(192b) Correlating Blending of Pulps of Different Qualities

T. J. G. Dattatreya¹ N. K. Bhardwaj, A. K. Ray

Department of Paper Technology, IIT Roorkee, Saharanpur 247 001, India

In this present study the physical and mechanical properties of pulp stock and paper made from mixture of different qualities were evaluated. The properties studied were : Canadian Standard Freeness, CSF, water retention value, WRV, apparent density/bulk, grammage, thickness, tensile index, burst index and tear index as well as double folds .  The pulp mix were obtained by blending different proportions of short and long fibre pulps. The values are then compared with those predicted from the models developed. The pulp fibres were collected from a nearby  integrated pulp and paper mill located in Northern India. The short fibre collected was obtained from a mixed hardwood chemical pulp (Eucalyptus Tereticornis and Poplar), and the long fibre pulp was a softwood chemical pulp (Pinus Roxburghii). The initial CSF value of the hardwood pulp was 460 mL, and those of the softwood pulps  380 mL and 270 mL, respectively . The hardwood pulp was blended with two types of softwood pulp varying in  the CSF value. Several blends of the two fibre types were performed by adding different proportions of long fibre pulps to the short fibre pulp. The pulp and paper properties of these blends were determined with TAPPI standard procedures. It is well known that the property development of long fibre pulp of any CSF value when mixed with short fibre pulps of different CSF values (even for the same CSF value) cannot be predicted easily. Therefore, this extensive investigation was undertaken to examine the hypothesis whether the pulp blends follow the super position principles of additivity. In order to fulfil the aim statistical regression equations were developed with least square techniques. Both linear and polynomial parabolic/quadratic type models were tried.  The correlations obtained prove  satisfactory with reasonable accuracy permitted in engineering estimates.