Model Updating and Damage Detection of a 3D Printed Model Frame

Abstract

Finite element modelling is the one of most commonly used method for structural analysis. However, there are some discrepancies between experimentally identified and finite element solutions due to modelling assumptions. Therefore, initial finite element model needs to be calibrated according to experimental results. Damage can occur in structures due to earthquake, wind, fire, corrosion, fatigue etc. effects. Reliable damage detection is important for the utilization of the structure after damaging events. In this study, two story-one bay frame is numerically modelled with using finite elements and modal analysis is performed using this model. For calibrating the initial numerical model, shake table tests are conducted on the 3D printed model. Using natural vibration frequencies and mode shapes, initial finite element model of the 3D printed model is calibrated by varying modulus of elasticity within an optimization problem. It is assumed that modulus of elasticity is the uncertain parameter. After that a controlled damage is introduced on one of the first story columns and shake table tests are again performed on the damaged model. Finally using the same optimization based model updating technique, the introduced damage is detected. Keywords— Model updating, damage identification, finite element modeling, 3D printed model.