Elastic Data Results

1
Elastic Data Results
Jorge Molina
2

• The sample analysed consisted of 31 runs taken
  at
• P1D 17.05 mm (8.98 ?) and P2D 13.80 mm (8.70
  ?)
• In total were used 6 cuts requiring
• Multiplicity 1 One or zero hits in each of 12
  planes (6 x detector)
• Multiplicity 2 One segment per plane
• Segment cut must exist 3 reconstructed segments
  per detector, ie all 6 segments must be nonzero
• Fiducial cut The distance yux-yuv between
  intersections of segments should not exceed in
  ? 2? of the gaussian distribution
• Background cuts events too far from the MC
  simulations are not considered
• Acceptance cut only the events that reaches the
  region allowed by simulation were considered

3
The table shows the efficiency of each cut
Note that most of the events registered were
coming together at the same time, that were
cleaned by the multiplicity cuts.
4
The events that survived all the six cuts have
the ? distribution
?
Events are peaked at zero as expected with a
resolution of ?? 0.01868
5
The correlation of the events that survived all
cuts are
The hitmaps in each detector are
P1D
P2D
6
The spatial resolution in each detector comes
from the difference between the segments
considered by the intersection of the uv segments
minus intersection of ux segments.
?P2D 155 ?m
?P1D 136 ?m
7
The dN/dt distribution of all events
reconstructed is
The fit shows the bins that will be considered
At higher values than 1.3 GeV2 points looks
doubtful At lower values the uncertainty in the
acceptance perturbs we avoid that effect
beginning at the fifth bin
8
The error in the x position were assumed to be
10 of the 3 mm offset found in previous
measurements. With the 0.3 mm relative variation
of the pots, data were reconstructed, corrected
by the unsmearing, and found that variation is
negligible. The error in the vertical position y
were determined changing the position of the pots
in ?0.5 mm in the calculation of the acceptance.
The dN/dt spectra measured then were divided by
these two values of the acceptance obtaining the
errors ? and ?- respectively.
The average error was calculated by ?m (?
?- )/2 Then this value was added in squared
with the statistical error to obtain the
systematic error in the vertical position ?y
(?2est ?2m )1/2
Note that the error bars in each bin of the dN/dt
spectra are determined in this way
9
The dN/dt spectra obtained with the events that
survived all cuts must be corrected by the factor
that takes into account the resolution of the
detector fres f(t)ideal/f(t)measured
To obtain this factor a exponential function were
used as ansatz, that was convolute with a linear
function that describes the variation of the
resolution in t The error bands shows that at
bigger values of t, the error in the
reconstruction is bigger
10
After the correction by fres, the slope of the
exponential function that fits the points went
from 3.631 to 4.015 GeV-2.
The reconstruction program pushes the events to
higher values of t
11
Error in the smearing process comes from the
determination of the parameters of the straight
function t m.t n that describes the change in
the resolution with t. With the uncertainty
?m, the unsmearing process were repeated changing
the parameter m by (m ?m) to obtain the new
value for the slope bm. The difference ?m b
bm was considered the systematic error due to the
parameter m. The same procedure were repeated
for the parameter n to obtain the error ?n b
bn .

• The contribution of the uncertainty in the
  vertical position to the final error were
  calculated in the same way that the uncertainties
  ?m and ?n. The spectra dN/dt measured were
  divided by the acceptance changed in 0.5 mm,
  then unsmeared and obtained the new slope by,
  which difference with b gives the term ?y by
  b that will contribute to the final error.

12
The final error in the slope b is calculated
through
Where each term contributes to the final error in
Finally the slope found is
13
The points corrected by the unsmearing procedure
were normalized by the points obtained by the
E710 experiment (which agrees with CDF for dN/dt
in the determination of the slope b at low t).
The results are in excellent agreement with the
model of M. Bloch showed in the figure
Warning error bars needs the contribution of the
unsmearing errors ? In progress
14
The x fibers (vertical ones) can be a tool to
study the beam behaviour.
Looking at the distributions before and after
multiplicity cuts reveals that in P1D are coming
more particles at the left side (they must be
rotated 180 for PD).
15
For P2D the situation is more critical. The
difference among the last fibers compared with
the rest is much bigger.
The multiplicity cuts cleaned up all the excess
of events. It means that the big majority of
events recorded has two or more particles
crossing the detector at the same time.
Measurements based in the timing revealed
already that particles are being produced in the
separator, may be due to beam-gas scattering or
to the misalignment of the beam that are hitting
somewhere in the middle between P1 and P2
16
Studying the hitmaps in both detectors in slices
of t we have
We suspect since 1.2

• Events in the t range

Total of events survived
17
Starting at t1.2 GeV2 ellipses begins to look
distorted. This is most evident in P2D, although
we can see in P1D the same effect, but more
reduced.

• Events in the t range

Total of events survived
18
1) Proton antiproton elastic scattering was
measured by the D0 Roman Pots. 2) Elastic data
samples contain diffractive and halo
backgrounds, most of which were removed by simple
cuts. 3) The four momentum transfer range
considered in the analysis were limited to the
region 0.96 lt t lt 1.3 GeV2. 4) Data are well
fitted by an exponential form, with the slope
corrected by the unsmearing procedure which used
a exponential as ansatz function. 5) The first
measurement of the dN/dt slope in the region
analysed, at c.m.s. energy of ?s 1.96 TeV gave
the result b 4.015 ? 0.193 GeV-2
19
6) The error in the measurement are mostly
dominated by the uncertainty in the determination
of the resolution in the reconstruction of the
trajectories. 7) The result of the slope found is
in good agreement with the phenomenological model
of M. Bloch based in measurements made by
experiments E710 and CDF at ?s 1.8 TeV. 8) Was
proposed a possible explanation for the behaviour
of the beam, specially in the region t gt 1.3
GeV2. 9) Both detectors, P1D and P2D registered
an excess of events in the outer part (relative
to the centre of the Tevatron), most of which
were cleaned by multiplicity cuts.