NuMI-B-87
                                                         May 5, 1995
                                                         Dan Crane
                                                         Gordon M. Koizumi
                                                         Anthony J. Malensek

          Trip Reports of the NuMI Beam Group to CERN, March 1995



Anthony J. Malensek
Engineering Physicist
708-840-3920
Research Division/Operations Department

and

Gordon M. Koizumi
Engineering Physicist
708-840-3068
Research Division/Operations Department
April 6, 1995


Dates of trip:	March 22, 1995-April 3, 1995
Destination:	CERN, Geneva, Switzerland
Purpose:	Gather information on the design of the CERN
	neutrino beam as well as deep underground 
	detector halls. 

ABSTRACT

After operating about 9 months, the CERN beam turned off about 3 months ago and
will be coming back on soon.  This was an opportune time that allowed us to
access the cave containing the neutrino beam line and two of the LEP
experimental halls.  Meetings were held with the designers and builders of the
operating systems which will be useful in our design of NuMI (Neutrinos at the
Main Injector).  Mechanical and electrical information was gathered about the
Horn focusing system, the target, beam monitors, alignment, and shielding.  In
addition the civil construction aspects of interest were the air handling
systems, the cranes, cable installation, fire protection, heat exchange,
physical movement, utilities layout in the shafts which lead from the surface to
the underground caverns, and size changes in the cavern after construction.

A)	Purpose.
The purpose of this trip was two fold.  One was to obtain detailed drawings and
operating experience of the focusing horns and the target used in the CERN
neutrino beam line.  The second was to obtain the conceptual design and
philosophy behind the building of large underground experimental hall located
deep underground.

B)	Summary of significant findings.

1)  Target.
a) Experience with Be has been good, although at high intensities (2 E13/few
msec), the downstream end of the Be rod which connects to the "10 cm
sub-assembly" fractured on the first couple of pieces.
b) They have run with a large radius (1 cm) carbon target with their Narrow Band
Beam at 2 E13/23 microsec.  When the target was analyzed after the run, it
showed no damage whatsoever.
c) They saw no difficulty with our proposal to use thin carbon rods in a
segmented target and under vacuum (cooling via black body only).
d) We have in mind a simpler version of their "10 cm sub-assembly", and would
like to go perhaps with a 20 cm sub-assembly.  Sub-assemblies are aligned via
machining.
e) Move entire target box with 4 motors (X, Y upstream and downstream) that are
external to the steel shielding.
2)  Horns.
a) The stripline can be air-cooled.
b) A good radiation hard insulator has been found--Arclex (made in UK); though
not liked by the machinists, it is more difficult than metals, but is
machinable.
c) The Al alloy that they use which is doped with small amounts of Cu, Mg, Zn,
etc. is commercially available, not a "special order" to mix a batch.
d) An ultrasonic test of the alloy before machining is essential to see that
there are no flaws.  The neck piece is machined by CERN and takes about 900 hrs.
e) A copy of TM-1928 (A. Visser's electrical design) was given to J.M. Maugain.
One of their engineers will give comments and will also do a "transformer
design", modeled after their system.
f) The watch words are "careful" and "important".  It is clear that engineer to
engineer contact is vital.  This is not a device that somebody designs on his
own without having previous experience.  Some examples of the care that is
needed are, geometric centering of the center conductor at several points in Z,
machining of the inner surface of the outer conductor, inner conductor is
pre-stressed, geometrical symmetry of the input current (they have 4 points
where the stripline feeds onto the conductor). 
3)  Recirculated air system.  The design for this is very important.  
a) It must minimize the release of short lived radioactive gases/particulate
into the environment by having a closed circuit air system with heat transfer
done on the surface.  The  air that is brought back to the surface is filtered
again to remove any radioactive dust that might have been picked up while in the
beam area. 
b) It helps to minimize humidity in the outside air from condensing inside the
cooler tunnels.  The dehumidification takes place at the surface. 
c) Pre-filtering of the air from outside with screens which could be cleaned
easily with water rather than replacement of the filters was important,
especially in summer when insects and leaves were significant problems in
addition to dirt.  The filtering/dehumidifying/heat transfer areas were huge.
Those for LEP were huge buildings although they were each responsible for 1/6 of
the LEP ring.  
d) Redundancy of the air system so that the entire beam  would not require a
shut down while portions of the air system was undergoing maintenance/repairs.
e) ODH problems increased by this system and necessary precautions have to be
taken because of it.
4)  Magnets were shown which had connections which were aligned by the means of
pins.
a)  These magnets could moved into place with overhead crane whose position
could be controlled by scales painted onto the walls of the enclosure.  These
indicators were accurate enough that the crane could align the magnets by
lowering the magnets onto pre-aligned pins. 
b) Electrical and water connections are made just by having the magnet lowered
onto the alignment pins.  Anthony wonders whether the E.P.B. and 3Q120 type
magnets would be heavy enough to permit good electrical contacts if this sort of
system is adopted at Fermilab.  Modifications to this system should be possible
which would permit use with lighter magnets.
5)  Crane system (5 ton capacity) in the Pre-target/Target Hall area had serious
problems as initially designed and had to be redesigned before dependable
operation could be carried out.
a)  The crane travels upwards going downstream at 2 deg. 39 min. slope on a cog
rail system.  The beam itself is pointed upwards at 2 deg. 26 min.
b)  The crane now uses a ribbon cable system to provide it with power.  The
original sliding contacts on rails system was a source of constant problems,
especially because this was in a very high radiation area.  The ribbon cable
system is dependable but requires space for the ribbon cable which is carried
along on its own track system along the edge of the tunnel.
c)  The position of the crane indicated by marks along the walls.
The crane has fine controls for position via a slow drive.  The slow drive is
well worth the cost because it minimizes radiation exposure as well as potential
damage of components during installation/repair.
6)  Rotating door, closed circuit TV and personal ID. badges with magnetically
encoded strips used to provide controlled access procedures for entry/exit from
beam area.  Controlled access is only at one point at the enclosure, i.e., not
all doors are controlled access points, others are emergency egress doors.
7)  Ground water irradiation problem was a very minor concern at the existing
neutrino facility.  The statement made was that the "ground water problem can be
ignored in most places".
8)  Rosin impregnated floor at the Pre-target/Target Hall was claimed to be an
important factor in keeping radiation contamination under control.  Gordon was
impressed at the cleanliness of the Pre-target/Target Hall Area.
9)  Low Na concentrations is highly desirable where personnel access to areas
with high losses are required shortly after the turn off of beam, e.g., in the
region near the target or the horns.  Even 2% Na concentrations are already bad.
Standard concrete has high concentrations of Na which makes it bad for high loss
area.  CaCO3 is desirable.  CERN uses marble for shielding in its present
neutrino area because of this.  
10)  Foam system used as fire protection system in the Pre-target/Target Hall.
A new foam barrier was being erected within the hall at the time of our visit.
11)  A 9 m diameter vertical access shaft is used at L3.  The L3 Experimental
Hall is approximately 50 m below the surface.  
a) It contained an elevator (which was the PRIMARY MEANS OF EGRESS EVEN IN
EMERGENCIES) capable of carrying a large number of people at one time.
b) Emergency stairs which were the SECONDARY EGRESS method.
c) Ventilation ducts.
d) 7 x 2 m drop shaft.  The drop shaft was serviced by an overhead crane which
was part of the building enclosing the drop shaft.
e) Ducts and cabling with service stairs/landings for the ducts/cabling.  The
ducts/cabling was exposed for access to the person using the stairs/landings.
Engineers (Hatch and Gregory) showing us around the L3 shaft insisted that the
service stairs/landings were important to have since changes/additions to the
cabling requirements were frequent enough that it would be a mistake not to
include the service stairs/landings.  The ducts were supported only at the top.
Apparently there was enough expansion and contraction due to temperature changes
that multiple attachments of the ducts down the shaft was considered to be a
significant problem due to expansion/contractions.  Welding of pipes had to be
done near the bottom of the shaft rather than from up near the top.  Ar gas was
used as part of the welding process for the ducts but the weight of the Ar gas
caused problems when the length of the pipes got too long.  This problem was
solved by welding near the bottom of the shaft with the Ar flowing upwards.
f)  There was a large room at the bottom of the vertical shaft which was an
emergency shelter/safe area to house those personnel in emergency while waiting
for the elevator.  The door from the room into the experimental hall was heavy.
Air pressure due to air flow direction (down the elevator/stairs shaft) made the
door which had to be pulled to get into the experimental area somewhat "heavy"
to open.
g) Access to the L3 vertical shaft elevator was by means of individual
identification badges with magnetic strip which had to be inserted into a slot
and then removed to allow a turn style to admit one individual through.  This
system recorded the identity of the person going down and kept track of the
total number of individuals down in the experimental hall.  The badges had to be
worn at all times in a location that was visible while accessing one of the
experimental halls (an electrician who was involved in an accident could not be
immediately identified motivated this requirement).  Upon exiting, the
identification badge had to be used to go through the turn style again to record
that person's egress.  Room had to be provided at the top of the shaft for this
access control system.
h)  Lots of sampling of the air to detect gas leaks are done in experimental
halls.  Volume and pressure readings are monitored as they go in and come out of
the halls to detect differences.  Pressure gradients are kept very small to
minimize leak problems.
i)  Heat exchangers used multiple stage system to handle the pressure problems
between the bottom of the shaft and the surface.  Efficient plate type stainless
steel heat exchangers, which are more expensive initially, were used since they
are more efficient and take less space down below in the cavity than tube type
heat exchangers.  The plate type heat exchangers are also stronger and can
handle pressure differentials better.  Importance of keeping the water clean in
the exchangers was stressed to minimize the maintenance problems.  The pumps and
exchangers are noisy and were allocated a special cavity in the case of LEP.
While LEP and associated experiments required megawatts of cooling, the size of
the cavity associated with the pumps and heat exchangers were impressively
large.
j)  Wet fog type of fire fighting system is being installed at L3.  This is
primarily a personnel protection system and not an equipment protection system.
Fog type system is superior to sprinkler system since the fog can be drawn into
the fire by the draft caused by the fire itself to help put the fire out while a
sprinkler always has areas it cannot protect because of shadowing effects.
k)  Site pipes up to the surface are in place and greatly aid alignment.
12)  Graham Stevenson had number of helpful suggestions for the present civil
design being proposed for NuMI.  Among them were:
a)  The double labyrinth legs to the elevator from the pre-target area in the
present sketch for the proposed NuMI Pre-target/Target Hall is probably not
needed.   Moving this access point further upstream than where the present
MacDraft drawings indicate would be helpful in reducing the amount of radiation
scattered off the target which could travel down the access corridor. 
b) MARS calculations should be done.  He also suggests angling of the drop shaft
rather than keeping it parallel to the Target Hall area to reduce volume needed
to be dug out.  
c) We probably don't need the removable concrete shielding wall shown in beam
dump picture.
d) He expressed concern about weather proofing of the top of drop hatches should
it be decided to locate the drop hatches outside of buildings.  Clearly he felt
that it would be better to house the drop hatches within a building.
e) We also discussed different philosophies concerning shielding of horns, the
replacement of failed horns, the removal of radioactive horns and shielding.
The discussion was centered on the relative merits of "close packed shielding"
with cranes which is being considered by Fermilab versus the use of a "train"
system which is being considered by CERN.  No final agreement was reached on the
optimum design but at this time the CERN and Fermilab philosophies seems to be
converging on similar designs from different directions.  Clearly more work must
be done in this area by those concerned about these matters both at CERN and at
Fermilab.  Close contact must be maintained by the groups on both sides of the
Atlantic with their counterparts to arrive at a design can do a good job, and
which is  cost effective, reliable and easy to use.  
13)  Cavity movement.
a)  Temperature changes introduced into the rock as a result of
cavern excavations can have serious effects on the dimensions of
the cavern.  The rock adjusts to new temperature conditions caused by the
presence of the experimental hall and shafts by expanding.  These changes in the
cavity size take time to reach some sort of equilibrium after the construction
of the cavity.
b)  Deformation of ALEPH/L3 type experimental hall of up to about 5 mm in radius
in 2 years were seen due to pressures from the surrounding rock.  Approx. 5 mm
deformations also seen in 4.3 m diameter tunnels due to pressures from the
surrounding rock, and moisture and temperature changes in the rock
14)  CERN design to Gran Sasso.
a)  CERN is planning to use conventional magnets for the SPS to LEP extraction
system.
b)  CERN assumes a 3 m diameter decay pipe in the range of 600 to 800 m length.
c)  The dump for the Gran Sasso beam would be about 100 m deep assuming it is
built.
d)  The prospect of installing a near detector hall at CERN is not very good
because of property rights issues.  Such a hall will be in Swiss territory where
construction will be more difficult than in French territory.  One option may be
to employ a location at Geneva Airport for a near detector hall.  This location
is used presently by helicopters, etc.  Since the land for this option is
already owned by a governmental agency, it may be possible to construct the near
detector hall at this location.
e)  CERN has given up on hadron monitors upstream of the beam dump in the
present neutrino area.  Beam monitoring of hadrons is done using low intensity
beam just upstream of the start of their decay pipe by instrumentation which can
be moved into the beam.  The hadron monitoring instrumentation is removed from
the beam for normal running.  Otherwise, the only monitoring of the beam are
muon monitors.

