Hot-extruded Cu-Zn-Si alloy bars with various chemical compositions were cold-caliber rolled down to 91.7% reduction at maximum. Heterogeneous nano-structure, in which coarse initial grains were subdivided mainly by mechanical twins and shear bands, was gradually developed with increasing reduction. The as-rolled bars exhibited an extraordinarily high tensile strength of 988 MPa at best with a reasonable ductility of 5.9%. The tensile strength was further slightly raised to 995 MPa by low-temperature annealing. 3D atom-probe analyses revealed dense Si segregation at twin boundaries and increase in the amount of segregation after annealing. Multi-scale simulation indicated that strain field formed by the Si segregation at twin boundary obstructed dislocation glide to develop geometrically necessary dislocations, which resulted in higher yield stress and work-hardening rate to cause higher tensile strength. It was found that strengthening mechanisms of heterogeneous nano-structured Cu-Zn-Si alloy bars are, therefore, complicatedly combined ones of grain refinement, work hardening, solid-solution hardening and grain-boundary segregation.

This Paper was Originally Published in Japanese in J. Japan Inst. Copper 63 (2024) 56–64.

Fig. 8 Concentration profiles of the main component elements in the Cu-20Zn-1.0Si modified alloy sample analyzed by means of 3D atom probe field ion tomography at around twin boundaries in the (a) as-91.7%-rolled and (b) peak-aged samples. Small amount elements added, Ni, P and Sn, were negligibly small to show in the above profiles. Broken arrows indicate the positions of twin boundaries.

In order to improve various properties of metals and alloys, research on ultrafine grains (UFGs) by severe plastic deformation (SPD) has been actively conducted around the world. As a result, it has been revealed that UFGs can improve not only mechanical properties but also variety of other physical properties [1–5]. However, the most SPD methods involve repeated processing by “shape-invariant processes”, which limit the applicable sample size, and the processing method is also complex, making it unsuitable for industrial mass production. And, therefore, alternative new methods have been demanded.

In recent years, Miura et al. have fabricated a complex “heterogeneous nano-structure (HN-structure)” comprising shear bands, deformation twins, and lamellar structures by means of simple cold rolling of extremely low stacking-fault energy (SFE) alloys, and have demonstrated that their mechanical properties are equal to or better than those processed by SPD [6–8]. HN-structures can be easily fabricated by using conventional rolling machine and applicable to long plate materials for industrial production, so they have attracted particular attention in recent years.

Copper alloys are ones of the low SFE materials, and the SFE value can be further reduced by adding elements. The lower the SFE, the easier the deformation twinning becomes [9], and it becomes easier to manufacture long sheets comprising UFGed structures with twin widths of several tens of nanometers [7, 8, 10]. Although the research on HN-structured copper alloys is lagging behind HN-structured stainless steels, strengths more than 1.5 times higher than the conventional mechanical properties have been achieved [8, 10]. However, these studies are on copper alloys as conductive functional materials, and there are few studies on them as structural materials.

In this study, we perform cold-caliber rolling on Cu-Zn-Si alloy bars with six different chemical compositions. The value of the SFE is reduced by addition of Si to the Cu-Zn alloy. And the effects of the amount of Zn and Si addition as well as area reduction on the microstructure and mechanical properties via caliber rolling followed by low-temperature annealing are systematically investigated. Furthermore, the essence of the strengthening mechanisms of the HN-structure is clarified by comparing the results of experiments and multi-scale simulations.

