High Intensity Proton Accelerator Facility in Japan
High Intensity Proton Accelerator Facility in Japan
Yoshiharu Mori
Yoshiharu
Mori
KEK, High Energy Accelerator Research Organization
KEK, High Energy Accelerator Research Organization
Abstract.
The accelerator complex of the joint project in Japan is described.
Abstract. The accelerator complex of the joint project in Japan is described.
INTRODUCTION
INTRODUCTION
One
frontiersofofaccelerator
acceleratorscience
scienceisisthe
the
One
of ofthethefrontiers
pathtoward
towardthethehighest
highestbeam
beampower.
power. Many
Many new
new
path
accelerators
basedononthis
thisphilosophy.
philosophy. InInproton
proton
accelerators
arearebased
accelerators,
the
high
beam
power
allows
production
accelerators, the high beam power allows production
a variety
intensesecondary
secondaryparticle
particlebeams
beamssuch
such
of of
a variety
of ofintense
kaons,
neutrons, muons,
, muons,neutrinos,
neutrinos.antitprotons
antitprotons ,,
as as
kaons,
neutrons
short-lived
radioactivenuclear
nuclearbeams.
beams. InInnuclear
nuclear
andand
short-lived
radioactive
particle
physics,
exampleusing
usingthese
thesesecondary
secondary
andand
particle
physics,
ananexample
beams
measurerare
rareprocesses
processessuch
suchasasneutrino
neutrino
beams
is is
to to
measure
oscillations
and
CP
violations.
In
addition,
sciences
oscillations and CP violations. In addition, sciences
and
technologies
other
than
particle
and
nuclear
and technologies other than particle and nuclear
physics
can
be
carried
out
by
using
these
secondary
physics can be carried out by using these secondary
beams.These
Thesesciences
sciencesand
andtechnologies
technologiesinclude
include a)a)
beams.
material and life sciences with neutron and muon
material and life sciences with neutron and muon
beam, b) accelerator -driven nuclear transmutation of
beam, b) accelerator -driven nuclear transmutation of
long-lived nuclides in nuclear waste, and others related
long-lived nuclides in nuclear waste, and others related
to the various secondary particles. In order to realize
to these
the various
secondary particles. In order to realize
requirements, the new high intensity proton
these
requirements,
the new high
intensity
facility is under construction
in Japan
as proton
a joint
facility
is
under
construction
in
Japan
as
joint
project of the Japan Atomic Energy Research aInstitute
project
of the
Atomic
Energy
Research Institute
(JAERI)
andJapan
the High
Energy
Accelerator
Research
(JAERI)
and the
High The
Energy
Accelerator
Research
Organization
(KEK).
location
of the facility
is the
Organization
(KEK).
The
location
of
the
facility
is the
JAERI/Tokai site. The project has evolved from
the
JAERI/Tokai
site. The
project
has of
evolved
Neutron Science
Project
(NSP)
JAERIfrom
and the
the
Neutron
ScienceFacility
Project(JHF)
(NSP)
of JAERI
Japan Hadron
project
of KEK.andIn the
Fig.
Japan
Hadron Facility
(JHF) in
project
of KEK.
In Fig.
1, expected
beam powers
the present
facility
are
1, compared
expected with
beamthose
powers
in the
present
available
in the
world.facility are
compared with those available in the world.
FACILITY OUTLINE
FACILITY OUTLINE
The facility comprises a 400 (600)-MeV
linac, a
3GeV
raid-cycling
synchrotron
and a linac,
50-GeV
The facility
comprises
a 400 (RCS)
(600)-MeV
a
synchrotron
(MR)synchrotron
as shown in (RCS)
Fig. 2. andA ahalf
of the
3GeV
raid-cycling
50-GeV
linac are injected
the RCS
while
are
synchrotron
(MR) astoshown
in ,Fig.
2. theAother
half half
of the
further
accelerated
up
to
600
MeV
by
the
linac are injected to the RCS , while the other half are
superconducting
The RCS
further
accelerated(SC)
up linac.
to 600
MeV provides
by thea
beam power of (SC)
1MW linac.
to the pulsed
spallation
neutron
superconducting
The RCS
provides
a
source
withofthe
repetition
of 25Hz,
whileneutron
the 50beam
power
1MW
to therate
pulsed
spallation
GeV with
MR provides
a beam
current
of 15 while
microA
source
the repetition
rate
of 25Hz,
thewith
50-a
period
of
about
3-3.6
s
to
either
the
particle
GeV MR provides a beam current of 15 microA withand
a
nuclear
or the neutrino
production
target.andIn
period
of physics
about 3-3.6
s to either
the particle
case ofphysics
the neutrino
total beamstarget.
are
nuclear
or thephysics,
neutrinotheproduction
In
case of the neutrino physics, the total beams are
1C
100
1000
10000
FIGURE 1. Beam power to be achieved by the facility.
FIGURE 1. Beam power to be achieved by the facility.
extracted with the one-turn extraction.
The time
extracted
the cycle
one-turn
The time
structure ofwith
the MR
and extraction.
bunch configurations
in
structure
of
the
MR
cycle
and
bunch
configurations
in
the MR are shown in Figs. 3 and 4.
the MR are shown in Figs. 3 and 4.
The phase I of the project was approved for the
The phasestarting
I of the
project
approved
for the
construction
from
April,was
2001.
The phase
I
construction
starting
from April,
will be completed
by March
2007. 2001.
In the The
phasephase
I the I
will
completed
by March
2007.
In RCS
the phase
I the
linacbewill
be constructed
only
for the
injection
linac
will
be
constructed
only
for
the
RCS
injection
(400MeV). The 50-GeV MR will be operated with an
(400MeV).
