Rapid algorithm to find optimal step excitations for modal testing of long-span roof structures. (15th December 2019)

• Record Type: Journal Article
• Authors: Wu, Xiao-shun; Xu, Si-dong; Feng, Long-kun
• Is Part Of: Engineering structures. Volume 201(2019)
• Publication Date: 2019-12-15
• Subjects: Modal testing -- Modal identification -- Closely spaced modes -- Ambient excitation -- Step excitation; Structural engineering -- Periodicals; Structural analysis (Engineering) -- Periodicals; Construction, Technique de la -- Périodiques; Génie parasismique -- Périodiques; Pression du vent -- Périodiques; Earthquake engineering; Structural engineering; Wind-pressure; Periodicals; 624.105
• DOI: 10.1016/j.engstruct.2019.109792
• Languages: English
• ISSNs: 0141-0296

Abstract

Highlights: Step excitations are proposed to trigger structural free vibration of long-span roof structures. Step excitation optimization triggers target mode without disturbance of its adjacent modes. Relationship between objective indicator vector and load distribution is established. A rapid bounded-variable least square solution method is adopted to optimize step excitations. Optimized step excitations improve the modal identification accuracy of closely spaced modes. Abstract: A rapid algorithm is proposed to obtain optimal step excitations for the modal testing of long-span roof structures (LRS). Step excitations of LRS are realized by suddenly removing gravity loads. The optimization of step excitations aims to strengthen the contribution of a target mode to the structural free vibration and to suppress the participation of its adjacent modes at the same time. In this way, the target mode can be well triggered without the disturbances of its adjacent modes. The simplified mathematical relationship between the objective indicator vector and the load distribution is established. By solving a bounded-variable least square problem, the optimal step excitations are rapidly acquired. The numerical study of a Geiger cable dome and both numerical and experimental studies of a spatial truss are carried out. The results show that the rapid algorithm can quickly complete the optimization of step excitations, and help the classic Ibrahim Time Domain method greatly improve the identification accuracy of closely spaced modes of LRS.