| PaperNO | Paper / Abstract |
|
G7-001
16:10
|
16:30
|
TOWARDS IMPLEMENTATION OF ACTIVE CONTROL SYSTEM USING ARTIFICIAL INTELLIGENT FOR FLEXIBLE STRUCTURES UNDER EARTHQUAKE EXCITATIONS
A trend towards more flexible structures and the necessity of more severe design constraints of motion for structures, such as sky scrapers or long span bridges has been observed. The advanced material science and Information Technology (IT), introduce the possibility of applying vibration control systems to civil structures. Because of the remarkable advances in core computing systems, the active vibration control on civil structures is progressing significantly both in its algorithm as well as in its implementation on flexible structures. Structural control heavily relies on computational instruments, that have limited processing power to calculate a large amount of data. Computing conventional optimal control algorithm on these instruments could result in a considerable response delay that can cause the structure to be unstable and uncontrollable especially for large structures. Artificial Intelligence algorithm, such as Artificial Neural Networks (ANN) is computationally more efficient and could be used as one of the solutions for a timely active control system.The main advantage of the ANN is its capability to learn direcly from data. It learns the mapping between input and output that could be used for the active control system. When the model and the learning algorithm are appropriately chosen, the ANN can be a robust tool for solving given problems. Their adaptability, robustness, and the inherent capability to handle most structural systems have been observed by many research groups.The purpose of the study is to evaluate the performance and the effectiveness of the ANN-based active control system strategy on a complex and large cable-stayed bridge structure. The ANN algorithm is used as the control force generator. The control forces were implemented on the deck and on the stayed cables of the bridge through some actuators. The learning process of the ANN was carried out by using the input-output data calculated using classic optimal control theory.To demonstrate the effectiveness of the proposed strategy, some numerical simulations were performed using a benchmark cable-stayed bridge subjected to certain recorded earthquake excitations.
Herlien D. Setio
Active-control, Artificial Neural Network, Cable-stayed Bridge
|
|
SE8-012
16:30
|
16:45
|
DEVELOPMENT AND VALIDATION OF SEISMIC-RESISTING DAMPERS: BUCKLING-RESTRAINED BRACE, SELF-CENTERING BRACE AND LEVER VISCOELASTIC WALL DEVICE
The added passive damping from the energy dissipation is a direct benefit to mitigate earthquake-induced hazards to buildings in small and strong earthquakes. This paper presents different passive energy dissipation devices verified from the simulation and test at NTU. First, a steel sandwiched buckling-restrained brace (SBRB) was developed to simplify and speed up the fabrication process. Unlike conventional BRBs that have a steel core embedded into a concrete-filled restraining member, the SBRB has a core plate between a pair of steel restraining member with or without concrete infill. This enables fast assembling in the shop and inspection after earthquakes by removing restraining members from the core. Secondly, a self-centering brace (SCB) that exhibits flag-shaped hysteretic response with minimal residual deformation was developed to reduce the residual deformation of the energy-dissipative brace to buildings in earthquakes. Multiple tests of a two-story, braced frame specimens were conducted at NCREE to validate the system response with SCBs, BRBs and a combination of these two different braces. Finally, in order to improve the seismic performance of building structures in small earthquakes, a new lever viscoelastic wall (LVEW) that combines a velocity-dependent VE damper and a displacement-dependent friction damper into a single steel wall damper was developed. Three full-scale steel wall specimens were designed and tested, dynamically and statically, to verify the lateral force versus deformation relationship in small and large earthquake motions.
