300 likes | 553 Views
Calorimetry optimised for jets. Henri Videau Jean- Claude Brient Laboratoire Leprince-Ringuet Ecole polytechnique - IN2P3/CNRS. Contribution to the session on jet calorimetry CALOR2002.
E N D
Calorimetry optimised for jets Henri Videau Jean- Claude Brient Laboratoire Leprince-Ringuet Ecole polytechnique - IN2P3/CNRS Contribution to the session on jet calorimetry CALOR2002
The physics programme for a coming electron linear collider is dominated by events with final states containing many jets, dijets from H, W, Z . We contend that, in the energy range under consideration, the best approach is to optimise the independent measurement of the tracks in the tracker, the photons in the electromagnetic calorimeter and the neutral hadrons in the calorimetry, together with a good lepton identification. This can be achieved with a good tracker and a high granularity calorimetry providing particle separation, through an efficient energy flow algorithm. But we do not contend that this is a universal panacea Studying that program from the calorimetric side on hardware and software issues is the goal of the CALICE collaboration
Jets at the linear collider _ radiative qq at 500 GeV WW at 800 GeV
The analytical energy flow approach has been widely used at LEP and the energy ditributions are quite similar ALEPH but LEP detectors had some draw- backs coil in the middle projective cracks poor longitudinal segmentation 2d digital read- out in the HCAL
Impact of the jet resolution on the physics programme Parametrising a = 0.3 Þ 6s 6 jets L=1 ab-1 a = 0.6 Þ 3s work in progress a 0.3 Þ 0.6 = loosing 45% of L in the separation ZZ / WW a 0.3 Þ 0.6 = loosing 40% of L a lot more work to assess the effect on all the programme
separation ZZ / WW 0.6 0.3
Reminder on the analytical energy flow basics The energy flow of a jet is written as the sum of its components the charged particles make about 60% of the energy and, being of rather low energy, are much better measured by the tracker Isolate the 10% neutral hadron energy Argument Þ Such a method relies more on separation of particles than on intrinsic energy resolution Þ far enough from the interaction point small radiation length small interaction length matched granularity Inside coil Compactness seeing the mips
To profit from that we need a good tracker, not so much on momentum resolution but good track efficiency, small rate of fake energetic tracks good V0 identification small rate of reinteraction a good electromagnetic calorimeter, not so much on resolution but good photon efficiency, even close to tracks small rate of fakes from hadronic debris good electron identification (prompt) a good hadron calorimeter to identify muons to disentangle neutral hadronic showers from charged ones to measure their energy
Elements for a solution concerning the calorimetry Density, good separation electromagnetic/hadron sandwich tungsten / silicium 24 X0 40 layers no projective cracks thickness < 20 cm ECAL cells matched to the Moli₩re radius ~ 1 cm2 good efficiency to mips. noise ~ 1/10 mips Radiator adapted to separation/resolution HCAL small cells read digitally A solution with scintillating tiles is also studied within CALICE
Effect of going from iron Resolution to expanded tungsten Separation
All the studies presented here have been done using Mokka an application on Geant4 http://polywww.in2p3.fr/tesla/calice_software.html Are these performances kept at high energies? The jet energies have been obtained in a multi step process knowing the extrapolation of the charged tracks reconstruct the photons in the Ecal subtract the cells of these photons identify the hadrons estimate the energy of the neutral hadrons through a neural net Different other approaches: thermodynamical or neural net This is the cornerstone of jet calorimetry
Seeing a W dijet impact on the first 4 X0 of the calorimeter in q f projection The square is 100 mrad wide X generated g's 8 + charged 4 * neutral had. 1 O reconstructed g's
Some results at 800 GeV on photons number of reconstructed g's versus number of generated g's
Energy distribution for true photons and reconstructed ones including fakes GeV
Energy distribution of generated true photons and reconstructed true photons A reconstructed photon is associated to a true one if more than 75% of its energy comes from it. GeV
Difference between the true photon energy and the reconstructed one per event. The fit is done with 2 gaussians. Norm1 101.88 Mean1 0.23 GeV s1 7.01 GeV Norm2 35.84 Mean2 - 0.02 GeV s2 18.49 GeV c2/dof = 1.1 GeV
Photon reconstruction efficiency at low energy GeV
Photon reconstructed energy versus true energy GeV
Energy photons/evt kin versus rec. Distribution of event photonic true energy and reconstructed
Distribution of event photonic true multiplicity and reconstructed
Neutral hadron energy distribution GeV
A more complete reconstruction of the jets at high energy is under way.
Few more informations about the digital HCAL solution. The sensitive detector A gas detector, compact, efficient to mip, high signal and cheap! Streamer or Geiger wire detector information from DHCAL subcollaboration: IHEP, Interphysica, LLR, MEPhI, Seoul U. RPC's
1x1 cm2 Pads outside Gap 1.2 mm Glass plates 1 mm TFE/N2/IB 80/10/10 Pads inside Efficiency to mip > 98% Signal on 50 W : 3 V
Scheme for a digital HCAL signal detection Fe or .. Chip PCB Pad Glass insulating layer Spacers resistive layer Fe conductive layer insulating layer
Read out scheme for a 64 channel chip ~ 64 million channels Cost ~ 0.2 Euros/ch Reading the chips through a token ring
Conclusions To extract the physics produced in an electron linear collider below 1 TeV, a measurement of the jet energies with a stochastic term at a level of 0.3 or below seems mandatory. Such a precision does not seem out of reach with an adequate calorimetric hardware and a proper software. We have a roadmap with hardware developments and prototypes (2004) and with software imagination Join the worldwide effort of the CALICE collaboration http://polywww.in2p3.fr/tesla/calice.html