Cooling flow regime of a plasma thermal quench

A large class of Laboratory, Space, and Astrophysical plasmas is nearly collisionless. When a localized energy or particle sink, e.g., in the form of a radiative cooling spot, is introduced into such a plasma, it can trigger a plasma thermal quench (TQ). Notice that injecting neon pellets to force a core TQ through radiation is ITER's standard approach of mitigating the thermal load of an incoming thermal quench in the event of a major disruption. Here we show that the electron thermal conduction in such a nearly collisionless plasma follows the convective energy transport scaling in itself or in its spatial gradient, due to the constraint of ambipolar transport. As the result, a robust cooling flow aggregates mass toward the cooling spot and the TQ of surrounding plasma takes the form of four propagating fronts that originate from the radiative cooling spot, along the magnetic field line. The slowest one, which is responsible for deep cooling, is a shock front. For the TQ in a tokamak, these fronts will turn the core TQ into four phases, each of which has its unique physics and duration. Specifically, the nearly collisionless TQ is governed by a fast scaling $T_{e\parallel}\propto t^{-2}$, while the collisional TQ is described by a slow scaling $T_{e\parallel}\propto t^{-2/5}$.