Page 1
Bob Laxdal, TRIUMF
F. Ames, R. Baartman, I. Bylinskii, Y.C.Chao, D. Dale, K. Fong,
E. Guetre, P. Kolb, S. Koscielniak, A. Koveshnikov, M. Laverty,
Y. Ma, M. Marchetto, L. Merminga, A.K. Mitra, N. Muller, R.
Nagimov, T. Planche, W.R. Rawnsley, V.A. Verzilov, Z. Yao, Q.
Zheng, V. Zvyagintsev
Status of Superconducting Electron Linac Driver for Rare Ion Beam
Production at TRIUMF
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 1
Page 2
• ARIEL Project – E-Linac Specification
• E-Linac design – Major components
• Status and commissioning results
• Future plans
Outline
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 2
Page 3
ARIEL and the e-Linac
Page 4
ARIEL Project (2010-2020)
ISAC-I
ISAC-II
•ISAC: World class ISOL facility
for the production and
acceleration of rare isotope
beams (RIB)
•Presently utilize one driver beam
at 500MeV and 50kW to create
RIBs for ISAC
•Now adding ARIEL to allow up to
three simultaneous RIB beams
•Add e-Linac (50MeV 10mA cw -
1.3GHz SC linac) as a second
driver to create RIBs via
photofission
•Add a second driver beam from
the cyclotron
E-Linac
500MeV
Cylotron
ARIEL
Page 5
Why electrons? Why 50MeV?
500MeV protons
50MeV electrons
Calculated in-target production for 10 μA, 500
MeV protons incident on a 25 g/cm2 UCx target
Calculated in-target production for 10 mA, 50
MeV electrons incident on a Hg converter and 15
g/cm2 UCx target
• the electron linac is a strong
complement to the existing
proton cyclotron
• Photofission yields high
production of many neutron rich
species but with relatively low
isobaric contamination with
respect to proton induced
spallation
• An energy of 50MeV is sufficient
to saturate photo-fission
production – fits the site footprint
and project budget
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 5
Page 6
E-Linac Specifications
• The ARIEL E-Linac specification – dominated by rf beam loading
– 10mA cw at 50MeV - 0.5 MW of beam power
– Choose five cavities 100kW of beam loaded rf power per cavity
– two couplers per cavity each rated for 50kW operation
– Means 10MV energy gain per cavity
• Linac divided into three cryomodules
– one Injector cryomodule (ICM) with one cavity
– two Accelerator crymodules (ACM1, ACM2) with two cavities each
– Installation is staged - Phase I – includes ICM and ACM1 for a required
25MeV/100kW demonstration by end of 2014
Gun
30MeV
50kW 50kW
2014
50kW
50kW
50kW 50kW
50kW 50kW
2018
50MeV 10MeV 50kW 50kW
ICM ACM1 ACM2
Page 7
The ARIEL e-Linac as a recirculator
• The linac is configured to
eventually allow a recirculating
ring for a multi-pass `energy
doubler’ mode or to operate as an
energy recovery linac for
accelerator studies and
applications
7
Klystron
Gallery
ICM
ACM1
ACM2
MEBT
10MeV
Dump
LEBT
E-Gun
30MeV
Dump
ERL
Dump
75MeV
100kW
Dump
HEBT
Sh
ield
ing W
all
Page 8
Accelerator Vault – Phase I
E-Gun
Vessel
ICM
E-Gun
HV
Supply
Cold
Box
Klystron Gallery
ACM1 MEBT
LEBT
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 8
Page 10
Electron Gun
• Thermionic 300kV DC gun – cathode
has a grid with DC supressing voltage
and rf modulation that produces
electron bunches at rf frequency
• Gun installed inside an SF6 vessel
• Rf delivered to the grid via a ceramic
waveguide
support Ceramic waveguide
SF6 Vessel
Rf tuner
Parameter Value
RF frequency 650MHz
Pulse length 160 (137ps)
Average current 10mA
Charge/bunch 15.4pC
Kinetic energy 300keV
Normalized emittance 5μm
Duty factor 0.01 to 100%
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 10
Page 11
ARIEL cavities
• The ARIEL cavities
– 1.3GHz nine-cell cavities
