1. Introduction
Traditional low temperature superconducting (LTS) materials like NbTi and Nb3Sn have been manufactured into cable-in-conduit conductors (CICC) [1], [2], and have been applied in the magnet system of fusion devices, such as the Experimental Advanced Superconducting Tokamak (EAST) [3], the Japan Trous-60 Super Advanced (JT-60SA) [4], the International Thermonuclear Experimental Reactor (ITER) [5], and so on. However, future fusion reactors require high magnetic-field (>20 T), long plasma confinement time, and lower construction and operating costs [6], [7]. The properties of LTS materials are limited under high magnetic-field due to their relatively low engineering current density (JE) and mechanical strength. LTS materials have relatively low upper critical field and insufficient pinning under high filed, and are thus not suitable for high-field (>20 T) applications. In addition, the cost of operating in liquid Helium is also high [7]. Fortunately, the remarkable progress of commercially available second generation (2G) high-temperature superconducting (HTS) tape with JE > 1000 A/mm2 at 4.2 K, 18-20 T [8], [9], and mechanical tensile strength > 998MPa at 77 K [10] has enabled more possibilities for the fusion industry.
Despite the excellent performances of 2G HTS tapes in high-field, single tapes cannot provide operating currents in orders of several kilo-amperes or withstand extremely large electromagnetic forces. Therefore, it’s necessary to combine multiple tapes into cables to maximize current capacity and enhance mechanical strength by structural materials. Three classical and distinctive cabling concepts have been proposed and developed, including the Twisted Stacked-tape Cable (TSTC) [11], the Conductor on Round Core (CORC) [12] and the Roebel Assembled Coated Conductor (RACC) [13]. Along with these concepts, several new cable structures have also been explored; for example, HTS Cross Conductor (HTS-CroCo) [14], Quasi-isotropic Strand (QI-S) [15], VIPER [16], SaMiT [17], FAIR [18], STARS [19], WISE [20], and others.
These cables offer currents in the range of several kA to more than 100 kA, meeting the requirements of the magnet system of fusion reactors. For applications in high fields, the mechanical properties of these cables are also a concern. As a result, the electromechanical performances of different HTS cables with different architectures have been widely studied [15], [17], [18], [19], [20], [21], [22], [23], [24]. It is well known that the critical current performance of the HTS cable is highly dependent on the applied stress and strain caused by various loads, such as tensile, bending and twisting load.
Among of them, the TSTC-type cable has the potential for very high current densities and is easily scalable to high currents as well [14]. Therefore, referring to the technical path of TSTC, a novel round strand comprised of an HTS tape stack and an enclosing copper tube has been developed. Inside the tube, solder Sn63Pb37 is used to fill all the gaps and ensure the electrical connection and mechanical strength. To characterize the electromechanical behavior of the round strand under the different load conditions, several samples have been prepared and their electromechanical characterization has been performed at 77 K.
The experimental investigations presented in this work focus on the effect of axial tensile stress/strain and fatigue cycles on the critical current (Ic) of the round strand at 77 K and self-field, which is extremely critical for the high-filed magnet design and operation. To further understand the reasons and mechanism of Ic degradation, the microscopic defects of the round strand samples were investigated before and after the experiment. The results obtained from these investigations provide valuable insights into the electromechanical limits of the round strand and pave the way for future efforts to enhance its mechanical strength.
