Minutes of CMSUK-ECAL meeting, 26-Sep-1996

Minutes of CMSUK ECAL meeting, Bristol 26/9/96
----------------------------------------------


1. New CMS offices at CERN  (D. Cockerill)

The building is ready, despite the protestations of the architect...
People will start to move offices soon. The xterminals which are
currently in 32-1A-06 will be shared amongst the new institute
offices, and some extra ones will be bought when things have settled
down after the move. It appears that the CMS secretariat do not have
up-to-date lists of UK collaboration members - a representative from
each institute should supply them with current information.


2. Preparations for CMSUK management committee  (P. Hobson)

The first meeting of the management committee is now scheduled - the
relevant people have received information on this already. The first
meeting is to be an exercise in "information collection", and deciding
upon the pro formas for future meetings. These are likely to include:
reports from the bugdet holder; reports on the ECAL and tracker from
Dave Cockerill and Geoff Hall; reports from institutional
representatives (verbal reports only required for the first meeting).
It was commented that these meetings should aim to present our work in
the wider context of CMS as a whole. To this end, Bob Brown is
preparing a "CMS overview" for the first meeting - comments are
welcome.

Preparation are also under way for the resource review board meeting
with George Kalmus/Ian Corbett on the 9th October.


3. Report from ECAL technical board meeting  (D. Cockerill)

A staging proposal for the ECAL has been put forward by H. Hofer for
discussion at the institution board meeting in November. He envisages
a two-stage construction for the ECAL: stage one would be the barrel
crystals and endcap preshower only. The lack of endcap could be
partially compensated for using a finely-segmented endcap HCAL. Stage
two, which would follow a review in 02/03, would complete the ECAL. In
this way, the completion of the ECAL by "2006-8" can be assured. It is
possible that a disproportionate capital (non-crystal) contribution
would also be required from institutes outside ETH until 2002.

It was pointed out that this proposal is disastrous from a physics
point of view! An ongoing discussion is apparently taking place
between Hofer and M. Dellanegra/J. Virdee. It was commented that the
UK has little leverage on financial decisions made within the ECAL
group, as our contribution is very small compared to that of ETH (who
are contributing around 40 MCHF towards the 80 MCHF ECAL budget), and
that we should use what political influence we have only when we are
in a firm position to make out contribution the ECAL effort.

RMB commented that we must continue to assume that 2005 will be the
LHC timescale, regardless of the political situation.


4. Report from technical board meeting  (R. Brown)

A handout of transparencies from a talk by P. Lecoq was distributed. A
few comments were made by RMB:

The major technical problem at the moment is the low-dose crystal
radiation damage. This varies considerably between crystals, but even
the best crystals so far experience around a 25% variation in light
yield. This is around an order of magnitude from the target value. The
implication is that there the scheduled "pre-production" order of a
large sample of crystals in 1997 will not take place (this is an ECAL
milestone). In effect, there has been a one-year slippage.

The other large problem is the neutron damage of the APD's. However,
it is now believe that these effects can be understood in term of a
simple silicon damage model for a wide range of diode structures. The
latest devices produced by Hamamatsu combine low capacitance with
relatively low charged particle sensitivity, but are still an order of
magnitude away from the required performance. This implies that the
ECAL may have to be run cold, though this may lead to "freezing-in" of
the fast radiation damage.


5. Report on reflective crystal wrappings  (D. Cockerill)

Some work performed on various types of thin-film reflectors for light
collection in the crystals was presented at Aachen.  The best
Al/polyester films give around a 15% decrease in light yield compared
to Tyvek. Silver films are slightly better. The current baseline
design is therefore a 25 micron aluminium film, incorporated into the
alveolar structure. A copy of the full report is available from Dave
Cockerill, and will be in the Aachen proceedings.

For the uninitiated: Tyvek is a strong white hydrocarbon polymer
material with a variety of industrial uses, which turns out to be an
efficient reflector for crystal wrapping... however, the thickness
uniformity is not good enough for use in the precision mechanical
structure of the CMS ECAL. The manufacturers are not interested in
improving this, as particle physics is not a major consumer of tyvek!

