IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 23, NO. 3, JUNE 2013 4201206
Optimization of Interstrand Coupling Loss and
Transverse Load Degradation in ITER Nb
3
Sn CICCs
A. Nijhuis, G. Rolando, C. Zhou, E. P. A. van Lanen, J. van Nugteren, R. P. Pompe van Meerdervoort,
H. J. G. Krooshoop, W. A. J. Wessel, A. Devred, A. Vostner, and I. Pong
Abstract—For the ITER Central Solenoid (CS), with Nb
3
Sn
CICCs that operate under fast ramping conditions, the selection
of the twist pitch lengths can have a significant impact on the
performance. The critical current and temperature margin are
influenced by the thermal contraction of the composite materials,
the transverse electromagnetic forces, and coupling currents. The
numerical cable model JackPot-ACDC is developed to calculate
the interstrand coupling loss for any time-dependent current and
magnet field for all strand trajectories in a CICC. It was a priori
predicted that the amount of coupling loss and critical current
degradation is subject to interference due to different subcable
twist pitches. Here test results are discussed of the ITER CS
conductor sample, manufactured according to the proposed de-
sign, optimizing the transverse load degradation, the temperature
margin, and the coupling loss.
Index Terms—CICC, CJackPot, coupling loss, ITER, S coil.
I. I NTRODUCTION
T
HE ITER CS conductor offers an exceptional challenge in
sustaining a large number of electromagnetic cycles and at
the same time restricting AC losses. The first tests on prototype
CS samples have shown that the design of such conductor is
not obvious since these exhibited significant degradation of
the transport properties in terms of current sharing temperature
(T
cs
) [1]. Already in 2006 it was predicted that the mechanical
performance of ITER TF CICCs could be improved by chang-
ing the cable pattern [2], which was confirmed experimentally
in 2007 and 2008 [3], [4]. The development of the JackPot-
ACDC cable model allowed to combine this concept with mini-
mization of the interstrand coupling losses for the CS conductor
[5]. Besides the input parameters of strand, cooling channel,
cable diameter and twist pitch sequence, JackPot requires
the strand strain scaling and interstrand resistivity parameters,
which are obtained from direct measurements. In June 2011
a JackPot prediction was presented recommending a pseudo
Long Twist Pitch cable pattern, limiting the scope for strand
Manuscript received October 9, 2012; accepted December 18, 2012. Date
of publication January 24, 2013; date of current version February 14, 2013.
The views and opinions expressed herein do not necessarily reflect those of the
ITER Organization.
A. Nijhuis, G. Rolando, C. Zhou, E. P. A. van Lanen, J. van Nugteren,
R. P. Pompe van Meerdervoort, H. J. G. Krooshoop, and W. A. J. Wessel are
with the University of Twente, 7500 AE, The Netherlands (e-mail: a.nijhuis@
utwente.nl).
A. Vostner, I. Pong, and A. Devred are with ITER Organization, 13115 St.
Paul-lez-Durance, France.
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TASC.2012.2235894
TABLE I
CSIO 1 & 2 CABLE DESIGN PARAMETERS
bending and allowing axial flexibility against thermal cycles
in combination with low coupling loss. The JackPot unprece-
dented prediction that increasing (lower stage) cabling pitches
and meanwhile reducing the AC coupling loss by selecting a
“close-to-one” ratio for the twist pitch sequence was adopted
by ITER for further investigation. Four conductor samples were
manufactured using different cable patterns and tested in Sultan
during spring 2012. The experimental results of the T
cs
and
AC loss measurements are presented and compared for all four
CSIO options. The data are directly taken from the Sultan server
without post processing. The results are compared, taking into
account cyclic degradation of T
cs
, influence of coupling loss
on the temperature margin during operation and the surplus
of the strand current, driving the strands into saturation due
to the severe time varying magnet field and current conditions
during the regular ITER 15 MA plasma scenario. The results
are compared with the predictions presented in 2006 and 2011
[2], [5] and with recent JackPot computations.
II. CSIO CONDUCTOR SAMPLES
In Table I the differences between the selected CSIO cable
patterns are listed in Table I [5], [6]. The CSIO1 and CSIO2
Sultan samples rely on US IT strands from the same supplier.
The cross section of both strand layouts, with Cu:nonCu 1.0
and 1.5, are shown in Fig. 1, further details can be found in
[1]. The tests of the witness samples from the CSIO1 Sultan
sample heat treatment pointed out that the I
c
of the strand with
Cu : nonCu =1.5 results into a substantially higher overall I
c
per triplet (31%), see Table II [1] and [7]. The specific CSIO1
& 2 strands are not tested on I
c
(B,T,ε) for the purpose of
strain scaling. The strain scaling parameters are taken from the
1051-8223/$31.00 © 2013 IEEE