Multi-scale model updating of a transmission tower structure using Kriging meta-method

A multi-scale model is often constructed using different finite elements and consists of a global scale model for the structural system and a few local scale models for critical structural components so that the multi-scale simulation can concurrently exhibit both global performance and local behavior of the structure. To ensure the multi-scale model can best represent the real structure, multi-scale model updating technique shall be developed accordingly. This paper thus presents a multi-scale model updating method for a transmission tower structure using the Kriging meta-model that actually is a surrogate for the multi-scale model. Firstly, the multi-scale model of a transmission tower is established by using beam elements to simulate global structure and solid elements to simulate local joints with bolt connections. Secondly, the multi-objective optimization problem that involves multiple objective functions is established to update key parameters of the multi-scale model so that the errors between the measured and predicted structural dynamic characteristics and multi-scale responses can be minimized. To improve the computational efficiency and accuracy of optimization, the Kriging meta-method is used to find the updated key parameters of the tower after a comparison with other meta-methods is made. Finally, the proposed method is applied to a physical transmission tower model, which has been tested in a laboratory, to demonstrate the feasibility and accuracy of the proposed model-updating method. The updated results show that the proposed updating method can improve the accuracy of the multi-scale model of the tower in both global and local structural responses.