PROBABILISTIC MODELS FOR PASSIVE PEOPLE IN HUMAN-STRUCTURE INTERACTION PROBLEMS

Abstract eng:
In the last two decades the models for human loading in structural systems have considerably changed. Human loading is no longer considered a simple mass as the full interaction between structures and humans alters the dynamic characteristics of the overall system. The most well known case of a human-structure interaction (HSI) problem occurred in England in 2000 when the Millenium Bridge showed extreme lateral vibrations after being open to traffic. The human in HSI problems has been mostly modeled in two ways: human as a mass model and human as a lumped masses-dampers-springs (MDS) system. Mass models are commonly used where the interaction between human and structure is minimal. MDS systems are used in cases where the expected interaction between the two systems is larger and more complex models are needed. Recently Ortiz et al (2014) proposed the use of models based in closed loop control systems to model the human. Feedback models follow the concept that the human body reacts to the motion of the structure and it could even add energy to the system. The implementations of the closed loop systems was performed using only one person with a straight knees position. In this paper traditional and controller based models are updated following a probabilistic approach based on Bayesian analysis. Experimental tests involving one, two and three people are performed using free vibration tests in a simple cantilever structure. Experimental transfer functions are used to update the models. The advantages and disadvantages of each type of model are discussed.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket