Spectroscopic and physical parameters of Galactic O-type stars - II. Observational constraints on projected rotational and extra broadening velocities as a function of fundamental parameters and stellar evolution

Context. Rotation is of key importance for the evolution of massive star, including their fate as supernovae or gamma-ray bursts. However, the rotational velocities of OB stars are difficult to determine. Aims. Based on our own data for 31 Galactic 0 stars and incorporating similar data for 86 OB supergiants from the literature, we aim at investigating the properties of rotational and extra line-broadening as a function of stellar parameters and at testing model predictions about the evolution of stellar rotation. Methods. Fundamental stellar parameters were determined by means of the code FASTWIND. Projected rotational and extra broadening velocities, v sin i and Theta(RT), originate from a combined Fourier transform and the goodness-of-fit method. Model calculations published previously were used to estimate the initial evolutionary masses, M-evol(init). Results. The sample O stars with MMevolinit greater than or similar to 50 M-circle dot, rotate with less that 26% of their break-up velocity, and they also lack slow rotators (v sin i less than or similar to 50 km s(-1)). For the more massive stars (M-evol(init) >= 35 M-circle dot) on the hotter side of the hi-stability jump, the observed and predicted rotational rates agree quite well; for those on the cooler side of the jump, the measured velocties are systematically higher than the predicted ones. In general, the derived Theta(RT) values decrease toward cooler T-eff, whilst for later evolutionary phases they appear, at the same v sin i, higher for high-mass stars than for low-mass ones. None of the sample stars shows Theta(RT) >= 110 km s(-1). For the majority of the more massive stars, extra broadening either dominates or is in strong competition with rotation. Conclusions. For OB stars of solar metallicity, extra broadening is important and has to be accounted for in the analysis. When appealing at or close to the zero-age main sequence, most of the single and more massive stars rotate slower than previously thought. Model predictions for the evolution of rotation in hot massive stars may need to be updated.