IJASEIT

International Journal on Advanced Science, Engineering and Information Technology, Vol. 9 (2019) No. 4, pages: 1302-1308, DOI:10.18517/ijaseit.9.4.8566

Response Surface Methodology Approach to the Optimization of Cyclone Separator Geometry for Maximum Collection Efficiency

Y Yunardi, Umi Fathanah, Edi Munawar, Bayu Pramana Putra, Asbar Razali, Novi Sylvia

Abstract

A Response surface methodology coupled with a Box-Behnken design experiment has been utilized to optimize geometry parameters of a cyclone as a gas-solid separator in an effort to obtain a maximum particle collection efficiency. Independent variables being optimized include seven geometry parameters of inlet height (a/D), inlet width (b/D), vortex finder height (S/D), vortex finder diameter (D_e/D), total cyclone height (H_t/D), cylinder height (h/D), and cone tip diameter (B_c/D). A number of 62 treatments were performed following Box-Behnken experimental design of seven factors and three levels (-1, 0 and +1). The response variable, the cyclone collection efficiency, was calculated in accordance with the Muschelknautz model using a spreadsheet software. The relationship between the response variable and independent variables was mathematically expressed according to a quadratic polynomial equation calculated with the aid of Design Expert software. The results of the research showed that among seven variables being investigated, there are only five cyclone geometry parameters which significantly affected the cyclone collection efficiency, including inlet height (a/D), inlet width (b/D), vortex finder height (S/D), vortex finder diameter (D_e/D) and total cyclone height (H_t/D). The optimization was then conducted to include these five variables that significantly affected the collection efficiency and neglected the remaining other two variables. The optimization computation was run in the Design Expert statistical software by setting a maximum possible value for the collection efficiency. The maximum collection efficiency of 91.244% was obtained when the independent variables of inlet height a/D=0.8, inlet width b/D=0.38, vortex finder height S/D=0.69, vortex finder diameter D_e/D=0.575 and total cyclone height H_t/D=3.12. Validation of this statistical finding was tested again and compared with the result of Muschelknautz model calculation to give a significantly small error of 0.82%.