Research on transport properties of liquid-solid two-phase flow in tapered pipes based on Witoszynski curve and machine learning optimization

In tapered pipes commonly used in engineering applications, increasing the particle mass flux and reducing specific energy consumption are key to improving the energy efficiency of liquid-solid two-phase conveyance. This study introduces the Witoszynski curve as the profile for tapered pipes and analyzes its transport characteristics in two-phase liquid-solid flow. The contraction ratio, contraction angle, and profile type were selected as optimization parameters, and the back propagation neural network (BPNN) and the genetic algorithm (GA) were employed as the prediction model and optimization algorithm. An optimization framework was developed to determine the optimal structural parameters for tapered pipes. The results show that using the Witoszynski curve as a contour guide allows particles to follow a smoothly varying curvature trajectory through the transition zone, thereby increasing particle velocity and improving flow field uniformity. Compared to the tapered pipes before optimization, the particle mass flux increased by 11.85%, and specific energy consumption decreased by 9.90%. Validation simulations conducted under various flow rates and with fluids of different viscosities demonstrate that the optimized model exhibits good robustness.