Axial strain characterization of the Nb3Sn strand used for China's TF conductor

Fusion Engineering and Design 86 (2011) 1–4

Contents lists available at ScienceDirect

Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes

Axial strain characterization of the Nb3 Sn strand used for China’s TF conductor夽 Bo Liu ∗ , Yu Wu, Fang Liu, Feng Long Institute of Plasma Physics, Chinese Academy of Sciences, Hefei 230031, China

a r t i c l e

i n f o

Article history: Received 10 April 2010 Received in revised form 3 July 2010 Accepted 5 July 2010 Available online 1 August 2010 Keywords: ITER Nb3 Sn Axial strain Improved deviatoric scaling model Pacman

a b s t r a c t Oxford Superconducting Technology (OST) has developed a type of Nb3 Sn strand which can be used in the ITER TF coils. The strand (billet number is OST 10424FE) is made by an internal tin process. The critical current (Ic ) has been measured subjected to uniaxial strain in the Twente University. The strand is destined for the cable-in-conduit conductors (CICC) of the China first short conductor sample for ITER toroidal field coil. For the uniaxial strain characterization, the voltage–current characteristics were measured with an applied axial strain from −0.8% to +0.5%. The strand appears to be fully reversible in the compressive regime during the axial strain testing, while in the tensile regime, the behavior is already irreversibly degraded when reaching the maximum in the critical current versus strain characteristic. The parameters for the improved deviatoric strain description are derived from the Ic data, giving the accuracy of the scaling with a standard deviation of 1.5 A, which is by far within the expected deviation for the large scale strand production of such a high Ic strand. Analysis of the relation between Ic , n index and the axial strain is reported. © 2010 Elsevier B.V. All rights reserved.

1. Introduction The magnet system for international thermonuclear experimental reactor (ITER) consists of 18 toroidal field (TF) coils, a central solenoid (CS), six poloidal field (PF) coils and 18 correction coils (CCs) (Fig. 1). All coils are superconducting. The TF coils generate the field to confine charged particles in the plasma, the CS coils provide the inductive flux to ramp up plasma current and contribute to plasma shaping, the PF coils provide the position equilibrium of plasma current (i.e. the fields to confine the plasma pressure) and the plasma vertical stability. The CCs allow correction of error field harmonics (up to 3 harmonics in toroidal and poloidal directions) due to position errors as well as from busbars and feeders [1,2]. The ITER TF coils require a higher-performance Nb3 Sn strand than that developed during the ITER model coil project, in order to recover the loss in conductor critical current caused by transverse electromagnetic forces acting inside the conductor. The minimum acceptable value for the critical current is 190 A made by a bronze process and 230 A internal tin process strand under a strain of 0.25%, a temperature of 4.2 K and a magnetic field of 12 T [3]. During operation the Nb3 Sn strands in the ITER TF coils will undergo strain, so they have to be characterized under relevant strain conditions to be able to predict their performances in the TF coils [4]. This paper

presents the results of strand characterization by using Pacman apparatus carried out at Twente University [5]. 2. Specification of the strand The high current density strand manufactured by Oxford Superconducting Technology (OST) follows the internal tin method and the billet number is OST 10424FE. The specification is given in Table 1. A cross-section of the strand before heat treatment (HT) and the filament bundle is shown in Fig. 2. Before the heat treatment, the chrome layer on the wire surface of all samples, except for some of the axial stress–strain samples, was removed by etching with a 37% solution of hydrochloric acid. The samples for all Ic measurements were heat treated in vacuum following the HT schedule used: 38 h at 210 ◦ C; 25 h at 340 ◦ C; 25 h at 450 ◦ C; 100 h at 575 ◦ C; 100 h at 650 ◦ C; with a ramp rate of 10 ◦ C/h The temperature was within 2 ◦ C deviation from the specified value. 3. Uniaxial strain 3.1. Experimental arrangement for Ic versus uniaxial strain testing in Twente University

夽 This work was supported by the Chinese ITER Special Program under Grant 2008CB717905. ∗ Corresponding author. Tel.: +86 15955108061. E-mail address: [email protected] (B. Liu). 0920-3796/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.fusengdes.2010.07.001

After the HT the sample was transferred carefully to the Pacman spring and was fixed with Sn–5 wt% Ag solder at about 500 K. Fig. 3 shows a view of the Pacman [5] spring with a wire sample soldered.

