A Comparative Study of Factorial and Central Composite Experimental Design in Modeling Chemical Pulping

摘要

We performed 27 kraft pulping trials of Anatolian Black Pine(Pinus nigra subsp.Pallasiana)according to a 3x3x3factorial experimental design,with duplicate experiments for each.The Factorial Experimental Design(FED)and Central Composite Design(CCD)were chosen to establish the regression models to predict some pulp properties (screened yield,total yield,kappa no,viscosity,holocellulose content and α-cellulose content).Independent process variables included 15-45 % sulphidity,14-22 % active alkali charge and 800-2400 Vroom's H-factor.Both 27 trials and selected 15 trials from 27 total pulping experiments according to the CCD experimental program were used for the data analysis.The results were fitted to a second order polynomial function and the prediction ability of the equations derived both FED and CCD methods were compared.The prediction of pulp properties such as kappa number,viscosity,total yield,screened yield,holocellulose and alpha-cellulose content,based on equations derived from CCD methods gave satisfactory results.The optimum kraft pulping conditions of Anatolian Black Pine were determined as 30% sulphidity,1600 H-factor and 18% active alkali charge.