The 9th International Conference on Matter and Radiation at Extremes (ICMRE2026) ( ICMRE2026)

Fully kinetic particle-in-cell (PIC) simulation of Z-pinch plasmas has attracted growing interest as a first-principles approach capable of capturing physics beyond the reach of magnetohydrodynamic (MHD) models, including two-fluid effects, magnetic reconnection, and anomalous resistivity. For many years, however, such simulations remained computationally prohibitive. This landscape is changing with the rapid growth of GPU computing power and the development of advanced solvers, making 2D and 3D simulations with controlled physical reductions increasingly feasible.

In this work, we report progress toward a full-device, fully kinetic Z-pinch simulation framework. Leveraging the multi-solver capability of our in-development code, we present direct comparisons among explicit, implicit, and hybrid PIC solvers within a unified platform applied to identical test cases. This enables a consistent assessment of solver behavior for canonical Z-pinch configurations under the same numerical and physical conditions. Our primary focus is on the explicit solver, which is conventionally regarded as too computationally expensive for practical Z-pinch modeling. We investigate whether controlled approximations — specifically electron mass scaling and a reduced speed of light — can meaningfully extend the usable regime of explicit PIC while preserving the essential dynamics of the pinch. These studies aim to identify the parameter windows in which such reductions remain physically acceptable and to characterize which observables are most sensitive to them.

Finally, we present ongoing developments in full-device modeling, in which the anode and cathode geometry and the external drive circuit are fully coupled to the plasma simulation.
 

PIC,Kinetic,GPU,Z-pinches