Move the pointer over a panel to inspect a sampled value.
What this shows: compression and deflection in a conditional, held-field plasma model. The density front keeps moving; a bright layer or a reduced flux at one instant is not a verified bow shock or steady shielding efficiency. The reference transmitted-flux estimate fails at least one declared numerical-sensitivity check. Treat it as exploratory; consult the report before comparing quantitative shielding performance.
How to read it: x follows the imposed flow, y points opposite lab vertical. The displayed slab is |z| ≤ 1 mm along the coil axis. The dashed line is the x = −7.25 mm readout plane; its numerical readout uses a central 6 × 6 mm footprint. White contours are |mean B| = 1, 2, 4 and 8 T. Velocity is hidden where slab density is below 2% of inflow. An asterisk marks a bulk-flux readout whose mean plane density is also below that threshold. Color scales clip at their displayed endpoints.
Model boundary: the imposed field contains the retained coil, array and feed/return network at the 500 ns FEM state. It is held constant. The plasma-test clock is separate from the Qin-1 discharge clock. No wire ablation, radiation, evolving ionization, absorbing metal surface or plasma-to-circuit feedback is included. The original-device run reached its wall-time limit before the requested 220 ns; all displayed cases have the common 200 ns output shown here.
Paired-feed case at 200 ns: mean density on the fixed readout plane is 1.29% of inflow, below the 2% threshold used for nearly empty cells. Its tiny residual flux is not a qualified shielding efficiency. About 4.89% of ion energy lies in the unqualified region near the hardware, mostly near the feed assembly. These are instantaneous cell-overlap diagnostics; no absorbing or responding conductor boundary was modeled.