Post on 09-Mar-2021
A. Sukhanov
Fermilab, TD SRF T&D
HOMSC 2018, Cornell UniversityOctober 1-3, 2018
Update on HOM studies for PIP II SC Linac
PIP-II Project
•Mission: deliver intense beam of neutrinos to the international LBNF/DUNE project
•Goals:
‣ Deliver >1 MW proton beam power from the Fermilab Main Injector over the energy range 60 — 120 GeV, at the start of LBNF operation
‣ Support ongoing 8 GeV program at Fermilab, including upgrade path for Mu2e
‣ Possibility of extension of beam power to LBNF >2 MW
‣ Possibility of extension to high duty factor/higher beam power
2
PIP-II Project
• PIP-II Linac
3
8
Figure 1‐1: Site layout of PIP‐II (north is to the right). New construction includes the linac enclosure, transfer line enclosure (including the beam abort area and a stub to facilitate future connection to the Muon Campus), linac gallery, utility building, and cryo/compressor building. The blue areas denote identified wetland areas.
PIP-II provides a variety of straightforward and cost effective upgrade paths. Delivery of more than 2 MW to the LBNF target will require replacement of the existing Booster. The most straightforward strategy would be to extend the 0.8 GeV linac to 1.5-2 GeV and to inject at this energy into a newly constructed 8-GeV rapid cycling synchrotron (RCS). Such a synchrotron could either be of a conventional sort, as currently deployed at the J-PARC facility in Japan, or could incorporate novel design features based on highly non-linear optical elements currently under study at Fermilab. In either case, the siting shown in Figure 1.1 is compatible with siting of such a RCS to the south of the linac, providing capabilities for injecting beam into the Main Injector.
Parameter
Particle species H-
Input beam energy 2.1 MeV
Output beam energy 800 MeV
Bunch repetition rate 162.5 MHz
RF pulse length pulsed-to-CW
Sequence of bunches Programmable
Average beam current 2 mA
Final rms norm trans. emittance
<= 0.3 mm-mrad
Final rms norm long. emittance
<= 0.35 mm-mrad
RMS bunch length < 4 ps
PIP-II Project
• PIP-II Linac
4
HWR SSR1 SSR2Elliptical 5-cell
Cavity Aperture, mm Eff. length, cm Eacc, MV/m Epeak, MV/m Bpeak, mT (R/Q), Ohm G, Ohm
LB650 88 70.3 16.9 40.3 74.6 341 193
HB650 118 106.1 18.8 38.9 73.1 610 260
CM #cav/CM #CM CM config CM length, m Q0 x1e10, @2K Rs, nOhm QL, x1e6
LB650 3 11 ccc 4.32 2.15 9.0 10.36
HB650 6 4 cccccc 9.92 3 8.7 9.92
LB650 HB650
HOM in PIP II (a.k.a. Project X)
• Presented at HOMSC 2012, published in Proceedings
5
Higher Order Modes in the Project-X Linac
V. Yakovlev, Fermilab
HOM in PIP II• Complex beam current spectrum• No HOM couplers and dumpers (QL up to 1e7)
• Investigation of HOM various effects in PIP II reported at HOMSC 12:‣ found no big issues
• Here we consider HOM effects in PIP II‣ new modes of linac operation (long trains of 162.5 MHz bunches, 10 mA)
- important for accelerator-driven sub-critical system demonstration‣ modifications to LB650 cavity design
- optimization of EM and mechanical parameters, production process
6
Frequency, MHz0 20 40 60 80 100 120 140 160 180 200
, mA
nI~
-210
-110
1
Frequency, GHz2.5 3 3.5 4 4.5 5 5.5
Ωm
ax
(R/Q
),
1
10
210
R0 31.89± 137.7
f0 0.1515± 1.425
R0 31.89± 137.7
f0 0.1515± 1.425
Beam Dynamics in PIP2 Linac with 10 mA
Motivation
• Transverse misalignment of SRF cavities and beam offset lead to excitation of dipole HOMs
‣ depends on bunch charge, mode (R/Q), Qext
‣ excited dipole HOMs introduce additional kick on beam particles
‣ increase of transverse beam emittance
‣ small effect in a steady state (CW beam, no variations in beam current and/or bunch timing pattern)
• Study transverse emittance variations due to transitions (beam turn on) and bunch charge variations in HB650 section of PIP2 linac
‣ (R/Q) of dipole modes in LB650 are much smaller compared to HB650
8
Dipole modes in HB650
• (R/Q) depends on beam velocity
•Mode 1376 MHz: (R/Q)=80kOhm/m2, Qext=1e7
9
Linac layout
•HB650 beta=0.92 section: 4 cryo-modules, 24 cavities (6 cavities per CM)
•Matrix tracking through linac elements
10
HB650 (6-cavity) Period
NC Doublet
CavityFieldMap
1.316m 1.525m
CavityFieldMap
CavityFieldMap
1.44m
SC Cryomodule
0.6m
0.2m
10.97m
CavityFieldMap
CavityFieldMap
CavityFieldMap
NC Doublet
1.429m
• 150mmspaceaddedforgatevalveflangeandmagnetflange.
