Feasibility of Roll-Threading Superelastic Cu-Al-Mn SMA Rods

Kise, Sumio; Kataoka, Nanami; Takamatsu, Ryo; Nishida, Minoru; Omori, Toshihiro; Kainuma, Ryosuke; Araki, Yoshikazu

doi:10.1061/(ASCE)MT.1943-5533.0003874

Open access

Technical Papers

Jul 15, 2021

Feasibility of Roll-Threading Superelastic Cu-Al-Mn SMA Rods

Authors: Sumio Kise [email protected], Nanami Kataoka [email protected], Ryo Takamatsu [email protected], Minoru Nishida [email protected], Toshihiro Omori [email protected], Ryosuke Kainuma [email protected], and Yoshikazu Araki [email protected]Author Affiliations

Publication: Journal of Materials in Civil Engineering

Volume 33, Issue 9

https://doi.org/10.1061/(ASCE)MT.1943-5533.0003874

PDF

Abstract

This paper examines the feasibility of roll-threading superelastic Cu-Al-Mn shape memory allow (SMA) rods. The key idea behind this study was to perform roll-threading between quenching and aging, where cold workability of the material is relatively high. To obtain excellent superelasticity, single-crystal superelastic Cu-Al-Mn SMA rods were prepared with 16-mm diameter and 300-mm length. Threadability was studied by dimensional inspection and surface observations. Mechanical properties of the threaded rods were examined by cyclic tension tests. Comparisons between rolled and cut threads were made by cutting each single-crystal rod into halves to obtain two rod specimens having the same crystal orientation, thus ensuring fair comparison by excluding the influence of variations in orientations of the single-crystal rod. From the study, it was demonstrated that roll-threading of Cu-Al-Mn SMA rods was possible if roll-threading is performed between quenching and aging. Since additional heat treatment from the ordinary manufacturing process is unnecessary, roll-threading of Cu-Al-Mn SMA rods is as easy as that of steel rods. It was also demonstrated that the roll threads had significantly superior fatigue resistance compared with the cut threads of the same size. These results suggest that superelastic Cu-Al-Mn SMA rods with rolled threads are suitable for mass production to be used in structural and earthquake engineering applications.

Introduction

In recent decades, application of superelastic shape memory alloys (SMAs), has gained significant attention within the structural and earthquake engineering community. SMAs, which have the property of shape recovery upon unloading, are being studied for their potential to improve the resilience of large-scale structures, like buildings and bridges, against earthquakes (Ozbulut et al. 2011; Chang and Araki 2016; Lecce and Concilio 2014). In such applications, splicing superelastic SMA rods to steel rods poses a major challenge. Welding is one of the most widely used techniques for splicing steel rods. Nevertheless, welding between steel and Ni-Ti SMAs, the most widely used SMA type, is difficult, and many issues need to be addressed for their use in structural engineering applications (Oliveira et al. 2017a, b). Mechanical splicing with threaded ends is also one of the most widely used techniques for splicing steel rods. Due to the hardness of the material, however, cut-threading Ni-Ti SMA rods is difficult, which makes it an expensive option (DesRoches et al. 2004; Tyber et al. 2007; McCormick et al. 2007). Recently, feasibility of mechanical splicing with headed ends has been studied as a cost-effective option for splicing Ni-Ti SMA rods to steel rods (Nakashoji and Saiidi 2014). Nonetheless, brittle fracture was often observed around the heat-treated headed ends, which remains an issue.