The specimens were 11.7%, 57.6%, and 72.1% caliber-rolled bars with a rhombus-shaped cross section (Fig. 1(a)) fabricated from hot-extruded bars of 20.0 mmϕ of Si-added Cu-Zn alloys, which were provided by San-Etsu Metals Co. Ltd. The chemical compositions are listed in Table 1. For simplicity, for an example, Cu-10Zn-2.5Si (mass%) alloy is hereafter referred to as 10Zn-2.5Si. In the experiment, the bars with a reduction of area of 72.1% were subjected to additional caliber rolling to reduce the area to 91.7% (cross section is flattened hexagonal, Fig. 1(b)). The 91.7% rolled bars were, then, annealed at temperatures between 523 K and 623 K to increase their strength. Vickers hardness tests and tensile tests using an Instron-type universal mechanical testing machine were performed on these samples to investigate their mechanical properties. The tensile tests were conducted at room temperature at an initial strain rate of $\dot{\varepsilon } = 1.4 \times 10^{ - 3}$ s⁻¹. Due to the size constraints of the bars after rolling, the gauge sizes were set to 4.0 × 1.5 × 0.5 mm³ or 6.0 × 3.0 × 0.5 mm³ (Fig. 2). Microstructural observations were carried out using a scanning electron microscope equipped with an orientation imaging microscope (OIM) and a transmission electron microscopy (TEM). About definition of the observation directions, it is described in Fig. 1. In addition, some samples were analyzed using a 3D atom probe tomography (analysis was outsourced to Toshiba Nanoanalysis Co. Ltd. and carried out using Ametech LEAP4000XSi) to measure the amount of grain-boundary segregation (GBS) of the added elements. Based on the crystal orientation information and data of GBS in the HN-structure, multi-scale simulation was performed to evaluate the effect of GBS on the mechanical properties of the HN-structure. The details of the strengthening mechanisms of the HN-structure were, then, discussed through both experimental and computational science approaches.

Fig. 1

Definition of directions of the caliber-rolled bars after reductions of (a) 11.7%, 57.6%, 72.1% and (b) 91.7%, where RD, ND1 and ND2 are the rolling direction and the longer and shorter axes normal to RD, and SD, LD, DD1 and DD2 are the shorter, longer and diagonal directions normal to RD.

Table 1 Chemical compositions of supplied Cu-Zn-Si alloys in mass%. Very small impurities less than 0.01 mass% are not shown here. The modified sample of 20Zn-1.0Si was also prepared by additions of Ni and P. Values of stacking fault energy (SFE) were evaluated using a method proposed by Denanot and Villain [11]. An asterisk symbol indicates that stacking fault energy was estimated ignoring the effects of small addition of Ni and P. Some of the initial grain sizes before rolling are also shown.

Fig. 2

Schematic illustrations of tensile specimens for the bars with reductions of (a) 11.7%, 57.6%, 72.1% and (b) 91.7%.

The microstructural changes during caliber rolling were observed using OIM. As an example, the microstructural changes of 20Zn-0.5Si and 30Zn-0.5Si are displayed in Fig. 3. It can be seen that the microstructure changes gradually with increasing area reduction. That is, i) at the area reduction of 11.7%, the initial structure remains still almost equiaxed, but fine structures that are identified as deformation twins by OIM begin to be formed within some grains, ii) at the area reduction of 57.6%, the initial grains elongate along the rolling direction (RD) and deformation twins are introduced within all grains, and iii) at the area reduction of 91.7%, shear bands develop in addition to dense deformation twins. However, at the regions of lower reduction, the changes in the microstructure appear inhomogeneous, and the fine-grained structure develops earlier especially near the surface. Therefore, the large difference in the twin distribution in the individual grains at the area reduction of 11.7% (Figs. 3(a), (g)) should be affected by the non-uniform development of the microstructure. In addition, at the regions of higher area reduction, the microstructure was too fine to analyze using OIM with the limited resolution. On the other hand, it is interesting to note that the weak {101} texture along circumferential direction of the round bar developed by hot extrusion did not change significantly even with increasing area reduction by caliber rolling. This is because the processing by caliber rolling, which reduces the area with rotating by 90 degrees at each pass of rolling, and the dense introduction of deformation twins and shear bands accompanied by large changes in the crystallographical orientation prevented the evolution of a sharp {101} rolling texture.

