Next: The isoscalar 1 mode
Up: Accuracy of solutions
Previous: Accuracy of solutions
The 0 isoscalar mode
In addition to the spurious state associated with nonconservation of particle-number by the HFB, the channel contains the important ``breathing mode''.
In Table 1 we display results from a run with
= for neutrons and
MeV for protons,
resulting in the inclusion of 310 proton quasiparticle states and the same number of
neutron states, with angular momentum up to .
The Table shows the QRPA energies and transition matrix elements of the
particle-number operator. The spurious state is below 200 keV, well separated from the other states, all of which have negligible ``number-strength''. The nonzero number strength in the
spurious state, like the nonzero energy of that state, is a measure of
numerical error.
If the space of two-quasiparticle states is smaller, with
MeV and
, the energy of the spurious state
and the number strength barely change.
Table 1:
The lowest-energy excited
states in Sn.
The second column shows the excitation energies
and the third column the squared matrix elements of the particle-number operator
between the th excited state and the ground state (=0).
|
Figure 2:
Isoscalar strength function
in Sn for (i) the single-proton energy cutoff
=100 MeV and the neutron-quasiparticle occupation
cutoff
(thin solid line); (ii)
=150 MeV and
(dotted line);
and (iii)
=200 MeV and
(thick solid
line). Results corresponding to (ii) and (iii) practically coincide.
|
Figure 2 shows the strength function
for the isoscalar transition operator, cf. [75],
|
(3) |
We have plotted three curves with successively more quasiparticle levels (from 246 proton
levels and 203 neutron levels to 341 proton levels and 374 neutron levels), with
cutoff parameters given in the figure caption. The major structures in the
strength function are stable. The error remaining after
to the gentlest truncation
is extremely small.
The dependence of the strength function on the box size and
quasiparticle cutoff is shown in
Fig. 3. The upper part of the Figure
(panels a-c) corresponds
to a constant smoothing width of =0.5 MeV. This relatively
small value is not sufficient to eliminate the finite-box effects
but it allows us to assess the stability of the QRPA solutions
as a function of .
The large structure corresponding to the giant monopole resonance (GMR)
is independent of box size no matter what the cutoff, but increasing the number of configurations magnifies the dependence on box size of local fluctuations in .
The lower part of the Figure (panels d-f) are smoothed more realistically, as in
Eq. (2). It is gratifying to see
that the resulting strength functions are practically identical,
i.e., the remaining dependence on and the cutoff is very weak.
Figure 3:
Isoscalar strength function
in Sn for the box radii: =20fm (solid line)
and =25fm (dotted line).
In (a), (b), and (c) the smoothing-width parameter is 0.5 MeV for all energies,
while in (d), (e), and (f) is given by Eq. (2). We use the same three sets of cutoff conditions as in
Fig. 2, namely
(i) in parts (a) and (d),
(ii) in parts (b) and (e),
and (iii) in parts (c) and (f).
|
The energy-weighted sum rule (EWSR) for the isoscalar mode [75] is given by
|
(4) |
where the expectation value is evaluated in the HFB ground
state. This sum rule provides a stringent test of self consistency in the QRPA.
In Sn, the right-hand side of Eq. (4) is 35215 MeV fm and the left-hand side
34985 MeV fm for all of the calculations of
Fig. 3;
the QRPA strength essentially exhausts the sum rule.
(The QRPA values of the EWSR in this paper are obtained by summing up to
MeV. )
Next: The isoscalar 1 mode
Up: Accuracy of solutions
Previous: Accuracy of solutions
Jacek Dobaczewski
2004-07-29