Proceedings of EPAC 2004, Lucerne, Switzerland
PERMANENT MAGNET GENERATING
HIGH AND VARIABLE SEPTUM MAGNETIC FIELD
AND ITS DETERIORATION BY RADIATION
T. Kawakubo, E. Nakamura, M. Numajiri, KEK, Ibaraki, Japan
M. Aoki, T. Hisamura, E. Sugiyama, NEOMAX Co. Ltd., Mishima-gun, Osaka, Japan
Abstract
L
8.8
(A)
(B)
175
86.3
100.1
Conventional high field septum magnet is fed by DC
current or pulse current. In the case of DC, the problem of
coil support is not very important, but the cooling of the
coil is a serious problem. While, in the case of pulse, the
problem of support is more important than that of cooling.
However, if the septum magnet is made of permanent
magnet, these problems are dissolved. And the cost for
electricity and cooling water becomes zero. Recently,
Nd2Fe14B sintered magnet which can generate high
magnetic field has been developed. Therefore, by using
this material, we made a permanent septum magnet which
has 1/4 scale of the real size and generates 1 [T] with the
variable range of +/- 10%. The magnetic field distribution
by changing the representative field in the gap is reported.
If this permanent magnet is set in an accelerator, the
deterioration of the permanent magnet by radiation will
be a serious problem. We also report the dependence of
the magnetic field generated by permanent magnet
samples on accumulated radiation from various types of
radiation source.
(C)
102.5
Figure 2: Schematic cross section of permanent septum
magnet in the case that outer structure fully removes
from the inner structure (unit: mm)
CONSTRUCTION OF MAGNET
This magnet is composed of three elements; iron, nonmagnetic steel and permanent magnet as shown
schematically in Fig. 1. NEOMAX-32EH is used for the
permanent magnet because of its high radiation resistance
as shown in the following section, and set with various
direction of magnetic pole as shown by arrows in Fig. 1.
The flatness of the magnetic field in the gap can be
obtained by shimming an iron block near the gap, and the
strength can be decreased to 80% by removing the outer
structure as shown in Fig. 2. The pictures when the outer
structure covers the inner structure (septum magnet) fully
and removes from it are shown in Fig.3.
iron
22.5
non-magnetic
steel
permanent
magnet
48.8
Inner Structure
Support Plate
Outer Structure
Figure 3: Pictures of permanent septum magnet
Left: Outer structure fully covers inner structure (Fig.1)
Right: Outer structure fully removes from inner
structure (Fig.2)
MAGNETIC FIELD
Magnetic field has been obtained not only by
measurement but also by calculation by removing the
outer structure with three steps. It can be said that the
calculations are in good agreement with the
measurements. X, Y and Z-axis are defined as horizontal,
vertical and longitudinal plane, respectively.
Horizontal distribution (X plane)
inner
structure
outer
structure
124.8
Figure 1: Schematic cross section of permanent septum
magnet covered fully by outer structure (unit: mm)
• Inside the core gap (at the centre of Y and Z plane) is
shown in Fig. 4.
• Leakage field out of the core gap (at the centre of Y
and Z plane) is shown in Fig. 5.
The horizontal width having the flat-top within 1% is
about 30mm, while the effective thickness of the septum
is 30mm. The leakage field is about 0.1% of the gap field.
1696
Proceedings of EPAC 2004, Lucerne, Switzerland
(L=0) Calculation
(L=37) Calculation
(L=120) Calculation
(L=0) Measurement
(L=37) Measurement
(L=120) Measurement
magnetic field were observed by inserting Hall prove into
the gap space.
1.2
Magnetic Field (By) [T]
1.0
Iron plate (shield for leakage field)
30mm
0.8
Permanent magnet
(upper part)
0.6
Flat top within the
difference of 1%
0.4
Iron (pole piece)
30mm 1.2mm
Acryl resin (for space)
0.2
Permanent magnet
(lower part)
0.0
10
20
8.8 (B) Inner Septum
30
40
Horizontal Plane (X) [mm]
57.5 60
50
(C) Inside Wall
Figure 7: test sample for radiation exposure
Figure 4: Horizontal magnetic field distribution in
the gap (note: (B) and (C) are shown in Fig. 2)
(L=0) Calculation
(L=0) Measurement
(L=37) Calculation
(L=37) Measurement
Permanent
magnet
name
(L=120) Calculation
(L=120) Measurement
Magnetic Field (By) [G]
5
(BH)max.
Residual
magnetic flux
density
Br
Intrinsic
coercive
force
HCJ
kJ/m3
374
342
303
279
255
T
1.42
1.33
1.26
1.21
1.15
MA/m
0.88
1.3
1.7
2.0
2.4
Energy product
(NEOMAX-)
47
44H
39SH
35EH
32EH
0
-5
-10
Leakage Field less than 0.1 % of
the Magnetic field in the gap
-15
Figure 8: Characteristics of permanent magnet
(1) γ-ray source
-20
-250
-200
-150
-100
Horizontal Plane (X) [mm]
-50
The test samples were irradiated by γ-ray from 60Co in
Japan Atomic Energy Research Institute (JAERITakasaki). As shown in Fig. 9, there is no
demagnetization in all samples.
0
(A) Outer Septum Wall
Figure 5: Horizontal leakage field distribution out
of septum (note: (A) is shown in Fig. 2)
Longitudinal distribution (Z plane)
47
(L=0) Calculation
(L=0) Measurement
(L=37) Calculation
(L=37) Measurement
35EH
Magnetic Field (kG)
(L=120) Calculation
(L=120) Measurement
6.0
5.5
5.0
0
1.2
1
2
3
4
5
6
7
8
9
10
11
12
13
Accumulated Dose of γ-ray (MGy)
1.1
1.0
Figure 9: Irradiation by γ-ray source
0.9
0.8
0.7
0.6
(2) Pure neutron source
14MeV pure neutrons were used to irradiate the
samples in the facility of Fast Neutron Source (FNS) of
JAERI-Tokai. The demagnetization of test samples is
shown in Fig. 10. The irradiated neutrons were obtained
by calculating the solid angle from neutron source.
