Investigation on evolution law of water flow deterioration caused by sedimentation in sewer pipelines: An approach based on fluid–structure coupling

With the increase in the length and age of urban sewer pipeline construction, black smelly water pollution caused by siltation and deposition has increased in sewer pipelines, affecting their flow capacity and increasing the risk of flooding. This presents a significant challenge to the water environment, urban environment, and human life and health. Hence, to investigate potential approaches for controlling pipeline siltation and deposition, governing equations and dynamic simulation model of fluid–structure coupling for silted pipeline are constructed by combining theory of fluid dynamics analysis, method of unit volume analysis, computational fluid dynamics (CFD), and discrete element method (DEM). Then, to predict sedimentation law of pipeline siltation particles with high accuracy, an adaptive punishment mechanism (APM) for intelligent prediction of particle settlement of pipeline sediment and hyperparameter optimization is adopted based on probability settlement function (PSF), non-dominated sorting genetic algorithm (NSGA) and multiscale bidirectional long short-term memory neural network (MBLSTM). By combining self-punishment mechanism based on PSF-NSGA-MBLSTM intelligent prediction of pipeline siltation particle settlement with CFD-DEM, a high-resolution numerical simulation method CFD-DEM-APM for hydraulic transport of siltation particles in sewer pipelines is proposed to describe the particle-phase and fluid-phase state transfer process. Experimental results show that the accuracy of CFD-DEM-APM is maintained within the 5%–11% range, which is far better than that of other algorithms. This study provides guidance on critical conditions for desilting and pipe inlet velocity/flow control to scour siltation and deposition of sewer pipelines for urban environment improvement and water pollution restriction.