The Spallation Neutron Source Project –
Physical
and Neutron
TechnicalSource
Challenges
The
Spallation
Project Physical and Technical Challenges1
1
Jie Wei22
Jie Wei
for
forthe
theSpallation
SpallationNeutron
NeutronSource
Source Collaboration,
Collaboration,USA
USA
Abstract.
power of
of 1.4 MW
Abstract.The
TheSpallation
SpallationNeutron
NeutronSource
Source(SNS)
(SNS)isisdesigned
designed to
toreach
reach an
an average
average proton beam power
pulsedneutron
neutronproduction.
production.This
Thispaper
papersummarizes
summarizesdesign
design aspects
aspects and
and physical challenges
challenges to the
forforpulsed
the project.
project.
INTRODUCTION
INTRODUCTION
TheSNS
SNSproject,
project,designed
designedtotoreach
reachananaverage
average beam
beam
The
power
above
1.4
MW
for
pulsed
neutron
production,
power above 1.4 MW for pulsed neutron production, isis
presentlyininthe
thefourth
fourthyear
yearofofa aseven-year
seven-yearconstruction
construction
presently
cycle
at
ORNL
(Fig.
!)[!].
The
accelerator
system
opercycle at ORNL (Fig. 1) [1]. The accelerator system operates
at
a
repetition
rate
of
60
Hz
and
an
average
current
ates at a repetition rate of 60 Hz and an average current ofof
− RF
mA.
consistsofofananHH~
RFvolume
volumesource
sourceofof48
48mA
mA
1.61.6
mA.
It It
consists
peak
current
at
6%
duty,
a
low-energy
beam
transport
peak current at 6% duty, a low-energy beam transport
(LEBT)housing
housinga afirst-stage
first-stagebeam
beamchopper
chopperwith
with±20
±20 ns
ns
(LEBT)
rise/fall
time;
a
402.5
MHz,
4-vane
radio-frequencyrise/fall time; a 402.5 MHz, 4-vane radio-frequencyquadrupole(RFQ);
(RFQ);a amedium-energy
medium-energybeam
beam transport
transport
quadrupole
(MEET)
housing
a
second-stage
chopper
(<
±10 ns
ns
(MEBT) housing a second-stage chopper (< ±10
rise/fall),
an
adjustable
beam-halo
scraper,
diagnosrise/fall), an adjustable beam-halo scraper, diagnosticsdevices,
devices,and
andmatching
matchingquadrupoles;
quadrupoles;aa402.5
402.5MHz,
MHz,
tics
6-tank drift-tube-linac (DTL) with permanent-magnet
6-tank drift-tube-linac (DTL) with permanent-magnet
quadrupoles; a 805 MHz, 4-module coupled-cavity-linac
quadrupoles; a 805 MHz, 4-module coupled-cavity-linac
(CCL); a 805 MHz, superconducting RF (SRF) linac of
(CCL); a 805 MHz, superconducting RF (SRF) linac of
medium- and high-/3 cavities accelerating the beam to
medium- and high-β cavities accelerating the beam to
the full energy; a high-energy beam transport (HEBT)
the full energy; a high-energy beam transport (HEBT)
for diagnostics, transverse and longitudinal collimation,
formatching,
diagnostics,
transverse
andand
longitudinal
energy
correction
painting; collimation,
and an accumatching,
energy
correction
and
painting;
and an
accumulator ring compressing the 1 GeV, 1 ms pulse
to 650
ns
mulator
ring
compressing
the
1
GeV,
1
ms
pulse
to
650
ns
for delivery onto the target through a ring-target beam
fortransport
delivery(RTBT).
onto the target through a ring-target beam
transport
(RTBT).
Table 1 lists major parameters. The energy acceptance
Table
1 listsis major
The energy
of
the ring
about parameters.
±50 MeV, mainly
due toacceptance
conditions
offor
thearing
is
about
±50
MeV,
mainly
due
toThe
conditions
tolerable H~
and
H°
stripping
loss.
back-up
forscenario
a tolerable
H− and toH0thestripping
loss.surface
The back-up
corresponds
case if the
field of
scenario
to the
if the(37.5
surface
field Exof
the SRFcorresponds
cavity is lower
thancase
expected
MV/m).
thetraSRF
cavity
is lower
than
MV/m).
Extunnel
space
(71 m)
is expected
reserved (37.5
to extend
the linac
tralength
tunnelforspace
(71 output
m) is reserved
to extend
the evolulinac
a higher
energy. Table
2 shows
length
forbeam
a higher
output energy.
Table
2 shows
evolution of
parameters
during the
cycle
including
extion
of beam
parameters
the cycle
including
expected
energy,
horizontalduring
(H), vertical
(V),
and longitupected energy, horizontal (H), vertical (V), and longitu-
FIGURE 1.1. Layout
Layout of
of the
the Spallation
Spallation Neutron
FIGURE
Neutron Source.
Source.
TABLE1.1. Spallation
Spallation Neutron
Neutron Source
Source primary
TABLE
primary parameters.
parameters.
Baseline
Baseline
Back-up
Back-up
1000
1000
±15
±15
11 + 12
11
+
12
33 + 48
33 +
48
27.5
β
=
0.61)
[MV/m]
27.5
Peak
field
E
(
p
AEP (J3 = 0.61) [MV/m
+2.5
∆E p fi
(βeld
= Ep
0.61)
±2.5
Peak
(J3 =[MV/m]
0.81) MV/m]
35
β
=
0.81)
[MV/m]
35
Peak
field
E
(
p
AEP (J3 = 0.81) [MV/m
+2 5/-T.5
∆E p (power
β = 0.81)
[MV/m]
1.4
Beam
on target,
Pmax [MW]+2.5/ − 7.5
Beam
Pmax [MW]
1.4
Pulse power
length on
on target,
target [ns]
695
Pulse
length
on target
695
Chopper
beam-on
duty[ns]
factor [%]
68
Chopper
beam-on
68
Linac macro
pulseduty
dutyfactor
factor[%]
[%]
6.0
Linac
macro pulseH~
duty
factor[mA]
[%]
6.0
Ave. macropulse
current
26
Ave.