Cu-based SMAs, including Cu-Al-Ni, Cu-Zn-Al, and Cu-Al-Be SMAs are alternatives to Ni-Ti SMAs (Miyazaki and Otsuka 1989; Sittner and Novac 2000; de Castro Bubani et al. 2013). However, one of the disadvantages of the Cu-based SMAs is their brittleness in bulk polycrystals (Humbeeck 2001; Recarte et al. 2002; Chen et al. 2009). Shape memory properties have been investigated in Cu-Al-Mn alloys with Al concentration higher than 20 atomic percent (at.%) and Mn concentration around 6 at.%, and it was pointed out that their ductility is low due to high degree of order (Mellor et al. 1986; Dutkiewicz et al. 1992; Kato et al. 2000; Mallik and Sampath 2007; Yang et al. 2019). To resolve the issues inherent in superelastic Ni-Ti SMAs—e.g., low cold workability (or brittleness), low machinability, and high material cost—superelastic Cu-Al-Mn SMA rods have been developed (Kainuma et al. 1996, 1998; Sutou et al. 2002, 2005, 2013; Omori et al. 2013; Kusama et al. 2017). Studies have also been done on how the material can be used for improving the response of civil structures, like buildings and bridges, against earthquakes (Shrestha et al. 2013; Hosseini et al. 2015; Araki et al. 2016; Varela and Saiidi 2017). Superelasticity is enhanced in a bamboo or columnar grained structure (Kainuma et al. 1998; Sutou et al. 2002; Liu et al. 2015), but the best superelasticity can be obtained in a single crystal in Cu-Al-Mn SMAs (Omori et al. 2020; Kise et al. 2021), which can be produced with cyclic heat treatment (Omori et al. 2013; Kusama et al. 2017). Similar to Ni-Ti SMAs, welding of Cu-Al-Mn SMAs to steel is difficult (Oliveira et al. 2016a, b, 2017a, b). On the other hand, cut-threading of superelastic Cu-Al-Mn SMA rods is as easy as that of steel rods due to their high machinability. To avoid brittle fracture at the cut threads, however, machining of SMA rods into a dog-bone shape is necessary, as demonstrated in superelastic Ni-Ti SMA rods (DesRoches et al. 2004; McCormick et al. 2007); this process requires significant time and cost and wastes substantial amounts of the material. To address these issues, the feasibility of mechanical splicing was studied between steel rods and Cu-Al-Mn SMA rods with headed ends (Kise et al. 2018). In the feasibility study, it was demonstrated that ductile fracture took place apart from the heat-treated headed ends by controlling the microstructure. This makes mechanical splicing of Cu-Al-Mn SMA rods with headed ends a promising option.

In mechanical splicing of steel rods, roll-threading is the most widely used technique due to advantages such as short labor time, low cost, and no waste of the material. Nevertheless, roll-threading superelastic Ni-Ti SMAs is difficult—Ni-Ti SMAs are strengthened by grain refinement to achieve excellent superelasticity, and additional heat treatment, such as annealing, is therefore needed to reduce the strength before roll-threading. However, similar to the heading (Nakashoji and Saiidi 2014), heat treatment increases the risk of brittle fracture around the boundary between the non-heat-treated center portion and the heat-treated end portions. As such, roll-threading superelastic Ni-Ti SMAs is not the best option for promoting cost effectiveness or improving mechanical properties. To the authors’ knowledge, no study has been reported on roll-threading of superelastic Ni-Ti SMAs.

On the other hand, the process of producing superelastic Cu-Al-Mn SMAs is significantly different from that of Ni-Ti SMAs. To produce superelastic Cu-Al-Mn SMAs, the material is solution-heat-treated and quenched from the high temperature around 1,173 K, and subsequently aged between 373 and 473 K (Kainuma et al. 1996, 1998; Sutou et al. 2002, 2005, 2013; Omori et al. 2013; Kusama et al. 2017). After quenching, the material has the ordered body centered cubic (BCC)

L 2_{1}

structure (Fig. 1) with lower degree of order (Kainuma et al. 1996), with low superelasticity and relatively high cold workability. After aging, on the other hand, the material has the higher degree of order, with excellent superelasticity and low cold workability. Our preliminary study attempted to roll-thread Cu-Al-Mn SMA rods after all heat treatment, including aging, was completed; however, the attempt was unsuccessful due to the excellent shape recovery property obtained by aging.

Fig. 1. BCC ${L 2}_{1}$ structure of Cu-Al-Mn alloy.

The objective of the present work is to study the feasibility of roll-threading superelastic Cu-Al-Mn SMA rods. The key idea behind this study is to perform roll-threading between quenching and aging, where cold workability of the material is relatively high. This method aims to roll-thread single-crystal Cu-Al-Mn SMA rods without losing their excellent superelasticity obtained by aging. Because no additional heat treatment from the ordinary manufacturing process is necessary, roll-threading is cost effective and suitable for mass production—this is the key difference from the heading process proposed previously (Kise et al. 2018). It has been long recognized that both the material cost itself and the processing cost for connection are the critical issues in commercializing SMAs in structural applications (Ozbulut et al. 2011; Chang and Araki 2016; Lecce and Concilio 2014). This work attempts to resolve this issue by developing a cost-effective method for connecting steel with Cu-Al-Mn SMAs, whose material cost is only a fraction of the more popular Ni-Ti SMAs. In this study, single-crystal superelastic Cu-Al-Mn SMA rods with 16-mm diameter and 300-mm length were prepared. Threadability was studied by dimensional inspection and surface observations using scanning electron microscopy (SEM). Mechanical properties of the threaded rods were examined with cyclic tension tests. Comparisons between rolled and cut threads were made by cutting each single-crystal rod into halves, to have two rod specimens with the same crystal orientation. This way, a fair comparison could be made by excluding the influence of variations in orientations of the single-crystal rods. It should be noted here that the use of huge single-crystal rods allows such comparisons, which is another key feature of this work. Microstructural analyses were also performed to study the fracture modes and strengthening mechanisms of the threaded portions.

