Slip
Slip Stacking
Stacking
Kiyomi
Kiyomi Koba
Koba and
andJim
JimSteimel
Steimel
Fermi National
National Accelerator
USA
Fermi
AcceleratorLaboratory,
Laboratory,Batavia,
Batavia,ILIL60510,
60510,
USA
Abstract. We
We have
have started
started beam
increase
proton
Abstract.
beam studies
studies for
for ‘slip
'slip stacking’[1]
stacking'[1]ininthe
theMain
MainInjector
Injectorininorder
orderto to
increase
proton
intensity
on
a
target
for
anti-proton
production.
It
has
been
verified
that
the
system
for
slip
stacking
is
working
intensity on a target for anti-proton production. It has been verified that the system for slip stacking is working
with low
low intensity
intensity beam.
beam. For
For aa high
feedforward
with
high intensity
intensity operation,
operation, we
weare
aredeveloping
developinga afeedback[2][3]
feedback[2][3]andand
feedforward
system.
system.
INTRODUCTION
INTRODUCTION
Fermilab Main Injector (MI) accelerates protons
MaintoInjector
accelerates
protons
and Fermilab
extracts them
a target(MI)
to produce
anti-protons.
and
extracts
them
to
a
target
to
produce
anti-protons.
In the operation cycle for the anti-proton production,
In bunches
the operation
cycle for
theBooster
anti-proton
production,
84
are injected
from
to MI.
Bunches
84
bunches
are
injected
from
Booster
to
MI.
are accelerated from 8GeV to 120GeV and Bunches
extracted
from target.
8GeV toFigure
120GeV
and extracted
toarehitaccelerated
the production
1 shows
a typical
to
hit
the
production
target.
Figure
1
shows
typical
operation cycle. The total beam intensity isa 4.5*
1012
operation cycle. The total beam intensity is 4.5*1012
particles per pulse (ppp) with a cycle of 1.5 sec. Main
particles per pulse (ppp) with a cycle of 1.5 sec. Main
rf
parameters
are
listed
in
Table
1.
rf
parameters
are
listed
in
Table
1.
Higher proton intensities are needed to get more
Higher proton intensities are needed to get more
anti-protons
and higher stacking rates in the
anti-protons and higher stacking rates in the
accumulator.
In
achieve this,
this, we
weare
areplanning
planning
accumulator. In order
order to
to achieve
toto use
a
scheme
called
"slip
stacking".
With
slip
use a scheme called “slip stacking”. With slip
stacking,
can be
be doubled
doubled
stacking, the
the intensity
intensity of
of the
the bunches
bunches can
by
at lightly
lightly lower
lower energy,
energy,
by injecting
injecting one
one bunch
bunch train
train at
another
train
at
lightly
higher
energy
and
bringing
another train at lightly higher energy and bringing
them
together.
them
together.
Since
different energies,
energies, MI
MI
Since two
two bunch
bunch trains
trains have
have different
must
aperture to
to accept
accept
must have
have an
an enough
enough momentum
momentum aperture
both.
of MI
MI isis +/-0.7%
+/-0.7% atat
both. The
The momentum
momentum aperture of
injection,
for each
each bunch
bunchtrain
train
injection, that
that is,
is, the rf frequency for
can
the original
original value.
value.
can be
be shifted
shifted by
by +/-3000Hz from the
TABLE 1. MI RF Parameters.
TABLE
1. MI RF
Parameters.
Rf frequency©
njection
52.811400MHz
RfNumber
frequency@njection
52.811400MHz
of rf cavities
18
Number
of rftotal
cavities
18
Maximum
rf voltage
1.6MV
Maximum
total rf voltage
1.6MV
@ injection
@Proton
injection
energy @ injection
8GeV
Proton
energy
@injection
8GeV
Harmonic
number
588
Harmonic
588
Transitionnumber
gamma
21.6
Transition
gamma
21.6
528.3m
Mean radius
Mean radius
528.3m
STACKINGPROCEDURES
PROCEDURES
AND
STACKING
AND
MEASUREMENT
RESULTS
MEASUREMENT RESULTS
Forslip
slipstacking,
stacking,wewe
three
different
rf systems
For
useuse
three
different
rf systems
and
follow
four
steps.