H− current
26
Linacmacropulse
ave. beam current
[mA][mA]
1.6
Linac
beam current
1.6
Ring rfave.
frequency
[MHz][mA]
1.058
Ring
frequency
1.058
Ringrfinjection
time[MHz]
[ms]
1.0
Ring
time [ms]
1.0
Ringinjection
bunch intensity
[1014]
1.6
14 ]
Ring
intensity
[10spread
1.6
Ringbunch
space-charge
tune
0.15
Ring space-charge tune spread
0.15
975
975
+15
±15
11 + 15
11++60
15
33
33 27.5
+ 60
27.5
+2.5
±2.5
27.5
27.5
+2.5
±2.5
1.7
1.7
699
699
68
68
6.0
6.0
32
32
1.9
1.9
1.054
1.054
1.0
1.0
1.9
1.9
0.20
0.20
Kinetic energy, E [MeV]
Kinetic energy, Ekk [MeV]
Uncertainty, AE (95%) [MeV]
Uncertainty, ∆Ekk (95%) [MeV]
SRF cryo-module number
SRF
SRFcryo-module
cavity numbernumber
SRF
number
Peakcavity
fi eld Ep
(]8 = 0.61) [MV/m]
dinal (L) acceptances and emittances, and controlled and
dinal
(L) acceptances
and emittances, and controlled and
uncontrolled
beam losses.
uncontrolledDESIGN
beam losses.
PHILOSOPHY
The primaryDESIGN
concern is PHILOSOPHY
that radio-activation caused by
The
primary
concern isbeam
that loss
radio-activation
caused by
excessive
uncontrolled
can limit the machine's
excessive
uncontrolled
beam lossBased
can limit
the machine’s
availability
and maintainability.
on operational
exavailability
and maintainability.
on operational
experiences, hands-on
maintenanceBased
demands
that the averperiences,
hands-on
maintenance
thebeam
average uncontrolled
beam
loss does demands
not exceedthat
1W
power
per tunnel-meter
[2]. Uncontrolled
losses
arebeam
usuage
uncontrolled
beam loss
does not exceed
1W
power per tunnel-meter [2]. Uncontrolled losses are usu-
1
SNS is managed by UT-Battelle, LLC, under contract DE-AC05OOOR22725
for thebyU.S.
Department
of Energy.
SNS is aDE-AC05partnership
SNS
is managed
UT-Battelle,
LLC,
under contract
of six national
laboratories:
Argonne,of
Brookhaven,
Jefferson,
Lawrence
00OR22725
for the
U.S. Department
Energy. SNS
is a partnership
Berkeley,
Los
Alamos, and
Oak Ridge.
of six
national
laboratories:
Argonne,
Brookhaven, Jefferson, Lawrence
2
Brookhaven
National
Berkeley,
Los Alamos,
andLaboratory,
Oak Ridge.and on a joint appointment with
2 the Oak Ridge National Laboratory for the SNS Project.
Brookhaven National Laboratory, and on a joint appointment with
the Oak Ridge National Laboratory for the SNS Project.
1
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
38
111
§sd
CO
H
SH
pg <D .S
^ O^
MeV
m
mA
mm
µm
µm
µm
µm
10−5 π eVs
10−7 π eVs
kW
W/m
µm
µm
Beam loss [W/m]
1.75609E-18
FE
58e
<1
44
44
480
24
400
24
^ro^iCS Tf CS
DTL
1
1000
248.0
9×104
200
480
24
480
24
19×105 /π
2×107 /π
62d
1
44
44
1000
169.5
38
50
26
0.26
26
0.26
1000
94.7
38
80
50
0.23
39
0.23
18
23
N/A
0.2
0.41
0.41
^ (
en
185.6
55.1
38
30
19
0.59
18
0.59
7.4
14
N/A
1
0.39
0.39
86.8
36.6
38
25
38
0.75
42
0.75
2.4
12
N/A
1
0.33
0.33
2.9
2.9
7.6
N/A
100 f
0.21
0.21
17
17
0.05a
70
0.2
0.2
Tf
S 8
H O
o r=- o o
la
^ «o 57 ^_r *>
!/3
Q
II
g
'« 3 S ffi d >
.d .d
.
£ o
gI
.
300
400
Length [m]
500
600
700
800
SNS
SNS addresses
addresses the
the above
above seven
seven issues
issues by
by adopting
adopting
aa low-loss
design
philosophy
[3].
Above
all,
low-loss design philosophy [3]. Above all, foreseen
foreseen
losses
losses are
are localized
localized to
to shielded
shielded areas
areas using
using 1)
1) adjustable
adjustable
scrapers
scrapers in
in the
the MEBT;
MEET; 2)
2) transverse
transverse and
and momentum
momentum
collimators
collimators in
in the
the HEBT
HEBT prior
prior ring
ring injection;
injection; 3)
3) twotwostage
stage transverse
transverse collimation
collimation and
and momentum
momentum cleaning
cleaning
with
with beam-in-gap
beam-in-gap(BIG)
(BIG) kicker
kicker in
in the
the ring;
ring; 4)
4) collimator
collimator
protection
protection in
in the
the RTBT,
RTBT, and
and 5)
5) beam-gap
beam-gap cleaning
cleaning with
LEBT
LEBTand
andMEBT
MEBTchoppers
choppersand
and ring
ring BIG
BIG kicker
kicker (Fig.
(Fig. 2).
2).
Emphasis
Emphasis isis also
also put
put on
on machine’s
machine's flexibility
flexibility and reliability.
liability. The
The SRF
SRF linac
linac allows
allows operation
operation with
with one
one failed
failed
cavity/klystron;
the
ring
accepts
±5%
variation
in linac
cavity/klystron; the ring accepts ±5%
output
output energy;
energy; aa wide
wide ring
ring tuning
tuning range
range avoids
avoids resoresonances;
nances; aa robust
robust injection
injection allows
allows independent
independent horizonhorizontal,
tal, vertical,
vertical, and
and longitudinal
longitudinal painting;
painting; adjustable
adjustable collicollimation
mation systems
systems accommodate
accommodate variable
variable beam
beam size;
size; and
and
design
designreserve
reserveand
andredundancy
redundancy ensure
ensure aa high
high availability
availability
(e.g.,
(e.g.,spare
spare cryo-module
cryo-module for
for aa quick
quick replacement,
replacement, power
supplies
supplies compatible
compatible with
with 1.3
1.3 GeV
GeVenergy,
energy,multi-foil
multi-foil exexchange,
change, spare
spare kicker
kicker power
power supply
supply (PFN),
(PEN), and
and aperture
aperture
clearance
clearancefor
for one-kicker
one-kicker failure).
failure).
