MEMORANDUM OF UNDERSTANDING
FOR THE 2008 MESON TEST BEAM PROGRAM
Test beam calibration of the MINERvA detector components
October 16, 2008
TABLE OF CONTENTS
2. PERSONNEL AND INSTITUTIONS ……………………………………………………………………...…...5
3. EXPERIMENTAL AREA, BEAMS AND SCHEDULE CONSIDERATIONS …………………………..….5
4. RESPONSIBILITIES BY INSTITUTION - NON FERMILAB……………………………...........................10
5. RESPONSIBILITIES BY INSTITUTION – FERMILAB……………………………………………………10
6. SUMMARY OF COSTS…………………………………………………………………………………………11
7. SPECIAL CONSIDERATIONS………………………………………………………………………...………13
APPENDIX I - HAZARD IDENTIFICATION CHECKLIST…………………………………………………...15
APPENDIX II – TERTIARY BEAMLINE DETAILS……….…………………………………………………...16
- - 2
This memorandum outlines the plan for beam time at Fermilab during Fall and Winter 2008-2009 to test the MINERvA detector components. This memorandum is intended solely for the purpose of providing a work allocation for Fermi National Accelerator Laboratory and the participating universities. It reflects an arrangement that is currently satisfactory to the parties involved. It is recognized, however, that changing circumstances of the evolving research program may necessitate revisions. The parties agree to negotiate amendments to this memorandum to reflect such revisions.
The experimenters primary need is to calibrate the MINERvA scintillator response (visible energy) to protons, pions, and electrons, to measure the resolution, and to reduce and then estimate the bias on the calorimetric shower energy reconstruction for these particles. This MOU describes the details involved in achieving that, as well as other, secondary goals of the beam test.
The experimenters require an upgrade of the the test beam capabilities to enable a usable rate for particles, especially pions, with momentum as low as 200 MeV/c. Assistance from the Meson Test Beam group and Fermilab (along with some MINERvA personel) is required to complete this upgrade.
The MINERvA experiment will make detailed neutrino nucleus cross section measurements over a range of energies (one to tens of GeV) and target nuclei (He, CH, C, Fe, Pb). This range is not yet explored completely or consistently, yet understanding these interactions is vital for current and upcoming neutrino oscillation experiments such as NOvA. MINERvA will systematically measure the detailed final states of low energy interactions, the calorimetric final states of higher energy interactions, and will be sensitive to how the cross sections evolve with the mass number A of the target nucleus as well as incident energy. In all cases, the reconstruction of pions and protons individually, as well as the total energy in hadronic showers, are vital to the MINERvA analysis goals.
The MINERvA test beam detector is a small version of the full detector. It will be approximately 1.1 meters square and roughly two meters long. The frame that contains the active components will be somewhat larger, approximately 3.5m x 3.5m x 3m in size. The scintillator readout is in the same UXVX orientation sequence as the full MINERvA detector. The frame holding the detector will importantly allow the experimenters to insert and remove lead and iron absorbers equivalent to the electron calorimeter (ECAL) and hadron calorimeter (HCAL) portions of the MINERvA detector. With no absorbers, the test detector will be like the fully-active MINERvA inner tracking detector.
The specific need for lower momenta particles is best demonstrated with the following two plots; they show the expected proton and charged pion spectra for the three primary processes MINERvA will study. The reconstruction of the neutrino interaction kinematics and the identification of the exclusive process both depend on the reconstruction of angle and energy of these hadronic products. The tracking capabilities of MINERvA give excellent angle resolution with little bias. Around 300 MeV/c is where there is a transition from pions that usually range out to pions that usually interact. Analysis of test beam particles will constrain the resulting
- - 3
shower resolution and bias.
Figure 1: Expected distribution of the leading pion (left) and proton (right) for quasi-elastic, resonance, and deep inelastic scattering events. It is the resonance and quasi-elastic hadron spectra that are particularly important for the MINERvA experiment. These estimates for the spectra and number of events for a CH target are for a NOvA-like medium energy beam with 12 20x 10 protons on target.
