Droplet size prediction in a microfluidic flow focusing device using an adaptive network based fuzzy inference system

Microfluidics has wide applications in different technologies such as biomedical engineering, chemistry engineering, and medicine. Generating droplets with desired size for special applications needs costly and time-consuming iterations due to the nonlinear behavior of multiphase flow in a microfluidic device and the effect of several parameters on it. Hence, designing a flexible way to predict the droplet size is necessary. In this paper, we use the Adaptive Neural Fuzzy Inference System (ANFIS), by mixing the artificial neural network (ANN) and fuzzy inference system (FIS), to study the parameters which have effects on droplet size. The four main dimensionless parameters, i.e. the Capillary number, the Reynolds number, the flow ratio and the viscosity ratio are regarded as the inputs and the droplet diameter as the output of the ANFIS. Using dimensionless groups cause to extract more comprehensive results and avoiding more experimental tests. With the ANFIS, droplet sizes could be predicted with the coefficient of determination of 0.92.