STAR Forward Tracking Alignment
Introduction
In this blog, we will keep track of the procedure and progress towards aligning the FTT (sTGC) and FST detectors in the forward tracking system at STAR. Currently, we are using the Millepede II minimization tool for a global alignment approach of all track fit and alignment parameters. A global alignment approach prevents bias from static track fit parameters during the fitting of the alignment parameters. The use of Millepede II is made easier by the GenFit generalized fitting package that is used in forward tracking. GenFit provides a set of classes for application of the general broken lines (GBL) fit method which is essential for the use of Millepede II. These packages and their tools will be outlined in this blog and all progress towards the alignment will be tracked.
Packages and Tools
Millepede II
Millepede II is split up into two sub programs, mille and pede. Mille helps to organize the data for the alignment and pede performs the actual alignment from the binary output files from Mille.
GenFit
GenFit is used in the standard forward tracking using a Kalman Filter Fitter method. This is an independent fit at each detector plane and thus we refit this track with the GBL method. GenFit provides several classes for implementing the GBL fit including: GblPoint, GblTrajectory, GblData, and MilleBinary. Below is an outline of these classes.
- GblPoint: Individual point object which contains the residual of the measurement and the track prediction, the covariance matrix of the measurement, the derivatives of the residuals with respect to the track parameters, and the derivatives of the residuals with respect to the alignment parameters.
- GblTrajectory: Collection of GblPoints. Can be fit to provide initial guesses to kinks at scatters. Provides interface to MilleBinary, which produces binary output file for input into pede.
- GblData: Data structure for GBL track data. GblTrajectory uses this class to organize its data.
- MilleBinary: Grabs necessary data from GblData and produces binary data file for pede.
Weekly update slides 10/18/2022 -STAR/system/files/userfiles/6551/Gavin_WeeklyUpdate_20221018(1).pdf
11/01/2022 - Several plots for predicted, measured and truth position of FST for alignment:
No Misalignment - STAR/system/files/userfiles/6551/misalignmentPlots_noMis.pdf
Rotate around z by 1 mrad - STAR/system/files/userfiles/6551/misalignmentPlots_rot0p001.pdf
Shift in sensors local u coordinate by 100 μm (more aligned with r position) - STAR/system/files/userfiles/6551/misalignmentPlots_shiftU100um.pdf
Shift in sensors local v coordinate by 100 μm (more aligned with phi position) - STAR/system/files/userfiles/6551/misalignmentPlots_shiftV100um.pdf
- gwilks3's blog
- Login or register to post comments