Synthesis of multiaxis stationary non-Gaussian shaped vibration

Abstract

Random vibration is present in many fields of engineering, although one where it has particular relevance is the industry of ground transportation. The interaction between the wheels and the road induces vibration that transmits to all the components in the vehicle, as well as to the passengers, producing mechanical damage, noise and discomfort. It is hence important to address the effect of vibration during the design and validation phases of development of the vehicle. For that purpose, a method for the synthesis of signals that can be used for numerical simulation or physical testing of the system being designed is needed. In this thesis, an algorithm for the synthesis of random vibration signals is developed, that takes into account the multiaxial and non-Gaussian nature of the process allowing individual adjustment of the kurtosis of the different degrees-of-freedom involved. The proposed algorithm avoids distortion of the spectrum when the kurtosis is being adjusted hence allowing for independent control of the spectral and kurtosis properties of the process. Theoretical expressions relating the control paramet-ers to the kurtosis and the spectrum of the process are provided. An additional benefit of the proposed algorithm is the possibility to modify the higher-order moments beyond the kurtosis.