Sie sind auf Seite 1von 4

3140 IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 17, NO.

2, JUNE 2007

AC Loss Characteristics of a Cylindrical High


Temperature Superconductor
Y. H. Ma, Z. Y. Li, K. Ryu, S. Choi, K. B. Park, I. S. Oh, H. J. Song, and G. Cha

Abstract—A cylindrical high temperature superconductor


(HTS) is often used in power applications such as power transmis-
sion cables and matrix fault current limiters. In commercialization
of these apparatuses, AC loss is a critical factor but not eluci-
dated completely because of complexities in its measurement, e.g.
non-uniform current distribution and phase difference between
currents flowing in an individual HTS tape. We have prepared
two cylindrical conductors composed of a Bi-2223 tape with
different critical current density and one hollow conductor made
of Bi-2212 bulk material. In this paper, the AC loss characteristics
of the conductors have been experimentally investigated and
numerically analysed. The results show that the measured losses
for all the conductors are not dependent on both arrangements
and contact positions of a voltage lead. This implies that most
Fig. 1. Cross-sectional view of the cylindrical conductors. a) Polygon model.
of loss flux is only in the conductors. The loss for the Bi-2223 b) Cylinder model.
conductor with high critical current density is in good agreement
with the numerically calculated one based on the polygon model,
whereas the loss measured for the Bi-2223 conductor with low
critical current density coincides with the calculated loss from the tapes due to degradation during handling, the non-uniform cur-
monoblock model. rent distribution between tapes, the tape’s arrangement, and the
Index Terms—AC loss, Bi-2212 bulk material, Bi-2223 tape, error-voltage picked up into a voltage lead by the out-of-phase
cylindrical high temperature superconductor. AC currents of adjacent tapes [2], [3]. In addition, transport AC
loss of the HTS conductor becomes a function of not only con-
ductor’s configuration, critical current density and n-value but
I. INTRODUCTION
also current’s amplitude and frequency [4]. Accordingly, it is
difficult to elucidate AC loss behavior of the conductor accu-
YLINDRICAL high temperature superconductor (HTS) is rately.
C frequently used in power devices such as power transmis-
sion cables and fault current limiters. Many studies on AC loss
In this work, we prepared the kA-class cylindrical conduc-
tors made of an HTS with different critical current density and
of the cylindrical HTS conductor have been done and hence n-value such as a Bi-2223 tape and Bi-2212 bulk material, and
there has been considerable progress [1]. However, AC loss investigated their AC loss characteristics experimentally. On the
behavior of the conductor composed of several HTS tapes in basis of the measured DC voltage-current characteristics of the
particular has not been established completely because of its HTS, we also analysed their losses numerically by taking ac-
complexities; for example, the uneven critical current of the count of actual geometry of the conductors. Here, we report the
experimental results and discuss them in detail.

Manuscript received August 29, 2006. This work was supported by a grant
from Center for Applied Superconductivity Technology of the 21st Century
II. NUMERICAL MODEL
Frontier R&D Program funded by the Ministry of Science and Technology, Re-
public of Korea.
The cross-sectional view of the cylindrical conductors for nu-
Y. H. Ma, Z. Y. Li, and K. Ryu are with the Department of Electrical merical analysis is depicted in Fig. 1. The governing equation
Engineering, Chonnam National University, Kwangju, 500-757 Korea (e-mail: in two dimensions is represented by
ramon800612@hanmail.net; y82610@hanmail.net; kwryu@chonnam.ac.kr).
S. Choi is with Magnet Development Group, Superconducting Materials
Center, National Institute for Materials Science, Tsukuba, Ibaraki 305-0003, (1)
Japan (e-mail: CHOI.Seyong@nims.go.jp).
K. B. Park and I. S. Oh are with Technology R&D Center, LS Industrial Sys-
tems, Chongju, 361-720 Korea (e-mail: kbparkd@lsis.biz; isoh@lsis.biz). where is the permeability of free space, is the z-compo-
H. J. Song is with the School of Nano Engineering, Inje University, Gimhae,
621-749 Korea (e-mail: hjsong@inje.ac.kr).
nent of magnetic vector potential, is the conductivity of
G. Cha is with the Department of Electrical Engineering, Soonchun- superconductor, and is the electric scalar potential, respec-
hyang University, Shinchang, Asan, Choongnam 336-746, Korea (e-mail: tively. The details about numerical computations of (1) are de-
gsoocha@sch.ac.kr). scribed in [5]. AC loss per unit length can be calculated by inte-
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. grating local energy dissipation over cross-section of HTS core
Digital Object Identifier 10.1109/TASC.2007.898326 as shown in Fig. 1.
1051-8223/$25.00 © 2007 IEEE

