The Performance Optimization And Experiment Study Of Magneto-rheological Engine Mount

Vehicle NVH problems caused by the engine have become increasingly prominentin the trends of automobile lightweight and high-power engine. To meet consumers’increasing comfort requirement, the transmition of vibration caused by engine to thevehicle body or frame have to be suppressed effectively, and vibration caused by theroughness road excitation to engine also needs isolation. But traditional engine mount isdifficult to meet the demand of ideal engine vibration isolation. Therefore, the researchof novel high-performance mount is necessary.A novel flow mode MR (magneto-rheological) engine mount is designed in thispaper which combines to the application of smart materials, and its performance isoptimized to meet the requirement of engine vibration isolation system better. So theseoptimization design theory and technology support have important theoretical andpractical value. The main contents include:①According to the ideal dynamic characteristics of the engine mount system, aMR engine mount with decoupler, disturbing plate and controllable damping channel isdesigned. Its operating principle is demonstrated, and the main structural parameters ofMR mount are determined.②According to the excitation conditions of engine, dynamic characteristic’spiecewise model is established by bond graph theory. The dynamic characteristic of MRmount is analysised, and the rationality of dynamic characteristic’s piecewise model isalso demonstrated.③Optimization variables of MR engine mount are selected by lumped parametermultiple interval sensitivity analysis, which is aim to lumped parameter’s characteristics.What’s more, the vibration isolation performance of MR mount is optimized base onnon-dominated sorting genetic algorithm (NSGA-II).④The sample of optimized MR engine mount is machined, and its static anddynamic characteristics is tested which verify the validity of MR mount’s theoreticalmodel and structural design methods.⑤Taking advantage of neural network’s nonlinear approximation ability, MRengine mount’s generalized regression neural network (GRNN) model and inversemodel are established base on experimental data. And the predictive ability of GRNNmodel and inverse model is analysised, which establish the foundation for engine semi-active vibration isolation control.