Fig. 3

Evolved microstructures at the center of bars of (a)∼(f) Cu-20Zn-0.5Si and of (g)∼(l) Cu-30Zn-0.5Si alloys caliber rolled to some reductions of (a), (g) 11.7%, (b), (h), 57.6% and (c), (i), 91.7%. (d), (e), (f), (j), (k), (l) are their corresponding inverse-pole figures with the maximum intensities. Observation was carried out either along ND1 or SD. The definition of observation directions is described in Fig. 1. (online color)

A more detailed OIM observation of the Cu-20Zn-1.0Si after 91.7% area reduction is exampled in Fig. 4. From Fig. 4(a), it was confirmed that the RD had crystallographical orientations from ⟨111⟩ to ⟨001⟩, and especially ⟨111⟩ had a strong accumulation with a maximum intensity of 7.0. The accumulation of ⟨111⟩ along RD became sharper with increasing area reduction, and this was observed in all the bars independent of chemical compositions. On the other hand, as shown in Fig. 4(b), at the surface normal to longer diagonal direction (LD) and normal to RD (Fig. 1(b)), which underwent locally and comparably large plastic deformation by the final rolling pass, showed a relatively strong accumulation of ⟨101⟩. Observation of the plane normal to SD (Fig. 4(d)) at higher magnification confirmed the introduction of a high density of Σ3 twin boundaries.

Fig. 4

Evolved microstructures in the Cu-20Zn-1.0Si alloy bar after 91.7% reduction, which were observed using OIM from different directions; (a) RD, (b) LD, and (c) SD. (d) is the imposed image of the area surrounded by dotted-square line in (c) showing Σ3 twin boundary distribution by red lines. (online color)

Examples of microstructures observed by TEM are shown in Fig. 5. The development of lamellar structure elongated in the direction of approximately 35 to 45 degrees from RD was observed, and these were judged to be composed of ultrafine deformation twins from the SAD pattern analysis (Fig. 5(b)). The twin-boundary spacings were not uniform, ranging from 5 to 200 nm, with an average of approximately 40 nm. In addition, the development of shear bands dividing the lamellar structure was confirmed in Fig. 5(c), and its interior was composed of somewhat equiaxed UFGs. The SAD pattern corresponding to the circle of the shear band showed a ring shape, and it was possible to determine that the grain boundaries of the UFGed structure that constitutes the shear band had medium to high angles. In addition, the formation of even finer secondary twins was observed inside the deformation twins that developed in a lamellar shape, as indicated by the arrows in Fig. 5(d), and the development of a hierarchical HN-structure was confirmed. The development of a similar microstructure was observed in all samples. This observation suggests that there is no significant difference in the formed microstructure at the large area reduction of 91.7% due to the small difference in SFE of the Cu-Zn-Si alloys employed in the present study. The HN-structure developed in the caliber-rolled bars was, yet, significantly different from that formed in the heavily flat-rolled sheets [6, 7, 10, 12], and the particularly characteristic “eye-shaped” twin domains were not formed.

Fig. 5

Typical TEM photographs of (a), (b) Cu-20Zn-1.0Si and (c), (d) Cu-30Zn-0.5Si (mass%) alloys caliber-rolled to 91.7%. SAD patterns in (b) and (c) were taken from the areas in dotted circles. An arrow in (d) indicates secondary twins.

Tensile test of each caliber-rolled bar was carried out at room temperature. Examples of the stress-strain curves are exhibited in Fig. 6. However, failure occurred at an area reduction of less than 72.1% in some alloy compositions. So those results are not included. The results of the tensile tests are summarized in Table 2. The high-Si-added alloy bars tended to break during rolling, which is thought to be due to the influence of inclusions that formed during casting and remained even after the subsequent heat treatment. On the other hand, the high-Si-added alloy bars showed the highest tensile strength at the low and medium area reduction regions. At the area reduction of 91.7%, the 20Zn-1.0Si and 20Zn-1.0Si modified achieved almost the same tensile strengths of 975 MPa and 988 MPa. However, no clear correlation was found between the SFE value and tensile strength.

Fig. 6

Samples of flow curves attained by tensile tests; (a) Cu-10Zn-2.5Si, (b) Cu-20Zn-1.0Si and (c) Cu-20Zn-1.0Si modified alloys (mass%). (online color)

Table 2 Summarized mechanical properties attained by tensile tests; yield strength (σ_Y), ultimate tensile strength (σ_UTS), fracture strain (ε_F). Asterisk symbols indicate the highest values achieved at each reduction.

