Font Size: a A A

Numerical Simulation Analysis And Optimization Design Of Cyclone Separator

Posted on:2018-09-10Degree:MasterType:Thesis
Country:ChinaCandidate:Y Y XuFull Text:PDF
GTID:2351330515954247Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
The effect of the cyclone separator geometry sizes(inlet height,inlet width,gas outlet duct diameter,gas outlet duct insert length,gas outlet duct length,cyclone height,cylinder height,cone tip-diameter)on the performance was investigated computationally.The results show that pressure drop of cyclone decreases with increasing inlet height or width;and the grade efficiency curves become flatter and cut off diameter d50 increases and the separating performance decreases.The effect of changing the inlet width is more significant in comparison with that of the inlet height.As for the gas outlet sizes,the changing of the gas outlet duct diameter have a significant influence on both the pressure drop and cut off diameter.Whereas the gas outlet insert length and gas outlet duct length and cone tip diameter have an insignificant effect on the pressure drop and collection efficiency(the cut off diameter).Cyclone height strongly affects the pressure drop as well as the collection efficiency(cut off diameter).The pressure drop and cut off diameter decrease with increasing height(cylinder height or cone height),and the changing of the cylinder height is more sensitive to pressure drop in comparison with cone height,but the separating performance of the cyclone separator is just the opposite.A dataset of 98 samples obtained from the measurements of pressure drop for different cyclone designs available in literatures is used in the present investigation to evaluate the prediction performance of the artificial neural network.To predict the complex non-linear relationships between the dimensionless pressure drop(Euler number)and the geometrical dimensions,back propagation neural network is developed and employed.Five mathematical models have been tested against the BPNN predict values.The result demonstrates that Shepherd-Lapple model show the same variation trend of the gas outlet duct diameter Dx,inlet height a,inlet width b with BPNN predict values,but but the prediction precision is low in terms of sensitivity and magnitude.The Casal-Benet model failed in predicting Euler number trend or magnitude.Both the Iozia model and Barth model failed in predicting inlet width.However,the effect of the geometrical parameters on the Euler number obtained from MM model is very close to that obtained from the BPNN prediction in the sensitivity and magnitude.Combining the BPNN prediction and MM model,the response surface method has been used to fit the two second order polynomials(Euler number Eu,the cut size d50).Analysis of variance was used to evaluate the significance of the coefficients of the second order polynomial.The result shows that the most significant geometrical parameters are gas outlet duct diameter,inlet width,inlet height and cyclone height.There are strong interaction between the effect of inlet height and gas outlet diameter,inlet width and gas outlet diameter,while other sizes are not significant.The statistical software Design expert8.0 has been used to optimize cyclone separator to obtain a new cyclone geometry sizes.A comparison of numerical simulation of the new design and the Stairmand high efficiency cyclone has been performed.The results show that the cut off diameterdso of the new geometry sizes cyclone separator is basically the same as the Stairmand high efficiency cyclone separator,while the pressure drop decreases by 38.3%.
Keywords/Search Tags:cyclone separator, numerical simulation, geometric size, BP neural network, mathematical model, response surface methodology, prediction, optimization
PDF Full Text Request
Related items