TPC GridLeak model changes for Run 17

Speaker : Gene Van Buren ( BNL )

Talk time : 12:20, Duration : 00:20
TPC GidLeak R&D was presented at an S&C meeting last August concerning modifications to the distortion model based on features learned from studies to reduce the GridLeak in the iTPC era (a STARNote is at the draft stage). Here are examples of the distortion plots for GridLeak between the two models: Δ(rφ) vs. φ:
(all units are [cm])

3D GridLeak (old) Full GridLeak (new)

At the time of that presentation, it was shown what impacts could be expected on tracks from modifying only the GridLeakdistortion. However, because TPC SpaceCharge distortions are calibration from signed DCAs which are also impacted y the GridLeak distortion, modifying the GridLeak distortion model causes the calibration to change the balance between these two (i.e. SpaceCharge will change to compensate for the signed DCA effects of the GridLeak change). Therefore, one needs to do the full calibration to see the final impacts...

For Run 17, both the old ("3D") and new ("Full") GridLeak modes were used in parallel calibrations of the pp510 dataset (work from Matt Posik). Differences between these two calibrations have been examined (Yuri Fisyak & me).

Impacts on Residuals:

Here is a look at the along-the-padrow residuals vs. radius from global tracks that have a primary partner (red) and the primaries (black), 3D is on the left, Full is on the right:
(NB: it is very likely that I still have a considerable amount of pile-up in these plots; it's hard to know if this explains some of the overall remaining residual structure.)
(all units are [cm])

Impacts on momenta:

It appears that for globals, the pT impact is roughly Δ(pT)/pT = 0.0075 * q * pT, and for primaries is roughly Δ(pT)/pT = -0.005 * q * pT (opposite directions). There are undoubtedly dependencies on track location (e.g. pseudorapidity) to be reviewed (this was demonstrated in the presentation last August).

Note also that this is possibly larger than known h-/h+ issues where the constant is more like 0.002, but isn't known to have luminosity dependence, so it will be valuable to get confirmation (possibly via TOF invariant mass) of whether this new model is hurting or helping.

Low statistics bump may be for TPC sector 20, which saw a larger correction change than any of the other sectors (remains to be confirmed). Differences between positively charged and negatively charged tracks beyond the sign flip are not yet understood.

Impacts on Efficiencies:

(NB: these are relative efficiencies, i.e. tracks found in one case divided by tracks found in either case)
Recover of TPC Sector 20 is apparent (though incomplete).

In 500 events, the calibrations find the following numbers of tracks passing quality cuts (see GridLeak 3D vs. Full comparisons for details on the quality cuts):
  Globals Primaries
3D 13474 13602
Full 13468 13605