Hello everyone,
I’m using Abaqus/Explicit with CEL (Coupled Eulerian–Lagrangian) to simulate a permeable cylindrical breakwater in a wave flume. The goal is to evaluate hydrodynamic loads and the structural response of the breakwater.

Issue: In some runs I observe strong local free-surface disturbances (jet/spray-like behavior and detached fluid islands) already during the gravity initialization step, i.e. before wave generation starts. Many simulations run fine, but occasionally the solution develops these disturbances early on (and rarely again later). So far I couldn’t identify a consistent root cause.

Question:
What are the most common reasons for such spurious currents / numerical instabilities in CEL (initial volume fraction setup, initial overlap, contact/interface settings, mesh resolution near the structure/free surface, stable time increment, boundary reflections, etc.)? Any recommendations to obtain a calm hydrostatic initial state would be highly appreciated.

Model (brief):

• Abaqus/Explicit, CEL (Euler: water/void or water/air; Lagrange: breakwater)
• Disturbance often appears at the beginning of the gravity step; rarely mid-run
• Attached figure shows the evolution of the issue

Thanks a lot for your help!