– End groups modified to
accommodate two 50kW couplers
and to reduce trapped modes
– Large (90mm) single chimney
sufficient for cw operation up to 50W
78mm 96 mm
Damper – (SS) Damper (Cesic)
Parameter Value
Active length (m) 1.038
RF frequency 1.3e9
R/Q (Ohms) 1000
Q0 1e10
Ea (MV/m) 10
Pcav (W) 10
Pbeam (kW) 100
Qext 1e6
QL*Rd/Q of HOM <1e6
Page 12
Injector Cryomodule
4K separator
2K separator
strongback
cavity
tuner
Power
coupler
Heat
exchanger
Houses
•one nine-cell 1.3GHz cavity
•Two 50kW power couplers
Features
•4K/2K heat exchanger with JT
valve on board
•Scissor tuner with warm motor
•LN2 thermal shield – 4K
thermal intercepts via syphon
•Two layers of mu-metal
•WPM alignment system Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 12
Page 13
Accelerator Cryomodule
• The ACM uses same basic design
as ICM but with two 1.3GHz nine cell
cavities each with two 50kW power
couplers
• There is one 4k/2k insert identical to
the ICM
• Physical dimensions
• L x H x W = 3.9 x 1.4 x 1.3 m
• 9 tons
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 13
2K phase
separator
cavity
Power
coupler and
support
WPM
bracket
4K-2K
Cryoinsert
Strongback
Heat
exchanger
4K phase
separator Support
Post
WPM
bracket Power coupler
Page 14
Cryogenics
• 4K liquid at 1.3 Bar delivered
in parallel to cryomodules
from supply dewar
• 4K levels are regulated by
LHe supply valve
• 2K levels are regulated by
JT valve in each CM
• 2K pressure is regulated by
2K exhaust valve on each
CM and trunk valve
upstream of SA pumps
4K – 1.3Bara
HP He
SA
Pumps
Compr.
MAIN
Supply
Dewar
Cold
Box
Heat
exchanger
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 14
Page 15
E-Linac RF Drive System
M
CAV CAV 1
M
M M
M
CAV 2
Klystron2 300kW
CAV1 CAV2
Klystron1 300kW
• For Phase I we
specify two 300kW
klystrons – one for
each cryomodule
• In the future one
300kW klystron will
drive ACM2
• we are looking for a
cost effective 1.3GHz
power source at
~150kW for the ICM
ICM ACM1
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 15
Page 16
Status and
Commissioning
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 16
Page 17
Progress
January 2014
July 2014
• Progress in the last year
• Cryogenics acceptance tests
complete
• E-Gun and LEBT installed and
commissioned – MEBT installed
• Two klystrons and HV supplies
installed and commissioned
• ICM assembled, installed and
commissioned
• ACM assembled and installed
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 17
Page 18
Electron Gun Status
• The electron gun and LEBT were
installed in February/March 2014
• Bias voltage of 325kV achieved
• 10mA cw achieved at 300kV
• Rf modulation with the ceramic
waveguide a success
• Macro pulsing demonstrated over a
broad range
• 100Hz-10kHz rep rates with duty
factors from 0.01-100%
• Transverse and longitudinal phase
space measured in LEBT
Ti Pierce Electrode Cu-Be Anode
SF6 Vessel Installed
350 kV, 16 mA HVPS Ceramic Waveguide
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 18
Page 19
LEBT Diagnostics
0 2 4 6 8 10 12
5.0
5.5
6.0
6.5
7.0
7.5
8.0
rms e
mitta
nce
[
m]
peak current [mA]
400 V
300 V
200 V
100 V
Deflector
Buncher
Screen
Solenoid
Diagnostic Box
rig Gun
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 19
• LEBT includes an
analyzing leg and
diagnostics to
characterize the gun
emittance and set the
matching for the ICM
• TM110 deflecting mode
cavity and high power
emittance rig
I=10mA
12.50
rms norm=7.6μm
Screen images downstream
of rf deflector show
manipulation of longitudinal
emittance with the buncher
cavity at different voltages.
