The present study focuses on the development of a novel venturi-based self-excited oscillation mixer that effectively utilizes the venturi effect to facilitate efficient abrasive intake while simultaneously ensuring effective prevention of backflow through the utilization of the systolic section within the venturi tube. It not only ensures uniform mixing of water and abrasive but also transforms the continuous jet into a pulsed one, thereby significantly enhancing exit velocity. The orthogonal experimental design method and single factor experiment method were employed to investigate the effects of inlet water pressure, water nozzle diameter, abrasive inlet angle, aspect ratio of the self-excited oscillation mixer, and abrasive pipe inlet diameter on the inlet pressure of the abrasive pipe and the velocity of the jet exit in the new mixing device. Approximate response surface models for these parameters were constructed using lsight optimization software, combining the results of orthogonal experimental simulation. By employing a multi-island genetic algorithm, we have globally optimized this innovative mixing device to determine its optimal performance parameters. Subsequently, comparative experiments were conducted to validate the performance of different mixing devices in descaling applications. Through experimental verification, it was found that the venturi-self-excited oscillation mixer exhibits excellent rust removal capabilities in steel plate tests compared to traditional self-excited oscillation mixers. These findings provide valuable guidance for the subsequent design and enhancement of abrasive water jet mixers.
To address the challenges of steel material loss (9%–12%) caused by iron oxide sheet formation and environmental contamination from conventional pickling rust removal methods, this study investigates mixture flow descaling (MFD) device as a potential alternative. While abrasive waterjets demonstrate descaling potential, their structural optimization mechanisms remain insufficiently explored. A Kriging-assisted modified non-dominated sorting genetic algorithm II optimization framework is proposed to enhance the descaling efficiency of the MFD device. Parametric analysis systematically evaluated how blade length, jet distance, and flow velocity collectively influence the descaling performance of the MFD device, with particular emphasis on quantifying interaction effects between these operational variables. Simulation and experimental validation were performed to verify these optimized results. After optimization, the simulation demonstrated that the erosion rate increased by 72.76%, and the descaling width increased by 81.97%. Additionally, the experimental results for surface roughness and rust curve provided further validation of the proposed algorithm's effectiveness. This research provides actionable design principles for optimizing water-slurry blasting systems in steel strip descaling applications.