Experiments in binary evolution

TL;DR

This white paper argues that constructing volume-complete samples of binary progenitors and their post-mass-transfer (MT) descendants is essential to constrain binary evolution, because the conserved quantity $o= ho/ au$ should hold across interaction stages ($ abla\sum o_{ m prog}=\sum o_{ m post-MT}$). It advocates using Gaia DR4+ and future spectroscopic surveys to derive atmospheric, fundamental, and orbital parameters for thousands of binaries across a broad period range, but acknowledges that time-resolved spectroscopy and high RV precision are required to solve orbits. A core methodological contribution is the diagnostic framework, where $q_{ m crit}$ can be constrained by ordering progenitors by $q$ and matching the inferred $o_{ m post-MT}$ to the low-$q$ tail of $o_{ m prog}$. The paper also proposes a practical survey architecture featuring a real-time scheduler that leverages population priors, a flexible multi-object spectrograph or telescope network, and dynamic exposure control to efficiently follow up target populations over many years. The practical impact is to enable robust tests of MT physics, CE processes, and the overall architecture of binary evolution.

Abstract

The majority of stars more massive than the Sun is found in binary or multiple star systems and many of them will interact during their evolution. Specific interactions, where progenitors and post-mass transfer (MT) systems are clearly linked, can provide yet missing observational constraints. Volume-complete samples of progenitor and post-MT systems are well suited to study those processes. To compile them, we need to determine the parameters of thousands of binary systems with periods spanning several orders of magnitude. The bottleneck are the orbital parameters, because accurate determinations require a good coverage of the orbital phases. The next generation of time-resolved spectroscopic surveys should be optimized to follow-up and solve whole populations of binary systems in an efficient way. To achieve this, a scheduler predicting the best times of the next observation for any given target in real time should be combined with a flexibly schedulable multi-object spectrograph or ideally a network of independent telescopes.