1 Introduction

The astrophysical significance of chlorine-like iron (Fe X) cannot be over-stated. Its lines are observed in many kinds of astronomical phenomena, from novae to cool stars, and so it plays an important role in modelling stellar atmospheres including that of the sun. In fact much interest stems from efforts to determine the precise mechanisms behind the origin of the solar wind using corona diagnostics. This has created much demand for accurate collisional data which this paper aims to address.

Previous collisional work on this ion include several distorted-wave (DW) calculations, including Blaha (1968, 1969), Krueger & Czyzak (1970), Nussbaumer & Osterbrock (1970), Mason (1975), Davis et al. (1976), Malinivsky et al. (1980), Mann (1983) and Bhatia & Doschek (1995). Some of these authors only calculated collision strengths for a few transitions, and in a previous paper in the IRON project series (Pelan & Berrington 1995) we compared new R-matrix calculations for 3s²3p⁵ ²P $_{3/2}^{\rm o}$ -²P $_{1/2}^{\rm o}$ with data from these sources. Mason also tabulated excitation data to the lowest 31 levels (i.e. from 3s²3p⁵ to the 3s3p⁶ and 3s²3p⁴3d levels), and Bhatia and Doschek tabulated collisonal data among 54 levels (i.e. including also the 3s3p⁵3d levels), giving the collision strength for a few medium-to-high energies. However, resonance structure is not normally included in these DW calculations, and we show in the present paper that this can have a significant effect on the calculated rates for excited transitions. Since Bhatia and Doschek give a good review of the earlier data, we confine ourselves in this paper to comparisons with them.

R-matrix calculations include the IRON project work of Pelan & Berrington (1995) on the ground-state fine-structure transition in Cl-like ions. R-matrix methods include resonances and channel coupling effects, and typically more target states are included in the model atom than just the required initial and final states, in order to obtain the effect of resonance structures converging to higher levels. Pelan and Berrington included the lowest 14 LS terms (i.e. all 3s²3p⁵, 3s3p⁶ and 3s²3p⁴3d terms), and used an algebraic transformation to intermediate coupling (Saraph 1978). Mohan et al. (1994) also reported a similar R-matrix calculation, however their calculation appeared to omit some resonance contributions and this was discussed more fully by Pelan and Berrington.

The present calculation is part of an international collaboration known as the IRON Project (Hummer et al. 1993), and extends the R-matrix calculation of Pelan & Berrington (1995) to tabulate data for Fe X from the lowest three levels to the lowest 31 levels. 180 target levels were actually included in the present R-matrix calculation, arising from 3s²3p⁵, 3s3p⁶, 3s²3p⁴3d, 3s3p⁵3d and 3s²3p³3d²configurations, as shown in Fig. 1. A full Breit-Pauli R-matrix (BPRM) treatment was adopted because term mixing among the 3s²3p⁴3d levels was considered too big for the algebraic transformation to be valid. This 180-level BPRM calculation is the biggest so far on this ion: the collision calculation required the setting up and diagonalizing of Hamiltonian matrices of order 6164, with 1104 coupled channels.

$\begin{figure} \par\begin{picture} (170,313) \put(5,0){\framebox (165,313)} \put... ...(5, 308){\line(1,0){80}} \put(5, 308){\line(1,0){80}} \end{picture}\end{figure}$

Figure 1: The model atom: calculated energies (Ryds) of the 180 levels from the 75 terms included for Fe X

Up: Atomic data from the