E-Gun transverse and longitudinal emittance measurements
t
E
Deflector and
buncher off
V=0 V=0.33Vo
V=0.44Vo V=0.7Vo V=Vo
See THIOC02
Page 20
0
100
200
300
400
0 200 400 600 800 1000
Specified
Measured
Cryogenics installation
Parameter Contract Measured
Liquefaction 288 L/hr 367 L/hr
Refrigeration 600 W 837 W
Cryo load at 14MV/m
and 150% of
estimated static load
• 4K system
• ALAT LL Cold Box, KAESER
(FSD571SFC) main compressor
(112g/s), Cryotherm - distribution
• Acceptance tests (with LN2 pre-
cooling) exceed all specifications
with comfortable margins
• Sub-atmospheric pumping
• Four Busch combi
DS3010-He pumping
units specified and
installed (1.4g/s @
24mBar each)
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 20
Refrigeration (W) L
iqu
efa
ction (
l/hr)
Page 21
High Power RF Installation
• Now installed • Two CPI 290kW cw 1.3GHz
klystrons
• Two 600kW 65kV klystron power
supplies from Ampegon
• Each klystron reaches specification
at the factory
• At TRIUMF – tests were limited by
available load or circulator – one
was operated to 250kW cw the
other to 150kW cw
• Delivered a peak power of 25kW
into a cold cavity at low duty factor
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 21
Page 22
Power coupler conditioning
•Power coupler conditioning
•Condition two couplers at once at
room temperature using 30kW IOT
•Two 50kW CPI couplers installed
on waveguide box and power
transmitted to a dummy load
•Preparation involves extended
bakeout (five days) at 100C with N2
flowing
•RF conditioning in both TW (18kW
cw) and SW mode (10kW pulsed)
with adjustable short (five days)
Conditioning Stand
20
22
24
26
28
30
32
34
36
08:2
4:0
0
10:4
8:0
0
13:1
2:0
0
15:3
6:0
0
18:0
0:0
0
20:2
4:0
0
22:4
8:0
0
01:1
2:0
0
03:3
6:0
0
06:0
0:0
0
08:2
4:0
0
10:4
8:0
0
Time
Tem
per
ature
(C
elsi
us)
0.0E+00
1.0E+03
2.0E+03
3.0E+03
4.0E+03
5.0E+03
6.0E+03
7.0E+03
8.0E+03
9.0E+03
1.0E+04
1.1E+04
1.2E+04
1.3E+04
1.4E+04
1.5E+04
1.6E+04
1.7E+04
1.8E+04
1.9E+04
2.0E+04
Pow
er(W
)
output vacuum port(Celsius)
output warm bellow(Celsius)
output box(Celsius)
input vacuum port(Celsius)
input warm bellow(Celsius)
input cold window(Celsius)
output cold window(Celsius)
output inner bellow(Celsius)
output inner bellow 2(Celsius)
input inner bellow(Celsius)
input inner bellow 2(Celsius)
forward power(W)
Page 23
ARIEL Cavities - PAVAC
Cavity Preparation Status
ARIEL1 BCP120, Degas at FNAL Installed in ICM1
ARIEL2 BCP, Degas at FNAL, 120Bake, HF rinse Installed in ACMuno
ARIEL3 120micron BCP Vertical test
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 23
Page 24
ARIEL Cavities
1.0E+08
1.0E+09
1.0E+10
1.0E+11
0 2 4 6 8 10 12
after process
before process
20W
Q
Ea (MV/m)
ARIEL2
• Cavity vertical cold tests in
ISAC-II before and after re-
process
• Both cavities reach the
specified gradient of 10MV/m
but at Qo=6e9
• For Phase I we have lots of
cryogenic power so derate
specification to Qo=5e9
• Strategy is to utilize ARIEL1
and ARIEL2 to characterize the
cryo-engineering of the
cryomodules and use ARIEL3
to optimize the process. 12/05/2014 ACOT May 2014 - Laxdal 24
1.E+08
1.E+09
1.E+10
1.E+11
0 2 4 6 8 10 12
After degas
before degas
20W
Q
Ea (MV/m)
ARIEL1
Page 25
Cryomodule strategy
• Jacket and install ARIEL1 in ICM
• Jacket and install ARIEL2 and install in ACM
together with a dummy cavity • We call the single cavity ACM configuration
ACMuno
• ACMuno
• Dummy cavity has all interface features including
helium jacket and DC heater
• All helium piping and beamline interconnects will
be final
• ACMuno allows a full cryogenics engineering
test plus two cavity beam acceleration to 25MeV
• The goal is to install the cryomodules for a
combined beam test in Sept. 2014 – cryogenic
engineering and funding milestone
ACM
ACMuno
Dummy cavity Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 25
Page 26
ICM Assembly
ICM mock-up – 2013
ICM top assembly
• Mock-up assembly
of ICM used to test
parts and
procedures
• Final assembly