In order to further increase the strength, the 91.7% rolled bars were annealed at low temperatures of 523 K, 573 K, and 623 K [10], and annealing-softening curves were plotted, after which tensile tests were conducted. Due to space limitations, as an example, only the annealing-softening curves at 573 K and the results of the tensile tests of the 20Zn-0.5Si bars annealed at 573 K are displayed in Fig. 7. Table 3 summarizes the results of all the tensile tests before and after annealing. At the early stages of annealing, there was 10–15% increase in hardness, followed by rapid softening. By the annealing to the maximum peak hardness at around 55 s, the tensile strength increased by 5%. However, it dropped rapidly because of occurrence of recrystallization by the prolonged annealing. While precipitation due to annealing was not confirmed, it is assumed that the temporary increase in strength is due to the GBS of added elements [6, 10]. This will be described in detail later. Using 3D atom probe tomography, the amount of GBS and the distribution of the added elements before and after annealing were investigated. The results are shown in Fig. 8. The grain boundaries investigated were “twin boundaries”, which have a high distribution density and are easy to analyze as the “same” type of grain boundary. As is clear from Fig. 8(a), a large amount of Si was already segregated at the twin boundaries after rolling, with the maximum segregation amount close to 7 atm% and the segregation width being more than 5 nm. Annealing made the GBS of Si more obvious and, hence, the difference in the concentration of Si in the matrices and at the twin boundaries became larger. The maximum amount of Si segregation at the grain boundaries in Fig. 8(b) was nearly 9 atm%. The reason why the difference in the maximum amount of segregation concentration before and after annealing was not so significant is because the annealing time for diffusion was too short. No GBS of the other impurities was detected. Although the segregation of the added elements to other low-angle lamellae or small to medium-angle boundaries was not investigated, it is assumed that higher concentrations of segregation of the elements should occur [6, 19].

Fig. 7

(a) Changes in hardness during annealing of 91.7%-caliber-rolled samples at 573 K and (b) the results of tensile tests of Cu-20Zn-0.5Si alloy bar before and after annealing. (online color)

Table 3 Summarized mechanical properties, yield strength (σ_Y), ultimate tensile strength (σ_UTS) and fracture strain (ε_F), achieved by tensile tests of 91.7%-rolled samples before and after annealing at 573 K for various periods of time to peak hardness, beginning of rapid softening (optional (opt.)) and full softening. Asterisk symbols indicate the highest values of the mechanical properties achieved at each annealing condition.

Fig. 8

Concentration profiles of the main component elements in the Cu-20Zn-1.0Si modified alloy sample analyzed by means of 3D atom probe field ion tomography at around twin boundaries in the (a) as-91.7%-rolled and (b) peak-aged samples. Small amount elements added, Ni, P and Sn, were negligibly small to show in the above profiles. Broken arrows indicate the positions of twin boundaries.

Cu-Zn-based alloys with extremely low SFE (6–13 mJ/m²) due to Si addition were caliber rolled at ambient temperature to produce structural Cu-Zn-Si alloy bars. With increasing area reduction, the initial coarse grains were subdivided mainly by deformation twins and shear bands, and an ultrafine and hierarchical HN-structures were developed. However, the feature was significantly different from that of the HN-structure containing “eye-shaped” twin domains formed by flat-roll rolling [6, 7, 10, 12], and the ultrafine-lamellar structure was complexly intersected from each other. This would be because of the successive change in the compression direction caused by the 90° rotation of the sample at each rolling pass, which introduced various variants of deformation twins. Zhou et al. reported the occurrence of multiple twins in a Cu-26Zn-2Si (mass%) alloy with an extremely low SFE (7 mJ/m²) after caliber rolling, which is consistent with the microstructure observed in the present study (Figs. 3, 4, 5) [13]. According to Narita and Takamura [9], the critical twinning stress τ decreases proportionally with the decrease in SFE, and is expressed as follows;

