D-optimal designs for the Cross viscosity model applied to guar gum mixtures

Abstract

D-optimal designs are developed using the Cross viscosity model for aqueous solutions of guar gum. Designs are developed for a 2 6 factorial design in the model parameters to study the effect of the model parameters on the experimental design. Analysis of this study shows that of the six experimental design points, two were independent of the six model parameters and the remaining four were dependent on only three of the model parameters. Additional analysis using ANOVA showed that only two of the model parameters, concentration and shear-rate exponents, statistically affect the design efficiency at the 0.05 level within the range of the model parameters studied. Furthermore, it is shown that incorrectly specifying the model parameters by as much as 50% reduces the design efficiency to 50–90%.