2. Experimental
2.1. Sample preparation of round strand
The round strand layout is comprised of HTS tapes, filling material (Sn63Pb37 solder) and the external copper tube. The geometric arrangement for strand is depicted in Fig. 1(a). The narrow tapes with 2 mm width are used to forming a square wire by stacking and soldering. Among these tapes, in order to reduce the bending strain of the strand, the superconducting layer of the upper 10 tapes is placed downwards, and the lower 10 tapes is placed upwards. The external copper tube has inner and outer diameters of 3 mm and 4 mm, respectively, providing both mechanical and electrical stability. Soldering ensures good conductivity between individual HTS tapes within the tape stack and between HTS stack and the copper tube. The tapes used in the round strands were commercially obtained from Suzhou advanced material research institute (Samri), which manufactured by the method of metal oxide chemical vapor deposition (MOVCD). The structural configuration of the HTS tapes is shown in Fig. 1(b). It mainly contains a 3.1 μm REBCO layer deposited on a 60 μm thick Hastelloy substrate, two 1 μm Ag layer on top of the REBCO layer and the bottom of the substrate layer, and finally electroplated 12 μm thick copper stabilizer for each side, respectively. The critical currents of tapes ranging from ∼ 70 A to ∼ 90 A. The main specifications of the round strand used are listed in Table 1.
Fig. 1. (a) Structure of the round strand: 3-D structure model (left) and the cross section (right), and (b) Structural configuration of the MOVCD REBCO tape. |
Table 1. Main Specifications of a Round Strand. |
| Parameters | Value |
|---|---|
| Thickness of tapes | 0.09 mm |
| Width of tapes | 2 mm |
| Thickness of REBCO layer | 3.1 μm |
| Number of tapes | 20 |
| Critical current (77 K, self-field) | ∼70-90 A |
| Tensile Strain | 0.45% |
| Tensile Stress | >550 MPa |
| Diameter of the strand | 4 mm |
| Inner diameter of copper tube | 3 mm |
| Thickness of copper tube | 0.5 mm |
| Thickness of the tape stack | 2 mm |
| width of the tape stack | 2 mm |
The key fabrication process is depicted in Fig. 2. Firstly, all tapes are uniformly pre-tined with eutectic Sn63Pb37 solder to facilitate the soldering process. Secondly, the tinned tapes are neatly stacked into the soldering mold and then compressed and solidified by heating the mold to about 200 ℃ and subsequently cooling it down. Lastly, the stacked square wire is inserted into the copper tube and filled with Sn63Pb37 solder to form a round strand. The length of the samples is approximately 500 mm. Photographs of the samples and cross-sections of the round strand are presented in Fig. 2(3).
Fig. 2. Fabrication process of HTS round strand sample using 2G HTS tapes. |
2.2. Monotonic axial tensile testing instrument
Three 500 mm long samples of round strand were meticulously prepared for the purpose of conducting monotonic axial tensile tests. The testing set-up for these round strand samples is visually depicted in Fig. 3. To ensure both electrical and mechanical connections, a specialized termination made of chromium zirconium copper, known for its exceptional mechanical strength and excellent electrical conductivity, was employed to securely clamp the samples at their ends. The copper terminals, with a length of 90 mm, were specifically chosen to facilitate the transfer of current, ensuring that it flows uniformly throughout the superconducting layer. To measure the elongation of the round strand samples, an epsilon's extensometer with a gauge length of 50 mm and a measuring range of ±10% was utilized. This extensometer was carefully positioned at the center of the samples, and the resulting signals were recorded separately. Additionally, in order to gather voltage data, a pair of voltage taps was directly soldered onto the sample surface at the same location as the extensometer.
Fig. 3. Illustration of the electrical connections and extensometer setup. |
To gain deeper insights into the behavior of critical current under increasing external tensile loading, our research involved conducting tests using a 100 kN low temperature tensile testing machine equipped with a cryostat (see Fig. 4(a)). The samples were subjected to tensile loading through the tension bar and load cell of the testing machine. Fig. 4(b) illustrates the setup of the experimental samples on the test fixture.
Fig. 4. (a) 100 kN low temperature tensile testing machine, and (b) tensile test setup. |
Prior to testing, the entire assembly was immersed in liquid nitrogen to maintain low temperatures. Tensile load was then applied gradually, with incremental increases, while measuring the V-I (voltage-current) curves at various strains. Importantly, the applied tensile loading was not released between critical current (Ic) measurements, ensuring a continuous assessment of the samples' behavior.