Step
1:
the
first
bunch
train
and follow four steps. Step 1: the first bunch train
is is
injected
from
Booster
on
the
central
orbit
and
captured
injected from Booster on the central orbit and captured
bythe
thefirst
firstrfrfsystem.
system.ToTomake
make
a room
second
by
a room
forfor
thethe
second
bunch
train,
the
first
bunch
train
is
then
decelerated
bunch train, the first bunch train is then decelerated
untilititcirculates
circulatesinside
insideofof
central
orbit.Step
Step
2: the
until
thethe
central
orbit.
2: the
secondbunch
bunchtrain
trainis isinjected
injectedonon
central
orbit
second
thethe
central
orbit
andand
capturedbybythe
thesecond
secondrf rfsystem.
system.Step3:
StepS:
bunch
captured
AsAs
bunch
trainshave
havetwo
twoslightly
slightlydifferent
differentenergies,
energies,they
they
trains
cancan
moverelative
relativetotoeach
eachother
otheras as
they
accelerated.Step
Step
move
they
accelerated.
when two
twobunch
bunchtrains
trainscoincide
coincideat atthethesame
same
4:4: when
longitudinal
areare
captured
by by
thethe
third
rf rf
longitudinallocation,
location,they
they
captured
third
system.
system.
Optimum
OptimumFor
ForFrequency
FrequencySeparation
Separation
it; it
FIGURE 1.
Typical operation cycle for anti-proton
FIGURE
1. Typical
operation cycle for anti-proton
production. The green trace and red trace show a total
production.
The
green
trace
and red trace show a total
intensity and momentum respectively.
intensity and momentum respectively.
Since
stacking,
Sincethere
therearearetwo
tworf rfvoltages
voltagesduring
during
stacking,
they
act
on
both
bunch
trains.
The
bunch
shape
hashas
they act on both bunch trains. The bunch shape
been
been measured
measuredtotodemonstrate
demonstratethat
thatthethefrequency
frequency
separation
thethe
second
rf systems
separationbetween
betweenthethefirst
firstand
and
second
rf systems
isisadequate.
adequate.
InInthis
bunch
train
waswas
injected,
thismeasurement,
measurement,one
one
bunch
train
injected,
then
two
rf
voltages
were
raised
and
the
frequency
then two rf voltages were raised and the frequency
separation
to to
1200Hz.
separationwas
wasincreased
increasedfrom
from400
400
1200Hz.Figure
Figure
2 shows the bunch shape at injection and at 150msec
2 shows the bunch shape at injection and at 150msec
after injection.
The signals, plotted here with
after injection.
The signals, plotted here with
5nsec/div, were obtained with a wall current monitor.
5nsec/div, were obtained with a wall current monitor.
It is clear that the frequency separation of 1200Hz is
It is clear that the frequency separation of 1200Hz is
enough to keep the bunch shape unchanged.
enough to keep the bunch shape unchanged.
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
223
r^=a
1
E^
the wall
wall current
currentmonitor,
monitor,reveals
revealsthetheprogress
progressof ofslipslip
the
the
wall
current
monitor,
reveals
the
progress
slip
stacking
from
the
beginning
to
the
end.
the
wall
current
monitor,
reveals
progress
ofofslip
stacking from the beginning to the end.
stacking
from
stacking
fromthe
thebeginning
beginning to
to the
the end.
end.
| ,/'••-
lf
'••.
