I OFL
ORNLP
1619
-
f
-
-
-
EFE
EEEFEEEE
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS - 1963
.
i.
.
.
.
5
3
1
.
.
.
.
.
.
.
ORNLt.1619
Con7-650922-9
Paper to be published in the Proceedings of the Magnet Technology
Symposium, Stanford University, Sept. 8-10, 1965
OCT 6 1965
SECTOR MAGNETS FOR A SEPARATED ORBIT CYCLOTRON*
R. S. Lord and E. D. Hudson
Oak Ridge National Laboratory
Oak Ridge, Tennessee
:
24
Summary
.7
.2
placed in a deflection sector at the appropriate
radius for the desired energy. An SOC can be
made to accelerate deuterons and alpha parti-
cles in addition to protons by adding trimming
coils to the pole tips. To so adapt a 50-MeV
proton SOC, the trimming coils should reduce
the field only about 5% at the maximum radius,
The Separated Orbit Cyclotron (SOC) consists
of individual sector magnets placed alternately
with rf cavities in a circular arrangement. In
each magnet sector the beam is bent and focused
by alternating gradient fields; ten to twenty pairs
of pole tips, depending on design, are mounted
on a common yoke and driven by a common coil
pair. A variable reluctance device under each
pole can be used to adjust the field to the desired
value, Model studies have been made, and a 50-
ton prototype sector magnet for a 50-MeV SOC
is being fabricated for further study.
The 10-50 MeV SOC
4
Introduction
The Separated Orbit Cyclotron (SOC) is an
accelerator having the unique capabilities of
high beam current, variable energy, CW oper-
ation, and essentially 100% beam extraction. The
SOC concept was first proposed by F.M. Russell'
of the Rutherford High Energy Laboratory,
England, while he was spending a year at the
Oak Ridge National Laboratory. Further details
of machine design are now being studied at Oak
Ridge, 2, 3, 4, 5, Rutherford, and Chalk River
Nuclear Laboratory ir Canada. Both Oak Ridge
and Chalk River are uiming toward building an
800-to 1000-MeV SOC with a proton beam current
up to 75 mA for the production of an intense
neutron flux. The beam power in such a machine
would be 60 MW, or more, and the thermal
neytron beam from the target would exceed
1010/cms-sec, a factor of ~10 higher than any
existing reactor.
The emphasis of the present study is on the
first stage SOC, for accelerating protons from
10 MeV to 50 MeV. This accelerator could be
built as a pilot model for the higher energy
stages; also, the high beam current would give
it unique capabilities for nuclear research up to:
50 MeV.
A model of a single sector of this first stage
is shown in Fig. 2. There would be twelve
sector magnets like the one shown; each contains
50 tons of steel, 3000 lbs of copper, and oper-
ates at 70 kW. The pole tips for eleven orbits
are mounted on a common yoke and driven by a
common pair of coils around each sector. This
type of construction is more economical than
individual magnets, especially for the closely
spaced orbits found in the higher energy stages.
The poles have the "doubleth configuration, that
is, there is a section of field with a positive
gradient immediately adjacent to a section of
field with a negative gradient for each pole pair
on any sector.
.
.
.
.
.
General Arrangement
.
!
The SOC consists of individual sector niagnets
placed alternately with evacuated rf accelerating
cavities in a circular arrangement, as shown in
Fig. 1. The beam is bent and focused in each
sector by alternating gradient fields shaped by
several pairs of pole tips. The maximum di.
ameter of the machine and the energy gain per
revolution, and hence the turn separation, are
chosen so that there is sufficient space for
separate sets of pole tips for guiding each indi.
vidual orbit. That is, the well defined beam
passes between any given pole pair only once as
it spirals outward from the injection radius to
the extraction radius. A final energy of 800 MeV
will require several stages, perhaps three SOC
stages connected in series with a linear acceler-
ator injector stage. Variable energy is obtain-
ed by extracting the beam with a 90° magnet
The radial gradient of the field in a 50-MeV
SOC will range from 1 to 2 KG/ in. and will be
constant for a given pair of poles. The beam
will be confined to a region 1 1/2 in. axially and
2 to 3 in. radially (see Fig. 3). The orbit
spacing for this stage will vary from 20 in. at
minimum radius to 10 in, at maximum radius.
In the higher energy stages the spacing may be
as small as 5 in, and the gradient as large as
3.5 kG/ in. For simplicity of construction the
average field under a pole will be 7 KG for all
stages. The correct integrated field around
each orbit is obtained by choosing the proper
length of pole tip along the beam path.
For constant gradient the profile of a pole
is a rectangular hyperbola in the region of
useful field (see Fig. 3). A bar of iron, or
neutral pole, placed on the low field side of the
gap widens the useful region of the field, as
shown in Fig. 4. The geometry of the narrow
gap side of the poles could also be altered to
extend the constant gradient region.
Magnetic Circuit
The flux path length in the yoke is longest
for poles located near the center of the yoke,
thus the average field under the poles shows a
variation with radius, see Fig. 5. To make the
.
AR
ar
"
.
...
.
.
.
EL
i
*
-
A
*Research sponsored by the U.S. Atomic
Energy Commission under contract with Union
Carbide Corporation.

