Second order approximations in model updating

The local quality of model parameterisation is one of the key ingredients of the methods used in the experimentally-based model updating of the mechanical systems. Traditional measures of the parameterization quality are based on the sensitivity of the primary observed structural responses to the changes in model parameters. This paper explores the use of second order information in model updating. While the use of these data is reported in the general literature on inverse problems and nonlinear regression, the use of this information is relatively unexplored in the field of model updating. It can be shown that there exists a direct link between regularisation approaches used in model updating and the second order derivatives originating from the extended part of the Taylor series approximation of the model data quadratic distance measure. In this sense, the second order derivatives act as regularising information in the model updating problem. The goal of the paper is to evaluate the use of second order derivatives in model updating. The theoretical concepts will be applied to the real model updating case study of a three story aluminium frame structure.