Report on centrifuge test techniques with Soil Structure Interaction

Soil-structure interaction plays an important role in the dynamic response of many types of structures. In the SERIES project the dynamic soil structure interaction is investigated for a single degree of freedom structure. This is an interesting problem where the dynamic response of the structure combines with the lifting-off of the base during rocking vibrations. Two types of single degree of freedom structures were considered, one with a lumped mass at the top and the other with a sway-frame. Several centrifuge tests were carried out changing the structures and soil densities. Again there was some overlap between the test specifications. The structures tested at UCAM had bearing pressures of 50 kPa and 100 kPa, while those tested at IFSTTAR had 100 kPa and 300 kPa. Similarly the stiffness of structure was changed in different centrifuge tests. The construction of the model structures was matched as closely as possible between the two centres. The dynamic response of the structures subjected to model earthquake loading was measured in several different ways. Some techniques of measurement were common between UCAM and IFSTTAR, while others such as advanced dynamic PIV measurements based on fast digital imaging were carried out at UCAM only. This technique is able give the soil displacement vectors at different points within an earthquake loading cycle. In all centrifuge tests, the accelerations within the soil and at different locations on the structure were measured using miniature accelerometers. The acceleration data was used to obtain the moment-rotation loops for the structure for different earthquakes. These were compared for structures of different stiffness, foundation soil placed at different relative densities and earthquakes of different intensity and frequency content. In this report details are given of the general theory of centrifuge modelling. This is followed by the experimental facilities, centrifuge model containers, instrumentation and data acquisition used at UCAM and IFSTTAR facilities. The details of the design of different model structures are also presented. Following this, the model preparation techniques including the air-pluviation devices used are presented. The technique of PIV analysis using digital images from a fast digital camera is introduced. Following this the experimental data obtained in each centrifuge test such as acceleration time histories, settlement time histories and the results of the analyses of the data in the form of moment-rotation loops are presented. Results from some of the selected PIV analysis are 2 also shown. Finally a brief summary of the research carried out in JRA 3.2 work package and the main outcomes are presented.