Novel Application of Adaptive Fixed Neighbourhood based SPH (AFN-SPH) Method to Reduce Computational Time in Meshfree based Plant Tissue Drying Models

Smoothed Particle Hydrodynamics (SPH) is a popular meshfree method used in fluid dynamics which can be used to model higher deformable boundaries. However, compared to conventional grid-based methods, SPH based meshfree methods fundamentally consume higher computational time mainly due to the Nearest Neighbour Particle Searching (NNPS) algorithm. This becomes an obvious problem when modelling large 2-D and 3-D plant tissues as there is a higher number of interactions between particles. Very few attempts have been recently reported to obtain efficient computational performance when modelling the drying phenomena of plant tissues. Therefore, this research aims to introduce a novel application of the Adaptive Fixed Neighborhood based SPH (AFN-SPH) method, where the fixed neighbourhood with a radius of three times the smoothing length, get updated timely to represent the dynamic changes of the problem domain. It was observed that AFN-SPH is beneficial particularly to simulate plant tissues in extreme drying. Both qualitative and quantitative good agreements were observed in results compared to the conventional SPH techniques. Further, around 30%-40% of time reduction was obtained for different tissues. Therefore, the results indicated that this method can be used as a computationally efficient, meshfree based modelling approach applicable for problem domains with deforming boundaries.