Normal Conducting RF Cavity Normal Conducting RF Cavity R&D for - - PowerPoint PPT Presentation
Normal Conducting RF Cavity Normal Conducting RF Cavity R&D for Neutrino Factory or R&D for Neutrino Factory or Muon Collider Muon Collider Thomas Jefferson National Laboratory Facility Thomas Jefferson National Laboratory Facility
Derun Li – NCRF R&D for NF and MC Page 1
(Sept. 11, 2006) Derun Li and Steve Derun Li and Steve Virostek Virostek Center for Beam Physics Center for Beam Physics Lawrence Berkeley National Laboratory Lawrence Berkeley National Laboratory
Derun Li – NCRF R&D for NF and MC Page 2
Derun Li – NCRF R&D for NF and MC Page 3
Introduction
– – Muon Ionization Cooling Muon Ionization Cooling – – International Muon Ionization Cooling Experiment (MICE) International Muon Ionization Cooling Experiment (MICE)
Experimental study program
– – 805 MHz cavity design 805 MHz cavity design – – Fabrication Fabrication – – Coupler design Coupler design – – Achievable accelerating gradient under a few Tesla magnetic fiel Achievable accelerating gradient under a few Tesla magnetic fields ds – – Button tests with different materials and coatings Button tests with different materials and coatings
Thin beryllium windows for RF cavity
– – Pre Pre-
stressed flat windows, grids – – Curved Be windows Curved Be windows
201 MHz Cavity Program
– – Cavity design Cavity design – – Fabrication techniques Fabrication techniques
Spinning
Extruding
E-
beam welding
Cleaning and electro-
polishing (EP)
– – Loop coupler and conditioning Loop coupler and conditioning – – Preliminary test results of the cavity Preliminary test results of the cavity
Some of the US MICE responsibilities
Summary
Derun Li – NCRF R&D for NF and MC Page 4
R&D program under auspices of the U.S. Neutrino Factory and Muon Collider Collaboration (NFMCC) Muon Collider Collaboration (NFMCC)
Three leading national laboratories (BNL, FNAL, LBNL) + JLab JLab and and
– – Production, acceleration and storage of intense muon beams
Production, acceleration and storage of intense muon beams (hardware and software) (hardware and software)
Technology and engineering solutions
Accelerator physics of intense muon beams
Support from DOE, NSF, Illinois State and U.S.-
Japan
R&D progress enhanced significantly by corresponding programs in Europe and Japan in Europe and Japan
Long term goals
– – Continue evaluating physics opportunities afforded by Continue evaluating physics opportunities afforded by intense muon intense muon beams beams from from Neutrino Factory Neutrino Factory through through a Muon Collider a Muon Collider
Near term goals
– – Muon cooling R&D (software and hardware) Muon cooling R&D (software and hardware) – – International International M Muon uon I Ionization
Cooling
Experiment ( xperiment (MICE MICE) ) – – Cost effective Neutrino Factory Design Cost effective Neutrino Factory Design
Derun Li – NCRF R&D for NF and MC Page 5
Neutrino Factory comprises Neutrino Factory comprises
Proton Driver
– – primary beam on production primary beam on production target target
Target, Capture, and Decay , and Decay
– – Create Create π π; decay into ; decay into µ µ
Bunching and Phase Rotation
– – Conditioning: reduce Conditioning: reduce Δ ΔE of bunch E of bunch
Cooling
– – Reduce transverse emittance Reduce transverse emittance – – MICE MICE
Acceleration
– – 130 130 MeV MeV ~ ~ 20 20– –50 50 GeV GeV
Storage Ring
– – Store for ~ 500 turns; long Store for ~ 500 turns; long straight straight
