Companion Star in GRO J1655-40

Orosz & Bailyn (1997) determined the spectral type of the companion star to be F3-F6 IV. This was achieved by comparing F-K V-III star (template) spectra to a summed, rest-frame spectrum of GRO J1655-40 from two epochs. The luminosity class is difficult to determine via this method. However, since a main-sequence F star would not be large enough to fill the Roche lobe, and the fact that F dwarf stars did not fit the GRO J1655-40 spectrum well, Orosz & Bailyn concluded that the companion star was most likely a subgiant. Although it was possible, they did not calculate a rotational velocity. Using their eclipsing light curve code to model BVRI photometry of GRO J1655-40, Orosz & Bailyn derived q = 2.99 0.08 (where their q = M₁/M₂) and i = 69.50 0.08^o. The code fits six free parameters and errors reflect those of the internal statistics. Systematic errors were not included as they were believed to be relatively small due to the goodness of fit to the light curves (residuals 0.02 mag or less). From q and i, they inferred the mass of the compact object to be 7.02 0.22 M and the companion star mass 2.34 0.12 M.

Shahbaz et al. (1999) measured the rotational velocity of the companion star in GRO J1655-40 using the same method as Orosz & Bailyn (1997). They found the spectral type F6 (they did not attempt to derive a luminosity class) and = 88.90 6.0 kms^-1. From this, they found q = 0.387 0.05, M₁ = 6.70 1.20 M and M₂ = 2.50 0.80 M.

Israelian et al. (1999) obtained spectra of an F6 III (HR870) and F7 IV (HR6577) star and produced synthetic spectra based on these templates assuming T_eff = 6,400K, = 3.7, [Fe/H] = 0.0 and microturbulence = 2 kms^-1. Broadening both observed and synthetic spectra and comparing the result to the spectrum of GRO J1655-40, they measured = 93 3 kms^-1. They did not derive a black-hole mass from this.

None of these groups have attempted to measure the effective temperature of the companion star, it has always been assumed from the spectral type. Israelian et al. (1999) measured various metallicities. In particular, they found an overabundance by a factor of 6-10 in the alpha elements O, S, Mg, Si, Ti and N. However, [Fe/H] = 0.1 0.2, i.e. almost solar. The overabundance of the alpha elements may be due to the supernova explosion which created the compact object; it is difficult to see how such abundances can be produced in the F star. Orosz & Bailyn (1997) and Shahbaz et al. (1999) did not attempt to derive the metallicity. The rotational velocity has been measured by Shahbaz et al. (1999) using spectra of real stars which span the spectral type and luminosity class of the companion star. Israelian et al. (1999) used two synthetic spectra based on their observed stars HR 6577 and HR 870.

We have derived the rotational velocity, effective temperature and metallicity of the companion star in GRO J1655-40 by fitting Kurucz model spectra to our observed spectrum. We believe the effective temperature should not be a fixed parameter as an overabundance of alpha elements in the companion star may lead to a different effective temperature which would otherwise be assumed for normal F stars.

© Copyright Astronomical Society of Australia 1997