Materials Science & Engineering A 831 (2022) 142218
Available online 20 October 2021
0921-5093/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Effects of strain rate on strain-induced martensite nucleation and growth in 301LN metastable austenitic steel
Lalit Pun
*, Guilherme Corr ˆ ea Soares , Matti Isakov , Mikko Hokka
IMPACT- Multiscale Materials Research Group, Engineering Materials Science, Faculty of Engineering and Natural Sciences, Tampere University, FI-33014, Tampere, Finland
A R T I C L E I N F O Keywords:
Strain-induced α′-martensite transformation Metastable austenitic stainless steel Strain rate effect
Digital image correlation Electron backscatter diffraction
A B S T R A C T
The effects of strain rate on strain-induced α′-martensite nucleation and growth were analyzed in this work.
Tension tests were performed at room temperature at strain rates of 2×10−4 s−1 and 0.5 s−1 using small polished specimens that fit inside a scanning electron microscope. The specimens were deformed incrementally, and microstructural evolution was tracked carefully at a specific location on the specimen surface. This approach allows the analysis not only of the spatial but also of the temporal evolution of the α′-martensite. Optical mi- croscopy images and electron backscatter diffraction (EBSD) measurements were taken for each plastic defor- mation increment. The size and number of α′-martensite particles were evaluated from the EBSD images, whereas local microlevel strains were obtained using Digital Image Correlation (DIC). According to the results, the number of nucleation sites for α′-martensite does not seem to be affected much by the strain rate. However, there is a notable strain rate effect on how the transformation proceeds in the neighborhood of freshly formed α′-martensite particles. At a low strain rate, repeated nucleation and coalescence leads to the notable growth of α′-martensite particles, whereas at a high strain rate, once nucleated α′-martensite particles remain as small isolated islands that do not markedly grow with further plastic strain. This phenomenon can be attributed to local microstructure-level heating caused by plastic deformation and exothermic phase transformation. This reduces the local growth rate of the α′-martensite particles in the vicinity of the above-mentioned islands, thus leading to a lower bulk transformation rate at higher strain rates.
1. Introduction
Austenitic stainless steels have been extensively studied over the years because they are good candidates for many applications mainly due to their desirable mechanical properties, corrosion resistance, toughness, and weldability. The deformation and/or transformation mechanisms of these steels are strongly influenced by the stacking fault energy (SFE) [1]. For example, the metastable austenitic stainless steel 301LN has a rather low stacking fault energy. In this alloy, the parent austenite phase with face-centered cubic (FCC) structure is relatively unstable during plastic deformation and readily transforms to near body-centered cubic (BCC) or body-centered tetragonal (BCT) α′-martensite [2]. The word ‘martensite’ within this work relates to α′-martensite having a BCC or BCT structure. In contrast, more alloyed steels, such as AISI 316, have a higher stacking fault energy, and therefore the phase transformation is suppressed and mechanical twin- ning is favored during deformation [3,4]. The general features of the
formation of α′-martensite during the plastic deformation of metastable austenitic steels are well known. Two transformation sequences are commonly reported in the literature [5–11], i.e., through an interme- diate epsilon martensite phase (γ→ε→α’) or directly from austenite to martensite (γ→α’). In the first transformation process, the ε-martensite with a hexagonal close-packed (HCP) structure acts as a precursor phase for the α′-martensite formation. However, according to Hedstr¨om et al.
[12], ε and α′-martensite form independently in AISI 301. It has also been shown that SFE influences the transformation sequence. Materials that have SFE < 18 mJ/m2 promote the indirect transformation sequence, whereas the direct sequence occurs in materials with SFE >
18 mJ/m2 [13]. Furthermore, it has been concluded that ε-martensite is favored by low temperatures [14,15].
Several microstructural features have been reported to be possible nucleation sites: the intersections of active slip systems with ε plates, the intersection of ε plates with a twin or grain boundary [6], grain and twin boundaries [7,13], isolated shear bands, grain boundary triple points,
* Corresponding author.
E-mail address: lalit.pun@tuni.fi (L. Pun).
Contents lists available at ScienceDirect
Materials Science & Engineering A
journal homepage: www.elsevier.com/locate/msea
https://doi.org/10.1016/j.msea.2021.142218
Received 21 June 2021; Received in revised form 12 October 2021; Accepted 18 October 2021
parallel micro shear bands [10], and intersections of shear bands con- sisting of bundles of faults and twins [16]. Additionally, there have been several attempts to observe the nucleation and growth of martensite in-situ during deformation using a transmission electron microscope (TEM). Suzuki et al. [17] performed in-situ TEM measurements on two kinds of specimens with grain sizes of 50 and 250 μm and observed α′-martensite forming at the junctions between two intersecting slip bands. Moreover, similar in-situ experiments done by Brooks et al. [18]
showed that martensite nucleates in one active slip system near defects where dislocations pile up. This nucleation process was observed to proceed with numerous pile-ups on parallel closely spaced γ - {111}
planes. In addition, in-situ TEM tensile tests on AISI304 steel performed by Kaoumi et al. [11] at room temperature at a strain rate of 10−3 s−1 revealed that the α′-martensite nucleates at the grain boundaries and grows towards the inside of the austenite grains with further straining.
It is well known that the α′-martensite phase transformation is affected by the loading conditions, such as temperature and strain rate.
