Track reconstruction efficiency was estimated using high quality tracks, defined as those that are produced at the primary vertex and generate at least 20 space points in the TPC. The reconstruction efficiency is defined as the fraction of those good quality tracks that are found by the pattern recognition program and that more than 90% of the space points assigned to the track were generated by it.
Figure 1 shows the track reconstruction
efficiency as a function of transverse momentum, 
, for
the five configurations considered with and without vertex
constraint.
As can be expected, the efficiency deteriorates when the
number of pad-rows decreases.
Figure 2 shows the the track reconstruction
efficiency as a function of rapidity, 
, for the
various configurations.
Figure 3 shows the momentum resolution as a function
of transverse 
obtained with the various pad-row configurations
for central Lund AuAu events. The momentum resolution was obtained
by fitting a gaussian to the relative difference
between the real and the measured momentum in the various 
intervals. As can be observed, the resolution obtained
when only pad-rows 21-45 or 30-45 are used without vertex constraint
is rather poor. The rest of the distributions look rather similar,
especially for the full TPC with even and third
pad-rows. Table 7 contains the momentum resolution
integrated over 
.
The momentum resolution as a function of 
in a larger range in
transverse momentum is shown in Figure 4. The results
shown were obtained from a sample of events with tracks generated
with a flat distribution in the 
range 0-40 GeV. The momentum
resolution was obtained from the comparison of reconstructed
and generated momenta. The dispersion of the distributions, rather
that a gaussian fit, was taken as the momentum resolution, since
some distributions are not gaussian.
In this case the results corresponding to pad-rows 21-45 and 30-45
without vertex constraint were not plotted since they are very
poor. The other configurations yield rather similar results.