-
Notifications
You must be signed in to change notification settings - Fork 5
Data‐driven Alignment
The plan for the FWD Tracking System alignment is to use zero-field data collected during Runs 22 and 23 for the alignment. We have roughly 40 million events in the zero-field alignment data set.
In the forward tracking software, we fit tracks using GenFit2 using the Kalman Filter method. These reconstructed tracks can then be refit with a General Broken Lines (GBL) track fit, where the track is described by the offsets of the hits at each plane. The GBL method allows us to describe an entire track trajectory with a singular set of parameters. The benefit of this GBL track fit is that we can use the Millepede II minimization program which can minimize track residuals by a simultaneous fit of local (track) and global (alignment) parameters.
- Brief update of status. Discusses hierarchical alignment.
- General status updated. No plots, but attempted to allow all sTGC and FST parameters to be free. Did not work (no constraints on the values).
- First look at residuals from FST first tracking in data. These are local residuals and therefore, the azimuthal and r distributions will not make much sense for the FST.
- Status/plan. Merged code from local alignment code version to the most up to date.
- Changed local coordinate system definition.
- FTT pentagon alignment w/ and w/o varying parameters that rotate or shift the module in the z-direction.
- Testing increased precision for sensitive portion of FTT GEANT geometry volume. No difference.
- Added an alignment TTree which keeps track of hit volume id and positions for MC, track, and reco. Showed some residuals for the FTT.
- Tested track sample size dependence of Millepede II alignment parameter errors.
- Static misalignment of FST with varied pT of tracks shows no significant pT dependence of alignment. Also looked at w/ and w/o B field and require all/at least 1 FST plane hit(s).
- How recovered alignment parameters for an individual sensor change with respect to other variances in parameters.
- Large update with residuals shown to the forward upgrade group.
- Updated inner FST sensor alignment results.
- Many distributions for FST hit/track residuals.
- Many distributions for FST hit/track residuals. With Misalignments.
- General discussion of how we approach the forward tracking system alignment with GenFit2 and Millepede II.
- Look at single inner FST sensor alignment and residuals.
We ran some of the following closure tests for FTT first tracking. We now wish to repeat these tests for FST first tracking with a FTT refit. The FST first tracking method has been found to recover more good quality tracks from data and improved ECAL and HCAL matching as opposed to the FTT first tracking.
Tests to run with MC:
- Field OFF:
- sTGC only shift - Shift a single pentagon module of the sTGC and test if we can recover the alignment parameters.
- FST sensor only shift - Shift a single silicon sensor of the FST and test if we can recover the alignment parameters.
- FST module (wedge) only shift - Shift an entire FST module without shifting the individual sensors. This will show that we can recover a set of larger structure parameters without individual sensor shifts. This is known as hierarchical alignment.
- FST module and sensor shifts - Shift both an entire FST module and the individual sensors.
- FST multiple shifts of modules and sensors - Shift multiple wedges and sensors and recover these parameters simultaneously.
- FST half shift only - Shift entire half of FST and recover only shifts in the half, rather than many shifts in the smaller structures.
- FST halves + modules shifts - Shift modules and halves simultaneously and recover parameters.
- FST halves + modules + sensors shifts - Shift sensors, modules, and halves simultaneously and recover all parameters.
- sTGC + FST Shift - We will shift all sTGC parameters and FTT parameters and recover all parameters.
- Field ON (same cases as above)