ECAL Studies for the CLICILD detector Jacopo Nardulli

ECAL Studies for the CLIC_ILD detector Jacopo Nardulli, CERN LCD CALICE Collaboration Meeting Heidelberg 14 -16/9/2011 ECAL Studies at CLIC 1/14

Why the ECAL is as it is ? • The LOI ILD Ecal structuring comes from an optimization from H. Videau available here ECAL Optimization from H. Videau • A few scenarios with different nr of layers and different absorber thickness are considered • Similar studies have also been done by M. S. Amjad and have been presented at the CALICE Meeting in Casablanca 16/9/2011 ECAL Studies at CLIC 2/14

Goal of this study • Attempt a similar optimization on the CLIC version of the ILD detector • To have an ECAL with less layers, therefore cheaper, but without degradation of the performance 16/9/2011 ECAL Studies at CLIC 3/14

Two possible studies • Fix the radiation length • But decrease the Nr of ECAL layers • Less expensive, energy resolution can be worse, as the pattern recognition • How to ? • Varying nr of layers and absorber thickness, • not the Si thickness • Vary the radiation length • And decrease the Nr of ECAL layers • Less expensive, overall performance changes 16/9/2011 • If I change the X 0, can have more leakage into the HCAL • Varying nr of layers, absorber thickness and studying performance as a function of the X 0 ECAL Studies at CLIC 4/14

Now only first study: Different ECAL models • Altering the number of layers and their absorber thickness in such a way that Total Absorber thickness in the detector remains the same. • The analyses included in this talk were done with Single Photon with θ and φ varying in the full range and Energies of 1, 100 and 500 Ge. V and using Pandora. PFA • Default model 20 layers in 1 st stack and 9 layers in 2 nd stack • Study in two steps, • first keeping the thickness of first and second stack also unchanged • then varying the ratio 16/9/2011 ECAL Studies at CLIC 5/14

Disclaimer • Here showing the Energy Resolution vs. the Energy and using Single Gaussian fits • To get the energy resolution I am NOT using the official Marlin. Processor which calculates rms 90 and mean 90 • So do not look at absolute numbers, but at the general trend of the plots 16/9/2011 ECAL Studies at CLIC 6/14

Results: changing Nr of layers in 1 st stack Mean Rec. Energy vs Input Energy 16/9/2011 ECAL Studies at CLIC 7/14

Results: changing Nr of layers in 1 st stack Mean Rec. Energy vs Input Energy Res. vs Input Energy 16/9/2011 ECAL Studies at CLIC 8/14

Results: changing Nr of layers in 2 nd stack Energy Res. vs Input Energy 16/9/2011 ECAL Studies at CLIC 9/14

Results: changing Nr of layers in 1 st and 2 nd stack Energy Res. vs Input Energy Preliminary conclusion: à Can go from 29 layers to 25 with hardly any loss in performance à Need to change ratio of X 0 in 1 st and 2 nd stack to further reduce nr of layers –now roughly 50%: 50%. Can try 40 -60 ! 16/9/2011 ECAL Studies at CLIC 10/14

Results: changing Nr of layers in 1 st and 2 nd stack and different ratios à Reducing the X 0 in 1 st stack and changing the ratio does not seem to help. à Now try something else 16/9/2011 ECAL Studies at CLIC 11/14

Results: just 1 stack Preliminary results à Using just 1 stack with less, but thicker, layers does not seem to degrade the energy resolution dramatically. 16/9/2011 ECAL Studies at CLIC 12/14

Conclusions à First attempt of an optimization of the ECAL for CLIC_ILD à First results show that we can go from 29 to 25 layers with a small loss in performance à Reducing the X 0 in 1 st stack and changing the ratio does not seem to help. à Using just 1 stack with less, but thicker, layers does not seem to degrade the energy resolution dramatically. 16/9/2011 ECAL Studies at CLIC 13/14

Next steps à Preliminary studies, lots of things not considered à i. e. is it right to use PFA ? à should I use Garlic ? à Is the energy resolution the main/only parameter to be used to understand possible loss of performance caused by having less layers ? à Finish 1 st study à More statistics à Use RMS 90 à Move to the 2 nd study à Performance as a function of the X 0 16/9/2011 ECAL Studies at CLIC 14/14

Backup 16/9/2011 ECAL Studies at CLIC 15/14

Results: changing nr layers in 2 nd stack Mean Rec. Energy vs Input Energy 16/9/2011 ECAL Studies at CLIC 16/14

Only 1 layer 16/9/2011 ECAL Studies at CLIC 17/14

Results: changing Nr of layers in 1 st and 2 nd stack Mean Rec. Energy vs Input Energy 16/9/2011 ECAL Studies at CLIC 18/14