Abstract:
Meshfree methods have become popular over grid-based numerical methods for solving many critical problem domains due to the advantage of using no fixed grids to discretise the simulation domain. Smooth Particle Hydrodynamics (SPH) is such meshfree technique mainly used for modelling fluid dynamics problems. It discretises a given fluid domain using a set of particles, which can move according to the governing equations where the properties of the fluid particles can be calculated using a unique kernel estimation. There, the Nearest Neighbouring Particle Searching (NNPS) process critically influences both the accuracy, and the computational time of the SPH based computation. Compared to conventional techniques used for this purpose, this research proposes a novel approach mainly targeting to reduce the computational time of SPH-based fluid dynamic simulations regarding plant tissues involving low Reynolds number flow conditions. Here, for a given fluid particle in the SPH scheme, particles located within a fixed distance range of three times the SPH smoothing length (h) are treated as neighbours, compared to dynamically computing the neighbours by searching the whole problem domain done in conventional approaches. This novel approach was tested on SPH-based plant tissue models having cell fluids, satisfying above flow conditions, and 33% to 44% computational time saving was observed, without compromising the model consistency for fresh tissues. The proposed approach has a higher potential of application for any meshfree-based fluid dynamic modelling application under above flow conditions, aiming to speedup the simulation process.
