MiniBooNE Beam Monte Carlo

1
MiniBooNE Beam Monte Carlo HARP Beryllium
Analysis
Neutrino Beams and Instrumentation 2005
Fermilab, July 11, 2005

• The GEANT4 MiniBooNE Beam Monte Carlo
• Primary Interactions
• Secondary Interactions and particle tracking
• Meson decays to produce neutrinos
• HARP Results and MiniBooNE
• Thin target cross section measurements
• Other measurements to be made at HARP

pion momentum (GeV/c)
p
pion angle (rad)
2
Creating a Neutrino Beam at MiniBooNE

• Neutrino beam created in the conventional
  way...
• Primary proton beam hits a fixed target to create
  secondary mesons
• Mesons focused by a high-current focusing horn
• Focused mesons allowed to decay in open decay
  region
• Neutrinos travel to a distant detector

See the several talks at this workshop by
MiniBooNE collaborators for details on the
various experimental features of the beam
line... ...our focus will be on the simulation
of this experimental beam line
Drawing not to scale
Decay region
1.8 m
25 m
50 m
450 m
3
Creating a Neutrino Beam at MiniBooNE
p-,K-
Relative neutrino fluxes
p,K
Log scale
nm
p
n
nm
n
ne
90 of all n flux
En (GeV)

ne 0.6 of all n's
Dominate flux channel critical to overall
normalization
Important intrinsic electron neutrino backgrounds
to oscillation signal
Non-negligible muon neutrino flux comes from kaon
decays

3 of all n's
3 of all n's
( )
4
Simulating a Neutrino Beam at MiniBooNE

• The MiniBooNE beam Monte Carlo is a GEANT4 based
  simulation
• Goals of the simulation
• Accurately predict the flux of nm, nm, ne, ne at
  the MiniBooNE detector per proton on target and
  unit area as a function of neutrino energy.
• Incorporate methods for estimating the
  uncertainties associated with these flux
  predictions.
• Features of the simulation
• Distinction between primary (8 GeV pBe)
  interactions and secondary (p,n,p,K Be,Al,etc)
  interactions.
• Flexibility to apply different hadronic models
  and hadronic cross sections to facilitate the
  direct input of external data as well as
  performing systematics studies related to the
  flux predictions.

5
Simulating a Neutrino Beam at MiniBooNE
primary pBe interactions

• Simulate optics of primary proton beam
• Parameters are monitored by beam position
  monitors
• Beam line simulation is used to extrapolate
  parameters to target face
• These are used to define features of proton beam
  in the MC
• Beryllium target geometry
• Diameter 1.0 cm
• Consists of 7 identical slugs
• Total length 71 cm 1.7l
• Be fins to facilitate target cooling

6
Simulating a Neutrino Beam at MiniBooNE
primary pBe interactions

• Two important pieces to the hadron rates in the
  simulation
• Total pBe inelastic cross section
• Differential cross section tables for
  each
  secondary important to the
  neutrino flux (p,
  n, p, p-, K, K-, K0)

determines rate of inelastic interactions
determines final state of individual
inelastic interactions
7
Simulating a Neutrino Beam at MiniBooNE
primary pBe interactions

• Two important pieces to the hadron rates in the
  simulation
• Total pBe inelastic cross section
• Differential cross section tables for
  each
  secondary important to the
  neutrino flux (p,
  n, p, p-, K, K-, K0)

determines rate of inelastic interactions
determines final state of individual
inelastic interactions
s(p-p)
pproton 8.9 GeV/c in sloped region of total
inelastic cross section Beryllium target ( Z
4, A 9.0128 ) is in a difficult region for
transparency, or A power law scaling fortunately
there are measurements of total cross sections
available
total
elastic
8
Simulating a Neutrino Beam at MiniBooNE
primary pBe interactions

• Two important pieces to the hadron rates in the
  simulation
• Total pBe inelastic cross section
• Differential cross section tables for
  each
  secondary important to the
  neutrino flux (p,
  n, p, p-, K, K-, K0)

determines rate of inelastic interactions
determines final state of individual
inelastic interactions
One can use built-in GEANT4 hadronic
models Binary Cascade Model, Bertini Model,
etc. Or differential x-section tables can be
generated from various sources MARS, GFLUKA,
etc., or custom Table used to generate
multiplicities and kinematics

...
pT
...
pL
9
Simulating a Neutrino Beam at MiniBooNE
primary pBe interactions

• Can use external data that is most similar (6-20
  GeV/c) to the MiniBooNE beam line configuration
  to develop a custom pBe -gt hadron production
  model
• Parametrize the data somehow to interpolate to
  exact MB experimental configuration ( pbeam
  8.9 GeV/c beryllium target )
• A popular form is the Sanford-Wang
  parametrization. An empirical (as opposed to
  physical) functional form developed to fit pBe
  production data between 10-34 BeV/c (hey, it
  was 1967)

