In the present study, the effects of 2 different oil types (soybean and olive oil) and 3 different gums (xanthan gum, sodium alginate, locust bean gum, and their blends) on the rheological and physicochemical properties (pH, titratable acidity, moisture, and color), overrun, melting rate, melting time, and sensory properties of frozen mellorine samples were determined. | Turkish Journal of Agriculture and Forestry Research Article Turk J Agric For (2014) 38: 745-757 © TÜBİTAK doi: Modeling of rheological properties of mellorine mix including different oil and gum types by combined design, ANN, and ANFIS models 1 2 1, 3 Salih KARASU , Mahmut DOĞAN , Ömer Said TOKER *, Erdal CANIYILMAZ Department of Food Engineering, Faculty of Chemical and Metallurgical Engineering, Yıldız Technical University, Esenler, İstanbul, Turkey 2 Department of Food Engineering, Faculty of Engineering, Erciyes University, Kayseri, Turkey 3 Department of Industrial Engineering, Faculty of Engineering, Erciyes University, Kayseri, Turkey 1 Received: Accepted: Published Online: Printed: Abstract: In the present study, the effects of 2 different oil types (soybean and olive oil) and 3 different gums (xanthan gum, sodium alginate, locust bean gum, and their blends) on the rheological and physicochemical properties (pH, titratable acidity, moisture, and color), overrun, melting rate, melting time, and sensory properties of frozen mellorine samples were determined. Apparent viscosity of all mix samples decreased with shear rate, meaning that mellorine mix samples showed shear thinning behavior. Mellorine mix samples showed Ostwald–de Waele flow behavior (R2 ≥ ). Viscous synergy indexes were calculated to determine if the gums had synergic interaction. The viscous synergy index value of the xanthan and locust bean gum combination was found to be approximately , indicating synergic interaction between them. The effects of different gums on the apparent viscosity values at 50 s–1 (η50) were satisfactorily modeled by a modified power-law model. The adaptive neuro-fuzzy inference system (ANFIS) model was also found to be sufficient to predict apparent viscosity values based on the oil type, gum concentrations, and shear rate (R2 = ). .