(aided by lessons
learned from mock-
up) - completed in
<1 month
ICM unit Complete (April 9, 2014) Top assembly into tank
Cavity hermetic unit (March 14, 2014)
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 26
Page 27
ICM Cold test
ICM craned into position ICM during cold test
Cold test complete Preparing cables and cryogenics
• ICM delivered to
cryogenic test area
• Established cool-down
protocol, vacuum
integrity and cryogenic
performance
• Tested thermal syphon
parameters
• Tuned couplers to
Qext~3x106
• Established cold
alignment
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 27
Page 28
ICM Move (April 28)
On April 28 the ICM was moved
from the clean room, craned
over ISAC-II hall, carted over to
proton hall loading bay, craned
down to e-hall and finally craned
into position, six weeks after
completion of the hermetic unit
ICM over ISAC-II
ICM on the move
Lowering ICM to the e-Hall ICM in position in the e-Hall
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 28
Page 29
10MeV Beam Test – June 2014
• 10MeV beam test
was an integration
test to validate
cryogenics, HLRF,
LLRF, e-Gun, LEBT,
ICM engineering and
synchronization
• The MEBT 10MeV
analysing leg served
as the destination for
the accelerated
beam
Page 30
Cold test results
50
60
70
80
90
100
0 5 10 15 20 25 30
2K Production efficiency
0
1
2
3
4
5
0 5 10 15 20 25 30
Heat Exchanger Temp. (degK)
Mass flow (g/s)
Ma
ss F
low
(g/s
) Te
mp
(K
) E
ffic
iency (
%)
Active Load (W)
Active Load (W)
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 30
Parameter Estimated Measured
4K static load (no syphon) 2 3
4K static load with syphon 6 6.5
2K static load 5 5.5
77K static load 100 <130
2K production efficiency 82% 86%
Syphon loop performance characterized –
works well – optimized in off-line cryostat tests
Early result – burst
disk works!
Page 31
ICM System Performance & Acceleration
• All systems functional • HLRF, LLRF, tuner, power couplers
• cavity phase lock is stable – couplers
balance – rf protection in place
• Confirmed tuning range – 400kHz
• Measured microphonics – very stable
• Successful acceleration achieved –
confirms rf integration and calibration -0.20
0.00
0.20
0.40
0.60
0.80
1.00
1.20
2.5 3 3.5 4 4.5 5 5.5 6
rf@600
rf@700
rf@800
rf@900
RF Calibration
E (MeV)
Microphonics
detuning spectra
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 31
Page 32
ICM Cavity Performance
• Q0 matches vertical test
so magnetic field
suppression is ok –
fundamental is not loaded
by the HOM dampers
• but …..
• gradient limited due to
strong field emission
• Detective work ensued
1.E+00
1.E+02
1.E+04
1.E+06
1.E+08
1.0E+08
1.0E+09
1.0E+10
1.0E+11
0 2 4 6 8 10
EINJ
Unjacketed
Field Emission
E (MV/m)
Qo
Fie
ld e
mis
sio
n X
-ra
ys
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 32
Page 33
Observations
1
10
100
1000
10000
100000
400 450 500 550 600 650 700 750
pos1
pos2
pos3
pos4
pos5
Expon. (pos2)
Expon. (pos3)
Expon. (pos4)
Expon. (pos5)
• Radiation measurements as
a function of monitor position
and rf set-point
• Results indicate that coupler
end of the cavity is the most
active by a factor of 5-10
• Further • Measurements of 7/9 and 8/9
fundamental modes suggest
that quench is in the end
groups
• Temperature sensors on
coupler side indicate some
heating during quench
Beam
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 33
Cavity field
X-r
ay flu
x
Page 34
Stainless steel HOM damper – coupler side
• Took ICM off line for
inspection
• Inspection revealed that
the SS damper tube that
fits inside the cavity at the
coupler end touched
down on the Nb cavity
causing scoring and
creating particulate
• Re-etched cavity and
assembled with added
support for HOM sub-
assembly
• ICM is now in re-
assembly and due on line
in two weeks
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 34
Page 35
ACMuno
ACMuno – ready for cooldown! Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 35
ACMuno assembly
proceeds through
June/July.
