in press                   Back to the articles list | Back to browse issues page

XML Persian Abstract Print


1- Assistant professor, Department of Civil Engineering, Ayatollah Boroujerdi University , nabavian@abru.ac.ir
2- Assistant professor, Department of Civil Engineering, P‌a‌y‌a‌m‌e N‌o‌o‌r U‌n‌i‌v‌e‌r‌s‌i‌t‌y, T‌e‌h‌r‌a‌n
3- Professor, Department of civil Engineering, Babol Noshirvani University of Technology
4- Associate professor, Department of Civil Engineering, Babol Noshirvani University of Technology
Abstract:   (314 Views)
High stiffness to weight ratio, ease and speed of handling, as well as having favorable architectural appearance cause that, double layer grids with ball joint system are widely used to cover large spans. A double-layer grid has a complex behavior due to a large number of elements and a particular type of joints; hence, structural identification of this type of structure is an important issue, which refers to the determination of natural frequencies, mode shapes, and damping ratios. These results are necessary to complete the structural health monitoring, finite element model updating and damage detection. Due to the limitations of input-output methods, modal parameters of civil engineering structures such as bridges, dams, tall buildings, and double layer grids are determined mainly by output-only modal identification. In output-only methods, the vibration parameters are determined based on the information acquired only from the structure’s output. In this work, physical model of a ball jointed double-layer grid with dimensions of 2.8 m at 2.8 m, which is supported on four steel pipes in four corners was made in the laboratory. The grid consists of 32 members connected together with 13 balls, each having ten threaded holes at different angles. each member consists of a middle pipe and connecting parts including conical piece, sleeve and high strength bolt at both ends of the pipe. The middle pipe has the nominal length, diameter and thickness of 120 cm, 7.64 cm and 0.35 cm, respectively. The horizontal center to center distance of adjacent balls in each layer of the grid is 1.414 m and the total height of the structure includes the column length (1.3 m) and the distance between the top and bottom layers (1 m), which is equal to 2.3 m in total. The approximate weight of the structure is 3532 N. All the members and the balls used in the grid are identical. After all the members of the grid have been assembled, the bolt at each joint is tightened in a series of steps by twisting the corresponding sleeve. Exciting the grid, its acceleration response was measured. The modal parameters were obtained using four output-only modal identification techniques; namely enhanced frequency decomposition (EFDD), curve-fit frequency domain decomposition (CFDD), data-driven stochastic subspace identification (SSI-DD) and covariance-driven stochastic subspace identification (SSI-Cov). Two types of excitations were used in output-only modal tests, namely direct and indirect excitations. Since the modal parameters obtained via input-output modal analysis have less uncertainty compared to the output-only modal analysis techniques, an input-output modal test was also performed and the results are considered as reference values. It deduced that parameters identified in the direct excitation, were more accurate compared to indirect excitation. The results showed that the natural frequencies and mode shapes of the double-layer grid were estimated with a high accuracy via the four methods. The greatest relative difference between the natural frequencies belonged to the second mode and equaled 2.07%. The dispersion of estimated damping was much higher compared to natural frequencies and mode shapes. The results indicated that identified damping in the direct excitation was lower than indirect one. Among the 4 methods, SSI-Cov had the least error in damping estimation of the double-layer grid. The values of estimated modal damping ratios were relatively low (fraction of 1%). The mean relative error of the identified parameters showed that the time-domain methods estimated the damping ratios with less error; While the frequency-domain methods identified natural frequencies and mode shapes with higher accuracy.
 
     
Article Type: Original Research | Subject: Civil and Structural Engineering
Received: 2024/02/21 | Accepted: 2024/07/10

Add your comments about this article : Your username or Email:
CAPTCHA

Send email to the article author


Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.