2.3. Axial fatigue testing set up
To investigate the behavior of the samples under cyclic tensile loading, an additional set of three samples was prepared for axial tensile fatigue testing. The experimental setup and cryostat used were the same as described in Section 2.1 and shown in Fig. 5. However, in this fatigue testing, the extensometer was not installed due to the cyclic stress load. The cyclic load was applied externally using a 100 kN low temperature fatigue testing machine (Fig. 5(a)) within the cryostat. As depicted in Fig. 5(b), the sample was mounted and then immersed in liquid nitrogen. Different amplitudes of cyclic tensile stresses were applied to investigate the effect on the critical current (Ic), using the tension bar and load cell of the fatigue machine.
Fig. 5. (a) 100 kN low temperature fatigue testing machine, and (b) fatigue test setup. |
3. Results and discussion
To assess the axial Tensile and fatigue behaviors, total six samples were manufactured, and all measurements were conducted at 77 K and in self-field. The applied tensile stress and resulting strain on the samples were determined using Equations (1), (2) [25], respectively:
$\sigma=\frac{F}{A}$
And
$\varepsilon=\frac{\Delta l}{\bar{l}}$
where σ is the tensile stress in MPa, F is the applied tensile force in N, A is the initial cross-sectional areas of the sample; ε is the tensile strain, $Δl$ is the elongation of the extensometer, l is the gauge length of the extensometer.
$E(I)=E_{C}\left(\frac{I}{I_{C}}\right)^{n}$
where Ec is the critical current criterion usually taken as 1 μV/cm and the exponent n ≫ 1. The n-index comes from the emperical power law in voltage-current measurements of superconductors. It gives an indication of how the electric field depends on the current in the transition region. An ideal superconducting-normal transition usually results in an infinite n-index. Here the behavior of n-index value from large to small can indirectly explain the damage behavior in the REBCO tape under the tensile deformation. Before the tests, the original Ic values of all samples were measured. The results ranged were between 1054A and 1090A for all samples, which aligned with the expected values based on the properties of the single HTS tape. The initial V-I curve of sample#1 to sample#6, as well as the fitted Ic and n-value, are shown in Fig. 6.
Fig. 6. The original V-I curve of sample#1 to sample#6, as well as the fitted Ic and n-values by the power law. |
3.1. Monotonic axial tensile testing results
The round strand samples (sample#1, sample#2, sample#3) were subjected to uniaxial tension at 77 K with self-field to determine the stress-strain characteristics and mechanical properties. The dependence of Ic on axial tensile stress under monotonic loading was measured, and the critical stress and strain were determined. The tensile force applied had an increment of approximately 32 MPa. The results of the normalized critical current (Ic/Ic0) and n values versus applied axial stress for different samples are shown in Fig. 7(a) and Fig. 7(b). The measurements revealed that sample#2 degraded first. As previously defined, the maximum and minimum critical tensile stresses for the three samples were 382 MPa and 300 MPa, respectively. In case of the tensile stresses exceeded the critical values, a sharp drop in the normalized critical current and n values was observed.
Fig. 7. (a) Normalized critical current versus axial tensile stresses for different samples, (b) n-value versus axial stress, (c) Axial tensile stresses and normalized critical currents as a function of the axial tensile strain, and (d) Axial tensile stresses versus the axial tensile strain for different samples. |
Fig. 7(c) illustrates the relationship between stress and Ic/Ic0 as a function of tensile strain for sample#1. The results indicate that the sample maintains 95% retention of Ic0 at tensile strains greater than 0.53%, with a corresponding tensile stress of approximately 382 MPa. This value exceeds the critical strain of 0.45% given by the tape manufacturer. The stress-strain curves for these samples are depicted in Fig. 7(d). By comparing these curves with the data presented in Fig. 7(a), it is evident that the critical tensile stresses and strains for sample#2 and sample#3 were approximately 300 MPa and 0.40%, and 350 MPa and 0.44%, respectively.