50 100 m
Derun Li – NCRF R&D for NF and MC Page 6
– – Components R&D:
Components R&D: NC RF cavities NC RF cavities, absorbers and solenoids , absorbers and solenoids
– – Normal conducting RF cavity studies:
Normal conducting RF cavity studies:
Experimental studies at 805 MHz using a pillbox cavity
Be windows R&D
– – Thermal and mechanical stabilities at high accelerating gradient
Thermal and mechanical stabilities at high accelerating gradients s
– – Scattering and limits
Scattering and limits
Tests are being conducting now at MTA (MUCOOL Test Area), FNAL FNAL
201 MHz cavity design, fabrication and tests and tests
– – Absorbers (ICAR, Japan Absorbers (ICAR, Japan-
US funding)
Absorbers, windows and safety issues
Design and FEA simulations of absorber windows
Absorber tests at MTA, FNAL
– – SC solenoids (magnets) SC solenoids (magnets) – – International MICE International MICE experiment experiment
Derun Li – NCRF R&D for NF and MC Page 7
High gradient RF cavities to compensate for lost longitudinal energy ergy – – Strong magnetic field to confine muon beams Strong magnetic field to confine muon beams – – Energy loss in LH Energy loss in LH2
2 absorbers
absorbers
Goal: Goal:
– – Development of NC 201 Development of NC 201-
MHz cavity operating at ~ 16 MV/m ~ 16 MV/m in a few in a few-
Tesla solenoidal solenoidal B B field field (~ 30 MV/m at 805 (~ 30 MV/m at 805-
MHz)
LH Absorbers RF Cavities
dx dE dx dE dx dE
Magnets
Derun Li – NCRF R&D for NF and MC Page 8
Muon beam is unstable, and has short decay time (~ 2 μ μs at rest) s at rest)
Muon beam is created with LARGE LARGE 6 6-
D phase space
– – Muon beam manipulation must be done Muon beam manipulation must be done quickly quickly including including cooling cooling
Requirements of RF cavity for muon beams
– – High cavity shunt impedance, high gradient and high field High cavity shunt impedance, high gradient and high field
Gradient at 201 MHz: ~ 16.5 MV/m {Kilpatrick criterion: 15 MV/m Kilpatrick criterion: 15 MV/m} }
Gradient at 805 MHz: ~ 30 MV/m {Kilpatrick criterion: 26 MV/m Kilpatrick criterion: 26 MV/m} }
– – RF cavity with closed iris RF cavity with closed iris (iris terminated by Be window) (iris terminated by Be window)
Higher shunt impedance
Independent phase control, higher transit factor
Lower peak surface field
Derun Li – NCRF R&D for NF and MC Page 9
Derun Li – NCRF R&D for NF and MC Page 10
– – Design and fabrication of the cavity Design and fabrication of the cavity – – Highest possible shunt impedance and high acceleration Highest possible shunt impedance and high acceleration gradient at the order of ~ 30+ MV/m gradient at the order of ~ 30+ MV/m – – Allowing for testing of Be windows with different thickness, Allowing for testing of Be windows with different thickness, coatings, and other windows as well coatings, and other windows as well
Copper windows, Be windows, Grids and curved Be windows
– – Study RF cavity operation and conditioning under the Study RF cavity operation and conditioning under the influence of strong external magnetic fields (a few Tesla) at influence of strong external magnetic fields (a few Tesla) at both the solenoid and gradient modes both the solenoid and gradient modes
– – Mechanical stabilities and thermal stress under RF heating Mechanical stabilities and thermal stress under RF heating – – Prototype and FEA modeling Prototype and FEA modeling – – Evolutions of Be windows Evolutions of Be windows
Derun Li – NCRF R&D for NF and MC Page 11