However, the effects of strain rate in particular are not completely un- derstood. It is generally accepted that during high strain rate deforma- tion, there is not enough time for effective heat transfer between the deforming material and the surroundings. Therefore, the heat energy resulting from the plastic deformation work and the latent heat of exothermic phase transformations cannot dissipate to the surroundings fast enough, which leads to so-called adiabatic heating and a material temperature increase. This is assumed to suppress the formation of α′-martensite at high strain rates, as reported by many authors [9,16, 19–23]. In addition, Murr et al. [16] found that α′-martensite content increases during high-rate tension at low strain levels, but at higher strains, the formation of martensite is suppressed, and adiabatic heating is thought to be dominant. Staudhammer et al. [21] reported that the minimum thickness of individual α′-embryos varies in the range of 50–70 Å for both a low strain rate (10−3 s−1) and a high strain rate (105 s−1). They showed that the distributions of embryo sizes at low strains for both strain rates are similar and concluded that suppression of α′-nucleation at higher strains and strain rates is a result of adiabatic temperature rise. Moreover, Lee et al. [24] discovered that α′-martensite nucleation can also take place in a single shear band without in- tersections during loading at strain rates ranging from 800 to 4800 s−1. These findings clearly indicate that the nucleation events of martensite particles still occur at high strain rates, but the underlying mechanisms for the reduction of the strain-induced martensite transformation at high-rate loading are not properly explained. Later on, Talonen and H¨anninen [15], and Curtze and Kuokkala [25] concluded that an in- crease in strain rate leads to an increase in SFE, which plays a key role in increasing the stability of the austenite. Isakov et al. [26] studied the effect of strain rate on martensitic transformation using interrupted tests and strain rate jump tests. It was concluded that in addition to bulk adiabatic heating, strain rate has a direct reducing effect on the trans- formation rate of martensite. V´azquez-Fern´andez et al. [27] uncoupled the effects of strain rate and adiabatic heating by experimentally simulating the macroscopic adiabatic heating occurring at high strain rates by heating the specimen continuously during a quasi-static tension test (without the external heating, the test would have been isothermal).
Based on these tests, it was concluded that the suppression of the phase transformation in high-rate deformation cannot be replicated at low strain rates with external heating, i.e., there is a direct strain rate effect in addition to the macroscopic adiabatic heating. Furthermore, they concluded that the increase in strain rate limits the nucleation of martensite to one variant only.
As discussed above, it has been traditionally concluded that the martensitic phase transformation is reduced at high strain rates due to adiabatic heating, but recent experimental evidence indicates that strain rate itself has also a direct role in the suppression of phase trans- formation. It is highlighted here that such conclusions have been drawn mainly by measuring the total (macroscopic) volume content of the α′-martensite. In contrast, less attention has been paid to studying how
the martensite nucleates and grows at the microstructural level at different strain rates. Usually, microstructural investigations are carried out after the deformation. That is, full deformation history for a particular location in the microstructure is not followed starting from the as-received state. Instead, conclusions are mainly drawn by pre- paring the samples after the experiments and comparing different specimens to one another and to a specimen at the as-received state. On the other hand, in-situ TEM experiments have helped in understanding the nucleation of martensite particles and their growth during plastic deformation. However, this kind of work is limited to a very small area and low strain rates only. Also, the thermal conditions of a very thin film would be different compared to a bulk specimen deformed at a high rate.
The thin film-like specimen may not produce similar adiabatic heating as the bulk counterparts. Because of many challenges associated with the experimental set-up and speed of the strain-induced martensitic transformations, in-situ TEM investigations of the phase transformation at high strain rates have not been reported. Instead, numerical analyses, such as micromechanical modeling [28–32] and the phase field method [33,34], have been used to understand the evolution of the micro- structure during martensitic transformations. Yeddu et al. [34] simu- lated the strain-induced martensitic transformations in stainless steels using a three-dimensional elastoplastic phase model and concluded that nucleation of martensite occurs in highly plasticized areas. However, the supporting experimental evidence is still limited. Thermodynamic cal- culations and simulations [25,35] have been carried out to explain the suppression of martensite transformation at high strain rates, but little experimental evidence is present to explain this phenomenon based on the microstructure evolution.
The present work aims to investigate the effects of strain rate on the α′-martensite nucleation and growth on the microstructural level. For this, incremental tensile experiments were carried out at different strain rates (2×10−4 s−1 and 0.5 s−1) on specimens, which had a pre-polished and carefully marked location on the surface. These specimens were small enough to fit inside a scanning electron microscope fitted with an electron backscatter diffraction analyzer. As a result, the evolution of an exact microstructural location could be followed throughout the defor- mation history. The test results clearly show that small α′-martensite particles nucleate readily at both strain rates, but the further growth and coalescence observed at the low strain rate is absent at the higher strain rate. As discussed in detail in the manuscript, this finding can be explained by referring to the local microstructure-level heating in the vicinity of the freshly nucleated α′-martensite particles.
2. Experimental procedure 2.1. Material and tensile experiments
The material investigated in this work was metastable austenitic stainless steel EN 1.4318 (AISI 301LN). The steel was provided in sheet form by Outokumpu Stainless (LTD) in 2B condition, in which the steel was cold rolled, solution annealed, pickled, and skin passed. The chemical composition of the steel is shown in Table 1. The SFE of the as- received steel was calculated using the thermodynamic model proposed by Curtze et al. [1]. The SFE of the 301 LN steel at room temperature is 20 mJ/m2.
