RSVP Simulations and Background Review

1
RSVP Simulations and Background Review 13 January
2005 Answers to the Review Committees Questions
2
1) Table listing all assumptions for running
conditions such as protons on target, hours of
running, etc...)
3
2) Time distributions of neutrons in beam.
t t0 Sqrt(M/2E)-1
Next microbunch starts here
Kaon delays relative to prompt photon times at
Z10m.
Distribution of neutron delays relative to prompt
photon times at Z10m.
4
3) What would be the effect of NOT running with
the 100 MHz cavity? (wider bunch, effect on
background)
Standard KOPIO microbunch configuration is 25 MHz
_at_ 150 kV and 100 MHz _at_ 150 kV, giving 180 ps RMS
width. A single 25 MHz cavity _at_ 150 kV can give
264 ps width. Unoptimized ratio gives 1.5
increase in background. With optimization
largest effect lt10
5
4) RMS variation of microbunch occupancy? What
is the origin of the step in the plot showing
E949 measurement?
Trigger time
Events from 0-50ns are enhanced since this data
was triggered by a coincidence of a beam kaon and
a decay pion, where the decay pion has to occur
within 50ns of the beam kaon. Calculate RMS of
peak Cherenkov rate for the 18 microbunches
recorded by the wave-form digitizers. RMS 4.2
Average the 18 Cherenkov Counter peaks
Cherenkov counter hits
Time since trigger (in ns)
6
5) Singles rates in ALL veto counters - versus
veto threshold - break down by source... -
electronics noise for APDs....
Average counter singles
APDs discussed on next slide
7
5 continued) Singles rates in ALL veto counters
- versus veto threshold - break down by
source... - electronics noise for APDs....

• Operating point
• Gain200,
• Temperature 10 C,
• se 0.3 MeV
• Electronic noise has gaussian
• distribution.

Note se above is 0.3 MeV total energy, i.e. 0.1
MeV visible. Noise contribution above 1 MeV
visible is negligible
8
6) How will you monitor/measure gamma
inefficiency?
KOPIO will monitor both single photon veto
inefficiency and pi0 veto inefficiency. To
monitor the single Photon Veto (PV) inefficiency,
high-statistics KL decays such as KL --gt 3 pi0
(Kp3) and KL--gt 2 pi0 (Kp2) will be used. For
example, in the Kp2 case events with 3 photons
converted in the preradiator (PR) will be
collected. The energy and direction of the
missing photon can be predicted by a fit to the
2pi0 hypothesis performed on the 3 measured
photons. The acceptance for these events is about
2.5. This includes the probability that 3
photons convert in the PR and the kinematic
selection. On average, the production rate of kp2
events in the decay volume is 4 x 10E3 events/s.
The collection rate will be approximately 100
events/s. The energy spectrum of the missing
photon is similar to the energy spectrum of the
missing photons in the Kp2 background. The pi0
veto inefficiency can be monitored by taking
advantage of the kinematically tagged Kp2 sample.
For example, suppose we have 600 events of
background (and 200 events of signal, not
relevant here) at the end of the experiment and
all the background is from Kp2. The kinematic
rejection for Kp2 is about 100 so this means that
we will have a sample of 60000 Kp2 events to work
with after the full pi0 veto has been applied.
These events can be used to monitor the quality
of the full (i.e. order 108) pi0 rejection
through the course of the experiment.
9
7) Fluxes of ALL particle types in beam
(including off collimators), Ks, K, Lambda, n,
KL, gamma.
2.9 x 108 KL/spill _at_ 100TP
(see Question 1) x 100 n flux for P(n) gt
750 MeV/c (from beam studies) x 20-50 g
flux for E(g) gt 10 MeV _at_ target (ditto) x
2 x 10-7 K/- in decay envelope
(from GEANT3) x 10-18 L in decay envelope (from
production/lifetime/stay clear) x 10-8 Ks in
decay envelope (from Ks lifetime/stay clear)
but then Ks ? p0 p0 are vetoed
like KL ? p0 p0
10
8) Provide breakdown of the estimates of KL --gt
K mu nu on a collimator.
11
9) Provide details of veto estimates for stopped
muons.
12
9 continued) and neutron halo.
The overall neutron random rate was calculated as
follows We measured the neutron flux down to 10
MeV using a target very like that of KOPIO.
This spectrum was put into GEANT to determine the
effect of the spoiler and the collimators.
Neutrons were generated from the target into a
solid angle much larger than the aperture of the
beam and followed by GEANT until they stopped or
reached our detector. The 4-momenta and
positions of at the entrance to the detector were
stored. These trajectories were then used to
generate the fluxes at points of interest along
the beamline. The stay-clear distances were set
so that the total rate of neutrons outside them
was not more than about 0.75MHz. So the implicit
assumption was that every one of the neutrons
that impinged on a detector element (except for
the catcher) would veto a coincident event. So
for example, if the veto gate were 10ns, this
would give a random veto fraction of 0.75. This
was somewhat costly in acceptance. Finally the
correction to the above for neutrons below 10 MeV
was determined by a separate GEANT/GCALOR
calculation in which the observed spectrum was
extrapolated downward and everything was followed
to 10-4 eV. This indicated that the total
neutrons were about twice as numerous as what we
were getting from the gt10 MeV result. So our
correction was increased by this factor.
13
10) In the KL --gt pi0 pi0 background estimate,
how do you include gamma fusion? (e.g. 2 gammas
close together)
14
11) What are the accidental and reconstruction
losses inthe preradiator? Dead wire due to
activity/occupancy?
The KOPIO preradiator drift chambers are made up
of small cells with anode wire pitch of 5 mm and
anode-cathode separation 2.5 mm. For the
Ar-Ethane (50-50) gas mixture the maximum drift
time is measured to be approximately 80 ns,
whereas a typical track takes approximately 50 ns
to be collected. (Using CF4 mixtures, which are
also under consideration, the maximum drift time
would be about 35 ns.) Simulations indicate that
maximum rates of 29 kHz/wire occur in the
horizontal wires nearest the neutral beam and the
average rate per wire is about 15 kHz. Dead time
due to pile-up is therefore expected to be lt
0.2.
 
