STATISTICAL OPTIMIZATION AND SENSITIVITY ANALYSIS OF RHEOLOGICAL MODELS USING CASSAVA STARCH

Models are sometimes employed to determine some parameters that can be used to distinguish between different types of food samples. Rheological models can be used to predict flow for severe conditions where it is difficult to determine the nature of the fluid flow, consequently it is essential to select the appropriate rheological models. This study aims to propose a rheological model that describes an ideal cassava starch rheological behavior and its influence on state variables such as concentration and temperature in order to validate the rheological models. In this study, five rheological models (namely; Power-law model, Robertson-stiff model, Herschel-Bulkey model, Prandtl-Eyring model and Bigham plastic model) were amended into various statistical model by adding the error variance (e). This study concludes that Herschel-Bulkley model and Robertson-stiff model closely explain the rheological patterns occurring during the production of cassava starch. The sensitivity evaluation of other rheological models demonstrate that the validity of Power-law model, Herschel-Bulkley model and Robertson stiff model is not notably influenced by changes in concentration and temperature of the cassava starch. Nevertheless, the Prandtl-Eyring and Bingham plastic models are noted to have less reliable prediction at lower temperature and higher concentration respectively.