A
:' \
s
^^^
;
x' 1
v
X
«
X
... j
X,---v
„i
dt»» 860H2
__^
FIGURE
2. Comparison
of the bunch
which
FIGURE
Comparison
of the
the bunch
bunch
shape,
which
isis is
FIGURE
2.2. Comparison
of
shape,shape,
which
from
a
wall
current
monitor,
at
injection
and
FIGURE
2.
Comparison
of
the
bunch
shape,
which
isafter
from aa wall
wall current
current monitor,
monitor, at
at injection
injection and
from
andat after
after
The bunch
on upper
left atwas
from a150msec.
wall
current
monitor,
at injection
and injection.
after
150msec.
The
bunch
on
upper
left
was
injection.
150msec.Other
The bunches
bunch on
upper
left 2msec.
was at The
injection.
at 2msec.
after
frequency
150msec.
Thewere
bunchatwere
on
upper
left was
atfrequency
injection.
Other
bunches
after
The
Other
bunches
were
at
after
2msec.
The
frequency
separation
was
400Hz(upper
middle),
600Hz(upper
Other bunches
were at after middle),
2msec. The
frequency
separation
was 400Hz(upper
400Hz(upper
600Hz(upper
separation
was
middle),
600Hz(upper
right),
left),
1000Hz(lower
middle),
separation
was800Hz(Lower
400Hz(upper
middle),
600Hz(upper
right),
800Hz(Lower
left),
1000Hz(lower
middle),
1200Hz(lower right).
right),
800Hz(Lower
left),
1000Hz(lower
middle),
right), 800Hz(Lower
left), 1000Hz(lower middle),
1200Hz(lower
right).
1200Hz(lower
right).
1200Hz(lower
right).
Simulation
FIGURE 3. Frequency (upper) and voltage (lower) as
FIGURE
3.3. Frequency
(upper)
and
voltage
(lower)
as as
FIGURE
Frequency
(upper)
and
voltage
(lower)
a function
of
time. Red
lines are
for
the first
rf andas
FIGURE
3. time.
Frequency
(upper)
and
voltage
(lower)
a function
of
Red
lines
are
for
the
first
rf
and
blue line of
aretime.
for the second
rf. are for the first rf and
aa function
Red
lines
function
Red rf.
lines
are for the first rf and
blue
line are of
for time.
the second
blue
line
are
for
the
second
rf.
blue line are for the second rf.
Simulation
Simulation
A simulation
study of the slip stacking has been
Simulation
A simulation
the slip stacking
has been Two
curried outstudy
with of
a simulation
code, ESME[4].
A Asimulation
study
ofofthe
has
been
simulation
theslip
slipstacking
stacking
hasTwo
been
curried
out
with each
a study
simulation
code,
ESME[4].
bunches
with
emittance
of 0.1 eV-sec
were
curried
out
with
a
simulation
code,
ESME[4].
Two
curried
out
with
a
simulation
code,
ESME[4].
Two
injected.
An
area
corresponding
to
one
bucket
area of
bunches each with emittance of 0.1 eV-sec were
bunches
each
with
emittance
of
0.1
eV-sec
were
was
kept
between
and
higher
energy
bunches0.1eV-sec
each
with
emittance
0.1bucket
eV-sec
were
injected.
An
area
corresponding
tooflower
one
area
of
injected.
An
area
corresponding
to
one
bucket
area
bunches.
They
then followed
the higher
four
stacking
steps
injected.
An area
corresponding
one
bucket
areaof
of
0.1eV-sec
was
kept
between
lowertoand
energy
O.leV-sec
was
kept
between
lower
and
higher
energy
and
were
captured
by
the four
third
rf system
0.1eV-sec
was
kept
between
lower
andstacking
higher
energy
bunches.
They
then
followed
the
stepswhich
provided
athen
bucket
area third
of
0.3four
eV-sec.
Atwhich
the
end of
bunches.
They
followed
the
stacking
steps
bunches.
