Loss Limitations and
Collimation
Angelika Drees, R. Fliller, W. Fu
The RHIC Collimation Sytem:
history and overview
RHIC Loss Limitations
Operational limits
Beam Dumps/Quench Limits
Soil Activation
Exp. Background
Summary
RHIC Collimator Configuration
RHIC was originally built with a 1-stage
collimation system only:
1 dual plane h/v scraper with 45 cm copper
jaws, linear motion in both planes, skew motion
only in horizontal
1 bent crystal collimator for studies in 1 ring
(yellow) only
The system was upgraded after the 2003
run because of high experimental
backgrounds and gap cleaning demands.
Crystal approach proved non sufficient.
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
RHIC overview: collimation system 2004 and upgrade
New blue 2ndary vertical
collimator (v2)
(capped area)
New yellow 2ndary
vert. Collimator (v2)
2000-2003:
1-stage system
including bent
crystal in 1 ring
2004:
Traditional 2-stage
system with 2
horizontal and 1
vertical secondary
collimators
2005:
Traditional 2-stage
system with 2
horizontal and 2
vertical secondary
collimators
Collimator Section Layout
In the shutdown
2003-2004 the
collimation system
was upgraded to a
conventional 2stage system
including new
individual
secondary
collimators for
both planes. The
new system was
first used in the
run 2004 for both,
Au and protons.
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Collimator Design
45 cm copper jaws
One side only
Rotatable, positioning: few mm
HH
V
cross-section of the primary collimator (dual plane)
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Crystal Collimation Attempt during
Fill 3061 (Au)
… If you want to know why we pulled out the crystal and installed traditional secondaries …
• STAR “Yellow Halo” signal during Crystal Collimation attempt
scraper is closer to the beam toward the bottom of graph
crystal is at 13.6 mm from beam center and channeling
• Scraper alone is more effective than the crystal and scraper together
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Loss Limitations
 Operational limit: keep allowable loss budget
(radiation safety), monitored hour by hour
 Quench limit:
magnet quenches due to accidental local losses during
ramp/store => BLMs
magnet quenches at beam dump due to debunched
beam => gap cleaning
 Soil activation (not under radiation protection),
depends on integrated yearly losses
 Experimental backgrounds: need ‘clean’ beams to
allow good signal/noise ratio in experiments and
keep false trigger rate small (dead time)
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Operational Limits: Monitoring of
Beam Loss/hour
dirty dump
alarm level is reset
max.
alarm level
1 hour
After every dump (operational or accidental) all RHIC loss monitors are analyzed. If
dump appears to be “dirty”, i.e. unmasked loss monitors register losses above a
certain level, injection into RHIC will be blocked for 1 hour (see figure).
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Loss Limitations
 Operational limit: keep allowable loss budget (radiation
safety), monitored hour by hour
 Quench limit:
magnet quenches due to accidental local losses during
ramp/store => BLMs
magnet quenches at beam dump due to debunched beam => gap
cleaning
 Soil activation (not under radiation protection), depends
on integrated yearly losses
 Experimental backgrounds: need ‘clean’ beams to allow
good signal/noise ratio in experiments and keep false
trigger rate small (dead time)
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Accidental Local Losses: BLM thresholds
Magnet quenches can be
caused by fast (ms) type
losses.
=> BLM trip levels
(thresholds)
• fill 4198
• fill 4118
Continuous scraping/losses
cause magnet quenches as
well. Trip levels don’t help.
 software integration
 or collimators
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Collimators as limiting Aperture
during Store
It is difficult to maintain limiting aperture position at all times – especially during ramp!