Finally,
Finally, the
the facility
facility isis designed
designed with
with the
the potential
potential to
to
reach
reach aa beam
beam energy
energy up
up to
to 1.3
1.3 GeV
GeV and
and aa beam
beam power
power
higher
higher than
than 22 MW,
MW, capable
capable of
of supplying
supplying aa second
second neuneutron
trontarget.
target.The
Thehigher
higher energy
energy can
can be
be reached
reached by
by upgradupgrading
ing the
thesuperconducting
superconducting RF
RF cavity
cavity gradient
gradient and
and klystron
klystron
power
power supplies,
supplies, and
and by
by filling
filling the
the presently
presently unoccupied
unoccupied
linac
linactunnel
tunnelspaces
spaceswith
with up
up to
to 99 additional
additional cryo-modules.
cryo-modules.
The ring
ring isis capable
capable of
of accommodating
accommodating the
the energy
energy and
and
The
power increase
increase without
without extensive
extensive hardware
hardware change
change –power
space isis reserved
reserved for
for two
two additional
additional extraction
extraction kickers,
space
andfor
forthe
the replacement
replacement of
of 22 injection-chicane
injection-chicane dipoles
dipoles to
to
and
0 stripping
satisfy HH°
stripping conditions
conditions [4].
[4].
satisfy
2.5
3.6
38
32
250
3.7
51
3.7
4.7
10
0.2b
2
0.27
0.27
2.5
3.8
38
7
r~«.
O
200
FIGURE
FIGURE 2.2. Expected
Expected beam
beam loss
loss across
across the
the SNS
SNS accelerator
accelerator
complex.
complex.The
Theuncontrolled
uncontrolled beam
beam loss
loss isis at
at 11W/m
W/m level.
level.
. .
Z O O O
0.065
0.12
47
O
100
Source:
Source:Data
Data from
from N.
N. Catalan-Lasheras,
Catalan-Lasheras, et
et al
al
387
64.2
38
80
57
0.41
55
0.41
7.2
17
N/A
0.2
0.41
0.41
cnooinOcno^H
RTBT
0
en
oo 0 0 °i ON ^. oo
RING
HEBT
SCL
5c
<1
0.46
0.46
Ring
HEBT
SRF 2
SRF 1
CCL
00 0 VO <
vo
q
5.00E-08 5.71E-07 9.37E-06
1.5
0.5
Ek (out)
Length
Peak current
Min. trans. aperture
Min. H acceptance
H emit. out, εun,rms
Min. V acceptance
V emit. out, εun,rms
Min. L acceptance
L emit. out, rms
Loss (control)
Loss (uncontroll)
H emit. out, εN,rms
V emit. out, εN,rms
si
DTL
o '"id
MEBT
•» S
|1
f
l
O <u
during
duringnormal
normaloperation
operation
2
CCL
RFQ
•8
High rad
areas
loss
3.26951E-18Uncontrolled
9.00E-08 1.91E-06
Uncontrolled
loss 1.41E-05
2.5
a a
LEBT
s •2-2
3
0
fi'S
I -I »I
<a ag= at a a
1000
150.8
9×104
Unit
1
5 <u£
13
RTBT
TABLE 2. Beam parameter evolution across the SNS accelerators. The aperture and acceptance do not include scrapers and collimators.
Notes are: a) corresponding to 27% chopped beam; b) corresponding to 5% chopped beam; c) beam loss on the transverse and momentum
collimators; d) including total 4% of beam escaping foil and 0.2and f) corresponding to 20% beam loss averaged over RFQ length.
Tolerated losses in Watt/m
a<D <Da
allyattributed
attributedtoto1)1)mismatch
mismatchupon
uponchange
changeof
oflinac
linacstrucstrucally
ture,lattice,
lattice,and
and frequency;
frequency; 2)
2) space-charge
space-charge effects
effects ininture,
cluding envelope
envelope and
and parametric
parametric resonances
resonances and
and nonnoncluding
equipartition inin the
the linac,
linac, and
and resonance
resonance crossing
crossing and
and
equipartition
instabilityenhancement
enhancementininring;
ring;3)
3)limited
limited physical
physicaland
and
instability
− and
0 stripmomentumacceptance;
acceptance;4)
4) premature
premature HH~
and HH°
stripmomentum
pingand
andring
ringinjection
injectionfoil
foilscattering;
scattering;5)
5)magnetic
magneticerrors,
errors,
ping
fringefields,
fields,and
andmisalignments;
misalignments;6)
6)instabilities
instabilities(resistive
(resistive
fringe
impedances due
due toto e.g.
e.g. extraction
extraction kicker,
kicker, and
and electron
electron
impedances
cloud);and
and7)7)accidental
accidentalloss
lossdue
duetotosystem
systemmalfunction
malfunction
cloud);
(ionsource
sourceand
andlinac,
linac,ring
ringextraction
extractionkickers).
kickers).
(ion
ACCELERATOR DESIGN
DESIGN CHOICES
CHOICES
ACCELERATOR
Superconducting vs.
vs. Warm
Warm Linac
Linac
Superconducting
The SRF
SRF linac
linac operating
operating at
at 805
805 MHz
MHz frequency
frequency acacThe
− beam
celerates the
the H
H~
beam from
from 186
186 MeV
MeV to
to top
top energy
energy
celerates
(Fig.3).
3).Comparing
Comparingwith
with the
the original
original normal-conducting
normal-conducting
(Fig.
(warm) CCL
CCL linac,
linac, the
the SRF
SRF linac
linac provides
provides aa high
high acac(warm)
celerating gradient
gradient (11
(11 -- 16
16 MV/m)
MV/m) capable
capable of
of reachreachcelerating
39
movable fixed
fixed
movable
movable
fixed
scraper collimators
collimators
scraper
scraper
collimators
Front End
LBNL
W Injector
injection
injection septum
septum
injection
septum
&
&bumps
bumps
&
bumps
FIGURE
and linac
linac stucture.
stucture.
FIGURE 3.
