Eventos Anais de eventos
COBEM 2023
27th International Congress of Mechanical Engineering
Predicting Friction Factors in Turbulent Flow of Herschel-Bulkley Fluids: A Radial Basis Function Neural Network Approach
Submission Author:
Glaucio Kenji Matoba , SP , Brazil
Co-Authors:
Glaucio Kenji Matoba , Daiane Iceri, Charlie van der Geest, Roney Thompson, Marcelo Souza de Castro
Presenter: Glaucio Kenji Matoba
doi://10.26678/ABCM.COBEM2023.COB2023-1123
Abstract
The accurate frictional pressure gradient is crucial for pipeline design systems in the industry. The prediction of frictional pressure gradient relies on factors such as pipeline diameter, mean velocity, fluid properties, and friction factor. However, predicting friction factors, especially for the turbulent flow of non-Newtonian fluids, remains challenging and prone to significant errors. In order to reduce these discrepancies, it is proposed an artificial neural network with a radial basis function (RBF) model to enhance friction factor prediction for non-Newtonian fluids with Herschel-Bulkley behavior. The RBF model was trained using a comprehensive database of 300 experimental data points. The model presented promising results, with an absolute error of approximately 10% compared to existing literature. This indicates the model's ability to provide reasonably accurate predictions for friction factors. Additionally, a statistical analysis was conducted to identify the most influential parameters of the friction factor, such as rheological parameters, Reynolds number, and pipe diameter. In summary, the proposed RBF model offers an effective approach to improve the estimation of friction factors for non-Newtonian fluids with Herschel-Bulkley behavior. It provides a valuable tool for pipeline system design, where accurate friction factor prediction is crucial. Moreover, the statistical analysis further enhances the understanding of the key parameters influencing friction behavior in non-Newtonian fluid flow.
Keywords
Herschel-Bulkley, Friction Factor, Turbulent flow, Radial Basis Function, artificial neural network