where γ_SF is the SFE value and b_T is the Burgers vector of the partial dislocation that triggers twinning. The SFE value of the Cu-Zn-Si alloy used in the present study is 6 to 13 mJ/m² (Table 1), which is about 1/2 to 1/3 of that of general 70/30 brass [14]. In other words, twinning can occur by lower applied stress of about 1/2 to 1/3 or even in grains having smaller Schmid factors. Even in the grains where deformation twins could not be introduced in the previous rolling pass, the Schmid factor changed significantly by the subsequent 90° rotation of the bars, and it is believed that deformation twins could be introduced by the next caliber rolling, resulting in the formation of a complex and ultrafine-lamellar structure. However, since systematic and detailed TEM observation of each sample was difficult, further research is required in the future regarding the twin density and twin/grain boundary spacing for each chemical composition.

Twin boundaries also act as an obstacle to the glide motion of dislocations, and it is known that strength increases with decreasing twin spacing according to the Hall-Petch relation [15–17]. These reports are consistent with the report by Flinn et al. that deformation resistance is proportional to the twin boundary density [18]. However, even when considering the effect of solid solution strengthening due to differences in the amounts of Zn and Si added, it is difficult to organize the mechanical properties summarized in Table 2 simply by SFE values. This point will be discussed in 4.2 below.

High concentration of Si segregation to twin boundaries was confirmed in the caliber-rolled bar, and the amount of segregation increased further by low-temperature annealing (Fig. 8), and the tensile strength was also increased. Miura et al. have shown that the tensile strength of HN-structured SUS316LN austenitic stainless steel increases by more than 300 MPa by low-temperature annealing, and that the increase in strength is due to the increase in the amount of segregation of trace elements to twin boundaries and low-angle lamellar grain boundaries [6]. Segregation is unlikely to occur at twin boundaries, which are low-energy grain boundaries, but the amount of segregation generally increases with increasing grain-boundary energy [19]. Thus, the amount of element segregation at grain boundaries is greatly affected by the grain boundary character, so twin boundaries were the subject of analyses in the present study. It was revealed that even at twin boundaries, where segregation is unlikely to occur, GBS of Si already existed before annealing and it became more pronounced by annealing (Fig. 8).

To confirm the effect of GBS on strength, the multi-scale simulation was performed using the crystallographical orientation, geometrical information of twin plane against the tensile axis of the caliber-rolled rods, and various material constants of Cu [20]. About the details of the analysis of the effect of GBS on mechanical properties using multi-scale simulation, it is reported elsewhere [6, 21, 22]. The area of the 3D multi-phase FEM analysis model including twin boundary was set to be 25 nm × 25 nm × 50 nm, and the twin boundary was located at the center of the height direction. Based on the measurement results in Figs. 5 and 8, the twin-boundary spacing and the GBS width in the model were conveniently set to be 25 nm and 5 nm respectively. To give the analysis model periodic boundary conditions in all directions, FEM analysis models were created for two types of twin orientations, where the twin boundary plane is parallel to {111} and the RD is parallel either to ⟨112⟩ or ⟨110⟩. In addition, the orientations of the upper and lower crystal grains are in a twin relationship (rotated 180° relative to the RD), but were randomly rotated by a maximum of 3° to reduce the symmetry of the crystal orientation relative to the loading axis. A tensile deformation of 5% was applied parallel to RD under periodic boundary conditions in all directions. The main material parameters were the initial dislocation density of 1 × 10¹¹ m⁻², the annihilation rate of dislocations of 1 × 10¹⁴ m², and the minimum length of immobilized dislocation of 30 nm.