C)	Follow-up activities.
The horn design has many subtle details that need to be directly communicated to
any engineer making modifications to an existing design or beginning a new one.
Just studying the drawings, it may not be obvious why a particular spacer of a
specific size is located where it is.  A subsequent trip by a designing engineer
on our end is recommended.  It is extremely useful to avoid known pitfalls.
Such a visit early in the design would make a valuable connection with people
who have the experience and can point out all the intricacies to make a top
notch design.  Any weak link in the horn system will manifest itself given the
fact that we will be pushing state of the art technology.  Because design flaws
have the potential of costing hundreds of thousands of dollars and delays
measured in tens of months it is essential to incorporate the existing long term
experience of the CERN.  
   On the last working day of our visit,  J. P. Quesnel mentioned that at one
point the existing CERN neutrino beam used a movable beam stop which occupied
about 1/2 of the decay pipe cross section.  He drew a sketch of the beam stop on
his board but he was not an expert on it.  He suggested we contact S. Rangod for
more details.  S. Rangod was not available for more discussions prior to our
departure.   More details on this beam stop will be sought out given that there
exists explicit interest in a similar device by J. Schneps of Tufts University
for use in the NuMI beam.  J. P. Quesnel also mentioned 3 other CERN individuals
that we might contact who had expertise on experimental halls: Jean-Luc Baldy, a
civil engineer concerned with deformations of experimental hall cavities;
Jean-Christophe Gayde, an engineer for experimental halls; and Christian
Lasseur, also an engineer.  Since we learned of these individuals on our last
working day, it was not possible to arrange a meeting while we were at CERN.
Contact with the  latter 3 individuals will be initiated soon by telephone or
e-mail.

