Tensile Deformation Behavior of High-Strength Nanostructured Cu–Si Solid-Solution Alloys Processed by Severe Plastic Deformation

Abstract

Tensile deformation behavior of high-strength nanostructured Cu–Si solid-solution alloys processed by high-pressure torsion (HPT) with 5 rotations was investigated at room and low temperatures. With increasing Si concentration, tensile strength of the nanostructured Cu–Si solid-solution alloys was significantly increased. The maximal tensile strengths were 980 MPa at room temperature, and 1350 MPa at 77 K in a Cu–2.04 wt.%Si alloy. This significant strengthening was achieved by grain refinement and increased dislocation density through severe plastic deformation (SPD) with the effect of Si addition on the decreasing stacking fault energy of the Cu–Si alloy. With increasing Si concentration, strain-rate sensitivity m of the nanostructured Cu–Si solid-solution alloys was decreased due to the increased dislocation density, resulting in accelerating plastic instability of tensile specimens, caused by the diminishing strain-rate hardening capacity after necking.

 

This Paper was Originally Published in Japanese in J. Japan Inst. Copper 59 (2020) 299–303. Some sentences were added in the main text of this paper. References of 2) and 5–11) were newly added.

Fig. 6 Strain-rate sensitivity vs. grain size in pure Cu and the Cu–Si solid-solution alloys processed by the HPT with 5 rotations. The open squares are the data obtained from the literature by Wei et al. for pure Cu.¹⁴⁾

1. Introduction

Stacking fault energy (SFE) of Cu solid-solution alloys monotonically decreases with increasing the concentration of solid-solution atoms of typical elements such as Al, Si, and Zn.^1,2) The SFE of Cu solid-solution alloys eventually approaches minimal values near the solid-solubility limits.¹⁾ In our previous study, these Cu solid-solution alloys were subjected to severe plastic deformation (SPD).³⁾ Marked increase in hardness after the SPD processing was observed in any of the above mentioned Cu solid-solution alloys with increasing the concentration of solid-solution atoms up to the solid-solubility limits.³⁾ Kunimine et al. conducted accumulative roll-bonding (ARB), which is a kind of the SPD processing, on Cu–Si solid-solution alloys with various Si concentration for grain refinement.⁴⁾ However, in the case of the Cu–Si solid-solution alloys with relatively high Si concentration or high equivalent strain after the ARB processing, it was difficult to obtain high-strength materials with tensile strength of over 600 MPa at room temperature (RT) due to the crack initiation and propagation in the materials during the ARB processing.⁴⁾

In the present study, Cu–Si solid-solution alloys with various Si concentrations up to near the solid-solubility limit were subjected to high-pressure torsion (HPT)^5,6) processing until the hardness of the alloys was saturated. The HPT processing is a kind of the SPD processing that allows us to apply giant strain to materials without any significant change in dimensions of workpiece materials. The relationships between microstructures and mechanical properties of the Cu–Si solid-solution alloys having various Si concentrations with saturated hardness achieved by the HPT processing were investigated through microstructural observation and tensile tests at RT and a low temperature in liquid nitrogen (77 K). Effects of Si concentration on the strain-rate sensitivity m of the SPD processed Cu–Si solid-solution alloys were also investigated. Nowadays, since brasses containing Bi and Si have been developed and used as substitute materials for free-cutting brasses containing Pb and Cd due to environmental issues and the RoHS directive, it would be useful to understand the relationships between microstructures and mechanical properties of the SPD processed binary Cu–Si solid-solution alloys as fundamental knowledge.

2. Experimental Procedure

Annealed pure Cu (4NCu) sheets with purity of 99.99 wt.% were used as a material. Ingots of a Cu–0.18 wt.%Si (Cu0.2Si), a Cu–0.73 wt.%Si (Cu0.7Si), and a Cu–2.04 wt.%Si (Cu2.0Si) alloy were made by plasma arc melting. The Si concentrations in these Cu–Si alloys were measured by inductively coupled plasma atomic emission spectroscopy (ICP-AES) after solution treatment. According to Gallagher’s report, the SFE value of pure Cu is 50∼60 mJ/m², and monotonically decreases to 15∼20 mJ/m² with increasing Si concentration up to about 2 wt.%.¹⁾ These ingots of the Cu–Si alloys were heat treated for homogenization, and then rolled into sheets. The sheets were sufficiently annealed and confirmed to be Cu–Si solid-solution alloys by X-ray diffraction (XRD) with an X-ray diffractometer (RINT-2500, Rigaku Corporation, Tokyo, Japan) using Cu-Kα₁ radiation (the wavelength λ = 0.15406 nm) under the operation condition for an accelerating voltage of 40 kV and a current of 200 mA. Disk shaped specimens with a diameter of 10 mm and a thickness of 0.8 mm were cut from the sheets of pure Cu and Cu–Si solid-solution alloys by a wire electrical discharge machine (EDM), followed by the HPT processing with 5 rotations (5R) under the condition of 6 GPa, 0.2 rpm and RT. In our previous study, it was found that the hardness of these pure Cu and Cu–Si solid-solution alloys was almost saturated and uniform at any position in the specimens processed by the HPT with 5R.³⁾

