1. Introduction
Austenitic stainless steel (ASS) has excellent corrosion resistance and processability and is widely used in the biomedical, electronic, chemical, electrical power, food, and nuclear industries. Besides that, the high demand for microparts has received much attention in recent decades. Furthermore, microforming technology has many issues, such as limitations in material applications and the requirement for high-cost mass production. When a plastic deformation is applied to the ASS, martensitic-induced transformation occurs in ASS. The martensitic phase volume fraction (Mf) increases in proportion to the increase in plastic deformation [
1,
2,
3]. Martensitic phase transformation (MPT) decreases the toughness but increases the strength of ASS [
4,
5]. ASS has a body center tetragonal (BCT) crystal structure because of the plastic deformation in ASS, which involves a face center cubic austenite phase at room temperature. Plastic instability, which can occur in thin metal foils, could be delayed, since the work hardening increases because of the MPT occurring in ASS. This means that MPT increases the strength and ductility of ASS [
6]. The existence of MPT is affected by the chemical composition, strain level, and grain size (Dg) of ASS thin foils [
7,
8,
9,
10,
11]. Xue et al. [
1] and Qin et al. [
12] found that the deformation in a stainless steel strip affects the Mf. The occurrence of the Mf can be controlled by the deformation of the stainless steel strip. After the plastic deformation, the occurrence of MPT in the surface is higher, compared to the interior of the stainless steel strip. When ASS is subjected to plastic deformation MPT, dislocation interaction, and twinning are formed. Twinning occurs more often on the surface, compared with the interior of ASS. The stacking fault energy (SFE) and austenitic phase increase as the temperature rises, thus limiting the MPT. The temperature rise is due to the increasing strain rate. The low strain and low temperature reduce the SFE and the stability of the austenite. This is why the MPT occurs more easily on the surface than on the interior. Furthermore, surface roughening also occurs more easily on the surface than on the interior of thin metal foils. An investigation on the correlation between MPT and surface roughening behavior in thin metal foils is required. Peng et al. [
13] and Qin et al. [
12] concluded that martensite transformation occurs because of the increase in the strain rate, which increases the local temperature. A high strain rate suppresses MPT. The Olson and Cohen model, called the OC model, represents the basic strain-induced MPT and explains the kinetic energy occurring in conjunction with MPT. The OC model explains that the martensitic nucleation occurs because of the shear band intersection during plastic deformation. Tomita et al. [
14] found that MPT nucleation occurs because of the shear band intersection during plastic deformation. When the strain rate increases, the shear band intersection, as the site of MPT nucleation, increases, but the probability of an MPT embryo decreases. This conclusion is considered to apply only to a constant temperature, but high temperatures caused by self-heating in tensile tests were ignored. Zandrahimi et al. [
15] found that the transformation of the austenitic phase into MPT in AISI 304 affects surface hardening, leading to the deterioration of wear resistance. An investigation of the surface roughness affected by MPT is required, because surface roughening is caused by grain deformation on the surface and affects the surface properties of thin metal foils. Qin et al. [
12] concluded that a high strain rate and temperature decrease the Mf, strength of thin foils, and ductility. Jeom et al. [
16] concluded that ε martensite occurs as a consequence of austenitic phase transformation in duplex stainless steel because of plastic deformation, which appeared when it was subjected to a low strain. The austenite phase transformed into ά martensite, which appeared when it was subjected to a high strain.
However, in the metal-forming and miniaturization processes, the prevailing phenomena, when the thickness is decreased the ratio of surface roughness to thickness increases, and this is called a non-uniformity thickness [
17,
18,
19]. The surface-roughening phenomena of sheet or thin material affect the mechanical properties of thin metal foils, such as the fracture behavior and necking phenomena. Surface roughening on a free surface occurs because of inhomogeneous grain deformation in polycrystalline metals. Thus, surface roughening is very important in the field of microforming technology using thin or sheet metal foils. Surface roughening affects the size effect of thin metal foils [
20]. The size effect affects the fracture behavior and flow stress and becomes more significant when the number of grains is decreased, especially in the fracture area [
21]. Meng et al. [
22] and Cheng Cheng et al. [
23] concluded that when the thickness of thin foil is uniform, the effect of free surface roughening on the failure behavior and flow stress is significant during plastic deformation. Stoudt et al. [
24] concluded that the increase of surface roughness depends on the Dg of Mg–Al alloy. Furushima et al. [
25] found that a decreasing thickness from 0.3 to 0.1 mm decreases the fracture strain in several kinds of sheet or thin metal foils, such as titanium and copper. Plastic deformation subjected to thin or sheet metal increases the surface roughness to thickness ratio. Rabee et al. [
26] evaluated the relationship between local microstructure and deformation-induced surface roughness, which needs to be clarified. Furushima et al. [
27] concluded that dimples do not occur on thin foils, such as copper, with a thickness of ≤0.1 mm. Surface roughening is considered to be one factor that has a huge effect on fracture behavior in copper thin foils when the thickness is ≤0.1 mm. When the thickness decreased in the same area, the surface roughness increased under uniaxial deformation. This means that when the number of grains decreases, the surface roughness increases because of the uniaxial tensile test at the same strain level. In addition, when the quantity of Dg is at least five, the fracture strain is low, and the surface roughness significantly increases under the uniaxial tensile test at the same strain level. Based on this conclusion, it could be predicted that when the Dg increases in the constant area, the surface roughness will increase because of the uniaxial stress state at the same strain level. The surface roughening phenomena with different types of Dg, besides copper metal, with thickness at least 0.1 μm, need to be investigated. Lei Zhang et al. [
28] made face center cubic (FCC) polycrystalline metals the object of their recent study on the evolution of surface roughness. Besides the FCC structure, it is very important to study surface roughening. Kengo Yoshida et al. [
29] found that surface roughness is mainly affected by Dg. When the Dg increases, the surface roughness to thickness (Ng) ratio decreases. The effect of surface roughness in metal foils with a Dg lower than 10 µm needs to be investigated. The role of surface roughness is very significant when the ratio of thickness to grain size decreases. Surface roughness becomes a defect when the ratio thickness to grain size decreases in thin metal foils. It is very important to study thin metal foils with a Dg below 10 μm. Shimizu et al. [
30] concluded that grain deformation affects surface roughness behavior. Furthermore, a different single grain deformation increases the surface roughness in sheet metal. It is very important to investigate the surface roughness with different types of Dg, which may have grain misorientation after deformation. Linfa Peng et al. [
31] found that with the increase of Dg, the density in the grain boundary decreases. Since the grain boundary density is not homogenous, inhomogeneous deformation occurs. It is very important to investigate surface roughening phenomena with a uniform Dg and a Dg of different sizes in thin metal foils.
Furushima et al. [
32] used pure copper C 10220-O, with a thickness of 0.05 mm. Surface-roughening phenomena occur because of a weak grain deformation. The inhomogeneous grain in polycrystalline metal shows the roughening phenomena. A weak grain will cause a lower flow stress in the grain. It is possible to study these phenomena without considering C 10220-O and rather focusing on weak and strong grains. P. Groche et al. [
33] found that by decreasing the Dg, the surface asperities become lower in fine grain because of the higher yield point, compared to coarse grain. The increasing and decreasing of the yield point are in line with the Hall–Petch formula. Cheng et al. [
23] found that surface roughening is a single factor that determines the failure of thin metal foils. It is very important to investigate surface roughening in thin metal foils, which may be affected by grain size.
