People Guide to e-Science Events Organisation News Publications

2010 2009 2008 2007 2006

Simulation on the NGS of binary star systems that make interstellar dust

Monday 21 June 2010

High-resolution near-infrared images of the evolved binary star system WR 140, located about 4,500 light-years away in the constellation Cygnus, suggested that the system is periodically forming and ejecting large arcs of dust. "In the new images, we can see the dust forming in the boundary region between the colliding stellar winds and being blown away at 6 million miles per hour," said Dr John Monnier of the Harvard-Smithsonian Center for Astrophysics in Cambridge, Massachusetts when he presented images of the last dust formation event at an American Astronomical Society meeting in Washington, D.C. in January 2002.

These binary systems consist of two hot, bright, massive stars swinging around each other, one of which is a Wolf-Rayet star -- a luminous star with little hydrogen that will soon explode as a supernova. Both stars have fast (3000 km/s), radiatively driven stellar winds which carry significant kinetic energy. Observations of WR systems with close circular orbits have found a handful where dust forms constantly as the two stars orbit each other. High-resolution near-infrared images of such systems have shown that, when we view these systems face on, we see a pinwheel-shaped nebula of dust surrounding the two stars.

The binary system WR 140, has a highly elliptical orbit. These stars orbit each other every eight years, approaching as closely as 2.5 astronomical units. (An astronomical unit is the average distance between the Earth and Sun.) Dust is formed only when the stars are at or near their closest approach. Only when the WR 140 stars are at their closest do the winds compact enough to allow dust formation. Thus, every eight years at their closest passage, an arc of dust is swept out into space as the energetic Wolf-Rayet wind blows away the newly formed dust.

The problem addressed by Ross  Parkin's study using the NGS was to explain how the spiral-shaped gas distributions must be considered in order to accurately account for observations at X-ray wavelengths since the absorption to X-rays will depend on what angle we see a binary system from. He has designed 3D models to reproduce the spiral-shaped gas distributions which produce the observed dust arcs. A major development in these models has been the inclusion of the driving force from the immense stellar radiation fields which accelerate the stellar envelopes to thousands of miles per second. The hydrodynamics of the binary stars were simulated for two different models of the mechanism driving the outflows. The simulations were then used to produce synthetic X-ray observational data in both cases. These simulated observations were then compared with the actual X-ray observations available to determine which model best produced the observed activity.

The X-ray emission from the super-massive Eta Carinae binary star in the constellation Carina, about 7500 light-years away, were then simulated using the 3D model of the wind-wind collision. The observable emission in the simulation was also compared to available X-ray data. However, there were significant discrepancies between the observed and model light curves and spectra through the X-ray minimum. Models incorporating an additional collapse/disruption of the wind-wind collision region and/or reduced preshock companion wind velocities brought the predicted emission and the observations into much better agreement.

These simulations have been run using the NGS over the last three years, producing several publications in the astronomy research literature. In the last six months alone, the NGS has provided 140,000 hours of computing time for this research, which would have taken over 16 years on a single desktop processor. The NGS is a collaboration of 24 UK universities offering grid access to computing and data resources and is coordinated by STFC e-Science as a national service. The NGS services are available to UK academic researchers to perform simulations and analysis of data in most research disciplines.