next previous
Up: The density and shock 2818


1 Introduction

The bipolar outflow NGC 2818 appears to possess many of the common characteristics of type I planetary nebulae, including anomalous abundances of He, C, and N, a high excitation spectrum (Dufour 1984), and a complex, filamentary sub-structure (e.g. Johnson 1969; Dufour 1984). It is however unusual among such nebulae in being located within a Population I galactic cluster of the same name; its position, extinction and radial velocity are all consistent with cluster membership (Tifft et al. 1972).

This should, in turn, permit a more precise analysis of the source physical characteristics, since cluster distances can be rather precisely determined (in principle) given a combination of stellar photometry and evolutionary theory. In reality, it appears that estimates for the distance of NGC 2818 have varied widely over the years, from 5 kpc (Shapley 1930), to 1.44 kpc (Barkhatova 1950), 3.2 and 3.5 kpc (respectively Tifft 1987 et al. 1972 and Dufour 1984), and more recently D= 2.3 kpc (Pedreros 1989).

Taking this latter (and more rigorous) value would imply a nebular mass $\approx 
0.22 \Rightarrow 0.48$ $M_\odot$ and unusually large dimensions $\Delta \alpha \times \Delta\delta= 1.44\times 0.67$ pc2, whilst the initial stellar mass was probably $2.2 \pm 0.3$ $M_\odot$ (based on observed main-sequence turn-off: Dufour 1984). This mass is consistent with the range expected for type I progenitors (Peimbert & Torres-Peimbert 1983). Such stars are expected to possess electron-degenerate helium cores immediately following the main-sequence evolution (Iben & Renzini 1983).

Bannerjee et al. (1990) and Meatheringham et al. (1988) find expansion velocities $V_{\rm exp} \cong 20$ to 27 km s-1 (although the estimates of Dufour (1984) appear to be $\sim$ twice as large). Bannerjee et al. also note a considerable complexity in the low-excitation line structure. Whilst Boffi & Stanghellini (1994) determine a filling factor $\varepsilon = 0.22$, both Kohoutek et al. (1986) and Dufour (1984) assign lower values $\varepsilon = 0.03$ to 0.055 on the basis of forbidden line/H$\alpha$ ratios, and inspection of direct nebular imaging. Pedreros (1989) apparently prefers an intermediate value $\varepsilon = 0.121$. Such results are consistent with imaging in suggesting a highly filamentary and/or fragmented shell; a characteristic which may also help to explain variations in the electron densities determined from forbidden lines, which appear to range between $n_{\rm e}$(S$^+)\cong 430$ cm-3 and $n_{\rm e}$(Cl$^{++}) \cong 1300$ cm-3 (Dufour 1984; see also Torres-Peimbert & Peimbert 1977). The mean density deduced from the H$\beta$ intensities is of order 117-140 cm-3 (Dufour 1984; Torres-Peimbert & Peimbert 1977).

Finally, the shell appears also to be associated with strong H2 S(1) $\lambda= 2.1~\mu$m quadrupole emission, suggesting the presence of (possibly UV excited) neutral gas within the principal ionised shell, and shock-excited material immediately adjacent (but external to) the southerly ionised limits to the source (Schild 1995).

It is apparent, therefore, that NGC 2818 is an unusually interesting example of the bipolar outflow phenomenon. We provide, below, the first detailed investigation of the excitation structure of this outflow, together with a fully-sampled map of electron density extending over most of the primary shell.


next previous
Up: The density and shock 2818

Copyright The European Southern Observatory (ESO)