DC also raised the issue of reflector uniformity along the crystals
(or the lack of it). This will form part of next year's testbeam work
at CERN - can we contribute at ISIS? A short report was given on the
work currently going on at ISIS: Currently, measurements are being
taken by scanning along one crystal. Can this be scaled up to 2x6
units (or even 6x6 units if these are used)? Comparison of the ISIS
results with GEANT simulation is good, and noise levels are as
expected. (A comment was made that the importance of electronic noise
in the ECAL may have been neglected until now). An upgrade of the ISIS
testbeam is in progress: a new Si/scintillator beam position
measurement system; time-of-flight apparatus to allow seperate
D/proton/neutron measurements; an automated scanning table system.
(See section 8).

Peter Hobson commented that the ISIS work could form a pilot project
for the regional centre quality assurance programme, with systematic
checks on the stability of measurements and so on. It was pointed out
that this required a committed and willing team of people to do the
work. PH agreed to come up with a proposal regarding this idea.


6. Report on August/September CERN testbeam results  (G. Davies)

The main point was the discrepancy in resolution measurements on the
same crystals between RAL (radioactive source/PM tube) and the X4
testbeam (real beam/APD). The two sets of results were compared with
monte-carlo simulation of a crystal with perfect uniform response,
giving a yield and 'added width' measurement for each. There appears
to be some correlation between the yield measurements from RAL and
CERN, but non in the added width results. These discrepancies could be
due to: different active area of the photodetectors (though the PM
used at RAL was masked); different wrapping of the crystals; different
spectral response of the photodetectors. An attempt was made to
compensate for the device area differences using a ray-tracing
simulation, and it seems that this is not the most important effect.
The crystals are/will be remeasured at X3 and PSI.

The radiation damage results from H4 look promising, but there are
problems with the stability of the 'reference' laser source! RMB
suggested using the OPAL system of laser and liquid scintillator,
providing a stable blue light source.


7. CMS bulletin (RMB)

This now exists, and is due to appear four times per year. Available
on WWW, or at CERN.


8. Discussion of ECAL spend

Dave Cockerill gave more details of the proposed work at ISIS. The
testbeam upgrade is under way (see above), which obviously requires
money. A new APD test box has also been built by R. Stephenson. It
includes temperature and bias voltage control, and a variety of
light/particle sources can be used. There is a possibility of placing
some APD's inside the ISIS ring to experience LHC-level neutron flux
at a variety of ambient temperatures.  These devices could be
monitored from the ISIS testbeam setup. It may be possible to enlose a
test box in the beam in order to measure APD gain continuously. The
work will hopefully confirm the correct temperature to run the devices
at in an 'accelerator-type' enviroment. This also needs funding.

PH asked whether the above projects counted as "regional centre
infrastructure". It would seem likely that the first does, but not the
destructive APD tests in ISIS.

A lively discussion followed concerning the reallocation of funds
between different projects in the UK ECAL expenditure breakdown. It is
accepted in the UK that the 0.13 Mpounds allocated for regional centre
infrastructure is not sufficient, but we cannot reallocate money from
mechanics/electronics/regional centre assembly to infrastructure under
CMS rules, even though the amounts allocated to each of these are
fairly arbitrary. John Connolly pointed out that around 100K of the
infrastructure budget is allocated to the purchase of an ACCOS machine
for crystal measurement - should this cost really be born by us?
Should the regional centres even be measuring the crystal properties
at all - perhaps the manfacturer should take the burden of quality
control? It was pointed out that the "usual" way of purchasing
materials was for the manfacturer to guarantee to conform to a
specification. However, in this case, there seems to be no other way
of testing the manufacturer's conformance.

It was generally felt that there were severe doubts in the UK
concerning the current ACCOS scheme. GD agreed to meet with JC, PH in
order to draft a letter expressing the UK feeling, which will be
circulated.

JC made a general comment that the amounts allocated to mechanics and
electronics look too small by at least 50% to cover our aspirations.
It was pointed out that there is more than one institute bidding for
some of our areas of interest, and that collaboration between
institutes seems inevitable. RMB noted that our growing feeling for
the real cost of items should be fed back into the cost book exercise.
There is a revision of the figures due in 1/97, so costings are
required by 11/96 at the latest.

