CALCULATION OF THE CRITICAL CURRENT IN PANCAKE-COILED

HIGH TEMPERATURE SUPERCONDUCTING TAPES

Frank Darmann
Australian Superconductors, MM Ltd, Eveleigh, NSW 1430, Australia.
e-mail: [email protected]
ABSTRACT
Progress in the development of long continuos lengths of High Temperature Superconducting
Tape (HTSC) has been rapid. In order to characterise these long length tapes, flat windings called
pancakes coils, are manufactured and the critical current of this coil is measured. The significant
magnetic fields generated by the coil are non uniform over the coil and across the width of an
individual turn. This phenomenon influences the critical current to a high degree. In order to
calculate the field free critical current, the DC magnetic field pattern of three flat pancake style
windings of HTSC Tape was calculated within each winding volume. The effect of the non-
uniform field across the tape width on the critical current of the pancake coil and on individual
turns of the coil was measured by employing current leads and voltage taps interleaved between
the turns of the coil. A model is proposed to explain the experimental observations along with a
method for correctly calculating the field free critical current from the measured critical current.
1. INTRODUCTION
The current world-wide production volume of Bi-
2223/Ag (HTSC) tape is accelerating, with a number
of companies claiming to produce high critical
current density (J
c
), long length tapes [1]. In order to
characterise these long lengths, a number of devices
have been constructed which measure the critical
current of small sections of tape in liquid nitrogen on
a spool-to-spool rig. These include both contact and
non-contact methods [2]. However, as a rapid check
of the overall critical current of a long length tape, a
transport current must be set up through the tape, and
the voltage measured across the whole length. Due to
the practical difficulties involved in handling such
long lengths of thin ribbon, the HTSC tapes must be
wound into a coil and the choices are usually limited
to either a multi-layer solenoid or pancake-type
winding. In this situation, the self-field generated by
the coil can be significant, and at 77 K the measured
critical current, (I
c
(P)), is not the field free critical
current, (I
c
(0)), which would be obtained if the tape
was measured in a straight length.
Methods have been reported in the literature on how
to correct for the coil field and obtain I
c
(0) from I
c
(P)
[3]. These methods calculate the load line for the coil
based on the greatest perpendicular component of
magnetic field found in the pancake and apply to this
the transport field performance of the tape. For a long
solenoid winding with a small annular thickness
compared to the length of the coil this type of analysis
is satisfactory because the magnetic field at each
individual turn of the winding may be shown to be
uniform across the width of the tape. In a pancake
winding however, the technique is not valid because
the magnetic field across the tape width is not
uniform, but varies from a maximum at the edges of
the pancake to a minimum in the middle, making a
suitable load line construction non-trivial.
2. MAGNETIC FIELD PATTERN OF A
PANCAKE WINDING
Figure 1 shows a schematic representation of a
pancake winding drawn in cylindrical r--z space,
along with a set of references axes relative to the tape
orientation (A-B-C). A pancake consists of a
concentric, flat, continuos winding of HTSC tape
with a suitable insulation separating each of the turns.
Typically, a number of turns, N, of the HTSC tape are
wound on an inner former of radius r
o
, where N is
usually of the order of 100 to 1000. Also included in
Figure 1 is a representation of the individual tape
turns at a particular radius. The tape width (2a) is
defined to be the physical extent of the internal
ceramic filaments across the HTSC tape, rather than
the HTSC tape width itself.
The pancake produces a cylindrical, symmetric,
magnetic field pattern within the volume of the
winding which can be represented in terms of a radial
and an axial component. The radial component of the
pancake field, B
r
, is perpendicular to the tape surface,
i.e. along the r direction, and the axial component, B
z
is parallel to the tape surface in the -z plane.
These components of the pancake field vary
throughout the winding typically as shown in Figure
2, where the magnetic field is normalised to the
transport current, I
trans
.
The radial component also varies across the width of
the tape, as shown in Figure 3, peaking at the edges,
I
z= +a
I
d
w = 2a
z= -a
r, (C)
z , (A)
I
, () , ()
zoomed area
of two tapes
z , (A)
r , (C)
r
i
2a
r
o
-0.003
-0.002
-0.001
0.000
0.001
0.002
0.003
425 445 465 485 505
R a d i a l P o s i t i o n , r ( m m )
M
a
g
n
e
t
i
c

f
i
e
l
d
.