The dislocation density of pure Cu and the Cu–Si alloy specimens processed by the HPT with 5R was evaluated by means of the Williamson-Hall method with the use of XRD.⁷⁾ First, peak diffraction angles θ and full widths at half maximum (FWHM) β of the X-ray diffraction peaks from the (111), (200), (220), (311), and (222) planes were measured. Annealed pure Cu and Cu–Si alloys were used as XRD peak-broadening references. Based on the following eq. (1):   

\begin{equation} \frac{\beta \cos \theta}{\lambda} = \frac{0.9}{D} + 2\varepsilon_{\text{m}} \frac{\sin \theta}{\lambda}, \end{equation} (1)

data points were plotted as the horizontal axis of 2 sin θ/λ and the vertical axis of β cos θ/λ. Here ε_m is the microstrain and D the crystallite size.⁷⁾ The ε_m and D were determined from the values of the slope and intercept of the straight line for the plotted data points, respectively. The dislocation density ρ was obtained by the following eq. (2):   

\begin{equation} \rho = \frac{2\sqrt{3} \varepsilon_{\text{m}}}{Db}, \end{equation} (2)

where b is the magnitude of the Burgers vector.⁷⁾

For microstructural observations, the HPT processed specimens were cut at a position of 2.8 mm from the center of the disk-shaped samples. The cross-sectional observations were performed by a field emission-scanning electron microscope (FE-SEM: JSM-7100F, JEOL, Tokyo, Japan). The average grain sizes of the specimens processed by the HPT with 5R for pure Cu and the Cu–Si solid-solution alloys were measured from backscattered electron (BSE) images by the linear intercept method.

Tensile tests were carried out at RT and 77 K in liquid nitrogen with a universal testing machine (AUTOGRAPH AG-X 10kN, Shimadzu, Kyoto, Japan) at a strain rate $\dot{\varepsilon }$ of 8.3 × 10⁻⁴ s⁻¹. The strain-rate dependence of the flow stress of the HPT processed pure Cu and the Cu–Si solid-solution alloys was evaluated by measuring the strain-rate sensitivity m through strain-rate jump tests during the tensile tests at RT. Here, the strain-rate sensitivity m is defined by the following eq. (3).   

\begin{equation} m = \frac{\partial \ln \sigma}{\partial \ln \dot{\varepsilon}} = \frac{1}{\sigma}\frac{\partial \sigma}{\partial \ln \dot{\varepsilon}} \end{equation} (3)

The strain-rate sensitivity m was experimentally determined from the following eq. (4) by using the amount of stress change Δσ = σ₁ − σ₂ of the flow stress caused by the suddenly changed strain rates from $\dot{\varepsilon }_{1}$ of 8.3 × 10⁻⁴ s⁻¹ to $\dot{\varepsilon }_{2}$ of 8.3 × 10⁻⁵ s⁻¹ immediately after the yielding during tensile deformation.   

\begin{equation} m \approx \frac{\sigma_{1} - \sigma_{2}}{[(\sigma_{1} + \sigma_{2})/2]\ln (\dot{\varepsilon}_{1}/\dot{\varepsilon}_{2})} \end{equation} (4)

The tensile tests were performed using miniature tensile specimens cut from the HPT processed disc-shaped samples with the wire EDM followed by polishing to a thickness of 0.5 mm as shown in Fig. 1.

Fig. 1

Drawing of a miniature tensile specimen taken from an HPT-processed disc.

