Statistical Approximations of Power Flow for a Vehicle Structure With Uncertainties

In this paper, an efficient approach is presented for probabilistic power flow analysis of complex structures with parameter uncertainties. A component-based technique is used to generate reduced order models of low- to mid-frequency vibration and power flow. The reduced order model (ROM) is built from component mode synthesis of finite element models, followed by a secondary modal analysis to reduce the number of degrees of freedom associated with the component interfaces. This formulation allows for efficient and accurate prediction of vibration transmission in complex structural systems. Then, this methodology is extended to the study of statistical approximations of the power flow for a structure with parameter uncertainties. Two techniques are employed for probabilistic power flow analysis: (1) a locally linear interpolation (LLI) method for low-dimensional random variable domains, and (2) an advanced mean value (AMV) method for multi-dimensional random variable domains. In addition, an extended version of the AMV method is presented for uncertainties characterized by a uniform distribution. These statistical approximations of power flow are first demonstrated for a simple example system, an L-shaped plate. They are then examined in more detail for an important engineering application, a ground vehicle model.Copyright © 2003 by ASME