The samples used for the tensile experiments were laser cut from a 2 mm steel sheet with a gauge section of 2.9 mm in width and 8 mm in length. Fig. 1 summarizes the experimental procedure carried out in this work. Uniaxial tensile tests were performed at room temperature at two different strain rates of 2×10−4 s−1 and 0.5 s−1. The tests were carried out using a servohydraulic Instron 8800 machine up to the total loga- rithmic strain of 0.18. The tension experiments were interrupted at the logarithmic strains of approximately 0.055, 0.1, 0.14, and 0.18. These interruption points were chosen based on the observation that the strain hardening rates at different strain rates are similar at the logarithmic strains of 0.055 and 0.1, but diverge at higher strains. For each
increment, the microstructures of the steel samples were analyzed using an optical microscope and EBSD. The main goal of these experiments was to compare the evolution of the microstructures of the material at these two different strain rates. For this purpose, a special specimen holder was manufactured, allowing the entire tension specimen to be placed inside the electron microscope for the EBSD investigations without the need to cut the tension specimen into smaller pieces. This way the same specimen could be repeatedly deformed and exactly the same area of the microstructure could be studied with the microscopes between the loading increments.
2.2. Microstructural characterization
Before mechanical testing, the gauge section of the tensile specimens was ground manually with silicon carbide grit papers. The samples were then electropolished using electrolyte A2 and subsequently etched in a 36% HNO3 +61% water solution. Moreover, four small indentations were made in the center of the gauge section of the specimen using a hardness tester to identify the same location where the microstructure was to be examined after each deformation increment. The optical and EBSD images were obtained from the undeformed sample surface as well as after each deformation increment. The optical images were obtained using a Leica DMi8 optical microscope. The EBSD data were obtained using a Zeiss Ultraplus UHR FEG-SEM system operating at an accelera- tion voltage of 20 kV, a tilt angle of 70◦, and a step size of 0.3 μm. Crystal orientation data and information on different phases were obtained bypost-processing the EBSD data using the AztecHKL software. Bound- aries with misorientation above 15◦were defined as high-angle grain boundaries. In addition, the bulk volume fraction of α′-martensite was measured after each deformation increment using the magnetic balance method. This technique determines the overall bulk content or the bulk volume fraction of the ferromagnetic phase, i.e., α′-martensite, by measuring the force that is required to detach a permanent magnet from the surface of the specimen. A detailed description of the method and its calibrations are described in Ref. [36].
2.3. Local DIC measurement and alignment of the strain and EBSD data Local strains were measured from the optical microscopy images of the specimens deformed to the logarithmic strains of 0.055, 0.1, 0.14, and 0.18. The parameters used in the strain calculations are presented in Table 2. The DIC calculations were made with LaVision DaVis 10.1.2.65016 software using both affine and 2nd order shape functions.
It was observed that the affine function gives a somewhat lower spatial resolution and thus effectively acts as a spatial low-pass filter. In contrast, the non-linearity of the 2nd order shape function was observed to better match the local deformation fields and thus result in higher spatial resolution. Therefore the 2nd order shape functions were used for the results presented here. The virtual strain gauge (VSG) size was approximately 13 μm, which is close to the average grain size of the material, i.e., the VSG was larger than the smallest grains found in the microstructure. Although this was not optimal, a compromise had to be made to use the smallest VSG possible while being able to maintain a good correlation in most parts of the microstructure. If a random speckle pattern was applied on the surface of the microstructure, the use of smaller subsets would be feasible, and the spatial resolution of the strain measurements would be enhanced. However, the use of artificial pat- terns would have compromised the EBSD measurements, and therefore Table 1
Chemical composition (wt.%) of the steel used in this work.
Material C Si Mn P S Cr Ni N Mo Fe
301LN 0.023 0.48 1.19 0.030 0.0003 17.4 6.5 0.138 0.10 Bal.
Fig. 1.Flowchart of the experimental procedure.
Table 2
Digital Image Correlation parameters.
Image size (pixels) 5453 ×3623
Camera noise (% of range) 1.13 – 3.50%
Image scale factor (pixel/μm) 20.45
Matching criteria WVSSDa
Interpolant 7-tap B-spline
Subset shape function Nonlinear 2nd order
Subset weighting function Gaussian
Subset size (pixel) 199
Step size (pixel) 66
VSG size (pixel) 265
Maximum iterations 999
Displacement resolution (μm) 1.94 ×10−4 – 8 ×10−3
Strain resolution (%) 0.006 – 0.163
Average correlation value 0.777 – 0.969
a WVSSD: weighted variables-based sum of squared differences.
this was not attempted. It is also important to clarify that a bicubic interpolation was used to estimate the strain between data points, and the presented full-field data should be taken as an average of the strain in that region instead of an absolute value for all points in the map.
Furthermore, the as-polished surface of the undeformed sample did not have enough features for the DIC algorithm to function properly, and the first reference image for the DIC analysis had to be that of the specimen deformed to the 0.055 logarithmic strain. Based on this and the fact that the microstructure evolves notably during deformation, thus affecting the performance of the DIC algorithm, the local strains presented in the results section are incremental, i.e., each strain field relates to the start of the increment, not to the as-received condition of the material. The EBSD maps containing the identified α′-martensite particles as well as the grain and twin boundaries were overlayed on top of the DIC strain maps to enable spatial comparison of the local strains and the local microstructure evolution. Any spherical aberration was corrected by dewarping the images so that the microstructural features would match on both DIC and EBSD maps.
3. Results
3.1. Mechanical response
Fig. 2(a) shows the stress-strain curves of the metastable austenitic stainless steel 301LN deformed at the strain rates of 2×10−4 s−1 and 0.5 s−1. The figure shows data from both incremental and continuous tests.
