A single star embedded in gas accretes at the Bondi rate — the maximum set by gravity, temperature, and density. The accretion is steady, predictable, and efficient. Add a second star in a tight orbit, and the naive expectation is that two gravitational wells capture more gas than one.
The binary chokes itself.
Orbital motion generates shocks as the two stars sweep through the surrounding medium. In adiabatic gas, these shocks produce entropy — heat that the gas cannot radiate away. The heated gas becomes buoyant, generating convective turbulence that opposes infall. The accretion rate drops to roughly ten percent of the Bondi prediction. The binary's own orbital energy, converted to thermal energy through shocks, creates a thermodynamic barrier to its own feeding.
The transition between efficient and self-limited accretion is governed by a single criterion: the ratio of the gas cooling time to the orbital period, scaled by a geometric factor. When cooling is fast (isothermal gas), shocks dissipate their heat before it can accumulate, and the binary accretes cooperatively near unity efficiency. When cooling is slow (adiabatic gas), entropy piles up, and the binary starves.
The result reframes binary accretion as a waste-heat problem. The binary is an engine that generates thermal pollution as a byproduct of gravitational work. If the environment can flush the waste, the engine runs. If not, the waste product — entropy — becomes a barrier as effective as any physical wall.
Systems that generate heat as a byproduct of their own consumption create a thermodynamic ceiling on further consumption. The transition from efficient to self-limited feeding is not gradual — it is controlled by whether the system can export its own waste faster than it produces it.