Estimating nonlinear wind-induced response of roof cable nets by aeroelastic experiments and ML modeling
Fabio Rizzo,Aleksander Pistol,Luca Caracoglia
DOI: https://doi.org/10.1016/j.ress.2024.110183
IF: 7.247
2024-05-09
Reliability Engineering & System Safety
Abstract:The paper examines the structural engineering challenges related to the assessment of the wind-induced vertical displacements of lightweight, hyperbolic-paraboloid cable-supported membrane roofs. Analysis and comparisons are conducted using three distinct methods for calculating the roof vertical, out-of-plane displacements. Finite-element model (FEM) analysis is employed using: ( i ) estimated static nodal forces obtained from wind pressure coefficients determined by aerodynamic wind tunnel tests on a rigid building model, and ( ii ) loads found from wind pressure coefficients, estimated through machine learning (ML) methods, and ( iii ) measured roof response found from aeroelastic wind tunnel tests on a flexible model. The study examines three different roof geometries (square, rectangular and circular) and two distinct membrane curvatures for each geometry. Furthermore, wind directionality (three mean-wind incidence angles) and Reynolds number effects (seven mean wind velocities) are studied. Comparisons show that the non-linear FEM analysis, based on estimated static wind loads, underestimates the roof displacements at the roof center, compared to the direct measurements on the aeroelastic model. The main contribution of the study consists of a novel application of ML models, and Artificial Neural Networks (ANNs) in particular, which are employed to correct roof displacement estimations, found by simplified aerodynamic test pressure measurements on rigid roof models. The goal is to better describe the complex fluid-structure interaction of the roof membranes, which can only be achieved by direct aeroelastic tests that are difficult to design and execute. Finally, the study demonstrates that ANNs can be used not only for preliminary design of the full-scale structure but also for construction of wind tunnel scale models.
engineering, industrial,operations research & management science