To BRAHMS collaboration
From  S.Sanders and F.Videbaek

Summary of Brahms Centrality Detector meeting held on Dec. 19, 1997:

The purpose of the meeting was to review the recent work on the 
Centrality Detector design and to decide which of the three different 
options that have been proposed for this design will be developed further.  
The three options that have been considered are 1) a pure Si strip detector 
array, 2) a pure scintillator tile array, and 3) a hybrid array composed of 
both Si strip detectors and scintillator tiles.  

As was clear from the last Collaboration Meeting, there is no support at 
present to proceed with a pure Si strip detector array.  Since it appears 
through GEANT simulations that a tile array will achieve the minimum 
requirements for the centrality measurement, the low-cost and simplicity of 
using a tile array argues strongly in its favor and  has focused the 
current discussion on options 2 and 3 above.  Recent measurements on the 
performance that can be achieved with the tile/fiber arrangement confirm 
the earlier results presented at the August collaboration meeting. A more 
thorough investigation into the cost of the tile/fiber array also leads to a 
consistent result to that presented in the August meeting.

Most of the discussion in the recent meeting focused on the 
feasibility of a hybrid array using a 
combination of Si strip detectors and scintillator tiles.  Recent measurement 
have shown the feasibility of using a relatively simple and modest cost 
electronics scheme for reading out the individual strip detector elements.  
It has also been discovered through a literature search that the UA2 
collaboration at CERN has used an array consisting of the same type of Si 
strip detectors as proposed for Brahms and has achieved good identification 
of single MIP events with this array.  A reasonably detailed mounting 
scheme for a hybrid array consisting of both Si strip detectors and 
scintillator tiles has been developed.  More details on the recent 
measurements concerning the Si strip detectors and the design for a hybrid 
detector mount can be found on the Brahms Web site.  Although the details 
of how the Si strip detectors might be read out differ somewhat from that 
proposed in the August meeting, the overall cost of a hybrid arrangement 
is essentally the same as that proposed in August. 

The advantages of the hybrid arrangement are seen as follows:
1)  The centrality measurement is less sensitive to corruption by 
anomalously large energy loss events.  These events can arise by the 
scattering of more ionizing fragments into the array and by secondary 
scatterings where the resulting particles traverse the array at an oblique 
angle.  The greater segmentation of the Si strip detectors and the small 
depth of the individual elements makes it less likely that a significant 
fraction of the array elements will be affected by such events.
2)  The Si strip detectors will  result in a centrality measurement with 42 
bins in pseudorapidity, as compared to the 8 bins possible with the 
scintillator tiles.  This greater segmentation should facility comparison with 
the measurements of the other RHIC experiments.  In particular, the 
greater segmentation will be important if the different experiments quote 
results for events gated on particle multiplicity within a limited range of 
pseudorapidity.
3)  By combining the results of the two independent measurements of 
particle multiplicity using the strip detectors and the scintillator tiles, it 
is possible to obtain a considerably better centrality measurement than would 
be achieved with either measurement alone. (This assumes the type of strip 
detector array geometries that are feasible within the cost restrictions of 
Brahms.)
For the above reasons, it was decided at the meeting to pursue further the 
hybrid arrangement for the centrality detector.  

A number of additional open questions were developed at the meeting:
-There is still a question as to the most effective trigger arrangement.  In 
particular, there is a question as to how many elements of the array need to 
be incorporated into a centrality trigger for event selection with Si+Si 
reactions.
-The mechanical support of the hybrid array from the floor needs to be 
more fully considered.  The current design shown on the Web site would 
interfere with part of the angular range covered by the forward 
spectrometer.  
-A significant improvement in the Si strip detector energy resolution may 
be achieved by incorporating a shaping amplifier after the preamp.  The 
best way to do this has to be determined.

Since the meeting a couple of additional questions has emerged in
discussions.
- The mass of the support structure, in particular that tied up in the
holder for si and pre-amp seems large. Way to reduce this should be
investigated.
- The large area and thickness on the tiles are responsible for the
anomolous events seen . Are there way to reduce this contribution to
make them less vunerable to this effec.e

Several "immediate action" items were identified at the meeting: 
-The order for the scintillator tiles should be placed with the Wayne State 
group in January. This will allow Wayne State to produce the tiles needed 
for Brahms before they start production of the tiles for Star.
-The  additional GEANT simulations to explore the question of how to develop a 
level 1 and/or level 2 trigger also need to be done soon since this relates to 
the overall electronics scheme and budget. 
-The Brookhaven group will  check on the types of shaping amplifiers
already developed and in use there to see if any may be appropriate
for use with the strip detectors.  
-Finally, the MOU's needed to start construction of the Centrality
Array need to be done soon to allow to the timely procurement of parts. 


The slides presented at the meeting are available on the web as
http://rsgi01.rhic.bnl.gov/export1/brahms/WWW/private/detectors/mult/centrality.html
This link can be found via the BRAHMS private home page.

If you have comments, questions on this issue please forward them to
us and YKLee.
Steve Sanders: sanders@kunucl.phsx.ukans.edu
Yung K. Lee: yklee@jhu.edu
F.Videbaek: videbaek@bnl.gov