APPENDIX

A)	Itinerary:

3/22	Departure from Batavia for CERN.
3/23	Arrival at CERN.
3/24	Meeting with G. Stevenson on radiation shielding and visit to the Neutrino
Target Cave and Service Building for the Neutrino Target Cave.  Meeting with M.
Hatch & R. Gregory on civil construction of deep underground structures followed
by visit to LEP 3.
3/27	Meeting with Alan Ball about neutrino detectors and visit to LEP 8 (the
Aleph Experiment), CHORUS, and NOMAD.
3/28	Meeting with J.M. Maugain, S. Rangod, and V. Falaleev on design and control
of the horns followed by a visit to the hardware.  Meeting with M. Ross and J.
Zazula on high intensity targets.Worked via telnet with Fermilab computers to
answer urgent questions related to work being done at Fermilab.
3/29	Meeting with P. Sievers on target design and lithium lenses.  Follow up
session with J.M. Maugain on horn power system.  Follow up session with
V.Falaleev.  More work via telnet on Fermilab computers to provide answers for
problems atFermilab.
3/30	Follow up meeting with G. Stevenson commenting on our proposed shielding
design.  Revisited NOMAD and CHORUS to learn answers to questions thought up
following initial visit to these experiments.  
4/1	Meeting with J.P. Quesnel about alignment of beam components and placement
of monuments.  Also discussed problems with deformations of temperature changes.
4/3	Return to Fermilab.

B)	List of persons contacted:
Allan Ball (Physicist)	
Graham Stevenson (Health Physicist)
Doug Morrison (Physicist)
Stephane Rangod (Engineer)
Jean-Marie Maugain (Engineer)
Valery Falaleev (Physicist)
Mauray Ross (Target System Engineer)
P. Sievers (Target & Li Lens Designer)
Jan M. Zazula (Target Flux/Heating Simulations/Calculations)
Mark Hatch (Ventilation Engineer)
(Charles) Ray Gregory (Gas Supply Engineer)
Jean-Pierre Quesnel (Optical Survey)

C)	Literature acquired:
A. Ball et al., "Design Studies for a Long Base-Line Neutrino Beam."
G. Acquistapace et al., "The West Area Neutrino Facility for CHORUS and NOMAD
experiments (94-97 operation)."
M. Modena et al., "Construction and Test of a Model of a Pulsed High Gradient,
Final Focus Quadrupole."
H. Butler et al., "Beam-Line Operational Using an Industrial Control System and
Distributed Object-Oriented Hardware Access."
D. Morrison, "Astrophysics Update for Particle Physicists."
S. Ye, "T9 Targets: Total Energy Deposited by a Single 450 GeV Proton on Each
Target"
Horn Assembly Drawing 253.03.548.0
Reflector General Assembly Drawing 253.02.1310
Horn Connections Drawing 253.02.4021
Horn Inner Conductor 1-4 Drawing EF 253 02 375.0
Horn Inner Conductor 2-4 Drawing EF 253 02 378.0
Horn Inner Conductor 3-4 Drawing EF 253 02 384.0
Horn Inner Conductor 4-4 Drawing EF 253 02 386.1
Station helium Ensamble Drawing 8034.9.500.0 B
Blindage Acier + Marbre Ensemble  Drawing 08SPSTNBDA90010
Support Cible Ensemble Drawing 08SPSTNBAC90032
Chariot Aval Ensemble Drawing 08SPSTNBAC90271
Chariot Aval Ensemble Drawing 08SPSTNBAC90261
Boite a Cible Ensemble Drawing 08SPSTNBAC90010
Supportage Boite a Cible Ensemble Drawing 08SPSTNBAC90020A