Massive Algols as whetstones for binary star evolution towards GW sources

An accurate understanding of gravitational wave (GW) sources depends on, amongst many others, our knowledge of their progenitor systems. Current predictions of GW progenitor populations suffer from uncertainties due to our lack of understanding of stellar and binary physics, such as mass transfer and internal mixing. To this end, Algol binaries are short-period eclipsing variables undergoing nuclear timescale mass transfer on the main sequence. Hence, these observed binaries form an ideal testbed to constrain our models. We compare predictions from ~100 000 detailed binary evolution models, computed with MESA, to observed populations of massive Algols (> 8 Msun) in the Magellanic Clouds and the Milky Way. We derive constraints on binary mass transfer efficiency, showing that efficient mass transfer is necessary at periods less than ~4 days while inefficient mass transfer may occur at higher periods. At the high mass end (> 25 Msun), we show that many Algols defy the conventional knowledge of binary evolution. During the nuclear timescale mass transfer, many mass donors remain more massive than their companions ('reverse Algols'), and nuclear timescale mass transfer may be interrupted or absent altogether. Due to the elevated luminosity-to-mass ratio from envelope stripping, many core-hydrogen-burning donors may develop Wolf-Rayet-type winds at luminosities where single stars would not. As most massive stars are empirically found to be in binaries, this significantly alters their evolution even during core hydrogen burning. We identified observational counterparts to these reverse Algols in the Large Magellanic Cloud and their contribution to the observed population of hydrogen-rich Wolf-Rayet stars on the main sequence.