Numerical Modelling of Tuned Liquid Dampers Using Implicit Smoothed Particle Hydrodynamics

Abstract

Tuned liquid dampers (TLDs) are one of the most common systems used to control the resonant response of buildings due to their simplicity and affordability. A TLD comprises a partially water-filled tank, which can be of different shapes, installed near the top of the building and tuned to the natural frequency of the building. Typically, the inherent damping of the TLD is improved by adding additional damping devices, such as screens. Studying the nonlinear flow of TLDs is imperative for designers in order to understand their response, and numerical modelling is essential for their effective design. Existing numerical models are typically restricted to a range of liquid depths, excitation amplitudes, tank-bottom geometries, and screen implementation configurations or require significant computational time and resources. Motivated by designer needs and existing limitations described above, this research aims to develop a computationally efficient numerical model to simulate TLDs equipped with screens without the current restrictions. The model is based on solving the free-surface flow of the TLD using the mesh-free Smoothed Particle Hydrodynamics (SPH) method. The model is complemented by a novel macroscopic screen model, which allows for larger computational resolution and a significant reduction in computational time compared to explicitly modelling the screens. Model results are validated using a wide range of experimental data, with a good agreement observed. The model is expanded to include tanks with irregular bottom geometries using an efficient particle-generating algorithm, and their response is studied under large harmonic excitation amplitudes. Finally, the model is used to investigate a realistic situation of a dual-function tank coupled to a structure to study its response under random excitation. It is found that the model efficiently captured the response of the structure under a range of excitation amplitudes using reasonable computational time and resources.