0.5
0.4
0.3
0.2
Permanent Magnet
Core Length
0.1
0.0
-250
-200
-150
-100
-50
0
50
100
150
-113.5 Longitudinal Plane (Z) [mm] 113.5
200
250
Figure 6: Longitudinal magnetic field distribution in
the core gap
7
DETERIORATON BY RADIATION
Demagnetization
47
44H
39SH
32EH
6
The Nd2Fe14B sintered magnet can generate high
magnetic field, however, there are many reports on its
demagnetization by radiation [1]. In order to choose the
material which is least affected by radiation, test samples
(see Fig. 7) of various types of permanent magnet (see Fig.
8) were irradiated by various radiation sources, and the
1697
Magnetic field [kG]
Magnetic Field (By) [T]
44H
6.5
Magnetic fields inside the gap along the longitudinal
plane (at the centre of X and Y plane) are shown in Fig. 6
by changing L (see Fig. 2). The effective length of the
magnet is almost same as the length of the permanent
magnet core.
5
4
3
0
10
20
Irradiated neutrons (*10
30
12
40
2
/cm )
Figure 10: Irradiation by 14 MeV pure neutrons
Proceedings of EPAC 2004, Lucerne, Switzerland
11 and 12, demagnetization per (1*1012 [neutrons/cm2]) in
Fig. 13 was calculated. As shown in Fig. 13, there is some
relationship between demagnetization and intrinsic
coercive force.
2
Demagnetization [kG] per 1*10 12 neutrons/cm (y)
(3) Radiation caused by beam loss of KEK-PS
Beam loss at the extraction points of KEK-PS 500MeV
booster ring and 12GeV main ring mainly generates γ-ray
and neutron. Since we know that there is no
demagnetization by γ-ray, we only measured the
accumulated neutron fluence by using the reaction of
27
Al(n, sp)22Na. In this method, the neutron having the
energy less than 20MeV can not be detectable because of
small cross section of reaction. Because there is an
atmospheric part in the main ring extraction beam line
near septum magnet, the beam loss per unit time is about
20 times larger than that of booster extraction. The test
samples were set near the extraction place of the booster
and main ring, respectively. The demagnetization of both
cases is shown in Fig. 11 and Fig. 12, respectively.
6.5
M ag n etic field [k G ]
47
44H
35EH
5.0
2
3
4
5
12
6
7
8
2
Irradiated neutrons (*10 /cm )
Figure 11: Irradiation by neutrons produced by beam
loss at the extraction of KEK-PS-500MeV booster
47
44H
35EH
y = 0.0476x
0.010
-3.5367
y = 0.033x
0.001
0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4
Intrinsic coercive force (Hcj) [MA/m] (x)
47 (PS)
47 (FNS)
2.5
32EH
6
Magnetic field [kG]
-4.0256
0.100
Ratio of re-bound of magnetic field [%]
7
FNS (14MeV)
(5) Re-bound of magnetic field
As shown in Fig. 11, some peaks indicate that the rebound of the magnetic field arises after stopping
irradiation in the booster ring, which saturates about two
weeks after. This phenomenon also arises in the main ring
and FNS. The ratio of re-bound magnetic field (Bre) to the
field (Bbef) just before the re-bound effect seems to be
roughly proportional to Bbef as shown in Fig. 14. More
data should be taken to make this relationship more clear.
5.5
1
KEK-PS (500MeV&12GeV)
Figure 13: Relationship between demagnetization and
intrinsic coercive force
32EH
6.0
0
1.000
5
4
3
35EH (PS)
32EH (FNS)
32EH (PS)
39SH (FNS)
1.5
0.5
3
2
44H (PS)
44H (FNS)
4
5
6
7
Magnetic Field just before Re-bound Effect [kG] (Bbef)
1
Figure 14: Re-bound of magnetic field of test
samples set in the KEK-PS booster and FNS
0
0
10
20
30
40
50
60
70
80
90
100 110
120 130
140
Irradiated neutrons (*10 12 /cm2 )
Figure 12: Irradiation by neutrons produced by beam
loss at the extraction of KEK-PS-12GeV main ring
(4) Demagnetization vs. intrinsic coercive force
The above measurements show that NEOMAX-32EH
has the most radiation hardness. In order combine the date
of FNS case and KEK-PS case, the irradiated neutron
number measured by 27Al(n, sp)22Na reaction should be
extended to the neutron number having the energy less
than 20MeV. From the neutron spectrum generated by the
high energy proton beam loss, the neutron number having
the energy larger than 1MeV is estimated to be three
times of the number with the energy larger than 20MeV.
Therefore, multiplying 3 to the irradiated neutrons in Fig.
Residual radiation of permanent magnet
The residual radiation of various kinds of permanent
magnet and metal were measured and analyzed [2]. The
radiation mainly comes from 54Mn and 60Co, and the
residual radiation of NEOMAX magnet is almost the
same level as that of stainless steel.
REFERENCES
[1] Since we do not have enough space to introduce the
papers, see the references of the following paper:, K.
Makita, et. al., Journal of the Magnetics Society of
Japan, 28, 326-329 (2004) (in Japanese)
[2] M. Numajiri, et. al., Proc. of the Fifth Workshop on
Environmental Radioactivity (2004) (in Japanese)
1698