10
Simulating a Neutrino Beam at MiniBooNE
primary pBe interactions

• Much of the available data was taken in the
  1970-80's.
• Low statistics
• Normalization errors often 20
• In some cases papers are missing important
  information
• Relevant data dominated by Brookhaven E910
  experiment (5) 6 GeV/c 12.3 GeV/c beam
• No available data at exactly MB beam energy

p
pT
xF
11
Simulating a Neutrino Beam at MiniBooNE
primary pBe interactions

• Sanford-Wang parameterization fit to external
  hadron data

12.3 GeV/c E910 beryllium Data

• Parametrization
• allows extrapolation from various data sets
  (different pbeam)
• fills in cross section tables beyond where there
  actually exists experimental data
• provides a method for estimating uncertainties on
  neutrino flux due to hadron production (vary
  parameters)
• Parametrization
• result can only be as good as the functional form
  is an accurate description of hadron production
  spectra in the relevant regions particularly
  scaling with pbeam.

12
Simulating a Neutrino Beam at MiniBooNE
primary pBe interactions

• So we try them all...
• Varying results most likely due to differences
  in
• - data used to tune the models
• - phenomenological differences in
  implementation.

p
pion momentum (GeV/c)
pion angle (rad)

• Simulate 8.9 GeV/c protons on the MiniBooNE
  beryllium target and look at pion production
  for 5 different hadronic interaction models
• MARS, GFLUKA, SW, Bertini, Binary
• Propagate through geometry and generate
  neutrino fluxes (from p) at MiniBooNE
  detector

13
Simulating a Neutrino Beam at MiniBooNE
primary pBe interactions

• MiniBooNE has collaborated with the HARP
  experiment (PS-214) at CERN to measure the
  primary pBe cross sections at exactly 8.9 GeV/c
  incident proton momentum
• Relieves the dependence on global hadronic
  model packages or the need to scale cross
  section data to our beam energy or target
  material
• But we will return to this...

14
Simulating a Neutrino Beam at MiniBooNE
secondary p,n,pBe,Al interactions

• Secondary hadrons from pBe interactions
  continue through significant amounts of
  different materials in the beam line
• Beryllium target is 71 cm long
• Horn is made of Aluminum
• Iron
• Concrete

• GEANT4 is used to track particles through the
  geometry of the target hall into the decay pipe

15
Simulating a Neutrino Beam at MiniBooNE
secondary p,n,pBe,Al interactions

• Can switch between different secondary models
  (GHEISHA, Bertini, Binary) and look at effect on
  neutrino flux
• Significantly smaller effect than primary model
• All models use same total total inelastic
  cross section tables

The ridiculous test of turning off p inelastic
interactions completely
16
Simulating a Neutrino Beam at MiniBooNE
Decay of mesons to produces neutrinos

• Beam MC -gt Re-decay program
• Two and three-body meson decays handled
  carefully by separate MC software
• Careful consideration of muon polarizations, etc
• Updated table of branching ratios and lifetimes
  used
• Re-decay program can be used to boost statistics
  for more rare processes

17
MiniBooNE Beam Hadron Production at HARP

• The first goal is to measure p production
  cross sections for Be at pproton 8.9 GeV/c.
• Additional measurements include
• p- production (important for anti-n
  running)
• K production (important for intrinsic n e
  backgrounds)
• Thick target secondary yields

50 l
100 l
18
HARP Beryllium Thin Target Results
Preliminary double differential p production
cross sections from the Be 5 target are available
0.75 lt pp lt 5 GeV/c 30 lt qp lt 210 mrad
Preliminary
qp (mrad)
pp (GeV/c)
Momentum and Angular distribution of pions
decaying to a neutrino that passes through the MB
detector.
19
HARP Beryllium Thin Target Results
MiniBooNE Neutrino Flux

• Use a SW parametrization fit to HARP data alone
  to generate pions in the MB monte carlo.

• A first look indicates that the results of using
  HARP cross sections are similar (within 10) to
  the SW fits used to date (more quantitative study
  needed)
• HARP results should significantly reduce the MB
  flux prediction uncertainty (full error studies
  to come as well)

p
p
Muon neutrinos at the MB detector
20
HARP Beryllium Thick Target Results

• Still need to settle the issue of the secondary
  interactions in the target.
• No need to extract complicated reinteraction
  rates in beryllium

Which (if any!) do we agree with in simple
pions/pot??
50 l
100 l
N.B. At HARP we have an entire matrix of useful
data. Proton, pion beams on Be, Al, other
nuclear targets, at a range of incident momenta
1.5 12 GeV/c
21
Summary Outlook

• GEANT4 MiniBooNE beam Monte Carlo provides a
  very flexible framework for using external data
  and various physics models to best simulate the
  neutrino beam.
• HARP (PS-214) will provide critical input to the
  MB simulation and has already begun to do so
  with a thin target Be cross section measurement
  at pbeam 8.9 GeV/c
• Further input from the HARP data set will
  include thick target yields, p- and kaon cross
  section measurements.