Although there is a slight difference in the stress-strain curves after the logarithmic strain of 0.10, it can be concluded that the material behaves rather similarly in both continuous and incremental tensile tests. The yield strength increases with the strain rate. Fig. 2(b) compares the respective strain hardening rates. The strain hardening curves shown here were calculated only from the continuous tests. The strain hard- ening rate at both strain rates is practically the same until the loga- rithmic strain of approximately 0.08, after which the hardening rate increases in the lower strain rate test. This is attributed to an increase in the α′-martensite content. In contrast, at the higher strain rate, the strain hardening rate remains at a lower level, which is explained via the decreased tendency toward the phase transformation [19].
The bulk volume fractions of the α′-martensite measured for the metastable AISI 301 LN steel are shown in Fig. 3. At the logarithmic strain of 0.055, the α′-martensite contents are comparable at both strain rates. However, as the plastic deformation proceeds, less martensite is observed in the specimens deformed at the higher strain rate than in the specimens deformed at the lower strain rate. At the logarithmic strain of 0.10, the difference in the bulk martensite content is already significant.
This is in correspondence with the observed differences in the stress- strain and strain hardening curves discussed above.
3.2. Microstructure evolution in the incremental tensile tests
The microstructures of the specimens deformed at different strain rates were studied using EBSD. The amount of non-indexed points, so- called zero solutions, for the undeformed metastable 301 LN speci- mens was 2%, but after each deformation increment, these non-indexed points increased gradually to around 10% after the fourth increment at the total logarithmic strain of 0.18 at both strain rates. In general, the zero solutions mainly appeared near phase and grain boundaries.
Fig. 4 shows the microstructural evolution of the metastable 301LN steel deformed incrementally at the low strain rate of 2×10−4 s−1. A large number of thermal twin boundaries represented by the white color are present in the microstructure. The grain boundaries between the austenite grains are shown in the brownish-red color. Few α′-martensite particles (Fig. 4(a)) were present in the microstructure in the as-received state or were probably produced during the sample preparation. Most of these α′-martensite particles were found at the grain boundaries (shown by ellipses and arrows) and also in one of the austenite grain interiors (inside the black rectangle). The arrows in Fig. 4 indicate some of the
Fig. 2. (a) The flow curves and the (b) strain hardening rate at room temperature for 301LN stainless steel at the strain rates of 2×10−4 s−1 and 0.5 s−1. Fig. 3.Bulk α′-martensite volume fraction as a function of the logarithmic strain at the strain rates of 2×10−4 s−1 and 0.5 s−1 for the metastable AISI 301LN steel.
new α′-martensite particles formed after each strain increment.
Fig. 4(b) depicts the microstructure of the 301LN metastable steel at the logarithmic strain of 0.055. For example, with an increase in the deformation, the α′-martensite particle initially present in an austenite grain interior (inside the black rectangle marked in Fig. 4(a) and (b)) grew larger. At this stage of the deformation, nucleation of new α′-martensite particles mostly takes place at the interior of twins, as well as near twin and grain boundaries. There are also color variations within some of the austenite grains corresponding to local differences in crystal orientations. These local misorientations occur from the pile-ups of dislocations generated during the plastic deformation of austenite grains [37].
Fig. 4(c) demonstrates the microstructure after the logarithmic strain of 0.103. With an increase in the logarithmic strain to 0.103, the already nucleated martensite particles grow in size. The α′-martensite particles nucleated at the grain boundaries extend towards the grain interiors. It can also be seen that the α′-martensite particles nucleated inside the austenite grains grow larger than those nucleated at the twin/grain boundaries. This is evident when comparing the growth of the martensite particles in areas highlighted by the circle and rectangle (inside the austenite grains) with the ones highlighted with the ellipse and hexagon (nucleated on the grain and twin boundaries) in Fig. 4(b) and (c). Also, it is interesting to notice that the nucleation of new α′-martensite particles at this stage of deformation mostly takes place in
the interior areas of austenite grains rather than at the twin/grain boundaries.
Further evolution of the microstructures at the logarithmic strain of 0.145 and 0.18 are presented in Fig. 4(d) and (e) respectively. New α′-martensite particles nucleate mostly inside austenite grains. Most importantly, there are multiple nucleation sites within an austenite grain. In the meantime, existing α′-martensite particles grow bigger.
Moreover, after the logarithmic strain of 0.18, α′-martensite particles have nucleated in virtually all of the austenite grains in the field of view.
The appearance of α′-martensite particles is blocky and irregular in shape.
Fig. 5 shows the microstructural evolution of the AISI 301 LN steel deformed incrementally at the strain rate of 0.5 s−1. The initial micro- structure of the undeformed specimen is depicted in Fig. 5(a), with ar- rows indicating some of the α′-martensite particles in the as-received state or possibly produced during sample preparation. All of the α′-martensite particles were present at the grain boundaries.
Fig. 5(b) shows the microstructure at the logarithmic strain of 0.055.
The α′-martensite particles appear at the grain boundaries and in the interior of the austenite grains. To follow the growth of the martensite particles with further deformation, an example of a martensite particle that had nucleated inside an austenite grain is indicated by the circle, while a martensite particle nucleated at the grain boundary is indicated by the rectangle. No martensite particles were found in the interior of Fig. 4.(a) EBSD crystal orientation map of the as- received 301LN steel superimposed with austenite grain boundaries (>15◦- brownish-red) and the white color showing detected <111>−60◦thermal twin boundaries. The microstructure deformed at a strain rate of 2×10−4 s−1 to the logarithmic strains of (b) εpl =0.055, (c) εpl =0.103, (d) εpl =0.145, and (e) εpl =0.18. The arrows indicate some of the new α′-martensite particles after each strain incre- ment. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
the twins/twin boundaries, which is in contrast to the observations made in the specimen deformed at the lower strain rate (Fig. 4). Also, existing martensite particles observed in Fig. 5(a) did not grow bigger.