(
T
/
A
)
Axial
Radial
-0.003
-0.0025
-0.002
-0.0015
-0.001
-0.0005
0
0.0005
0.001
0.0015
0.002
0.0025
0.003
z
M
a
g
n
e
t
i
c

F
i
e
l
d
(
T
/
A
)
-a + a
0
(centre of tape)
Radial component
of magnetic field
Axial component
of magnetic field
Figure 1 : Depiction of two tapes (lower figure) within the winding of a HTSC coil, (above)
along

Figure 2. Typical magnetic field components

across a pancake radius. The radial component
is valid only for the edge of the pancake/tape.
Radial Position, r (mm)
and falling to zero in the exact centre line of the
pancake.
The radial component of the pancake field therefore
has a plane of asymmetry about the centre of the
pancake, in the r plane. The axial component,
however, is essentially uniform across the tape width,
and forms a plane of symmetry about the centre line
of the pancake in the r plane.
In order to map the magnetic field pattern
numerically within the pancake, software was written
which divided the pancake profile into a square grid,
having sides r and z, in 2-dimensional cylindrical
space, with each grid point representing an imaginary
thin circular turn. The magnetic field pattern,
B(r,z)/I
trans
(units = TA
-1
) due to each square grid
could then be calculated exactly in terms of its axial
and radial component shown by equation (1) [4],
where f(r,z) is a known function of elliptical
integrals. The magnetic field at any grid point, (r,z)
within the square grid was then calculated by the
superposition of the contribution from each pancake
grid point excluding the one at (r,z), as shown in
equation (2). In this way, the inevitable singularity
problem associated with this technique was avoided.
(1)
(2)
where
(3)
and where B
r
= Radial field Component (Tesla/Amp)
B
z
= Axial field Component (Tesla/Amp)
In the case of a pancake winding, the tape width is
very much smaller than the coil diameter and it can
be shown that the radial component of field, |B
r
(r,z)|,
varies in a linear manner across the tape width, as
shown in Figure 3. The maximum field at any radius,
r, is denoted by B
e
(r), and occurs at the edge of the
tape. In this case, the solution to equations (1) and (2)
will then be of the form shown in equation (4)].
(4)
where B
e
(r) and B
z
(r) are field constants which
depend only on the radial position of each turn in the
pancake.
3. MAGNETIC FIELD PERFORMANCE OF
HTSC TAPES
The effect of a uniform magnetic field on the critical
current density of HTSC tapes is demonstrated by
plotting the transport critical current density, in the
presence of the magnetic field, J
c
(B), against the
magnitude of the applied magnetic field, B
o
at 77K.
These measurements are conventionally taken over
short lengths (~5 cm) using a spatially-uniform
magnetic field, applied at an angle, , in the AB and
CB planes. In addition, the normalised current
density, J
c
(B)/J
c
(0), where J
c
(0) is the critical current
density in the absence of any applied field, is used
here. Typically the base 10 logarithm of J
c
(B)/J
c
(0) is
plotted against B
o
.
At 77 K, a characteristic weak link-strong link
regime is evident, [5,6] with a severe reduction in the
critical current occurring in fields 40 mT applied in
the C direction. It has been found that the component
of the magnetic field along the C direction dominates
the critical current for angles 7
o
[5]. Empirical
and experimental equations have been proposed to fit
the observed relationship [6], however, for the
purposes of the analysis required here, a simple 4
th
order polynomial fit shown by equation (5), which is
valid up to 50 mT is sufficient.
(5)
Where

