%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Minutes of the Nstar Meeting, April
25, 1998
Submitted by Ralph Minehart Agenda:
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
9:00 Status of replay - progress towards production of DST tapes.
Joe Manak Joe reported on new developments in the cooking of data. Linux
systems
have been moved to single CPU mode to avoid driver problems.
Sun and AIX are being moved from native compilers to gnu s.
He estimated that the Linux systems can analyze 20 M events/day,
the Suns can analyse 20 M/day, and the AIX machines can analyse 4M/day.
The code has undergone massive revisions. The Linux boxes have been
suffering from an
instability due to an ethernet problem, but they should be usable in
about a week.
The Suns have a compiler problem that prevents use of optimization.
The plans for
cooking are to use hit based tracking, to filter the events by requiring
at least one
negative particle, and to write out raw banks and a selected set of
reconstruction banks.
9:15 Status of GSIM -
D. Weygand Dennis reported on the status of the GSIM project that is
being headed by Maurik Holtrop. A list of the projects, completed,
on-going,
or not yet spoken for, is available on the web.
9:30 Acceptance studies with GSIM -
Will Brooks, K. Joo Will reported on the activities of the GSIM focus
group,
consisting of a core group each devoting about 1/3 of their time to
the project
and a group of contributing members. The group is conducting regular
virtual meetings.
They have gotten time-based tracking to work on gsim events. Their
efforts to speed up
gsim by a factor of 10-100 is about half done. Other goals are to improve
the quality
of the simulations by improving event generators, developing code for
background and
out-of-time simulations, etc. They also will coordinate and manage
long-term simulation runs,
that will be carried out at many places off-site. He estimated that
3% statistics calculations
would require 20 dedicated cpu's running for 4.7 months for inelastic
simulations and 48 cpu-days
for elastic simulations. He showed simulation results obtained
with CELEGS at 1.6 GeV and 60% field, and compared them to data. K.
Joo showed some
systematic errors in the recsis reconstruction of the simulated events.
10:00 Other acceptance studies -
Cole Smith K. Joo showed preliminary results on work to determine acceptance
using
radiated AO events as input to gsim and recis. His calculations were
at 1.6 GeV and
58% field. The number of events that made it through gsim and recsis
were rather small,
a problem that requires further investigation. He showed many plots
comparing thrown
data to accepted data. Typically, W distributions showed acceptance
fall-off at large W.
The electron scattering angle distributions fell off more linearly
in angle than the thrown
data which were more expeonential in character. Similar statement could
be made for
the scattered elecron momentum. The 2 phi dependence in the resonance
center of mass
angle was considerably washed out by the gsim-recsis calculation. In
general, this looks
like a good start, but the problem with the gsim-recsis through-put
needs to be fixed.
Cole Smith reported on acceptance studies at 1.6 GeV using CELEGS as
the event generator.
Two dimensional plots of theta vs momentum were similar to results
obtained from Run 7893.
Plots of theta vs phi for the simulated data showed interesting differences
between hit-based
and time-based tracking. One peculiarity is a sharp peak in the center
of the phi distribution
for each sector for negative tracks, using time-based tracking. He
also observed this peak in
the CLAS data. For positive particles a peak rather than a dip, is
observed in the center of
the phi distribution for each sector. Cole believes that the problem
may be due to an enhancement
of electron events near the edge of each sector due to scattering from
the mini-torus.
More studies are needed.
11:00 Particle ID -
Hovanes Egiyan Hovanes presented the results of the timing studies that
made use of the RF to sharpen the start times used in calculating time-of-flight.
Corrections
were made to the zero time for each scintiallator strip. After making
the corrections excellent
alignment of the proton and pion mass peaks was obtained at all three
energies
(1.6, 2.4, and 4.0 GeV).
When all sectors were combined a mass plot shows pion, kaon, proton
and deuteron peaks.
After cutting
on K+e- events, he also obtained a prominent lambda peak and a small
Sigma-0 peak.
11:20 Elastic cross section -
Gagik Gavalian G. Gavalian reported his analysis of elastic
scattering at 2.4 GeV and 4.0 GeV. He cut the elastic peak at W=1.1
GeV and he made fairly
tight phi cuts on the scattered electron in order to have a well-defined
acceptance.
The cross sections were plotted separately for each sector and compared
to world data.
Without radiative corrections, our data typically agree with the world
data to about 10%.
Radiative corrections will be of the order of 15%.