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Collimator use during the Ramp
• Losses around
RHIC on 2 ramps
during FY01
top: beam scraping
at abort kickers
center: scrapers
inserted
bottom: lattice
• Using the
collimators reduced
losses at abort
kickers by x100
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Loss Limitations
 Operational limit: keep allowable loss budget (radiation
safety), monitored hour by hour
 Quench limit:
magnet quenches due to accidental local losses during
ramp/store => BLMs
magnet quenches at beam dump due to debunched beam => gap
cleaning
 Soil activation (not under radiation protection), depends
on integrated yearly losses
 Experimental backgrounds: need ‘clean’ beams to allow
good signal/noise ration in experiments and keep false
trigger rate small (dead time)
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Gap Cleaning: Motivation and
Method
 gap cleaning is necessary due to the extensive
debunching of HI beams (quenching risk)
 method:
debunched beam is excited transversely (continuously
during the store: 1Hz) using damping kickers
the collimation system absorbs the large amplitude
particles (in addition to halo)
 debunched beam should be lost in collimator
area, efficiency relies on collimator performance
to avoid increase of exp. backgrounds
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Excitation frequency for Cleaning
default procedure:
excite around
betatron tune of
bunched beam
(measured
automatically at the
end of ramp)
optional procedure:
perform automated
tune scan to find
resonant tune of
debunched beam and
excite with this
frequency
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Cleaning ON vs. cleaning OFF:
56 x 56
fill 4471, Feb. 06 2004
start of cleaning
~5 109
yellow:
cleaning off,
started around 3:30
to allow clean beam
dump
blue:
cleaning on,
debunched beam is
continuously excited
and absorbed by
collimators
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
1 Hz excitation on standard BLMs:
Normal cleaning levels are < 1% of trip
Level => acceptable
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Remaining debunching with Cleaning ON:
Debunching rate clearly
correlated with bunch
current at the beginning of
the store. For 4h store
length, an upper limit of
1.4*109 ions per bunch
results to maintain less than
5*109 debunched beam at
the end of store (for safe
beam dump).
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Loss Limitations
 Operational limit: keep allowable loss budget (radiation safety),
monitored hour by hour
 Quench limit:
magnet quenches due to accidental local losses during ramp/store =>
BLMs
magnet quenches at beam dump due to debunched beam => gap cleaning
 Soil activation (not under radiation protection), depends on
integrated yearly losses
RHIC has to operate such that leachate (from activated soil)
cannot contain concentrations of 3H or 22Na that exceed 5% of
the drinking water standard
Method: either cap the collimator area (prevent leaching) or use
removable soil samples inside tunnel near collimators/dumps to
monitor activation level. 22Na is easily measured.
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Fill 4581: Au ions (100 GeV), blue gap cleaning was off
The debunched beam is lost at
the collimators:
deb.rate = 2*10^7 ions/min
loss rate (total beam) is:
loss rate = 3*10^7 ions/min
(lost anywhere: triplets, dump ...)
total: 2-5*10^7 ions/min
Fill 4471, Feb. 06 2004, Au ions, 100 GeV
The debunched beam is lost at
the collimators:
deb.rate = 2.3*10^7 ions/min
loss rate (total beam) is:
loss rate = 3*10^7 ions/min
total (at collimators):
2.5-5*10^7 ions/min
Fill 4870, Mar 24 2004, Au ions, 100 GeV
The debunched beam is lost at
the collimators:
deb.rate = 2.4*10^7 ions/min
loss rate (total blue beam) is:
3*10^7 ions/min
loss rate (total yellow beam) is:
8*10^7 ions/min
total (at blue collimators):
2.5-5*10^7 ions/min
total (at yellow collimators):
2.5-10*10^7 ions/min
Beam losses with proton beams
Typical “bad” store:
Loss rate: 10 10^11 = 5 10^9 p/min
Assumptions for Soil Activation
Calculation
 Au Operation:
total loss/min (from gap cleaning): 2*107 ions/min
assume all is lost at collimators, 10% at new V2
-> 2*106 ions/min during the store
assume average of 100 h/week at store (best week last run)
add 0-3 106 ions/min @ v2 => 2-5*106 ions/min
 proton Operation:
total loss/min (from bad lifetime store during high intensity
run): 5*109 p/min
assume all is lost at collimators, 10% at new v2
-> 5*108 p/min during the store
assume average of 80 h/week at store (best avg. last run)
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Calculate Soil Activation (Na-22) from Loss
Data
Calculate for v2:
given the assumptions and using an estimation with the bare collimator/no
shielding (there is some concrete shielding @ v2), for Na-22, assuming 15
weeks for each of the Au and proton ions (ie. 30 weeks in total) we get:
8.7 (18) pCi/l and 11.0 pCi/l
for the case of Au-ions and protons respectively. This adds up to
19.7 (29) pCi/l (after 1year),
which is just the same (or x1.5 of) the 5% drinking water limit (20 pCi/l)
for Na-22.