Figure 9 shows the calculated true stress-true strain curves when tensile deformation is performed along RD. The solid lines are the calculation results when the effect of GBS is considered, and the dashed lines not considered. It is evident from Fig. 9 that both the yield stress and the work hardening rate increase when GBS is considered. The maximum tensile stress of RD//⟨112⟩ is larger than that of RD//⟨110⟩. It can be seen that the work hardening rate of RD//⟨112⟩ changes significantly at around the true strain of 0.01. The change in the work hardening rate would be because of the activation of secondary slip systems. As is clear from the results of simulation in Fig. 9, the GBS significantly raises yield strength and work hardening rate, and the deformation stress at the strain of 5% was increased by about 33%. It has been theoretically explained that the yielding of UFGed materials is controlled by the emission of dislocations from grain boundaries [23]. On the other hand, molecular dynamics calculations of high-entropy alloys have revealed that GBS makes it difficult for dislocations to escape from the grain boundaries, resulting in an increase in yield strength [24]. These results attained by theoretical research strongly suggest that the strengthening due to GBS in HN-structures would be due to the suppression of dislocation emission from the grain boundaries.

Fig. 9

Result of multi-scale simulation of tensile properties along ⟨112⟩ and ⟨110⟩ of heterogeneous nano-structured Cu-20Zn-1.0Si modified alloys with and without grain-boundary segregation of Si, which were described by GBS and non-GBS. (online color)

The distribution of geometrically necessary dislocations (GN dislocations) at the strain of 2% is displayed in Fig. 10, showing a cross section at the center of the analysis region. Without GBS, dislocations accumulate only at twin boundaries. However, with GBS, it can be understood that the segregation promotes non-uniform deformation around the twin boundaries and it propagates into the grain interior. This non-uniform deformation is assumed to be due to the strain field generated by GBS. The increase in GN dislocation density near the twin boundaries impedes the dislocation motion for slip deformation, resulting in an increase in the work hardening rate and deformation stress. In the HN-structured Cu-Zn-Si alloy, GBS of Si has already occurred during caliber rolling. And it has become clear that GBS as well as the other strength mechanisms such as grain refinement, work hardening, and solid solution strengthening possesses extremely strong effect on the tensile strength. It is reported that the segregation of solute elements to grain boundaries in UFGed structures during severe deformation is due to the drag of dislocations of solute atoms [25]. Low-temperature annealing increased the amount of segregation to cause further increase in strength.

Fig. 10

Distribution (m⁻²) of geometrically necessary dislocations at a strain of 2% at around twin boundaries with and without Si segregation calculated by means of multiscale simulation, where tensile directions are (a) ⟨112⟩ and (b) ⟨110⟩. Twin boundaries, shown by thin gray lines, were aligned at the centers in the above analytical models. (online color)

Nevertheless, early recrystallization and softening occurred during annealing due to the low thermal stability of the UFGed structure [6, 10, 26], and sufficient time was not provided for the diffusion of Si to the grain boundaries. However, the addition of very small amounts of Ni and P resulted in a slight improvement in the heat resistance and, therefore, the highest tensile strength was achieved.

On the other hand, it is easy to estimate that sufficient segregation of Si to the high densely developed grain boundaries in the HN-structure becomes difficult when the added amount of Si is low. If we hypothetically consider a simplified model with average twin-boundary spacing of 40 nm and GBS width of 5 nm based on experimental data, for example, even if the entire amount of Si added at 0.5 mass% segregates to the grain boundaries, the GBS will only be 4 mass% (8.3 atm%). In the HN-structure, there are sub-boundaries, medium- and high-angle grain boundaries, and finer secondary twins at higher density, which allow larger amounts of segregation. Segregation of the entire amount of Si to the grain boundaries is, however, thermodynamically impossible. In other words, the actual amount of segregation to the twin boundaries must be significantly lower than 4 mass% (8.3 atm%). Therefore, it is easy to infer that a deficiency of Si occurs in the alloys with 0.5 mass% Si addition due to the segregation of Si to various grain boundaries, and the amount of segregation to the twin boundaries is considerably limited. And it can be understood that this deficiency of the added Si is the reason why the strength of the alloys with 0.5 mass% Si added is the lowest. In short, a larger addition of Si is required to increase the strength of HN-structured Cu-Zn-Si alloys by utilizing of the mechanism of GBS. Excepting alloys with 0.5 mass% Si addition, there is a relatively good correlation between tensile strength and SFE value, and there is a clear tendency for strength to increase as the SFE value decreases. The above results can serve as important design guidelines for improving the thermal stability of HN-structured copper alloys and increasing their strength through GBS of solute elements.