TheGeV.
50-GeV MR Although
will be operated
with an
energy of 30
the neutrino
energy
of target
30 GeV.
Although
production
is not included
in phase the
I, theneutrino
effort
will be immediately
started
for the
approval
the
production
target is not
included
in phase
I, theofeffort
phasebeII immediately
in order to start
the experiments
for the of
longwill
started
for the approval
the
base line
and the ADS.
phase
II inneutrino
order toexperiments
start the experiments
for the long-
baseInlineTable
neutrino
experiments
and the
ADS. of the
1, the
main beam
parameters
accelerator
listed.
Theparameters
H- beams, which
In Tablecomplex
1, the are
main
beam
of the
are
produced
with
the
negative
ion beams,
sourcewhich
are
accelerator complex are listed. The Haccelerated
to
400
MeV
by
the
linac.
The
beams
are produced with the negative ion source are
are
chopped with
a chopping
56 %.
The two
accelerated
to 400
MeV byrate
the of
linac.
The beams
are
buckets inwith
the RCS
are waiting
chopped
a chopping
ratefor
of the
56 beam
%. injection.
The two
The
injection
continues
for
500
microsec,
the
buckets in the RCS are waiting for the beamwhile
injection.
magnet
system
of
the
RCS
is
operated
with
sinusoidal
The injection continues for 500 microsec, while the
waves ofsystem
a frequency
of 25isHz.
After with
the beams
are
magnet
of the RCS
operated
sinusoidal
accelerated
to
3
GeV,
the
beams
are
fast
extracted
waves of a frequency of 25 Hz. After the beamsfor
are
most of times to the muon production target and
accelerated
to 3 GeV, the beams are fast extracted for
neutron production target located in a series in the
most of times to the muon production target and
neutron production target located in a series in the
CP642, High Intensity and High Brightness Hadron Beams: 20th ICFA Advanced Beam Dynamics Workshop on
High Intensity and High Brightness Hadron Beams, edited by W. Chou, Y. Mori, D. Neuffer, and J.-F. Ostiguy
© 2002 American Institute of Physics 0-7354-0097-0/02/$ 19.00
23
In addition to the spallation neutron source and the
In addition to the spallation neutron source and the
nuclear
and particle
particlephysics
physicsexperimental
experimentalareas,
areas,thethe
nuclear and
accelerators
provide
the
beams
to
the
experimental
accelerators provide the beams to the experimental
areas for
forthe
themuon
muonscience
scienceand
andthe
theaccelerator-driven
accelerator-driven
areas
nuclearwaste
wastetransmutation
transmutationsystem
system(ADS).
(ADS).
The
muon
nuclear
The
muon
productiontarget
targetisislocated
locatedinina aseries
seriesin infront
frontof of
production
thethe
neutron
production
target
within
the
Materials
and
Life
neutron production target within the Materials and Life
Science
Experimental
Area
as
mentioned
above.
Thus,
Science Experimental Area as mentioned above. Thus,
the requirement
requirementfor
forthe
theaccelerator
acceleratorcomplex
complexfrom
fromthethe
the
muonscience
scienceisisonly
onlythe
thebeam
beampulse
pulselength
length
(100
muon
(100
ns)ns)
in in
this case.
case. Fortunately,
Fortunately,this
thisrequirement
requirementis isconsistent
consistent
this
withthe
thepresent
presentaccelerator
acceleratorscheme.
scheme.
with
Material
and Life Science Experimental Hall. Every
Material and Life Science Experimental Hall. Every
three
second,
to the
the
three second,however,
however, the
the beams
beams are
are extracted
extracted to
MR.
The
two
buckets
among
the
nine
buckets
in
the
MR. The two buckets among the nine buckets in the
MR
at aa time.
time.
MRaccept
acceptthe
thetwo
twobunches
bunches from
from the
the RCS
RCS at
This
4. After
After
Thisisisrepeated
repeatedfour
four time
time as
as shown
shown in
in Fig.
Fig. 4.
the
is
thelast
lasttwo
two bunches
bunches are
are injected
injected ,, the
the ramping
ramping is
immediately
started,
and
the
beams
are
slowly
immediately started, and the beams are slowly
extracted
Physics
extractedfor
for0.7
0.7s.s.totothe
theNuclear
Nuclear and
and Particle
Particle Physics
Experimental
Area
in
one
case
after
the
beams
are
Experimental Area in one case after the beams are
accelerated
The beams,
beams,
acceleratedup
uptoto the
theenergy
energy of
of 50
50 GeV.
GeV. The
ininother
extraction
other case,
case, are
are extracted
extracted with
with one-turn
one-turn extraction
kicker
kickerand
andseptum
septummagnets
magnets to
to the
the neutrino
neutrino production
production
target.
target.The
Thetotal
totalmain
mainring
ring cycle
cycle in
in the
the slow extraction
modeisisabout
about3.6
3.6sec.
sec.
mode
Forthe
theADS,
ADS,the
thecontinuous
continuouswave
wave
(CW)
beams
For
(CW)
beams
areare
ultimately
ultimatelyrequired.
required.The
Thelinac
linactotobebeoperated
operatedin inboth
both
the
the pulsed
pulsed mode
modeand
andthe
theCW
CWmode
modehas
hasbeen
beenonce
once
proposed
cancan
proposedfor
forthe
theNSP,
NSP,since
sincethe
theCW
CWproton
protonlinac
linac
be
operated
in
the
pulsed
mode.
However,
this
kind
of of
be operated in the pulsed mode. However, this kind
linac
current
in in
thethe
linacisisvery
veryexpensive.
expensive.IfIfthe
thepeak
peakbeam
beam
current
CW
or or
more,
than
CWmode
modeisislower,
lower,bybya afactor
factorofofthree
three
more,
than
that
that inin the
the pulsed
pulsed mode,
mode,the
thelinac
linacmay
maybebemore
more
expensive
and
pulsed
expensivethan
thantwo
twolinacs,
linacs,that
thatis,is,CW
CWone
one
and
pulsed
one.
For
this
reason,
the
ADS
experiments,
which
areare
one. For this reason, the ADS experiments, which
costly
compatible
with
the
pulsed
neutron
source,
are
costly compatible with the pulsed neutron source, are
limited
thethe
limitedtotothose
thosewhich
whichuse
usethe
thepulsed
pulsedbeams.
beams.Since
Since
pulse
length
the
3-GeV
beam
from
the
RCS
is
too
pulse length the 3-GeV beam from the RCS is too
short for the ADS experiment, the 600-MeV linac is
short for the ADS experiment, the 600-MeV linac is
necessary for the ADS experiment.
Pacific Ocean
50 GeV Synchrotron
Nuclear and Particle
Physics Experiments
3 GeV Synchrotron
Nuclear
Transmutation
Materials and Life
Science Experiments
Linac
Neutrino Beams to
Super Kamiokande
necessary for the ADS experiment.
FIGURE 2. Layout of the facility.
Accelerator Complex
Accelerator Complex
FIGURE 2. Layout of the facility.
a) Linac
a) Linac
TABLE 1 Main Beam Parameters
The linac comprises a volume-production type of
Thesource,
linac comprises
volume-production
type of
H- ion
a 50-keV alow-energy
beam transport
Hion
source,
a
50-keV
low-energy
beam
transport
(LEBT), a 3-MeV, 324-MHz Radio-Frequency
(LEBT), a(RFQ)
3-MeV,
Radio-Frequency
Quadrupole
linac, a324-MHz
50-MeV, 324-MHz
DriftQuadrupole
(RFQ)
linac,
a
50-MeV,
324-MHz
DriftTube Linac (DTL), a 200-MeV, 324-MHz
Separated
Tube(SDTL),
Linac (DTL),
a 200-MeV,
324-MHz
Separated
DTL
and a 400-MeV,
972-MHz
high-energy
DTLas(SDTL),
a 400-MeV,
972-MHz high-energy
linac
shown inand
Fig.
5. Table 2 summarizes
the main
linac as shown
Fig.will
5. Table
2 summarizes
the main
parameters.
The in
linac
be operated
at a repetition
parameters.
will be
operated
at a repetition
rate
of 50Hz.The
Thelinac
400-MeV
beam
is transported
and
rate of to
50Hz.
The at400-MeV
beam
injected
the RCS
a repetition
rate is
of transported
25 Hz. At theand
other
halfto of
the 400-MeV
injected
thethe
RCSrepetition,
at a repetition
rate of 25 beam
Hz. Atisthe
further
accelerated
600MeV the
and400-MeV
is used for
the is
other half
of the to
repetition,
beam
ADS.
further accelerated to 600MeV and is used for the
TABLE 1 Main Beam Parameters
Linac
Linac
Energy for RCS injection
400 MeV
Energy for RCS injection
400 MeV
Energy for ADS
600 MeV
Peak Current
50mA
Beam Pulse Length
500ms
Repetition
Rate
RCS
50 Hz
Energy for ADS
Peak Current
Beam Pulse Length
Repetition Rate
600 MeV
50 mA
500 ms
50 Hz
RCS
Extraction Beam Energy
3 GeV
Extraction
RepetitionBeam Energy
325GeV
Hz
Repetition
Average Beam Current
25
333Hz
mA
Average
Beam
Current
Extraction
Scheme
333mA
Fast
Extraction
Scheme
MR
Fast
MR
Extraction Beam Energy
50 GeV
Extraction
BeamCurrent
Energy
Average Beam
50
15 GeV
mA
Repetition
Average
Beam Current
0.3 Hz
15mA
Chopping Rate
56 %
Extraction Scheme
Repetition
Fast,Hz
and Slow
0.3
RFQ,
DTL, Rate
SDTL Frequency
Chopping
324
56MHz
%
Extraction Scheme
Fast, and Slow
ACS,
Frequency
MHz
RFQ,SCC
DTL,
SDTL Frequency 648324
MHz
ADS.
TABLE 2 Main parameters of the 600-MeV linac.
TABLE
2 Main parameters of the
600-MeV linac.
Energy
600 MeV
Energy
Repetition
50600
Hz MeV
Repetition
Beam
Pulse Length
50µsHz
500
Beam Pulse Length
ACS, SCC Frequency
24
500 jis
648 MHz
Peak
Current
Peak
PeakCurrent
Current
50mA
50
50mA
mA
Average
Current
Linac
LinacAverage
AverageCurrent
Current
1.25
mA
1.25
1.25mA
mA
symmetry
perhaps
important.This
Thisisisisone
oneofof
ofthe
the
symmetry
symmetryisis
isperhaps
perhapsimportant.
important.
This
one
the
reasons
for
developing
the
Annular-Ring
Coupled
reasons
for
developing
the
Annular-Ring
Coupled
reasons for developing the Annular-Ring Coupled
Structure
(ACS)
for
thehigh-energy
high-energylinac
linacstructure.
structure.
Structure
Structure(ACS)
(ACS)for
forthe
the
high-energy
linac
structure.