Theythen
the
four
stacking
steps
and
were
captured
byfollowed
the
rf
system
these
processes,
we
found
a single
bunch
emittance
and
byby
third
rfrf system
which
andwere
were
captured
the
third
system
which
provided
acaptured
bucket
area
ofthe
0.3
eV-sec.
At
the of
end
of
0.3
without
beam
loss, indicating
anend
emittance
provided
a bucket
ofof0.3
eV-sec.
ofof
provided
aeV-sec
bucket
area
eV-sec.
At
the
end
these
processes,
wearea
found
a 0.3
single
bunchAt
of the
emittance
growth
of
50
%.
these processes, we found a single bunch of emittance
these
processes,
found
a single
bunchanofemittance
emittance
0.3
eV-sec
withoutwebeam
loss,
indicating
0.3
eV-sec
0.3
eV-sec
beamloss,
loss,indicating
indicatingananemittance
emittance
growth
of without
50without
%. beam
Measurement
growth
50%.
growthof of
50 %.
The frequency
and voltage as a function of time are
Measurement
shown in Measurement
Figure
3 for the first and the second rf
Measurement
The frequency
and18voltage
functionone
of time
are for
systems. Of
cavitiesasata53MHz,
was used
shown
in
Figure
3and
for
the first
and
the of
second
rf
The
frequency
voltage
function
ofsystem.
timeare
areThe
the
first system,
another
for
the
second
The
frequency
and
voltage
asas
a afunction
time
separation
was
kept
at the
1200Hz.
showninfrequency
inOf
Figure
first
and
the second
second
systems.
18 cavities
atthe
53MHz,
one
was
usedThe
forrfrffirst
shown
Figure
3 3forforthe
first
and
train
was at
injected
on one
the
central
orbit
the
first bunch
system,
another
for
the second
The
systems.
cavities
at53MHz,
53MHz,
onesystem.
wasused
used
forwith
systems.
OfOf1818
cavities
was
for
nominal
frequency
at
0.13msec
and
captured
by
frequency
separation
was
kept
at
1200Hz.
The
first
the
first
system,
another
for
the
second
system.
The
the first system, another for the second system. The the
first separation
rf
system
of 62kV.
The
intensity
bunch
train
was
injected
on
central
orbit
with
frequency
was
keptthe
1200Hz.
Thewas
firstlow,
frequency
separation
was
kept
atat1200Hz.
The
first
12
0.4*10
ppp, injected
in 0.13msec
orderonto the
reduce
the orbit
beam
loading
nominal
frequency
at
and
captured
by
the
bunch
train
was
central
with
bunch train
was Atinjected
onthe
thefrequency
central oforbit
with rf
effects.
this
time,
the
second
first
rf frequency
system
of at62kV.
Theand
intensity
wasby
low,
nominal
frequency
at0.13msec
0.13msec
and
captured
by
the
nominal
captured
the
12 system was 1200Hz higher than the first rf system.
0.4*10
ppp,
in oforder
to reduce
the
beamwas
loading
firstrf rf
system
of62kV.
62kV.
The intensity
intensity
was
low,
first
system
The
low,
The first bunch was then decelerated to the frequency
12 12ppp,
effects.
At this
time,
frequency
of beam
the
second
rf
0.4*10
orderthe
reduce
the
beam
loading
0.4*10
ppp,
ininorder
totoreduce
the
loading
1200Hz
lower
than
the
original
value.
After one
system
was
1200Hz
higher
than
the
first
rf
system.
effects.
At
this
time,
the
frequency
of
the
second
cycle the
of 66.7msec,
second
trainrfrfwas
effects. Booster
At this time,
frequency the
of the
second
The
first
bunch
was
then
decelerated
to
the
frequency
FIGURE 4. Measurement results of the whole slip
system
was
1200Hz
higher
than
the
first
rf
system.
injected
on the
central
by the
system was
1200Hz
higher
thanorbit
the and
first captured
rf system.
stacking cycle. The lower figure shows the expanded
1200Hz
lower
than
the
original
value.