On the other hand, the number of martensite particles that formed during the higher strain rate deformation is comparable to that observed at the lower strain rate. Some of the new α′-martensite particles that formed during deformation are indicated by arrows. As can be seen, the size of the particles observed at the higher strain rate is much smaller than those observed at the lower strain rate (Fig. 4).
Fig. 5(c) shows the microstructure at the 0.10 logarithmic strain. The martensite particles have grown very little compared to Fig. 5(a) and (b). New martensite particles nucleated in the microstructure in a similar manner as described above for Fig. 5(b). Some examples of the new martensite particles are indicated by arrows. A similar evolution of the microstructure continues in Fig. 5(d) and (e), which show the microstructure after deformation to the 0.14 and 0.177 logarithmic strains, respectively. In Fig. 5(e), a large number of small α′-martensite particles can be seen. The martensite islands found inside the austenite grains have grown larger, but the martensite particles present on the grain boundaries have not grown much.
3.3. Local strains in the microstructure
The local strains in the microstructures of the AISI 301 LN samples were determined using DIC based on the optical images taken after the strain increments. The strain data were superimposed with the EBSD data. In Fig. 6(a), the black lines indicate grain and twin boundaries and the blue color represents the α′-martensite particles of a specimen deformed at the lower strain rate of 2×10−4 s−1 to the global logarithmic strain of 0.055. Fig. 6(b) illustrates the local strains in the loading di- rection within the microstructure as the global logarithmic strain is increased by 0.048. During this deformation increment, some of the grains have deformed locally by a logarithmic strain increment of 0.10, which is approximately twice the applied global strain increment (0.048), indicating a very heterogeneous local strain distribution.
Similar observations of strain localization within some austenite grains have also been reported by Das et al. [38]. As an example, the local strains in Fig. 6(b) are strongly accumulated in two areas, which are denoted by black rectangles. In these areas, new martensite particles are found when deformation proceeds, and the existing particles also grow bigger. Moreover, there is also a region (indicated by the black ellipse) in Fig. 6(b), where large local deformation has taken place but no Fig. 5.EBSD crystal orientation maps of the (a) as-received 301LN steel and after deformation at a strain rate of 0.5 s−1 to the logarithmic strains of (b) εpl =0.055, (c) εpl =0.10, (d) εpl =0.14, and (e) εpl =0.177. The annotations and color coding are according to Fig. 4. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
martensite is found. Furthermore, the local strains in some of the austenite grains are lower than the applied global strain or even reaching zero (some of them are indicated by the white ellipse areas).
Nevertheless, the majority of austenite grains deformed locally by amounts close to the applied global strain increment.
The microstructures and local strain maps for the further deforma- tion increments are presented in Fig. 6(c) and (d). In the locations indicated by the rectangles in Fig. 6(c), more martensite particles are found with further deformation, and the particles again grow bigger.
The local strains in those areas represented by the rectangles are approximately twice compared to the global or average deformation increment. In Fig. 6(d), areas indicated by the rectangles have large local strains and a strong increase in the number and size of martensite par- ticles. Also, areas indicated by the white ellipse in Fig. 6(d) show an area of austenite grains with near zero local strains.
Fig. 6(e, f, g, and h) illustrate the microstructural and local defor- mation evolution for the AISI 301LN at the higher strain rate of 0.5 s−1. Similar deformation hot spots that were observed at the lower strain rate can also be observed at the higher rate deformation. The martensite
particles are often found in and around these regions of high local deformation. However, it is quite obvious that the size of the martensite particles is smaller, and their growth is more limited at the higher strain rate. The martensite particles appear in the regions of large local deformation, but their size does not seem to grow despite the continued deformation.
4. Discussion
From the EBSD data, it was observed that the growth of the α′-martensite particles occurs by repeated nucleation and coalescence similarly as reported by Murr et al. [16]. The resulting martensite par- ticles appear irregular and blocky. A closer examination of these large and blocky α′-martensite particles (Fig. 4(e)) reveals that they are made up of several small α′ particles that have formed close to each other.
Moreover, some of the larger martensite islands are composed of several small segments of α′ that are oriented differently. In such cases, the indexing algorithm used here does not count them as a single α′particle.
Also, if there is a grain boundary within the α′-martensite island, the Fig. 6.Local strains in the loading direction superimposed with EBSD images at two different strain rates. The blue color in the micrographs represents α′-martensite particles, the black lines represent grain boundaries, and the continuous coloring represents the local strains obtained for each loading increment. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
algorithm again separates the island into different martensite particles.
For these reasons, areas of individual α′-martensite particles were additionally measured using ImageJ software. The measurements were carried out on an area sized 9295 μm2.
The statistical distribution of different-sized particles is shown as histograms in Fig. 7. The gray bars in the Figure represent the number of different-sized martensite particles observed in the micrographs of the specimen deformed at the strain rate of 2×10−4 s−1, whereas the red bars represent similar data obtained at the strain rate of 0.5 s−1. In Fig. 7 (a), it can be seen that there is a large number of small martensite par- ticles at the beginning of plastic deformation at both strain rates. At the logarithmic strain of 0.055, the largest α′-particle at the higher strain rate has an area of 3 μm2, whereas at the lower strain rate, growth by coalescence has produced some large martensite particles with an area of ~16 μm2. Because of the repeated nucleation and coalescence, the martensite particles formed at the lower strain rate continue to grow extensively with plastic deformation, but at the higher strain rate of 0.5 s−1, the growth of these particles is comparatively low. This can be seen in Fig. 7(c); after 0.14 logarithmic strain at the low strain rate, α′-martensite islands with a total area of up to ~100 μm2 are visible in the microstructure but are absent at the higher strain rate. Fig. 7(d) shows the histogram after the fourth increment at the logarithmic strain of 0.18. At this amount of plastic deformation, the number of small α′-martensite particles formed at the lower strain rate differs only slightly from that at the high rate, but large α′-martensite islands are observed only at the lower strain rate. As noted above, at the lower strain rate some of the small martensite particles have nucleated very close to each other to form large α′-martensite islands.
