SIMBAD references

In a previous paper a model was presented to calculate the thermal emission of molecules around a central star. The model includes a self-consistent determination of the gas kinetic temperature, photoelectric heating, cooling by water molecules and the constraint that the presence of dust puts on the molecular excitation. The model is applied to the CO(1-0) and CO(2-1) observations of the OH/IR stars OH 32.8-0.3 and OH 44.8-2.3 (abbreviated to OH 32.8 and OH 44.8). Both come from the sample observed by Heske et al. (1990) who noted that in the less extreme OH/IR stars (like OH 44.8) the mass loss rate derived from infrared properties agrees reasonably well with that estimated from the CO emission but that in extreme OH/IR stars (like OH 32.8) the mass loss rate derived from the infrared is an order of magnitude larger than that derived from CO emission. For a dust opacity at 60µm of 228 cm²g^–1 the best model for OH 44.8 has the following parameters: {dot}(M)=9.0x10^–6M_☉/yr, dust-to-gas ratio {PSI}=0.0035 and mean dust grain size a=0.14µm. The derived mass loss rate is insensitive to the adopted opacity. The results are relatively insensitive to any model assumptions. For OH 32.8 no model is found that fits the observed line profiles for a constant mass loss rate throughout the envelope. For a grain size of a=0.125µm, an opacity of 228 cm²/g (following the result for OH 44.8) and a mass loss history in which the mass loss rate drops by a factor of 10 for radial distances larger than a critical distance R_c, the following model reproduces the observed intensities: (present-day) {dot}(M)=2.0x10^–5M_☉/yr, {PSI}=0.015 with R_c∼1.3x10¹⁷cm (corresponding to a timescale of about 2800 years). Models with {dot}(M)>4.0x10^–5M_☉/yr cannot be made to fit the observations, models with {dot}(M)<2.0x10^–5M_☉/yr probably can, but result in higher dust-to-gas ratios ({PSI}~{dot}(M)^–1). The distinction made by Heske et al. (1990) between moderate OH/IR stars (like OH 44.8) and extreme OH/IR stars (like OH 32.8) can be understood as follows: the CO shell in the extreme OH/IR stars is so large that the outer part samples a previous phase of lower mass loss, several 10³ yrs ago. Finally, I comment on the possibility that in extreme mass losing stars the temperature in the outer parts of the circumstellar shells drops below the cosmic background radiation temperature. Based on the models for the two OH/IR stars I derive that this occurs if {dot}(M)_–5>4.8Q_0.01L₄^4/3v₁₀^1/3, where Q_0.01 is the effective absorption coefficient in units of 0.01, {dot}(M)_–5 is the mass loss rate in 10^–5 M_☉/yr, L₄ the stellar luminosity in 10⁴ L_☉ and v₁₀ the expansion velocity of the shell in 10 km/s. This relation is expected to be valid for oxygen-rich stars and standard values for the dust opacity and the photoelectric heating rate.