According to previous research, a weak grain deformation has a huge effect on surface roughening. The higher strain level applied to thin metal foils, such as ASS SUS 304 and SUS 316, increased the surface roughness. When the number of grains equals at least five, the surface roughness increases significantly because of the uniaxial tensile stress state. Surface roughness with a constant thickness and various sizes of Dg has not yet been investigated. The correlation between surface roughness and MPT in a grain is still not clear. It can be predicted that if the grain has a high MPT, the grain strength will increase, and grain deformation will become more difficult, compared to a grain with a lower MPT. The grains of SUS 304 or SUS 316 with a lower MPT change into weaker grains, compared to grain with a higher MPT. The purpose of this study is to understand and briefly clarify how MPT affects grain strength, which will be shown by determining the surface roughness behavior of thin metal foils. In this study, the authors used SUS 304 and SUS 316 thin metal foils with various sizes of Dg. SUS 304 and SUS 316 were used to clarify how MPT affects the surface roughness both in SUS 304, which consists of a complicated phase, and SUS 316, which consists of a more uniform phase. No study has yet been conducted on the surface-roughening behavior and phase transformation in various thin metal foils with a FCC and body center cubic (BCC) structure and with different sizes of Dg below 10 μm. First, the aim of this study is to investigate how MPT affects the surface roughening behavior in SUS 304 and SUS 316 thin metal foils. The second aim is to investigate the effect of a fine Dg on surface-roughening behavior, compared to a coarse Dg. In this study, ASS SUS 304 and SUS 316 are subjected to a uniaxial tensile stress state step by step, the surface roughening behavior is measured, and the MPT, local grain misorientation, and deformation are analyzed using Scanning Electron Microscope–Electron Backscaterr Diffraction (SEM-EBSD). Materials and Research Methods
2. Materials and Research Methods
2.1. Materials
Thin metal foils SUS 304 and SUS 316 were heat treated and rolled into a 0.1 mm thickness. The thin metal foils were obtained from Komatsu Seiki Koshakuso Co. Ltd., Suwa City, Nagano, Japan. The chemical compositions of the material were detected by an X-ray fluorescence spectrometer, as listed in
Table 1 and
Table 2. Different compositions of stainless steel SUS 304 and SUS 316 thin foils materials were used so that the effect of chemical composition, such as nickel element chromium element and carbon element, on the MPT could be investigated. It is well known that the nickel element is an austenite stabilizer, and the chromium element is a ferrite stabilizer. This study investigates how the austenitic stabilizer affects the MPT induced by plastic deformation. Furthermore, based on the chemical composition, SUS 304 thin metal foils consist of more complicated phases in their microstructure than SUS 316 thin metal foils. The microstructure affects the MPT formation and the occurrence of surface roughening.
According to the chemical composition, shown in
Table 1 and
Table 2, the quantity of carbon in SUS 304 is higher than that in SUS 316. This means that the quantity of martensite phase in SUS 304 will be higher than in SUS 316 because of plastic deformation. The martensitic phase and carbide compound may form in the grain matrix, which will increase the strength of thin foils. Because of plastic deformation or heating and quenching to which the thin metal foils of SUS 304 are subjected, they will affect the MPT and Mf. Plastic deformation affects the dislocation motion. The dislocation motion in the slip band of lattice crystal changes the crystal structure from FCC to BCT (body center tetragonal). The quantity of nickel in SUS 316 is almost 4–5% higher than that in SUS 304 thin metal foils. From this, it could be inferred that the ductility and toughness of SUS 316 is higher than those of SUS 304, because nickel is an austenitic former element that has mechanical properties such as strength and ductility. Moreover, the quantity of the chromium element, such as the ferrite former, in SUS 304 is higher than that in SUS 316 thin metal foils. This means that the chromium carbide more easily forms in the grain matrix or boundary and increases the strength of thin metal foils when plastic deformation is applied. The higher the nickel elements in stainless steel, the more difficult the MPT or the more difficult the transformation of austenite into martensite, because nickel is an austenitic stabilizer—it needs more energy to change the austenitic phase into martensite.
This study used different sizes of Dg to investigate the surface roughness behaviors with various sizes of Dg. It is known that surface roughness increases proportionally with a grain larger than 10 µm in size, when plastic deformation is applied to thin metal foils. However, it is still not clear yet how the surface roughness behaves in thin metal foils with a Dg lower than 10 µm, especially in SUS 304 and SUS 316, which have a BCC and FCC crystal structure. It is very important to clarify the surface roughness behavior and mechanism in thin metal foils SUS 304 and SUS 316 with various sizes of Dg lower than 10 µm, such as 0.5, 1.0, 1.4, 1.5, 3.0, and 9.0 µm, in size.
