Results from particle beam tests of the ATLAS

Results from particle beam tests of the ATLAS liquid argon endcap calorimeters n Beam test setup n Signal reconstruction n Response to electrons § CALOR 2004 Perugia, Italy Mar 29 – Apr 2, 2004 Electromagnetic Scale n Response to pions § weighting using energy density ATLAS HEC: Canada, China, Germany, Russia, Slovakia ATLAS EMEC: France, Russia, Spain Michel Lefebvre University of Victoria Physics and Astronomy

ATLAS Endcap LAr Calorimeters Electromagnetic endcap with presampler 2 m Hadronic endcap Forward See M. Vincter’s talk M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 2

Electromagnetic Endcap Calorimeter n EMEC absorber structure § Pb absorbers arranged radially, no azimuthal cracks § folding angle and wave amplitude vary with radius § inner and outer wheels n EMEC readout structure § layer 0 (presampler) = 0. 025 0. 1 § layer 1 (front): 2 to 4 Xo = 0. 025/8 0. 1 § layer 2 (middle): 16 to 18 Xo = 0. 025 § layer 3 (back): 2 to 4 Xo = 0. 050 0. 025 M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 3

HEC-EMEC beam test configuration n H 6 beam area at the CERN SPS § § e , , beams with 6 Ge. V E 200 Ge. V. Here report on e , . 90 o impact angle: non-pointing setup (not like ATLAS) beam position chambers optional additional material upstream (presampler studies) front face of HEC seen through the EMEC HEC PS+EMEC M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 4

Main goals of the HEC-EMEC beam test n Determination of the hadronic calibration constants in the ATLAS region 1. 6 < | | < 1. 8 n Development of hadronic energy reconstruction methods n Monte Carlo simulation validation and extrapolation to jets Other goals are to test • detector operation • electronics • software framework M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 5

Signal reconstruction n Optimal filtering § need known physics signal shape § discrete ( t = 25 ns) measurements (signal + noise): § autocorrelation matrix from noise runs: § estimate signal amplitude S with § minimize § solution is given by the optimal filtering weights n Signal shape § obtained directly from data § or obtained from calibration pulses and detailed knowledge of difference between signal pulse shape and calibration pulse shape M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 6

HEC calibration: ADC to n. A n Calibration pulse height § crucial to understand the channel-by-channel variation in the difference in pulse height and shape between data and calibration signals § electronics modeling § predict signal pulse from calibration pulse to about 1% calibration signal (points) electronics function fit (line) data signal (points) prediction (line) fit residua M. Lefebvre, CALOR 04 residua ATLAS LAr Endcap Calorimeters 7

Electronic noise n Electronic noise obtain directly from data § EMEC: use muon data and remove hit cells § HEC: use first 5 time samples (which are out of signal region) 327 Me. V EM scale 25. 8 Me. V EM scale M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 8

Clustering n Cell-based topological nearest neighbor cluster algorithm 180 Ge. V pion § clusters are formed per layer using neighbours (that share at least one corner) § Eseed > 4 noise § |Ecell| > 2 noise § include neighbour cells with |Ecell| > 3 noise M. Lefebvre, CALOR 04 n. A ATLAS LAr Endcap Calorimeters 9

Electrons: geometrical corrections n -dependent correction required 3 absorbers § electric field and sampling fraction nonuniformities § non-pointing setup § well understood n Other smaller -dependent corrections neglected in this analysis M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 10

Electrons: EMEC electromagnetic scale n Needed as reference for hadronic calibration Include 2% -dependent n Obtained from beam test data geometrical response corrections where The leakage is only outside the cluster, hence measurable. It is < 3% for Ebeam > 30 Ge. V signal shape uncertainties and dependent corrections which have not been applied n MC simulation: See D. Salihagic’s talk M. Lefebvre, CALOR 04 Linearity better than 0. 5% ATLAS LAr Endcap Calorimeters 11

Electrons: energy resolution impact point J Note: nonpointing setup!! possibly some dependence, due to variation of sampling fraction and weak dependence of electric field M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 12

Pions: response n Use HEC EM scale from previous TB, modified by new electronics, and EMEC EM scale obtained here n Example: 120 Ge. V pions in EM scale HEC EMEC M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 13

Pions: cluster weighting n EMEC and HEC are non-compensating calorimeters § corrections (weights) are required (over the EM scale constants) § various weighting methods are being investigated n Cluster weights as a function of EM energy density total leakage vs E density leakage outside detector: use MC § the weights should be obtained from MC. . . not yet available § we consider the (H 1) form EM energy over cluster volume M. Lefebvre, CALOR 04 leakage outside cluster: use data 200 Ge. V pions ATLAS LAr Endcap Calorimeters 14

Pions: test of cluster weighting procedure n 30 Ge. V pions with no energy deposited in the HEC § test the procedure without the need for MC (except for part of lateral leakage) § only EMEC weights required § data agrees well with the proposed weights form M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 15

Pions: cluster weights n Obtain weights through the minimization of where noise is the total electronics noise; cluster noise and electronics noise contribution to the leakage estimate Energy dependence of weights C 1 and C 2 strongly correlated; C 2 fixed to 1500 cm 3/Ge. V M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 16

Pions: energy resolution EM scale Clear improvement when using cluster weighting constant term compatible with zero n Weighting also attempted at cell level: similar results M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 17

Pions: e/ ratio n Effective e/ ratio § obtained from the cluster weighting function § composite calorimeter: e/h has no direct interpretation. . . with this warning: § -: e/h = 1. 69 0. 1 using Groom’s with Eo’ = 1 Ge. V and m = 0. 85 n MC simulation: See D. Salihagic’s talk M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 18

2004 HEC-EMEC-FCAL beam test n Address the | | interface region absorption length budget interface around | | 3. 2 M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 19

2004 HEC-EMEC-FCAL beam test n Summer 2004 HEC-EMEC-FCAL combined beam test n Focus on energy reconstruction in the 2. 8 < | | < 3. 2 region § special mini-HEC modules to fit in test beam cryostat § cold and warm tail catchers § beam starts in May M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 20

Conclusions n ATLAS LAr EMEC-HEC beam tests, 1. 6 < | | < 1. 8 § e , , beam with 6 Ge. V E 200 Ge. V. Results reported: e , n Electronics calibration method to be used in ATLAS § optimal filter weights § detailed electronic calibration procedure for ADC to n. A § development of the related software tools n Test of first steps toward an hadronic calibration strategy § clustering; to be improved including 3 D clusters and pileup § cluster and/or cell weighting n Remaining calibration tasks § use of validated Monte Carlo simulations § jet reconstruction and particle identification in jets n Upcoming HEC-EMEC-FCAL beam tests, 2. 8 < | | < 3. 2 § three-calorimeter forward region M. Lefebvre, CALOR 04 ATLAS LAr Endcap Calorimeters 21