project x reference design overview
DESCRIPTION
Project X Reference Design Overview. Sergei Nagaitsev April 12, 2011. Reference Design. Reference Design Capabilities. 3 GeV CW superconducting H- linac with 1 mA average beam current. Flexible provision for variable beam structures to multiple users - PowerPoint PPT PresentationTRANSCRIPT
Project X Reference Design Overview
Sergei Nagaitsev
April 12, 2011
Reference Design
Apr 12, 2011 - S. Nagaitsev Page 2
Reference Design Capabilities
• 3 GeV CW superconducting H- linac with 1 mA average beam current.– Flexible provision for variable beam structures to multiple users
• CW at time scales >1 msec, 15% DF at <1 msec– Supports rare processes programs at 3 GeV– Provision for 1 GeV extraction for nuclear energy program
• 3-8 GeV pulsed linac capable of delivering 300 kW at 8 GeV – Supports the neutrino program– Establishes a path toward a muon based facility
• Upgrades to the Recycler and Main Injector to provide ≥ 2 MW to the neutrino production target at 60-120 GeV.
• Day one experiment to be incorporated utilizing the CW linac
Þ Utilization of a CW linac creates a facility that is unique in the world, with performance that cannot be matched in a synchrotron-based facility.
Apr 12, 2011 - S. Nagaitsev Page 3
Reference DesignProvisional Siting
Apr 12, 2011 - S. Nagaitsev Page 4
Pulsed Linac
CW Linac
Pulsed 3-8 GeV Linac based on ILC / XFEL technology
Apr 12, 2011 - S. Nagaitsev Page 5
Reference design: accelerator scope
• Warm cw front end 162.5 MHz, 5 mA (H- ion source, RFQ, MEBT, chopper)
• 3-GeV cw SCRF linac (325, 650 MHz), 1-mA ave. beam current
• Transverse beam splitter for 3-GeV experiments
• 3-8 GeV: pulsed linac (5% duty cycle), 1.3 GHz
• Recycler and MI upgrades
• Various beam transport lines
5% du
ty cy
cle
Pulsed dipole
Changes to the reference design since Sep 2010
• Eliminated a 1.3-GHz section from the CW linac (Nov 2010)– Larger aperture, lower stripping losses, fewer cavities, cost neutral
• Changed the RFQ frequency to 162.5 MHz (Jan 2011)– Easier chopper, lower RFQ power, larger aperture
• Introduced a LEBT chopper (Nov 2010)
• Ion source operates at a constant current, 5 mA max (Nov 2010)
• Limit the max. cryo loss per cryomodule to 250 W (2K) (Nov 2010)– Per cavity – 25 W
• Established pulsed 1.3 GHz linac configuration (Nov 2010)– 25 MV/m gradient– 1 mA beam current in 4.3 ms pulses– 1 Klystron per 16 cavities (2 CM)– 8 ms, 10 Hz
Apr 12, 2011 - S. Nagaitsev Page 6
Apr 12, 2011 - S. Nagaitsev Page 7
Linac beam current
• Linac beam current has a periodic time structure (at 10 Hz) with two major components.
0 2 4 6 8 10
Pulseddipole
OFF
ON
Beamcurrent, mA
0
1
Time, ms
4.3 ms flattop
Choppingfor injection
Choppingfor 3-GeV program
100
0
20
40
60
80
100
120
140
160
180
200
-10-8-6-4-20246810
Matching
DC switch dipole to Experimental Area
Pulsed switch dipole to RCS
DC dipole
ARC ARC
DC dipole
Linac Dump
Momentum Dump
Momentum Dump
Buried Beam Pipe
CollimationCollimation
Beam to Recycler
Chopping for injection
• RF frequency at injection into the Recycler : ~50 MHz
• Chopper needs to provide a kicker gap (~200 ns per 11 µs) and needs to remove bunches that fall into “wrong” phase of ring rf voltage.
