An experimental investigation on soda pulping of industrial bagasse was carried out in the laboratory with or without the presence of oxygen. The effects of temperature, percentage of alkali, and inhibitor like magnesium sulphate on yield, rejects, and kappa number were examined. One sample was also tested for EDTA in place of Mg~(++). The results were statistically evaluated. A new factor, h, proportional to H factor is proposed. The statistical models are built for yield, rejects and kappa number in terms of temperature, concentration, and proposed new h factor.From detailed analysis, it is found that yield of soda- oxygen pulp can be greatly enhanced by using inhibitors like MgSO4. Even E.D.T.A. can act as an inhibitor. Yield can be increased by about 15 - 20% in Soda - Oxygen process (Mg++ inhibitor) compared to soda pulp of bagasse. Maximum yield is obtained at 150°C and 14% alkali charge for cooking time of lhour and oxygen pressure of 5 bar. Center cook (150°C, 14%) was beaten to 45 °SR and sheets made yielded tensile of 3.39 kN/m and burst to be 196.13 kpa.m~2/g.An analytical model defining h proportional to H factor has been developed. Pulping experiments were carried out based on Statistical design of experimentation. The results indicate the following:(1) Soda - oxygen pulp yield can be greatly enhanced by using inhibitors like MgSO4.(2) Even E.D.T.A. can act as an inhibitor.(3) Yield can be increased by about 15 - 20% Soda - Oxygen (Mg++ inhibitor) compared to soda pulp of bagasse.(4) Maximum yield is obtained at 150°C and 14% alkali charge for cooking time of lhour and oxygen pressure of 5 bar.(5) Center cook (150°C, 14%) was beaten to 25°SR and sheets made gave tensile of 3.39 kN/m. Burst was found to be 196.13 kpa.m~2/g.Based on the data statistical multivariable nonlinear regression equations are developed for yield, rejects, and kappa Number. Two groups of models are attempted. Both the groups of statistical models are compared.The models predicted data agrees very closely with the experimental data with R~2 = 1. Therefore the models are found to be excellent. It is very difficult to distinguish between the two groups of models. Any one of them can be employed for prediction.
展开▼
