Refine
Document Type
- Doctoral thesis (2)
Has Fulltext
- yes (2)
Is part of the Bibliography
- no (2)
Language
- English (2)
Keywords
Institute
The accelerated urbanization has led to increasing tension on urban land use. In this context, more and more slender high-rise buildings are being built worldwide in pursuit of better economic benefits. However, these structures are susceptible to wind excitation due to their lower first natural frequency. Different passive, semi-active, and active damping systems have been developed to reduce wind-induced structural vibration. Among them, the tuned mass dampers are widely used and proved as a very effective method in practice. However, this system requires a large additional damping mass. This also causes additional reinforcement, which increases the cost and carbon footprint. A huge space near the top story of the building is needed for the installation. In this research, a novel system named distributed-Multiple Tuned Facade Damping (d-MTFD) system is proposed by using specially designed parallel moveable Double-Skin Facade (DSF) outer skin as damping mass. These moveable facade elements can be installed on the upper stories of the high-rise building. Smooth-running guide rail systems are used to achieve the parallel moveability. Multi-objective optimization based on the Genetic Algorithm (GA) is applied to reduce the maximum top floor acceleration (Objective I) and to reduce the maximum facade relative displacement (Objective II) simultaneously. The optimization results for the passive and semi-active systems are presented in the form of the Pareto front. The trade-off between these selected two competing optimization objectives is observed. This approach was first validated in a simulation using a 306 m tall reference building for a wind speed of 13.5 m/s at 10 m above ground level with a return period of 10 years. Acceptable peak accelerations at the top story for hotel use and a maximum facade relative displacement of less than ±0.5 m could be achieved for the benchmark building with the d-MTFD system. For semi-active control, the variable damping coefficient can be achieved by using stepper motors in generator mode. The electrical damping coefficient can be continuously adjusted by the developed power electronics. In addition, electrical energy can be generated and stored in a battery. A full-scale prototype with one parallel moveable facade element was built. Based on the prototype, the functionality of the semi-active control using a stepper motor and its energy harvesting performance was tested by applying Hardware-in-the-Loop (HiL) simulations. Greybox system identification was used to estimate some parameters (spring stiffness, friction, etc.) in the connection. The accurate system identification results ensure further validation using HiL simulations. The HiL simulations successfully demonstrated the feasibility of a self-powered semi-active d-MTFD system.
The linear parameter-varying framework for active vibration control of elastic kinetic structures
(2024)
Lightweight design is essential for preserving resources, and in addition, transformability can be beneficial for specific applications where multi-functionality is required. Multi-functionality contributes to saving resources by allowing multiple purposes to be served effectively and by allowing adaptation to changing needs. Elastic kinetics are a recent approach to design transformable lightweight structures with a stable transformation process, realized by elastic bending of structural members. Their lightweight and flexible design comes at the cost of increased sensitivity to static and dynamic disturbances. However, most of the current research concentrates on transformation principles instead of on disturbance mitigation. This research focuses on dynamic disturbance mitigation for transformable lightweight structures based on elastic kinetics using active control. The dynamic behavior of such structures is transformation state dependent due to geometric nonlinearities, which is not easily captured by linear time-invariant methods. For this reason, modeling and control design are performed in the linear parameter-varying (LPV) framework. Thereby, two distinct LPV modeling methodologies are investigated, one based on finite element models and another utilizing experimental system identification techniques. Furthermore, LPV control strategies are introduced, which allow to adapt to the varying structural dynamics encountered in elastic kinetics. To validate the efficacy of the modeling and control approaches, a physical demonstrator resembling a segment of an elastic kinetic roof structure is realized within a laboratory environment. This demonstrator structure is used for experimental studies to assess the accuracy and performance of the introduced methods. Utilizing experimental data from the demonstrator structure, a proposed LPV system identification methodology can be applied to derive separate LPV models for the bending and the torsional dynamics of the demonstrator structure. Based on the derived LPV models, a decoupled active vibration control for bending and torsional dynamics by means of an LPV output-feedback is designed. The designed control is implemented on a real-time environment, and its effectiveness is demonstrated experimentally for fixed and for varying operating conditions on the realized demonstrator structure.