Harnessing toroidal neutral flows to enhance particle exhaust

Ordered toroidal neutral flows naturally arise in tokamaks through momentum exchange with anisotropic edge plasma flows. Such neutral flows have been experimentally diagnosed in Alcator C-Mod during detachment [1][2] and explained by theory [3][4]. Recent measurements in ASDEX Upgrade—albeit in attached plasmas—support the same physical picture [5], alongside edge plasma modelling [6]. Existing SOLPS-ITER simulations for DTT, ITER and EU-DEMO further confirm the dominant presence of ordered, toroidal neutral velocities in the km s⁻¹ range, implying a substantial directed particle flux.

Despite this body of evidence, toroidal neutral flows have never been deliberately exploited for particle exhaust. Instead, particle exhaust in present-day tokamaks and stellarators is predominantly pressure-driven: neutral particles, undergoing largely random motion in the poloidal plane, are captured only probabilistically by pump inlets. It works—but it relies on randomness.

In contrast, our Direct Simulation Monte Carlo (DSMC) models of neutral kinetics show that intercepting such ordered motion can, in principle, substantially enhance exhaust performance across all neutral species. The strong dependence on flow velocity and density highlights the need for direct experimental assessment.

In this seminar we therefore propose a possible strategy to validate toroidal neutral flows in ASDEX Upgrade—uniquely suited to host a first validation via DIM-II. After discussing our proof-of-principle results, we will outline a prototype concept and associated diagnostics capable of (1) capturing, (2) measuring the toroidal neutral flow, and (3) benchmarking against conventional pressure-driven capture.

This seminar is intended as a starting point: an idea to test a fundamentally different exhaust paradigm—one that leverages order rather than randomness.

[Updated 18.02.:] (For the curious: plotting |vzdena|/mass, |vzdena|/mass/pdena and vzdena / √(vxdena² + vydena² + vzdena²) from SOLPS fort.46—and for molecules, with due care for neutral–neutral collisions and units—already tells an interesting story… Q&A encouraged!)

[1] Pitcher 1999 J. Nucl. Mater. 266–269 1009

[2] Welch 2001 Phys. Plasmas 8 1253

[3] LaBombard 1997 J. Nucl. Mater. 241–243 149

[4] Brunner 2013 J. Nucl. Mater. 438 S1196

[5] Gradic 2018 Plasma Phys. Control. Fusion 60 084007

[6] Rozhansky 2022 Nucl. Mater. Energy 33 101316