– 50% of bunches are removed ( ion source at 2 mA)– 80% of bunches are removed ( ion source at 5 mA)
Apr 12, 2011 - S. Nagaitsev Page 80 1 2 3 4 5 6 7 8 9 102-
1-
0
1
2
f t( )
g t( )
h t( )
t
Recycler RF
162 MHz bunches to be removed
Apr 12, 2011 - S. Nagaitsev Page 9
Chopping and splitting for 3-GeV experiments
1 msec period at 3 GeVMuon pulses (16e7) 81.25 MHz, 100 nsec at 1 MHz 700 kWKaon pulses (16e7) 20.3 MHz 1540 kWNuclear pulses (16e7) 10.15 MHz 770 kW
Separation scheme
Ion source and RFQ operate at 4.2 mA75% of bunches are chopped at 2.5 MeV after RFQ
Transverse rf splitter0
2
4
6
8
10
12
14
16
18
0.0 0.1 0.3 0.4 0.5 0.6 0.7 0.9 1.0 1.1 1.2 1.4 1.5 1.6 1.7 1.9 2.0
Num
ber
of io
ns p
er b
unch
, (e7
)
Time, us
Beam after splitter
Apr 12, 2011 - S. Nagaitsev Page 10
0
2
4
6
8
10
12
14
16
18
0.0 0.1 0.3 0.4 0.5 0.6 0.7 0.9 1.0 1.1 1.2 1.4 1.5 1.6 1.7 1.9 2.0
Num
ber
of io
ns p
er b
unch
, (e7
)
Time, us
0
2
4
6
8
10
12
14
16
18
0.0 0.1 0.3 0.4 0.5 0.6 0.7 0.9 1.0 1.1 1.2 1.4 1.5 1.6 1.7 1.9 2.0N
umbe
r of
ions
per
bun
ch, (
e7)
Time, us
0
2
4
6
8
10
12
14
16
18
0.0 0.1 0.3 0.4 0.5 0.6 0.7 0.9 1.0 1.1 1.2 1.4 1.5 1.6 1.7 1.9 2.0
Num
ber
of io
ns p
er b
unch
, (e7
)
Time, us
10 MHz bunches
20 MHz bunches
1 MHz pulses
Apr 12, 2011 - S. Nagaitsev Page 11
Ion source: TRIUMF-typeH- DC ion source
Delivery expected: May 2011LBNL + FNAL will test emittance etc.
Apr 12, 2011 - S. Nagaitsev Page 12
Proposed LEBT configuration
• LEBT will likely have a chopper to provide 0.5 ms gaps, needed for a switching magnet at 3 GeV
RFQ (162.5 MHz)
CW RFQ:
Several design
studies by LBNL
Good experience at ANL and LBNL
Apr 12, 2011 - S. Nagaitsev Page 13
Apr 12, 2011 - S. Nagaitsev Page 14
MEBT design: 5 mA at 162.5 MHz beam
=0.25 ∙μm; z,n=0.3 ∙μm
325 MHz warm bunching cavities
MEBT chopper
Apr 12, 2011 - S. Nagaitsev Page 15
RT Bunching CW cavities
Apr 12, 2011 - S. Nagaitsev Page 16
Effect of longitudinal emittance (L/=0.52)
Apr 12, 2011 - S. Nagaitsev Page 17
Bunch Length (3) vs. Synch. Phase
SSR0
MEBT
Apr 12, 2011 - S. Nagaitsev Page 18
Prelim Conclusions: Front End
• Transverse rms (norm) emittance 0.25 mm-mrad
• Range of acceptable longitudinal emittances is close to equipartition, z/ ≈ 0.8-1.2, =0.25 ∙μm
• No need for using 162.5 MHz cavity (copper) in MEBT.
