2025 年 66 巻 2 号 p. 160-164
The hot-extruded bars of a Cu-Zn-Si alloy were subjected to groove rolling. Groove rolling was performed using rolls with grooves of different sizes so that the cross-sectional area of the bars after rolling decreased. The specimen bars were rotated by 90 degrees along the rolling direction after each rolling pass, and rolling was continued up to 68% reduction in cross-sectional area. With increasing rolling reduction, the initial coarse-equiaxed grains macroscopically elongated parallel to the rolling direction. Within the grains, ultrafine mechanical twins were high-densely introduced, which promptly and drastically fragmented the initial coarse grains to develop heterogeneous nanostructure consisting of micro- and nano-meter-ordered mechanical twins. The tensile strength of the as-hot-extruded bar of 464 MPa increased to 704 MPa after 42% reduction, and further increased up to 940 MPa after 68% reduction. The remarkable increase in strength is attributed to the evolution of the heterogeneous nanostructure developed by grain subdivision owing to dense mechanical twinning. It can be concluded, therefore, that the strength of the Cu-Zn-Si alloy bars was significantly increased by the evolution of the heterogeneous nanostructure even by relatively small area reduction of 68%, i.e., equivalent strain of 1.13.
This Paper was Originally Published in Japanese in J. Japan Inst. Copper 63 (2024) 14–19. The caption of Fig. 4 was slightly modified.
Copper alloys, known for their high strength and electrical conductivity, are widely used in electrical and electronic components for various electronic devices and automobiles. In recent years, the rise of Internet of Things devices, along with the increasing automation and electrification of automobiles, has led to a growing demand for downsizing various electronic devices. Consequently, copper alloys used in such devices are required to possess higher strength. However, there is a trade-off between strength and electrical conductivity. As a result, numerous studies and development efforts have focused on enhancing both strength and conductivity simultaneously. Grain refinement is an effective method for improving the strength of copper alloys without significantly compromising their conductivity [1]. It has been reported that metallic materials with ultrafine grains of size ≤1 µm can achieve an optimal balance between strength and ductility [2]. The severe plastic deformation (SPD) method is a commonly employed technique to produce ultrafine-grained metallic materials by applying intense plastic strain. However, most SPD methods are batch processes based on shape-invariant deformation. Due to their process complexity, limitations on product size, and the need for specialized manufacturing equipment, SPD methods have seen limited application in industrial mass production.
Recent research by Miura et al. reported that exceptionally high strength and sufficient ductility can be achieved in face-centered cubic metals with very low stacking fault energy, such as Cu-Zn and Cu-Be alloys [3–5], by employing simple and heavy cold rolling. This improvement in mechanical properties is attributed to the development of a heterogeneous-nano (HN) structure during cold rolling. The HN structure is a complex arrangement comprising deformation-induced, nanometer-scale component microstructures, including conventional lamellae elongated along the rolling direction, deformation twin domains, and shear bands [6]. Most grain boundaries in the HN structure are low-angle or coherent Σ3 boundaries, which offer considerable thermal stability [7]. It has been demonstrated that age hardening can be effectively utilized by first introducing a HN structure into precipitation-hardenable Cu alloys through cold rolling after solution treatment, followed by subsequent aging heat treatment to superimpose the precipitation hardening effect [8–10]. Thus, while the HN structure can be developed through relatively simple cold rolling, it significantly enhances the mechanical properties of metallic materials. Moreover, since this method requires no substantial changes to existing manufacturing facilities, the HN structure is being recognized as a novel microstructural control method suitable for industrial mass production.
However, simple cold rolling is primarily suitable for producing sheet or plate materials. Therefore, recent efforts have focused on producing HN-structured bars and tubes [11]. Groove rolling is a promising technique for manufacturing HN-structured bars. As depicted in Fig. 1, groove rolling utilizes rolls with a pair of grooves to produce bars.
Schematic illustration showing groove rolling. (online color)
In this study, a Cu-Zn-Si alloy was used as a model material to explore the enhancement of its mechanical properties through groove rolling. In addition, the relationship between the microstructure and the mechanical properties of the groove-rolled specimens was investigated.
In this study, Cu-26Zn-2Si (mass%) hot-extruded bars with a diameter of 10 mm were used. Based on stacking fault energy (SFE) data for Cu-Zn and Cu-Si alloys [12], binary systems, the SFE value of the present alloy, calculated using the method of Denanot and Villain [13], is extremely low, approximately 7 mJ/m2, which is about one-third of the value for conventional 70/30 brass (20 mJ/m2) [14]. The extruded bars underwent groove rolling at room temperature until a maximum area reduction of 68% (an equivalent strain of 1.13) was achieved. As illustrated in Fig. 2, the specimens were rotated 90° along the rolling direction after each rolling pass. Table 1 shows the area reductions and corresponding equivalent strains of the rolled specimens. Hereafter, specimens will be referred to according to their area reduction (AR): AR-22%, AR-42%, AR-57%, and AR-68%. The rolling direction of the final pass is denoted as RD, the direction parallel to the rolls as TD, and the direction perpendicular to both as ND.