There are two things that drive the schedule. The earliest the MINERvA test beam detector could be ready for beam is Winter 2008-2009. With no summer shutdown in 2008 the experimenters request an engineering run, probably in January 2009, which requires that the main beam components have already been installed. The experimenters would continue assembly of the detector and commissioning using cosmic rays. This commissioning will follow the same procedure as will be used for the MINERvA tracking prototype which will take place in Summer-Fall 2008. The other schedule driver is the time needed to accomplish the Meson Test Beam upgrade to create a suitable tertiary beam that can produce lower momenta hadrons at moderate rates. This process is expected to be a joint effort between MINERvA and MTest personnel prior to November 2008, with the bulk of the commissioning taking place in October 2008. In addition, by Summer 2009, the components needed for the MINERvA test beam detector must be available for re-use in the full MINERvA detector. Based on very preliminary estimates of the rates, the experimenters anticipate a six week run in the test beam will be adequate. This includes some running in low-momentum tunes of the existing secondary beamline as well as the new tertiary beam.
The MINERvA experiment is funded by the DOE and the NSF. The test beam activities are funded primarily by an NSF-MRI plus the use and re-use (pre-use) of MINERvA detector components funded by the DOE.
- - 4
2. PERSONNEL AND INSTITUTIONS
PI and Group Leader: Richard (Rik) Gran (University of Minnesota
Lead Experimenter in charge of beam test: Rik Gran (Unviersity of Minnesota Duluth)
Deputy Experimenter in charge: Julian Felix (Universidad de
Fermilab liaison: Jorge Morfin (FNAL)
The members of the group which will take part in the beamline design, installation, data
taking activity and dismantling at Fermilab are:
Rik Gran (University of Minnesota Duluth)
Cody Rude (University of Minnesota Duluth)
Emily Draeger (University of Minnesota Duluth)
Julian Felix (University de Guanajuato)
Gerardo Zavala (University de Guanajuato)
Zaidy del Rosario Urrutia (University de Guanajuato)
Aaron Higuera (University de Guanajuato)
Carlos Perez (PUCP, Peru)
Carmen Araujo (PUCP, Peru)
Heidi Shellman (Northwestern University)
Bruno Gobbi (Northwestern University)
Lee Patrick (Northwestern University)
Jorge Morfin* (Fermilab)
Kevin McFarland*,** (University of Rochester)
Dave Pushka (Fermilab)
Jim Kilmer (Fermilab)
Paul Rubinov (Fermilab)
Other MINERvA students, staff, and faculty may participate as shift takers for this experiment,
and as technical experts on the detector subsystems MINERvA will install and operate.
* Co-spokespersons for the MINERvA experiment.
** PI on the NSF-MRI grant that supports the activity in this MOU. Gran is a co-PI.
- - 5
3. EXPERIMENTAL AREA, BEAMS AND SCHEDULE CONSIDERATIONS
3.1.1 The tests will take place in MT6-2A and MT6-2B.
3.1.2 Office space including two desks and network connectivity will be provided for the
duration of T-977, including pre-beam setup and commissioning time. 3.1.3 Space needed in or nearby the beams location, prior to assembly in 2A, to stage the
components of the detector. Ideally this space will be available for several months
prior to running as the different components of the detector become available. 3.1.4 The experimenters will need to be able to access the detector during the run period in
order to insert or remove the absorbers.
3.1.5 The steel absorbers will weigh 600 pounds. This will require an overhead crane or
gantry crane for configuration changes.
3.1.6 The experimenters need the ability to rotate the detector to take data at 25 degrees to
normal incidence. The detector stand will be designed to allow for this. 3.1.7 If the beam profile demands it, or depends on the momentum selection, the
experimenters need to be able to reposition the detector in the optimum location in
3.1.8 This experiment will require the design and construction of a tertiary beamline that
will bend particles to the right of the nominal secondary beam, and the beam
instrumentation and the detector will be placed in this area. The plan for this
tertiary beamline is outlined in Appendix II.
3.1.9 The experimenters will also take some running in the secondary beam in the area
MT6-2B, thus preparations are needed for two separate stand locations.