The
axialsymmetry
symmetryalso
alsoimply
implythe
theeasy
easymanufacturing
manufacturing
The
Theaxial
also
imply
the
easy
manufacturing
and
the
mechanical
thestructure.
structure.
and
andthe
themechanical
mechanicalstability
stabilityofofthe
the
structure.
Average
Current
after
chopping
700
jiA
Current
Average
Currentafter
afterchopping
chopping 700
700µA
µA
Total
Length
Total
Length
350
350
350mmm
(393
mmwith
with
debuncher)
(393
(393m
withaaadebuncher)
debuncher)
We
developed
the
smallest-possible
electro
Wehave
havedeveloped
developedthe
thesmallest-possible
smallest-possibleelectro
electro
quadrupole
magnets.
The
electromagnet
coils
are
quadrupole magnets.
magnets. The
The electromagnet
electromagnet coils
coils are
are
produced
using
the
electroforming
method
producedbybyfully
fullyusing
usingthe
theelectroforming
electroformingmethod
method
andthe
thewire
wirecutting.
cutting.InInthis
thisway,
way,itititbecomes
becomespossible
possible
and
this
way,
becomes
possible
usea afrequency
frequencyofof324
324MHz
MHzfor
forthe
theDTL
DTLstarting
starting
totouse
324
MHz
for
the
DTL
starting
from
from3 3MeV.
MeV.Definitely,
Definitely,the
theklystrons
klystronscan
canbe
beused
usedfor
for
from
the
klystrons
can
be
used
for
this
this frequency.
frequency. However,
However, the
thehuge
huge power
power feeding
feeding
However,
the
huge
power
feeding
systemisisnecessary
necessaryfor
forexciting
excitingthese
theseelectromagnets.
electromagnets.
system
for
exciting
these
electromagnets.
Anotherproblem
problemarising
arisingfrom
fromthe
thehigh
highaccelerating
accelerating
Another
arising
from
the
high
accelerating
frequency
frequencyis
thatthe
theaccelerating
acceleratingenergy
energyof
ofthe
theRFQ
RFQ
frequency
isisthat
that
the
accelerating
energy
of
the
RFQ
linac
isisquite
linacis
quitelimited
limited(2
2.5MeV
MeVfor
for~400MHz),
~400MHz),
linac
quite
limited
(2(2~~~2.5
2.5
MeV
for
~400MHz),
since
the
sincethe
thefour-vane
four-vanetype
typeof
theRFQ
RFQcannot
cannotexceed
exceed
since
four-vane
type
ofofthe
the
RFQ
cannot
exceed
four
times
as
long
as
the
free-space
wave
length.
four
times
as
long
as
the
free-space
wave
length.
This
four times as long as the free-space wave length.This
This
problem
is
solved
by
the
invention
of
the
p-mode
problem
is
solved
by
the
invention
of
the
p-mode
problem is solved by the invention of the p-mode
stabilizing
stabilizingloop
loop(PISL),
(PISL),which
whichisisisalso
alsoused
usedfor
forthe
the
stabilizing
loop
(PISL),
which
also
used
for
the
SNS.
The
SNS.The
ThePISLs
PISLseliminate
eliminateany
anyeffect
effectof
thedeflecting
deflecting
SNS.
PISLs
eliminate
any
effect
ofofthe
the
deflecting
field,
resulting
field,resulting
resultingin
thehigh
highquality
qualityof
theaccelerating
accelerating
field,
ininthe
the
high
quality
ofofthe
the
accelerating
and
focusing
fields.
and
focusing
fields.
and focusing fields.
extraction
Extraction
Extraction
at at
50GeV
50GeV
44
22
Injection Injection
at at 3GeV
3GeV
rise-time:
rise-time:
300nsec
300nsec
11
50
50 %
%
9 9buckets
buckets
The
anan
Themedium-energy
medium-energybeam
beam
transport
(MEBT)
also
The
medium-energy
beamtransport
transport(MEBT)
(MEET)isisisalso
also
an
important
component
ininthe
proton
linac,
ininparticular
important
component
the
proton
linac,
particular
important component in the proton linac, in particular
for
the beam
from the
forthe
theinjector
injectorlinac.
linac.First
Firstofofall,
for
the
injector
linac.
First
of all,
all, the
the beam
beam from
from the
the
RFQ
RFQ should
should bebe matched
matched toto the
the DTL
DTL both
both
RFQ should be matched to the DTL both
longitudinally
longitudinallyand
andtransversely.
transversely. Second,
Second, this
this isis the
the
longitudinally and transversely. Second, this is the
place
placewhere
whereone
onecan
canchop
chopthe
thebeam,
beam,which
whichthe
thering
ringRF
RF
place where one can chop the beam, which the ring RF
separatrix
separatrixcannot
cannotaccept
acceptfor
foritsitsphase.
phase.The
Thechopping
choppingisis
separatrix
cannot do,
accept
for the
itschopping
phase.
Thefield
chopping
is
very
since
should
verydifficult
difficulttotodo,
sincethe
choppingfield
should
very
difficult
to
do,
since
the
chopping
field
should
rise
riseand
andfall,
fall,respectively,
respectively,ininbetween
betweenthe
thetwo
twobunches.
bunches.
rise
and fall,
respectively,
indeflected
between
the
two
bunches.
Otherwise,
the
beams
Otherwise,
the
beamspartly
partly
deflectedby
bythe
thechopper
chopper
Otherwise,
the
beams
partly
deflected
by
the
would
totothe
wouldbebeaccelerated,
accelerated,eventually
eventuallygiving
givingrise
risechopper
the
would
be accelerated,
giving has
rise
tobeen
the
high-energy
beam
The
high-energy
beam loss.
loss.eventually
The RF
RF chopper
chopper
has been
high-energy
beam
loss.