Pillbox cavity Pillbox cavity Be (or Cu) windows Be (or Cu) windows Waveguide + window Waveguide + window Coupler Thermo Thermo-
couples
Three more view ports on the equator Three more view ports on the equator
Derun Li – NCRF R&D for NF and MC Page 12
– – 38 M 38 M /m (Z /m (Z0
0); 32
); 32 M M /m (ZT /m (ZT2
2); {definition used:
); {definition used: Z = V Z = V2
2/P
/P} }
Q0
0 = 18,800
= 18,800
– – Critical coupling at Critical coupling at c
c = 1.0
= 1.0 – – Accelerating gradient of Accelerating gradient of 30 MV/m 30 MV/m requires requires 2 MW 2 MW peak power peak power – – 350 watts 350 watts average on cavity body, average on cavity body, 52 watts 52 watts on two windows
( (66 watts if Be windows 66 watts if Be windows) at a duty factor of 1.8x10 ) at a duty factor of 1.8x10-
4 (12
(12 μ μs s pulse length and 15 Hz repetition rate) pulse length and 15 Hz repetition rate)
Derun Li – NCRF R&D for NF and MC Page 13
Derun Li – NCRF R&D for NF and MC Page 14
– – Before tuning: Before tuning: f f = 803.198 MHz, = 803.198 MHz, c
c = 0.12
= 0.12 – – After tuning: After tuning: f f = 805.486 MHz, = 805.486 MHz, Q Qext
ext = 12,800
= 12,800 – – Shipped to Alpha Braze, California for brazing Shipped to Alpha Braze, California for brazing – – Measurements after final brazing (before shipping to FNAL): Measurements after final brazing (before shipping to FNAL):
Derun Li – NCRF R&D for NF and MC Page 15
– – Frequency tuning (shortening the gap) Frequency tuning (shortening the gap) – – Coupler tuning (widening coupling slot and shortening the transi Coupler tuning (widening coupling slot and shortening the transition tion waveguide) waveguide) – – Good agreements between simulations and measurements Good agreements between simulations and measurements
Coupling slot
Derun Li – NCRF R&D for NF and MC Page 16
MWS model of the J-Lab Cavity Excitation pulse E-Field decay
Method has been benchmarked again against measurements for the JLab HOM damped cold test cavity
as a function of time inside the cavity
Derun Li – NCRF R&D for NF and MC Page 17
Slope k
Derun Li – NCRF R&D for NF and MC Page 18
12 MW klystron with capability of 50 μs and 15 Hz Open cell cavity Waveguide Input Waveguide directional coupler Superconducting solenoid up to 5 T
Klystron collector Shielded Cave
Derun Li – NCRF R&D for NF and MC Page 19
Derun Li – NCRF R&D for NF and MC Page 20
gradient (54 MV/m surface field)
window damage
gradient of 30 MV/m with no magnetic field and reached up to 40 MV/m
surface have been tested versus magnetic fields up to 4 Tesla – – No surface damage was found on No surface damage was found on the the Be windows Be windows – Little multipacting was observed; accelerating gradient limit is a function of the external magnetic field
40 40 37 36 34 32.4 31.7 31 28.8 26.74 26.4 25.9 22 40 40 25.75 23.25 22.5 21.5 20.9 16.5 15 13.5 5 10 15 20 25 30 35 40 45 1 2 3 4 5 Magnetic Field Tesla G radient M V/m Damaging Sparks Safe limit
We have conducted experimental studies at 805 MHz for nearly We have conducted experimental studies at 805 MHz for nearly three years at Lab G and has resumed recently at MTA, FNAL three years at Lab G and has resumed recently at MTA, FNAL
Derun Li – NCRF R&D for NF and MC Page 21
These windows (surface damage) have been analyzed at ANL and JLab
Derun Li – NCRF R&D for NF and MC Page 22
SC Solenoid SC Solenoid 805 MHz pillbox cavity 805 MHz pillbox cavity inside the SC solenoid inside the SC solenoid Up to 12 MW Up to 12 MW peak power peak power