3. SNS
SNS front-end
front-end and
and
linac
stucture.
ing
ing aa higher
higher beam
beam energy,
energy, encounters
encounters less
less beam
beam loss
loss and
and
halo
scraping
due
to
its
larger
bore
radius,
is
immune
halo scraping due to its larger bore radius, is immune to
to
one cavity/klystron
operates at
at aa better
better vacuum,
vacuum,
cavity/klystron failure,
failure, operates
and is expected
and availabilavailabilexpected to
to have
have higher
higher reliability
reliability and
of
two
types
of
SRF
cavities
allows
ity. The selection
SRF
cavities
allows
for
selection of two types of
allows for
for
economic savings
savings and
and future
future energy
energy upgrades.
upgrades. On
On the
the
other hand, the
the relatively large
large phase
phase slip
slip requires
requires dedetailed error-sensitivity analysis.
analysis. The
The choice
choice of
of cavity
cavity gegeometric
/3
β value
value is
is based
based on
on aa smooth
smooth transition
transition from
from the
the
ometric β
warm
warm section
section linac,
linac, aa maximized
maximized final
final output
output energy,
energy, and
and
ββ secaa comfortable
high-/3
secseccomfortable transition
transition from
from mediummedium- to
to highhighalso
choose
tion with tolerance to
to one
one cavity
cavity failure.
failure. We
We
also
choose
We
constant-gradient, continuous
constant-gradient,
continuous focusing
focusing to
to maximize
maximize the
the
accelerating field
accelerating
field strength
strength [5].
[5].
Considering the
Considering
the tight
tight construction
construction schedule,
schedule, aa modermoderate peak
peak surface
ate
surface field
field of
of 27.5
27.5 MV/m
MV/m is
is chosen
chosen for
for the
the
mediumβ
cavity.
Benefiting
from
electro-polishing,
medium-/3
medium-β cavity. Benefiting from electro-polishing, aa
higher peak
peak field
higher
for the
field of
of 35
35 MV/m
MV/m is
is assumed
assumed for
for
the highhighβ
cavity.
In
order
to
reduce
uncertainties
in
RF
/3
β cavity. In order to reduce uncertainties in RF controls
controls
of an
β
of
an ion
ion (((/3
< 1)
detuning, micromicroβ<
<
1) beam
beam under
under Lorentz
Lorentz detuning,
detuning,
microphonics, beam
phonics,
beam transients
transients and
and injecting
injecting energy
energy offset,
offset, we
we
decide to
to drive
decide
drive each
each cavity
cavity with
with its
its own
own klystron
klystron using
using
independent amplitude
independent
amplitude and
and phase
phase control.
control.
extraction
extractionkickers
kickers
extraction
kickers
beam
beam gap
gap kicker
kicker
extraction
extractionseptum
septum
extraction
septum
beam
beam
RF
RF
RF
\
/
instrumentation
instrumentation
instrumentation
FIGURE
FIGURE4.
4. SNS
SNSaccumulator
accumulatorring
ringlayout.
layout.
FIGURE
4.
SNS
accumulator
ring
layout.
magnetic
magneticerrors
errorsdue
dueto
toeddy
eddycurrent,
current,ramping,
ramping,saturation,
saturation,
magnetic
errors
due
to
eddy
current,
ramping,
saturation,
and
and power-supply
trackingis
isnon-trivial.
non-trivial.The
Thestudy
studyconconand
power-supply tracking
tracking
is
non-trivial.
The
study
concluded
that
the
required
RCS
design
is
technically
that the
the required RCS
RCS design
design is
is technically
technicallymore
more
cluded that
more
demanding
andless
lesscost
costeffective
effective [4].
[4].
demanding and
and
less
cost
effective
[4].
Permanent
magnets
were
Permanent magnets
magnets were
wereconsidered
consideredas
asan
anoption
optionfor
for
Permanent
considered
as
an
option
for
the
ring
magnets.
Electromagnetic
the accumulator
accumulator
ring
magnets.
Electromagnetic
magnets
accumulator ring magnets. Electromagnetic magnets
magnets
were
chosen instead,
instead, given
given the
the uncertainty
uncertainty in
in the
the linac
linac
were chosen
chosen
instead,
given
the
uncertainty
in
the
linac
energy.
energy. This
This choice
choice is
is especially
especially appropriate
appropriateto
toaccomaccomenergy.
This
choice
is
especially
appropriate
to
accommodate
modatelater-adopted
later-adoptedSRF
SRFlinac.
linac.
modate
later-adopted
SRF
linac.
Ring’s
Ring's FODO-doublet
FODO-doubletLattice
Lattice
Ring’s
FODO-doublet
Lattice
The
The four-fold
four-fold symmetric
symmetric ring
ring lattice
lattice contains
contains four
four
four-fold
symmetric
ring
lattice
contains
four
dispersion-free
straights, each
each housing
housing injection,
injection, collicollidispersion-free straights,
straights,
each
housing
injection,
collimation,
mation, RF,
RF, and
and extraction,
extraction, as
as shown
shown in
in Fig.
Fig. 4.
4.◦Each
Each
mation,
RF,
and
extraction,
as
shown
in
Fig.
4.
Each
◦ horachromatic
arc
consists
of
4
FODO
cells
with
90
achromatic arc
arc consists
consists of
of 44 FODO
FODO cells
cells with
with 90
90°
horachromatic
horizontal
izontalphase
phaseadvance.
advance.
izontal
phase
advance.
After
optimization, the
the ring
ring lattice
lattice has
has doublet
doublet
After optimization,
optimization,
the
ring
lattice
has
doublet
straights
[3].
The
lattice
combines
the
FODO
straights [3].
[3]. The
The lattice
lattice combines
combines the
the FODO
FODO strucstrucstraights
structure’s
ture's simplicity
simplicity and
andease
easeof
ofcorrection
correctionwith
withthe
thedoublet
doublet
ture’s
simplicity
and
ease
of
correction
with
the
doublet
structure’s
structure's flexibility
flexibility for
for injection
injection and
and collimation.
collimation. InInstructure’s
flexibility
for
injection
and
collimation.
Injection
jection at
at
dispersion-free
region
allows
independently
at aaa dispersion-free
dispersion-free region
regionallows
allowsindependently
independently
adjustable
adjustable
painting
in
the
transverse
(with
orbit
bumps
adjustable painting
painting in
in the
the transverse
transverse(with
(withorbit
orbitbumps
bumps
in
the
ring)
and
longitudinal
(with
an
energy-spreading
in the ring)
ring) and
and longitudinal
longitudinal (with
(with an
an energy-spreading
energy-spreading
phase-modulated
phase-modulated RF
RF
cavity
in
the
HEBT)
directions
RF cavity
cavity in
in the
the HEBT)
HEBT) directions
directions
for
a
robust
operation.
The
12.5
m-long
for a robust operation. The
uninterrupted
The 12.5
12.5 m-long
m-long uninterrupted
uninterrupted
straight
straight
section
with
flexible
phase
advance
further
straight section
section with
with aaa flexible
flexible phase
phase advance
advance further
further
improves
collimation
efficiency.
