Numerical simulation of drifting sand
Two-phase flows are involved in many industrial and natural flow phenomena varying from as specific as the transport of crude oil in pipelines to as general as the dispersion of pollutants in the atmosphere. Numerical modelling based on Computational Fluid Dynamics (CFD), has attracted the attention of scientists and engineers from a wide range of backgrounds over recent decades during which these models have been extensively developed, analysed and applied to many practical applications. Wind blown particles such as sand or snow and their resulting accumulation around buildings, roads, oil field installations and security fences causes severe structural and design problems. These are traditionally addressed based on previous experience, full-scale field investigation or using scale model wind tunnel experiments, all of which incur high cost. In this study, wind blown particles are considered as a two-phase flow system. A finite volume based CFD code is developed using two-phase flow theory and is employed to numerically simulate the drifting of sand and snow around obstacles of different geometry. The model solves the governing transport equations in three dimensional space. Three different approaches are investigated to represent and solve the secondary flow phase, particles, within the flow field; a particle tracking model, based on a Lagrangian reference frame and the homogenous and the mixture models, based on an Eulerian reference frame. The capabilities and limitations of each of these models are investigated for flow fields involving drifting particles around obstacles of different geometry. Particles transported by wind both in suspension and saltation are modelled based on the physical characteristic and the threshold condition of the particle. Their effect on the flow field is incorporated through separate source terms contributing to the particle transport equation. The Eulerian based models are coupled with the Fractional Area/Volume Obstacle Representation (FAVOR) as a mean of representing the solid boundary formed by deposited particles separating the flow field from the accumulation zones. The FAVOR treatment allows the flow field to respond to the changes in the geometry of the deposition regions and further calculations take into account the erosion and deposition processes that have previously occurred. The model can be calibrated to match specific flow conditions through several controlling parameters. These controlling parameters are identified and analysed for four distinct case studies. Model results are compared with field and wind tunnel observations available in the literature and with field measurements conducted as a part of this study in the desert of the State of Kuwait. Qualitatively good agreement between the model and the observations is obtained in two as well as three dimensions. Although the mixture and particle tracking models show the potential capability to simulate such flow systems, the homogenous model is found to be the most appropriate model due to its relative simplicity compared to the mixture model and its lower computational cost compared to the Lagrangian particle-tracking model. In conclusion, a practical CFD tool has been developed and validated, incorporating novel physical and numerical models. The tool can be utilised by scientists and engineers to further understand the real world problem of drifting sand and snow in urban and industrial environments.