Curved Be windows
Button study
─ Completed design and
Completed design and fabrication fabrication
─ ─ Ratio of peak surface
Ratio of peak surface field (on the button) field (on the button) versus accelerating versus accelerating field on field on-
axis ≈
≈ 1.7 with
1.7 with ~ 0.5 MHz shift ~ 0.5 MHz shift
─ ─ New materials
New materials
─ ─ many buttons
many buttons
─ ─ SS, Cu with
SS, Cu with coatings, coatings, … …
─ ─ Different coatings
Different coatings
Demountable button
Derun Li – NCRF R&D for NF and MC Page 23
Derun Li – NCRF R&D for NF and MC Page 24
– Transparent to muon beams – Perfect electric boundary to RF field – No detuning to cavity frequency
– Pre-stressed flat Be windows – Pre-curved Be windows – Grids
A pre-curved Be window: 0.25 mm thick and 21 cm in radius
Window profile evolutions
Derun Li – NCRF R&D for NF and MC Page 25
design and prototype for 805 MHz pillbox cavity (Ph.D thesis at IIT)
Tests at MTA soon Field enhancement between 1.4 & 3.6 RF Heating on tubes
4x4-Connected 3.60 4x4 -Waffle 2.30 1.80 6x6 -Waffle 1.64 1.40 1.39 6x6 Middle- Concentrated/Waffle 1.40
Tube DIA (cm)
Grid
0.50 1.00 1.25 1.50
Grid Model Electric Field Magnetic Field Maximum Surface Field Enhancement
First prototype of solid Al grid
Derun Li – NCRF R&D for NF and MC Page 26
– – The graphite die in Al fixture (room or high temperatures) The graphite die in Al fixture (room or high temperatures) – – 10 S.S sheets (10 mils) and 3 Be foils (10 mils) have been order 10 S.S sheets (10 mils) and 3 Be foils (10 mils) have been ordered for ed for the pre the pre-
form tests – – Halogen lamp heating tests may be conducted at the 805 MHz low Halogen lamp heating tests may be conducted at the 805 MHz low power test cavity to benchmark the FEA models power test cavity to benchmark the FEA models
Derun Li – NCRF R&D for NF and MC Page 27
Succeeded in the S.S. window with Cu frame for 805 MHz cavity Pre-formed at room temperature by holding foil edge then braze the Cu frame A finished curved S.S. window with brazed Cu frame
Derun Li – NCRF R&D for NF and MC Page 28
However, we believe the curved Be windows can be formed at highe However, we believe the curved Be windows can be formed at higher r temperature at that time. It turns out that the windows can be temperature at that time. It turns out that the windows can be formed and formed and brazed at Brushwellman Company brazed at Brushwellman Company without significant cost increases without significant cost increases. We . We have succeeded in making these windows and have three (805 MHz have succeeded in making these windows and have three (805 MHz cavity) and two (for 201 MHz) of them ready for use for high pow cavity) and two (for 201 MHz) of them ready for use for high power tests. er tests.
Derun Li – NCRF R&D for NF and MC Page 29
Two windows available now
– – 21 21-
cm and 0.38-
mm thick – – “ “Good Good” ” braze (between annular frames and foil) braze (between annular frames and foil) – – Achieved ~ 95 % of the designed profile Achieved ~ 95 % of the designed profile – – Thin Ti Thin Ti-
N coatings
Windows installed pointing to the same direction in the cavity
Ready for HP tests
42-cm
Derun Li – NCRF R&D for NF and MC Page 30
Derun Li – NCRF R&D for NF and MC Page 31
– – Prototype cavity for MUCOOL/MICE Prototype cavity for MUCOOL/MICE
Designs
Couplers and ceramic RF windows
Engineering
Fabrication: cavity body, ports, tuners, etc.