In this study, average tensile strength of 344 MPa and tensile strain of 0.47% were obtained for axial tensile tests in liquid nitrogen. The detailed experimental results of the round strand samples are shown in Table 2.
Table 2. The detailed experimental results of the round strand samples. |
| Parameters of 95% Ic0 | Sample#1 | Sample#2 | Sample#3 | Average value |
|---|---|---|---|---|
| Critical tensile stress (MPa) | ∼382 | ∼300 | ∼350 | ∼344 |
| Critical tensile strain (%) | ∼0.53 | ∼0.4 | ∼0.44 | ∼0.47 |
Based on the test results, it was observed that the superconducting layer of all three samples suffered significant damage under various tensile loads. Notably, sample#2 exhibited a much lower critical tensile stress and strain compared to the other samples. This could be attributed to two main factors. Firstly, incomplete air discharge during the filling and forming process of sample#2 may have led to the formation of numerous large cavities or gaps, thereby weakening the overall tensile strength of the round strand. Secondly, it is possible that the round strand experienced external nonuniform tensile and extrusion stresses or other loads during the forming process, resulting in high residual stress within the sample. Further investigation into the degradation of Ic will be presented in section 4.
3.2. Axial fatigue testing results
To investigate the effect of cyclic axial tensile stress on Ic in liquid nitrogen, three additional samples (sample#4 to sample#6) were utilized. Ic measurements were performed at different tensile loads, with the stress amplitude cycled between the peak stress and 10% of the peak stress. The value of the peak stress was chosen according to the results of the axial tensile test. The peak stress for the fatigue test of sample #4 was selected to be approximately 60% of the average critical tensile stress of samples#1 to #3. For sample #5, the peak stress was chosen to be the average of the critical tensile stress of samples#1 to #3, aiming to study the damage of the strand under extreme conditions. In the case of sample #6, the peak stress was set as 1.1 times the average of the three tensile samples, in order to study the fatigue characteristics under overload conditions. In order to ensure the safety of the future fusion devices or other devices during operation, these extreme performance tests of the strand are essential, which can provide important references for the design of conductors and devices. In addition, the fatigue tests involved subjecting the samples to tensile loads for up to 100,000 cycles, and the loads were cycled at frequencies of up to 10 Hz in a liquid nitrogen environment. The V-I curves were measured at intervals of every 2000 cycles till 50,000 cycles, and at intervals of every 5000 cycles after. The V-I curve has been provided in section3 (Results and discussion).
Fig. 8(a) illustrates the relationship between Ic/Ic0 and the number of cycles for sample#4, sample#5, and sample#6 at different tensile load amplitudes. For sample#4, a peak tensile stress of 200 MPa was applied, which corresponds to approximately 60% of the average critical tensile stress value of sample#1 to #3. No significant degradation of Ic/Ic0 was observed even after 100,000 cycles. For sample#5, tested at a peak stress of 340 MPa (close to the average critical stress of samples#1 to #3), exhibited a stable Ic performance up to 36,000 load cycles. However, fracture occurred suddenly at approximately 37,800 cycles, specifically at the upper clamping terminal, due to stress concentrations and higher frequency (as shown in Fig. 8(c)). The damage to the copper former of the strand resulted from stress concentration caused by excessive preloading during sample installation. As the peak stress increased to approximately the critical tensile stress of sample#1 (∼380 MPa), rapid decay of Ic was observed after exceeding 16,000 loading cycles. The critical current decreased from approximately 95% to 70% of its initial value at 55,000 cycles. Fig. 8(b) displays the n-values against the number of cycles for sample#4 to #6 at different peak stresses. The evolution of the n-values mirrors the change in Ic/Ic0, both indicating the performance change of the samples to various fatigue loads.
Fig. 8. (a) Normalized critical current as a function of the number of axial tensile cycles at different peak stresses, (b) The n-value of different samples, and (c) Fracture photo of sample#4 at near the upper terminal. |
As a preliminary conclusion, fatigue properties of the round strand cable have been investigated. The peak stresses of the fatigue tests are about 90% of the critical stresses of the samples. Yet no significant degradation was observed after over 30,000 cycles, indicating robust mechanical behaviors.
