RPC Muon Reconstruction CMS RPC Workshop Aug 31

RPC Muon Reconstruction CMS RPC Workshop Aug. 31 – Sept. 1, 2020 Sandro Fonseca (UERJ) Michael Tytgat (Ghent U. )

Introduction • RPC Muon Reconstruction is a new sub-group under the RPC DPG group • The main goal is to coordinate the activities and plans concerning improvements of the contribution of RPC for local and global muon reconstruction • Helping with the adaptation and implementation of specific models • Providing support of RPC POG coordination team concerning these new developments • Interacting with and reporting to Muon POG • Twiki page: https: //twiki. cern. ch/twiki/bin/view/CMS/RPCMuon. Reconstruction • Meeting foreseen once per month; Tuesday 3 -4 pm CERN time • • Kick-off meeting happened on July 28, 2020: https: //indico. cern. ch/event/942204/ Mailing list: cms-dpg-rpc-muon-reconstruction@cern. ch August 31, 2020 M. Tytgat - RPC Workshop 2

Central Software Repository • Using a Git. Lab to help us to organize an evolution of tasks and subprojects for each group (UERJ-Rio, Ghent, CBPF), so we will have visibility of each task under each team https: //gitlab. cern. ch/cms-rpc-muonreco August 31, 2020 M. Tytgat - RPC Workshop 3

Initial Set of Tasks and Projects Initial set of tasks defined by Brieuc, Andres & Sandro August 31, 2020 M. Tytgat - RPC Workshop 4

RPC Upgrade Extracted from Brieuc slides: https: //indico. cern. ch/event/839288/contributions/3542843/attachments/1909167/3154106/Brieuc_Francois_RPC. pd August 31, 2020 M. Tytgat - RPC Workshop 5

RPC timing to improve muon object To update the muon timing producer in CMSSW (Task 1) The code already exists, so the idea is to modify the path where they use RPC to fill the muon time. A first step will be to check that muon time is not using phase 2 sub-bx time. If that is not the case, modify the code and make a performance study to show that we get an improvement in the muon time. Bring CMSSW muon timing code at same level for RPCs as for CSC/DT Responsibles: Michael Tytgat (Ghent U. ) Status: Work in progress Reports: • RPC DPG Working Meeting (13 May 2020): https: //indico. cern. ch/event/918670/contributions/3860930/attachments/2037416/34116 81/RPCMuon. Timing. pdf • RPC DPG Working Meeting (8 July 2020): https: //indico. cern. ch/event/937362/contributions/3938066/attachments/2070536/34757 91/RPCMuon. Timing_8 jul 2020. pdf August 31, 2020 M. Tytgat - RPC Workshop 6

Correct RPC timing using Muon object Improvement for offline reconstruction of Phase 2 Correct the RPC timing using the muon object (Task 2) The muon direction can be used to extrapolate the track to the RPC and the location of the hit along the strip. This can be applied in barrel and endcap (muon is a global object) Muon Track CMSSW code needs to be in place first (see previous task) Responsibles: Michael Tytgat & Amrutha Samalan (Ghent U. ) Status: Work in progress August 31, 2020 M. Tytgat - RPC Workshop 7

RPC timing to improve muon object Improvement for offline reconstruction of Phase 2 To use the RPC timing (rec hit timing or cluster timing, no need to use the digis) in the muon object (Task 3) Examples (the focus here is on endcap CSC but everything mentioned should be also studied for other regions): - if there is a muon that has a hit far away in time, we remove the hit, and redo the fit of the muon. This could require to match RPC and CSC information to be able to remove also the matched CSC segment. - Forbid building muons with hits too far away in time. - When deriving muon quantities, give a smaller weight to hits far away in time. Responsibles: Gilvan Alves & Carsten Hensel (CBPF) Status: preliminary exploration of the task August 31, 2020 M. Tytgat - RPC Workshop 8

RPC/CSC Super Segments Try to build super segments with RPC and CSC at the local reco stage (Task 4) Improvement for offline reconstruction of Phase 2 CSC Local reconstruction reference • • • Proof of concept to evaluate if there are benefits to using the RPC to decide CSC ambiguity Keeping in mind that CSC side not had any clear trouble from ghost segments We can use a cone which axis is the CSC segment, and check its projection on each adjacent RPC chamber; CSC segment could be counted or not, depend on the projected cross section of the cone onto the RPC area (blue area) Responsibles: Sandro Fonseca & Mapse Barroso (UERJ) Status: preliminary update on the task by the end of August 31, 2020 M. Tytgat - RPC Workshop 9

