Inverse parameter identification and model updating for Cross-laminated Timber substructures

Abstract

Finite Element (FE) model updating is crucial for identifying key parameters in structural design and improving predictive accuracy. Despite extensive research on advanced FE procedures approved for user applications, persistent disparities remain in real-world scenarios, especially for complex materials like wood. Capturing accurate mechanical characteristics with traditional models poses challenges in sustainability projects. This study introduces a derivative-free model updating procedure using a Single-Objective Optimisation (SOO) incorporating observed and predicted natural frequencies and vibration modes. The objective function optimises tuning parameters to minimise discrepancies between predicted and observed outcomes. The focus is on Cross-laminated Timber (CLT), a composite wooden structure gaining traction as a sustainable alternative to materials like reinforced concrete and steel. However, the mechanical properties of CLT can vary due to inherent variability in wood’s mechanical characteristics. This research identifies sensitive mechanical properties — longitudinal Young’s modulus, internal shear moduli, and rolling shear modulus of CLT — using a model updating procedure based on a comprehensive set of data from Experimental Modal Analysis (EMA). The study provides mathematical algebraic derivations of the updating procedure and a step-by-step implementation algorithm to facilitate practical application in structural engineering.