1. Proposal for UK involvement in CALICE
Paul Dauncey
Imperial College
for the CALICE-UK groups
13 May 2002
UK CALICE Involvement

2. The UK people 19 names; 5 institutes
Birmingham; C.M.Hawkes, N.K.Watson
Cambridge; C.G.Ainsley, M.A.Thomson, D.R.Ward
Imperial; D.A.Bowerman, W.Cameron, P.D.Dauncey, D.R.Price, O.Zorba
UCL; J.M.Butterworth, D.J.Miller, M.Postranecky, M.Warren
Manchester; R.J.Barlow, I.P.Duerdoth, N.M.Malden, D.Mercer, R.J.Thompson
13 May 2002
UK CALICE Involvement

3. A future linear collider
A future e+e– linear collider (LC) with:
High centre-of-mass energy; 500 – 1000 GeV
High luminosity; 3 – 10 x 1034 cm-2 s-1
is regarded as a high priority for the future of HEP.
It is strongly supported worldwide:
EFCA; July 2001.
HEPAP/Snowmass; September 2001.
ACFA; September 2001.
as well as in the UK:
UK; Blair report, September 2001.
and the technical feasibility was demonstrated by:
TESLA TDR; March 2001.
13 May 2002
UK CALICE Involvement

4. Physics at a linear collider
The earliest date for a LC to start would be 2012, so it has to be seen in the context of the LHC:
Higgs; mass will be known (if it exists).
Existence of SUSY; some particles identified if so.
Top quark mass; known to GeV
The physics programme at a LC would complement that from the LHC (as LEP did after UA1/2) with precision measurements:
Fundamental quantities; e.g. Higgs and top mass.
Distinguish models; e.g. details of SUSY spectrum.
13 May 2002
UK CALICE Involvement

5. Physics at a LC (2)
Particular highlights of a LC physics programme include:
Higgs; mass to 0.1%, width to 5%, fermion and gauge boson couplings, spin and parity, self-coupling.
SUSY; masses, couplings, mixing parameters, charged Higgs, separation of close-lying states.
Strong Symmetry Breaking; if no Higgs, study WLWL and ZLZL scattering with nnW+W– and nnZZ final states.
Top; mass to 0.2 GeV with a threshold scan.
LC physics programme could take around 10 years – –
13 May 2002
UK CALICE Involvement

6. Calorimetry at a LC
Many states of interest result in quarks and hence hadronic jets; combining these give masses needs:
Angles; straightforward to measure accurately.
Energies; much more complicated.
LEP experience showed jet energies are best measured using “energy flow” algorithms:
Explicit association of tracks and clusters.
Prevents double counting of track/cluster energies.
Needs a “tracking calorimeter” with fine granularity.
Intrinsic calorimeter energy resolution is a
secondary effect.
Aleph achieved DE/E = 60%/E in the central region.
13 May 2002
UK CALICE Involvement

7. TESLA TDR Si-W ECAL
TESLA TDR specified a silicon-tungsten (Si-W) sampling electromagnetic calorimeter:
40 layers, between
0.4 and 1.2X0
(radiation lengths).
24X0 total thickness.
32 million channels.
13 May 2002
UK CALICE Involvement

8. Si-W properties Tungsten has:
Small Moliere radius; ~ 9 mm; gives narrow
showers and so reduces overlaps. The effective
Moliere radius depends also on gap and pixel sizes.
Short radiation length; X0 ~ 3.5 mm; depth of ECAL
can be kept small ~ 20 cm.
Small radiation/interaction length; good
longitudinal separation of EM and hadronic
showers.
Silicon diodes also have good properties:
Dimensions; gaps and pixels can be kept small
Signals; reasonable size and simple to use
13 May 2002
UK CALICE Involvement

9. TDR ECAL issues TDR tried to specify a “perfect” physics ECAL:
No amplification inside calorimeter volume
No cooling pipes needed
Keep gaps small
Si pixel sizes 1cm x 1cm
Matched to Moliere radius
Large number of channels to calibrate
Electronics space restricted
Requires significant electronics integration; analogue, digital and optical
13 May 2002
UK CALICE Involvement

10. TDR ECAL cost The main figure of (de)merit is the cost…
ECAL total cost of 133 Meuros ~ 90 Mpounds:
Silicon wafers; 70% of the cost:
Effectively only depends on the total area,
i.e. number of layers.
Pixel size is almost irrelevant to cost.
Coil size; ~2 Meuros per extra cm:
Gap size directly impacts size (multiplied by
a factor 40)
Cost/performance optimisation is needed:
Complex, multi-parameter space.
13 May 2002
UK CALICE Involvement

11. TDR performance
For hadronic jets, TDR calorimeters give DE/E = 33%/E with further improvement possible from better algorithms.
DE/E = 60%/E DE/E = 30%/E
For Z or W to two
jets, this gives
sZ/W ~ GZ/W
Distinguishes
nnW+W– and nnZZ
final states
Photons have
DE/E = 1%+10%/E – –
13 May 2002
UK CALICE Involvement

12. What needs to be known How does jet resolution degrade with:
Number of layers; obvious cost factor for Si wafers
Pad size; determines number of channels and hence electronics cost
Number of dead channels; wafer yield is major cost factor for Si wafers
Inter-layer gaps; can cooling pipes be inserted? Can electronics be inside the ECAL?
Resolution/calibration; how good does it need to be and how will it be measured?
These require an accurate simulation to answer
13 May 2002
UK CALICE Involvement

