle Scholar
- 47.Zehetbauer MJ, Steiner G, Schafler E, Korznikov A, Korznikova E. In Horita Z (ed) Proceedings of the 3rd international conference on nanomaterials by severe plastic deformation, Fukuoka, Japan (2005) Mater Sci Forum 503–504: 57–64Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.