Sensitivity analysis and optimization of hydraulically interconnected suspension parameters of mining dump trucks

Here, to improve driving stability and smoothness of tri axle mining dump trucks, a new suspension system was proposed with front axle left-right interconnected and middle and rear axle X-shaped cross interconnected. By using impedance matrix transfer method and system boundary conditions, a whole vehicle mechanical-hydraulic coupled model was established. The state matrix of the coupled system was solved to obtain various modes of vehicle body-wheel motion. Modal characteristics of the new suspension and those of traditional suspension were compared, and tire dynamic load distribution was analyzed. The system acceleration transfer matrix and random road surface input matrix were used to obtain the optimization objective function. Finally, Morris method was used to analyze sensitivities of suspension system structural parameters to body bounce, pitch and roll modal vibrations, and obtain sensitive parameters affecting suspension performance. Multi-objective optimization of these sensitive parameters was performed with NSGA-II algorithm. The results showed that compared to traditional suspensions, pitch and roll modal frequencies of the new suspension increase by 8. 3% and 35. 2%, respectively, so the new suspension has better anti-pitch and anti-roll performance; meanwhile, dynamic load distribution of middle and rear wheels is improved; initial inflation pressures of accumulators, initial inflation volume of front accumulator, inner diameter of damping hole and area ratio of upper chamber and lower chamber of cylinder barrel are the main factors to affect riding comfort; after optimization, peak acceleration responses of the vehicle's bounce, pitch and roll modes are reduced by 50. 3%, 58. 2%, and 59. 8%, respectively, RMS values of these accelerations under different vehicle speeds are reduced to effectively improve the comprehensive performance of the suspension. © 2024 Chinese Vibration Engineering Society. All rights reserved.