2

B. Liu et al. / Fusion Engineering and Design 86 (2011) 1–4

Fig. 1. The ITER coils. Fig. 3. Overall view of the Pacman spring with the sample soldered. Table 1 The structure parameter of Nb3 Sn strand. Parameter

Numeric values

Diameter Cr plating Twist pitch Cu/non-Cu ratio Twist direction Ic (4.2 K, 12 T, 10 ␮V/m, A) n (4.2 K, 12 T)

0.82 mm ± 3 ␮m 1.0 ␮m 17 mm 0.933 Right 271.9 27.6

The Pacman spring is a circular bending beam made of Ti alloy with a T-shaped cross-section. The strain on the outer surface of the beam is controlled by the torque applied to the spring and it is measured by the strain gauges attached to the spring. The rotation induced by the motor–worm gear combination at room temperature is transferred to the low-temperature region through a set of concentric tubes, coupled mechanically with the two revolving halves of the Pacman spring support. The temperature variations are applied by placing a polyimide insulator cup over the Pacman, thus creating a helium gas vol-

Fig. 2. The cross-sectional area of Nb3 Sn strand before heat treatment.

ume. Cooling and heating paths on the spring are symmetrical. Two thermometers are used so that the temperature gradient over the measured sample length can be minimized. The temperature can be balanced well within ±20 mK. The set-up is inserted into the bore of a superconducting solenoid, with the magnetic field perpendicular to the sample. 3.2. Experimental results Ic versus uniaxial strain testing In total about 60 VI curves as a function of strain (−0.8% to +0.5% applied), temperature (4.2, 6, 8, 10, and 12 K) and magnetic field (6, 8, 10, 12 and 14 T) were measured per sample during the experimental phase. The maximum critical current in the experiment exceeds 500 A. The critical current criteria is Ec = 10 ␮V/m [6]. The results of the critical current measurements as a function of strain, temperature and magnetic field are summarized in Figs. 4–6. 3.3. n-value The n-value is commonly used to characterize the sharpness of field-current transition in superconductors, which is related to the performance degradation and the filaments uniformity. Like Ic of strands, the n-value varies with field, temperature and strain for superconducting wires. Fig. 7 shows the n-value plotted as a function of applied strain for the strands of OST.

Fig. 4. Critical current as a function of applied strain.

B. Liu et al. / Fusion Engineering and Design 86 (2011) 1–4

3

Table 2 The parameter of strand scaling law. Parameter

Numeric values

Ca1 Ca2 ε0,a εm Bc2m (0) Tcm C1

44.32 0 0.304% −0.048% 32.86 16.08 22,300

The improved deviatoric scaling model for the critical current in the Nb3 Sn composites can be written as C1 2 × s(ε) × (1 − t 1.52 ) × (1 − t 2 ) × h0.5 × (1 − h) Ic (H, T, ε) ∼ = 0 H (1) Fig. 5. Critical current as a function of temperature at 6, 8, 10, 12 and 14 T, at zero applied strain.

3.4. An improved deviatoric scaling model We adopt the University of Twente’s improved deviatoric scaling model to confirm the relation about Ic , axial strain, magnetic field and temperature [7].

t=

T , Tc∗ (ε)

h=

H ∗ (T, ε) Hc2

∗ Hc2 (T, ε) ∼ = H×c2m (0) × s(ε) × (1 − t 1.52 ) ∗ Tc∗ (ε) = Tcm × s(ε)1/3

 s(εaxial ) =



Ca1

× (1 − Ca1 × ε0,a )

εshift =

C

a2

2

2

(εshift ) + (ε0,a ) − −1



2

(εaxial − εshift ) + (ε0,a )

2



 − Ca2 × εaxial

+1

× ε0,a

2 − C2 Ca1 a2

εaxial = εapplied + εm

Fig. 6. Critical current as a function of magnet field at 4.2, 6, 8, 10, and 12 K, at zero applied strain.

Here 0 is the magnetic permeability of vacuum, H is the magnetic ∗ is the inhomogeneity averaged, effective, critical magfield, Hc2 ∗ netic field, Hc2m is the inhomogeneity averaged maximum critical magnetic field (at εaxial = 0), T is the temperature, Tc∗ is the inhomo∗ is the geneity averaged critical temperature defined at H = 0, Tcm inhomogeneity averaged maximum critical temperature (at H = 0), ε is the strain, εapplied is the applied axial strain, εm is the (thermal) pre-compression strain, εshift is the axial difference between a three-dimensional deviatoric strain minimum and the position of the maximum in axial strain sensitivity results, ε0,a is the remaining strain component, C1 is the constant, Ca1 is the second invariant axial strain sensitivity, Ca2 is the third invariantaxial strain sensitivity. ∗ ∗ and four (0) and Tcm Three superconducting parameters C1 , Hc2m deformation-related parameters Ca1 , Ca2 , ε0,a and εm have to be determined experimentally. This can be done by applying a least square fit of (1) to the entire dataset. The results of the best found fit to the Ic data obtained below the irreversible strain limit (i.e. +0.2%) are presented in Fig. 4, and the scaling parameters are listed in Table 2. 4. Discussions

Fig. 7. n index as a function of applied strain.

