Computational fluid dynamics-based structure optimization of ultra-high-pressure water-jet nozzle using approximation method

TL;DR

This paper develops a CFD-based optimization framework for ultra-high-pressure water-jet nozzles, integrating an enhanced sparrow search algorithm and SVM prediction to improve hydrodynamic performance.

Contribution

It introduces an LTC-SSA algorithm and LTC-SSA-SVM model for more accurate and diverse optimization of UHP water-jet nozzle structures.

Findings

Optimized nozzle structures show significantly improved hydrodynamic performance.

LTC-SSA outperforms standard SSA in avoiding premature convergence.

The combined LTC-SSA-SVM model enhances prediction accuracy of wall shear stress.

Abstract

Since the geometry structure of ultra-high-pressure (UHP) water-jet nozzle is a critical factor to enhance its hydrodynamic performance, it is critical to obtain a suitable geometry for a UHP water jet nozzle. In this study, a CFD-based optimization loop for UHP nozzle structure has been developed by integrating an approximate model to optimize nozzle structure for increasing the radial peak wall shear stress. In order to improve the optimization accuracy of the sparrow search algorithm (SSA), an enhanced version called the Logistic-Tent chaotic sparrow search algorithm (LTC-SSA) is proposed. The LTC-SSA algorithm utilizes the Logistic-Tent Chaotic (LTC) map, which is designed by combining the Logistic and Tent maps. This new approach aims to overcome the shortcoming of "premature convergence" for the SSA algorithm by increasing the diversity of the sparrow population. In addition, to improve the prediction accuracy of peak wall shear stress, a data prediction method based on LTC-SSA-support vector machine (SVM) is proposed. Herein, LTC-SSA algorithm is used to train the penalty coefficient C and parameter gamma g of SVM model. In order to build LTC-SSA-SVM model, optimal Latin hypercube design (Opt LHD) is used to design the sampling nozzle structures, and the peak wall shear stress (objective function) of these nozzle structures are calculated by CFD method. For the purpose of this article, this optimization framework has been employed to optimize original nozzle structure. The results show that the optimization framework developed in this study can be used to optimize nozzle structure with significantly improved its hydrodynamic performance.