0 < B
o
< 50 mT is the component of magnetic
field along the C axis, and the k
i
values are constants.
The relatively smaller effect of the component of field
parallel to the tape surface will be ignored in all
subsequent analysis.
4. CURRENT DISTRIBUTION WITHIN HTSC
TAPES
( )
( )
B r z
I
f r z
trans
,
,
B r z B r z
r z r z
( , ) ( ' , ' )
( , )

B r B z ( , ) ( , ) ( , ) r z
B
r z r z
r
Z
+
B r z ( , ) ( ) ( ) r z I B r
z
a
B r
trans e z

1
]
1
+

'

J ( B)
J (0)
C
C

k B
i
i
0
4
Figure 3 : Typical variation in field
components across the tape width at a
particular radius within the pancake.
It is well known that the transport current distribution
across the width of a HTSC tape is not uniform but
rather varies in a complicated manner due to the
internal structure of the filament arrangement [7]. An
expression is required which models this effect when
the level of transport current through the tape
approaches the value of the critical current.
Equation (6) is an approximation to the normalised
zero-field current distribution across the tape width.
In proposing this equation, a constant local J
c
is
assumed everywhere within the tape and an elliptical
envelope is used to modulate the current across the
width. The reason for this stems from the fact that the
amount of superconductor across the width of the tape
will vary in an elliptical manner. This is most evident
in monofilamentary tape cross sections, but is also
valid for multifilamentary tapes, as shown in Figure
4.
I z
I z
z
a
( )
( )

0
1
2
2
for {-aza} (6)
where I(z) is the current per unit width (units: A/m)
at the position z across the tape width, and I(z=0) is
the current per unit width at the position z = 0, the
centre of the tape.
5. THEORY AND CALCULATION OF THE
FIELD FREE CRITICAL CURRENT
By using equations (2) and (3), the worse-case load
line may be determined for each turn, i.e. the load
line that would be valid for a radial field applied
uniformly across the tape width and equal to that at
the edge. In this manner, a maximum field free
critical current, i
co
, can be found for each short
section, and an upper limit of I
c
(0) for the whole coil
may be estimated.
To obtain an accurate I
c
(0) value, another approach
must be adopted as follows. The tape width, w = 2a,
is divided into portions of width z. Each element
experiences a magnetic field determined by the
position z across the width, B(z)|
r,z
, each of which is
assumed to carry a current density I(z) (units: A/m)
when the transport current, I
trans
= I
c
(P). A load line
for each z element may then be drawn, and the field-
free critical current for each element calculated. The
field-free critical current for a particular turn at
radius r, defined as i
co
(r), may then be calculated by
summation of the results over the filamentary extent
2a, as described by equation (7). This equation can be
experimentally validated by measuring both i
cB
and
i
co.
c
co
a
a
i
r
i
r a
Jc B r z
Jc
I z
I z
dz
( )
( )
( ( , ))
( )
.
( )
( )
{ }

1
2 0 0
(7)
Where i
c
(r) = the measured in-field critical current of
the turn at radius r; i
co
(r) = field-free critical current
of the turn at radius r; B(r,z) = the magnetic field at
position z with the turn at radius r; J
c
(B(r,z)) is the
critical current density as a function of field, B,
(which itself is a function of the position z within the
turn at radius r - hence the nomenclature); J
c
(0) =
field free critical current density; and I(z)/I(z=0)
represents the normalised current distribution across
the tape width and is given by equation (6).
The composite expression J
c
(B(r,z))/J
c
(0) in equation
(7) describes the current distribution across the tape
width in the non-uniform magnetic field of the coil.
By combining equations (4)-(6) a final expression to
obtain i
co
(r) is then given by equation (8).
cB
co
i
r
i
r a
k
i B r
z
a
z
a
dz
i
cB e
i a
( )
( )
( ( )

1
]
1
1

'

