TPC status report Marian Ivanov 8 th March

TPC status report Marian Ivanov 8 th March 2011 1

Outlook TPC reconstruction and the new QA selection criteria Coverity TPC calibration and reconstruction – – 8 th March 2011 New d. Edx calculation New space point calibration 2

Outlook TPC calibration strategy – – – To be presented on Thursday https: //twiki. cern. ch/twiki/bin/viewauth/ALICE/PWG 1 Calibration. Strategy https: //alisoft. cern. ch/Ali. Root/trunk/TPC/doc/calib TPC PWG 1 QA – – 8 th March 2011 Automatic extraction of the QA trend graphs on the WEB (currently at the GSI web page ) http: //www-alice. gsi. de/TPC/PWG 1 train/ 3

TPC reconstruction and the new QA selection criteria (0) http: //indico. cern. ch/contribution. Display. py? contrib. Id=6&co nf. Id=127310 One of the major systematic error in the physics analysis is introduced by selection criteria on the number of clusters. Based on experience with TPC data of 2009 and 2010, it is known that TPC gain and baseline variation influence the distribution of found clusters up to 30% level. This variation is time, energy loss, and running condition dependent and it is only partially described in the MC. However, this variation influences the TPC tracking performance onlyto the level of 1 − nmiss /nf ind. Ongoing work on integration of the new selection criteria into 8 th March 2011 4

8 th March 2011 5

8 th March 2011 6

Coverity Our records show you currently have 111 defects assigned to you (PWG 1 and TPC) Removing of the remaining defects ongoing Main source – obsolete code to be removed – 8 th March 2011 After the start of the Pb. Pb pass 2 7

New d. Edx calculation (0) Experience from the 2010 data – the d. Edx calculation not sufficient robust in respect to the variation of the gas gain Problem: The TPC is working with the gas gain ~ 6600 while in the TDR we expected to run at the gain 20000. Significant fraction (x – momentum, gain, position and PID dependent) of the cluster lost, because did not pass the threshold of electronic. ==> The old d. Edx algorithm did not calculates the truncated mean from 0 up to 0. 6 but from x-0. 6*(1 -x) The gain calibration equalize the d. Edx for MIP particles. Second order corrections needed to follow the Bethe-Bloch. Such correction time dependent. 8 th March 2011 8

New d. Edx calculation - Pb. Pb The Altro baseline restoration algorithm was not switched on The validation of the algorithm took 2 weeks. Decided to continue data taking with the same setup. Consequences – the baseline depends on the current event multiplicity (track occupancy). Observed d. Edx – linearly decreasing with the occupancy. Solution – Adding the baseline correction (as function of occupancy) as a additive factor to the observed d. Edx. Calibration of the baseline shift done in the Pass 0. 8 th March 2011 9

New space point calibration Transition form the physical + effective” correction to the pure physical correction ( more details during the calibration meeting on Thursday) Improvement of TPC-ITS matching form RMS 0. 5 mm down to 0. 2 mm Physical correction TPCEx. Bshape ~ 0. 2 -1 cm (magnetic field inhomogenity) Field cage rod shift ~ up to 0. 7 cm in critical sector 10 -11 Ex. B twist ~ 1 mm (angle between E and B field) Inter detector alignment ~ +-0. 5 mm TPC-ITS (B+ and B-) Sector alignment ~ rms ~ 0. 1 mm 8 th March 2011 10

Work in progress Make the space point calibration algorithm “user friendly running” on the grid to be repeated after each LHC period Open issue cosmic tracks and high pt V 0 s (un-binned information) Alignment of the outer detectors - by-product of the TPC calibration (constrains for A side/C side ) Global calibration/alignment? The TPC space point correction classes finalized. To be used also in the MC soon. 8 th March 2011 11