The evolution of single stars on and away
from the main sequence is well understood. A
degenerate core is formed in a star as the star
leaves the main sequence and expands to a giant
with a radius typically 50 - 500
Ro . Observationally it is
known that most stars are members of binary
systems, and among these many have orbital periods
less than 100 y. It can happen, therefore, that
the expanding envelope of the primary of a binary
system can reach to the secondary. As this
happens, the primary fills its Roche tidal lobe
and transfers matter to the secondary; if the
primary has a radiative envelope the rate at which
this occurs exceeds the Eddington limit of the
secondary, which therefore repels the incoming
gas, forming a common envelope around the two
stars. Friction within the envelope causes the
stars to spiral towards each other until the
energy and angular momentum extracted from the
binary orbit and transferred to the envelope are
sufficient to eject the common envelope as a
planetary nebula, leaving a short period binary
comprising a white dwarf and a main sequence
star.
This mechanism of producing short period
binaries containing white dwarfs, proposed by
Ostriker and by Paczynski (1976), is the probable
origin of the class of objects known as
Cataclysmic Variable Stars (CVs), which encompass
the classical novae, dwarf novae, novalike
variables and a variety of related objects.
Evidence has been accumulating for forty years
(Crawford & Kraft 1956, Warner 1995a) that
every CV consists of a secondary star (usually a
dwarf, but a few systems contain giants) filling
its Roche lobe and transferring mass to a white
dwarf primary. In systems of normal chemical
composition the orbital periods lie between 75
mins and ~250 d, with the majority having
. A
few hydrogen-free systems are known for which 17
mins < Porb < 50 mins.
It should be noted that CVs are very compact
binary systems: for h
such a binary would fit inside the Sun.