Comparing
improves
with
the
improves collimation
collimation efficiency.
efficiency. Comparing
Comparing with
with the
the
original
original
all-FODO
lattice,
matching
between
the
arcs
original all-FODO
all-FODO lattice,
lattice, matching
matching between
between the
the arcs
arcs
and
and
the
straights
increases
the
arc
acceptance
by
50%
and the
the straights
straights increases
increases the
the arc
arc acceptance
acceptanceby
by50%
50%
with
the
same
magnet
aperture
(Fig.
5).
with the same
same magnet
magnet aperture
aperture (Fig.
(Fig. 5).
5).
Accumulator
Accumulator Ring
Ring vs.
vs. RCS
RCS
During the
During
the first
year of
first year
of construction,
construction, aa study
study was
was
performed comparing
performed
comparing the
the present
present structure
structure of
of fullfullenergy linac
energy
linac plus
plus accumulator
accumulator ring
ring to
to aa rapid-cyclingrapid-cyclingsynchrotron
(RCS)
design:
a
60
Hz,
400
MeV linac
linac feeds
feeds
synchrotron (RCS) design: a 60 Hz, 400 MeV
two,
vertically
stacked
RCSs
accelerating
the
proton
two, vertically stacked RCSs accelerating the proton
beam to
to 2
challenge to
to the
the
beam
2 GeV
GeV energy.
energy. The
The biggest
biggest challenge
RCS
design
is
from
the
stringent
(1
W/m)
beam-loss
criRCS design is from the stringent (1 W/m) beam-loss criterion: although
by aa factor
factor of
of 5,
5, still
still only
only 0.4%
0.4%
terion:
although relaxed
relaxed by
uncontrolled
loss
is
allowed
for
a
2
MW
beam
power
uncontrolled loss is allowed for a 2 MW beam power
assuming
assuming 90%
90% collimation
collimation efficiency.
efficiency. On
On the
the other
other hand,
hand,
among
existing
rings
the
lowest
loss
of
about
0.3% is
is
among existing rings the lowest loss of about 0.3%
achieved at
at LANL’s
LANL’s PSR,
800 MeV
MeV accumulator,
accumulator, as
as
achieved
LANL's
PSR, aa 800
opposed to
to typical
of aa few
few to
of percent
in
opposed
typical losses
losses of
to tens
tens of
percent in
RCSs
(e.g.
ISIS,
FNAL
and
AGS
Boosters).
RCSs (e.g. ISIS, FNAL and AGS Boosters).
As opposed
opposed to
the RCSs
operating at
at
As
to the
the accumulator,
accumulator, the
RCSs operating
30
Hz
require
a
high
RF
voltage
(about
400
kV per
per ring
ring
30 Hz require a high RF voltage (about 400 kV
at 1.4
1.4 -- 1.9
1.9
MHz)
for
fast
acceleration,
large
magnet
at
1.9 MHz)
MHz) for
for fast
fast acceleration,
acceleration, aaa large
large magnet
magnet
aperture
aperture to
to accommodate
accommodate the
the space
space charge
charge at
at aa lower
lower enenergy,
ergy, ceramic
ceramic vacuum
vacuum pipes
pipes with
with detailed
detailed RF
RF shielding,
shielding,
and high-performance power supplies. Minimization of
and high-performance power supplies. Minimization of
CHALLENGES
CHALLENGES
&
LESSONS
LEARNED
CHALLENGES &
&LESSONS
LESSONSLEARNED
LEARNED
Front
Front
End
&
Warm
Linac
Front End
End &
&Warm
WarmLinac
Linac
Tight
focusing used
for chopping
and antichopTight optical
optical
optical focusing
focusing used
used for
for chopping
chopping and
and antichopantichopping
in
a
long
MEBT
is
a
source
of
beam-halo
generping in
a
long
MEBT
is
a
source
of
beam-halo
in a long MEET is a source of beam-halo genergeneration.
Studies
show
that
even
without
the
antichopper,
ation.
ation. Studies
Studies show
show that
that even
even without
without the
the antichopper,
antichopper,
partially
particles are
still mostly
contained by
partially deflected
deflected
deflected particles
particles are
are still
still mostly
mostly contained
contained by
by
40
6.23 QQyy== 6.20
6.20
QQxx==6.23
6
βx
βy
5
β
1/2
1/2
[m ]
ηx
4
3
2
4
1
3
1
η [m]
2
0
0
20
40
S [m]
60
80 -1
80
Time: Sun Dec 1918:38:10 1999 Last file modify time: Fri Dec 1713:56:46
Time: Sun Dec 19 18:38:10 1999 Last file modify time: Fri Dec 17 13:56:46 1999
FIGURE5.5. SNS
SNS ring
ring lattice
lattice super-period
super-period of
of FODO/doublet
FODO/doublet
FIGURE
structure.
The
lattice
periodicity
is
4.
structure. The lattice periodicity is 4.
the envelope
envelope of
of the
the nominal
nominal unchopped
unchopped beam
beam [6].
[6]. The
the
The
MEET
quadrupoles
are
thus
made
independently
adMEBT quadrupoles are thus made independently adjustable
so
that
alternative
optics
can
be
realized,
avoidjustable so that alternative optics can be realized, avoidingtight
tightfocusing
focusing atatthe
the antichopper
antichopper or
or MEBT
MEET chopper.
chopper.
ing
Permanent-magnet
quadrupoles
are
used
in
Permanent-magnet quadrupoles are used in the
the DTL
DTL
due to the tight geometry (402.5 MHz starting at 2.5
due to the tight geometry (402.5 MHz starting at 2.5
MeV), although electromagnetic quadrupoles could be
MeV), although electromagnetic quadrupoles could be
used at DTL tank 3 and beyond. During 1999, the aperused at DTL tank 3 and beyond. During 1999, the aperture of CCL was reduced from 4 to 3 cm for cost savings.
ture of CCL was reduced from 4 to 3 cm for cost savings.
Later when SRF linac is adopted, simulated beam loss ofLater when SRF linac is adopted, simulated beam loss often occurs near the end of CCL as the focusing strength
ten occurs near the end of CCL as the focusing strength
is reduced to match the SRF optics.
is reduced
to match the
SRF performance
optics.
A key challenge
in linac
is to minimize
A
key
challenge
in
linac
performance
minimize
beam emittance growth and centroid jitterisintoboth
transbeam
emittance
growth
and
centroid
jitter
in
both
transverse and longitudinal directions upon ring injection,
reverse
and
longitudinal
directions
upon
ring
injection,
reducing foil traversal, scattering and radio-activation. The
ducing
foil
traversal,
scattering
and
radio-activation.