Large, pre-
curved and thin 42-
cm diameter Be windows
– – Commissioning and operation Commissioning and operation
RF Conditioning and background (for MICE only)
High gradients (~ 16 MV/m) and high RF power
With the thin Be windows
With external magnetic fields
Derun Li – NCRF R&D for NF and MC Page 32
De De-
mountable Pre-
curved Be windows to terminate Be windows to terminate RF fields at the iris RF fields at the iris 2 2o
tilt angle Spherical section at the equator to Spherical section at the equator to ease addition of ports ( ease addition of ports (± ± ~ 6 ~ 6o
) Elliptical Elliptical-
like (two circles) nose to reduce peak surface field reduce peak surface field 6 6-
mm Cu sheet allows for uses of spinning technique and mechanical spinning technique and mechanical tuners similar to SCRF ones tuners similar to SCRF ones Stiffener ring Stiffener ring
Derun Li – NCRF R&D for NF and MC Page 33
– – Frequency: 201.25 MHz Frequency: 201.25 MHz – – β β = 0.87 = 0.87 – – Shunt impedance (definition used: VT Shunt impedance (definition used: VT2
2/P): ~ 22 M
/P): ~ 22 MΩ Ω/m /m – – Quality factor (Q Quality factor (Q0
0): ~ 53,500
): ~ 53,500 – – Be window radius and thickness: 21 Be window radius and thickness: 21-
cm and 0.38-
mm
– – ~ 16 MV/m peak accelerating field ~ 16 MV/m peak accelerating field – – Peak input RF power ~ Peak input RF power ~ 4.6 MW 4.6 MW per cavity (85% of Q per cavity (85% of Q0
0, 3
, 3τ τ filling) filling) – – Average power dissipation per cavity ~ Average power dissipation per cavity ~ 8.4 kW 8.4 kW – – Average power dissipation per Be window Average power dissipation per Be window ~ ~ 100 watts 100 watts
Derun Li – NCRF R&D for NF and MC Page 34
Spinning of half shells using thin Spinning of half shells using thin copper sheets and e copper sheets and e-
beam welding to join the shells to join the shells Cavity design uses pre-curved Be windows, but also accommodates different windows
Derun Li – NCRF R&D for NF and MC Page 35
Spinning a bowl Spinning a bowl Spinning tools Spinning tools
Derun Li – NCRF R&D for NF and MC Page 36
CMM scans to measure the spun profiles of shells spun profiles of shells
Frequency and Q measurements
Copper tapes for better RF contacts contacts
Profile measurements: Profile measurements: 3 CMM scans per half shell conducted 3 CMM scans per half shell conducted at 0 at 0o
, 45o
, 90o
, respectively
201 MHz Muon Cavity Shell #1 CMM Profiles
4 8 12 16 20 10 20 30 40 50 60 Radial Dimension (cm) Axial Dimension (cm)
Measured frequency of shell-1: 196.97 MHz (simulated frequency: 197.32 MHz)
Derun Li – NCRF R&D for NF and MC Page 37
Derun Li – NCRF R&D for NF and MC Page 38
Mechanical cleaning of the cavity inner Mechanical cleaning of the cavity inner surface right after e surface right after e-
beam welding of the stiffener ring (above) the stiffener ring (above)
Derun Li – NCRF R&D for NF and MC Page 39
Cavity and fixture system is mounted and assembled on a mounted and assembled on a plate and placed on the welder plate and placed on the welder sliding table sliding table
External structural weld is near full penetration and is achieved full penetration and is achieved in three offset passes in three offset passes
A final cosmetic/vacuum weld is performed on the inside of is performed on the inside of the joint with the cavity the joint with the cavity mounted on a horizontal rotary mounted on a horizontal rotary table table
Derun Li – NCRF R&D for NF and MC Page 40
Possible improvement: Possible improvement: Anneal around the Anneal around the extruding area + combination between pilot hole extruding area + combination between pilot hole dimensions and lid heights, dimensions and lid heights, … …
Derun Li – NCRF R&D for NF and MC Page 41
We have successfully developed techniques to extrude ports We have successfully developed techniques to extrude ports across e across e-
beam welded joints.
