Development of Servo-Hydraulic Earthquake Actuators for use on Geotechnical Centrifuges

Earthquakes inflict severe damage to infrastructure. While UK is relatively safe from earthquakes the UK based Construction Industry operates worldwide often performing designs in earthquake prone areas including South-East Asia, Taiwan, Indian sub-continent and other countries. Also the UK based Insurance industry and Risk Management industries often need predictions on the performance of engineered infrastructure. In recent years the use of dynamic centrifuge modelling to investigate various problems in the earthquake geotechnical engineering area has established itself as an attractive tool both to understand the fundamental mechanics at work as well as establishing the performance of a given earthquake resistant design. Worldwide the number of centrifuge centres that are able to carryout earthquake testing on centrifuge models is on the increase with University of California, Davis, RPI in New York and Hong Kong University of Science and Technology upgrading their earthquake actuators and LCPC in Nantes, France and Korean Hydraulic Research Labs recently adding earthquake actuators to their centrifuge facilities. Earthquake modelling on centrifuges started at Cambridge in late 1970's with the famous 'bumpy road' system coming into operation in 1979. In 1994 a Stored Angular Momentum (SAM) based earthquake actuator was developed which could impart strong earthquakes at single tone bursts. This actuator was very successful at Cambridge leading to significant amount of research both fundamental research looking at liquefaction phenomena of soils as well as investigating specific boundary value problems and produced 10 PhD's and significant number of journal and conference publications. The SAM earthquake actuator system though very successful needs to upgraded. The real need in our research capabilities at this time is to have the ability to produce multi-frequency earthquake motions which are a more realisitc representations of the real earthquakes recorded in the field. The analytical ablities in Cambridge have been significantly boosted by the development of Wavelet analyses methods that allow us to analyse multi-frequency inputs in time-frequency space. This ability gives us the distinct advantage to follow the dynamic respone of complex non-linear soil-structure systems. These theorectical developments and the needs of the current research community and the UK based practicising engineers all confluence into the requirement of a servo-hydraulic earthquake actuation system. The medium term aim at Cambridge is to develop a powerful 2-D servo-hydraulic earthquake actuator that is able to generate horizontal and vertical shaking of centrifuge models. However such a system is quite complex and it is considered prudent to approach this development in stages. This is necessary for us to gain experience with fast acting servo-hydraulic valves. This proposal aims to develop a 1-D servo-hydraulic earthquake actuator as the first stage in the development of a 2-D earthquake actuator. This will help us gain the necessary experience in the usage of fast acting servo-valves in the high gravity field, carrying out design and implementation of the appurtant systems such as hydraulic powerpacks, control systems, system safety devices. These systems need to be incorporated into rather specialist requirements of the Cambridge Centrifuge facility which has a limit of 1 ton on the payload (shaking systems + centrifuge model) and the special swing up and lock mechanisms. The project will have a definite deliverable in the form of 1-D earthquake actuator with a shaking force of about 7 tons and will be used in its own right to trigger powerful earthquakes and gives us the ability to model realistic earthquake loading.