Pass schedule of groove rolling process.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/MT-D2024007tab01.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
The microstructures of the rolled specimens were observed using scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Observations were conducted along TD. Backscattered electron (BSE) images were captured using a retractable BSE detector at an acceleration voltage of 20 kV and a probe current of 14 mA. Crystallographical orientation data from the same area as the BSE images were analyzed using electron backscattered diffraction (EBSD), with an acceleration voltage of 20 kV and a beam step size of 50 nm. Specimens for SEM observation were mechanically polished using #2000 emery paper and then electro-polished at 10 V and 243 K for approximately 90 seconds, using a 30% nital solution (methanol:nitric acid = 7:3 by volume) and a SUS304 stainless steel cathode. TEM observations were performed at an accelerating voltage of 200 kV, and specimens for TEM observation were prepared using the ion-thinning method.
Tensile tests were conducted at room temperature at an initial strain rate $\dot{\varepsilon }$ of 10−3 s−1. Dogbone-shaped specimens, with gauge dimensions of 5 mm (length) × 2 mm (width) × 0.5 mm (thickness), were cut out using a wire electric discharge machine, aligning the tensile axis parallel to the RD of the rolled bars. The surfaces of these specimens were mechanically polished using #2000 emery paper.
Figure 3 shows the SEM-BSE images of the specimens before and after groove rolling, focusing on areas located at the center of the bar. The as-extruded specimen exhibited a coarse, approximately equiaxed grain structure with an average grain size of 55 ± 12 µm (Fig. 3(a)). Annealing twins were frequently observed within these grains. After groove rolling, as shown in Figs. 3(b) and 3(c), the grains were elongated along the RD. The aspect ratio of the grains increased from 1.3 in the AR-22% specimen to 3.7 in the AR-68% specimen, which represents the maximum area reduction. Additionally, sharp linear contrasts were observed within the grains, indicated by arrows in Fig. 3(c).
SEM-BSE images of (a) as-extruded, (b) AR-22% and (c) AR-42%.
Figure 4(a) presents a higher magnification SEM-BSE image of the AR-42% specimen. The coherent Σ3 boundaries, indicated by red lines, were superimposed on the band contrast map obtained from EBSD analysis (Fig. 4(b)). The twins observed in Fig. 4(b) were significantly finer than the annealing twins seen in the as-extruded specimen, indicating that these fine twins were deformation twins introduced during groove rolling. The twin boundary spacing decreased as the area reduction increased (Table 2), demonstrating that the initial coarse grains were instantly subdivided by the deformation twin bands, resulting in grain refinement. Moreover, as highlighted by the arrows in Fig. 4(a), densely packed linear contrasts were observed within some twin bands, suggesting further subdivision by these contrasts. To further investigate these linear contrasts with ultrafine spacing, TEM observations were conducted, as exemplified in Fig. 5. Similar to Fig. 4, the initial coarse grain was fragmented by twin bands. Additionally, it was confirmed that a different variant of twins formed, further fragmenting the twin bands. For clarity, the original twins are referred to as primary twins, and the subsequent variants as secondary twins. The ultrafine-grained structure developed during groove rolling, characterized by different variants of twins heterogeneously dividing the initial coarse grains, is defined as a “hierarchical HN structure”. This structure is distinct from the HN structure, which consists of lamellae, deformation twin domains, and shear bands typically formed by conventional heavy rolling.
(a) SEM-BSE image of AR-42% specimen. (b) Band contrast map taken at the same area in (a). The bold red lines in (b) indicate coherent Σ3 boundaries. (online color)
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/MT-D2024007tab02.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
(a) Bright-filed TEM image of AR-42% specimen. (b), (c) selected-area diffraction patterns taken from the areas indicated in the squares in (a), respectively. (online color)
Table 2 provides the average boundary spacing of primary (dp) and secondary (ds) twins. To explore the relationship between the formation of secondary twins and primary twin bands, the fraction (fs) of primary twin bands containing secondary twins per unit length was analyzed. The fs values are listed in Table 2. As the area reduction increased, both dp and ds decreased, while fs showed a slight increase. Furthermore, it was observed that the spacing of primary twins containing secondary twins (ws) was more than double that of primary twins without secondary twins (wno-s), indicating a tendency for secondary twins to form in wider primary twin bands.
3.2 Mechanical propertiesTensile tests were conducted along the RD on both the as-extruded specimen and the groove-rolled specimens with varying area reductions. Each test condition was repeated three times to ensure accuracy. The representative stress-strain curves are displayed in Fig. 6, and the average values of 0.2% yield stress (σ0.2), ultimate tensile stress (σUTS), and fracture elongation (εf) are summarized in Table 3. As the area reduction increased, both σ0.2 and σUTS showed a significant increase, while εf decreased. Notably, for the AR-68% specimen, which experienced the maximum area reduction in this study, the σUTS value reached 940 MPa, more than double that of the as-extruded specimen.