The
RF
chopper
has
been
devised,
devised,and
anddeveloped
developedfor
forthis.
this.Another
Anotherdifficulty
difficultyinin
devised,
and
developed
for this.
Another
difficulty
in
thechopper
chopper
thatany
anyscraper
scraper
stoppercannot
cannot
stand
the
isisthat
ororstopper
stand
the
chopper
is of
that
any
scraper
or beams.
stopper
cannot
stand
thebeam
beamloss
loss
ofallall
the
chopped
beams.The
Thebeams
beams
the
the
chopped
the
beam
losschopped
of all the
chopped
beams.
The linac,
beams
willbe
bepartly
partly
chopped
before
entering
theRFQ
RFQ
linac,
will
before
entering
the
will
be
partly
chopped
before
entering
the
RFQ
linac,
deceleratingthe
thebeam
beambelow
belowthe
theenergy
energyacceptance
acceptance
bybydecelerating
by
decelerating
the
beam
below
the
energy
acceptance
theRFQ.
RFQ.
ofofthe
ime:
rise-timsee:c
ri1s1e0-t4nsec
n
1104
33
FIGURE
cycle.
FIGURE4.4.
4. Time
Timesturucture
sturucture
the
Main
Ring
cycle.
FIGURE
Time
stuructureofof
ofthe
theMain
MainRing
Ring
cycle.
sec )
1n c z
581n2seMHz)
58 .7 MH
1
(
72
(1. 0%
1%
10
c c
se sez) z)
8n 8nHMH
7
59.6579M
6
.
(1 (1
FIGURE
FIGURE3.3. Bunch
Bunchconfigurations
configurationsininthe
the50-GeV
50-GeVMR.
MR.
FIGURE 3. Bunch configurations in the 50-GeV MR.
Anotherfeature
featureofofthe
thelinac
linacdesign
design isisthat
thatthe
the
Another
Another transition
feature
of(200
theMeV
linacfrom
design
is tothat
the
longitudinal
transition
(200
MeV
from
SDTL
toACS)
ACS)
longitudinal
SDTL
longitudinal
MeV from
SDTL(50
to MeV
ACS)
separatedtransition
fromthe
the(200
transverse
transition
(50
MeV
isisseparated
from
transverse
transition
is
separated
from
the Ittransverse
transition
(50
MeV
from
DTLtoto
SDTL).
Itisiswell
wellknown
known
thatthe
thebeam
beam
from
DTL
SDTL).
that
from
tobeam
SDTL).
It isdegradation
well
known arise
that
loss DTL
and beam
quality
degradation
arisethe
the
loss
and
quality
atatbeam
the
loss
and beam
quality of
degradation
arise atgive
the
transitions.
Theseparation
separation
ofthe
thetwo
twotransitions
transitions
give
transitions.
The
transitions.
The separation
twothe
transitions
give
moreflexibility
flexibility
orderof
avoid
the
mismatching
ususmore
ininorder
totothe
avoid
mismatching
us
more
flexibility
in
order
to
avoid
the
mismatching
the transition,
transition, which
which gives
gives rise
rise toto the
the halo
halo
atat the
at
the transition,
which
gives rise
tothe
the
halo
formation.
should
emphasized
that
the
linac
formation.
ItItshould
bebeemphasized
that
linac
isis
formation.
Itinjector
should
that stringent
the
linacreis
usedasasananinjector
theemphasized
RCS.The
Themost
most
stringent
re
used
totobe
the
RCS.
used
as an of
injector
The
most stringent
re
alignment
of0.05
0.05to
0.1RCS.
mmis
isnecessary
necessary
for the
the
alignment
~ ~the
0.1
mm
for
alignment
0.05 ~ 0.1InIn
mm
is context,
necessary
foraxial
the
quadrupoleofmagnets.
magnets.
this
context, the
the
axial
quadrupole
this
quadrupole magnets.
of the RFQ.
Weare
aredeveloping
developingthe
theion
ionsources
sourcesboth
bothwith
withand
and
We
We cesium.
are
developing
sourcesthe
both
and
without
cesium.
firstthe
weion
attempted
the
ionwith
source
without
AtAtfirst
we
attempted
ion
source
without
cesium.
At
first
we
attempted
the
ion
source
without cesium,
cesium, that
thatis,is, purely
purelyvolume
volume production,
production,
without
without
that is, purely
volume
sincewe
wecesium,
prefer cesium-free
cesium-free
ion source
sourceininproduction,
order toto
since
prefer
ion
order
since
we
prefer
ion
source
in limit
orderofofto
avoidthe
thepossible
possiblecesium-free
decreaseininthe
thedischarge
discharge
limit
avoid
decrease
avoid
the possible
inthe
thepeak
discharge
limit
of
thefollowing
following
RFQ.decrease
However,
the
peak
beamcurrent
current
the
RFQ.
However,
beam
following
RFQ.
However,
peaktoto
beam
current
ofthe
the
cesium-free
ion
sourceisisthe
limited
mA
ofthe
cesium-free
ion
source
limited
1616mA
soso
of
cesium-free
ion source
iscesium-free
limited toion
16source
mA so
far.the
Further
improvement
thecesium-free
ion
source
far.
Further
improvement
ofofthe
far. Further improvement of the cesium-free ion source
In this context, the axial
25
has been started last March. The beam transmission
is under way. On the other hand, the cesium-seeded
is under
Ondeveloped
the other as
hand,
the cesium-seeded
ion
sourceway.
being
a back
up (of course,
ion source being developed as a back up (of course,
useless, if the RFQ cannot allow the use of the cesium)
useless, if the RFQ cannot allow the use of the cesium)
has already produced sufficient peak beam current.
has already produced sufficient peak beam current.