The
warm DTL operating at 402.5 MHz is expected to be less
warm
DTLtooperating
at 402.5
expected
to be
less
sensitive
vibrational
noisesMHz
than ismost
existing
linacs
sensitive
to
vibrational
noises
than
most
existing
linacs
operating at 200 MHz. A tight RF control (<0.5% amoperating
at 200
tightwarrants
RF control
(<0.5%
amplitude and
0.5° MHz.
phase A
error)
tolerable
energy
plitude
andbefore
0.5◦ phase
error)
warrants
energy
variation
the beam
enters
energytolerable
correction
and
variation
theinbeam
enters[7].
energy correction and
spreadingbefore
cavities
the HEBT
spreading cavities in the HEBT [7].
Source: Datafroml.Hofmann
Source: Data from I. Hofmann
FIGURE 6. Analytical resonance chart showing Instability
FIGURE 6. Analytical resonance chart showing Instability
growth rate due to space-charge coupling resonance. The efgrowth rate due to space-charge coupling resonance. The effects become important only when the transverse and longitufects become important only when the transverse and longitudinal tunes are on resonance, and when the emittances differ
dinal tunes are on resonance, and when the emittances differ
signifi cantly. The dash lines indicate equipartition.
significantly. The dash lines indicate equipartition.
trum is possible because of the pulsed time structure of
trum is possible because of the pulsed time structure of
the beam and the fact that the beam frequency shifts with
the beam and the fact that the beam frequency shifts with
variable ring energy and repetition rate (e.g. for some
variable ring energy and repetition rate (e.g. for some
two-target operation scenarios). Fortunately, transverse
two-target
operation
transverse
and
longitudinal
(beamscenarios).
break-up) Fortunately,
instabilities are
minor
and
longitudinal
(beam
break-up)
instabilities
are minor
issues for an ion beam in the presence of a cavity-toissues frequency
for an ionspread
beam [9].
in the
presence
of aarecavity-tocavity
HOM
dampers
implecavity
frequency
spread
[9].
HOM
dampers
are[10].
implemented only for the purpose of power dissipation
mented
only
for
the
purpose
of
power
dissipation
[10].
The SRF linac performance is limited by the availThe
SRF linac
performance
the available
klystrons
power
(550 kW). isUplimited
to 40%byRF-power
able
klystrons
power
(550
kW).
Up
to
40%
RF-power
is reserved for compensation of cavity errors (Lorntz deis reserved
for compensation
ofloss,
cavity
errors (Lorntz
detuning,
microphonics,
coupling
frequency
setting),
tuning,
microphonics,
coupling
loss,
frequency
setting),
klystron loss, and missing-cavity tuning. To reduce such
klystron loss,
tuning.
reducecryssuch
overhead,
eachand
SRFmissing-cavity
cavity is equipped
withToa piezo
overhead,
each
SRF
cavity
is
equipped
with
a
piezo
crystal driven fast tuner to compensate for the Lorentz force.
tal driven fast tuner to compensate for the Lorentz force.
Superconducting RF Linac
Superconducting RF Linac
Ring and Transport
Ring and Transport
Using only two types of cavity /3 for over 800 MeV of
Using onlycompromises
two types of cavity
β for over 800
of
acceleration
the equipartition
law.MeV
Spaceacceleration
compromises
the
equipartition
law.
Spacecharge coupling can cause transverse and longitudinal
charge
coupling
canwhen
causethe
transverse
emittance
exchange
emittanceand
ratiolongitudinal
meets resemittance
exchange (Fig.
when6)the[8].
emittance
ratio depending
meets resonance conditions
In addition,
onance
conditions
(Fig.
6) [8]. Inspace-charge
addition, depending
on the level
of initial
mismatch,
parameton
initial mismatch,
ricthe
halolevel
mayofdevelop
in the linac.space-charge
Efforts were parametmade to
ric
halo may
develop in the
linac. Efforts
were made
reserve
an economically
affordable
large aperture,
and to
reserve
economically
affordable
large
and to
reservean
tunability
in the MEET,
CCL
andaperture,
SRF linac.
reserve
tunability
in the MEBT,
and on
SRF
Effects
of higher-order
modesCCL
(HOM)
thelinac.
cavities
isEffects
anotherofissue.
Overlapping
of (HOM)
beam and
spechigher-order
modes
on HOM
the cavities
is another issue. Overlapping of beam and HOM spec-
Solid-steel, as opposed to laminated-steel, was seSolid-steel,
to laminated-steel,
was
selected
for most as
ringopposed
and transport
magnet cores for
cost
4 for cost
lected for
most ring and
transport
magnet
cores
savings.
Individually,
good
field quality
(<10~
relative
−4
savings.
Individually,
field quality
(<10excessive
relative
error
at full
acceptance)good
is achieved.
However,
error
at
full
acceptance)
is
achieved.
However,
excessive
(up to 0.25%) magnet-to-magnet variation is found
in
(up dipole
to 0.25%)
magnet-to-magnet
is found in
the
transfer
function and its variation
current dependence,
theshown
dipoleintransfer
anddipoles
its current
dependence,
as
Fig. 7 function
[11]. These
are shimmed
to
as shown
in Fig.
These
shimmed
achieve
below
10~74 [11].
variation
fordipoles
1 GeV are
operation,
andto
achieveaccording
below 10to−41.3
variation
for 1 GeV operation,
and
sorted
GeV measurement
data to minisortedorbit
according
to strength.
1.3 GeV measurement data to minimize
corrector
mize orbit corrector strength.
41
i|i
Hi
•
I
I
11Ml
•I-4iii|
— I mtfpm
•-i^
r?:^r*S?
llii;|Hi:|sll;
I
I
I
I
I
I
1
llilill
•Bl
iiiiiiiii
isisis .SHiHi::*HiHi::*Hi?3feissis
III
;***; ;*;*; 11
10 10
10
8 8
8
6 6
6
4 4
4
2 2
2
0 0
30 310 10 20 20 30 30 40 40 50 50
30
f20
[MHz]
[MHz]
3 10
20
30
40
50
ff[MHz]
f [MHz]
ii|1111
iiiiiiiiii: ***
Nil
I
I
HiiHiiisH
H
H
Imag Zy [kΩ/m]
Imag Zy [kΩ/m]
Imag Zy [kΩ/m]
—!
ijil
11111
IIIIII
Source:
DataData
fromfrom
P. Wanderer,
A.A.Jain,
etetalet
Source:
P. Wanderer,
A.