3. Results

3.1 Effects of Si concentration on microstructure and dislocation density of the HPT processed Cu–Si solid-solution alloys

Figures 2(a)–(d) show SEM-BSE images of the microstructures obtained by observing the cross-sections of the HPT processed specimens with 5R for (a) 4NCu, (b) Cu0.2Si, (c) Cu0.7Si, and (d) Cu2.0Si. The average grain sizes were 175 nm for 4NCu, 140 nm for Cu0.2Si, 110 nm for Cu0.7Si, and 60 nm for Cu2.0Si, respectively. With increasing Si concentration, the average grain size was monotonically decreased as shown in Fig. 3. In addition, the relationships between the dislocation density ρ of the HPT processed Cu–Si solid-solution alloys with 5R and Si concentration are shown in Fig. 3. The dislocation density ρ increased with increasing Si concentration, and was saturated at approximately 5.0 × 10¹⁴ m⁻². These results can be attributed to the decrease of the SFE from 50∼60 mJ/m² for pure Cu to 15∼20 mJ/m² for Cu2.0Si.^1,2)

Fig. 2

SEM-BSE images of cross-section in the HPT processed discs with 5 rotations for (a) 4NCu, (b) Cu0.2Si, (c) Cu0.7Si, and (d) Cu2.0Si.

Fig. 3

Grain size and dislocation density vs. Si concentration in pure Cu and the Cu–Si solid-solution alloys processed by the HPT with 5 rotations.

3.2 Effects of Si concentration on tensile deformation behavior of the HPT processed Cu–Si solid-solution alloys at room and low temperatures

Figures 4(a) and (b) show nominal stress-nominal plastic strain curves of pure Cu and the Cu–Si solid-solution alloys processed by the HPT with 5R, tested at (a) RT and (b) 77 K. Tensile strengths at RT were 470 MPa for 4NCu, 670 MPa for Cu0.2Si, 890 MPa for Cu0.7Si, and 980 MPa for Cu2.0Si. Tensile strengths were significantly increased with increasing Si concentration. It could be said that the strengthening was achieved by the significant grain refinement and increased dislocation density after the HPT processing due to the decrease of the SFE with increasing Si concentration in the Cu–Si alloys, as can be seen from the results in Fig. 3. Significant increase in strength after the SPD processing by controlling the SFE can be also seen in other Cu solid-solution alloys.^8–11) Regarding the total elongation, it was about 35% for 4NCu. However, the total elongation decreased with increasing Si concentration, and reached about 18% for Cu2.0Si. It should be noted that the elongation obtained in the tensile tests greatly varies depending on the dimensions of tensile specimens.¹²⁾ In particular, the miniature tensile specimens used in this study tend to show a relatively large elongation.¹²⁾ The tensile strengths at 77 K were 670 MPa for 4NCu, 920 MPa for Cu0.2Si, 1200 MPa for Cu0.7Si, and 1350 MPa for Cu2.0Si. The tensile strengths at 77 K became 1.35 to 1.43 times higher than that at RT due to the increased effective stress caused by the decreased activation energy for plastic deformation at the low temperature. Furthermore, it was confirmed that the sharp drop of the flow stress in the nominal stress-nominal plastic strain curves at 77 K became more significant with increasing Si concentration since the Si solute atoms might act as strong obstacles to mobile dislocations.

Fig. 4

Nominal stress–nominal plastic strain curves of the HPT processed specimens with 5 rotations for 4NCu, Cu0.2Si, Cu0.7Si, and Cu2.0Si at (a) RT, and (b) 77 K.

3.3 Effects of Si concentration on strain-rate dependence of flow stress of the HPT processed Cu–Si solid-solution alloys

Figure 5 shows the effects of Si concentration on the amount of flow-stress change Δσ obtained by strain-rate jump tests using the HPT processed specimens with 5R. The amount of flow-stress change Δσ was 27 MPa for 4NCu, and gradually increased with increasing Si concentration and reached 41 MPa for Cu2.0Si. Figure 5 also shows the strain-rate sensitivity m derived from the amounts of flow-stress change Δσ. As Si concentration increased, the value of m gradually decreased from 0.025 for 4NCu to 0.016 for Cu2.0Si. Thus, the addition of solute Si leads to reducing the m value of the SPD processed Cu. Lu et al. have been reported that the m value was about 0.06 for pure Cu having a grain size of about 10 nm and a strength of 1000 MPa produced by electrodeposition method.¹³⁾ Using the m value of 0.06, the value of Δσ can be estimated as about 150 MPa for the electrodeposited Cu. On the other hand, although the HPT processed Cu2.0Si specimen with 5R also exhibited almost the same strength of about 1000 MPa in the present study, the value of Δσ was 41 MPa with the m value of 0.016. Therefore, it can be understood that the value of Δσ of the HPT processed Cu2.0Si specimen with 5R is about one-fourth or one-third times of the value of Δσ for the electrodeposited Cu with the same strength of 1000 MPa.

Fig. 5

Amount of stress change and strain-rate sensitivity vs. Si concentration in pure Cu and the Cu–Si solid-solution alloys processed by the HPT with 5 rotations.