• SSR0 section at 325 MHz works for 162.5 MHz RFQ option
• Unanswered issues:– Chopper driver– Beam absorber (10 kW)– Absorber sputtering and blistering rates
SRF LinacTechnology Map
Apr 12, 2011 - S. Nagaitsev
b=0.11 b=0.22 b=0.4 b=0.61 b=0.9
325 MHz2.5-160 MeV
b=1.0
1.3 GHz3-8 GeV
650 MHz0.16-3 GeV
Section Freq Energy (MeV) Cav/mag/CM Type
SSR0 (bG=0.11) 325 2.5-10 18 /18/1 SSR, solenoid
SSR1 (bG=0.22) 325 10-42 20/20/ 2 SSR, solenoid
SSR2 (bG=0.4) 325 42-160 40/20/4 SSR, solenoid
LB 650 (bG=0.61) 650 160-460 36 /24/6 5-cell elliptical, doublet
HB 650 (bG=0.9) 650 460-3000 160/40/20 5-cell elliptical, doublet
ILC 1.3 (bG=1.0) 1300 3000-8000 224 /28 /28 9-cell elliptical, quad
CW Pulsed
Page 19
325 MHz spoke cavity families
20
cavity type βGFreq MHz
Uacc, max
MeVEmax
MV/mBmax
mTR/Q,
ΩG,Ω
*Q0,2K
109
Pmax,2K
WSSR0 β=0.114 325 0.6 32 39 108 50 6.5 0.5SSR1 β=0.215 325 1.47 28 43 242 84 11.0 0.8SSR2 β=0.42 325 3.34 32 60 292 109 13.0 2.9
Parameters of the single-spoke cavities
SSR0 – design,
prototyping
SSR1 – prototyping, testing
SSR2 - design
Apr 12, 2011 - S. Nagaitsev
21
Focusing Periods in SSR sections:
800 mm Focusing Period:SSR0: (sol+cav) = 610 mmSSR1: (sol+cav) = 800 mmSSR2: (sol+cav+cav+60 mm) = 1600 mm
Apr 12, 2011 - S. Nagaitsev
650 cavities
• 650 MHz, 5-cell cavity:– Similar length as for ILC-type cavity;– About the same maximal energy gain per cavity;– The same power requirements;
• Benefits compared to 1.3 GHz ILC-type cavity:– Higher accelerating efficiency smaller number of cavities and RF sources:– Beam dynamics
• 2-fold frequency jump instead of 4-fold easier transition• Smaller beam losses;• Less effect of cavity focusing (~1/ λ)
• Trade-offs: – more serious problem with microphonics, but still may be manageable;– Larger diameter (comp to 1.3), higher cost per cavity;– additional rf frequency -> infrastructure.
Page 22Apr 12, 2011 - S. Nagaitsev
650 MHz cavities
Parameter LE650 HE650β_geom 0.61 0.9R/Q Ohm 378 638G-factor, Ohm 191 255Max. Gain/cavity (on crest) MeV 11.7 19.3Acc. Gradient MV/m 16.6 18.7Max surf. electric field MV/m 37.5 37.3Max surf. magnetic field, mT 70 70Q0 @ 2°K 1010 1.5 2.0P2K max [W] 24 29
650 MHz: β=0.61 650 MHz: β=0.9
Page 23Apr 12, 2011 - S. Nagaitsev
Apr 12, 2011 - S. Nagaitsev 24
LE Cryomodules
11325.12
8878.49
CMs lengths are shown from the first cavity iris to the last cavity iris.
HE Cryomodules3238.09
3282.03
Apr 12, 2011 - S. Nagaitsev Page 25
650 MHz Cryomodule (Tesla Style-Stand Alone, 250 W @ 2K)
End PlateBeam
Vacuum VesselCold mass supports (2+1)
Power MC (8)
Apr 12, 2011 - S. Nagaitsev Page 26
Cavity string & 300mm pipe
CW linac: Envelopes
Apr 12, 2011 - S. Nagaitsev Page 27
Energy Gain per Cavity
Apr 12, 2011 - S. Nagaitsev Page 28
• Single cavity per power source
– Solid State, IOT
3 GeV CW LinacCryogenic Losses per Cavity
• ~42 kW cryogenic power at 4.5 K equivalent
Apr 12, 2011 - S. NagaitsevPage 29
Solenoid Magnetic Field
Apr 12, 2011 - S. Nagaitsev Page 30
Apr 12, 2011 - S. Nagaitsev Page 31
Quadrupole (doublet) Gradient
35 0 100
35 0
Doublet (FD)
Apr 12, 2011 - S. Nagaitsev Page 32
Errors and misalignments
• Misalignments ±1 mm for all elements (specification ±0.5 mm )• RF jitter of 0.5 ° x 0.5 % in the front-end & 1 ° x 1% RF jitter in the
high-energy part was implemented. • 100 seeds and 1 million macro-particles per seed. • 1 corrector, 1 BPM per solenoid/douplet/quad; BPM resolution=30 μm• Beam centroid is corrected to ±1mm; Emittance increase < 20%. • The uncorrected seeds predict losses above 100 W/m; corrected – no
losses
The red curves are the 100 seeds without correction and the blue curves are the 100 seeds corrected
Beam Centroid off-set Beam Transverse Emittance
Apr 12, 2011 - S. Nagaitsev Page 33
H- stripping Summary• Stripping on residual gas is < 0.1 W/m if pressure is better than 10-8
Torr - Assuming gas contains 50% H2, 25% O2, 25% N2
• Magnetic stripping is well below 0.1 W/m even for unrealistic 5 mm beam offset.
• Stripping on blackbody radiation is not an issue for SC linac.