Jain,
Source:
Wanderer,
Jain,
alal
Source: Data
Data from
from P.
P. Wanderer,
A. Jain,
et al
Real Zy [kΩ/m]
Real
RealZZy y[kΩ/m]
[kΩ/m]
lllilli
sisisiisisi
S*s*
i*.s;s; HiiiSHiiSHi
Real Zy [kΩ/m]
Real Zy [kΩ/m]
Real Zy [kΩ/m]
lliipilpil
llllfllllllllOllllll;
25£2
termination
25Ω
termination
25Ωtermination
termination
25Ω
12
12
12
CMD5005
10
CMD5005
10
CMD5005
C2050
10
C2050
8
C2050
8
8
6
6
6
4
4
4
2 2
2
0 0
30 3310 10
10 2020
20 3030
30 404040 505050
f [MHz]
3 10
20
30
40
50
[MHz]
ff[MHz]
f [MHz]
10 10
10
8 8
8
6 6
6
4 4
4
2 2
2
0 0
30 3 10 10 2020 3030 4040
20
30
f [MHz]
ff[MHz]
3 10
20
30
40
[MHz]
f [MHz]
Imag Zy [kΩ/m]
Imag
ZyZ[kΩ/m]
Imag
[kΩ/m]
y
Open
termination
Open
termination
Open
termination
Open
termination
12
1212
10
1010
8
8 8
6
6
6
4
4
4
2 2
2
0 0
30 3 31010
10 2020
20 30 30
30 40 40
40 50 50
50
f [MHz]
3 10
20
30
40
50
ff[MHz]
[MHz]
f [MHz]
Integral Transfer Function at 1..0 <3**t 1« SDlf Dipoles
5050
50
FIGURE
7.
of
integral
transfer
function
ofofof
SNS
FIGURE
7.Variation
Variation
of
integral
transfer
function
SNS
Source:
Data
from
D.D.
Davino,
H.H.H.
Hahn
Source:
Data
from
Davino,
Hahn
FIGURE
7.
Variation
of
integral
transfer
function
SNS
Source:
Data
from
Davino,
Hahn
FIGURE
7.
Variation
ofand
integral
transfer
function
of
SNS
Source:
Data
from
D.D.Davino,
H. Hahn
ring
dipole
magnets
before
after
shimming.
The
measurering
dipole
magnets
before
and
after
shimming.
The
measurering
dipole
magnets
before
and
after
shimming.
The
measurering
dipole
magnets
before
and
after
shimming.
The
measureFIGURE
9.
Comparison
of
bench-measured
coupling
FIGURE
9.
Comparison
of
bench-measured
coupling
ment
current
corresponds
1 GeV
beam
energy.
ment
current
corresponds
to 1to
GeV
beam
energy.
FIGURE 9.9. Comparison
Comparison ofof bench-measured
bench-measured coupling
coupling
ment
energy.
FIGURE
ment current
current corresponds
corresponds to
to 11 GeV
GeV beam
beam energy.
impedancefor
foropen
openand
and2525
25ΩQ
ΩPFN
PFN
termination,
and
high
impedance
impedance
for
open
and
PENtermination,
termination,and
andhigh
high
impedance
for
open
and
25
Ω
PFN
termination,
and
high
(1600)andandmedium
medium(100)
(100)permeability
permeabilityferrite
ferriteofofthe
thering
ring
(1600)
(1600)
and
medium
(100)
permeability
ferrite
of
the
ring
(1600)
and
medium
(100)
permeability
ferrite
of
the
ring
extraction-kicker
assembly.
extraction-kicker
assembly.
——— neutralization factor within the beam radius
extraction-kicker
assembly.
extraction-kicker
assembly.
0.25
next-generation, high-poweraccelerator
acceleratorfacility.
facility.
next-generation,
next-generation,high-power
high-poweraccelerator
acceleratorfacility.
facility.
next-generation,
high-power
I am
indebted
tomy
mycolleagues,
colleagues,especially
especiallythose
thoseparparI Iam
indebted
to
amindebted
indebtedtotomy
mycolleagues,
colleagues,especially
especiallythose
those parI
am
ticipating
SNS
accelerator-physicsdiscussions.
discussions. participating
in in
SNS
accelerator-physics
ticipating in
inSNS
SNSaccelerator-physics
accelerator-physicsdiscussions.
discussions.
ticipating
0.20 _ beam current (a.u.) . .
0.15
REFERENCES
REFERENCES
REFERENCES
REFERENCES
0.10
0.05
0.00
1900 2000 2100 2200 2300 240O 25OO 26OO 27OO
Time (ns)
Source:
M. Pivi,
M. Furman
Source:
DataData
fromfrom
M. Pivi,
M. Furman
Source: Data from
Pivi, M. Furman
Furman
from M. Pivi,
FIGURE
Simulation
of single-bunch
electron
multipactFIGURE
8. 8.Simulation
of single-bunch
electron
multipacting with
secondary-emission
yield
of 2.
FIGURE
8.
Simulation
of single-bunch
multipactFIGURE
8.a peak
Simulation
multipacting
with
a peak
secondary-emission
yield
of electron
2.
ing with
with aa peak
peak secondary-emission
secondary-emission yield of 2.
ing
Main
challenges
include
meeting
target
Main
ringring
challenges
include
meeting
thethe
target
re-requirements
on the
peak
current
density,
minimizing
unMain
ring
challenges
include
meeting
the targetunreMain
ring
quirements
onchallenges
the
peak
current
density,
minimizing
controlled
beam
loss,
and
controlling
collective
effects
quirements
on
the
peak
density,
minimizing
unquirements
on
the
peak
current
uncontrolled beam loss, and controlling collective effects
(space charge,
instabilities,
electron cloud
(Fig.effects
8)) [4].
controlled
beam
loss,
collective
controlled
beam
loss,
and controlling
effects
(space
charge,
instabilities,
electron cloud
(Fig. 8))
[4].
Efforts
are made
to minimize
leading
sources
of
beam(space
charge,
instabilities,
cloud
(Fig.
8))
[4].
(space
charge,
instabilities,
electron
8))
[4].
Efforts are made to minimize leading sources of beamcoupling
impedance
(Fig. 9 [12]),
andsources
to enhance
Landau
Efforts
are
made
of
beamEfforts
are
made
to
minimize
leading
beamcoupling impedance (Fig. 9 [12]), and to enhance Landau
damping
[4].
coupling
impedance
coupling
impedance
(Fig. 9 [12]), and to enhance Landau
Landau
damping
[4].