.
1
,
RELEASED TOR WYOUNCTMENT
IN NUCLEAR SCIENCE ABSTRACTS
.
.
- -
.
:
-*.
-
:
-
72
.
,
:**
VIKTYR Nhid..
+
79
7
4-7
. FW
H
L
VMVA
...,
..
.-
-.
.
--.--.-.-
.
-.
-
.-.
field under all poles the same, the higher fields
are reduced to the value of the lowest field by
de coupling the poles from the yoke by the appropri-
i
ate amount. In the simplest case this can be done
by interposing a gap between the root of the pole
1.
References
F. M. Russell, Nucl. Inst, and Meth. 23
229-230 (1963).
E, D. Hudson, R. S. Lord, and R. E.
Worsham, IEEE Trans, on Nucl. Science
NS-12 No. 3, 489-493 (1965).
R. S. Lord and E. D. Hudson, IEEE Trans.
on Nucl. Science NS-12 No. 3, 373-376
(1965).
R. E. Worsham, IEEE Trans. on Nucl.
Science NS-12 No. 3, 644-647 (1965).
N. F. Ziegler, IEEE Trans. on Nucl. Science
NS-12 No. 3, 128-132 (1965).
Nucleonics 22 , 22-23 (December 1964).
6.
average field between the poles of a 1/8 scale
model is also shown in Fig. 5. More complex in
construction but easier to adjust is the system
shown in Fig. 6. When the two iron plates are
aligned in the field, the coupling between the
yoke and the pole is maximum, and when they
are displaced from each other the coupling, and
hence the gap field, is at a minimum. The effect
of the coupling device is shown in Fig. 7.
Fabrication and Tolerances
The poles will be attached to the upper and
lower yokes by clamps that ride in T-slots, as
shown in Fig. 8. At each of the two T-slots
there is a raised rail, 0.020 in. high, which
forms the pole-mounting surface. It is this
surface that determines the spacing of the magnet
poles. The magnet will require fairly tight
tolerances because small errors in magnetic field
can make a fairly large difference in the space
required for the beam. The pole mounting sur-
faces will be flat to within 0.002 inch, and the
gap tolerance between pole mounting surfaces
will be £0.002 inch. The method of machining
the pole tips is as yet undecided. It will depend
on the balance of cost against tolerance that can
be achieved. Possible methods include numeri.
cally controlled machining with either a ball end
mill or a cylindrical cutter or grinding wheel,
cutting with a template follower, or milling with
a cutter having the shape of the pole.
The water-cooled main coils will be wound
from hollow rectangular copper conductor. There
is a potential radiation damage hazard to the
insulation of both the main coils and trimming
coils if even a very small fraction of the beam is
lost within the machine. To insure against this
hazard only highly radiation resistant materials
will be used in the coil construction. These
include the inorganic materials such as glass,
asbestos, ceramics bonded with silicone resins,
and/ or epoxies.
A full-scale prototype sector magnet for a
10-50 MeV SOC is being fabricated. It will be
used to study the details of the interrelationships
between aperture, gradient, mean field level, and
element separation. Other magnet problems to
be investigated are tolerances on positioning the
various elements, alignment and adjustment
methods, and end effects, including effects at
the transition between positive and negative
gradients. The 350-800 MeV stage will require
24 sector magnets each weighing 180 tons. It is
hoped that many of the design problems for this
800-MeV stage can be solved with the 50-ton,
50-MeV prototype magnet.
11
ws
Khi
.
.
.
-
..
. wie
t.
*:
1-311
tr
.
hi
.
.
.
*
fr:11
STYL
PENY!
..,-
'
!
.
1
.
..
...::
:.
1,
-
.
*.
*
:
?'
"
..
*
T*
.
-
-*.
.
.*
t1
.
.
!
FT
!
!
15
4
MAY
:
1
..4;
.
.
1.
1 ...
-
.
.
!
!
.!
Y
.
.
..
.
-
-
.
.
...
.
..
..
..
.
.
.
.
.
.
.
V?
FH
N