11:40 Inclusive ep cross section -
Stepan Stepanyan Stepan obtained absolute cross sections for inclusive
inelastic
inelastic scattering and compared the results to radiative corrected
calculations
using the Bodek parameterization. The agreement between the CLAS data
and
the parametrization
is excellent at both 2.4 and 4.0 GeV.
1:00 N* data/analysis (overview) -
Volker Burkert Volker reviewed the status of the N* program, using as
his
theme the question of whether we will be able to do the physics we
set out to do.
Using recsis and SEB he included small momentum corrections to the
electron to make
the W peaks independent of the electron scattering angle. He showed
plots of W at 1.6 GeV
for which the elastic width in W was 10.6 MeV(sigma). He then adjusted
the hadron momentum
with a linear function so that the neutron missing mass in pi-plus
production was independent
of the the hadron momentum. In plots of missing mass for ep to epX,
he found both a pi-0 peak and an eta peak. By selecting events with
a detected photon
he enhanced the eta peak. Further enhancement was obtained by selection
of events with
2 or more detected photons.
At 2.445 GeV the width of the elastic peak was 18 MeV.
Studies of epX final states again revealed the pi-zero and eta missing
mass peaks,
and an omega peak as well.
He also showed an rf corrected mass plot of positive particles
which had a strong K+ peak. A missing mass plot of epK events shows
peaks well
above background at the lambda and sigma-0 mass, with mass resolution
of 10-12 MeV (sigma).
At 4.045 GeV, the elastic peak has a width of 33.5 MeV. Peaks are seen
at the Delta(33),
at 1500 and 1700 MeV. Sector 2 had an unusually large width due to
problems with having
to fit to 5 super-layers. Missing mass plots for epX show substantial
production of both
eta and omega, with considerable enhancement of signal to noise when
detection of 1 gamma
is required. In plots of events with 2, 3, or 4 detected gammas, there
is an indication of
eta' production.
He then looked at some physics aspects of the data. At 2.445 GeV, the
width of the second
peak for combined production of p-pi0 and n-pi+ events was smaller
than expected from the PDG.
He studied eta production using as background the events on each side
of the eta peak.
With background subtraction he obtained a peak for N*(1535) in a W
plot which was about
8 times a fairly flat tail extending to higher W. This tail is a question
mark.
Using a background subtraction like the one used for eta production
he looked for
He also studied (e p pi+) final states. The missing mass peaks at the
Delta(33).
There is also evidence for a peak at N*(1520). He concluded that there
is a lot of interesting
1:30 pi-0 production in Delta -
Kyungseon Joo K. Joo reported on a study of the effect of the elastic
tail in the (e'p pi0)
1:50 pi+ production -
Hovanes Egiyan Hovanes reported on analysis of 4M triggers from Feb.
98, 1.6 GeV runs.
2:30 p-eta - Richard Thompson Richard reported on analysis of
2:50 p-omega -
Alan Coleman Alan looked at (e p pi+) final states. After selecting
events with the
3:10 Break
3:30 Large Angle Calorimeter, calibrations, multi-hadron analysis
M. Anghinolfi Marco reported on the data analysis projects at Genova.
3:50 K-lambda, K-sigma -
Reinhard Schumacher, Ken Hicks Ken Hicks reported on work of their group
to study
4:45 Radiative corrections -
Ralph Minehart reported that the AO generator with radiative corrections
is now generating events
5:00 New ideas:
resonance contributions around W from 1800 to 2100 in e-p-omega final
states.
He showed t distributions for 1.7
are that the production is mostly non-diffractive, with a peak around
t=-1.2 GeV**2.
When the missing mass for the ep part of this final state is calculated
there are
noticeable peaks at Delta(1700) and N*(2250).
physics in the existing E1 data, with plenty of statistical significance.
It should\ be possible
to do detailed amplitude analysis, but first we need a lot of simulation
data for acceptance
calculations, and we need good modeling of the physics background when
subtractions are needed.
final states. The missing mass plot shows both a pi0 and a photon peak,
and they are not
well enough resolved to make a clean separation between elastic tail
and pi0 production.
He could select a rather clean set of elastic tail events nad showed
the W plot for them.
The characteristic rise in cross section at large W was apparent. After
using a phi cut to
eliminate the elastic tail he plotted phi-(center of mass) distributions.
The elastic tail
contributes a large narrow peak around phi=0, so it can be mostly cut
out. He compared the data
to simulation. In this case the data look more like the event generator
distribution than like the distribution obtained after subjecting the generaed
events to gsim-recsis.
