Upload
arnold-wilkins
View
214
Download
0
Embed Size (px)
Citation preview
Recent DHCAL Developments
José Repond and Lei XiaArgonne National Laboratory
Linear Collider Workshop 2013University of Tokyo
November 11 – 15, 2013
J. Repond - Imaging Calorimeters
2
The DHCAL
Description
54 active layers Resistive Plate Chambers with 1 x 1 cm2 pads → ~500,000 readout channels Main stack and tail catcher (TCMT)
Electronic readout
1 – bit (digital) Digitization embedded into calorimeter
Tests at FNAL with Iron absorber in 2010 - 2011
Tests at CERN with Tungsten absorber 2012
1st time in calorimetry
J.Repond DHCAL
3
DHCAL Construction Fall 2008 – Spring 2011
Resistive Plate Chamber
Sprayed 700 glass sheets Over 200 RPCs assembled → Implemented gas and HV connections
Electronic Readout System
10,000 ASICs produced (FNAL) 350 Front-end boards produced
→ glued to pad-boards 35 Data Collectors built
6 Timing and Trigger Modules built
Assembly of Cassettes
54 cassettes assembledEach with 3 RPCs
and 9,216 readout channels
350,208 channel system in first test beam Event displays 10 minutes after closing enclosure
Extensive testing at every step
J.Repond DHCAL
4
Testing in BeamsFermilab MT6
October 2010 – November 2011 1 – 120 GeV Steel absorber (CALICE structure)
CERN PS
May 2012 1 – 10 GeV/c Tungsten absorber (structure provided by CERN)
CERN SPS
June, November 2012 10 – 300 GeV/c Tungsten absorber
Test Beam Muon events Secondary beam
Fermilab 9.4 M 14.3 M
CERN 4.9 M 22.1 M
TOTAL 14.3 M 36.4 M
A unique data sample
RPCs flown to GenevaAll survived transportation
J.Repond: DHCAL
5
Recent developments
Improved Resistive Plate Chambers
1-glass design High-rate RPCs
High voltage distribution system
Gas recirculation system
Resistive paint
Resistive paint
Mylar
1.2mm gas gap
Mylar Aluminum foil
1.1mm glass
1.1mm glass
-HV
Signal padsG10 board
Typical RPC design
6
1-glass RPCs
Offers many advantages
Pad multiplicity close to one → easier to calibrate Better position resolution → if smaller pads are desired Thinner → t = tchamber + treadout = 2.4 + ~1.5 mm → saves on cost Higher rate capability → roughly a factor of 2
Status
Built several large chambers Tests with cosmic rays very successful → chambers ran for months without problems Both efficiency and pad multiplicity look good
Efficiency Pad multiplicity
J. Repond - The DHCAL
7
Rate capability of RPCs
Measurements of efficiency
With 120 GeV protons In Fermilab test beam
Rate limitation
NOT a dead time But a loss of efficiency
Theoretical curves
Excellent description of effect
Rate capability depends
Bulk resistivity Rbulk of resistive plates (Resistivity of resistive coat)
Not a problem for an HCAL at the ILCB.Bilki et al., JINST 4 P06003(2009)
J.Repond: DHCAL
8
10-18
10-15
10-12
10-9
10-6
10-3
100
103
106
109
1012
Fu
sed
qu
arz
Tefl
on
Poly
eth
yle
ne
Am
ber
Para
ffin
Ru
bb
er
Su
lfu
rM
ica
Porcela
inS
tyro
foam
Ru
tile
(TiO
2)
Ru
tile
(TiO
2)(
Z)
Gla
ssM
arb
le Bak
eli
teD
isti
lled
wa
ter
Fu
sed
sil
ica
Beryll
ium
ox
ide(B
eO
)+In
trin
sic s
ilic
on
Sa
pp
hir
e+
Ferrit
e(F
e2O
3)+
Dry s
oil
Ple
xig
las+
Gall
ium
arse
nid
e (
GaA
s)+
Intr
insi
c g
erm
an
ium
Tell
uriu
mC
arb
on
Grap
hit
eC
ast
iron
Mercu
ry
Nic
hro
me
Sil
icid
es
Germ
an
sil
ver
Sil
icon
ste
el
Poly
sili
co
nL
ead
Tan
talu
mT
inP
lati
nu
mIr
on
Nic
kel
Zin
cT
un
gst
en
Brass
Alu
min
um
Gold
Cop
per
Sil
ver
Material
(S
)
C. Pecharromán X. Workshop on RPC and related Detectors (Darmstadt)
Here is the problem
There is a gap between 10-9 and 10-3
J.Repond: DHCAL
9
Available resistive plates
108 109 1010 1011 1012 1013102
103
104
105
106
Floatglass
Semi-conductive
glass
Ceramics
Streamer mode
Warm glass
BeijingCBM Requirement
Beijing
INR+CBM
lip Coimbra
ALICE-muon
LHCb
ATLAS
Warsaw
CMS-forward
CMS-barrel
CERN+Bologna
CERN+Rio
Dresden
STAR
ALICE-TOF
Lip+USC
Max
Counting R
ate(
Hz/
cm2)
Volume Resistivity(cm)
J.Repond: DHCAL
10
Where to use high-rate RPCs
ILC – Hadron calorimeter (close to beam pipe)
CLIC – Hadron calorimeter (forward direction – 2γ background)
