PARAMETER IDENTIFICATION OF NONLINEAR CONSTITUTIVE LAWS BY AN UNSCENTED KALMAN FILTER

Summary. In this paper we deal with the calibration of nonlinear constitutive laws via the recently proposed unscented Kalman filter (UKF). This nonlinear filter does not require the computation of the gradient of the equations governing the dynamics of the structural system; furthermore, the statistics of the system state variables-model parameters and nodal displacements- are accurately propagated in time up to the third order. In case of the numerical simulation of spalling in ceramics, we compare the performances of the UKF and of the usually employed extended Kalman filter (EKF).