After the arc discharge power supply is upgraded, the
After the arc discharge power supply is upgraded, the
peak beam current was increased in proportion to the
peak beam current was increased in proportion to the
arc power up to 70 mA (above the required value) with
arc power up to 70 mA (above the required value) with
an
aperture size of 8 mm(|). The emittance measured is
an aperture size of 8 mmφ. The emittance measured is
small
on
small enough.
enough. At
At present
present the
the effort
effort is
is concentrated
concentrated on
the
increase
in
its
lifetime,
which
is
one
half
of
the
the increase in its lifetime, which is one half of the
required
requiredvalue.
value.
has
been the
started
The beam
through
RFQlast
wasMarch.
in agreement
withtransmission
the designed
through the RFQ was in agreement with the designed
value.
value.
b) Rapid Circulating Synchrotron
b) Rapid Circulating Synchrotron
We have chosen the lattice with three-folding
We have chosen the lattice with three-folding
symmetry. We need three long straight sections. One
symmetry. We need three long straight sections. One
is dedicated to the long RF acceleration section,
is dedicated to the long RF acceleration section,
another to the injection and collimation, and the other
another to the injection and collimation, and the other
to
the extraction.
extraction. The
The latter
latter two
twosections
sectionswill
willsuffer
suffer
to the
from
a
lot
of
radioactivity,
in
particular,
the
from a lot of radioactivity, in particular, the
injection/collimation
section.
It
is
preferable
to
keep
injection/collimation section. It is preferable to keep
the RF
RF section
section apart
apart from
from these
theseradioactive
radioactivesections,
sections,
the
since the
the maintenance
maintenance ofof the
the RF
RF components
components are
are
since
usually required
required more
more frequently
frequently than
than other
other
usually
components. The
Thecircumference
circumferenceofofthe
theRCS
RCSisislimited
limited
components.
by
two
factors.
One
is
the
beam
pulse
length
less
by two factors. One is the beam pulse length ofofless
than
1
ms
for
the
neutron
production,
and
the
other
than 1 ms for the neutron production, and the other isis
the circumference
circumference ofofthe
theMR.
MR.As
Asseen
seenfrom
fromFig.
Fig. 6,6,the
the
the
present circumference
circumference for
for the
the MR
MR isis perhaps
perhaps the
the
present
maximum, ifif one
oneattempts
attemptstotokeep
keepthe
theMR
MRwithin
withinthe
the
maximum,
campus. IfIf one
one increases
increases the
the circumference
circumference ofof the
the
campus.
RCS, the
thenumber
numberofofthe
thebeam
beamtransfer
transferfrom
fromthe
theRCS
RCStoto
RCS,
the
MR
must
be
decreased,
resulting
in
the
decrease
the MR must be decreased, resulting in the decrease inin
the beam
beam current
currentofofthe
theMR.
MR. Once
Oncethe
thecircumference
circumference
the
of the
the RCS
RCSisisthus
thuslimited,
limited,the
thethree-folding
three-foldingsymmetry
symmetry
of
should be
be taken
taken inin order
order toto keep
keep one
onelong
longstraight
straight
should
section for
forthe
thesufficient
sufficient RF
RFacceleration.
acceleration.Although
Althoughthe
the
section
advantages and
and the
the disadvantages
disadvantagesofofthe
thethree-folding
three-folding
advantages
symmetry have
have been
been investigated
investigatedinincomparison
comparisonwith
with
symmetry
the four-folding
four-folding symmetry,
symmetry, we
wefinally
finallydecided
decidedtotouse
use
the
the
the three-folding
three-folding symmetry
symmetry partly
partly for
for this
this reason.
reason.
Another
Another reason
reasonisisthat
thatthe
thelattice
latticewith
withthe
thethree-folding
three-folding
symmetry
symmetry isis geometrically
geometrically matched
matchedtotothe
thelandform
landform
rather
rather than
thanthe
thefour-folding
four-foldingsymmetry.
symmetry.
248.79m
3.0 ra
3.1m
27.1m
50keV3MeV
91.2m
50.1 MeV
5.7 MW
108.3m
190.8 MeV
23.6 MW
400 MeV
43.8 MW
600 MeV
(10 MW)
FIGURE.5
FIGURE.5 Layout
Layout of
of the
the linac
linac
The
under
The Lorentz
Lorentz detuning
detuning which becomes dynamic under
the
compensated.
thepulse
pulse operation
operation should
should be
be accurately compensated.
The SCC
SCC has
has been
been recently
recently power-tested with the
The
the
samepulse
pulse mode
mode as
as required.
required. The detuning is periodic
same
periodic
from pulse
pulse to
to pulse.
pulse. The
The amount of the static detuning
from
detuning
was in
in agreement
agreement with
with the
the simulation [14]
was
[14] within aa
few percent.
percent.
This detuning
detuning will be accurately
few
This
compensated, ifif one
one uses
uses a system of one SC
compensated,
SC cavity
cavity
per one
one klystron.
klystron. However,
However, a system of two SC
per
cavities per
per one
one klystron
klystron is only competitive in cost
cavities
cost
with the
the normal
normal conducting
conducting (NC)
(NC) system,
system, if
with
if one
one uses
uses
the SCL
SCL for
for the
the acceleration
acceleration from
from 200
200 MeV
the
MeV to
to 400
400
MeV.
