<p>Accurately modelling melting pool dynamics, including surface tension and Marangoni force, is essential for reliable numerical investigations of laser material processing operations. However, conventional Smooth Particle Hydrodynamics (SPH) approaches often struggle to estimate these forces due to the truncated kernel support and unified kernel partition. To address these limitations, a novel geometry-based scheme was proposed for the estimation of surface tension and Marangoni force. Using numerical benchmarks, the initial comparison of the developed scheme against conventional approaches revealed that the proposed framework could be more accurate, particularly in estimating the surface curvature and surface temperature gradient. Following these validations, the proposed framework was applied to model two different laser material processing operations, namely laser micro texturing and laser powder bed fusion (LPBF). In the simulations conducted for laser micro texturing, the Marangoni force and surface tension were found to be the primary factors influencing the generation of textured patterns. Furthermore, the simulated melt pool shape and surface profile demonstrated good agreement with experiments. When implemented for LPBF, the proposed framework was shown to produce numerically stable outcomes and the simulated melt depth also demonstrated reasonably good agreement with experimental data. Based on the obtained results, it is argued that the proposed SPH scheme can offer an accurate numerical framework for modelling melting pool dynamics for a wide range of laser material processing applications.</p>