CERN engineering forum report

From outside CERN:
Phil Burrows, Toshiaki Tauchi, Junji Urakawa,
Hitoshi Yamamoto, Hiroshi Yamaoka, Wolfgang Lohmann,
Norbert Meyners, Franz Czempik
Ariella Cattai, D Forkel-Wirth, J. Osborne, M. Butin,
A. Herve, A. Gaddi, D.Campi, G. Faber, J-L. Baldy,
M. Battistin, S. Stapnes, F. Hahn, M. Oriunno,
M. Kotamaki, G. Perinic, C. Schaefer, H. Gerwig,
C. Lasseur, S. Roesler, H. Vincke, H. Postema,
P. Petagna, J. Baechler, M. Nessi, L. Linssen + more

On Oct. 12, we had a 5 hour tour of ALICE, ATLAS,
and CMS. At ALICE, the tour was centered around the
TPC which was not the focus of this visit, but still
very interesting for some of us. (such as 50% occupancy
tracking etc.) ALICE is using the left-overs from L3
as much as possible, including the yokes, the coil etc.
The hall is also recycling from L3.

At ATLAS, the M. Nessi gave us the tour. ATLAS is
a very large detector which sits in a cavern it
barely fits, and assembled in situ as a big ship.
Each end of the detector is serviced by separate
shafts, and served by two (-if I remember correctly)
65t cranes. The HCAL is also assembled underground -
at least about 90% of it. Muons are inserted into
the gaps of the huge toroid magnets piece by piece
in situ. So, it can be said that it is truly a
completely underground assembly. There are large number
of mid-size halls where electronics, cooling water system,
transformers are located. Everywhere, people were
working hectic. Also, there are several personnel
lifts moving constantly and working on assembly
of the Big Wheel etc. They swing quite largely
and seemed like they can hit the delicate part of
the detector often. M. Nessi was loathing the
decision 15 years ago to use toroids.

CMS, on the other hand, was quite calm may be because
it was after 4pm, but probably not entirely because of
it. The detector was assembled more or less completely
on surface, and being tested. In the movie, ECAL was to
be assembled underground, but in reality it is going to
be lowered underground nearly completed. There was a
jig to install HCAL sitting next to the detector. It seems
that the assembly procedure of CMS in general was conceived
with more care and innovation than that of ATLAS.
The air pad had gasket that touches the floor and the
air does not leak much normally. The entire air pad
is serviced by one small compressor and pressure vessels
that can fit in a 10 ft cube. The air pad can go over
a step of a few mm, and you would hear noise of air
leaking then. (no activation of airpads were seen
during the tour). It was recommended by Alain that
steel sheets are used under the air pads so that
air cannot get into cracks of concrete which may lift
the entire concrete slab. The CMS hall has a floor
that is tilted by 1.2% since the LEP itself is tilted.
Still, the airpads can be used without much trouble,
he says. The 2000t gantry crane with 4 hooks was already
there and ready to do the work. The first piece will go down
in two weeks. The gantry crane used existing vertical
parts and the horizontal part on top was made for this
building. The shaft cover is 2m thick and holds 2000t,
and can be opened or closed in 30min. CMS has a large
service cavarn - smaller in cross section but longer
then the detector cavarn - which contains counting
rooms, cooling plants, transformers etc. There is a
separate service shaft that is more or less filled
up with elevator, cables, ventillation pipes etc.
There are quite a bit of spaces under the CMS floor
which contains cables, power supplies etc. Originally
these spaces were not used for anything, but now it is
fully utilized. I asked where they intended to put those
items under floor originally. There was a laugh. The cable
chains were used extensibly, and they seem to work quite

On Oct. 13th. talk sessions began at 9am and speakers
were mostly from CERN, except the LDC talk and the talk
that explained why we were there (I showed slides made
by Andrei). You can see the talks at

My internet connection went down the night before, and
did not get the most up to date slides from Andrei,
so the talk I gave used the penultimate slides with
modifications based on discussions I had with Andrei.
I left the most up to date slides on the web since
it is better annotated, even though it is not the
actual slides used.

On the push-pull issue, there was not much comments
from CERN. It is probably understandable since we are
in an uncharted territory regarding the really successful
push-pull scheme.

There were many important information obtained this
time, and we should be grateful that CERN people have
put so much time and effort to help us. At the end of
the session, when we told them that we will make good
use of the information, there was big applades and
table thumping.

---Addition by Wolfgang----
For the tests of subdetectors several alcoves are added to the surface
hall, to ensure calme and stable working conditions
The infrastructure built for the surface tests will to a large extend be
used for the operation of the detector in the beam.

Are pads are used for the movement of detector parts over larger
distances. When detector parts are combined, the last ~10 cm
are moved using grease pads. The latter are better suited for approaching
large masses to millimiter distances without damage.

Each slice of the central CMS detector is supplemented by
galleries mounted outside on the iron yoke. These galleries
contain racks mainly for low power supply of the FE electronics.
Flexible cable chains are used for electrical, gas and cooling supplies
of each slice and contain all signal cables.
These cable chains allow the shift of each detector slice
along the beam axis, important for maintainance.

---Addition by Norbert----
- CMS can open each side by 10m.

- The service cavern at CMS is 80m long, 15m wide with 4
levels. Two cable distribtion of 1.8m height and two for
electronic, cryo, gas, and other supplies.

- 20% of the ATLAS muon chambers can not be exchange any

- It takes about 20 days to access the inner detector of

It may no be explicitly mentioned, but
- CMS and ALTAS have a survey tunnel/balcony with a
stretched wire and a hydrostatic leveling system connecting
one final focus system to the other and connecting the
detector to the machine.

---Addition by Phil----
* The radiation safety issues are 'moving targets', and even a conservative margin can disappear later as the exposure limits are tightened by the regulatory bodies. This could be of special relevance for the 'self-shielding' concept in the push-pull scenario. For what it's worth, one of the CERN engineers I talked to was sceptical about our simulations to date of beam losses in the detector, given that realistic cracks in the detector may not yet be fully worked out.

* Two access shafts to the IR hall, for safety reasons, seems mandatory. We might have a problem with this in our current model of 1 shaft per detector side of the single IR hall, especially if there is a shielding wall in the middle.

* 'Redundancy of cranes' impressed me. A single crane needs maintenance and cannot be scheduled with full efficiency. Having 2 cranes allows for more flexibility. Even having two 'independent' hooks on a single gantry allows for flexibility in the event of breakdown, delays, shift crew issues, etc.

* I was impressed by the 'underground labyrinth' of passages below CMS that allows for cableways, and rising platforms to service the detector in the open position.

* There seemed considerable scepticism to the fast-turnaround push-pull model. One person privately doubted we could do it in less than 2 weeks, and there was general mirth that it could be done in a few days.

Overall I was very impressed with the presentations and the wisdom gained by our CERN colleagues. They provided an excellent meeting for us, at a busy time for them. They did us a great service.