3.2 BEAM REQUIREMENTS
3.2.1 Secondary Beam: particles between 1 GeV/c and 10 GeV/c. Species tagged to be
pions and protons for the main calorimetry work, but also electrons and muons for
calibration purposes. The HCAL configuration of this detector will stop 1.2 GeV/c
3.2.2 Tertiary Beam: particles between 200 MeV/c and 1000 MeV/c. Species tagged to be
pions and protons for the main calorimetry and kinematic work, but also electrons and
muons for calibration purposes. See Appendix II for more tertiary beam details. 3.2.3 Intensity Needed: 200 to 1000 particles per 4 second spill. 3.2.4 Beam size needed: ~100 cm2 (10 cm diameter or square). 3.2.5 See table for other beam parameters required, followed by detailed description.
parameter Low p High p Units item
Momentum (p) 200 10000 MeV/c 3.2.2 Resolution (showers) NA 5% 3.2.6 ?p/p
Resolution (stopping) 1.0% NA 3.2.7 ?p/p
Bias in momentum 1.0% 1.0% 3.2.8 ?p/p Spot size 10 10 cm 3.2.4
Purity 95% 95% After PID 3.2.10
Rate 200 1000 Per spill 3.2.11
- - 6
3.2.6 For runs where the experiment is collecting hadron showers, especially for middle and high momenta, it is expected that shower fluctuations will be large. The experimenters need a resolution on the incident particle momentum that is significantly less than this, on the order of dp/p = 5%.
3.2.7 For the lowest momentum runs, the experimenters will collect a large number of stopping hadrons. In this case the experimenters will use the data to measure the stopping range. The experimenters would like a much tighter resolution: 1.0%. For stopping muons with 1000 MeV/c momentum, the experimenters expect an intrinsic resolution of 4% in the HCAL configuration of our detector. In fact, it is expected that this resolution goal will be the most difficult to achieve for the lowest momentum beams, especially from the effects of multiple scattering. Alternatively, it may be satisfactory to know the shape of the momentum resolution accurately, or to calibrate and know the particle momenta across the beam spot using upstream tracking chambers in the beamline and MINERvA's intrinsic tracking capabilities, even if the variation is larger than the desired value.
3.2.8 In both cases above, the experimenters require the absolute bias to be small. The target for the resulting uncertainty in the total bias is 2% for MINERvA analysis, so the test beam's intrinsic bias needs to be smaller than this for most of the momentum range, similar to the level of statistical error. The bias will necessarily be larger at the lowest momentum due to multiple scattering in the beamline components. At moderate and higher momenta it will probably be driven by ability to model the decay muon and electron backgrounds for the analysis. The analysis can tolerate a bias close to 1%, at which point it becomes a major contributor to the total error budget. The experimenters would like help in performing an aggressive program of cross-calibration between the secondary and tertiary beam, as well as redundant measurements of the beamline to help manage this uncertainty.
3.2.9 The experimenters are assuming a few cm beam size at 1000 MeV/c. For a larger spot size, especially in cases in the tertiary beam where the momentum spread in the beam or the backgrounds dominate the flux off center, the experimenters may purposely want to use scintillator to trigger on a smaller spot. [This trigger system will be the responsibility of MINERvA.]
3.2.10 Purity. The experimenters aim to obtain a purity of 95% after PID tagging, and enough information to estimate the remaining backgrounds and subtract them. This will require the standard MTEST instrumentation (Cerenkov, TOF, tracking) in the secondary beam to tag and trigger on the resulting particles. At the lowest momenta in the tertiary beam, the backgrounds will be from electrons and also muons coming from pion decay, that the experimenters estimate might intrinsically be as high as 80% of the particles in the beam. To get the purity required, the PID instruments should allow rejection of wrong-species particles with 99% efficiency. Decay muons will probably be the dominant impurity at the lowest momenta; rejecting these may require information from the MINERvA detector itself. Depending on the beam scenario, the experimenters may want to apply the PID tags off-line, but more frequently will want to form a trigger from them. The electronics to form the trigger are the responsibility of the experimenters, though they may use logic modules and space in crates in the MTest control room.
