This paper aims to evaluate how different numerical parameters affect the accuracy and reliability of hybrid smoothed particle hydrodynamics-finite element method (SPH-FEM) simulations for the blast response of full-scale reinforced concrete beams.
A numerical study was conducted using LS-DYNA, combining the SPH method to model the explosive and the FEM for the reinforced concrete structure. 81 simulations were performed, varying SPH particle count (800,000; 1.6 million; 3.2 million), FEM mesh size (10, 15 and 20 mm) and concrete models (CSCM, RHT and K&C), across scaled distances ranging from 0.4 to 0.8 m/kg1/3. The results were validated against full-scale experimental data.
The study demonstrates that the structural response is non-linearly sensitive to both SPH and FEM parameters. Finer meshes tend to overpredict peak accelerations, as does a reduced number of SPH particles. Contrary to established practice, which advocates the use of a high number of SPH particles, this paper shows that such an approach is not universally valid and that optimal particle density depends on the FEM mesh size employed.
This paper focuses on a specific blast setup and geometry, so findings may have limited generalisability. Further research should cover other scaled distances.
The paper provides valuable guidance for engineers and researchers using SPH-FEM methods in blast analysis, identifying optimal combinations of mesh size and SPH resolution.
This study addresses a gap in the literature regarding the combined effects of SPH and FEM discretisation parameters in blast simulations. Unlike previous works, this research systematically explores their interaction and impact on accuracy and efficiency, providing guidance for future SPH-FEM modelling.
