Interpretive Modeling of plasma evolution during fueling experiments at CMFX

Abstract

The Centrifugal Mirror Fusion Experiment (CMFX) is an axisymmetric magnetic mirror with a central cathode which generates an azimuthal, radially sheared, supersonic \( E \times B \) flow. The induced rotation stabilizes, confines, and heats the plasma. The diagnostic set on CMFX is sparse, giving limited insight to the state of the plasma. In this work, we developed a time-dependent interpretive analysis framework that uses applied voltage, input power, and measured neutron yield rate to infer evolving plasma conditions throughout a discharge. The 0D MCTrans++ code serves as the core physics model, incorporating centrifugal effects, viscous heating, and angular momentum confinement to infer plasma parameters from operating conditions and experimental observables. An iterative Newton's method was implemented to solve for the plasma state evolution consistent with experimental measurements averaged over successive time intervals. The interpretive analysis was applied to experiments comparing different fueling strategies, revealing a path to improved performance via several short puffs of fuel spread across the discharge. This insight led to operations at voltages up 70 kV. Deuterium neutron yields up to \(1.5 \times 10^7\) n/s were measured, and ion temperature was inferred to reach 950 eV. Until CMFX gains a more complete diagnostic set, this interpretive analysis framework provides useful insight into the evolution of centrifugal mirror plasmas.

Figures (3)

• Figure 1: Schematic layout of CMFX showing the major components of the machine and the magnetic field geometry (not to scale). Two large-bore LTS magnets produce the mirror field. The electric field is generated by the central cathode and tungsten-coated ring electrodes, which also serve as plasma limiters. Insulating end plates prevent electrical shorting of the radial electric field along magnetic field lines. Red field lines indicate where the field was measured to validate the calculated field model.
• Figure 2: Measured and inferred evolution of CMFX plasma under different fueling schemes - blue, single 1 ms fuel puff; green, single 0.25 ms fuel puff, purple, double 0.25 ms fuel puff. The red dashed vertical lines indicate when the fuel was injected into the machine. Top two rows show measured applied bias voltage, current, input power and neutron yield rate, while remaining plots show plasma parameters inferred by MCTrans++ interpretive model: ion temperature, electron density, azimuthal velocity, mach number, charge exchange loss rate, neutral density, energy confinement time, and triple product.
• Figure 3: Time traces of measured and inferred plasma parameters at 70 kV. The first two panels show the average traces of applied bias voltage, current draw, input power and neutron yield measured for 30 repeated discharges. The remaining panels show the results of our interpretive physics modeling: ion temperature, electron density, azimuthal velocity, mach number, charge exchange loss rate, neutral density, energy confinement time, and triple product. Red vertical lines indicate when gas is injected into machine.