Prediction of electrodiffusio-osmotic transport of shear-thinning fluids in a nanochannel using artificial neural network

Abstract

We have numerically investigated the electrodiffusio-osmotic (EDO) transport of non-Newtonian electrolytic solution, governed by an externally applied electric field and concentration difference, in a charged nanochannel connected with two reservoirs. We have examined the EDO transport characteristics by varying electrical, chemical, and rheological parameters. The relative augmentation in net throughput due to EDO transport is compared to the pure electro-osmotic flow and is found to be greater than unity [reaches up to the order of ∼O(103)] for the considered range of concentration difference and flow-behavior index. As shown, the EDO throughput with concentration difference follows an increasing–decreasing trend at the smaller nanochannel height (<10 nm), while exhibiting an increasing trend at the higher nanochannel height (>10 nm). Notably, the net flow for shear-thinning fluid gets fully reversed at higher concentration differences and for a higher value of zeta potential. In the second part of the work, we discuss the use of an artificial neural network (ANN) essentially to predict the net EDO throughput from the nanochannel. The ANN model considered here is of a single-hidden-layer feedforward type. For activation, we used a sigmoid-purelinear transfer function between the layers. Additionally, the Levenberg–Marquardt algorithm is used to perform the backpropagation. To predict the volume flow rate per unit width, we have used four input features: concentration difference, flow-behavior index, nanochannel height, and zeta potential. We have established that an ANN model with eight neurons in the hidden layer accurately predicts the flow rate per unit width with a very small root mean squared error. The inferences of this analysis could be of huge practical importance in designing the state-of-the-art nanodevices/systems intended for offering finer control over the underlying transport.