Surrogate-based aerodynamic optimization for enhancing the shelter effect of porous fences on a triangular prism

Abstract

To further improve the shelter effect of porous fences on the wind erosion of a triangular prism pile of coal, an aerodynamic optimization procedure has been developed, by coupling computational fluid dynamics (CFD) solver, kriging model and genetic algorithm. Two-dimensional Reynolds- Averaged Navier-Stokes equations with the RNG k-ε turbulence model were used for numerical simulation of the shelter behaviour. The single and multi-objectives were minimizing the sum of absolute mean pressure coefficients on the surface of the prism, and/or the total peak velocity ratio values around the prism. Two case studies focusing on the optimization design of a single front fence and dual fences were presented, in order to determine the optimal fence parameters (height and porosity). The weighted-sum method was utilized to solve multi-objective problems, which finds the Pareto-optimal front one by one by changing the objective weighting coefficients. The results showed that the optimal design of porous fence provided a remarkable shelter performance improvement compared with the near optimal design. It was also found that the weighting coefficient had a significant influence on the optimal results, and a good trade-off between the two conflicting objective functions has been achieved by selecting suitable fence parameters.