Materials and Methods

Materials Preparation

The ingot of Cu-17 at.% Al-11.4 at.% Mn alloy, with 110-mm diameter and 700-mm length, was prepared from 99.99% Cu, 99.99% Al, and 99% Mn by melting in a high-frequency vacuum induction furnace. Then the ingot was hot-forged down to 40 mm diameter at 1,073 K. A long rod of 16-mm diameter was obtained by repeating cold-rolling and drawing at the cold working rate of 30%, with an intermediate annealing at 793 K for 60 min. The long rod was cut into rods of 300-mm length and then heat treated under air to produce single-crystal rods. In the heat treatment, the rods were first cooled down from the

β

phase at 1,173 K to the

α + β

phase at 773 K at the rate of

1 K / \min

. After keeping the temperature at 773 K for 1 h, the temperature was heated up to the

β

phase at 1,173 K at the rate of

2 K / \min

. This process was repeated several times, and the temperature was kept at 1,173 K for 6 h (Omori et al. 2013; Kusama et al. 2017). Then the rods were quenched with water. After the heat treatment, the rods were etched using an acid ferric chloride solution to locate the grain boundaries.

As shown in Fig. 2, seven single-crystal rods were selected wherein no grain boundaries were confirmed. Fig. 3 shows the crystal orientations of the seven rods measured by electron backscattered diffraction (EBSD). Five rods (#1–#5) were used to study the difference in superelasticity and fatigue resistance of rods with rolled and cut threads. One rod (#6) was used to examine the strengthening mechanism at the rolled threads. One rod (#7) was used to study the difference in threadability at the as-quenched and as-aged states.

Fig. 2. Photographs of the single-crystal Cu-Al-Mn alloy bars after heat treatment and etchings.

Fig. 3. Grain orientations along the longitudinal direction of each bar specimen are plotted on the inverse pole figure. The contour is plotted using the theoretical isotransformation strain obtained by Sutou et al. (2002).

From each single-crystal rod, two rods were obtained by cutting into halves for the comparison studies. Thus, there was a total of 14 rod specimens of 150-mm length. The diameter of all the rod specimens was 12.4 mm, except the threaded portions. The pitch and length of the M14 threads were 2 and 25 mm, respectively. In each pair of the rod specimens, roll-threading was performed on one specimen and cut-threading was performed on the other specimen. Superelasticity and fatigue resistance were examined using the rod specimens obtained from #1 to #5 rods. The rod specimens with rolled threads were called Specimens R1–R5, and those with cut threads were called Specimens C1–C5. Fig. 4 shows the photographs of the roll-threading machine and the roll-threading process of a Cu-Al-Mn SMA rod. Fig. 5(a) shows the schematic drawing of threading rod specimen. Fig. 5(b) shows the rod specimens before roll-threading, and Figs. 5(c and d) show the rod specimens after roll- and cut-threading. Specimens R6 and C6, having rolled and cut threads, respectively, were used for studying strengthening mechanisms at the threaded portions. The rod specimens without and with aging were called Specimens AQ7 (as-quenched) and AA7 (as-aged), respectively. Neither threading nor reduction of section was performed to Specimens AQ7 and AA7. All the specimens except AQ7 were aged at 423 K for 0.5 h.

Fig. 4. Photographs of roll-threading process: (a) thread rolling machine; (b) initial state; (c) intermediate state; and (d) final state.

Fig. 5. Schematic drawing and photographs of bar specimens: (a) drawing; (b) before roll-threading; (c) after roll-threading; and (d) after cut-threading.

