5. Projected rotational velocities
The spectral lines of PG 1605+072 are considerably broadened, which we attribute to stellar rotation and derive v sin i = 39 km/s, by fitting the strongest metal lines. In Fig. 3 we compare a section of the spectrum of PG 1605+072 to that of the pulsating sdB star Feige 48 observed with the same instrumental setup. However, oscillations in these objects should be associated with motions which could lead to line broadening similar to the rotation effect. For a radial pulsation with a sinusoidal velocity curve a radius change of R/R=3% for PG 1605+072 would be required. In comparison Feige 48 is very sharp-lined (v sin i8 km/s) corresponding to R/R0.6% if sin i=1. Although the intensity amplitudes of the pulsations in Feige 48 are considerably smaller than for PG 1605+072 we regard rotation as the more likely reason for the broadening of the lines of the latter. Time resolved spectroscopy should allow to disentangle both effects as well as to search for temperature changes associated with the pulsations.
Assuming a mass of 0.5 the radius of R=0.28 follows from the gravity. Since sin i cannot be constrained the corresponding rotation period of PG 1605+072 must be smaller than 8.7h. PG 1605+072 displays the most complex power spectrum with more than 50 frequencies identifiable (Kilkenny et al. 1999), 39 being bona fide normal pulsation frequencies.
Usually rotation becomes manifest in the power spectrum by the characteristic splitting into equidistantly spaced multiplet components as is observed e.g. for the pre-white dwarf PG 1159-035 (rotation period: 1.4 d, Winget et al. 1991). Such multiplet's, however, have not been identified for PG 1605+072. Fast rotation introduces higher order terms that result in unequally spaced multiplet components. Recently, Kawaler (1999), was able to identify the five main peaks by considering mode trapping and rotational splitting. He predicted that PG 1605+072 should be rapidly rotating (130 km/s). The measured v sin i= 39 km/s, hence, is a nice confirmation of Kawaler's prediction.
Rotation is interesting also from the point of view of stellar evolution. PG 1605+072 is probably already in a post-EHB phase of evolution (Kilkenny et al. 1999) and will evolve directly into a white dwarf, i.e. will shrink from its present radius of 0.28 to about 0.01. Hence PG 1605+072 will end its life as an unusually fast rotating white dwarf if no loss of angular momentum occurs. Isolated white dwarfs, however, are known to be very slow rotators (e.g. Heber et al. 1997).
© European Southern Observatory (ESO) 1999
Online publication: July 26, 1999