4. Microscopic characterization
In order to investigate the reasons and mechanisms of Ic degradation in monotonic axial tensile testing and fatigue testing, samples exhibiting the best and worst properties from the axial tensile and fatigue experiments were picked for microstructural analysis. Microscopic analysis samples were prepared in the voltage monitoring section. Each section has one sample for cross-sectional view, and another for longitudinal cross-sectional view in length of 20 mm. The defects present in these samples, including fabrication imperfections and damage caused by mechanical loading, were thoroughly examined using metallographic microscopy and scanning electron microscopy (SEM).
Figs. 9 and 10 present the representative cross-section and longitudinal section of samples after axial tensile and fatigue tests, as observed through metallographic microscope analysis. In sample#1, the gaps, voids, and cracks are relatively small. However, in sample#2 and sample#4, these defects are more abundant and larger in size, while sample#6 exhibits moderate levels of such defects. It is worth noting that these lower-density and small defects resulting from fabrication imperfections do not significantly impact the electrical performance during the axial tensile and fatigue tests.
Fig. 9. Microscopic cross section of (a) sample#1, (b) sample#2, (c) sample#4 and (d) sample#6. |
Fig. 10. Microscopic longitudinal section of (a) sample#1, (b) sample#2, (c) sample#4 and (d) sample#6. |
On the other hand, sample#2 and sample#4 exhibits numerous large cavities and voids between the copper former and HTS tapes along its axial length. This observation suggests that sample#2 experienced an imbalance or nonuniform axial tensile stress during the experiment, which is a crucial factor contributing to the Ic degradation at a relatively low tensile stress level. In contrast, despite the existence of numerous large cavities and voids that resemble those observed in sample#2, sample#4 does not exhibit a reduction in its critical current carrying capacity when subjected to lower-level fatigue cyclic loads. This indicates that these defects come from the manufacturing process and have no significant impact on the performance of the strand under low stress conditions. These cavities, voids, and certain cracks primarily originated during the manufacturing process due to imperfect filling and forming techniques. In order to mitigate the effects of these defects on the axial tensile and fatigue strength of the round strand, a refined approach to the manufacturing processes is suggested. The vacuum impregnation method can be employed to remove air from the interior of the copper tube, consequently minimizing the presence of voids and gaps. Alternatively, additional techniques for filling can be explored, such as the utilization of fine copper wire as a filling material.
To gain further insights, SEM analysis was conducted on YBCO layers within the microscopic range of 5 to 100 μm. Figs. 11 and 12 display the representative cross-section and longitudinal section of the tensile and fatigue tested samples. The results revealed only a small amount of delamination and voids in the cross-sections of both monotonic and cyclic loading tested samples, as depicted in Fig. 11(a) and (c). Similar defects were also observed in the longitudinal sections as shown in Fig. 12(c). These few delamination and voids mainly result from thermal stress and the solder filling process during the molding process. However, these preexisting delamination and voids from the manufacturing process were significantly exacerbated by the fatigue test in sample#4 and sample#6.
Fig. 11. Microscopic cross section of (a) tensile sample#1, (b) tensile sample#2, (c) fatigue sample#4, and (d) fatigue sample#6. |
Fig. 12. Microscopic longitudinal section of (a) tensile sample#1, (b) tensile sample#2, (c) fatigue sample#4, and (d) fatigue sample#6. |
According to the observation of microcracks in tensile and fatigue samples, it was found that the REBCO superconducting layer of sample#1 and sample#2 had a high density and number of microscopic cracks, followed by sample#6, and sample#4 had the least number of cracks, as shown in Fig. 11(b), (d) and Fig. 12(a), (b) and (d). Meanwhile, it was observed that the number and density of microscopic cracks among different samples are related to the extent of Ic degradation in the sample. The greater the number of cracks within this layer, the more pronounced the decline in performance. These findings provide a plausible explanation for the degradation of Ic.