Derun Li – NCRF R&D for NF and MC Page 42
Local annealing is achieved by repeatedly achieved by repeatedly passing a diffuse e passing a diffuse e-
beam around port around port
Softening of copper must be local to preserve cavity be local to preserve cavity
Port pulling tool is used in a horizontal orientation a horizontal orientation
A weld prep is machined into the port lip using a into the port lip using a numerically controlled mill numerically controlled mill
Derun Li – NCRF R&D for NF and MC Page 43
Loop coupler at critical coupling Prototype coupling loop design uses standard
Parts were joined by torch brazing Coupling loop has integrated cooling lines ☺ Two SNS style RF windows mfg. by Toshiba received (no cost to us !) Two couplers with RF windows Bellows connection required on MICE cooling channel (Study-II) for thermal and dimensional reasons
Loop Coupler Design Loop Coupler Design
Ceramic RF window
Derun Li – NCRF R&D for NF and MC Page 44
rotating the loop
the loop
Ceramic RF window
Water cooling line
Derun Li – NCRF R&D for NF and MC Page 45
Conditioning started during at SNS, ORNL SNS, ORNL
Good vacuum ~ low 10-
8 T
T
Achieved 600 kW 600 kW in TW mode in TW mode (matched load) (matched load)
Achieved 10 kW average power 10 kW average power (~ 9 kW for nominal NF (~ 9 kW for nominal NF parameters) parameters)
Achieved 2.4 MW peak power 2.4 MW peak power in in SW mode (at variable short SW mode (at variable short positions) positions)
Ceramic windows work perfectly within two weeks of the within two weeks of the conditioning conditioning RF Power RF Power Two couplers Two couplers
Derun Li – NCRF R&D for NF and MC Page 46
The thermal solution provides temperature distribution throughout the cavity and the beryllium window The peak temperature
the beryllium window (86 ºC) FEA helps to determine designs for: → Cooling tubes → Be window thickness
Derun Li – NCRF R&D for NF and MC Page 47
Requirement:
conduction
body
and temperature We have developed the technique and achieved the design goal Silicon-Bronze with helium gas torch + argon gas flowing in the cooling tubes
Derun Li – NCRF R&D for NF and MC Page 48
First measurement: First measurement: f f ~ 199.5 MHz with ~ 199.5 MHz with β βC
C (max) ~ 5
(max) ~ 5
Derun Li – NCRF R&D for NF and MC Page 49
surfaces are ready for electropolishing
rotary buffing wheel and a cavity rotation fixture
Derun Li – NCRF R&D for NF and MC Page 50
electropolish process
rebuffing was required
Derun Li – NCRF R&D for NF and MC Page 51
Derun Li – NCRF R&D for NF and MC Page 52
blank-offs, diagnostic spool, window cover plates, gate valve and window pump-out tubes
assembled state and packaged in a custom made crate for shipping to the MTA
Couplers
Derun Li – NCRF R&D for NF and MC Page 53
couplers were installed in a portable clean room
Derun Li – NCRF R&D for NF and MC Page 54
f = 199.578 MHz Q0 = 49,000 ~ 51,000 (better than 90% of the design value) Two couplers ─ balanced ─ coupling adjustments
Derun Li – NCRF R&D for NF and MC Page 55
Mechanical tuning plates at four locations locations
Dial indicators to measure displacement between Al plates displacement between Al plates
Tuning measurement in air – – Equivalent to MICE cavity Equivalent to MICE cavity under vacuum under vacuum
Adjusted up to 2-
mm with 8 steps
mm each
Measured tuner sensitivity
– – ~ 78 kHz/mm ~ 78 kHz/mm
Calculated tuner sensitivity
– – 115 kHz/mm 115 kHz/mm – – Disagreements are due to Disagreements are due to deflection of the Al plates deflection of the Al plates Dial indicators
Derun Li – NCRF R&D for NF and MC Page 56
Loop power coupler Loop power coupler Portable clean room Movable cavity support The cavity 201 MHz coaxial RF power line RF probes Vacuum pump
Derun Li – NCRF R&D for NF and MC Page 57
9 Torr
Derun Li – NCRF R&D for NF and MC Page 58
Coupling Coils Coupling Coils Eight 201 MHz cavities Eight 201 MHz cavities Power couplers Power couplers
Derun Li – NCRF R&D for NF and MC Page 59
Three quarter section of 3-D view of the RFCC module
RFCC is US MICE MICE deliverable deliverable
Spectrometer solenoid
Derun Li – NCRF R&D for NF and MC Page 60
NFMCC has continued to make good technical progress in R&D for intense muon beam accelerator for intense muon beam accelerator
R&D Programs – – RF cavity R&D RF cavity R&D – – Be window R&D Be window R&D – – SC magnets and system integration SC magnets and system integration
Experimental studies at 805 MHz using the pillbox cavity are being conducting now at MTA being conducting now at MTA
The 201 MHz test cavity fabrication completed; tests will continue ue – – Reached 16 MV/m quickly without magnetic field Reached 16 MV/m quickly without magnetic field
Continue to develop and test hardware for μ μCOOL and MICE COOL and MICE
Derun Li – NCRF R&D for NF and MC Page 61