Wireless instrumentation and metadata for geotechnical centrifuges

Abstract

Centrifuges are used in geotechnical earthquake engineering to simulate complicated phenomena such as nonlinear soil-structure interaction and liquefaction induced by earthquakes. Cost efficient alternatives to large scale testing, centrifuge modeling has become widely accepted to analyze experimentally all kinds of geotechnical problems. This thesis explores the use of wireless technologies to improve data transfer and control remotely data acquisition systems of centrifuges. These improvements greatly reduce electrical noises and increase the high-frequency performance of data acquisition. This thesis also contributes to the introduction of metadata for documenting the experimental results and processes of centrifuge testing. The first part of the research work consisted of re-assembling the centrifuge of the University of Southern California, modifying the mechanical centrifuge balancing system, replacing outdated parts with more modern ones, and upgrading the hydraulic shaker that simulate in-flight earthquake shakings. The second part focused on the high-speed wireless instrumentation system that overcomes the limitations imposed by traditional slip ring connections. The new centrifuge system, including renovated mechanical parts and new wireless instrumentation, was tested by carrying out a series of dynamic tests on a laminar box, which simulates the responses of saturate soil deposits to earthquake motions. The experimental results indicate an overall good performance of the complete centrifuge system. The third part of the research work deals with the application of novel information technology methods for documenting experimental results as well as experimental processes. Metadata are introduced to describe all experimental steps from the construction of the soil specimen to the final reports. The metadata model, which is proposed for centrifuge modeling, explains the data collected during the experiment through an additional layer of data called "metadata." These metadata, also referred to as data about data, are totally new to centrifuge testing and are likely to become critical for describing increasingly complicated centrifuge experiments. The proposed metadata model, which is applied to generate web reports and computer visualization, leads to powerful ways of exchanging information among researchers.