This research was financially supported by the NEDO “Strategic Energy Saving Technology Innovation Program”. The alloy rods were kindly provided by San-Etsu Metals Co., Ltd. We would like to express our gratitude to them.

• 1)  Z. Horita (ed.): Nanomaterials by Severe Plastic Deformation, (Trans Tech Publications, Switzerland, 2005).
• 2)   N.  Tsuji,  Y.  Ito,  Y.  Saito and  Y.  Minamino: Strength and ductility of ultrafine grained aluminum and iron produced by ARB and annealing, Scr. Mater. 47 (2002) 893–899. doi:10.1016/S1359-6462(02)00282-8
• 3)   H.  Miura and  W.  Nakamura: Microstructure and mechanical properties of Mg–8Al alloy fabricated by room-temperature multi-directional forging, Philos. Mag. Lett. 93 (2013) 601–607. doi:10.1080/09500839.2013.827800
• 4)   H.  Ueno,  K.  Kakihata,  Y.  Kaneko,  S.  Hashimoto and  A.  Vinogradov: Enhanced fatigue properties of nanostructured austenitic SUS 316L stainless steel, Acta Mater. 59 (2011) 7060–7069. doi:10.1016/j.actamat.2011.07.061
• 5)   Y.  Nakao and  H.  Miura: Nano-grain evolution in austenitic stainless steel during multi-directional forging, Mater. Sci. Eng. A 528 (2011) 1310–1317. doi:10.1016/j.msea.2010.10.018
• 6)   H.  Miura,  C.  Watanabe,  Y.  Aoyagi,  Y.  Oba,  M.  Kobayashi and  N.  Yoshinaga: Age-hardening mechanisms of heterogeneous-nanostructured SUS316LN stainless steel fabricated by heavy cold rolling, Mater. Sci. Eng. A 833 (2022) 142531. doi:10.1016/j.msea.2021.142531
• 7)   H.  Miura,  Y.  Iwama and  M.  Kobayashi: Comparisons of Microstructures and Mechanical Properties of Heterogeneous Nano-Structure Induced by Heavy Cold Rolling and Ultrafine-Grained Structure by Multi-Directional Forging of Cu–Al Alloy, Mater. Trans. 60 (2019) 1111–1115. doi:10.2320/matertrans.MF201904
• 8)   H.  Miura,  Y.  Chiba,  M.  Kobayashi,  C.  Watanabe and  N.  Muramatsu: Effects of Be Content and Heat-Treatment Conditions on the Heterogeneous Nanostructure and Mechanical Properties of Cu–Be Alloys, Mater. Trans. 63 (2022) 1431–1436. doi:10.2320/matertrans.MT-D2022001
• 9)  N. Narita and J. Takamura: Dislocations in Solids, Vol. 9, ed. by F.R.N. Nabaro, (Elsevier, 1992) pp. 141–169.
• 10)   H.  Miura,  K.  Arai,  M.  Kobayashi,  Y.  Sumino and  C.  Watanabe: A Basic Research for Conquest Drawbacks and Further Improvement of Properties of Heterogeneous-Nano Structured Copper Alloys, Mater. Trans. 64 (2023) 967–973. doi:10.2320/matertrans.MT-D2022011
• 11)   M.F.  Denanot and  J.P.  Villain: The stacking fault energy in Cu-Al-Zn alloys, Phys. Status Solidi A 8 (1971) K125–K127. doi:10.1002/pssa.2210080248
• 12)   C.  Watanabe,  S.  Kobayashi,  Y.  Aoyagi,  Y.  Todaka,  M.  Kobayashi,  N.  Sugiura,  N.  Yoshinaga and  H.  Miura: Heterogeneous Nano-structure and its Evolution in Heavily Cold-rolled SUS316LN Stainless Steels, ISIJ Int. 60 (2020) 582–589. doi:10.2355/isijinternational.ISIJINT-2019-445