Compare with Measurement (from soil sample):
activation level at primary collimator calculated: 200 pCi/l.
measured: 13 pCi/l
assuming 100% of all losses at primary collimator is conservative
the 15+15 weeks of running is conservative
the operating efficiency is overestimated
distance collimator-soil is underestimated
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Loss Limitations
 Experimental backgrounds: need ‘clean’ beams to allow
good signal/noise ratio in experiments and keep false
trigger rate small (dead time).
 Collimator positioning should be reliable and efficient for
background reduction
 Time at the beginning of the store is precious because
this has the highest luminosity
-> Collimation and positioning of collimators should be FAST
(and precise!)
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Collimation during Fill 1759 (Au)
• FY01 run has low
beam current
intensities
• 1-stage system
• only PHENIX
benefits (some)
from collimation
• there is generally
no/little need for
collimation during
the FY01 run
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Collimation during Fill 3094 in FY03
(d-Au)
• All experiments but
PHOBOS benefit
from collimation
• vertical scraper
retraction (vert.
lines) clearly raises
background
• reduction rates are
between 2 and 5
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Automatic Steering Algorithm
RHIC has 5 jaws per ring, most
allow both, linear motion and
angular motion (to parallelize
with beam). Potentially time
consuming!
=> 18 degrees of freedom
(+ 4 more next run)
Requires automation (3 steps):
1. Move to STDBY position
(based on BPM readings)
2. Move Closer to beam
(based on loss monitor
feedback, serial)
3. Remove Halo/Store
(based on lattice functions,
parallel)
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Collimation during Fill 4854 (Au) in
the blue ring
Serial collimator
steering (mode: Move
Closer), following
parallel mode does not
improve backgrounds.
Vertical lines denote
when each collimator
moves. Background
improvement approx.
x6.
Note: secondary
vertical collimator
quite efficient.
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Collimation during Fill 4436 (Au)
Parallel collimator
steering (mode: “Store”),
using lattice fct. and
assumed emittances.
Vertical lines denote
when collimators start
and stop moving
simultaneously.
Background improvement
approx. x10.
Note: procedure stops
automatically when
desired background
levels are reached.
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Compare RHIC data from 1-stage and 2-stage collimation system
Fill 3254
1-stage
Fill 4854
2-stage
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Comparison of Background Reduction
Rates
Average ratio of uncollimated to collimated background for the STAR
and PHENIX detectors (sensitive to beam direction) over 6 stores in
2003 and 2004. In both cases PHOBOS had a b* of 3m.
Background
signal
d-Au 2003
Au-Au 2004
b* = 3m
5
b*=3m
11
STAR Yellow
1
3
PHENIX Blue
4
11
PHENIX Yellow
1
4
STAR Blue
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004
Summary
 Operational Limit
about 1 normal ‘fill’ /hour in uncontrolled areas
 Magnet Quenches
store/ramps: BLM thresholds pull permit (fast and
slow modes)
beam dump: continuous gap cleaning
 Soil activation
maintain 5% of DWS (i.e. 20 pCi/l 22Na) by caps or
soil sample monitoring in uncapped areas
 Exp. Background
traditional 2-stage collimation system achieves
average background reduction of x10
Angelika Drees
ICFA-HB2004 Workshop, Bensheim, Oct. 18, 2004