In the LEGS beam test that established the
preradiator proof-of-principle for photon
reconstruction, the event reconstruction
efficiencies were found to be 97 for the cathode
strip planes and 95 for the anode wire planes.
In this test measurement, at least three
consecutive hits were required and the anodes
(cathodes) of all chambers were aligned
vertically (horizontally). The preradiator
GEANT3 simulation uses hits produced in chambers
described above which alternate the directions of
anodes and cathodes. No cuts are applied to
select tracks other than the requirement that 3
to 5 of the initial chambers involved in the
event were hit. The reconstruction efficiency
determined from the simulations was 96,
consistent with those of the measurements cited
above. (This inefficiency has a small overlap
with the inefficiency included in the self-veto
correction).
 Although the effect of missing hits due to
activity or occupancy appears to be minimal, we
plan to study the resolution and efficiency for
cases where one of the initial chamber hits is
missing.
15
12) How is background rejection affected if veto
thresholds have to be raised? 3 MeV? 5 MeV?
(Visible energy is 30 of total)
Table of inefficiencies
Threshold (visible E) Egamma 1 MeV
3 MeV 5 MeV ------ -----
----- ----- 5 0.35
1.0 1.0 15 0.05
0.93 20
0.67 30 28E-4 34E-4
46E-4 50 13E-4 22E-4
41E-4 70 6.2E-4 9.9E-4
16E-4 90 4.1E-4 4.7E-4
5.9E-4 110 2.0E-4 2.3E-4
2.9E-4 130 0.88E-4 1.0E-4
1.3E-4 150 0.58E-4 0.73E-4
0.81E-4 190 0.26E-4 0.26E-4
0.26E-4 250 2.20E-5 300
1.46E-5 400 7.50E-6 500
4.45E-6 600 3.73E-6 700
3.43E-6 800 3.17E-6 900
3.05E-6 1000 2.94E-6
5 MeV 3 MeV 1 MeV
Backgnd ratios for 3MeV/1 Mev 1.6/-0.3 and
5MeV/1MeV 5.9/-1.1 Note 5 MeV threshold is
10x that used by E949
16
13) Provide details of KL flux estimate?
We assume the cross section for KL is the average
of the K and K- cross sections. We used a
parameterization of K? invariant cross sections
on heavy nuclei that was checked against those
measured by AGS E802 for pA ? K?X at 14.6 GeV/c
on Be, Al, Cu and Au at 5, 14, 24, 44, 58
degrees. The parameterization was constructed as
follows 1. We first made a fit to the
invariant cross section of pp production of K?
EK ds(pp ? KX) B(s) (1-xR)n(s)
exp(b-vb2c2(s)p-2) dpK
2. We next obtained the invariant
cross-sections on Be via where IV(A) A and
is a function of the weighted pp and np
inelastic cross-sections 3. The Be results were
multiplied by a factor (APt/ABe)a(xF) to obtain
the K production cross section off Pt. 4.
Overall scaling factors were obtained by
comparing the predicted small angle
cross-sections to agree with data at 23 GeV
17
KL flux (continued)
Then predictions for wide angle K? production at
14.6 GeV/c on Au were compared with E802 data.
Very good agreement was found (see below for an
example). The cross sections could then be
confidently extrapolated to 25.5 GeV/c
Double differential cross section for production
of K by p on Au at 14.6 GeV/c
Measured
Predicted
18
14) Reconstruction losses for cluster
finding?Tracking estimates? (I.e. we have the
impression realistic losses are not included.)
Photon conversion included in reconstruction
efficiency. This loss included in self-veto.
19
Simulation/Software Personnel Availability