Test Methods

The test matrix for each specimen is shown in Table 1. To examine threadability, dimensional inspection was performed for the rolled and cut threads. In the dimensional inspection, the major diameter was directly measured, and the pitch and minor diameters were checked using go/no go ring gauges. The arithmetic average roughness of the thread surface was measured using a surface roughness measurement tester. SEM analysis was also performed to observe the surface of the threads.

Table 1. Test matrix for each specimen

Tests	Specimen number and state of specimen in tests AQ (as-quenched), AA (as-aged)
Tests	#1	#2	#3	#4	#5	#6	#7
Crystal orientation	AQ	AQ	AQ	AQ	AQ	AQ	AQ
Ring gauge tests of rolled/cut threads	AQ	AQ	AQ	AQ	AQ	AQ	—
SEM observation at rolled/cut threads	—	—	—	—	—	AQ	—
Cyclic tension tests of bar specimens with rolled/cut threads	AA	AA	AA	AA	AA	—	—
Cyclic tension tests of bar specimens without threads	—	—	—	—	—	—	AQ
Cyclic tension tests of bar specimens without threads	—	—	—	—	—	—	AA
Transformation temperature	—	—	—	—	—	—	AQ
Transformation temperature	—	—	—	—	—	—	AA
TEM observation	—	—	—	—	—	—	AQ
TEM observation	—	—	—	—	—	—	AA
Hardness at rolled/cut threads	—	—	—	—	—	AA	—
SEM-EBSD observation at rolled threads	—	—	—	—	—	AA	—

To compare the mechanical properties of rod specimens with rolled and cut threads, two types of cyclic tension tests were performed at room temperature. First, cyclic tension with monotonically increasing amplitude was applied to Specimens R1–R5 with rolled threads and Specimens C1–C5 with cut threads to study their superelasticity, and to obtain the material constants shown in Table 2. In the cyclic tension tests, each rod specimen was elongated to the target strain and then unloaded to 0 stress. Only 1 cycle of loading was applied at each target strain. The target strain was increased from 1% to 10% with an increment of 1%. During the whole loading cycle, the strain rate was kept constant at

3 \times 10^{- 4} s^{- 1}

. Strain was measured using a noncontact video extensometer with gauge length of 50 mm. Fig. 6(a) shows a rod specimen with nuts for grip and Fig. 6(b) shows the test setup.

Table 2. Young’s modulus

E

, transition stress

σ_{t}

, superelastic strain

ε_{s e}

, residual strain at 5% applied strain

ε_{r \cdot 5}

, and residual strain at 10% applied strain

ε_{r \cdot 10}

Sample	$E$ (GPa)		$σ_{t}$ (MPa)		$ε_{s e}$ (%)		$ε_{r \cdot 5}$ (%)		$ε_{r \cdot 10}$ (%)
Sample	Rolled	Cut	Rolled	Cut	Rolled	Cut	Rolled	Cut	Rolled	Cut
#1	43	33	180	181	7.8	7.4	0.30	0.29	1.10	0.32
#2	43	30	160	165	7.8	7.6	0.05	0.05	0.10	1.12
#3	21	19	188	175	8.4	8.6	0.11	0.10	0.12	0.14
#4	60	57	222	235	6.5	—	0.10	—	—	—
#5	27	25	202	212	6.4	7.0	0.17	0.20	0.54	1.09

Fig. 6. (a) Bare specimen with nuts; and (b) test setup.

Second, cyclic tension tests with constant amplitude were performed to compare the fatigue resistance of the rod specimens with rolled and cut threads. Here, the target strain amplitude was fixed to 5% during all loading cycles. In each loading cycle, forced displacement was applied until the target strain amplitude was reached, and then the specimen was unloaded to 0 stress. The strain rate was controlled to be

1.2 \times 10^{- 3} s^{- 1}

during all loading cycles. Loading was terminated after 500 loading cycles were applied or when the specimen fractured.

No standard is available for cyclic tension tests of superelastic SMA rods. In the first set of cyclic tension tests, the number and amplitude of loading cycles were determined to obtain the fundamental material constants shown in Table 2 in linear elastic and superelastic responses. The strain rate was selected to be slow enough to realize isothermal condition. In the second set of cyclic tension tests, on the other hand, the number and amplitude of loading cycles were determined to reflect the conditions used in seismic design of buildings and bridges (Shrestha et al. 2013; Hosseini et al. 2015; Araki et al. 2016; Chang and Araki 2016; Varela and Saiidi 2017). The strain rate was determined to be as fast as possible within the capacity of the loading machine.

