Minutes of the Physics Commissioning Meeting, Sep 15

[Thorsten Oest]

Minutes of the Physics Commissioning Meeting, Sep 15
====================================================

o VDS (Martin Braeuer)
- CATS reconstruction (transparencies by S. Masciocchi):
Data from the rate scan (run 9769-9789) were studied using
the CATS reconstruction package.
* Alignment check: The beamspot was found at the same position
for tracks from different quadrants.
* The number of tracks found in sector 1 and 4 is reduced as
expected due to the missing SL 1-3. In average 16 tracks/event
were found at 10 MHz.
* The #tracks/event scales linear with the target rate. The
10,15,20 MHz runs with outer 1 wire show a reduced number
of tracks. This is not yet understood.
* Comparison with MC: The beamspot is somewhat broader in MC:
MC : sigma_x = 229 mu, sigma_y = 809 mu
data : sigma_x = 190 mu, sigma_y = 615 mu
The #tracks/event, #hits/track, #unused hits/event and the
chi2 distribution for tracks in data are in reasonable agreement
with MC.

- Martin showed plots with reconstructed vertices. The interaction
region is clearly visible in three dimension. In a 7 wire run
vertices off all wires are visible. The wire position (x,y,z) was
determined for the station 1 wires. Deviations of the order of a
mm from the nominal z position were observed:
z position
from vertices nominal target position
above 1 -14.8 mm -13 mm
below 1 -14.6 mm -14 mm
inner 1 -8.5 mm -9 mm
outer 1 -11.4 mm -9 mm
This is not in contradiction with the nominal target position given
by the target group since their measurement was not expected to be
better than a mm.

The geometrical acceptance for the present VDS setup were found
to be:
phi = 2.3 - 5.4, theta = 0.02 - 0.2 and
phi = -0.8 - 2.3, theta = 0.02 - 0.1

The alignment was checked after pot movements. The constants
are stable within about 1 mu.

o Alignment
- RICH internal alignment (Denis Dujmic)
The RICH internal alignment has been studied using
conversion behind the magnet. The photon direction as measured
by the ECAL cluster and target constraint can be compared with
the RICH direction for the electron/positron pair.
In previous studies large deviations were found in data as
well as in MC. The reason for this is a phi dependent correction
which has to be applied for charged tracks (because of the
deflection in the magnet). For conversion this leads to a
systematic shift in phi. Taking this into account the internal
RICH alignment was found to be correct within 0.75 mrad.
Work is underway to study the vertical alignment using VDS
tracks and to disentangle the effects by spherical and planar
mirrors

- Global alignment (Thorsten Oest)
Global alignment constants (VDS/ECAL/RICH) were determined for
the present VDS position. The constants of all runs are
stored in a database and will be applied automatically
in the offline analysis (from ARTE-03-02-r1 on).
Systematic effects have been studied by looking at runs with
all 8 wires. The constants determined from the different runs
are in agreement. An internal wire misalignment in x,y with
respect to the target position in the event stream was found.
above 1 1.09 mm above 2 -1.34 mm
below 1 - below 2 0.08 mm
outer 1 -0.05 mm outer 2 0.73 mm
inner 1 0.07 mm inner 2 -0.47 mm

- Survey measurements (Achim Spengler)
From the survey measurements it turns out that the
* ECAL is shifted in horizontal direction with respect to the
proton axis by 8.8 mm
* the RICH is rotated by 0.85 mrad (horizontal)
* the magnet is rotated by 0.53 mrad (horizontal)
The measurements are in agreement with the software alignment.
In the last alignment meeting the decision was taken that this
should be corrected in the Christmas shutdown.
There was also a self-made (by HERA-B) measurement of the VDS
position with respect to the magnet (distance between the VDS
vessel and the magnet). A 3 mrad vertical rotation was measured
for the VDS (the measurement has large errors). This is in rough
agreement with a value of 1.5 mrad from the software alignment
although there a rotation with respect to the ECAL was measured.

o data reconstruction (T. Oest)
There are a few new packages added to the rc99 reconstruction:
* CATS can be called by the kuip command SUBD VDS ON CATS
It can be run in parallel with HOLMES to compare both
algorithms.
* global alignment constants are read in automatically
* track reconstruction: MARPLE matching, RANGER refit
* RICH particle ID (rire package)
* primary vertex reconstruction (vt package)

M. Bruinsma mentioned that the reconstruction & data quality is
running online on the 4LT farm now.

-------------------------------------------------------------------
| Thorsten Oest Phone : [49] (40) 8998 3008 |
| Uni HH / HERA-B Fax : [49] (40) 8998 4018 |
| Geb. 61 /103a e-mail: Thorst...@desy.de |
| |
| Luruper Chaussee 149 |
| D-22761 Hamburg |
-------------------------------------------------------------------