HBSec9on
Cavity Gradient and Quadrupole Fields
•HB650 Vacc(beta_G) = 20 MV
•HB650 Quad gradient 12 T/m
11
0
5
10
15
20
25
0 20 40 60 80 100 120
Am
plitu
de (M
V)
Cavity #
MEBTHWRSSR1SSR2
LB650HB650
0
5
10
15
20
0 20 40 60 80 100 120 140 160G
radi
ent (
T/m
)
Position of the Center (m)
MEBTHWR
LB650HB650
Model• 60 pC bunches with 10% variation, 162.5 MHz bunch frequency
‣ train of 5e6 bunches
• Transverse misalignment of cavities R.M.S. = 0.5 mm
• Matrix tracking through linac
• Consider one dipole mode with highest (R/Q)
• Each bunch introduces voltage V[i] = jcq(R/Q)x
‣ bunch sees half of this kick voltage
• Total kick seen by bunch is sum of voltages from all previous bunches with proper time dependent factors: ~exp(jwTb)*exp(-wTb/2Q)
• Simulate 100 linac configurations (time consuming)
‣ random variations of transverse displacement of cavities and HOM frequency with 1 MHz RMS
•Calculate effective RMS emittance at the end of linac from variations of (x,x’) of bunches
‣ compare to nominal emittance 0.3 mm*mrad (relative emittance)
12
Tracking in linac
• bunch trajectories in linac
13
z , m0 5 10 15 20 25 30 35 40 45 50
x , u
m
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
Effective emittance
• Effective emittance at the end of linac for single linac configuration as a function of bunch number
14
bunch
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000310×
rela
tive
emitt
ance
cha
nge
0
5
10
15
20
25
30
-610×
Maximum emittance growth
Initial emittance growth after turn on, scales ~ Qext
Emittance variations due to Qb variations
Maximum emittance growth
• 100 configurations of linac with random cavity misalignment of 0.5 mm
‣ Median value of relative emittance growth is 4e-5
15
Relative emittance growth0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
-310×0
2
4
6
8
10
Longitudinal emittance
• 10 mA (presented in LINAC’12, TUPB054)
16
Relative emittance growth, !"z
Pro
babi
lity
Blue – old design Red – new design
10 mA
New Design of LB 650 MHz 5-cell Cavities
New Designs of LB650 5-cell Cavities
•New designs of LB650 5-cell cavities have been developed at FNAL, INFN and VECC
‣ Optimization of EM and mechanical properties, production process
• 3-D RF simulation of cavities of each design (A. Lunin)
‣ calculate frequency spectra, QL, (R/Q) of monopole, dipole and quadrupole modes
18
Monopole modes
•QL and (R/Q)
19
Frequency, GHz0 0.5 1 1.5 2 2.5
LQ
310
410
510
610
710
810
910
1010FNALINFNVECC
Frequency, GHz0 0.5 1 1.5 2 2.5
, O
hmz
(R/Q
)
-210
-110
1
10
210
310FNALINFNVECC
Dangerous trapped mode in VECCdesign: f=1.663 GHz, (R/Q)=18 OhmThis mode is relatively far from mainbeam current harmonics of 162.5 MHz
Dipole modes
•QL and (R/Q)
20
Frequency, GHz0 0.5 1 1.5 2 2.5
LQ
310
410
510
610
710
810
910
1010FNALINFNVECC
Frequency, GHz0 0.5 1 1.5 2 2.5
2 ,
Ohm
/mt
(R/Q
)
-210
-110
1
10
210
310
410
510
610
710
810
Most possibly dangerous modes are:VECC: 0.974 GHz, (R/Q) = 38 hOhm/m**2df = 1 MHz FNAL: 0.973 GHz, (R/Q) = 35 kOhm/m**2df=2 MHz
Linac Layout
• LB650 beta=0.61 section: 11 cryo-modules, 33 cavities (3 cavities per CM)
21
LB650 (3-cavity) Period
CavityFieldMap
0.6m
0.2m
NC Doublet
1.145m 1.184m
CavityFieldMap
CavityFieldMap
0.9m
SC Cryomodule
5.371m
NC Doublet
1.258m
• 150mmspaceaddedforgatevalveflangeandmagnetflange.
