The chemical engineering industries are utilizing the bottom inlet cyclone separator with venturi for separating the particles from an air/gas medium. For improving the performance of this equipment, important geometrical features such as venturi inlet width, total height of the cyclone and body height of the cyclone are considered for optimization. Central composite design was used in response surface methodology (RSM) to fit the regression equation. This regression equation was evaluated by analysis of variance (ANOVA). Then, this polynomial equation was optimized by particle swarm optimization (PSO) for minimizing the cut-off diameter. These optimized results were compared with genetic algorithm (GA) results. Based on this optimized result, an experimental setup was created for validation purpose. The experimental results were compared with GA and PSO results. A good agreement was obtained between these results. The magnesium particles were utilized for predicting the cut-off diameter of the new design. The Stokes number of this new design was less when compared with the mathematical model. The new design gives better performance when compared with the mathematical model. The numerical simulation was executed for predicting the particle collection efficiency, cut-off diameter and flow pattern inside the cyclone. The results were compared with the mathematical model and venturi inlet tangential entry cyclone.