The OST strand appears to be quite sensitive to the tensile strain. Due to the extreme strain sensitivity in tensile applied strain direction, particular care must be paid in determining the peak at zero intrinsic strain. When the applied strain is larger than +0.2% in the tensile direction, the Ic and the n-value are irreversibly degradated for the OST strand. This may due to the formation of breakages and voids form which would degrade the transport properties at a high stress state.

4

B. Liu et al. / Fusion Engineering and Design 86 (2011) 1–4

The experimental data below +0.2% applied tensile strain were used to determine the scaling parameters for the improved deviatoric scaling mode as given in Table 2. The result of the best fit for the critical current shows a good match with the measured data. The standard deviation of the error between improved deviatoric scaling mode model parameterization and experimental results amounts to 1.5 A. As Taylor mentioned [8] the n(ε) should reach a maximum at εm similarly to Ic (ε), but the peak of n-value in Fig. 7 has a deviation in comparison with that in the curve of Ic (ε) versus applied strain. The peak value of n(ε) is about 29.6 compared with the applied zero-strain value 28.2. To the authors’ knowledge, there is not a commonly agreed explanation for the phenomenon. A speculation is that, under the compressive strain, the cracked, separated pieces of the brittle Nb3 Sn filaments get better contact again after being compressed, leading to a partial recovery of the critical current. Another speculation is a possible contribution from strain release in the radial direction of the filaments due to the plastic deformation of the matrix material. According to the ITER specification the n-value of strands should exceed 20. In the range −0.6% < ε < 0.15%, the n-value of OST strands actually satisfies the ITER request [9]. 5. Conclusions The OST 10424FE Nb3 Sn for the ITER TF coils has been characterized in detail using Pacman apparatus. The internal tin strands show susceptibility to applied strain. Scaling relations which can express these strand performance were obtained. For applied strain in the

range of interest, say −0.6% < ε < 0.15%, the n-value of OST strands satisfy the ITER request. Acknowledgement The authors would like to thank A. Nijhuis (University of Twente) who offered the experimental data for this paper. References [1] N. Mitchell, D. Bessette, R. Gallix, The ITER magnet system, IEEE Trans. Appl. Supercond. 18 (2) (2008) 435–440. [2] N. Mitchell, P. Bauer, D. Bessette, Status of the ITER magnets, Fusion Eng. Des. 84 (2009) 113–121. [3] Y. Nunoya, T. Isono, N. Koizumi, Characterization of ITER Nb3 Sn strands under strain-applied conditions, IEEE Trans. Appl. Supercond. 18 (2) (2008) 1055–1058. [4] A. Nijhuis1, Y. Ilyin, W. Abbas, Axial and transverse stress–strain characterization of the EU dipole high current density Nb3 Sn strand, Supercond. Sci. Technol. 21 (2008), 065001 (10pp.). [5] A. Godeke, M. Dhalle, A. Morelli, A new device to investigate the axial strain dependence of the critical current density in superconductors, Rev. Sci. Instrum. 75 (2004) 5112–5118. [6] R. Flükiger, D. Uglietti, B. Seeber, Asymmetric behavior of J(c)(epsilon) in Nb3 Sn wires and correlation with stress induced elastic tetragonal distortion, Supercond. Sci. Technol. 18 (2005) 416–423. [7] Y. Ilyin, A. Nijhuis, E. Krooshoop, Scaling law for the strain dependence of the critical current in an advanced ITER Nb3 Sn strand, Supercond. Sci. Technol. 20 (2007) 186–191. [8] D.M.J. Taylor, S.A. Keys, D.P. Hampshire, E–J characteristics and n-values of a niobium–tin superconducting wire as a function of magnetic field, temperature and strain, Physica C 372–376 (2002) 1291–1294. [9] P.X Zhang, M. Liang, X.D. Tang, Strain influence on Jc behavior of Nb3 Sn multifilamentary strands fabricated by internal tin process for ITER, Physica C 468 (2008) 1843–1846.