4. Discussion

4.1 Strengthening mechanism of the SPD processed Cu–Si solid-solution alloys

Summarizing the experimental results described in the section 3, the decrease of the SFE of the Cu–Si solid-solution alloys with increasing Si concentration^1,2) results in suppression of the dynamic recovery during the SPD processing. This makes it possible to apply a larger strain to the materials by the SPD processing. As a result, the grain refinement of the Cu–Si alloys was accelerated, and furthermore, the dislocation density was also increased. Therefore, it can be said that the two strengthening mechanisms mainly caused the remarkable increase in strength. On the other hand, the strain-rate sensitivity m significantly decreased with increasing Si concentration. The reason for this will be considered below. Figure 6 presents the relationship between the strain-rate sensitivity m and grain size. It is well known that the m value remarkably increases with the increase of strength as the grain size becomes finer in the case of face-centered cubic (fcc) metals such as Cu and Ni.^13–15) The reported data for pure Cu¹⁴⁾ are plotted in Fig. 6. The dominant plastic deformation mechanism in coarse-grained pure Cu has been considered as the cutting of forest dislocations by mobile dislocations. The m value for this deformation mechanism has been reported as at most about 0.006.¹⁵⁾ On the other hand, in the ultrafine-grained materials with the grain size of less than 1 µm and nanocrystalline materials with the grain size of less than 100 nm, Kato et al.¹⁶⁾ reported that the significant increase in the m value with decreasing the grain size was attributed to the change in the dominant plastic deformation mechanism to the interaction between dislocations and grain boundaries (dislocation bowing-out and its depinning at grain boundaries).^17,18) As described in section 3.3, Lu et al. have reported that the electrodeposited nanocrystalline Cu with the grain size of about 10 nm exhibited the tensile strength of 1000 MPa and the m value of about 0.06.¹³⁾ The important point here is that the high strength of the electrodeposited nanocrystalline Cu was achieved only by the grain refinement strengthening. In contrast, the high strength of about 1000 MPa of the HPT processed Cu2.0Si with 5R was achieved by not only grain refinement strengthening with the average grain size of about 60 nm but also the dislocation strengthening caused by the high dislocation density. Therefore, it can be understood that the cutting of forest dislocations inside grains by the bowing-out dislocations from grain boundaries also contributed to the dominant plastic deformation mechanisms in addition to the interaction between the bowing-out dislocations and grain boundaries as mentioned above. That is the reason why the m value of the Cu–Si alloys decreased with increasing Si concentration and reached 0.016 for Cu0.2Si although the grain size became finer as shown in Fig. 6. It can be said that the strengthening mechanism of the high-strength Cu–Si solid-solution alloys processed by the SPD is different from that of the electrodeposited nanocrystalline Cu.

Fig. 6

Strain-rate sensitivity vs. grain size in pure Cu and the Cu–Si solid-solution alloys processed by the HPT with 5 rotations. The open squares are the data obtained from the literature by Wei et al. for pure Cu.¹⁴⁾

4.2 Reason for decrease in total elongation of the SPD processed Cu–Si solid-solution alloys

The total elongation of the high-strength Cu–Si solid-solution alloys processed by the SPD decreased with increasing the Si concentration at both RT and 77 K, as shown in Fig. 4. The reason for this will be described below. Equation (5) is well known as a condition for the onset of necking in tensile tests.^19,20)   

\begin{equation} \frac{1}{\sigma}\left(\frac{\partial \sigma}{\partial \varepsilon} \right) + m \leq 1 \end{equation} (5)

Here, the first and second terms describe the contribution to the ductility by the strain hardening and the strain-rate hardening, respectively. It means that the necking of the tensile specimen starts when the sum of the two terms becomes smaller than the right-hand side. The strain-rate sensitivity m as the second term in eq. (5) also contributes to the improvement of the ductility due to the retardation of fracture caused by the strain-rate hardening in the local necking region after the onset of necking in the tensile specimen.

In the case of the high-strength Cu–Si solid-solution alloys processed by the SPD, almost no effect of the Si addition on the strain hardening rate during uniform tensile deformation was observed since the HPT processing was carried out until the strength was saturated. Therefore, there is little difference in the uniform elongation of the SPD processed pure Cu and the Cu–Si alloys. On the contrary, it should be noted that the post-uniform elongation decreased with increasing Si concentration. This result should be caused by the decrease in the strain-rate sensitivity m with the increase of Si concentration, that is, the diminishing strain-rate hardening capacity in the local necking region during the post-uniform deformation. As a consequence, the plastic instability of the tensile specimens with higher Si concentration was further promoted, resulting in accelerating the fracture.