• 13)   H.  Zhou,  G.  Chen,  Y.  Li,  N.  Koga,  C.  Watanabe and  H.  Miura: Microstructure and Mechanical Properties of Cu–Zn–Si Alloy Bars Produced by Groove Rolling, J. Japan Inst. Copper 63 (2024) 14–19. doi:10.34562/jic.63.1_14
• 14)  L.E. Murr: Interfacial Phenomena in Metals and Alloys, (Addison-Wesley, Reading, 1975), pp. 131.
• 15)   R.  Armstrong,  I.  Codd,  R.M.  Douthwaite and  N.J.  Petch: The plastic deformation of polycrystalline aggregates, Philos. Mag. 7 (1962) 45–58. doi:10.1080/14786436208201857
• 16)   W.  Charnock: The influence of grain size on the nature of portevin-lechatelier yielding, Philos. Mag. 18 (1968) 89–99. doi:10.1080/14786436808227311
• 17)   T.  Morikawa and  K.  Higashida: The Influence of Deformation Twinning on Yield Strength in Cold-rolled FCC Ferrous Alloy, Tetsu-to-Hagané 91 (2005) 834–838. doi:10.2355/tetsutohagane1955.91.11_834
• 18)   J.E.  Flinn,  D.P.  Field,  G.E.  Korth,  T.M.  Lillo and  J.  Macheret: The flow stress behavior of OFHC polycrystalline copper, Acta Mater. 49 (2001) 2065–2074. doi:10.1016/S1359-6454(01)00102-1
• 19)   R.  Monzen,  T.  Kuze,  O.  Matsuda and  H.  Miura: Antimony-induced embrittlement of [011] symmetric tilt boundaries in copper bicrystals, Metall. Mater. Trans. A 34 (2003) 773–779. doi:10.1007/s11661-003-1005-0
• 20)   J.  Oda and  H.  Okada: Design Method of Materials Composed of Some Ingredients by Using Neural Network, Materials System 14 (1995) 51–56. doi:10.34401/zairyosystem.14.0_51
• 21)   Y.  Aoyagi,  T.  Tsuru and  T.  Shimokawa: Crystal plasticity modeling and simulation considering the behavior of the dislocation source of ultrafine-grained metal, Int. J. Plast. 55 (2014) 43–57. doi:10.1016/j.ijplas.2013.09.009
• 22)   Y.  Aoyagi,  C.  Watanabe,  M.  Kobayashi,  Y.  Todaka and  H.  Miura: Crystal Plasticity Simulation on Effect of Heterogeneous-nanostructure Induced by Severe Cold-rolling on Mechanical Properties of Austenitic Stainless Steel, Tetsu-to-Hagané 105 (2019) 262–271. doi:10.2355/tetsutohagane.TETSU-2018-078
• 23)   M.  Kato,  T.  Fujii and  S.  Onaka: Dislocation Bow-Out Model for Yield Stress of Ultra-Fine Grained Materials, Mater. Trans. 49 (2008) 1278–1283. doi:10.2320/matertrans.MRA2008012
• 24)   K.  Shiotani,  T.  Niiyama and  T.  Shimokawa: Atomistic study on dislocation emission from segregated grain boundaries in high-entropy alloys, Phys. Rev. Materials 7 (2023) 123606. doi:10.1103/PhysRevMaterials.7.123606
• 25)   X.  Sauvage,  A.  Ganeev,  Y.  Ivanisenko,  N.  Enikeev,  M.  Murashkin and  R.  Valiev: Grain Boundary Segregation in UFG Alloys Processed by Severe Plastic Deformation, Adv. Eng. Mater. 14 (2012) 968–974. doi:10.1002/adem.201200060
• 26)   M.M.  Abramova,  N.A.  Enikeev,  X.  Sauvage,  A.  Etienne,  B.  Radiguet,  E.  Ubyivovk and  R.Z.  Valiev: THERMAL STABILITY AND EXTRA-STRENGTH OF AN ULTRAFINE GRAINED STAINLESS STEEL PRODUCED BY HIGH PRESSURE TORSION, Rev. Adv. Mater. Sci. 43 (2015) 83–88.