Representatives from each institute gave brief reports on their likely
regional centre infrastructure costs:

Richard Head (Bristol): The main cost will be the conversion of a room
in the laboratory into a clean/temperature controlled enviroment. No
costs are available on this yet. We should look into the possibility
of common purchases between institutes, in order to achieve economies
of scale. Figures will be obtained for the next UK ECAL meeting. It
was pointed out that costings are very difficult to obtain until each
institute's exact responsibilities are finalised. In any case, the
money for clean rooms, etc, does not need to be spent until maybe six
months before the startup of the regional centre.

Peter Hobson (Brunel): Main costs are with the APD testing programme.
This is possibly a quite short-term spend. No real figures yet, but
probably around 10K. Figures next meeting.

John Connolly (RAL): Most money will go on handling/geometrical
measurement equipment, plus whatever facilities are required to test a
2x6 module. This assumes untested crystals in, 2x6's out. No firm
costings possible until the regional centre details are fixed (18
months?).

Gavin Davies (IC): The construction work for the assembly area is
almost finished, but nothing further can be committed until the
regional centre plan is known. An up-to-date estimate of the costs,
and information on the likely split of funding between IC and CMSUK,
will be available next meeting.

Helen Heath (Bristol): The costs of the CRISTAL database system will
be assessed ASAP, but this is difficult until the system
specifications have been finalised. It also depends upon whether the
ACCOS/CRISTAL machines can be one and the same thing, and upon the
reliability of UK networking, and so on. Perhaps 15K for a workstation
will be needed.


9. Report on ACCOS meeting  (G. Davies)

The meeting was apparently poorly attended. Hopefully, the tender for
machines will go out soon, and a decision on the supplier will be made
by the November ECAL week. The machine should arrive at CERN after
tests around 9/97. There is a possibility that IC could be involved in
supplying some mechanics for the machines, but there are no plans to
become actively involved in this for at least a year, possibly not at
all, depending on money.

J-M LeGoff made a report at the above meeting on the CRISTAL project.
A decision has to made on the operating system to be used for ACCOS
computers - NT is cheaper, but unix is a proven system. CRISTAL will
have to interface to either, but this may not be easy in the UK RC if
we wish to run both types of software on the same machine.

Some figures are now available on the ACCOS machine: the cost is
around 150KCHF, with no climate control. Power consumption will be
around 2kW.


10. Report on mechanical design  (J. Connolly)

A report on the status of the ECAL mechanical support design was
circulated. Briefly, there are two "models" being developed (not
"prototypes", as we don't want to be committed at this stage), the
costs for these would be perhaps 2-3K for the first, and certainly
less than 10K for the second. A general point is that the 2x6-based
design is not considered to be stable. It was emphasised that an
extremely robust design is required in order to ease assembly in the
regional centres, installation, etc. Monthly meetings with people from
Ecole Polytechnique will be taking place from now on. EC are to
produce some 2x6's to their design (with embedded aluminium) for
various appraisals.


11. Presentations on physics topics

Dave Newbold briefly assured the meeting that physics simulation work
is taking place at Bristol, in the context of the calorimeter trigger
development.

Peter Hobson gave a talk on the possibility of detecting anomalous
triple vector boson couplings at CMS. A parton-level first look
indicated that we could measure something (but not with no ECAL
endcaps!). Further analysis will take place, probably by Kate Mackay
at CERN.

Gavin Davies expressed an interest in working on higgs/SUSY search
topics when time permits. Two new CMS people are joining the IC group:
David Britton, as a lecturer, and Elizabeth Martin, as an advanced
PPARC fellow.

Bruce Kennedy presented a study of the Z' discovery potential at CMS.
Again, this would appear to be a promising channel at first sight -
and one that can be attempted with high luminosity but a staged ECAL.
How much we can see strongly depends on the Z' mass, and the assumed
model.


12. AOB

Next meeting will be at Imperial College, 10.30 on Friday 29/11/96


Dave Newbold 30/9/96