Ping-Ting Chung, Yu-Ting Ling, Chun-Hsiang Huang, Wen-Hao Tseng, Steven Tsuang, Luh-Maan Chang, Yung-Hsiang Chen, Chung-Che Chou
Dual-core self-centering brace (DC-SCB), Lever viscoelastic wall (LVEW), Sandwiched buckling-restrained brace (SBRB)
|
|
SE8-013
16:45
|
17:00
|
Performance of friction-pendulum bearing systems subjected to near-fault ground motions
A series of shaking table tests were conducted using a friction-pendulum isolation system subjected to four sets of ground motions. Set 1 (Set 2) includes pulse-like records with pulse periods ranged between 0.5 and 2 (2 and 6) seconds; Set 3 are non-pulse-like records with average spectral accelerations similar to that of Set 2; and Set 4 are synthetic gourd motions spectrally matched to the response spectra of the records of Set 2. The displacement demand of the tested isolation system for Set 2 is much greater than that for Set 1 despite that both sets are all pulse-like records. Also, similar average spectral accelerations for Sets 2, 3, and 4 over a wide period range did not result in similar average responses of the isolation system.
Ya-Heng Yang, Yu-Chen Lin, Chang-Ching Chang, Yin-Nan Huang
base isolation, friction pendulum bearing, near fault, pulse-like ground motion, Shaking table test
|
|
SE8-016
17:00
|
17:15
|
BUILDING MASS DAMPER DESIGN BASED ON OPTIMUM DYNAMIC CHARACTERISTIC CONTROL APPROACH
A new seismic design approach, the Building Mass Damper (BMD), which comes from a combination of mid-story isolation and tuned mass damper (TMD) design concepts, recently attracts immense attention. Primarily because that the use of partial structural mass of the building as an energy absorber in the BMD design can overcome the drawback of limited response reduction due to insufficient added tuned mass in the conventional TMD design. In this study, an optimum building mass damper (OBMD) design approach, namely optimum dynamic characteristic control approach, based on a simplified three-lumped-mass structure model is proposed to seismic protection of both the superstructure (or tuned mass) and the substructure (or primary structure) respectively above and below the control layer. A series of sensitivity analyses and experimental studies on different parameters, including mass, frequency, and damping ratios, of a building designed with a BMD system were conducted. The test results verify the practical feasibility of the BMD concept as well as the effectiveness of the proposed OBMD design. Furthermore, by comparing with the numerical results of a mid-story isolated counterpart, it is demonstrated that the proposed OBMD design can have a comparable and even better control performance.
Kuo-Chun Chang, Bo-Han Lee, Shiang-Jung Wang, Wei-Chu Chuang
building mass damper, numerical study, optimum design, sensitivity analysis, Shaking table test
|
|
G7-013
17:15
|
17:30
|
PERFORMANCE IDENTIFICATION OF THE BI-AXIAL DYNAMIC TESTING SYSTEM
The bi-axial dynamic testing system (BATS) was established in the Tainan Laboratory of the National Center for Research and Earthquake Engineering (NCREE) in 2017. It is presently one of the few advanced large-scale testing machines that possess dynamic compression and shear testing capabilities. It is globally beneficial for not only research and development of seismic isolation technology but also for performing prototype and production tests on full-scale seismic isolators. Identifying its essential parameters, including the effective mass and friction coefficients, and its dynamic performance in an explicit manner is imperative before it can service the public. Therefore, a series of tests, including triangle wave cyclic loading tests with varied horizontal displacements and velocities as well as sin wave cyclic loading tests with varied horizontal displacements and excitation frequencies, were conducted to clearly understand the crucial parameters and dynamic performance of the BATS. Three stages with different test conditions and specimens were schemed as follows: (1) Triangle wave and sin wave cyclic loading tests on the bare testing system without applying any compression load were performed to identify the relation between horizontal velocities and average friction coefficients as well as the effective mass of the BATS; (2) Triangle wave and sin wave cyclic loading tests on flat sliding bearings under different vertical compression loads were performed, and a simple linear regression method was adopted to identify the average friction coefficients of the BATS at different horizontal velocities; (3) Through comparing the identification results with the test results of full-scale friction pendulum bearings and further discussions, the rationality and applicability of the identified effective mass and average friction coefficients of the BATS can be further demonstrated.
Chung-Han Yu, Cho-Yen Yang, Wang-Chuen Lin, Chiung-Lin Liu, Shiang-Jung Wang, Jenn-shin Hwang
bi-axial dynamic testing system, dynamic performance, essential parameter
|