PROPERTIES OF DISCONTINUOUS AND NOVA-AMPLIFIED MASS-TRANSFER IN CATACLYSMIC VARIABLES

We investigate the effects of discontinuous mass loss in recurrent out burst events on the longterm evolution of cataclysmic variables (CVs), Similarly we consider the effects of frictional angular momentum loss (FAML), i.e. interaction of the expanding nova envelope with the seco ndary. The Bondi-Hoyle accretion model is used to parametrize FAML in terms of the expansion velocity nu(exp) of the nova envelope at the lo cation of the secondary; we find that small nu(exp) causes strong FAML . Numerical calculations of CV evolution over a wide range of paramete rs demonstrate the equivalence of a discontinuous sequence of nova cyc les and the corresponding mean evolution (replacing envelope ejection by a continuous wind), even close to the mass-transfer instability. A formal stability analysis of discontinuous mass transfer confirms this , independent of details of the FAML model. FAML is a consequential an gular momentum loss that amplifies the mass-transfer rate driven by sy stemic angular momentum losses such as magnetic braking. We show that for a given nu(exp) and white dwarf mass the amplification increases w ith secondary mass and is significant only close to the largest second ary mass consistent with mass-transfer stability. The amplification fa ctor is independent of the envelope mass ejected during the outburst, whereas the mass-transfer amplitude induced by individual nova outburs ts is proportional to it. In sequences calculated with nova model para meters taken from Prialnik & Kovetz, FAML amplification is negligible, but the outburst amplitude in systems below the period gap with a whi te dwarf mass similar or equal to 0.6 M. is larger than a factor of 10 . The mass-transfer rate in such systems is smaller than 10(-11)M. yr( -1) for similar or equal to 0.5 Myr (similar or equal to 10 per cent o f the nova cycle) after the outburst. This offers an explanation for i ntrinsically unusually faint CVs below the period gap.