Unveiling a New Kind of Stable Star System: Machine Learning Provides the Answer
A new type of star system could reveal how they form, according to a study
More than half of the high-mass star reside in multiple systems. Due to the complex orbital interaction, physicists find it difficult to determine how stable and long lasting these systems are. A team of astronomers recently applied machine-learning techniques to simulations of multi-star systems and discovered a new arrangement of stars.
The three-body problem is a well-known problem in classical mechanics. Newton’s laws can calculate the forces between objects and their evolution with ease, but there is no analytic solution for a third mass object. To solve this problem, over the years, physicists have devised various approximation systems to study such systems. They concluded that the vast majority are unstable.
It turns out, however, that the galaxy is full of many multiple-star systems. Over half of the massive stars are part of a binary system, and some of them even belong to quadruple or triple star systems. The systems are obviously very stable. They would have disintegrated long before we could observe them. We are unable to assess how these systems organise themselves or what orbital options they have due to the limitations of our instruments.