Results

Threadability

From the results of the dimensional inspection, a minor difference was observed between the maximum and minimum major diameters (13.3 and 13.0 mm) of rolled threads, while such a difference was not observed in the major diameter (13.6 mm) of cut threads. The deviation from the true circle observed in the rolled threads arose from the anisotropy in plasticity of single-crystal Cu-Al-Mn SMAs (Kise et al. 2021). All the threaded rod specimens with rolled and cut threads passed the go/no go ring gauge tests by ISO 1502 (ISO 2008), which indicates that pitch and minor diameters of the threads satisfied the standard requirements.

Fig. 7 shows the comparison of the surfaces of the rolled and cut threads. Figs. 7(a and b) show the rolled and cut threads, respectively. Figs. 7(c and d) show the SEM views of the regions indicated by white rectangles shown in Figs. 7(a and b). For the rolled threads, the average and standard deviation values of the arithmetic average roughness obtained from three measurements were 0.24 and 0.03 μm, respectively. These values for the cut threads were 0.46 and 0.10 μm, which are significantly larger than those for the rolled threads. These results agree with the SEM view shown in Figs. 7(c and d), where the surface of the rolled thread is much smoother than that of the cut thread.

Fig. 7. Comparison of the surface of rolled and cut threads: (a) photo of rolled thread; (b) photo of cut thread; (c) SEM view of rolled thread; and (d) SEM view of cut thread.

Mechanical Properties

Fig. 8 shows the stress–strain curves under cyclic tension loads with monotonically increasing strain amplitude. Fig. 9 shows the photographs of the rod specimens after the cyclic tension tests. Table 2 shows the values of the representative material constants obtained from the cyclic tension tests. Fig. 10 shows the definitions of the variables written in Table 2. Excellent superelasticity can be seen in each rod specimen, as shown in Fig. 8. It can also be seen from Fig. 8 that each pair of the rod specimens cut from the same rod—e.g., R1 and C1—have similar stress–strain curves. The variations in the stress–strain curves of the single-crystal rods (#1–#5) can be explained by their dependence on the crystal orientation (Sutou et al. 2002; Omori et al. 2020; Kise et al. 2021). Specimen C4 fractured before 3% target strain. In testing Specimen R4, the cyclic loading was terminated before 8% target strain because the tensile load reached the loading capacity of the testing machine.

Fig. 8. Cyclic stress–strain curves of the bar specimens with (a) rolled threads; and (b) cut threads.

Fig. 9. Photographs of the bar specimens after the tension tests to examine superelasticity: (a) bar specimens with rolled threads; and (b) bar specimens with cut threads.

Fig. 10. Schematic illustration of the definitions of the material parameters under cyclic tension loads.

Figs. 11 and 12 show the stress–strain curves at representative loading cycles obtained from the fatigue tests. None of the rod specimens with rolled threads fractured for 500 cycles, while all the rod specimens with cut threads fractured before reaching 500 cycles. Fig. 12 also shows the number of the loading cycles at which fracture occurred. These results clearly demonstrate the superior fatigue resistance of rolled threads to cut threads in superelastic Cu-Al-Mn SMA rods.

Fig. 11. Stress–strain curves at representative loading cycles for the bar specimens with rolled threads.

Fig. 12. Stress–strain curves at representative loading cycles for the bar specimens with cut threads.

Discussions

Threadability

This section discusses in more detail the workability in roll-threading Cu-Al-Mn SMA rods before and after aging. Fig. 13 shows the stress–strain curves obtained from the tension tests of Specimens AQ7 (as-quenched) and AA7 (as-aged). Here, the loading protocols of the tension tests are the same as those for the tension tests of R1–R5 to study their superelasticity, shown in the section “Test Method.” The only difference is that no threading was performed to Specimens AQ7 and AA7. Fig. 14 plots the relationship between the recoverable and applied strains for Specimens AQ7 and AA7. From these figures, it is observed that plastic deformation took place in Specimen AQ7 after the strain reached 5%, while little plastic deformation can be seen in Specimen AA7 even after the strain reached 10%. These results confirm that workability in roll-threading of the Cu-Al-Mn SMA rod before aging is significantly higher than that after aging. Although excellent cold workability in as-quenched polycrystal Cu-Al-Mn SMAs has been reported by Kainuma et al. (1996) and Babacan et al. (2018), a comparison of superelastic and plastic responses between the single-crystal rod specimens before and after aging was made for the first time in the current study.