High-performance
beam diagnostics is needed to acdamping
[4].
damping
[4].
High-performance
beam diagnostics is needed to accommodate the large
variation
of beam
and
High-performance
beam
diagnostics
is parameters,
needed to
acHigh-performance
toand
accommodate
the large variation
of beam parameters,
for
machine
protection
across
the
entire
facility.
Lasercommodate
the
large
variation
of
beam
parameters,
and
commodate
large variation
of beam
and
for
machine the
protection
across the
entire parameters,
facility. Lasermonitors
are under
test for
implementation
forwire
machine
protection
across
thepossible
entireimplementation
facility.
Laserfor
machine
protection
across
the
entire
facility.
Laserwire
monitors
are
under
test
for
possible
in monitors
the SRF linac
for a clean
operation,
and luminescence
wire
are
under
testoperation,
for
possible
wire
monitors
test
for
possible
implementation
in
the
SRF monitors
linacare
forunder
aare
clean
andimplementation
luminescence
profile
under
test
to
reduce
space-charge
in
the
SRF
linac
for
a
clean
operation,
and
luminescence
in
the
SRF
linac
for
a
clean
operation,
and
luminescence
profile
monitors
are
under
test
to
reduce
space-charge
and
electron-cloud
complications
in
the
ring.
profile
monitors are
are
under test
test to
to
reduce
profile
monitors
under
reduce
space-charge
and
electron-cloud
complications
in the
ring.space-charge
and electron-cloud
electron-cloud complications
complications
in
and
in the
the ring.
ring.
SUMMARY
SUMMARY
SUMMARY
By adopting superconducting
RF technology for the
SUMMARY
By linac
adopting
superconducting
RF
technology
thedeand
by
fully
optimizing
the
accumulatorfor
ring
By adopting
adopting
superconducting
RFaccumulator
technology
for
the
By
superconducting
RF
technology
for
thetolinac
andthe
bySpallation
fully
optimizing
ring
design,
Neutronthe
Source
project, half
way
linacthe
andSpallation
by fully
fully optimizing
optimizing
the accumulator
ring
delinac
and
by
the
accumulator
ring
sign,
Neutron
Source
project,
half
way
towards its completion, is meeting the challenge to debe a
sign, the
the
Spallation
Neutron
Source
project,
half
way
wards
its Spallation
completion,
is meeting
theproject,
challenge
be toa
sign,
Neutron
Source
halfto
way
towards its
its completion,
completion, is
is meeting
meeting the
wards
the challenge
challenge to
to be
be aa
42
Holtkamp,
EPAC
(2002)
1. 1.N.N.
Holtkamp,
EPAC
(2002)
N.
Holtkamp,
EPACet(2002)
(2002)
N.
Holtkamp,
EPAC
N.
Catalan-Lasheras
et
SNS
NotesSNS/AP/7
SNS/AP/7(2001)
(2001)
2.1.1.2.N.
Catalan-Lasheras
al,al,
SNS
Notes
2.
N.
Catalan-Lasheras
et
al,
SNS
Notes
SNS/AP/7
2.
N.
Catalan-Lasheras
et
al,
SNS
Notes
SNS/AP/7
(2001)
3.
J.
Wei
et
al,
Phys.
Rev.
ST-AB
3,
080101
(2000)
3. J. Wei et al, Phys. Rev. ST-AB 3, 080101 (2000) (2001)
3.4.J.J.J.Wei
Wei
etet
al,
Phys.
Rev.
ST-AB
080101
(2000)
al,
Phys.
Rev.
ST-AB
3,3,
080101
(2000)
J.Wei
Wei
al,
EPAC’00,
981;
PAC
(2001)
319;
EPAC(2002)
(2002)
4.3.
etet
al,
EPAC’00,
981;
PAC
(2001)
319;
EPAC
4.5.J.J.J.Stovall
Wei
etet
al,et
EPAC'OO,
981;
PAC
(2001)319;
319;EPAC
EPAC(2002)
(2002)
J.Wei
Stovall
al,
LINAC
2000,
p.605
et
al,
EPAC’00,
981;
PAC
(2001)
5.4.
al,
LINAC
2000,
p.605
Stovall
al,SNS
LINAC
2000,p.605
p.605
Jeon
104050000-TD0010-R01
(2002)
J.J.D.
Stovall
etetal,
al,
LINAC
2000,
6.5.6.D.
Jeon
et et
al,
SNS
104050000-TD0010-R01
(2002)
D.
Raparia
al,
Workshop
HB2002,FNAL
FNAL(2002)
(2002)
D.
Jeon
al,et
SNS
104050000-TD0010-R01
(2002)
D.
Jeon
etetet
al,
SNS
104050000-TD0010-R01
7.6.7.D.
Raparia
al,
Workshop
HB2002,
8.
I.
Hofmann,
et
al,
PAC
(2001)
2902
D.
Raparia
et
al,
Workshop
HB2002,
FNAL
(2002)
7.
D.
Raparia
et
al,
Workshop
HB2002,
FNAL
(2002)
8. I. Hofmann, et al, PAC (2001) 2902
Sundelin
al
PAC’01
1984;
D.Jeon
Jeonetetal,al,PAC’01
PAC’012063
2063
I.I.R.
Hofmann,
PAC
2902
Hofmann,
al,
PAC(2001)
(2001)
2902
9.8.9.R.
Sundelin
etetet
alal,
PAC’01
1984;
D.
10.
S.
Kim,
et
al,
PAC
(2001)
1128
9.
R.
Sundelin
et
al
PAC’01
1984;
D.
Jeon
et
al,
PAC’01
2063
R.
Sundelin
et
al
PAC'Ol
1984;
D.
Jeon
et
al,
PAC'Ol
2063
10. S. Kim, et al, PAC (2001) 1128
11.P.S.
P.Kim,
Wanderer
et
al,
EPAC
(2002)
10.
etetal,
(2001)
1128
10.
S.Wanderer
Kirn,
al,PAC
PAC
(2001)
1128
11.
et
al,
EPAC
(2002)
12.P.
D.
Davino et
(2002)
11.
Wanderer
etetal,
al,
EPAC
(2002)
11.
Wanderer
al,EPAC
EPAC
(2002)
12.
D.P.Davino
et al,
EPAC
(2002)
12.
12. D.
D.Davino
Davinoetetal,
al,EPAC
EPAC(2002)
(2002)