1
1
0
4 2
2
0
(8)
which is valid for an individual turn within a
pancake. Once the validity of equation (8) has been
experimentally verified for individual turns, I
c
(0) of
the pancake may be obtained from I
c
(P) by applying
equation (8) to the turn of the pancake which
experiences the maximum edge radial magnetic field
constant, B
e
(max). For a pancake, this occurs in the
Figure 4. Cross sectional image of tape, with fitted elliptical envelope to filaments
2a = 3.0 mm
inner turns away from the extreme outer and inner
diameter of the pancake. Equation (9) must then be
evaluated in order to determine the field-free critical
current of the whole pancake, I
c
(0).
(9)
EXPERIMENTAL RESULTS
Three different lengths of HTSC tapes were fabricated
using the powder-in-tube route with dimensions (0.27
0.02) x (3.4 0.2) mm
2
. Table 1 shows the
geometrical details of the three pancakes. After heat
treatment, these were wound into pancakes with 50
m thick polypropylene insulation used to separate
the turns. A Kontron image analysis system
connected to a Carl Zeiss Axiolab A microscope was
used to capture an image of a cross section of each
tape in the BC plane.
Pancake Length Turns 2r
i
2r
o
(m) (mm) (mm)
1 155 140 300 394
2 180 163 300 415
3 877 297 850 1028
Table 1: Details of the three pancakes tested
Current leads and voltage contacts were connected to
the ends of each pancake winding so that I
c
(P) could
be measured using the standard 4 probe technique. In
addition, several pairs of voltage contacts and current
leads were interleaved between the tape and
insulation at various selected radii. In this way, i
co
(r)

,
as well as i
cB
(r)of selected turns could be directly
determined. The small field caused by a single turn
was considered negligible compared to that of the
whole pancake. The pancake coils were immersed in
liquid Nitrogen, and a voltage criterion of 1 V/cm
was used to determine the critical current value at 77
K.
The magnetic field performance data required to
solve equation (5) was collected using an Oxford
Instruments 9T magnet, with 2 circle goniometer
capability. The sample was rotated about the A axis
until the magnetic field was determined to be
perpendicular to the tape surface. The critical current
was measured in applied fields of magnitude 1 T at
77 K and the fitting constants in equation (5) were
calculated using a least squares fitting algorithm.
RESULTS AND DISCUSSION
The transport field performance of a short length of
tape is shown in Figure 5. Table 2, shown as an inset
in Figure 5 lists the derived constants used in
equation (5) to model the field performance data up to
fields of 50 mT.
Figure 4 showed a cross sectional image of the tape
employed to wind the three pancakes. An
approximate elliptical envelope to the filamentary
distribution is
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52
Magnetic Field, B (mT)
J
c
(
B
)
/
J
c
(
0
)
Data taken at 77 K
With DC Magnetic Field // C axis
shown. Figure 6 shows the resulting approximate
field free current distribution, I(z)/I(z=0)

, as
described in
equation (5), compared to the field free current
distribution in a short sample of tape, measured using
micro hall probe analysis [7]. Both curves are shown
normalised to I(z=0) and are taken for the case where
the transport current is equal to the critical current of
the tape sample. These results confirm the suitability
of equation (7) for estimating the current density
distribution across the tape width.

