Space Charge Issues in the SNS Linac and Ring

clanmurderUrban and Civil

Nov 15, 2013 (4 years and 8 months ago)

420 views

Space Charge Issues in the SNS
Linac and Ring

by

Sarah Cousineau,

on behalf of the SNS project


Oak Ridge National Laboratory, USA


Coulomb05 Workshop,

Senigallia, Italy, September 12


16, 2005

The SNS Accelerator Complex

Oak Ridge Accelerator
Systems:

Integration,
Installation, Commissioning,
Operation

At peak : ~500 People

worked on the construction

of the SNS accelerator

~177 M$


~60 M$

~113 M$

~20 M$

~63 M$

~106 M$

The Spallation Neutron Source


The SNS is a short
-
pulse neutron source
with a single
-
purpose
mission of neutron
science, under
construction at
ORNL


SNS will be the
world’s leading
facility for neutron
scattering research


The peak neutron
flux will be ~20

100x
ILL


The SNS will begin
operation in 2006


SNS is funded
through DOE
-
BES
and has a Baseline
Cost of 1.4 B$


It will be a short drive
from HFIR, a reactor
source with a flux
comparable to the
ILL


SNS Design Parameters


Average proton power on target:
1.44 MW



Beam energy:
1 GeV


Pulse parameters:
1
-
ms pulse, 60 Hz

repetition rate (6% duty)


Beam current:


26 mA average macropulse

current


38 mA peak H
-

current


1.6 mA average linac beam

current


Ring accumulation:
1 ms

pulse compressed
to
695 ns in 1060

injected
turns


Ring intensity:
1.5x10
14

protons

945 ns

period

Chopping structure

38

mA

26

mA

1 ms long

Macropulse structure

20 to 50
m
s

ramp

time

time

mini
-
pulse

SNS Accelerator Complex

945 ns

1 ms macropulse

Current

mini
-
pulse

H
-

stripped
to protons

Current

1ms

Front
-
End:
Produce a 1
-
msec long,
chopped,
low
-
energy
H
-

beam

LINAC:
Accelerate
the beam to
1 GeV

Accumulator Ring:
Compress 1 msec
long pulse to 700 ns

Deliver
beam to
Target

Chopper
system makes
gaps

Ion Source

2.5 MeV

1000 MeV

86.8 MeV

CCL

SRF,
b
=0.61

SRF,
b
=0.81

186 MeV

387 MeV

DTL

RFQ

Ion Source Beam Distributions


H
-

ion source, capable of >38 mA and 60 Hz operation.


Emittance measurements show that the beam leaving the ion source
has “wings” or tails; caused by optical nonlinearities in electrostatic
lenses.


Measured distributions used to generate input distribution for
simulations.

1.0
3.0
5.0
7.0
9.0
-30.0
-15.0
0.0
15.0
30.0
45.0
60.0
75.0
x
x'
0.0
2.0
4.0
-41.0
-20.0
1.0
22.0
43.0
64.0
y
y'
Horizontal

Vertical

MEBT Halo Scrapers



MEBT scrapers allow cleaning
of beam tail coming from source.

No scraping

With scraping

Matching the lattice with space charge


Linac lattice must be matched for a specific current.


Challenge during commissioning runs, when beam current and
quad settings were often changing.

SC Matched case:
20mA beam

with
20mA lattice
.

SC Mismatched case:
38mA beam

with
20mA lattice
.

MEBT + DTL

MEBT + DTL

Linac Beam Profile Measurements

DTL3 wirescanner


best matched case

DTL3 wirescanner


various quad settings


Measured profiles as function of last two MEBT quad setpoints.


Even nominal (design) cases shows some “halo” (could be ion source tails).
Here we define halo as any non
-
Gaussian tails.


See that non
-
gaussian tails can be “tuned” by matching into DTL1.

Linac Beam Profile Measurements



Main commissioning concern is rms beam size, not halo.



Preliminary studies underway with Parmilla to investigate halo for future operation.



Benchmarked results: Qualitatively good (trends agree), quantitative fair.

Example benchmark of DTL3 wirescan data

Linac Beam Profile Measurements


Small halo seen for design case, large halo for mismatched
case.


Halo seems to be large in DTL, but dies of in CCL.


Simulations show this is because core grows and consumes
halo in CCL.


Design quad settings

Mismatched quad settings

Mismatched Beam Simulations


For design lattice settings,
previous studies show
strong dependence on
initial distributions.



For mismatched case, final
distribution out of warm
linac is independent of the
initial distribution.


