In a conventional type Ia supernova, the deflagration flame will transition to a detonation flame as it enters the lower density outer layers of the white dwarf. The detonation flame then consumes the entire white dwarf and the energy release from thermonuclear burning causes the whole star to explode violently as a type Ia supernova. The difference between a deflagration flame and a detonation flame is that the latter propagates faster than the local speed of sound in the white dwarf. As a result, a detonation flame is able to consume the entire white dwarf since the material in front of the detonation flame is unable to “see” the approaching flame front.
However, it is possible to have a failed-detonation scenario where the deflagration flame fails to transition to a detonation flame. This may explain a peculiar subset of type Ia supernovae that are characterised by low ejecta velocities, low luminosities and low ejecta masses. A failed-detonation type Ia supernova occurs when enough mass is burnt during the deflagration phase such that the conditions necessary for the deflagration flame to transition to a detonation flame cannot be achieved and the white dwarf fails to detonate. In this scenario, thermonuclear burning during the deflagration phase delivers energy to the white dwarf, causing the star to expand and then contract. Because too much energy is delivered to the white dwarf, it is unable to attain high enough densities and temperatures to launch a detonation flame during maximum contraction.
For a failed-detonation type Ia supernova, the white dwarf will remain intact as the deflagration is too weak to completely unbind it. However, the white dwarf will now have a lower mass as the failed-detonation event is expected to produce a few tenths of a solar mass of ejecta. The thermonuclear fusion processes occurring within the deflagration flame results in ejecta that is rich in intermediate-mass elements (magnesium, silicon and sulphur) and iron-group elements (iron, cobalt and nickel). A significant proportion of the heavy elements are expected to fall back to the white dwarf and gravitationally settle to form an iron/heavy-core at its centre. The end result is an iron/heavy-core C-O white dwarf.
Due to the highly asymmetric nature of the outburst, the white dwarf will receive a kick velocity of a few 100 km/s. Even so, the large orbital velocities found in most binary star systems suggest that even a kick velocity of a few 100 km/s may be insufficient to unbind the binary. However, for binary systems consisting of a white dwarf accreting matter from an evolved star such as a red giant, the natal kick velocity is likely to unbind the system because of the large binary separation between the white dwarf and the red giant. It is also possible for the natal kick from the asymmetric outburst to launch the white dwarf towards its companion star and this should produce very interesting results.
George Jordan IV, et al., 2012, “Failed-Detonation Supernovae: Sub-Luminous Low-Velocity Ia Supernovae and Their Remnant-Kicked Iron-Core White Dwarfs”, arXiv:1208.5069v1 [astro-ph.HE]