-1.5 -1 -0.5 0 0.5 1 1.5
z / a
Table 2:
i = 0, ki = 1.0512E+00
i = 1 ki = - 7.0410E-02
i = 2, ki = 2.89521E-03
i = 3, ki = -5.5570E-05
i = 4, ki = 3.9580E-07
Figure 5. Normalised magnetic field
performance of a short length of tape.
Elliptical
Approximation
Obtained by
micro Hall Probe
analysis
I z
I z
( )
( ) 0
z/a
Figure 6: Comparison between the adopted
approximation of the current distribution and
that measured, showing the close
approximation.
Table 3 shows the measured i
cB
and i
co
values for a
selection of turns within each pancake and the I
c
(P)
results for each pancake.
Also included in Table 3 are the calculated field
constants, B
e
(r) for the radial component of field in
equation (2). For the purposes of these calculations,
step sizes of 0.05, 0.1, 0.2 and 0.5 mm were used
with r = z, i.e. a square grid. The results obtained
for the step sizes of 0.05, 0.1, and 0.2 mm were
consistent within t 2 %. A step size = 0.5 mm was
found to be too coarse for the dimensions of pancakes
and gave results which were about t 20 % different
from those obtained with a finer mesh.
The results in Table 3 show that the calculation of i
co
from i
cB
at each turn was in good agreement with the
measured i
co
, validating the use of equation (8) to
calculate I
c
(0) from I
c
(P). The uniformity of i
co
in the
turns had to be reasonable for a good approximation
of I
c
(0) to be obtained.
Diameter
of turn (2r) (mm)
Be(r)
(mT/A)
icB(r)/ico(r)
(measured)
icB(r)/ico(r)
(calculated)
Pancake 1
325 1.7 0.70 0.68
332 1.7 0.67 0.65
374 1.7 0.8 0.79
{ Ic(P) = 12.5 A}
{ Ic(0) = 19.2 A}
Pancake 2
170 1.7 0.51 0.49
175 1.7 0.51 0.48
{ Ic(P) = 10.0 A}
{ Ic(0) = 20.0 A}
Pancake 3
888 2.0 0.68 0.64
940 2.0 0.54 0.57
962 2.0 0.79 0.79
992 2.0 0.62 0.65
1002 1.8 0.66 0.66
{ Ic(P) = 5.5 A}
{ Ic(0) = 9.6 A}
The intergrand in equation (8) represents the current
(units: A/m) in a particular turn as modified by the
presence of the magnetic field of the pancake. A plot
of this intergrand is shown in Figure 7 compared
with
the field free current density distribution
approximation. As expected, the relatively high
magnetic fields depress the current density at the edge
of the pancake. In the extreme case of a very large
edge component of radial field, the current density at
the edges of the tape would be completely suppressed,
and the current would only flow in the very centre of
the tape, being spatially-limited by the magnitude of
B
e
(R).
CONCLUSION
The field free critical current of individual turns
within a pancake winding have been measured and
predicted accurately by using a current distribution
model valid across the tape width. A method for
calculating the field-free critical current of a pancake
coil from the measured critical current has been
derived.
ACKNOWLEDGEMENTS
The authors would like to the CSIRO/TIP for the
magnetic field and micro Hall probe measurements.
REFERENCES
[1] M. Apperley, F. Darmann, R. Zhao, G.
McCaughey and T. P. Beales, to be published in
IEEE Trans. Appl. Supercond.

[2] D. Bentzon, D. Suchon, P. Bodin, Z. Han, P.
Vase, P. Skov-Hansen, R. Bruun, and J. Goul,
Appl. Superconduct., 2, 158 pp 1299-1302, 1997

[3] Z. Bodin, P. Han, M. Vase, M. D. Bentzon, P.
Skov-Hansen, R. Bruun, and J.Goul, Appl.
Supercond. 2, 299 (1997).

[4] M.N Wilson in Superconducting Magnets,
Chapter 3, p. 36, (Oxford University Press,
1983)

[5] Sasaoka and J. Sato, Cryogenics (1997) 37, 409.

[6] Horvat, W. G. Wang, R. Bhasale, Y. C. Guo, H.
K. Liu and S. X. Dou, Physica C 275, 327
(1997).

Table 3: Comparison of the measured and
calculated values of critical currents of several
individual turns within 3 different pancake coils
Figure 7. Effect of pancake field pattern on current
distribution across tape width at three different
turns within the pancake.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
N
o
r
m
a
l
i
s
e
d

C
u
r
r
e
n
t






Field free
approximation
i cB.Be = 0.01 T
i cB.Be = 0.02 T
i cB.Be = 0.04 T
[7] J.Herrmann, N. Savvides, K.-H. Muller, R.
Zhao, G. McCaughey, F. Darmann and M.
Apperley, Physica C 305, 114 (1998).