MEBT

DTL5

CCL4

DTL3

Nominal

Parabolic

Other Linac High Intensity Beam Challenges


Beam loading a major issue at high intensity (
≥ 20 mA)


Adaptive feed forward necessary for good bunching, low losses.

Beam loading prevented normal
acceleration of beam with >20mA peak
current

Beam loading effect eliminated by
means of Adaptive Feed Forward.

SCL Commissioning Results


4K commissioning run: Reached 860 MeV, 180us, 20mA on Aug 21


4 low energy SCL cavities out of tuning range at 4K.


Missing cavities lead to high losses in transition region from doublet to
FODO structure.

Commissioning the SCL relied on tuning on losses!

SCL Commissioning Results


2K commissioning run: Reached 910 MeV on Aug 30
th
!


Losses way down with missing upstream cavities included.

Beam is Ring Ready!

SNS Ring: Loss
-
loss design philosophy


The ring was designed with a low loss philosophy.


Designed centered around mitigating losses from:


Injection.


Extraction.


Space charge.


Other collective effects: impedances, e
-
p, etc.

Ring will require uncontrolled losses

0.01% of the total
beam intensity.

600 Turns

200 Turns

1060 Turns



Injection painting scheme
optimized to minimize space
charge.



Paint with hole in the center to
help create uniform density.



Also try to keep circulating beam
foil intercepts to a minimum (~6
foil hits per proton).

No Space Charge


1060 Turns

Phase
-
Space Painting with Space
-
Charge

Horizontal Phase
-
Space: P
x

vs. X

Phase
-
Space Painting with Space
-
Charge

Real Space: Y vs. X

600 Turns

200 Turns

1060 Turns

Correlated painting scheme chosen over anti
-
correlated because:



Smaller number of foil hits.



Less space charge halo observed in simulation.



Footprint suites target requirements
.

No Space Charge


1060 Turns

Lattice Tune Chosen to Avoid Resonances



Design lattice tune for 1.4 MW operation: Q
x
=6.23, Qy=6.20.



Intensity limitation for this tune is half
-
integer coherent resonance.



Chromaticity adds another

Q = 0.07 in spread.



Sextupoles in ring for correcting chromaticity.

N=0.5*10
14



263 turns

N=1.0*10
14



526 turns

N=2.0*10
14



1052 turns

SC Tune Footprint

ORBIT Simulation of Baseline Accumulation Scenario



Beam broadening from space charge observed:


Paint to


= 165

, space charge broadens to 175



Emittance Distributions

Emittance (


mm mrad)


Fraction larger than emittance

No Space Charge

With Space Charge

No Space Charge

With Space Charge

Alternative Lattice Tune



Alternative lattice tune studied: Q
x
=6.4, Qy=6.30.

SC Tune Footprint

Crosses resonances
:

3Q
x

= 19; normal sextupole

2Q
x

+ Q
y

= 19; skew sextupole

Sextupole correctors available in
the ring for correcting sextupole
errors.

Anticipated Loss Distribution in the SNS Ring



Space charge induced beam halo will be intercepted by collimation system.



Final loss distribution determined by collimation system (except for injection,
extraction losses).



Losses > 1 W/m in collimation straight.

Simulated Loss Pattern in Ring

The Remaining Commissioning Schedule

Jan/05

Jan/06

Jan/07

Linac, to linac dump

25/Jul


Sept/05

HEBT/Ring/RTBT, to extr. dump

2/Jan


19/Feb/06 (47 days)

CD
-
4 deadline

30/Jun/06

RTBT, to target

1/Apr


28/Apr/06

Compared to original plan, ring commissioning will be
performed within a much smaller time frame and with a
very reduced suite of diagnostics
.



Post


CD4 Intensity Ramp
-
Up

Accelerator Availibility and Operation
3
50
150
300
800
1200
1400
1400
1400
1400
50
75
80
85
90
91
92
93
94
95
0
500
1000
1500
2000
2500
3000
'06 2
'07 1
'07 2
'08 1
'08 2
'09 1
'09 2
'10 1
'10 2
'11 1
Years
Ops Hours
40
60
80
100
%
User Ops hours
Accel. Phyics
Beam Power/kW
Reliability (%)

We will commission
the beam with low
intensity, ~2
×
10
13
ppp
(10mA, 1 Hz).



We will ramp up beam
power gradually.




Should reach 1.4 MW
by 2010.



Plans for second
target station in
~2010.