Additionally, according to the observation of SEM images, the electro-mechanical strength of a strand was found to depend primarily on the fracture behavior in the REBCO layer. It is worth noting that these microcracks tend to propagate primarily along the width of the tape, regardless of whether the loading is monotonic or cyclic. The main microcrack direction of the round cable is similar to the microcrack observed in the fatigue test sample of REBCO tape mentioned in Refs. [27], [28]. The main reason for this phenomenon is that HTS tape is a layered structure composed of multiple materials, and the REBCO layer is a brittle ceramic material. During the axial tensile or cyclic process, the main Hastelloy substrate layer and Ag layer experienced a larger axial stress-strain, which predominantly influenced the mechanical strength of the REBCO layer [27], [28], [29], [30]. As a result, the REBCO layer is more likely to develop microcracks in the width direction of the tape. On the other hand, defects inside the copper tubes, solder, and tapes of the round strand will be further enlarged, and the Ic will be reduced under the influence of axial tension [30]. Furthermore, the microcracks propagate along the width direction of the REBCO layer, which increases the internal resistance of the tape, posing a significant obstacle to the current flowing in the longitudinal direction of the REBCO layer. When the current is too large, it will be diverted to other layers or structural materials, resulting in a decrease in the overall critical current carrying capacity of the round strand.
5. Conclusions
In this paper, a new HTS round strand based on the technical path of TSTC for the high-field magnet system of future fusion reactors was reported. The electromechanical behavior of the round strand was investigated under monotonic axial tensile and axial cyclic load conditions in liquid nitrogen.
(1) The effect of axial tensile stress-strain on the Ic of round strands was measured at 77 K and self-field. The Ic of the tensile-testing samples remained relatively stable until the stress-strain limit was reached. Beyond this limit, the critical current experienced a sharp degradation with an increasing axial tensile load. In the tensile testing experiment, the round strand exhibited good mechanical strength, with an average tensile strength of 344 MPa and a tensile strain of 0.47%. The maximum tensile stress and strain are as high as 382 MPa and 0.53%, respectively. However, sample#2 only achieved a tensile stress of approximately 300 MPa and a strain of 0.4%.
(2) In the case of axial fatigue testing, the round strand demonstrated good tolerance to various tensile loads. For 100,000 cycles, no significant Ic degradation was observed when testing within the range of 20-200 MPa. The Ic of the sample tested at an average tensile stress of 340 MPa also showed no significant reduction up to 36,000 cycles before fracture occurred. However, when the peak stress approached the maximum critical tensile stress of 380 MPa, rapid Ic degradation occurred after 16,000 cycles.
(3) The microscopic studies revealed that the growth of the preexisting cracks and voids in the transverse direction of the tapes is the primary cause of Ic degradation, and the number and density of microcracks within the REBCO layer are directly related to the extent of Ic decay. Besides, the microcracks tend to propagate predominantly along the width of the tape, irrespective of whether the loading is monotonic or cyclic.
To further enhance the mechanical strength of the round strand, it is recommended to refine the manufacturing processes. The vacuum impregnation method can be used to eliminate air from the interior of the copper tube, thereby reducing the occurrence of cavities, voids, and gaps. Meanwhile, the solder filling and forming process should be optimized to reduce delamination and voids. This can be achieved by implementing various techniques, such as utilizing different heat treatment methods, adjusting the cooling rate, or incorporating stress relief features. Furthermore, to enhance the mechanical strength of the strand, one can consider using a high-strength stainless steel former or filling it with ultrafine copper wire. Finally, more experimental studies are needed to gain a better understanding of the tensile and fatigue characteristics of the strands.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
This work has been supported by the Southwestern Institute of Physics (SWIP) under project number 202101XWCXRZ001 and 2021XWCXRZ002.
