Reply to the initial set of comment and questions

1
Reply to the initial set of comment and questions

• Thanks to all reviewers for your attention and
  for your thoughtful and penetrating comments and
  questions. We find them extremely helpful.
• Some of the questions relate to issues already
  studied or discussed and we wish to share our
  thoughts on them, others
• Many of the issues raised, as well as lot of
  others, are addressed in FLARE notes
  http//www-off-axis.fnal.gov/notes/notes.html.
  See in particular materials from the FLARE
  workshop, November 2004.

2
Thermodynamics of the large argon tank

• Axi-symmetric model using ANSYS (Z. Tang. FLARE
  note 31)

Liquid flow
Temperature gradient Note full scale 0.113oC
3
Will argon freeze at the bottom of the tank due
to 5 atm pressure?

• No
• "Thermophysical Properties of fluids. Argon,
  ethylene, parahydrogen, nitrogen, nitrogen
  trifluoride and oxygen", in the Journal of
  Physical and Chemical Reference Data, Volume 11,
  1982, Supplement No. 1.
• R. Schmitt freezing temperature at the
  pressure at the tank bottom is 83.91oK. Actual
  temperature is 87.3oK (FLARE note 31).

4
Argon receiving, quality control, purification
systems

• Critical set of issues, clearly. More work needed
  to have them under full control, clearly. Design
  and specification process has started ( FLARE
  notes 24,26,27,29)
• Experimental effort on proving validity of the
  underlying assumptions (purification power of
  commercial filters, effect of impurities on the
  electron lifetime, composition of impurities,
  out-gassing rates, time dependence) underway (PAB
  setup, Lab 3)

5
Initial purification system

• Design throughput 200 t/day
• Oxygen load _at_1ppm delivered argon purity 200
  g/day. May be more. Probably will be less..
• 24/7 operation for 9 month

6
Main tank28 t/hour re-circulation and
purification system

• Phase I initial purge 100-200 tons of LAr ( 2
  weeks) (vessel not evacuated)
• Very rapid volume exchange (several hours) gt
  rapid purification
• Main issue very large oxygen capacity required
• Milestone achieve gt10 ms lifetime
  before continuing the fill process
• Phase II filling
• Purity level determined by balance of the
  filtering vs. impurities introduced with the new
  argon
• Phase III operation
• Low rate of volume exchange (74 days)
• Removal (mainly) of the impurities introduced
  with argon
• Balance between purification and out-gassing
• In this phase out-gassing of tank walls, cables
  and other materials becomes a visible factor,
  although still very small.
• Tank walls, materials, cables must not
  contain quantities of slowly out-gassing
  contaminants way beyond expectations.

7
Wasnt electron lifetime of ICARUS T600 limited
by cables outgassing ??

• And doesnt this indicate that cables, walls,
  etc.. may be a limiting factor for a very large
  detector ???
• Not necessarily. Probably not. Observe rate of
  lifetime improvement in ICARUS doubles at 40
  days, compared to 20 days (outgassing 1/t)

8
Signal size vs. drift distance vs. purity

• ICARUS signal 15,000 el, S/N6
• FLARE design signal 22,500 el, S/N8. Required
  purity 3x10-11 (oxygen equivalent)
• Significant margin. 2 m drift distance does not
  offer major improvement

Noise level
9
Additional tank for repairs ?

• What if argon in the main tank gets poisoned?
• Install more purification units. Piping must be
  sized to allow for that
• Once the tank is filled with Liquid Argon there
  is no practical possibility of repairs of any
  failed equipment inside. Frequent suggestion
  build another tank to enable transfer of LAr,
  access and repair.
• It is an interesting suggestion requiring
  detailed risk and cost-benefits analysis.
• Design goal minimize the probability of a
  requirement for access
• Minimize the number of components inside the tank
• Robust, failure proof components and construction
  techniques
• In-situ testing to make failures very improbable
• Minimize the impact of an improbable failure(s)
  a nuisance rather than a disaster (example
  broken wire)
• Likely outcome all of the above notwithstanding
  some committee will insist on it. Observation
  spare tank must match the size of the main
  detector tank.

10
Rightsizing of the tank (experiment?)

• ICARUS is building 1200 t detector. A leap to
  50,000 tons is too ambitious.
• One monolithic (sort of) detector is too risky.
  Minimize the risk of unforeseen failures by
  having several smaller detectors
• You have to build prototypes to learn how to
  build such a detector. They must be relevant to
  the ultimate detector construction.
• How does the detector cost scale with size? What
  are the cost drivers? Constant costs vs.
  volume-related.
• A lot of wisdom and practical experience
  speaking..