Figure 1 shows a typical microstructure with various sizes of Dg, from fine grain to coarse grain. According to the Hall–Petch equation, it is well known that Dg affects the mechanical properties of materials. Materials that have a fine grain will have a higher tensile and yield strength than materials that have a coarse grain. The surface roughness in thin metal foils of SUS 304 and SUS 316 will be different, because the thin metal foils with various sizes of Dg are subjected to the same tensile stress state.
The entire dimension shown in
Figure 2a is in millimeters. Based on
Figure 2a, the surface roughness behavior is investigated in points A, B, C, and D. The purpose of measuring the surface roughness in various areas is to compare the surface roughness behavior in thin metal foils because of plastic deformation and its mechanism. The thin metal foils shown in
Figure 2a are of commercial SUS 304 and SUS 316 with a 4.0 mm width, 0.1 mm thickness, and 20.0 mm gauge length. The standard used is DIN 50125. The sample made of the dog bone type is shown in
Figure 2a.
2.2. Method
Before the materials are subjected to the uniaxial tensile stress state, they are cleaned using ethanol and combined with ultrasonic vibration for 30 min to increase the cleaning of the surface. The samples are subjected to the uniaxial tensile stress state for five steps, with constant strain. After the samples are subjected to the uniaxial tensile stress state, the surface roughness is measured using a confocal laser microscope, the Keyence Confocal Laser Microscope (VE 8800, Keyence Co., Japan). The uniaxial tensile test was conducted using a commercial tensile test machine, Autograph AG-IS 50 kN, produced by Shimadzu Corp., Japan, with capacity of 50 kN.
The surface roughening behavior of SUS 304 and SUS 316 with various sizes of Dg were investigated using a uniaxial tensile test. The uniaxial tensile test was run step by step using a constant strain over the yield point of the thin metal foils. The gauge length of the thin metal foils was 20 mm. The width and thickness of the thin foil were 4 mm and 0.1 mm, respectively. The bevelled radius was 3.6 mm. In order to maintain consistency, the uniaxial tensile test was conducted using the commercial tensile test machine, Autograph AG-IS 50 kN (Shimadzu Corp., Japan). The strain rate of the uniaxial tensile test was 1.6 × 10−3 s−1. The measured and observed surface roughness used different sizes of Dg, different materials, and a constant thickness. The elongation was measured optically with a video non-contact extensiometer (DVE-201, Shimadzu Corp., Japan), since the contact extensiometer could not be affixed to the metal foil. The surface roughness during the deformation was measured using a uniaxial tensile testing machine that halted at every step. In the universal tensile testing machine, the sample was subjected to a uniaxial tensile stress state at each step, and the tensile test was stopped. Then, the sample was taken out from the clamp tensile testing machine, and the surface roughness behavior was measured using a Keyence Confocal Laser Microscope (VE 8800, Keyence Co., Japan). The area of the surface roughness measurement was in the center points of A, B, C, and D, according to the direction of the rollers at each step using contact surface roughness measurement. The surface roughness was measured for five steps in the same position, with constant strain. Besides the surface roughness, the Rz value was evaluated. The area of the surface roughness measurement had a length of 0.7 mm.
The microstructure behavior was investigated using the SEM SU-70, produced by the Hitachi High Technology corporation, in normal mode. The acceleration voltage was 5 kV, the emission current 16 μA, and the working distance 10 mm. The phase transformation was investigated using an EBSD Digi View (EDAX) in field-free mode, with an acceleration voltage of 15 KV, emission current of 16 μA, and working distance of 20 mm. The sample was observed in a 30 μm × 50 μm area. The step (resolution) of the EBSD machine is 0.1 μm, and the pixel binning is 8 × 8.
The Dg was measured using the JIS G0551 standard. The technical methods for measuring Dg were to draw a line (length is L) on a photomicrograph, as shown in
Figure 2b. Then, the number of crystals on the line were counted, and the end of the line was counted as ½.
Figure 2b is an analogy of the crystal structure photomicrograph. Last, we calculated the Dg using the following formula. Average grain size =
.