This study explores a new perspective on triboelectrification that could potentially lead to the development of a non-destructive approach for the rapid characterization of powders. Sieved yellow pea powders at various particle sizes and protein contents were used as a model system for the experimental charge measurements of the triboelectrified powders. A tribocharging model based on the prominent condenser model was combined with a Eulerian-Lagrangian computational fluid dynamics (CFD) model to simulate particle tribocharging in particle-laden flows. Further, an artificial neural network model was developed to predict particle-wall collision numbers based on a database obtained through CFD simulations. The tribocharging and CFD models were coupled with the experimental tribocharging data to estimate the contact potential difference of powders, which is a function of contact surfaces' work functions and depends on the chemical composition of powders. The experimentally measured charge-to-mass ratios were linearly related to the calculated contact potential differences for samples with different protein contents, indicating a potential approach for the chemical characterization of powders.
Keywords: artificial neural network (ANN); computational fluid dynamics (CFD); rapid characterization; tribocharging.