Muon. Time in CMSSW (Task 2) • Current Muon. Time (→ time of flight) is driven by DT, CSC & Ecal • Muon RPCTime is currently based on bunch crossing number, i. e. 25 ns resolution only • Reco. Muon/Muon. Identification package - Muon. Timing. Filler. cc: RPC time is determined using (average of) RPCRec. Hits matched to standalone muon tracks and depends on RPCRec. Hit time. Error = bunch crossing time (time. Error < 0) = RPCRec. Hit time (time. Error >= 0) August 31, 2020 M. Tytgat - RPC Workshop 10

Looking at collision and MC data • Data sample: Run 2 2018 pp collision data from Single. Muon stream • MC sample: Phase. IITDRSpring 19 DR/Mu_Flat. Pt 2 to 100 -pythia 8 -gun/GEN-SIM-DIGI-RAW, with and/or without Min. Bias pileup from same production (using PU 200, 25 ns) • Need RPCRec. Hits and standalone muons • GEN/SIM level info (sim. RPCHits, gen. Particles …) to match RECO muons and RPCRec. Hits to generator level muons • Start from RAW data or GEN-SIM MC, and rerun entire chain until RECO level • Using CMSSW_10_6_0_patch 2 (same as for MC sample production) August 31, 2020 M. Tytgat - RPC Workshop 11

Muon. RPCTime in Run 2 data Run 2 2018 pp collision data from Single. Muon stream Supposed to be time of muon at CMS IP assuming it originates from IP (i. e. going from “inside” to “outside”) August 31, 2020 M. Tytgat - RPC Workshop 12

Muon. RPCTime in Phase-2 MC Phase. IITDRSpring 19 DR/Mu_ Flat. Pt 2 to 100 -pythia 8 gun/GEN-SIM-DIGI-RAW No pile-up included yet August 31, 2020 M. Tytgat - RPC Workshop 13

RPC vs. CSC/DT time RPC time CSC/DT time RPCs time can now finally compete with CSC/DT … August 31, 2020 M. Tytgat - RPC Workshop 14

GEN-RECO Matching Perform matching of RPCRec. Hits to generator level muons in MC sample, i. e. check which RPC hits originate from generated primary muons standalone muons RPCRec. Hits Track. Rec. Hits generated muons raw. Id() pdg. Id() det. Unit. Id() RPCSim. Hits track. Id() Sim. Tracks genpart. Index() Gen. Particles Ø Check if and how matching depends on RPCRec. Hit time August 31, 2020 M. Tytgat - RPC Workshop 15

GEN-RECO Matching efficiency for: • all RPCRec. Hits: 96. 5% • RPCRec. Hits assigned to standalone muons: 99. 7% August 31, 2020 M. Tytgat - RPC Workshop 16

BACKUP

Recap – muon reconstruction • Muons are reconstructed using the information from inner trackers and muon detectors • Tracker muons (or similar variants) using inner tracks, matching segment/rec. Hit • Standalone muons using muon detectors only • Global muons combining inner tracks and standalone muons • Are merged into a reco: : Muon collection • Basic elements to build muon tracks are Rec. Hits • DT/CSC – build segments, position and direction in their local coordinates • RPC – build ”hit”, position in their local coordinates • Use the Kalman filter algorithm, propagation is done with non-uniform B-field information and interaction with materials • RPC is also participating in muon reconstruction • note - no “segment”, local y-position error by strip length 18 RPC in muon reconstruction John Goh (Hanyang Univ. ) 2017. 02. 21 August 31, 2020 M. Tytgat - RPC Workshop 18

RPC local reconstruction • Current RPC detector is designed to give the position of a muon on the detector surface, with fast response and high efficiency • Strips are fired when a charged particle passes through the detector volume • Charge collected on a strip is converted to a digitized signal, RPCDigis • RPCDigis contain strip number, bunch crossing #, detector(roll) ID • Multiple strips can share charge from a charged particle • RPC local reconstruction is basically a clustering of RPCDigis • Cluster adjacent RPCDigis in a roll, in same bx • Compute position and its error, keep the number of clustered RPCDigis • Do the clustering with consideration of masked strips • For phase-II – same algorithm, compute time & 2 D position 19 RPC in muon reconstruction John Goh (Hanyang Univ. ) 2017. 02. 21 August 31, 2020 M. Tytgat - RPC Workshop 19

RPC in muons reconstruction • Standalone muons • Starts from finding muon seed using CSC/DT segments • Tracking is done using DT/CSC segments and RPCRec. Hits • Do track fit twice, inside-out and outside-in and contraine to point at the beam spot • Global muons • Inner- and standalone. Muon tracks are matched by propagating at a common surface • Run the Kalman filter algorithm to perform a combined track fit • Tracker muons / RPCMuons • Extrapolate inner track to muon stations to match DT/CSC segments (TRK) or RPCRec. Hits (RPCMuon) • Track parameters are unchanged, muon hits are used only for the identification 20 RPC in muon reconstruction John Goh (Hanyang Univ. ) 2017. 02. 21 August 31, 2020 M. Tytgat - RPC Workshop 20