Fig. 13. Stress–strain curves: (a) Specimen AQ7 (as-quenched); and (b) Specimen AA7 (as-aged).

Fig. 14. Relationship between the recoverable strain and the applied strain of Specimens AQ7 (as-quenched) and AA7 (as-aged).

To examine the change in the degree of order directly, transmission electron microscopy (TEM) observations were performed. Figs. 15(a and b) show the selected area electron diffraction pattern (SAEDP) taken from as-quenched and an as-aged specimens, respectively. Although the ordered reflections of the

L 2_{1}

structure are already observed in the as-quenched specimen, as indicated by the arrow in Fig. 15(a), their intensity is very weak. On the other hand, much stronger and sharper ordered reflections can be observed in the as-aged specimen. These results suggest that the degree of order in the as-quenched specimen was lower, and that in the as-aged specimen increased significantly. These observations in the present single-crystal specimens are consistent with those for polycrystalline specimens reported by Kainuma et al. (1996).

Fig. 15. Selected area electron diffraction patterns: (a) Specimen AQ7 (as-quenched); and (b) Specimen AA7 (as-aged).

Fig. 16 shows the results of the differential scanning calorimetry (DSC) measurements. The austenite finish temperature

A_{f}

of the as-quenched specimen was 165.1 K, while that of the as-aged specimen was 206.2 K. Because the

A_{f}

of each specimen is much lower than the room temperature, the residual deformation caused by roll-threading is plastic deformation, and hence cannot be recovered by heating. It can be also observed from Fig. 16 that the transformation temperatures of the as-quenched specimen were lower than those of the as-aged specimen. This result for the single-crystal specimens is consistent with the literature (Kainuma et al. 1995; Babacan et al. 2018), wherein Kainuma et al. (1995) reported that the increase of transformation temperature after aging is caused by the increase of the degree of order in polycrystal Cu-Al-Mn alloys. As shown in Fig. 13, the transformation stress of Specimen AQ7 is about 300 MPa, while that of Specimen AA7 is about 200 MPa. Such a difference between the transformation stresses of Specimens AQ7 and AA7 is due to the difference in the transformation temperature in these specimens discussed above. In general, plastic deformation takes place when the critical shear stress for slip is lower than the transformation stress in SMAs (Miyazaki and Otsuka 1989). Hence, the higher transformation stress is another possible reason for the plastic deformation observed in Specimen AQ7. These results support the suitability of roll-threading at the as-quenched state from the viewpoint of crystal structure. One may argue that it is better to perform roll-threading at the

α + β

phase before the solution-heat treatment because Cu-Al-Mn alloy has excellent cold workability (Kainuma et al. 1996). However, this was intentionally avoided here because the rolled threads can deform significantly during the solution-heat treatment, and the high strain at the rolled threads can prevent crystal growth to obtain single crystals (Kise et al. 2018).

Fig. 16. DSC curves and transformation temperatures ( $M_{f}$ , martensite finish temperature; $M_{s}$ , martensite start temperature; $A_{f}$ , austenite finish temperature; $A_{s}$ , austenite start temperature) of (a) Specimen AQ7 (as-quenched); and (b) Specimen AA7 (as-aged).

Mechanical Properties

Figs. 11 and 12 demonstrate that the fatigue resistance of rolled threads is superior to that of cut threads. To explain the reasons for this finding, the strengthening mechanism is examined in the rolled threads and the fracture modes in the cut threads.

Strengthening Mechanism in Rolled Threads

To examine the distribution of strength in rolled threads, Vickers hardness was measured for Specimen R6. For comparison, Vickers hardness was also measured for Specimen C6 with cut threads. Each specimen was cut into halves along the longitudinal axis. Fig. 17 shows the distribution of the measured Vickers hardness values in the rolled and cut threads. Contour lines are also shown in Fig. 17(a) to show the distribution of the hardness values more clearly. Vickers hardness values measured at the center axis of the rolled and cut threads, not shown in the figure, were HV228 and HV227, respectively. Significant increase can be observed in the hardness at the surface of the rolled thread. On the other hand, little variation can be seen in the hardness of the cut thread.

Fig. 17. Distribution of the measured values of Vickers hardness at the cross section in (a) rolled thread; and (b) cut thread.

