Optimizing shape for hydraulic characteristics of Myring portable measuring flume

Water resources can be efficiently and precisely allocated in the irrigation districts of modern agriculture, particularly in the context of the ever-increasing water scarcity. However, some operational challenges still remained on the flow measurement facilities that are currently deployed in terminal channels. Furthermore, conventional devices are frequently susceptible to severe sediment deposition, excessive head loss, and low measurement accuracy due to the complex hydraulic conditions and silt-laden water in the networks. Precise water metering and equitable water allocation are substantially limited to increase the maintenance costs and labor intensity. This study aims to optimize a portable flow measuring flume using the Myring streamline profile, and specifically engineer it for the high-precision water quantity measurement in irrigation networks. Among them, the Myring profile was originally recognized in the field of underwater vehicles, due to its exceptionally low-drag characteristics. A throatless flume structure was also constructed after optimization. Smooth flow transitions effectively minimized the flow resistance after aerodynamic design. Moreover, the formation of the vortices was prevented to reduce the sediment accumulation. The hydraulic performance of this device was enhanced to integrate the Computational Fluid Dynamics (CFD) simulations with the physical model experiments. The geometric parameters were identified as the total length of the flume l, the sharpness factor of the contraction section n, and the departure angle of the diffusion section θ. An advanced framework of optimization was established to effectively navigate the complex and multi-parameter space. A Sobol sequence sampling was also utilized to minimize the head loss ratio. A uniform and representative initial sample space was constructed for the global coverage of the parameter range. Subsequently, a coupled surrogate model was employed to integrate the Back Propagation (BP) neural network with the Particle Swarm Optimization (PSO) algorithm. While the BP network was used to map the highly non-linear relationship between geometric parameters and hydraulic efficiency. The PSO algorithm was also introduced to perform a global search, effectively overcoming the conventional BP networks to trap in local optima. The hydraulic performance of the optimal structure was then verified in a rectangular channel using FlOW-3D numerical simulation and physical tests. A contraction ratio ε and a flow rate were also covered the range of 0.50 to 0.66 and 25 to 55 l/s, respectively. The results indicate that the superior hydraulic performance was achieved in the Myring streamline flume after PSO-BP optimization, compared with the conventional ones. The exceptional energy conservation was also obtained to significantly minimize the head loss in a range of only 2.00 to 3.00 cm. The upstream backwater height was effectively controlled between 1.57 and 3.87 cm, thereby minimizing the impact on the upstream channel’s conveyance. Furthermore, the upstream Froude number remained consistently below 0.50 over all working conditions. A stable subcritical flow reduced the surface fluctuations for the reading stability. The average relative error in the flow measurement was found to be 1.8%, indicating a high degree of precision suitable for trade metering. In conclusion, the portable flume fully complied with the standard specifications for flow measurement in irrigation districts. This device can offer a simplified structure, portability, robust sediment transport capacity, minimal head loss, and high measurement accuracy, significant value for widespread application. The finding can provide a reliable technical solution for smart irrigation in precision agriculture.