Authorized licensed use limited to: UNIVERSIDADE DE SAO PAULO. Downloaded on March 19,2010 at 12:32:39 EDT from IEEE Xplore. Restrictions apply.
MA et al.: AC LOSS CHARACTERISTICS OF A CYLINDRICAL HIGH TEMPERATURE SUPERCONDUCTOR 3141

TABLE I
SPECIFICATIONS OF THE CYLINDRICAL CONDUCTORS

Fig. 2. Photographs of the cylindrical conductors. (a) Conductor consisting of


the Bi-2223 tapes. (b) Bi-2212 hollow conductor.

fabricated two Bi-2223 conductors with different critical current


densities. The other is a hollow conductor made of Bi-2212 bulk
material, which is often used as an element in a superconducting
fault current limiter (SFCL). The Bi-2212 hollow conductor has
very low n-value in comparison with the Bi-2223 conductor. Its
photo is also shown in Fig. 2(b). The main parameters for two
kinds of the cylindrical conductors are summarized in Table I.
The parameters used in this numerical analysis are listed in Sinusoidal transport current was provided to the cylindrical
Table I. Even if many Bi-2223 filaments in the conductors A conductors by means of a power supply with 1500 A and 10 V.
and B are embedded in normal matrix, core including all the The current of each isolated tape in the Bi-2223 conductor was
filaments can be treated as a single elliptical superconductor manipulated uniformly by changing the impedance of an ex-
because of electromagnetic coupling among the filaments. The ternal reactor. Moreover, we adjusted the phases of currents in
normal matrix can be neglected in loss evaluation because AC the tapes using the same reactor and made them coincide with
loss in the matrix is minor portion of the total loss in the Bi-2223 the phase of the total current in the Bi-2223 conductor to mini-
tapes. The conductivity of superconductor is determined from mize phase errors when the inductive voltage induced in voltage
the voltage-current characteristics, which are investigated ex- leads is compensated by a cancel coil [2]. AC loss was evalu-
perimentally in this work. It is given by ated by integrating the product of transport current and voltage
between voltage taps during one period. The distances between
voltage taps are 500 mm for the Bi-2223 conductor and 160 mm
(2) for the Bi-2212 hollow conductor, respectively. All experiments
in this work were made in liquid nitrogen.
where is the conductivity, the z-component of elec-
tric field, the critical electric field of , the IV. RESULTS AND DISCUSSIONS
critical current density, and the n-value of the superconductor.
Although the critical current density and the n-value usually de- A. Cylindrical Conductor With High Critical Current Density
pend on the local magnetic field [5], [6], their field dependence We first investigated the influence of voltage lead’s arrange-
is not considered in this analysis because of minor contribution ment and frequency on AC loss in the conductor A. Fig. 3(a)
to the AC loss except for the case of large magnetic field. The shows the AC losses measured at 100 Hz for three different ar-
n-values and critical currents measured at self-field are summa- rangements of the rectangular voltage leads positioned on the
rized in Table I. tape of No. 5: their heights from the conductor surface are 0 mm,
3 mm, and 6 mm, respectively as in Fig. 2(a). Fig. 3(b) indicates
the frequency dependence of AC loss. It can be seen from Fig. 3
III. EXPERIMENT
that the measured AC losses are independent of both voltage
Two kinds of the cylindrical conductors were prepared for lead’s arrangement and frequency. This means that most of loss
this work. One is a single-layer conductor composed of several flux exists inside the cylindrical conductor and measured AC
Bi-2223 tapes placed on a circular plastic former, which are ar- losses are hysteretic. On the basis of the result in Fig. 3(a), we
ranged parallel to the conductor axis. We insulated the tapes in chose 0 mm-lead to evaluate AC loss of the cylindrical conduc-
the conductor electrically, in order to exclude complicated phe- tors correctly in all of the following measurements.
nomena due to current redistributions on the way of the con- Fig. 4 shows the AC losses of the conductor A measured for
ductor. To minimize the effect of gap length on AC loss char- ten voltage leads contacted to individual tape as shown in Fig. 1
acteristic of the Bi-2223 conductor, we made the gap between when current in each tape is uniform. For comparison with the
neighboring tapes to become nearly zero [3]. Its photograph is measured data, we plotted the two calculated losses for an av-
illustrated in Fig. 2(a). In order to investigate the influence of erage critical current of 1270 A in the conductor A: the dashed
tape’s critical current density on AC loss of the conductor, we line is based on numerical analysis of (1) with a polygon model

