We present the results of hydrodynamical simulations of the disk photosphere irradiated by strong X-rays produced in the inner most part of the disk. As expected, the irradiation heats the photosphere and drives a thermal wind. To apply our results to the well-studied X-ray transient source GRO J1655-40, we adopted the observed mass of its black hole, and the observed properties of its X-ray radiation. To compare the results with the observations, we also computed transmitted X-ray spectra based on the wind solution. Our main finding is: the density of the fast moving part of the wind is more than one order of magnitude lower than that inferred from the observations. Consequently, the model fails to predict spectra with line absorption as strong and as blueshifted as those observed. However, despite the thermal wind being weak and Compton thin, the ratio between the mass-loss rate and the mass accretion rate is about seven. This high ratio is insensitive to the accretion luminosity, in the limit of lower luminosities. Most of the mass is lost from the disk between 0.07 and 0.2 of the Compton radius. We discovered that beyond this range the wind solution is self-similar. In particular, soon after it leaves the disk, the wind flows at a constant angle with respect to the disk. Overall, the thermal winds generated in our comprehensive simulations do not match the wind spectra observed in GRO J1655-40. This supports the conclusion of Miller et al. and Kallman et al. that the wind in GRO J1655-40, and possibly other X-ray transients, may be driven by magnetic processes. This in turn implies that the disk wind carries even more material than our simulations predict and as such has a very significant impact on the accretion disk structure and dynamics.
Download