Therefore,
the
feasibility
of
the
400-MeV
MeV. Therefore, the feasibility of the 400-MeV SCL
SCL
as an
an RCS
RCS injector
injector is
is dependent
dependent upon
upon how
as
how similar
similar the
the
detuning of
of the
the two
two cavities
cavities are
are to
to each
each other.
other. It
It is
detuning
is
recently realized
realized that
that the
the high
high field
field gradient
gradient imposes
recently
imposes
further severe
severe phase-amplitude
phase-amplitude control
control for
further
for the
the same
same
deviation
of
the
beam
energy.
For
the
same
reason
deviation of the beam energy. For the same reason as
as
thelarger
larger acceptance,
acceptance, the
the random
random kick
kick or
or walk
walk and
the
and the
the
synchrotron oscillation
oscillation during
during the
the course
course of
synchrotron
of the
the
acceleration
through
the
higher
field
gradient
cavities
acceleration through the higher field gradient cavities
becomes larger
larger in
in the
the direction
direction of
of the
the Dp/p
Dp/p in
becomes
in the
the
longitudinal
phase
space.
Under
the
presence
of
longitudinal phase space. Under the presence of the
the
Lorentz detuning
detuning the
the field
field control
control of
of the
Lorentz
the SCL
SCL is
is
obviously much harder than the NC linac. For this
obviously
much harder than the NC linac. For this
reason, we have finally decided to use the NC linac up
reason, we have finally decided to use the NC linac up
to 400. The commissioning of the 3-MeV RFQ linac
to 400. The commissioning of the 3-MeV RFQ linac
Other
Other features
features incorporated
incorporatedininthe
thelattice
latticedesign
designare
are
as
as follows.
follows. First,
First,the
thestraight
straightsections
sectionsare
aremade
madeofofnonodispersive
dispersive lattice
lattice inin order
order toto avoid
avoid the
thesynchro-beta
synchro-beta
coupling.
coupling. Second,
Second, the
the transition
transitionenergy
energyisischosen
chosenfarfar
above
the
operation
energy.
The
disadvantages
above the operation energy. The disadvantagesofofthe
the
RCS
RCS have
have been
been discussed
discussed inin Sec.
Sec. 1.2
1.2inincomparison
comparison
with
with the
the AR
AR scheme.
scheme. The
The RCS
RCS design
designshould
shouldsolve
solve
these
these problems.
problems. First
Firstofofall,
all, the
thespace
spacecharge
chargelimit
limitonon
the
the beam
beam current
current should
should be
be increased
increased asas much
much asas
possible.
possible. For
For this
thispurpose,
purpose,we
weattempt
attempttotoincrease
increasethe
the
beam
beam emittance
emittance asas large
largeasaspossible,
possible,keeping
keepingthe
thegap
gap
of
of the
the bending
bending magnets
magnets fixed.
fixed. Perhaps,
Perhaps,this
thisisismost
most
cost-effective
method.
Specifically,
we
set
cost-effective method. Specifically, we setthe
thegap
gapofof
the
the bending
bending magnets
magnetsasas210
210mm.
mm.Then,
Then,we
weattempted
attempted
to
make
the
beta
function
there
as
small
to make the beta function there as smallasaspossible.
possible.
As
As aaresult,
result,we
wecould
couldkeep
keepthe
thephysical
physicalaperture
apertureofof486
486
πn mm
mrad.
The
collimator
acceptance
was
chosen
mm mrad. The collimator acceptance was chosen
two third as wide as the physical aperture in order to
two third as wide as the physical aperture in order to
restrict the beam loss at the collimator. Finally, the
restrict the beam loss at the collimator. Finally, the
painting emittance of 216 π mm mrad was chosen two
painting emittance of 216 n mm mrad was chosen two
26
third as wide as the collimator acceptance. This
implies that the emittance growth is allows up to 1.5
times after the injection.
harmonics into the RF accelerating field. Although
the emittance growth should be carefully estimated on
the basis of the beam simulation, the tune shift of 0.16
looks reasonable for keeping the emittance growth
within 1.5 times. Taking all of these features into the
lattice design, we have eleven families of magnet
power supplies. The precise tracking of each of
families is necessary. It is really a technical challenge
how to precisely track this large number of families of
the magnets.
TABLE 3. Main parameters of the 3-GeV
synchrotron.
Energy
3GeV
Beam Intensity
8.3 x 1013 ppp
Repetition
25 Hz
Average Beam Current
333mA
Beam Power
1.0 MW
Circumference
348.3 m
Magnetic Rigidity
3.18-12.76 Tm
One of the most difficult problems inherent to the
high-energy RCS was solved by the innovative
development of the accelerating cavity loaded with
magnetic alloy(MA) [16-18], one of which is
FINEMET. This cavity can generate the field gradient
of over 50 kV/m (potentially over 100 kV/m) which is
several times as high as conventional ferrite-loaded
cavities. For this reason the RF system becomes a
reasonable size even for the high-energy RCS. Further
power test and beam test of the MA-loaded cavities are
being continued after several successful experiments.
As an injector the RCS has to match its beam
longitudinally for the injection to the MR. For this
reason the transition gamma should be much higher
than 3 GeV, although the ring circumference becomes
longer than the low transition gamma lattice. In
addition the beam should be elongated in order to
avoid a fast blow up just after the injection.