5. Conclusions

In the present study, the effects of Si concentration on the tensile deformation behavior of the high-strength nanostructured Cu–Si solid-solution alloys processed by the HPT processing was investigated at RT and 77 K. The important findings are as follows.

1. (1)    With increasing Si concentration in the HPT processed Cu–Si solid-solution alloys, the grain size became significantly finer and the dislocation density increased.
2. (2)    With the increase of Si concentration in the HPT processed Cu–Si solid-solution alloys, the tensile strength remarkably increased and reached the maximal values of 980 MPa at RT and 1350 MPa at 77 K. This strengthening was mainly caused by the grain refinement and increased dislocation density through the SPD.
3. (3)    The increase of Si concentration in the SPD processed Cu–Si solid-solution alloys led to decreasing the strain-rate sensitivity m. This promoted the plastic instability due to the diminishing strain-rate hardening ability in the local necking region after the initiation of necking during tensile deformation, resulting in the decrease of the total elongation of the SPD processed Cu–Si solid-solution alloys.

Acknowledgments

This research was supported by a Grant-in-Aid for the 2016–2017 academic year from Japan Institute of Copper. This work was also supported by JSPS KAKENHI Grant Number JP16K18259. The authors are grateful to Prof. Nobuhiro Tsuji, Kyoto University, for use of the HPT processing machine.

REFERENCES

• 1)   P.C.J.  Gallagher: Metall. Trans. 1 (1970) 2429–2461.
• 2)   C.B.  Carter and  I.L.F.  Ray: Philos. Mag. 35 (1977) 189–200.
• 3)   T.  Kunimine and  M.  Watanabe: Mater. Trans. 60 (2019) 1484–1488.
• 4)   T.  Kunimine,  T.  Fujii,  S.  Onaka,  N.  Tsuji and  M.  Kato: J. Mater. Sci. 46 (2011) 4290–4295.
• 5)   K.  Edalati and  Z.  Horita: Mater. Sci. Eng. A 652 (2016) 325–352.
• 6)   Z.  Horita,  Y.  Tang,  T.  Masuda and  Y.  Takizawa: Mater. Trans. 61 (2020) 1177–1190.
• 7)   H.  Adachi,  Y.  Miyajima,  M.  Sato and  N.  Tsuji: Mater. Trans. 56 (2015) 671–678.
• 8)   X.X.  Wu,  X.Y.  San,  Y.L.  Gong,  L.P.  Chen,  C.J.  Li and  X.K.  Zhu: Mater. Des. 47 (2013) 295–299.
• 9)   X.X.  Wu,  X.Y.  San,  X.G.  Liang,  Y.L.  Gong and  X.K.  Zhu: Mater. Des. 47 (2013) 372–376.
• 10)   Y.L.  Gong,  C.E.  Wen,  Y.C.  Li,  X.X.  Wu,  L.P.  Cheng,  X.C.  Han and  X.K.  Zhu: Mater. Sci. Eng. A 569 (2013) 144–149.
• 11)   T.  Kunimine: Mater. Sci. Forum 1016 (2021) 780–785.
• 12)   Y.H.  Zhao,  Y.Z.  Guo,  Q.  Wei,  T.D.  Topping,  A.M.  Dangelewicz,  Y.T.  Zhu,  T.G.  Langdon and  E.J.  Lavernia: Mater. Sci. Eng. A 525 (2009) 68–77.
• 13)   K.  Lu,  L.  Lu and  S.  Suresh: Science 324 (2009) 349–352.
• 14)   Q.  Wei,  S.  Cheng,  K.T.  Ramesh and  E.  Ma: Mater. Sci. Eng. A 381 (2004) 71–79.
• 15)   T.  Kunimine,  N.  Takata,  N.  Tsuji,  T.  Fujii,  M.  Kato and  S.  Onaka: Mater. Trans. 50 (2009) 64–69.
• 16)   M.  Kato,  T.  Fujii and  S.  Onaka: Mater. Trans. 49 (2008) 1278–1283.
• 17)   M.  Kato: Mater. Sci. Eng. A 516 (2009) 276–282.
• 18)   T.  Kunimine,  T.  Aragaki,  T.  Fujii,  S.  Onaka and  M.  Kato: J. Mater. Sci. 46 (2011) 4302–4307.
• 19)   E.W.  Hart: Acta Metall. 15 (1967) 351–355.
• 20)   Y.M.  Wang and  E.  Ma: Mater. Sci. Eng. A 375–377 (2004) 46–52.