Concentration Boundary Layer in Membrane Degumming: A CFD Model and Neural Network Approach

The use of crude vegetable oils without degumming, during biodiesel production, might decrease the conversion rate and be difficult to separate glycerol from biodiesel. Ultrafiltration is promising technology for gum removal from crude vegetable oils. However, since the oil constituents have very close molecular weight, degumming process by membrane is relatively difficult. In order to understand the membrane degumming process of corn-oil, a prediction of concentration boundary layer thickness was calculated using a CFD model. An artificial neural network is developed to learn the relationship between Reynolds and Schmidt numbers of feed solution which affects the boundary layer thickness along the membrane tube.