--------------3B8251FBE79A9E8D9031BABE
Content-Type: text/plain; charset=us-ascii; name="upgrade_05_01_98.txt"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="upgrade_05_01_98.txt"
M E M O R A N D U M
TO: Distribution
FROM: B. Mecking
SUBJECT: Hall B Upgrade Meeting Notes
DATE: 6-May-1998
MINUTES OF THE 1-MAY-1998 HALL B UPGRADE MEETING
1. Constrained kinematical fitting (Mark Ito)
Instrumenting the target region of CLAS with a full coverage detector will
allow us to know the direction of all charged particles in an event. For
charged particles heading for the torus coils this will be the only
tracking information available. In order of increasing complexity, one can
distinguish the following cases:
a) The final state is completely known with the exception of a single
charged particle for which only the direction is known. This is a
relatively simple case since, in principle, the 4-vector of that
particle can be determined from 4-momentum conservation. In practice
knowing the direction will lead to a 2-C fit and improve the overall
resolution.
b) The final state is completely known with the exception of two charged
particles for which only the directions are known. This is a more
complicated case since the 4-vectors of these particles can no longer
be determined from 4-momentum conservation. Only the 3-momenta can be
determined without making any mass assumptions.
The details of the kinematical fitting procedure used are described in
Technical Note 309
KINFIT: A Program for Non-Linear
Least-Squares Fitting with Constraints
Mark M. Ito
The process
ep -> e' p pi+ pi-
was used as an example to illustrate the power of kinematical fitting
and the influence of measurement errors. It was assumed e' and p had
been fully reconstructed, and only the directions were known the
pions. To proceed with the fitting, the assumption was made that the
missing particles were pions (generally a good assumption in case of a
hydrogen target, especially if one nucleon has already been detected).
This again reduces the problem to a 2-C fit.
As a first result, knowing the directions of the 2 pions with an
accuracy of 5 mrad (rms) reduces the uncertainty in the determination
of the two-pion mass by a factor of 2.5 compared to not knowing the
directions.
2. Potential layout of the CLAS inner detector (B. Mecking)
A sketch of a potential layout of the CLAS inner detector was presented
and discussed. The arrangement consists of the following components
(starting from the outside heading in the direction of the target):
a) Photon detector in front of the coils
This detector covers an angular range up to about 90 degrees (seen
from the standard target position. Radial space allotments is about 18
cm. Width approximately 16cm (slightly bigger than the width of the
torus coils). Potential technical solution: dense crystals with light
readout that is insensitive to magnetic fields).
Interesting engineering problem: this detector can not be moved in as
one piece -> needs to be inserted in pieces.
b) Complete coverage for charged particles
This detector is located in the present Region I area (i.e. the
polarized target can still be used). It covers an angular range up to
about 130 degrees (seen from the standard target position. Radial
space allotments is about 18 cm. Desired properties are:
- minimum amount of mechanical structure (thickness) within the open
gaps. More thickness is allowed right in front of the photon
detector.
- high rate capability to cope with electromagnetic background (note
that the mini-torus will have to be replaced by other magnetic
shielding arrangements which may not be quite as effective).
- preservation of sector structure for ease of triggering
Potential technical solutions will be discussed at the next meeting.
c) Particle ID layer
This topic had been discussed by Paul Stoler at the previous meeting.
If particle ID could be accomplished in front of the tracking (instead
of behind it) it would reduce the size of this detector dramatically.
A potential technical solution is based on low index of refraction
aerogel material (n = 1.02 - 1.04) as the Cerenkov medium followed by
a focusing mirror system which reflects the Cerenkov light onto a
position sensitive photon detector.
This detector could not coexist with the present polarized target.
Next upgrade meeting:
Friday, 8-May-1998 9:30 - 10.50 a.m. (Seminar at 11:00 a.m.)
Trailer City Room 84
Topic: Recent developments in tracking detectors (Howard Fenker)
Distribution: CLAS Collaboration
--------------3B8251FBE79A9E8D9031BABE--