7
mena
,
Vri.ip-HITE
i
*
. L'
e
YYYmpi
.
10-50 MeV SOC
12 SECTOR 75 mA
*
ICET
*
Fig. 1 Model of the 10-50 MeV SOC showing magnet sectors alternating with accelerating cavities. The beam is
extracted at various energies with a moveable 90° magnet, which requires the omission of one cavity.
*.
.
*

LOW
.
.
:..
.
•
T
2
MY
-
.
TC
.
-
,
-,
I
.
:..*...
.
.:
.
7.-
.
"
*
",
**
10-50 MeV SOC
12 SECTOR 14 MA
I
Wizy
".
..!
!
..
.
.
Fig. 2 Model of a single sector of the 10-50 MOV SOC. In the magnet sector the beam is confined to the evacuated
boam tubes between the pole tips,
*
::
.
VERNET
CIR
.
..
?!
ORNL-DWG 65-8727

-FIELD-COUPLING
SHIMS
POLE TIP
-NEUTRAL POLE
h
Wh
.
BEAM. PIPE
More
.
OR
S...-..:.
..
.
*
-TRIMMING COIL
- YOKE
TI
-
012345
INCHES
.
Fig. 3 Cross section of a portion of a SOC magnct showing some details of the pole tip construction,
ORNL-DWG 65-8357
TTC

BEAM APERTURE
WITHOUT
NEUTRAL POLE
POLE TIP
C
.
NEUTRAL
POLE-
n
*
BEAM APERTURE
WITH NEUTRAL POLE
♡
A
.
2.9-in. USEFUL FIELD
WITH NEUTRAL POLE
O
CER
IES
MAGNETIC FIELD (KG)
WITHOUT
NEUTRAL POLE
H
EN
'
o
1.7-in. USEFUL FIELD
WITHOUT NEUTRAL POLE
i
t
dur.
+_
.
"
- WITH NEUTRAL POLE
.
.
N
...
- In
O
1
2
3
4
5
6
inches
Fig. 4 Field and beam aperture, with and without the noutral pole.
".
.
.
"
f.
''
1'
-.
.
.'
"
''
-
-
ORNL-DVIG 65-5725
WITHOUT GAP
Looooo
MAGNETIC FIELD (KG)

Sot
WITH GAP

POLE BASE GAP (in.)
100
125
150
175 200 225
RADIUS (in.)
250
275
300
i
Fig. 5 Magnetic field in the beam aperture (1/8 scale model), with and without a corrective gap between the base of
the poles and the yoke.
ORNL-DWG 65-8356
WWW
UPPER YOKE

UPPER YOKE
ܥܥܥܠܠܠܠܠܠܠܥܕ

POLE
POLE
GAP -
---- GAP mam


POLE
POLE
-FIELD-
COUPLING
SHIMS
LOWER YOKE
OWER YOKE
MINIMUM COUPLING
MAXIMUM COUPLING
Fig. 6 Field coupling shimo adjusted for minimum and maximum coupling.
1
.-
I
...
.
.
!
:
FETY.
IT! 29*
.
.
TUT
3.
.."
ST
t Di
l
.
"!
he
W
a
T
.
..
..
LKiirit
V-
toid
H*
**14*
ORNL-DWG 65-5837

GAP FIELD (KG)
A4
6.6
.
0
100
10 20 30 40 50 60 70 80 90
PERCENT OF POLE BASE GAP FILLED WITH IRON
.HU
GW;
WALIMU
TLV
The area of polebase gap filled with iron
Fig. 7 Gap ficld as a function of the adjustment of the field coupling shim.
includes the area of the non-moveable side rails.
ORNL-DWG 65-5916

,
i
2
-
.
T.
FIELD-COUPLING SHIMS
a
.
POLE TIP
nine
i
we
TRIMMING
fem
.
.
COILS
.
.
.
dewind me that
fa
in
entation
Sem
To.. YOKE
TAT
NEUTRAL
POLE
.
C
1.
.
Fig. 8 A pole assomhly mounted on the yoke,
tt
i
fr
y
?
-
tit
..
.
EN
A
WE
14,
Pi
ren
22
SGVETYRTE
2 . W
KORTTI
2
2 .
Se

..
!
RSS
*
END
.
17
PEL
S.MP3
er de
"!
NO:
BY
AT
*
What
RE
..
D
&
SO
DATE FILMED
1029/65
PI
.
1
W
.
a
.