11
Rightsizing of the experiment

• Technical solutions and construction techniques
  are likely to similar for tanks above 10 kton.
  Linear dimensions scale with cube root of the
  volume (1.7 for 10/50 kton case).
• Most of the site-related, argon receiving and
  purification costs are almost independent of the
  size. We are in process of understanding the
  costs of smaller detectors.
• Scenario I build four tanks (15 kton each), use
  one as a holding tank.
• Scenario II
• Begin with 15 kton tank as a Phase I of an
  off-axis experiment.
• Demonstrate the construction, purification,
  performance. Determine the running conditions on
  the surface and measure potential backgrounds for
  proton decay and supernova detection.
• Depending on the experience, proceed with Phase
  II by building more of 15 kton detectors or jump
  into 50 kton tank
• Reduce the initial risk and provide clear path
  towards the ultimate program of studies of
  neutrino oscillations
• Physics potential of the Phase I is at least
  comparable to all other putative experiments

12
Efficiency/background rejection

• What is it? How is it determined? How sure are
  you?
• Why is it so much better than OOPS (Other Options
  Perceived to be Simpler)?
• Are you planning to scan all events in the
  experiment?
• Can you fish out events out of the ocean of
  cosmic ray-induced stuff?
• When will you have fully automatic reconstruction
  program ?

13
ne Appearance Experiment, A Primer

• At an off-axis position in the nominal NUMI beam,
  if no oscillations
• 100 ev/kton/year of nm CC events
• 30 ev/kton/year of NC events
• 0.5 ev/kton/year of ne CC events
• All of the above for neutrinos with energy 1.5,
  3 GeV
• For CC events the observed energy is that of the
  interacting neutrino (DE/E 10) .
• For NC events the observed energy of only 1/6
  of events falls into the signal region.
  Troublesome sample of NC events is thus 5
  ev/kton/year
• Turn on oscillations sample of nm CC events is
  reduced from 100 to 10. The nt resulting from
  oscillations do not CC interact (below
  threshold). Some of the nm CC events may show up
  as ne CC events - signal.
• Physics potential of an experiment depends on the
  number of identified signal ne CC events.

14
Experimental Challenge

• Maximize Mxe
• Where
• M detector mass
• e efficiency for identification of ne CC events
• While maintaining hgt20/ e (to ensure NC bckg lt
  0.5 ne CC bckg)
• Where h is the rejection factor for NC events
  with observed energy in the signal region
• Why is it hard to achieve high e
• Y-distribution electron energies ranging from 0
  to En
• Low(er) electron energies emitted at large angles
• Why is it hard to achieve high h
• p0s produced in the hadronic shower, early
  conversions and/or overlap with charged hadrons
• Coherent p0 production

15
Tools

• Neutrino event generator NEUGEN3. Derived from
  Soudan 2 event generator. Used by MINOS
  collaboration. Hugh Gallagher (Tufts) is the
  principal author.
• GEANT 3 detector simulation trace resulting
  particles through a homogeneous volume of liquid
  argon. Store energy deposits in thin slices.
• LAIR (Liquid Argon Interactive Reconstruction),
  derived from MAW (Robert Hatcher), derived from
  PAW.
• Project energy depositions onto the wire planes
• Bin the collected charge according to the
  integration time
• Ignore (for now) edge effects, assume signals
  well above the electronics noise
• Assume two track resolution (2 ms)
• Event display (2D, 3 projections)
• Interactive vertex reconstruction
• Interactive track/conversions reconstruction
• 3D event display (J. Kallenbach). Early stages of
  development.
• Prototypes of automatic event classification
  software

16
Early results (MSU, C. Bromberg)

• Algorithm for electron ID
• Charged track originating at the vertex and
  developing into EM shower (at least 3 consecutive
  hits with more that 1.5 MIP of ionization)
• EM shower starting no earlier than 1.5 cm from
  the vertex
• Less than 4 photon conversions in the event
• Fine longitudinal and transverse granularity of
  the detector of critical importance.
• And the answer is
• e 82-6 (41 events out of 50
  events accepted)
• h gt 15 (66 C.L.) (35 events out of 35
  rejected)
• Work in progress. Quite some fun. Come and join,
  room for major contributions

17
(Double?) Blind Scan Analysis at Tufts

• A random collection of signal and backgrounds
  events scanned by undergraduate students trained
  to recognize electron neutrino interactions
  (assign likelihood from 1 to 5)
• Sample of electron candidates) (score gt 3)
  scanned by experts-physicists (still flying
  blind)
• Several examples of events identifiable
  (according to scanners) thanks to superior
  granularity and resolution of the detector

18
It is important to have a good detector

• High-y, low energy (170 MeV) electron easily
  recognized by all scanners
• ? Key to achieving high signal efficiency

19
It is important to have a good detector

• Coherent p0 production
• Easily recognized as a conversion
• A key to keeping background low

20
And the bottom line is

• ne identification efficiency, e76-11 (13 out
  of 17)
• NC rejection factor, h53 (3 out of 159)
• No nm CC backgruond (0 out od 17)
• It was the first try. More scanning underway.
  Improvements expected.
• Automated analysis software under construction.
• WARNING possibly addictive.