After
onetrains
The
first
bunch
was
then
decelerated
to
the
frequency
second
rf
system.
After
slipping,
both
bunch
The first bunch was then decelerated to the frequency
signal at just after recapture by the third rf.
Booster
cycle
ofthan
66.7msec,
the second
train
1200Hzwere
lower
theoriginal
original
value. all
After
one
captured
by
the
third
system,
18 was
cavities.
1200Hz
lower
than
the
value.
After
one
injected
on
the
orbit
and
captured
by the
BoosterFigure
cycle
66.7msec,
the
second
was
4, of
acentral
mountain
range
picture
of thetrain
signals
from FIGURE 4. Measurement results of the whole slip
Booster cycle of 66.7msec, the second train was
second
system.
After slipping,
bunch by
trains
injectedrf on
the central
orbit andboth
captured
the
injected
on the bycentral
orbitsystem,
and captured
by the
were
captured
thirdslipping,
all 18
cavities.
second
rf system. theAfter
both
bunch
trains
second
rf4, system.
After
slipping,
both
bunch trains
Figure
a mountain
range
picture
of
the
from
were captured
by the
third
system,
allsignals
18 cavities.
were captured by the third system, all 18 cavities.
Figure 4, a mountain range picture of the signals from
Figure 4, a mountain range picture of the signals from
stacking
cycle.
The lower figure
shows
expanded
FIGURE
4. Measurement
results
of the
FIGURE
4.
Measurement
results
ofrf.the
the whole
whole slip
slip
signal
at just
after
recapture
the third
stacking
cycle.
The
lowerby
figure
shows
the expanded
stacking cycle. The lower figure shows the expanded
signal at just after recapture by the third rf.
signal at just after recapture by the third rf .
224
BEAM
LOADING
COMPENSATION
BEAM LOADING
LOADING COMPENSATION
COMPENSATION
BEAM
For
high
intensity
operation,
beam loading
loading
For high
high intensity
intensity operation,
operation, beam
For
compensation
is
the
key
issue.
A
simulation
was
compensation isis the
the key
key issue.
issue. A
A simulation was
compensation
carried
out
to
compare
the
beam
with
and
without
carried out
out to
to compare
compare the
the beam
beam with
with and without
carried
compensation
of
beam
loading
effects.
Two trains
of
compensation of
of beam
beam loading
loading effects.
effects. Two
trains
of
compensation
12
12 went
84
bunches
with
the
total
intensity
of
5*10
12
84
bunches
with
the
total
intensity
of
5*10
went
84 bunches with the total intensity
through
the
slip
stacking
process.
through the
the slip
slip stacking
stacking process.
process. At
At the
the end
end of
of
through
stacking,
with
no
compensation,
the
bunch shape
shape
stacking, with
with no
no compensation,
compensation, the
the bunch
stacking,
became
distorted
and
many
particles
were
left outside
outside
became distorted
distorted and
and many
many particles
particles were
were left
became
of
the
bucket.
When
there
was
compensation
of
40 dB
dB
ofthe
thebucket.
bucket. When
When there
there was
was compensation
compensation of 40
of
of
loop
gain,
there
were
no
particle
losses.
ofloop
loopgain,
gain,there
there were
were no
no particle
particle losses.
losses.
of
FIGURE
6.
Measurement
of
the
beam
spectra
from
FIGURE
FIGURE 6.
6. Measurement
Measurement of
of the
the beam
beam spectra
spectrafrom
fromaaa
gap
voltage
monitor.
On
the
left
graph,
the
green
and
gap
gap voltage
voltage monitor.
monitor. On
On the
the left
left graph,
graph, the
thegreen
greenand
and
red
traces
show
without
and
with
feedback
system,
red
red traces
traces show
show without
without and
and with
with feedback
feedback system,
system,
respectively.