LBSec9on
Cavity Gradient and Quadrupole Fields
• LB650 Vacc(beta_G) = 12 MV
• LB650 Quad gradient 9 T/m
22
0
5
10
15
20
25
0 20 40 60 80 100 120
Am
plitu
de (M
V)
Cavity #
MEBTHWRSSR1SSR2
LB650HB650
0
5
10
15
20
0 20 40 60 80 100 120 140 160G
radi
ent (
T/m
)
Position of the Center (m)
MEBTHWR
LB650HB650
Model• 30 pC bunches with 10% variation, 162.5 MHz bunch frequency
‣ train of 1e5 bunches
‣ about 50% of bunches removed for injection at 44.705 MHz
• Transverse misalignment of cavities R.M.S. = 0.5 mm
• Matrix tracking through linac
• Consider one dipole mode with highest (R/Q)
• Each bunch introduces voltage V[i] = jcq(R/Q)x
‣ bunch sees half of this kick voltage
• Total kick seen by bunch is sum of voltages from all previous bunches with proper time dependent factors: ~exp(jwTb)*exp(-wTb/2Q)
• Simulate 100 linac configurations (time consuming)
‣ random variations of transverse displacement of cavities and HOM frequency with 1 MHz RMS
•Calculate effective RMS emittance at the end of linac from variations of (x,x’) of bunches
‣ compare to nominal emittance 0.3 mm*mrad (relative emittance)
23
bunch
0 10 20 30 40 50 60 70 80 90 100310×
rela
tive
emitt
ance
cha
nge
0
0.01
0.02
0.03
0.04
0.05
0.06-310×
Effective emittance
• Effective emittance at the end of linac for single linac configuration as a function of bunch number
24
Maximum emittance growth
Initial emittance growth after turn on, scales ~ Qext
Emittance variations due to Qb variations
Maximum emittance growth, VECC design
• 100 configurations of linac with random cavity misalignment of 0.5 mm
‣ Median value of relative emittance growth is 5e-5
‣ 95% of configurations have relative emittance growth less than 1e-3
25Relative emittance growth
0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007 0.0008 0.0009 0.0010
5
10
15
20
25
30
35
40
45
Maximum emittance growth, FNAL design
• 100 configurations of linac with random cavity misalignment of 0.5 mm
‣ Median value of relative emittance growth is 2.5e-5
‣ 95% of configurations have relative emittance growth less than 1.5e-4
26
Relative emittance growth0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007 0.0008 0.0009 0.0010
10
20
30
40
50
60
70
80
Conclusion
• Study effects of dipole HOM excitation on transverse beam dynamics in PIP2 linac with 10 mA peak current
‣ Considered mode with largest (R/Q)=80 kOhm/m2, f=1376 MHz, Qext=1e7
‣ 0.5 mm random cavity misalignment
‣ 10% bunch charge variations
• Relative emittance growth is 4e-5 - should not be a problem in PIP2 linac with 10 mA
• Study effects of dipole HOM excitation on transverse beam dynamics in LB650 section of PIP2 linac with different designs of LB650 cavities
‣ Considered modes with largest (R/Q)
- VECC design, mode 0.974 GHz, (R/Q)=38 Ohm/m**2, df=1MHz
- FNAL design, mode 0.973 GHz, (R/Q)=35 Ohm/m**2, df=2 Mhz
‣ 0.5 mm random cavity misalignment
‣ 10% bunch charge variations
• Relative emittance growth is 5e-5 for VECC design and 2.5e-5 for FNAL design
• No considerable effects are expected for INFN design
27