CMS – Hadron calorimeter (forward direction)
Current forward calorimeters inadequate for high-luminosity running
PbWO4 Crystals Scintillator/Brass + Quartz fibers/Steel
To start in year ~2023Luminosity of 5 x 1034 cm-2
(> x10 higher than now)
J.Repond: DHCAL
11
High-rate Bakelite RPCsBakelite does not break like glass, is laminated but changes Rbulk depending on humidity but needs to be coated with linseed oil
Gas inlet Gas outlet
Gas flowdirection
Fishing line
Sleeve aroundfishing line
Additional spacer
Use of low Rbulk Bakelite with Rbulk ~ 108 - 1010 and/or Bakelite with resistive layer close to gas gap
Several chambers built at ANL
Gas
Resistive layer for HV
J.Repond: DHCAL
12
High-rate Bakelite RPCsBakelite does not break like glass, is laminated but changes Rbulk depending on humidity but needs to be coated with linseed oil
Gas inlet Gas outlet
Gas flowdirection
Fishing line
Sleeve aroundfishing line
Additional spacer
Use of low Rbulk Bakelite with Rbulk ~ 108 - 1010 and/or Bakelite with resistive layer close to gas gap
Several chambers built at ANL
Gas
Resistive layer for HV
Noise measurement: B01
1st run at 6.4 kV Last run, also 6.4kV, RPC rotated 900
Readout area
Fishing lines
(incorporated resistive layers)
J.Repond: DHCAL
14
Noise measurementsApplied additional insulationRate 1 – 10 Hz/cm2 (acceptable)Fishing lines clearly visibleSome hot channels (probably on readout board)No hot regions
Cosmic ray testsStack including DHCAL chambers for trackingEfficiency, multiplicity measured as function of HVHigh multiplicity due to Bakelite thickness (2 mm)
J.Repond: DHCAL
15Tests carried out by University of Michigan, USTC, Academia Sinica
GIF Setup at CERN
J.Repond: DHCAL
16
First results from GIF
Source on
Source off
Background rate
Absolute efficiency not yet determined
Clear drop seen with source on
Background rates not corrected for efficiency drop
Irradiation levels still to be determined (calculated)
J.Repond: DHCAL
17
Development of semi-conductive glass
Co-operation with COE college (Iowa) and University of Iowa
World leaders in glass studies and development
Vanadium based glass
Resistivity tunable Procedure aimed at industrial manufacture (not expensive)
First samples
Very low resistivity Rbulk ~ 108 Ωcm
New glass plates
Rbulk ~ 1010 Ωcm produced Plates still need to be polished Production still being optimized
18
High Voltage Distribution System
Generally
Any large scale imaging calorimeter will need to distribute power in a safe and cost-effective way
HV needs
RPCs need of the order of 6 – 7 kV
Specification of distribution system
Turn on/off individual channels Tune HV value within restricted range (few 100 V) Monitor voltage and current of each channel
Status
Iowa started development First test with RPCs encouraging Work stopped due to lack of funding Size of noise file
(trigger-less acquisition)
19
Gas Recycling SystemDHCAL’s preferred gas
Development of ‘Zero Pressure Containment’ System
Work done by University of Iowa/ANL
Status
First parts assembled…
Gas Fraction [%] Global warming potential (100 years, CO2 = 1)
Fraction * GWP
Freon R134a 94.5 1430 1351
Isobutan 5.0 3 0.15
SF6 0.5 22,800 114
Recycling mandatory for larger RPC systems
J.Repond: DHCAL
20
Summary
After successful testing of the DHCAL at Fermi and CERN
Further improvements to the active medium and its supplies
Development of 1-glass RPCs (design validated!) Development of low-resistivity bakelite/glass (ongoing, but encouraging) Development of a high-voltage distribution system (stalled) Development of a gas recirculation system (new concept, being assembled)
J.Repond: DHCAL
21
Backup
J.Repond: DHCAL
22
CMS forward calorimeter
Driven by successful application of PFAs to CMS analysis
Proposal to replace forward calorimeters with an IMAGING CALORIMETER
Several members of CALICE have been contacted by CMS
J.Repond: DHCAL
23
Formidable challenge
Charged particle flux
In calorimeter volume up to 50 MHz/cm2 at shower maximum
Total dose
Fluences of 1016 neutrons