3.2.11 Intensity. The experimenters want ~200 particles per spill at the lowest momenta, up to ~1000 per spill at higher momenta. These values are for the actual species of interest (e.g. Pions), not for the combination of selected species and background, and will be the dominant challenge at the lowest momentum tunes in each beam. There will likely be a
- - 7
practical upper limit to the physical rate in the detector the experiment can handle
because of the MINERvA DAQ electronics, currently expected to be at least 100 Hz if
the experimenters use the currently planned dual/parallel readout system. The
Accelerator Division will make a best-faith effort in achieving the desired rates, and the
experimenters will work with the AD personnel to understand what is ultimately
3.2.12 The experimenters are assuming a 4 second spill duration, arriving every one minute,
for 14 hours per day (840 spills per day, not including shot setup or down time). The
request is actually for integrated spills with the minimum intensity shown, for a duration
of four weeks: 20,000 spills total with at least 200 species-selected particles per four-
second spill. This will give suitable data at the lowest rate, lowest momentum beam
tunings in one 12 hour period. Arrangements to achieve the same integrated triggers at
a slower or faster rate could be satisfactory.
3.2.13 Beam placement. The test beam needs to be steered within a few centimeters of the
center of the MINERvA test beam detector. The detector stand will allow the
experimenters to move detector to the optimal position in the beam.
3.3 EXPERIMENTAL CONDITIONS
3.3.1 The space in MT6-2A will be cleared in order to prepare for the installation of the
detector via the hatch. The space needs to remain clear enough to allow repositioning
the detector at two angles each in the secondary and tertiary beam. 3.3.2 A tertiary beamline is required to achieve particle momenta below 1 GeV/c. The design
and construction of this beamline is a joint effort between MTEST and the MINERvA
3.3.3 The experimenters are evaluating several possible TOF systems, including the standard
units available at MTest, and some alternatives identified by Erik Ramberg. If the
capabilities of the standard 150 ps TOF system are found to be inadequate, the
experimenters will consider what steps to take and amend this MOU as necessary. 3.3.4 High resolution tracking chambers will be needed in the tertiary beamline. The
experimenters are considering the “Fenker” chambers already available at MTEST and
also wire chambers identified from the HyperCP experiment for their use. The
experimenters will work with the MTEST personnel to recondition and test these. 3.3.5 Some care will be needed to keep the temperature in the experimental area from
fluctuating wildly, especially since the plan is to run during the winter. At a minimum,
the facilitity certainly needs to minimize opening exterior doors and avoid opening doors
into the experimental area during running. A temperature readout provided by Mtest
will be needed during data taking periods.
3.3.6 When the 14 hours of beam time is done for the day, the experimenters will usually leave
the system on and keep the electronics “warm” overnight, if the requirements of ORC
have been satisfied. If other activity in MTEST permits, the experimenters will likely
switch to a cosmic trigger and collect muons during the beam off hours. The detector
will have adequate safety protection for the racks and electronics, and the collaboration
will maintain a shift person during these times if required to do so (but prefer not to, if a
shift person is not needed).
3.3.7 The experimenters will provide the computers needed for DAQ, data analysis, and disks
for on-location data storage.
3.3.8 Fermilab will provide 600 Gigabytes of space on backed-up managed disk server that
will be used for long term access to the test beam data. The experimenters require
- - 8
network access to transfer these data, preferably continuously, or at a minimum after
3.4.1 The beam and environmental variables will need to be inserted into the data stream. 3.4.2 The experimenters require suitable electronics to form a PID trigger. These already
exist as part of the MTest facility.
3.4.3 The MINERvA electronics will open a gate, can take a trigger input which will decide
between reading out the contents of that gate or clearing the contents and starting a new
gate. There is roughly 10 microseconds time from the opening of the gate to when the
trigger condition must be present.
3.4.4 The conditions that generate a trigger especially (but possibly also beam quantities)
will need to be synchronized between the external components and the MINERvA
DAQ while reading out hundreds of triggers per spill. Probably this is best done via a
counter/timestamp system latched into each data stream separately. The
experimenters will work with MTEST personnel to provide a suitable interface
between all the components involved.
3.4.5 The experimenters will use the same custom front end electronics and DAQ system
developed for the MINERvA experiment. The front end electronics are mounted on
the PMT boxes in a frame right next to the scintillator, and are read out through VME
and a PC. A low voltage system provides 48 V to Cockroft-Walton bases on the
PMTs, so no exposed HV parts are present. All these components will be provided by
3.5.1 The time schedule is dictated by the availability of an operational tertiary beamline and
the MINERvA test beam detector scintillator planes. As of this writing, the former is
expected to be ready by the end of October, 2008, the latter by January 2009. 3.5.2 A few months will be required to assemble and commission the MINERvA test beam
detector, including the detector engineering run.