On the
the right graph,
graph, the blue
blue and green
green
respectively.
respectively. On
On the right
right graph, the
the blue and
and green
traces show
show without
without and
and with
with feedforward
feedforward system,
system,
traces
traces show without and with feedforward system,
respectively.
respectively.
respectively.
SUMMARY
SUMMARY
In order to achieve an increase in
in proton intensity,
intensity,
In order to achieve an increase in proton
proton intensity,
Fermilab
Main
Injector
will
use
a
stacking
Fermilab Main Injector will use a stacking process
process
"slip stacking”.
stacking".
be doubled
doubled
called
called “slip
“slip stacking”. The
The intensity
intensity will
will be
be doubled
at aa slightly
slightly lower
by
by injecting
injecting one
one train
train of
of bunches
bunches at
at a slightly lower
and lower
energy,
another
at
a
slightly
higher
energy
and
then
energy,
another
at
a
slightly
higher
energy
then
brining them together for the final capture.andBeam
brining them together for the final capture. Beam
studies have stared for this process and we have
studies
have stared
this for
process
and intensity,
we have
already verified
that, for
at least
low beam
intensity,
already verified
that,
at
least
for
low
beam
intensity,
higher
the stacking procedure works as expected. For higher
the stacking procedure works as expected.
higher
of the
theFor
intensity operation, development work of
feedback
intensity
operation,
development
work
of
the
feedback
feedforward system is under way.
and feedforward
and feedforward system is under way.
FIGURE
5. Simulation results comparing the beam
FIGURE 5.
FIGURE
Simulation results comparing the beam
after
the
slip
stacking
with (right) and without (left)
after
the
slip
stacking with
(left) aa
after the slip stacking
(right) and without (left)
compensation of beam loading
effects.
effects.
compensation of beam loading effects.
Development Of Beam Loading
Development Of Beam Loading
Compensation
Compensation
We are
are developing feedback
feedback and feedforward
feedforward
We are developing
feedback and feedforward
systems
systems for beam
beam loading
loading compensation. Figure 6, in
systems
for beam loading compensation.
Figure 6, in
which
which the
the center frequency
frequency is the fundamental,
fundamental, shows
which
the
center
frequency
is
the
fundamental,
frequency
frequency spectra from
from a gap voltage monitor.shows
The
frequency
spectra
from
a gap
voltage
monitor.
The
gain
gain for feedback
feedback system
system was
was -20dB.
–20dB. One
One can
can see
see
gain
was –20dB.
can see
from
thisfeedback
that the
feedforward
system One
reduces
the
fromfor
thesystem
feedforward
from
this that
the feedforward
system reduces the
revolution
sidebands
by 20dB.
revolution
revolution sidebands by 20dB.
REFERENCES
REFERENCES
1.
1.
S. Shukla
Shukla et
et al,
al, “Slip
"Slip Stacking
Stacking in
in the
the Fermilab
Fermilab Main
Main
S. Shukla Snowmass’96,
et al, “Slip Stacking
in the Fermilab Main
Injector.",
Snowmass'96,
1996.
Injector.”,
June 1996.
Injector.”, Snowmass’96, June 1996.
"Improving the linearity of
2. J.Dey and J.Steimel, “Improving
2. ferrite
J.Dey and
J.Steimel,
the linearity
of
feedback",
PAC2001
loaded
cavities“Improving
using feedback”,
PAC2001
ferrite loaded cavities using feedback”, PAC2001
"Narrowband Beam
Beam
3. J.Dey, J. Steimel and J. Reid, “Narrowband
3. Loading
J.Dey, J.Compensation
Steimel and inJ. the
Reid,
“Narrowband
Beam
Fermilab
Main Injector
Loading Compensation
in the Fermilab Main Injector
Cavities", PAC2001
Accelerating
Cavities”,
Accelerating Cavities”, PAC2001
"Users Guide to ESME”,
ESME", 2000.
4. J. MacLachlan, “Users
4. J. MacLachlan, “Users Guide to ESME”, 2000.
225