Uncertainty Analysis in Structural Dynamics

Abstract

There has been a lot of recent interest on the effects of uncertainty in the design of engineering structures. This is primarily due to the fact that the properties of the structure are uncertain would lead to uncertainty in the dynamic response. Two main factors that can result in these changes in structural properties are uncertainty and manufacturing variability. The former can arise as the property of an individual system can change with time due to variation of environmental conditions, loads, wear and etc. Similarly, the latter is inevitable as it is obvious that mass-produced products are never identical. The uncertainty and variability can lead to changes in the dynamic response, such as natural frequencies and frequency response functions. Consequently, there can be issues concerning safety, reliability and performance of any system. It is thus important to develop modelling and analysis techniques to take into account the uncertainty effects on the dynamics of a structure. One possible approach to quantify the uncertainties in the dynamic response is to use Monte Carlo simulation. However, the associated computational cost can be high as many repetitions may be required. It can be higher when there are multiple uncertainty parameters. Consequently, there is a need for more advanced analysis approaches to address the issues. This talk will focus on uncertainty analysis of structural dynamic problems. The research background will firstly be introduced. A variety of methods, including the polynomial chaos expansion method and a Gaussian Process based emulator approach will be presented. The application of some uncertainty analysis methods will be demonstrated with case studies.