SNS Power Upgrade

1.107
1.098
1.058
Ring rf frequency [MHz]
0.2
0.15
0.15
Ring space
-
charge tune spread,

Q
sc
683
691
695
Pulse length on target [ns]
70
70
68
Chopper beam
-
on duty factor [%]
6.0
6.0
6.0
Linac beam macro pulse duty factor [%]
65
42
26
Average
macropulse
H
-
current [mA]
92
59
38
Peak Current from front end system
3.9
2.5
1.6
Linac average beam current [mA]
5.0
3.0
1.4
Beam power on target,
P
max
[MW]
12 + 8 (+1 reserve)
12 + 8 (+1 reserve)
12
SRF
cryo
-
module number (high
-
beta)
1.6
1.0 / 1060
35 (+2.5/
-
7.5)
27.5 (+/
-
2.5)
33+48
11
1000
Baseline
2.5
1.0 / 1100
31
27.5 (+/
-
2.5)
33+80 (+4 reserve)
11
1300
Upgrade
3.8
1.0 / 1110
34
27.5 (+/
-
2.5)
33+80 (+4 reserve)
11
1400
Ultimate
Ring bunch intensity [10
14
]
Ring injection time [ms] / turns
Kinetic energy,
E
k
[MeV]
Peak gradient,
E
p
(
b
=0.81 cavity) [MV/m]
Peak gradient,
E
p
(
b
=0.61 cavity) [MV/m]
Number of SRF cavities
SRF
cryo
-
module number (med
-
beta)
1.107
1.098
1.058
Ring rf frequency [MHz]
0.2
0.15
0.15
Ring space
-
charge tune spread,

Q
sc
683
691
695
Pulse length on target [ns]
70
70
68
Chopper beam
-
on duty factor [%]
6.0
6.0
6.0
Linac beam macro pulse duty factor [%]
65
42
26
Average
macropulse
H
-
current [mA]
92
59
38
Peak Current from front end system
3.9
2.5
1.6
Linac average beam current [mA]
5.0
3.0
1.4
Beam power on target,
P
max
[MW]
12 + 8 (+1 reserve)
12 + 8 (+1 reserve)
12
SRF
cryo
-
module number (high
-
beta)
1.6
1.0 / 1060
35 (+2.5/
-
7.5)
27.5 (+/
-
2.5)
33+48
11
1000
Baseline
2.5
1.0 / 1100
31
27.5 (+/
-
2.5)
33+80 (+4 reserve)
11
1300
Upgrade
3.8
1.0 / 1110
34
27.5 (+/
-
2.5)
33+80 (+4 reserve)
11
1400
Ultimate
Ring bunch intensity [10
14
]
Ring injection time [ms] / turns
Kinetic energy,
E
k
[MeV]
Peak gradient,
E
p
(
b
=0.81 cavity) [MV/m]
Peak gradient,
E
p
(
b
=0.61 cavity) [MV/m]
Number of SRF cavities
SRF
cryo
-
module number (med
-
beta)
SNS Power Upgrade Technical Issues



Accelerator Issues


Cryomodule Acquisition Approach



Nine (9) HB Cryomodules


SRF Facility to Support Acquisition Approach


Full Production vs Maintenance/Testing


Front End/Ion Source R&D



Reliability/Higher Current


75 mA


Ring Injection Issues Dump Upgrade



Foil issues: Need new material, multiple foils, or laser stripping.


New Magnets: Scaled for 1.3 GeV


Injection Dump: Capacity of 150kW may need upgrade to 300kW.


Target Issues


Target module designed for 1 MW


R&D (Bubble Injection) to extend to > 2 MW


Accelerator Physics R&D projects


Laser stripping proof of principle experiments.


Active feedback system experiments at PSR.



Active Feedback System for Upgrade



Active feedback system
planned for intensity upgrade.



Can stabilize e
-
P and other
instabilities resulting from
collective effects
.



Proof
-
of
-
principle
experiments done at PSR in
spring, 2005 (SNS, Argonne,
LANL, Indiana University
collaboration).

Active Feedback System

Pick
-
up

Kicker

Circulating beam

Courtesy
C. Deibele

First results of e
-
P feedback experiments

Courtesy S.
Henderson

Results summary:



Instability suppression
observed.



During normal operation,
high RF voltage used to
suppress instability.



With damper on, RF
voltage could be reduced
by 13% to 18%.



System still needs
optimization.

Summary



SNS is on track for completion in 2006.




So far, SNS warm and superconducting linac has been
commissioned.
All major beam commissioning milestones
have been met.




Have observed some space charge and high intensity affects
in during linac commissioning.




Space charge effects will become more apparent during
ramp up to high intensity.




SNS has been approved for a beam power upgrade to 3 MW
beginning in ~2010.