He found 122K events with an electron and a pi+. He looked at a W bin
of .15 GeV width
around the Delta(1232) esonance and binned the data into 4 Q2 bins.
He showed plots of
theta and phi distributions. These data must be corrected for acceptance
before drawing
physics conclusions.
(ep --> e' p eta) events. He used particle id from time of flight,
accepted only electrons
in the EC fiducial region. Electrons were selected as negative particles
with tracks
matching the EC to within 40 cm, matching the CC to within 8 degrees,
with E/p >0.7,
and with EC(inner) > 168 MeV. Protons were selected as positive partices
with a mass
between 0.8 and 1.2 GeV. Selecting e'p events satisfying these cuts,
he obtained a missing
mass plot with an eta peak at 550.4 +- 0.4 MeV, and a resolution of
18 MeV. A plot of
cos(theta*) is isotropic. The distribution of phi* is symmetric around
180 degree where
it has a deep deep. For the phi angle he chose the y axis perpendicular
to the electron
scattering plane. He selected background events using the missing mass
range from 0.4 to
0.5 GeV. T
he background distribution of cos(theta*) was forward peaked with a
forward to backward
ratio of about 2:1. The phi* distribution was very similar to that
of the eta events.
The eta events appear to be peaked around the S11(1535) mass with a
shoulder on the high W side.
squared missing mass of this three particle system greater than 0.1
GeV**2 and with
W > 1.72 GeV, he plotted the square of the missing mass for the ep
system. This shows
a strong peak at the omega mass. The missing mass plot for the subset
of events that
also include 2 detected photons also shows the omega peak but with
relatively more
background.
The effective mass of the two gammas shows very little enhancement
at the pion mass.
He showed two dimensional plots of the ep missing mass squared vs.
t for different
missing mass and W cuts. This analysis shows that it should be possible
to study
p-omega final states, but more data need to be looked at.
Befcause of problems in older cooked data they re-did the reconstruction
of
a large sample of data using release (1-14) of recsis. They used a
homemade program
to transfer the information from BOS format to HBOOK. These files were
then transferred
to a Sun server in Genova. Many input files were combined into the
analyzed output ntuple.
They sutdied 2' + two hadron events. Missing mass plot of (e'p pi+)
and (e'p pi-) events
showed evidence for missing pion and missing (2 pi) state. Missing
mass spectrum of (e'pi+ pi-)
events has a large peak at the nucleon mass. A plot of the invariant
mass of the (p pi+)
system is dominated by the Delta(1232). The invariant mass of the (pi+
pi-) system at 4GeV
shows some evidence for rho-0 state. The W spectrum for (e'p pi+ pi-)
final states does not
yet show definitive peaks, but there are a number of resonant states
so resolving them will
require more sophisticated analysis. They also have made an attempt
to measure the pi-
acceptance for the (e' p pi+ pi-) events. The statistics so far are
too low for any definitive
statements. The percnetage of 2 pion events in 4 GeV runs is very low
(1/1000 triggers).
They will continue to investigate with additional data, and are thinking
of doing long GSIM runs
in Genova. They have also done studies to look for pi0's in multipion
states by studying 2 gamma events.
They see in the invariant mass of the two photons a peak at the pion
mass.
They have made a cut on this peak and done a number of studies of these
events.
exclusive electroproduction of Kaons. The goal is measure coupling
constants and form
factors and to compare them to calculations. He discussed the use of
W and t dependence
to sort out the constributions of the various N* states. In this regard
the isospin selectivity
of lamda and sigma-0 is useful. Reinhard Schumacher reported on his
results to study hyperon
production with the E1 data. They have looked at the 2.4 GeV data (about
5% of the total).
From this they conclude that they can expect to accumulat 17000 lambda's
and 9,500 sigma-0's.
The hyperon mass resolution is about 35 MeV(FWHM). The Q**2 dependencies
of the lambda and sigma
production are similarly. He showed some detailed distributions of
the data, from which he drew
the following conclusions. The W and t distributions show different
resonance structure for the
lamda and sigma formation. Distributions of the c.m. decay angle show
forward peaking in the
lambda decay and nearly isotropic distribution for the sigma decay.
The dependence on phi may
indicate different LT and TT structure in the two systems. He also
reported evidence that the
lambda is negatively polarized, as in photoproduction, and the sigma
is postively polarized.
that can be used as input to GSIM.