Fig. 18 shows the results of the microstructure analyses of the rolled thread. From the SEM view shown in Fig. 18(a), the fiber flow can be observed. Figs. 18(b and c) show the inverse pole figure (IPF) and the grain reference orientation deviation (GROD) map obtained from the EBSD analysis. From these figures, large residual strains can be observed. The white arrows shown in Fig. 18(c) indicate the estimated direction of plastic deformations caused by roll-threading. From Figs. 17 and 18, it can be inferred that the plastic deformations due to roll-threading increased the strength of the roll thread.

Fig. 18. Microstructure of the rolled thread: (a) SEM view; (b) IPF map; and (c) GROD map.

Fracture Mode in Cut Threads

Fig. 19 shows the fracture surface of the specimens with cut threads. Figs. 20(a and b) show the SEM view of Specimens C1 and C4. The arrows in Fig. 19 indicate the fracture origins. It can be observed from Fig. 19 that fracture took place at the boundary between the threaded and straight portions in all the rod specimens. The fracture modes can be classified into the following two types: (1) brittle cleavage fracture, characterized by the river pattern shown in Fig. 20(b), took place in Specimen C4 at the first cycle; and (2) fatigue fracture occurred in Specimens C1, C2, C3, and C5 at the 335th, 226th, 14th, and 406th cycles, respectively. As shown in Fig. 19, the fracture surface of the specimens with fatigue fracture can be divided into three regions (Becker and Lampman 2002). In Region I, a crack propagated along the plane about 45° from the longitudinal axis, where the shear stress is maximum. In Region II, a crack propagated along the plane 90° from the longitudinal axis. Striation, typically seen in fatigue fracture surface of metals, can be seen in Regions I and II. The rough surface in Region III indicates that the final brittle fracture took place in this region.

Fig. 19. Photographs of the fractured section of the bar specimens with cut threads.

Fig. 20. SEM views of the fracture surface: (a) Specimen C1; and (b) Specimen C4.

Conclusions

This work examined the feasibility of roll-threading superelastic Cu-Al-Mn SMA rods. The following conclusions can be drawn from the study. (1) Roll-threading of Cu-Al-Mn SMA rods is possible if roll-threading is performed between quenching and aging. Because no additional heat treatment from the ordinary manufacturing process is necessary, roll-threading of Cu-Al-Mn SMA rods is as easy as that of steel rods. (2) The lower degree of

L 2_{1}

order and lower stability of martensite contribute to the threadability in the as-quenched condition. (3) The rolled threads have significantly superior fatigue resistance compared to cut threads having the same size, due to work hardening. No fracture took place in the rod specimens with roll threads during the fatigue tests with 500 cycles of 5% target strain. On the other hand, brittle or fatigue fracture took place in the rod specimens with cut threads. These results suggest that roll-threading superelastic Cu-Al-Mn SMA rods is feasible and suitable for structural and earthquake engineering applications.

Data Availability Statement

Some or all data, models, or code that support the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgments

This research was supported by Fund for the Promotion of International Research No. 16KK0149, provided by the Japan Society for the Promotion of Science (JSPS). The authors would like to thank Dr. Shigekazu Yokoyama and Dr. Toyohiko Higashida of Sekisui House, Ltd., and Mr. Kouji Ishikawa of Furukawa Techno Material Co., Ltd. (FTM), for their helpful discussions. The authors appreciate assistance with in the experiments from Mr. Yuji Oizumi, Ms. Satomi Norimitsu, and Mr. Kenji Uruma of FTM.

References

Araki, Y., K. C. Shrestha, N. Maekawa, Y. Koetaka, T. Omori, and R. Kainuma. 2016. “Shaking table tests of steel frame with superelastic Cu-Al-Mn SMA tension braces.” Earthquake Eng. Struct. Dyn. 45 (2): 297–314. https://doi.org/10.1002/eqe.2659.

Abstract

Introduction

Materials and Methods

Materials Preparation

Test Methods

Results

Threadability

Mechanical Properties

Discussions

Threadability

Mechanical Properties

Strengthening Mechanism in Rolled Threads

Fracture Mode in Cut Threads

Conclusions

Data Availability Statement

Acknowledgments

References

Information

Published In

Copyright

History

Authors

Affiliations

Metrics

Citations

Download citation

Cited by

Figures

Other

Share

Copy the content Link

Share with email

Share

Request Username

Create a new account

Change Password

Password Changed Successfully

Verify Phone

Congrats!