Abstract

This thesis delves into the performance of balloon-type cross-laminated timber buildings. The study focuses on hybrid buildings with balloon-type cross-laminated timber shear walls and steel frames. The effects of mass and vertical geometric irregularities in hybrid buildings are evaluated using Modal Response Spectrum Analysis. The findings reveal that mass irregularities markedly influence inter-story drift, mainly when mass is incorporated at different levels, and that the location of mass addition is vital to the building's seismic response. The vertical geometric irregularity of the building impacts the seismic response uniquely in two orthogonal directions.
Further, the study explores the lateral performance of unbonded post-tensioned balloon-type CLT shear walls. Traditional sensitivity analysis and machine learning models were employed to identify the critical parameters influencing the lateral and uplifting response of the unbonded post-tensioned balloon-type cross-laminated timber shear wall under lateral loading. Various machine learning algorithms were utilised to investigate the best performance model for predicting the critical parameters while applying Shapley Additive explanations (SHAP) to elucidate better the factors influencing lateral and uplifting responses.
The thesis offers valuable perspectives on integrating advanced computational techniques with traditional structural engineering practices, fostering the development of resilient and sustainable building designs.