A study on vibration prediction of diesel engine based on virtual material modeling method of joint surfaces

Abstract

The vibration prediction model of diesel engine is generally established by finite element method (FEM). In FEM, the components are firstly represented by mesh models, which will be connected to develop a combined model through joint surfaces. The processing method of joint surfaces is key to the accuracy of the combined model, for joint surfaces play an important role in transmitting force and vibration. For this purpose, an equivalent model of transversely isotropic virtual material is proposed to connect the mesh models of the components in this paper. In this model, the different connection characteristics of both normal and tangential are taken into consideration. In addition, the calculation models of the virtual material layer parameters, such as elastic modulus, shear modulus, Poisson’s ratio, thickness, density, and damping coefficient, are established in theory. At the same time, the modal test is introduced to acquire the natural frequency and damping ratio for identifying the unknown parameters of the virtual material layer, by using the BP neural network algorithm. After that, the complete finite element model of a diesel engine assembly can be established, and the vibration prediction is made in time domain, where cylinder gas pressures, piston slap forces, main bearing forces, and valve train forces are taken into consideration comprehensively. Furthermore, the experiment of the engine vibration response is carried out as well. The vibration acceleration level of engine feet and the vibration severity of engine block are adopted to evaluate the structure vibration response of the diesel engine. The results show that there is a good consistency between the prediction results and the experiment results in frequency domain. The consistency of the vibration acceleration level in 1/3 octave is up to 8 kHz, and that of the vibration severity is up to 1 kHz. And the results verify the accuracy of the vibration prediction model of the diesel engine, which is based on transversely isotropic virtual material modeling method of joint surfaces.