On the physics of Circinus X-1. II. Eccentricity, mass transfer rates, secular orbital effects, possible γ-ray burst, and a possible origin for circinus X-1

Abstract

Simultaneous radio and X-ray observations of the flare from Circinus X-1 on 1978 February 1-5 are reported and then accounted for within the frame work of the basic physical picture given in Lecture I. The latest data hone the parameters detailing the basic physical processes occurring. The 5 GHz radio data are used to derive an orbital eccentricity, eccentricity, e ≅ 0.72±0.01 for the Cir X-1 binary system.Observational evidence is presented here for both stellar wind and Roche lobe overflow accretion of matter on to the compact star.The steady-state stellar wind mass loss from the OB supergiant is estimated to be some 5 × 10−6 Mg yr−1, 3.5 × 10−8 M0 yr−1 but only lasts ∼2–3 d near periastron.The corresponding mass loss is some 10−10 MD per orbit. For the first 0.5 d of this time, before Roche lobe overflow commences, stellar wind accretion dominates the mass transfer.Temporal ephemerides for periastron in the orbit are estimated using both the radio data and the X-ray data. Very close agreement is found, the ephemerides differing by only 0.05 d. The X-ray cut-off time on 1978, 1st , February agrees well with the estimated time when stellar wind mass transfer first generates enhanced radio emission at 5 GHz. Our model for Cir X-1, used in conjunction with the 5 GHz data, yields a period for the orbit of the compact star round the primary star. The shape of the radio light curve towards the end of the flare gives insight into the physical conditions in the radio emitting region when mass transfer by Roche lobe overflow abruptly ceases. our model for Cir X-1 is extended in this paper to include: (i) effects caused by libration of the orbit of the binary system, which results from the primary star's spin axis being inclined to the orbital angular momentum vector; (ii) variations in the X-ray and radio flare light curves caused by a bulge in the rotating OB supergiant. These effects are discussed in terms of the available observations. We also speculate on the possibility of seeing periodic γ-ray bursts from Cir X-1, and on a possible origin for the Cir X-1 system.