Page 36
Future
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 36
Page 37
• Present to Dec. 2014 – Continue beam tests at 25MeV up
to 100kW
• Early 2015 – Assemble a second ICM with
ARIEL3 and test in e-Hall as part of a collaboration with VECC
– Remove ACMuno and complete with ARIEL4
• 2018 – funding dependent – Complete second accelerating
module (ACM2) to complete e-Linac
– Fabricate, process and test two more cavities
– Install 150kW RF system for ICM
37
ARIEL e-Linac Completion
Page 38
Summary
• The ARIEL e-Linac initial phase is nearing
completion • Cryogenic, rf and service installations
complete
• The 300kV E-Gun has met specification • being used presently to commission the
LEBT and MEBT
• The ICM initial cold tests demonstrated the
cryo-engineering matches specifications • a problem with the coupler side damper tube
reduced performance
• The ACMuno is on-line and cryogenic tests
will begin this week • The second cavity will be added after the
cryo-engineering is confirmed and initial
beam commissioning with ICM and ACM is
complete
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 38
Page 39
Owned and operated as a joint venture by a consortium of Canadian universities via a contribution through the National Research Council Canada
Propriété d’un consortium d’universités canadiennes, géré en co-entreprise à partir d’une contribution administrée par le Conseil national de recherches Canada
Canada’s national laboratory for particle and nuclear physics Laboratoire national canadien pour la recherche en physique nucléaire
et en physique des particules
Thank you!
Merci!
TRIUMF: Alberta | British Columbia | Calgary |
Carleton | Guelph | Manitoba | McGill | McMaster |
Montréal | Northern British Columbia | Queen’s |
Regina | Saint Mary’s | Simon Fraser | Toronto |
Victoria | Winnipeg | York
Thanks, Merci
Page 40
ICM / ACM Cryogenic Circuits
June 14, 2011 CEC-ICMC Spokane - R. Laxdal - Injector Cryomodule for e-Linac 13
4K Supply
4K Return
2K Return
77K Supply 77K Exhaust
4K Cooldown
JT Valve
Cooldown Valve
4K intercepts
syphon circuits
• 4K/2K insert designed to fit in a
separate test cryostat prior to
cryomodule assembly
• One 4K circuit feeds the heat
exchanger and JT valve for 2K
supply
• One 4K circuit feeds the bottom of
the cold mass through a cooldown
valve for initial cooling
• One 4K circuit cools thermal intercepts via a self regulating
thermal syphon circuit – flow is governed by the heat load
and the LHe level in the 4K reservoir
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 40
Page 41
Syphon Loop
Sept. 1, 2014 MOIOC01 - Laxdal - TRIUMF e-Linac 41
Syphon loop • Demonstrated that the loop
turns on and off depending on
head pressure and heat load –
self regulating
• Existence of flow is diagnosed
by a temperature sensor on
the return column
• Demonstrated that the heat
load to 2K is effectively
intercepted by syphon loop
cooling
• Lesson learned – beware of
creating convection in the 4K
reservoir – heat source
Q
m mm
Page 42
RF Protection
60(opt)ext
opt
101Q so
10000 0at and 10000
W10,kW100 MV/m,10 mA,10
regime operating Typical
opt
cav
beam
cavbeama
Q
bP
Pb
PPEI
)(100J/msec matched be lcavity wil The :ie
1e6 to1e10 from go willQ
100kW20n
10m0.25W
0.25W 40
W10 So
normal. goes cavity) of 40
1 (ie
cell one of 25% assume :estimationQuench
0
Normalequench zon
SCequench zon
P
P
• Need fast protection to kill the rf in
case of rf transient • PMTs on couplers
• Quench detection circuit
• Developed transient model for
quench • COMSOL – quench zone analysis
• Mathematica model used to study
rf transients
• Conclude that for all beam duty
factors the cavity gradient gives the
cleanest indication of a quench
• 42