To take account of the deviations from Keplerian motion that are present in the radial velocity curve of GP Vel, we adopted a somewhat different approach to that employed by Barziv et al. (2001). Instead of trying to average the velocity excursions over many orbital periods, we chose to carry out a comprehensive radial velocity study with maximum phase coverage over several consecutive orbital cycles. In this way we aimed to track the velocity excursions closely and thus model and ultimately remove them.
The observations of GP Vel reported here were made using the 74-inch telescope at Mount Stromlo Observatory, Canberra, Australia. In order to cover two complete orbits, twenty-one continuous nights were obtained between 1996 February 1 and 1996 February 21. Only two nights were completely lost (the 9th and 17th) although most of the first night was spent trying different setup configurations, so this too produced little data. Typically, four sets of three spectra were obtained on each of the remaining nights, with exposures mostly of 1000s. In total 180 spectra of GP Vel were obtained.
The echelle and CCD camera were mounted at the coudé focus. All
observations used the 31.6 lines mm-1 echelle grating, the 158 lines mm-1 cross disperser grating, the 81.3 cm focal length camera and the
thinned 2k 
2k Tektronix CCD, with a slit width of 2 arcsec. Around 70 separate echelle orders fell onto the detector spanning a wavelength range 3370-5970 Å. To shorten the readout time, 
binning was used on the
CCD for the first seven nights, but this was changed to 
binning
(i.e. binning only in the cross-dispersion direction) for the remainder of
the run.
The radial velocity standards HR 1829 (a G5II star) and HR 3694 (a K5III-IV
star) were also observed on most nights, as well as a B0Ia star
(
Ori) to provide a template spectrum for the cross correlation. We
also obtained the usual complement of flat fields and bias frames as well as
thorium-argon arcs each time the telescope was moved.
Copyright ESO 2003