13. The CALICE collaboration
Formed to study issues of calorimetry for a LC. Currently 96 physicists, 17 institutes, 7 countries:
Spokesperson; J.-C. Brient, LPNHE Ecole Polytechnique
Steering Board chair; R.-D. Heuer, Hamburg/DESY.
Studying both ECAL and HCAL, as energy flow requires integrated approach:
ECAL; Si-W option  UK interest
HCAL; tile scintillator or “digital” RPCs
Two separate efforts:
“Physics prototype”; beam test  UK interest
“Technical prototype”; mechanical TDR structure
13 May 2002
UK CALICE Involvement

14. Physics prototype Beam test of ECAL and both HCAL options. ECAL;
Total 30 layers, 24X0;
10 x 0.4X0
10 x 0.8X0
10 x 1.2X0
Total 9720 channels;
3x3 wafers/layer
6x6 channels/wafer
Active volume;
18 x 18 x 18 cm3
13 May 2002
UK CALICE Involvement

15. UK proposal We propose to work in two areas:
Readout and DAQ; for the physics prototype. We would provide the readout electronics for the ECAL. As the tile HCAL might be able to use the same boards, we propose to supply those also. We would also provide the DAQ for the whole system.
Simulation studies; on the development of energy flow algorithms and the impact of the calorimeter design. In addition, we would work on the ECAL cost/performance optimisation.
These both clearly lead to analysis of the beam test data
13 May 2002
UK CALICE Involvement

16. Proposed readout system
Short timescale, so aim for simplicity and robustness, not high performance or any major technical development.
All in VME; readout boards directly connected to wafer/VFE PCBs, trigger board to distribute trigger in crate
13 May 2002
UK CALICE Involvement

17. Readout board Reads out 6 PCBs; total of 15 readout boards.
18 channels from each VFE chip digitised by 16-bit ADC; 648 total.
DACs for calibration pulse to VFE inputs.
FPGA for board control, VFE and VME interfaces.
Board reads 1296 bytes per event; 19 kBytes total for ECAL.
13 May 2002
UK CALICE Involvement

18. Trigger and test boards
The trigger board handles trigger to readout boards:
Distributes trigger to readout boards via J2 backplane.
Vetos further triggers until readout complete.
Possibly distributes triggers to HCAL readout.
In addition, a test board is needed:
Connects to readout board through PCB cable.
Supplies all VFE output to readout board, checks all VFE input from readout board.
Conceptually an “inverted” readout board with swapped DACs  ADCs; similar implementation.
13 May 2002
UK CALICE Involvement

19. Data acquisition
Needs to readout whole physics prototype, not just ECAL.
Data volumes:
ECAL; 9720 channels, ~ 19 kBytes.
Tile HCAL; around 1200 channels, ~ 2 kBytes.
Digital HCAL; around 400k channels, ~50 kBytes.
VME maximum limit will be around 1 kHz; aim for 100 Hz.
Around 107 to 108 events expected.
One to two months data taking.
Total data volume of order 1 Tbyte.
13 May 2002
UK CALICE Involvement

20. Simulation studies
Simulation studies form an integral part of this work:
Develop energy flow algorithms based on both calorimeters and the inner tracking detectors.
CALICE simulation uses GEANT4; need to verify both the electromagnetic and hadronic interaction descriptions.
Investigate proposed changes to the calorimeter structure or readout and their effect on energy flow resolution.
Optimise the ECAL performance within a more realistic cost envelope; check effects of reducing the number of silicon layers, resolution, dead channels, etc.
13 May 2002
UK CALICE Involvement

21. Milestones
The end point is fixed; a beam test early in The schedule to get to this is:
System specification;
Complete; end June 2002
Readout board;
Prototype design complete; end December 2002
Prototype fabrication complete; end February 2003
Prototype tests complete; end June 2003
Production design complete; end July 2003
Production fabrication complete; end September 2003
Production tests complete; end December 2003
13 May 2002
UK CALICE Involvement

22. Milestones (2) Test board; Design complete; end February 2003
Fabrication complete; end April 2003
Trigger board;
Design complete; end August 2003
Fabrication complete; end October 2003
Tests complete; end December 2003
All prototyping completed within FY02/03.
All production completed within FY03/04.
13 May 2002
UK CALICE Involvement

23. Equipment request The estimated cost of the equipment is as follows:
NRE; total = £2k
Readout boards; £2800 x 22 boards = £62k
Trigger boards*; £1200 x 3 boards = £4k
Test board; £2900 x 1 board = £3k
Cables*; £30 x 100 cables = £3k
PC and disk; £4k PC, £8k disk = £12k
VME interfaces; £4k PCI/VME, £3k extender = £7k
VME crates; £5k x 2 crates = £10k
The total is £103k, of which £23k would be spent in FY02/03 and the remaining £80k in FY03/04.
(*Cost depends on external factors and is less certain)
13 May 2002
UK CALICE Involvement

24. Engineering effort
The estimated engineering effort needed is as follows:
Readout boards; 18 months
Trigger boards; 6 months
Test board; 6 months
Layout and fabrication; 4 months
The total is 34 months, of which 17 months would be needed in FY02/03 and the remaining 17 months in FY03/04. The University groups can provide 18 months of this; we request the rest from RAL TD:
Board design; 12 months, 6 months in each FY
Layout and fabrication; 4 months, 2 months in each FY
13 May 2002
UK CALICE Involvement

25. Summary
Excellent calorimetry is vital to the physics programme of a future linear collider. Algorithms and simulation both need work to be able to design the calorimeters needed.
Within the ECAL, there are many interesting problems to be solved. The baseline TESLA TDR solution may not be optimal and can probably be made cheaper.
The UK has a critical mass of bodies to get involved; it
Has made itself known to the CALICE collaboration
Has carved out a role in the short term
Is keeping options open for the longer term
Needs funding to proceed further
13 May 2002
UK CALICE Involvement