• Intra-beam stripping is ~0.1 W/m
Apr 12, 2011 - S. Nagaitsev Page 34
3 – 8 GeV acceleration
• Pulsed linac based on the ILC technology– 1.3 GHz, 25 MV/m gradient, ≤5% duty cycle – considering 8-30 ms pulse length– ~250 cavities (28 ILC-type cryomodules) needed.– Simple FODO lattice– 1 Klystron per 2 CM’s
Apr 12, 2011 - S. Nagaitsev Page 35
Cavity Energy Gain
Eacc = 25 MV/m: L=1.038m
Apr 12, 2011 - S. Nagaitsev Page 36
Envelopes (sigma)
Apr 12, 2011 - S. Nagaitsev Page 37
Summary of studies for pulsed linac• Beam losses are smaller than for CW linac
- Intra-beam stripping is well below 0.1 W/m- Magnetic stripping is small for reasonable beam displacement (<20mm)
• LLRF control- Simulated for scheme with 1 klystron feed 2 CM- VS control at the level below~0.5 % and 0.5 deg (individual cavity error
~10% and 10 deg) allows to keep energy jitter at 8 GeV below 10 MeV. (needed for injection)
Injection
• Need to accumulate 26 mA-ms protons in the MI/Recycler
• For a stationary foil, a single 26-ms pulse would destroy the foil.
• 6 pulses at 10 Hz give enough time for radiative cooling between pulses
Apr 12, 2011 - S. Nagaitsev Page 38
Foil efficiency
Apr 12, 2011 - S. Nagaitsev Page 39
Recycler injection
Apr 12, 2011 - S. Nagaitsev Page 40
Alternative injection scheme
• A possibility to inject in a single 26-ms pulse is highly desirable.– Eliminates the need for Recycler (as an accumulator)– Potentially allows for a lower linac energy (6-7 GeV)– Requires long-pulse linac operation
Apr 12, 2011 - S. Nagaitsev Page 41
Recycler parameters
Apr 12, 2011 - S. Nagaitsev Page 42
Number of particles 1.7×1014
Longitudinal emittance 0.6 eV sMomentum spread (100%) ±2.5×10-3
Transverse emittance (100%) h=v
25 mm·mrad
Harmonic number 588Number of bunches 548Main RF Frequency 52.811 MHzMain RF Voltage 750 kVSecond Harmonic RF Voltage 375 kVBetatron tunes Qh/Qv 20.45/20.46
Betatron tune chromaticity -20Synchrotron tune (maximum) 0.0067Transverse acceptance 40 mm·mradMomentum acceptance ±3.2×10-3
Phase-space painting
Apr 12, 2011 - S. Nagaitsev Page 43
Motivation– Gaussian beam G =3– Single RF harmonic at 53 MHz B =5
DQ= -0.3
– Uniform beam G =1– Longitudinal painting B =2
DQ= -0.04
BGN
N
b
320
14
2r-Q
mrad mm25107.1
D
Longitudinal painting
• Linac long. emittance: 1.3e-4 eV-s
• Recycler long. emittance: 0.6 eV-s
Apr 12, 2011 - S. Nagaitsev Page 44B=2.2
0 1 2 3 4 5 6 7 8 9 102-
1-
0
1
2
f t( )
g t( )
h t( )
t
Recycler RF
162 MHz bunches to be removed
Transverse painting
• Transverse painting is designed to:– Minimize the number of secondary passages and foil heating– To make correlated x-y painting (K-V distribution) with radius increase
for each next pulse
• Injected beam does not move on foil
• Closed orbit describes almost a quarter of circle (forward and back)
Apr 12, 2011 - S. Nagaitsev Page 45
Recycler modifications
• New rf systems:– 53 MHz and 106 MHz (fixed frequency)
• New injection system
• Possibly TiN vacuum pipe coating to mitigate electron cloud
Apr 12, 2011 - S. Nagaitsev Page 46
MI modifications
• New RF systems– 53 MHz and 106 MHz (with freq. sweep for acceleration)– 2.7 MV/turn at 36-degree synch phase
• γt jump system• Possibly TiN vacuum pipe coating to mitigate electron cloud
Apr 12, 2011 - S. Nagaitsev Page 47
Summary
• Project X design concept is well developed and has been stable for more than 1 year.
• We have an RD&D plan to take us through CD2. Key R&D issues (among others)
– Chopper section
– Recycler injection
– SRF cavities
Apr 12, 2011 - S. Nagaitsev Page 48