Authorized licensed use limited to: UNIVERSIDADE DE SAO PAULO. Downloaded on March 19,2010 at 12:32:39 EDT from IEEE Xplore. Restrictions apply.
3142 IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 17, NO. 2, JUNE 2007

Fig. 4. AC loss vs. peak transport current for different contact positions in the
conductor A with high critical current density.

Fig. 3. (a) Voltage lead’s arrangement and (b) frequency dependence of the AC
losses in the conductor A with high critical current density.

as in Fig. 1(a) and the solid line based on monoblock model with
a cylindrical cross section:

Fig. 5. AC loss vs. peak transport current for different contact positions in the
(3) conductor B with low critical current density.

where is the critical current of the hollow conductor, and


are inner and outer radii, and is the peak transport current, 0 mm-lead at 100 Hz are indicated in Fig. 5. The calculated ones
respectively. from both monoblock model of (3) and numerical analysis of (1)
Despite different contact positions of the voltage leads, the with the polygon model are also plotted in Fig. 5 together. As
AC losses measured in the conductor A are identical. This im- shown in Fig. 5, the measured AC losses are almost identical re-
plies that our measured loss data are reliable. Furthermore, it can gardless of different contact positions as in Fig. 4. It is seen that
be seen from Fig. 4 that the measured losses are larger than the the AC loss characteristic of the conductor B is different from
ones calculated from monoblock model but in good agreement that of the conductor A: the measured losses of the conductor
with the numerically calculated losses based on the polygon B are smaller than the ones calculated from numerical analysis
model. This suggests that the electromagnetic behavior of the but coincide with the calculated losses from monoblock model.
conductor A, consisting of Bi-2223 tapes with high critical cur- This means that the electromagnetic behavior of the conductor
rent density, is similar to that in the polygon model as shown in B with low critical current density is similar to that of the cylin-
Fig. 1(a). drical model as shown in Fig. 1(b), although geometric arrange-
ment of the Bi-2223 tapes are analogous to the polygon model
B. Cylindrical Conductor With Low Critical Current Density as in Fig. 1(a).
To investigate the effect of critical current density on AC loss It is useful to compare AC loss of the conductor B consisting
characteristics of the cylindrical conductors, we measured the of several Bi-2223 tapes with that of an ideal hollow conductor,
AC losses of the conductor B, which consisted of the Bi-2223 conductor C, with similar critical current density: the critical
tapes with low critical current density, for uniform current distri- current densities are 27.5 in the conductor B and
bution. The losses measured for three contact positions with the 12.7 in the conductor C, respectively. We investigated

Authorized licensed use limited to: UNIVERSIDADE DE SAO PAULO. Downloaded on March 19,2010 at 12:32:39 EDT from IEEE Xplore. Restrictions apply.
MA et al.: AC LOSS CHARACTERISTICS OF A CYLINDRICAL HIGH TEMPERATURE SUPERCONDUCTOR 3143

Fig. 7. Frequency dependence of the AC loss in the conductor C with low


n-value.

