In this section we present relativistic BP results for the maxwellian averaged collision strength, , with radiation damping included. The effects of relativity and radiation damping on the maxwellian averaged collision strengths are examined.

The maxwellian averaged collision strength, or the
effective collision strength, is given by:

where is the collision strength for excitation from level
*i* to level *j*, averaged over a maxwellian distribution of outgoing
electron energies above the excitation threshold
of the level *j*, at temperature *T*. This slowly varying function
of temperature can then be used to obtain the rate coefficient,
, for electron impact excitation,

where is the energy difference between levels *i* and *j*
and is the
statistical weight of level *i*.

**Table 1:** Labeling of energy levels included in the calculation and the
comparison of the observed and calculated energies (in rydbergs)
in Fe XXII

**Table 2:** Comparison of the effective collision strengths obtained
from
the relativistic Breit Pauli R-matrix method with the 45 level target
expansion (45 BP) and with the 15 level expansion (15 BP), and
by the term coupling method with the 8 term (15 level) target expansion
(TCC) for Fe XXII. All three sets of the results
include radiation damping effect

The values for the 105 transitions between
the 15 *n*=2 levels (see Table 1 (click here)), the 450 transitions from these
15 *n*=2 levels to the 30 *n*=3 levels and the 435 transitions
between these 30 *n*=3 levels are tabulated for the same 32 *z*-scaled
temperatures as those for the B-like ions in Paper III.
We also calculated a set of results for the 105 transitions between
the 15 *n*=2 levels using a 15-level BP target expansion (15BP).
In order to show the physical effects mentioned above,
we compare in Table 2 (click here) the results for 2 temperatures,
*T* = 441 000 and 2 205 000 K, obtained with the
45-level target expansion (45BP), the 15BP, and the
TCC calculations with the 8-term target (that corresponds to the same 15
levels as in 15BP)
for transitions from the ground level
to the fourteen excited *n*=2 levels.
The two sets of BP results for excitation to
the low-lying levels are larger than
those with the TCC (those for the lower temperature
enhanced more), indicating the effect of the more
extensive resonances due to relativistic level splitting in the BP
calculations.
For the intercombination transitions to the levels
(indices 3, 4 and 5 in the tables), the results with the BP
are considerably larger than those with the TCC, obviously due
to the relativistic intermediate coupling.
For transitions to the high-lying *n*=2 levels, namely the levels
(indices 11 to 15 in the tables),
the values for 45BP are generally higher
than the 15BP, especially for the higher temperature
*T* = 2 205 000 K, indicating the effect of resonances due to
the *n*=3 levels.

**Figure 1:** Comparison of the present effective collision strengths
by the relativistic Breit-Pauli R-matrix method with the 45-level target
(the solid lines) and the 15-level target (the dash-dotted lines),
by the term-coupling method (the dashed lines) and by
the relativistic distorted-wave method (the dotted lines) for
the transitions from the ground level to
**a)** , **b)** ,
**c)** and **d)**

To further illustrate these effects,
in Fig. 1 (click here) we compare the values from the 45BP, 15BP, TCC and RDW
(Zhang & Sampson 1994a) calculations for transitions from
the ground level to the , and the three
levels.
It is seen that the values do not differ very much between the 15BP,
the TCC and the RDW for the higher
temperatures. The 45BP results are mostly higher than the three other sets
at these temperatures, indicating the effect of the *n*=3 resonances.
For lower temperatures, the two sets of BP results
are the same. However, they differ significantly
from the TCC and the RDW results, especially for the intercombination
transitions, showing that the relativistic effects affect
the low-energy resonances. Obviously the RDW results are too low due to
absence of
resonance contributions. The TCC results are also low
as explained by Fig. 1 (click here) of Zhang & Pradhan (1995a). From
that figure
we see that the TCC method takes relativistic effects into account only
when the electron energies are greater than the target thresholds
coupled to the upper level of the transition through intermediate
mixing, while the BP method, in contrast,
includes the relativistic effects ab initio for
the entire energy region. Here
it is again shown that the TCC method is not an adequate approximation
for the relativistic effects in highly charged ions.

To study the radiation damping effect on the rate coefficients, we have also calculated the values with this effect turned off. By examining the two sets of results, it is found that, although radiation damping affects collision strengths near a few target thresholds for some transitions, especially for the forbidden and the intercombination transitions, the effect on the rate coefficients is minimal for most transitions, only reducing the results by 1% or 2%, except for three transitions, , and . For these three transitions, not only is the radiation damping effect on the collision strengths large, but its contribution to the rate coefficient is enhanced since the scattered electron energies, which enter the exponential term in Eq. (1), are small near the next target threshold where radiation damping is most pronounced. In Fig. 2 (click here) comparisons of the BP values with and without radiation damping are made for four transitions, including the above three transitions and the transition . For the former three transitions the radiation damping effect reduces the values by about 8 - 20% at low temperatures, but for the latter transition the reduction is just about 2% (although its effect on the collision strength can be seen in Fig. 2 (click here) of Zhang & Pradhan 1995a).

The values for the 105
transitions, with *n*=2, are tabulated in Table 3 at 32 *z*-scaled temperatures
(*T* in K, so that actual *T* values are obtained by multiplying
by = 441). Results for the 450 *n*=2-3 transitions
and for the 435
transitions with *n*=3 are tabulated in Table 4 and Table 5, respectively.
Tables 3, 4 and 5 are not reproduced here, but
are available in electronic form
at the CDS, or via ftp from the authors at zhang@payne.mps.ohio-state.edu.

**Figure 2:** Radiation damping effects on the effective collision strengths
for the transitions **a)** ,
**b)** ,
**c)** and
**d)** .
The solid lines indicate the results with radiation damping and
the dashed line those without radiation damping

web@ed-phys.fr