As discussed above, there is a clear difference in the morphology of the strain-induced α′-martensite at the two studied strain rates; even
though the number of small α′-particles is comparable for both strain rates, formation of large α′-islands via repeated nucleation and coales- cence takes place only at the lower strain rate. This is an important finding and is discussed in the following from the standpoint of local microstructural adiabatic heating. Firstly, it should be noted that the microstructural data in this study was collected from incremental loading experiments. As the material is cooled down after every increment, the bulk adiabatic heating in the incremental high strain rate experiments is rather low, and the bulk thermal conditions are closer to those of the continuous isothermal low strain rate tests. In fact, the bulk heating of the same AISI 301 steel specimens during high rate deformation was measured by V´azquez-Fern´andez et al. [39]. Using high-speed infrared cameras, they were able to demonstrate that the bulk heating of the specimens at the strain rates of 0.1 s−1 and 85 s−1 after the plastic deformation of 5% was only around 5 ◦C. However, the local- or microstructural-level conditions can still be adiabatic, and the temper- ature of the material can increase considerably on the microstructural level. This, together with the heat released locally by the exothermic phase transformation, can lead to the formation of local ‘hot spots’
where the temperature is higher than elsewhere in the material, which at lower strain rates does not occur because the heat is effectively dissi- pated to the surroundings. As a result, even though the martensite particles are formed, their growth is impeded by the local heating, and thus the particles remain smaller, leading to a lower global phase transformation rate. Furthermore, the experimental evidence obtained in this work suggests that plastic deformation is not uniformly distrib- uted in the microstructure. The strain-induced martensitic phase trans- formation leads to significant local strain heterogeneity in the microstructure as predicted by micromechanical modeling [30− 32].
The deviation in the local strain state from the applied global one is due
Fig. 7.Size distribution histograms of the α′-martensite particles measured from the micrographs taken between the deformation increments.
to increasing amounts of heterogeneously distributed martensite in the microstructure [30]. Also, the evidence presented in this work shows that the α′-martensite phase transformation tends to focus on the regions with large local strains. Similar results have been presented by Yeddu et al. [34] and Rafaja et al. [40], who concluded that martensite nucleation occurs in highly plasticized areas.
5. Conclusions
The effects of strain rate on the microstructural evolution of 301LN metastable austenitic stainless steel were investigated by incremental tensile tests at low and high strain rates. Microstructural characteriza- tion was carried out using optical microscopy and EBSD. The conclu- sions of the work can be summarized as follows:
- New α′-martensite particles form in the microstructure during deformation at both strain rates. The growth of martensite particles occurs by repeated nucleation and coalescence.
- The martensite particles grow much larger during deformation at the lower strain rate of 2×10−4 s−1, whereas at the higher strain rate of 0.5 s−1, the growth of the particles is minimal. Large α′-martensite islands formed by the coalescence of small particles are absent at the high strain rate deformation.
- The deformation is quite inhomogeneous on the microscale. Some areas barely deform at all, whereas some other areas deform twice as much as compared to the global average strain.
- New martensite particles are often observed in the regions of large plastic strain.
- The reduced growth of the martensite particles at high strain rates can be explained by local adiabatic heating. The martensite particles form in areas with a large plastic strain, and if the strain rate is high, the local adiabatic heating together with the latent heat from the phase transformation increases the local temperature of the material and stalls the phase transformation in that part of the microstructure.
On the global scale, this is observed as a reduced phase trans- formation rate at high strain rates.
Data availability statement
All data included in this study are available upon request by contact with the corresponding authors.
CRediT authorship contribution statement
Lalit Pun: Conceptualization, Methodology, Investigation, Data curation, Software, Writing – original draft. Guilherme Correa Soares: ˆ Investigation, Writing – review & editing. Matti Isakov: Conceptuali- zation, Supervision, Validation, Writing – review & editing. Mikko Hokka: Project administration, Funding acquisition, Conceptualization, Supervision, Validation, Writing – review & editing.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This study was financed by Tampere University graduate school. We would like to thank Dr. Mari Honkanen for the EBSD measurements of the samples. This work made use of the Tampere Microscopy Center facilities at Tampere University.
References
[1] S. Curtze, V.-T. Kuokkala, A. Oikari, J. Talonen, H. H¨anninen, Thermodynamic modeling of the stacking fault energy of austenitic steels, Acta Mater. 59 (2011) 1068–1076, https://doi.org/10.1016/j.actamat.2010.10.037.
[2] M.J. Sohrabi, M. Naghizadeh, H. Mirzadeh, Deformation-induced martensite in austenitic stainless steels: a review, Arch. Civil Mech. Eng. 20 (2020) 1–24, https://
doi.org/10.1007/s43452-020-00130-1.