Fig. 6. AC loss vs. peak transport current for different contact positions in the
conductor C with low n-value. V. CONCLUSION
In this work, we prepared the kA-class cylindrical conduc-
tors made of an HTS with different critical current density and
the AC losses of the conductor C made of Bi-2212 bulk for n-value such as a Bi-2223 tape and Bi-2212 bulk material, and
various voltage leads placed around an outer circumference at investigated their AC loss characteristics.
180 Hz as shown in Fig. 1(b). The results are shown in Fig. 6. The measured losses of the conductor A, which consists of
The calculated losses from both monoblock model of (3) and a Bi-2223 tape with high critical current density, are in good
numerical analysis of (1) with the cylinder model are also agreement with the numerically calculated ones based on the
depicted in Fig. 6. It can be seen from Fig. 6 that the measured polygon model. However, the AC losses of the conductor B with
AC losses are the same regardless of contact positions of the low critical current density coincide with the calculated losses
voltage leads as shown in Figs. 4 and 5. However, the AC loss from the monoblock model. Moreover, the measured losses of
characteristic of the conductor C is significantly different from the Bi-2212 hollow conductor disagree with the calculated ones
those of the conductors A and B: the measured losses of the from both the monoblock model and the numerical analysis with
hollow conductor disagree with the calculated ones from the the cylindrical model. This is due to low n-value of 8.5 in the
monoblock model. Moreover, they are smaller than the losses Bi-2212 hollow conductor compared with the Bi-2223 conduc-
from numerical analysis of (1) with the cylinder model as in tors of A and B.
Fig. 1(b). This is due to the frequency dependence of transport
AC loss in a high temperature superconductor, which is not REFERENCES
usually taken into account in numerical analysis: transport [1] R. C. Duckworth, M. J. Gouge, J. Caughman, J. W. Lue, J. A. Demko,
loss decreases with increasing frequency [4]. Especially, this J. Tolbert, C. L. H. Thieme, and D. T. Verebelyi, “On the effect of
NiW on the inductance and AC loss of HTS cables,” IEEE Trans. Appl.
frequency dependence enhances with decreasing the n-value of Supercond., vol. 15, no. 2, pp. 1578–1582, 2005.
the superconductor. In our conductor C, the measured n-value [2] K. Ryu, S. D. Hwang, S. W. Kim, and S. Hahn, “Influence of tape’s
is 8.5 and small compared with 10.3 and 15.2 of the conductors critical currents and current distributions on ac loss measurement in a
multi-tape conductor,” IEEE Trans. Appl. Supercond., vol. 15, no. 2,
A and B as listed in Table I. pp. 1611–1614, 2005.
To confirm whether or not the measured AC losses illustrated [3] R. Inada, Y. Nakamura, A. Oota, T. Fukunaga, and P. X. Zhuang, “Ef-
in Fig. 6 are true, we investigated the AC losses for several fect of wire arrangements on AC transport losses in cylindrical conduc-
tors composed of Ag-sheathed Bi2223 tapes,” Physica C, vol. 412–414,
frequencies from 90 Hz to 180 Hz. The results are indicated pp. 1148–1153, 2004.
in Fig. 7. We could not get reliable loss data for frequencies [4] D.-X. Chen and C. Gu, “Alternating current loss in a cylinder with
around power frequency of 60 Hz because of the short distance power-law current-voltage characteristic,” Appl. Phys. Lett, vol. 86, pp.
2525041–2525043, 2005.
of 160 mm between voltage taps and the large noise from our [5] S. Choi, W. Nah, J. H. Kim, J. Joo, Y. H. Jo, and K. Ryu, “AC transport
power supply. We can see from Fig. 7 that even if test frequen- current loss of horizontally attached Bi-2223/Ag tapes,” IEEE Trans.
cies are a little high, the measured losses of the conductor C Appl. Supercond., vol. 14, no. 2, pp. 1894–1897, 2004.
[6] S. Choi, W. Nah, J. H. Kim, J. Joo, Y. H. Jo, and K. Ryu, “AC transport
agree with each other. This implies that our measured AC losses current loss of vertically stacked Bi-2223/Ag conductors,” Physica C,
are very reliable. vol. 412–414, pp. 1158–1162, 2004.

Authorized licensed use limited to: UNIVERSIDADE DE SAO PAULO. Downloaded on March 19,2010 at 12:32:39 EDT from IEEE Xplore. Restrictions apply.

Das könnte Ihnen auch gefallen