Lattice Cell Structure (3-Cell FODO x 2module arc
+ 3-Cell Straight) x 3
Typical Tune
(6.72, 6.35)
Momentum Compaction Factor 0.012
(no transition below 3 GeV)
Transition g
9.17
Total Number of Cells
27
The Number of Bend. Magnets
24
Magnetic Field
0.27 - L I T
The Number of Quadrupoles
60
Maximum Field Gradient
4.6 T/m
Harmonic Number
2
RF Frequency
1.36-1.86 MHz
c) Main Ring
The striking feature of the MR lattice is the choice
of the imaginary transition gamma. The imaginary
transition gamma is realized by the missing bend
method, in which the beta modulation is relatively
small. The missing bend structure generates the
negative dispersion at bending magnets, resulting in
the imaginary transition gamma. Similarly to the RCS,
we make the dispersionless straight section in order to
avoid the synchro-betatron coupling. The beam
emittance at the injection is chosen as 54 n mm mrad,
since it corresponds to a Laslett tune shift of ^ 0.14
with a bunching factor of 0.27 and a form factor of
1.7. The sizes of the magnets are quite reasonable by
this choice. One serious problem is that this injection
emittance is based upon the following assumption. The
painting emittance at the 400-MeV injection to the
RCS is 144 7i mm mrad, which grows by a factor of
1.5, being adiabatically damped to 54 p mm mrad.
Since the RCS collimator acceptance is 324 n mm
mrad, the extracted beams from the RCS can extend
beyond this value of emittance. The beams located
between 54 n mm mrad and 324 n mm mrad will be
lost at the collimator located at the beam transport
from the RCS to the MR. If the emittance growth is
more than estimated, the beam loss at the collimator
Ave. Circulating Beam Current9 ~ 12.4 A
RF Voltage
467 kV
RF Voltage per Cavity
42 kV (14 kV/gap)
The Number of RF Cavities
11 (+1)
Painting Emittance at Injection
216 n mm.mrad
Collimator Acceptance
324 n mm.mrad
Physical Aperture
486 n mm.mrad
Beam Emittance at Extraction
81 n mm.mrad
Bunch. Factor with 2nd harmonicO.41
Incoherent Tune Shift
0.16
Bunch. Factor without 2nd harm.0.27
Incoherent Tune Shift
0.24
The measure of the space charge effect is
represented by the incoherent Laslett tune shift
(spread). The value of the tune shift for the beam
power of 1 MW is 0.24 with a bunching factor of 0.27,
while it will come down to 0.16, if the bunching factor
is improved to 0.41 by introducing the second
27
will limit the beam current. The emittances related to
this matter are listed in Table. 5.
TABLE 5. Emittance and acceptance (n mm mrad)
in the MR cycle
Unnorm. Norm.
It is also noted that the painting emittance for the
MR injection cycle is by a factor of 1.5 less than that
for the neutron production cycle. In other words, we
have to change the painting emittance depending upon
different cycles. The Laslett tune shift for the MR
injection cycle is larger (- 0.22 ) than that for the
neutron production cycle (- 0.16).
50 GeV
Beam Intensity
3.3 x 1014 ppp
Repetition
0.3 Hz
Average Beam Current
15mA
Beam Power
0.75 MW
Circumference
1567.5 m
Magnetic Rigidity
12.8-170 Tm
17.3-22.3
96 (5.85 m )
Magnetic Field
0.143 ~ 1.9 T
Total Number of Quads.
216 (0.86,1.26,
Maximum Field Gradient
18T/m
Harmonic Number
9
RF Frequency
1.67-1.72 MHz
Beam Current (fundamental)
19 ~ 25 A
RF Voltage
280 kV
RF Voltage per Cavity
47kV(16kV/gap)
Number of RF Cavities
6
Beam Emittance at Injection
54 71 mm mrad
220
486
54
220
54
120
54
220
54-81
81
330
81
330
81
extraction (50GeV)
6.1
The slow extraction scheme is most difficult issue
to solve for this kind of high-intensity, high-energy
proton synchrotron. Only the one percent beam loss is
allowed during the slow extraction process. An
electrostatic septum (80 mmf tungsten wires with
rhenium) is being developed for this purpose. The
voltage of 230 kV, which is higher than the necessary
value of 170 kV, has been already supplied to the
electrodes. Although the beam simulation results
satisfy the above requirement, the further improvement
in the beam loss simulation will be necessary for
increasing the margin, which is needed for this kind of
the beam loss/radioactivity elimination.
The RF system of the MR will also use cavities
loaded with the same MA as that of the 3-GeV ring.
However, the Q value will be optimized for the MR.
The tunabliity of the Q value by cutting the MA core,
which is also developed for this project, is another
important advantage of the MA-loaded cavity.
1.76,1.86m)
11
54
10
1.46,1.56,1.66,
Number of Quadrupole Family
Extraction
324
extraction (30 GeV)
Momentum Compaction Factor -0.001 (imaginary gT)
Number of Bending Magnets
146
injection
+ 3-Straight Cell + 2-matching cell)
Typical Vertical Tune
144
MR
+ Insertion(2-matching cell
22.3
Injection
BT from RCS to MR
Lattice Cell Structure Arc(3-Cell DOFO x 8 module)
Typical Horizontal Tune
Physical
RCS
TABLE 4. Main parameters of the 50 GeV Main
Ring.
Energy
Collimator
Summary
The accelerator scheme for the high-intensity
proton accelerator facility project in Japan is
described. This facility has several unique featres.
First, the RCS scheme is chosen for the MW proton
machine producing the pulsed spallation neutrons.
Second, the MR aims to be a MW proton machine also
for the several-10 GeV region.
Beam Emittance at Ext.(30 GeV) 10 n mm mrad
Beam Emittance at Ext. (50 GeV)6.1 n mm mrad
REFERENCES
[1] The Joint Project for High-Intensity Proton Accelerators,
KEK Report 99-4, JHF-99-3 and JAERI-Tech 99-056
(1999).
28