Eclipserende dubbelsterren in het tijdperk van TESS

Astronomy is experiencing an unprecedented era of abundance of observational data. The TESS space mission has brought us high-quality brightness measurements of stars over the entire sky, observing large swaths of the sky at the same time over periods of 27 days. The satellite beams down the full camera images, taken at regular time intervals, to astronomers on Earth. This increases enormously the scope of the science output of the mission, beyond the core science and guest observer programs. Several data reduction pipelines are in operation, producing millions of light curves from those images. This has posed the challenge of analysing vast quantities of data. In this thesis, we take advantage of the abundance of processed light curves by exploring the statistical properties of a large sample of eclipsing binary stars for empirical constraints on the theory of tides in close binaries. An eclipsing binary is a system of two stars orbiting around the common centre of mass, for which the Earth is in the same plane as the orbit. This means the stars periodically align with our line of sight, blocking a portion of the light from their companion, resulting in an eclipse. This chance alignment has elevated eclipsing binaries to a position of importance in astrophysics, as it allows for a precise determination of various orbital and stellar properties. In addition to the extrinsic variability in the light curve caused by eclipses, stars can show intrinsic variability. This may come in the form of pulsations, caused by various phenomena that drive the star to oscillate with certain frequencies. Variability can also come from stellar winds or eruptions, and another type of extrinsic variability is caused by star spots on the rotating surface. Stellar pulsations have risen to the foreground over several decades, due to their own link to the stellar properties. This link can be compared to bells of different shapes and sizes that ring with a different sound, which makes them identifiable. Therefore, eclipsing star systems that also have at least one pulsating component provide a unique opportunity for a strong synergy. One of our aims is to collect a large sample of pulsating eclipsing binaries for their astrophysical value. We develop a classification framework for eclipsing binaries, taking into consideration the effect of pulsations. We employ a series of conditions based on the behaviour of time derivatives to identify and isolate different signals while ensuring minimal bias. We further apply machine learning techniques to aid in the rapid and robust identification of eclipses in complex astronomical time series. This methodology has been key in collecting two large new eclipsing binary samples, that include systems with and without intrinsically variable components. The first is a catalogue of 3155 binaries on the hot side of the main sequence, with stellar spectral types O, B and early-A. These stars are of particular interest for their scarcity and uncertain physics relative to stars of lower effective temperatures. About a quarter of targets are found to show significant pulsations to the level of precision of TESS. The second sample in addition includes A- and F-type stars and consists of 69 057 eclipsing binary candidates. The light curves of these candidates are characterised for an investigation of the statistical properties of the sample. To manage the large data quantity and complexity of signals, we develop a methodological framework for the automated analysis of eclipsing binary light curves, which takes into account any additional variability present. The output includes frequencies and amplitudes of sinusoids found in the light curve, that describe the intrinsic variability, if present. This method also provides estimates of fundamental properties of the systems, including the orbital period, the sum of scaled radii and the orbital eccentricity. We obtain a final sample of 14 573 targets across spectral types O, B, A, and F, with parameter estimates. The culmination of this work is to use the properties of this large sample to test the theory of stellar tides. The theory of tides predicts a certain critical orbital period and scaled orbital separation that marks the transition from eccentric orbits at longer periods to circular orbits at shorter periods. Where this transition happens strongly depends on the dissipative mechanism considered in the theory. We take three different measurements of the critical orbital period and critical scaled separation across four effective temperature ranges. These measurements constrain the range of the most to the least circularised binary systems in the sample. Differences in age among a sample cause some smearing of the eccentric-to-circular transition line. However, we find that age alone can not fully account for the observed spread. Selective mechanisms of dissipation, including inertial waves and gravity waves, may explain the observations. Altogether, this thesis provides advances in three areas of eclipsing binary research. The identification of eclipses in vast databases of space-based light curves benefits from a new classification scheme that is capable of finding shallow eclipses among high amplitude pulsations. We contribute to the analysis of eclipsing binary light curves, and light curves of intrinsic variability, with a novel and fully automated pipeline that can aid individual target analyses and enable the investigation of large samples. We deliver empirical constraints on the circularisation of close binaries, from an unprecedented sample of systems observed by TESS. All three contributions have a clear benefit for future studies. Our conclusions on orbital circularisation will help steer development of a better theory of binary tides.