[3] I. Karaman, H. Sehitoglu, H.J. Maier, Y.I. Chumlyakov, Competing mechanisms and modeling of deformation in austenitic stainless steel single crystals with and without nitrogen, Acta Mater. 49 (2001) 3919–3933, https://doi.org/10.1016/
S1359-6454(01)00296-8.
[4] S.J. Wang, T. Jozaghi, I. Karaman, R. Arroyave, Y.I. Chumlyakov, Hierarchical evolution and thermal stability of microstructure with deformation twins in 316 stainless steel, Mater. Sci. Eng., A 694 (2017) 121–131, https://doi.org/10.1016/j.
msea.2017.03.073.
[5] P.L. Mangonon, G. Thomas, The martensite phases in 304 stainless steel, Metall.
Trans. 1 (1970) 1577–1586, https://doi.org/10.1007/BF02642003.
[6] G.B. Olson, M. Cohen, A mechanism for the strain-induced nucleation of martensitic transformation, J. Less Common. Met. 28 (1972) 107–118, https://doi.
org/10.1016/0022-5088(72)90173-7.
[7] F. Lecroisey, A. Pineau, Martensitic transformations induced by plastic deformation in the Fe-Ni-Cr-C system, Metall. Mater. Trans. B 3 (1972) 391–400, https://doi.
org/10.1007/BF02642042.
[8] J.Y. Choi, W. Jin, Strain induced martensite formation and its effect on strain hardening behavior in the cold drawn 304 austenitic stainless steels, Scripta Mater.
36 (1997) 99–104, https://doi.org/10.1016/S1359-6462(96)00338-7.
[9] A. Das, S. Sivaprasad, M. Ghosh, P.C. Chakraborti, S. Tarafder, Morphologies and characteristics of deformation induced martensite during tensile deformation of 304 LN stainless steel, Mater. Sci. Eng., A 486 (2008) 283–286, https://doi.org/
10.1016/j.msea.2007.09.005.
[10] A. Das, S. Tarafder, Experimental investigation on martensitic transformation and fracture morphologies of austenitic stainless steel, Int. J. Plast. 25 (2009) 2222–2247, https://doi.org/10.1016/j.ijplas.2009.03.003.
[11] D. Kaoumi, J. Liu, Deformation induced martensitic transformation in 304 austenitic stainless steel: in-situ vs. ex-situ transmission electron microscopy characterization, Mater. Sci. Eng., A 715 (2018) 73–82, https://doi.org/10.1016/j.
msea.2017.12.036.
[12] P. Hedstr¨om, U. Lienert, J. Almer, M. Od´en, Stepwise transformation behavior of the strain-induced martensitic transformation in a metastable stainless steel, Scripta Mater. 56 (2007) 213–216, https://doi.org/10.1016/j.
scriptamat.2006.10.009.
[13] Y.F. Shen, X.X. Li, X. Sun, Y.D. Wang, L. Zuo, Twinning and martensite in a 304 austenitic stainless steel, Mater. Sci. Eng., A 552 (2012) 514–522, https://doi.org/
10.1016/j.msea.2012.05.080.
[14] G.B. Olson, M. Cohen, Kinetics of strain-induced martensitic nucleation, Metall.
Trans. A 6 (1975) 791, https://doi.org/10.1007/BF02672301.
[15] J. Talonen, H. H¨anninen, Formation of shear bands and strain-induced martensite during plastic deformation of metastable austenitic stainless steels, Acta Mater. 55 (2007) 6108–6118, https://doi.org/10.1016/j.actamat.2007.07.015.
[16] L.E. Murr, K.P. Staudhammer, S.S. Hecker, Effects of strain state and strain rate on deformation-induced transformation in 304 stainless steel: Part II. Microstructural study, Metall. Trans. A 13 (1982) 627–635, https://doi.org/10.1007/BF02644428.
[17] T. Suzuki, H. Kojima, K. Suzuki, T. Hashimoto, M. Ichihara, An experimental study of the martensite nucleation and growth in 18/8 stainless steel, Acta Metall. 25 (1977) 1151–1162, https://doi.org/10.1016/0001-6160(77)90202-4.
[18] J.W. Brooks, M.H. Loretto, R.E. Smallman, In situ observations of the formation of martensite in stainless steel, Acta Metall. 27 (1979) 1829–1838, https://doi.org/
10.1016/0001-6160(79)90073-7.
[19] J. Talonen, H. H¨anninen, P. Nenonen, G. Pape, Effect of strain rate on the strain- induced γ→ α′-martensite transformation and mechanical properties of austenitic stainless steels, Metall. Mater. Trans. 36 (2005) 421–432, https://doi.org/
10.1007/s11661-005-0313-y.
[20] S.S. Hecker, M.G. Stout, K.P. Staudhammer, J.L. Smith, Effect of strain state and strain rate on deformation-induced transformation in 304 stainless steel: Part I.
Magnetic measurements and mechanical behavior, Metall. Mater. Trans., A 13 (1982) 619–626, https://doi.org/10.1007/BF02644427.
[21] K.P. Staudhammer, L.E. Murr, S.S. Hecker, Nucleation and evolution of strain- induced martensitic (bcc) embryos and substructure in stainless steel: a transmission electron microscope study, Acta Metall. 31 (1983) 267–274, https://
doi.org/10.1016/0001-6160(83)90103-7.
[22] P.J. Ferreira, J.B. Vander Sande, M.A. Fortes, A. Kyrolainen, Microstructure development during high-velocity deformation, Metall. Mater. Trans. 35 (2004) 3091–3101, https://doi.org/10.1007/s11661-004-0054-3.
