MVD Internal Alignment

1
MVD Internal Alignment
Tokyo meeting, 24/06/2002 Takanori Kohno

• Introduction
• ?2 definition
• Alignment procedure
• Residuals before alignment
• Summary

2
Introduction

• In order to exploit the best position measurement
  with MVD, relative positions between MVD sensors
  has to be known precisely.
• Two geometries exists now. (design and survey
  geometries)
• With the survey geometry, position resolution of
  40µm is achieved.
• The aim of the alignment is to reduce the
  uncertainty to about 20 µm.
• This study is based on 280k cosmic events taken
  in JADE hall last year. Since there were no
  magnetic field, tracks were fitted with a
  straight line.

3
?2 definition
Alignment parameters shift and rotations.
shift SX, SY, SZ rotation RX, RY,
RZ for every sensors.
Q
P
px, py, pz direction of the track at the
intersection point.
Since theres no measurement on z-axis ? no
sensitivity to SZ.
4
Matrix calculation
5
Alignment Procedure

• Loop over all sensors.
• Loop over all tracks and hits.
• (Fit the track by excluding particular hit.)
• Find the intersection of the track and the
  sensor.
• Calculate the momentum of the track at the
  intersection.
• Calculate the residual of the hit and the track.
• Update matrix M and vector V.
• Determine the alignment parameters (shift and
  rotation).
• Update the geometry of sensors all at once.
• Check whether the alignment parameters have
  converged.
• If not, goto 1.

In this study, tracks were fitted only once with
all hits associated with it.
6
Residual distributions
7
Some ideas for further study

• First of all, perform the iterative procedure and
  see how it works.
• After each iteration, see the following
  quantities.
• Residual distribution.
• Alignment parameters (shift and rotations).
• Maybe useful to fix the relative positions of
  sensors in the same half-module. ? this enables
  2D measurement for each detector element to be
  aligned.
• Move on to helix tracks. ? tracks with higher
  momentum due to CAL and B field.