It is a well accepted idea that Planetary Nebulae ( P.N. ) formation is due to mass ejection from red giant envelopes.

According to the original model, as was proposed by Roxburgh and Lucy in 1967, and established about a year later by Paczynski and Ziolkowski (1968), red giant envelopes become dynamically unstable above a certain boundary luminosity, while their total energy, including the ionization energy, which should be available upon expansion, is positive. This energy is sufficient for ejecting the entire envelope with a velocity similar to those observed in P.N.

In the light of these facts, it was quite natural to look for “one shot” ejection mechanisms in unstable red giant envelopes, using full non-adiabatic dynamical codes. This has been done several times since the adiabatic analysis carried out by Paczynski and Ziolkowski (1968), but with minor success, at least in those cases where natural initial conditions were used.

It was found that as soon as the radial expansion of the envelope becomes comparable to its initial radius, the entire radiative zone, which is located above the partial ionization zone, turns to be almost completely transparent. Thus, most of the released recombination energy is radiated directly out of the star, instead of pushing mass shells outwards. Since this energy leakage is very fast, motion is quickly reversed without resulting in any mass loss. Nevertheless, it was pointed out by Smith and Rose (1972) and later by Wood (1974) that there is a possibility for some mass ejection due to shocks generated within the contracting envelope.