[23] J.A. Lichtenfeld, C.J. Van Tyne, M.C. Mataya, Effect of strain rate on stress-strain behavior of alloy 309 and 304L austenitic stainless steel, Metall. Mater. Trans. 37 (2006) 147–161, https://doi.org/10.1007/s11661-006-0160-5.
[24] W.S. Lee, C.F. Lin, Morphologies and characteristics of impact-induced martensite in 304L stainless steel, Scripta Mater. 43 (2000) 777–782, https://doi.org/
10.1016/S1359-6462(00)00487-5.
[25] S. Curtze, V.-T. Kuokkala, Dependence of tensile deformation behavior of TWIP steels on stacking fault energy, temperature and strain rate, Acta Mater. 58 (2010) 5129–5141, https://doi.org/10.1016/j.actamat.2010.05.049.
[26] M. Isakov, S. Hiermaier, V.-T. Kuokkala, Effect of strain rate on the martensitic transformation during plastic deformation of an austenitic stainless steel, Metall.
Mater. Trans. 46 (2015) 2352–2355, https://doi.org/10.1007/s11661-015-2862-z.
[27] N.I. V´azquez-Fern´andez, T. Nyyssonen, M. Isakov, M. Hokka, V.-T. Kuokkala, ¨ Uncoupling the effects of strain rate and adiabatic heating on strain induced martensitic phase transformations in a metastable austenitic steel, Acta Mater. 176 (2019) 134–144, https://doi.org/10.1016/j.actamat.2019.06.053.
[28] G. Reisner, E.A. Werner, F.D. Fischer, Micromechanical modeling of martensitic transformation in random microstructures, Int. J. Solid Struct. 35 (1998) 2457–2473, https://doi.org/10.1016/S0020-7683(97)00149-2.
[29] V.I. Levitas, A.V. Idesman, G.B. Olson, Continuum modeling of strain-induced martensitic transformation at shear-band intersections, Acta Mater. 47 (1998) 219–233, https://doi.org/10.1016/S1359-6454(98)00314-0.
[30] F.D. Fischer, G. Reisner, E. Werner, K. Tanaka, G. Cailletaud, T. Antretter, A new view on transformation induced plasticity (TRIP), Int. J. Plast. 16 (2000) 723–748, https://doi.org/10.1016/S0749-6419(99)00078-9.
[31] V.I. Levitas, A.V. Idesman, G.B. Olson, E. Stein, Numerical modelling of martensitic growth in an elastoplastic material, Philos. Mag. A 82 (2002) 429–462, https://doi.
org/10.1080/01418610208239609.
[32] F. Lani, Q. Furn´emont, T.V. Rompaey, F. Delannay, P.J. Jacques, T. Pardoen, Multiscale mechanics of TRIP-assisted multiphase steels: II. Micromechanical modelling, Acta Mater. 55 (2007) 3695–3705, https://doi.org/10.1016/j.
actamat.2007.02.015.
[33] H.K. Yeddu, V.I. Razumovskiy, A. Borgenstam, P.A. Korzhavyi, A.V. Ruban, J. Ågren, Multi-length scale modeling of martensitic transformations in stainless steels, Acta Mater. 60 (2012) 6508–6517, https://doi.org/10.1016/j.
actamat.2012.08.012.
[34] H.K. Yeddu, T. Lookman, A. Saxena, Strain-induced martensitic transformation in stainless steels: a three-dimensional phase-field study, Acta Mater. 61 (2013) 6972–6982, https://doi.org/10.1016/j.actamat.2013.08.011.
[35] H. Mirzadeh, A. Najafizadeh, Correlation between processing parameters and strain-induced martensitic transformation in cold worked AISI 301 stainless steel, Mater. Char. 59 (2008) 1650–1654, https://doi.org/10.1016/j.
matchar.2008.03.004.
[36] N. Vazquez Fernandez, M. Isakov, M. Hokka, V.-T. Kuokkala, Effects of adiabatic heating estimated from tensile tests with continuous heating, in: Conference Proceedings of the Society for Experimental Mechanics Series, 2018, https://doi.
org/10.1007/978-3-319-62956-8_1. Cham.
[37] N. Gey, B. Petit, M. Humbert, Electron backscattered diffraction study of ε/α′ martensitic variants induced by plastic deformation in 304 stainless steel, Metall.
Mater. Trans. 36 (2005) 3291–3299, https://doi.org/10.1007/s11661-005-0003-9.
[38] Y.B. Das, A.N. Forsey, T.H. Simm, K.M. Perkins, M.E. Fitzpatrick, S. Gungor, R.
J. Moat, In situ observation of strain and phase transformation in plastically deformed 301 austenitic stainless steel, Mater. Des. 112 (2016) 107–116, https://
doi.org/10.1016/j.matdes.2016.09.057.
[39] N.I. Vazquez-Fernandez, G.C. Soares, J.L. Smith, J.D. Seidt, M. Isakov, A. Gilat, V.
T. Kuokkala, M. Hokka, Adiabatic heating of austenitic stainless steels at different strain rates, J. Dyn. Behav. Mater. 5 (2019) 221–229, https://doi.org/10.1007/
s40870-019-00204-z.
[40] D. Rafaja, C. Ullrich, M. Motylenko, S. Martin, Microstructure aspects of the deformation mechanisms in metastable Austenitic steels, in: H. Biermann, C.
G. Aneziris (Eds.), Austenitic TRIP/TWIP Steels and Steel-Zirconia Composites:
Design of Tough, Transformation-Strengthened Composites and Structures, Springer International Publishing, Cham, 2020, pp. 325–377, https://doi.org/
10.1007/978-3-030-42603-3_11.
