The annular linear induction electromagnetic pump (ALIP), with advantages such as an axisymmetric structure and high sealing performance, has become a key delivery device in fourth‑generation nuclear power liquid metal‑cooled reactors. However, its relatively low efficiency limits engineering applications, and the end effect is a critical factor leading to performance degradation. To mitigate the end effect and reveal the underlying optimization mechanisms, this study takes an ALIP for actual production as the research object and conducts coupled multi‑physics simulations and structural optimization. First, a three‑dimensional finite element model coupling electrical, thermal, and fluid fields is established in COMSOL. The model is validated through no‑load experiments (with active power deviation ≤5%), confirming its accuracy. Second, the end effect is quantitatively decomposed into a static end effect (evaluated by pulsating magnetic field) and a dynamic end effect (evaluated by eddy‑current magnitude), clarifying its influence mechanisms on magnetic field distortion, flow stability, and energy loss. Based on this analysis, three optimization schemes are proposed: a conical frustum structure at the inner core outlet (ALIP‑cone), a rounded chamfer structure at the inner core outlet (ALIP‑circle), and a gradient arrangement of winding turns at the ends (ALIP‑coil). Simulation results show that all three schemes effectively improve the continuity of the magnetic circuit at the ends and reduce the reverse Lorentz resistance density. ALIP‑coil achieves the best suppression of the static end effect, with a significant reduction in total harmonic distortion (THD). ALIP‑circle exhibits the most notable improvement in head near the rated flow region. After optimization, the pump efficiency shows a marked increase compared to the standard model. By quantifying the impact of the end effect, proposing targeted optimization schemes, and elucidating their working mechanisms, this study provides theoretical support and engineering reference for the further comprehensive and efficient design of ALIPs.
Qiu et al. (Fri,) studied this question.