3D models of radiatively driven colliding winds in massive O + O star binaries - II. Thermal radio to submillimetre emission

In this paper, the thermal emission over cm to submillimetre (submm) wavelengths from the winds in short-period O + O star binaries is investigated (potential non-thermal emission is presently ignored). The calculations are based on 3D hydrodynamical models which incorporate gravity, the driving of the winds, orbital motion of the stars and radiative cooling of the shocked plasma. The thermal emission arises from the stellar winds and the region where they collide. We investigate the flux and spectrum from a variety of models as a function of orbital phase and orientation of the observer, and compare to the single star case. The emission from the wind-wind collision region (WCR) is strongly dependent on its density and temperature, being optically thick in radiative systems and optically thin in adiabatic systems. The flux from systems where the WCR is highly radiative, as investigated for the first time in this paper, can be over an order of magnitude greater than the combined flux from identically typed single stars. This excess occurs over a broad range of wavelengths from cm to submm. In contrast, when the WCR is largely adiabatic, a significant excess in the thermal flux occurs only below 100GHz.

In circular systems with (near) identical stars, the observed variability in synthetic light curves is typically less than a factor of 2. Eccentric systems may show order of magnitude or greater flux variability, especially if the plasma in the WCR transitions from an adiabatic to a radiative state and vice-versa - in such cases, the flux can display significant hysteresis with stellar separation. We further demonstrate that clumping can affect the variability of radio light curves.

We investigate the spectral index of the emission, and often find indices steeper than +0.6. Synthetic images display a variety of morphologies, with the emission sometimes resembling an intertwined `double-helix'. We conclude by comparing our results to observations. The predictions made in this paper are of interest to future observations with the next generation of radio and submm telescopes, including the EVLA, e-MERLIN, ALMA and the Square Kilometre Array, and future upgrades to the VLBA.