Nominal stress – nominal strain curves of specimens groove-rolled to various area reductions.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/MT-D2024007tab03.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
After groove rolling, the initial coarse grains in the specimens were subdivided by deformation twins, and these twin bands were further fragmented by higher-order twins, resulting in a “hierarchical HN structure”. The introduction of high-density multiple deformation twins can be attributed, in part, to the multidirectional nature of the groove rolling process. Cai et al. conducted compression tests on copper single crystals with various orientations and found that mechanical twinning is highly dependent on the orientation of the compression axis [15]. Li et al. studied the effect of rolling methods on the mechanical properties of HN-structured Cu-Zn alloys [16]. They demonstrated that two-directional rolling, which involves rotating the specimens by 90° along the RD during the early stages of cold rolling, achieved a better strength-ductility balance compared to conventional unidirectional rolling. This implies that altering the rolling direction early in the process can induce twins even in grains that are not typically prone to mechanical twinning, thereby increasing the volume fraction of deformation twin domains in the final HN structure. In groove rolling, the specimens were rotated by 90° after each rolling pass, which facilitated the formation of deformation twins and encouraged the development of multiple twins, similar to the findings of Li et al. Additionally, Narita and Takamura reported that the critical stress (τc) for twinning decreases as the SFE decreases [17]:
| \begin{equation} \tau_{\text{c}} = \gamma_{\text{SF}}/2b_{\text{T}} \end{equation} | (1) |
Here, γSF represents the SFE, and bT is the magnitude of the Burgers vector of partial dislocations that cause mechanical twinning. According to this equation, τc decreases with a reduction in SFE. In the present Cu–Zn–Si alloy, mechanical twinning can occur with only one-third of the applied stress required for twinning in 70/30 brass, or even within grains with lower Schmid factors. Moreover, in grains where mechanical twinning did not occur during the previous rolling pass, the Schmid factor changes significantly with a 90° rotation of the specimen. This facilitates the introduction of deformation twins in subsequent rolling paths. By subdividing the coarse initial grains through the formation of high-density deformation twins, an ultrafine HN structure with a boundary spacing of <100 nm (as shown in Table 2) was achieved, even with a relatively low amount of deformation (equivalent strain of 1.13, as indicated in Table 1), compared to what is typically required in the SPD method.
The strength of metallic materials generally increases as grain size decreases, as described by the Hall-Petch law [18–20]. Twin boundaries also serve as obstacles to dislocation movement, and it has been demonstrated that strength increases as twin boundary spacing decreases [21–23]. Therefore, if the effective grain size of the specimens after groove rolling is considered equivalent to the secondary twin boundary spacing ds, it can be inferred that strength increases as ds decreases with increasing area reduction (Table 2). Conversely, the fracture elongation εf decreased with increasing area reduction (Table 3). This reduction in ductility can be attributed to grain refinement, caused by the fragmentation of initial grains by deformation twins, and an increase in dislocation density during groove rolling. Zhang et al. performed simple rolling on a Cu–20%Zn-1.2%Si alloy until a 90% thickness reduction (equivalent strain of 2.3) at 77 K, resulting in an ultrafine-grained specimen with an average grain size of 440 nm [24]. σUTS and εf of the ultrafine-grained specimen were 940 MPa and ∼9%, respectively. The AR-68% specimen from this study demonstrated a superior strength-ductility balance compared to the ultrafine-grained specimen reported by Zhang et al. (Table 3). Notably, these superior mechanical properties were achieved through groove rolling with a relatively low area reduction of 68% (equivalent strain of 1.13) at room temperature, as opposed to the high reductions achieved through cryogenic rolling.
Flinn et al. reported that the flow stress of materials with high-density twins is proportional to the number density of twin boundaries [25]. In the present study, the number density of twin boundaries increased with increasing area reduction during groove rolling (Table 2), which coincided with increased strength. This observation can be qualitatively explained by the findings of Flinn et al. However, due to the “inhomogeneity” of the hierarchical HN structure, the twin boundary spacing and density have not yet been quantitatively evaluated. Therefore, it is necessary to identify and quantify the specific microstructural factors within the hierarchical HN structure that contribute to the observed strength enhancement. Further strengthening could potentially be achieved by increasing the volume fraction of higher-order twins and decreasing the boundary spacing, both of which are related to the effective grain size. Additionally, optimizing the rolling process to enhance these microstructural characteristics will be a focus of future work.
A Cu-Zn-Si alloy was subjected to groove rolling to examine its microstructure and mechanical properties. The key findings are summarized as follows:
This research was supported by a research grant from the Copper Society of Japan for 2022. The authors also acknowledge the Grant-in-Aid for Scientific Research (Grant # 20H02461). Alloy specimens were provided by Kitz Metal Works.