3.5.3 A few days setup in the beamline, followed by six weeks running (or more, depending
on the particle rates achieved).
3.5.4 The experimenters prefer the 14 hour per day schedule.
3.5.5 Many configuration changes will require access to the experimental area, though the
run plan will seek to minimize and optimize the frequency of these changes. 3.5.6 Runs will be taken with no absorber, with iron, and with lead, corresponding to the
inner detector, hadron calorimeter, and electromagnetic calorimeter. 3.5.7 Runs will be taken in the secondary beamline and the tertiary beamline. 3.5.8 Runs will be taken at normal incidence and 25 degrees incident. 3.5.9 The nominal plan is to take runs in the tertiary beam will cover the range 200 to 1500
MeV/c, and settings in the secondary beam of 1000, 1200, 1400, 1600, 1800, 2000,
3000, 4000, 5000, and 10000 MeV/c for the HCAL configuration. The tertiary beam
will cover 1000 MeV/c and below, the secondary beam will cover 1000 MeV/c and
above, and both will provide data at 1000 MeV/c. Since the tertiary beam design does
not allow tuning, the plan is to take a wide range of momenta in one or only a few
configurations covering momenta up to 1600 MeV/c. In the secondary beam, the
- - 9
experimenters request beam tunes at 10000, 5000, 4000, 3000, 2000, 1800, 1600, 1400,
1200, and 1000 MeV/c. The experimenters understand the pion content falls
drastically in the secondary beam at these lowest momenta, and the beam may be
essentially electrons only at 1000 MeV/c.
3.5.10 The highest momentum tunes (1600 to 10000 MeV/c) will not be needed for the Pb
ECAL runs or the tracker runs.
3.5.11 Some runs will be taken in a mixed tracker-ECAL-HCAL configuration that most
closely resembles the MINERvA detector.
3.5.12 The inner detector configuration of the test beam is not designed to contain most
particles. The experiment only needs a few runs with through going, and possibly
stopping particles to calibrate dE/dx.
3.5.13 The experiment will certainly need to run the beams with the opposite polarity. 3.5.14 Some special runs with selected stopping and through going muons will be important.
At 1200 MeV/c the HCAL detector will stop muons. The experimenters will need to
obtain data at this momentum tune, and 1000 MeV/c to cross-calibrate the secondary
and tertiary beamlines.
3.5.15 The experiment may be required to split the proposed six week run into two separate
periods, to allow for more flexible scheduling in the Meson Test Beam Facility.
4. RESPONSIBILITIES BY INSTITUTION - NON FERMILAB
The experimenters will take care of and monitor their detector when it is on, including times
when it is operating in cosmic mode while other users have the beam. MINERvA experts are
also responsible for their own DAQ and data handling, and for coordinating with MTEST to get
access to the MTEST data stream. The experimenters will be responsible for coordinating the
effort to calibrate the new tertiary beamline.
The personnel from Northwestern (Bruno Gobbi, Heidi Schellman, Lee Patrick) and a student
from Guanajuato (Aaron Higuera) will participate in the reconditioning of the HyperCP
tracking chambers. Personnel from Guanajuato, Lima, and Duluth, plus Bruno Gobbi from
Northwestern, and possibly others within the MINERvA collaboration, will participate in the
installation and commissioning of the tertiary beamline and the components, especially Carlos
Perez who has already been working on beamline simulations. Carmen Araujo (Lima) and
Zaidy Urrutia (Guanajuato) are working on the Time of Flight system, including the use of the
existing MTest system as well as a new, larger area replacement. This MOU will be amended
at that time. Ph.D. student Zaida Urrutia (Guanajuato) who will be active through the analysis
phase. Senior personnel will be playing many roles and include Bruno Gobbi, Rik Gran, Jorge
Morfin, and Julian Felix. All the above are at or will be at Fermilab extensively during the
Summer and Fall 2008.
5 RESPONSIBILITIES BY INSTITUTION - FERMILAB
5.1 FERMILAB ACCELERATOR DIVISION:
5.1.1 Use of MTest beam as outlined in Section 3.
5.1.2 Maintenance of all existing standard beam line elements (SWICs, loss monitors, etc)
instrumentation, controls, clock distribution, and power supplies.
- - 10