Holistic seismic design of clustered soil-foundation-structure systems using a large laminar box
Introduction
The 2010-2011 Canterbury earthquake sequence, the 2010 Maule Earthquake in Chile as well as the 2011 Tohoku earthquake in Japan, has shown that saturated loose sandy ground is extremely prone to liquefaction1-4. In the case of Canterbury earthquakes the scale of devastation caused by liquefaction, especially to shallow foundations of individual residential dwellings, was without parallel in NZ history (e.g. Figure 1). While the displacements at the surface of a sand deposit can be easily observed, the cause of such surface deformation lies underground. While a body of knowledge has been gathered over the past 40 years, the subsurface nature of the causes of the phenomenon has prevented a full understanding. Consequently, the intrinsic reasons for the damage to shallow foundations, surface-mounted structures and underground facilities, are hidden and are not well understood. Recent simulations of field shaking e.g. using T-Rex4 and explosives do not correctly simulate the liquefaction process because of the different nature of the ground excitations. In the current research at the University of Auckland Centre for Earthquake Engineering Research (UACEER), shake table tests using a laminar box and centrifuge tests will be performed to enable a more realistic simulation of earthquake loading and the resulting soil behaviour. The results obtained from the simulations will provide insights into the failure of structural foundations from soil liquefaction that has clearly been identified by the Royal Commission of Enquiry as needing urgent resolution5.
In conventional design practice earthquake-induced soil-foundation-structure interaction (SFSI) is hardly considered. If SFSI is considered at all, then usually linear soil behaviour is assumed. The influence of adjacent structures on the nonlinear seismic behaviour of a group of soil-foundation-structure systems is as good as ignored. Yet, the significant impact of nonlinear soil behaviour on the overall structural seismic performance prevails4, particularly in regions with poor soil, was observed in almost all major earthquakes. The need for an urgent resolution of building damage, due to nonlinear soil induced foundation failure, has been identified by the Royal Commission of Enquiry. Although it is known that the soil supporting one building and that supporting multiple buildings behave differently, the effect of multiple building foundation-soil systems has hardly been considered due to the complexity. An integrated consideration of a closely coupled soil-building system is virtually unknown. These deficiencies will be addressed in the research at UACEER. A clearer understanding of the whole system dynamics will lead to more accurate design simulation of the system response and thus improved economic and resilient aseismic design.
Figure 1. Impact of liquefied soil on structures
Simulation of liquefiable sand
To simulate the impact of liquefiable sand on structures a laminar box6, 7 with dimensions of 2 m x 2 m x 2 m was built at UACEER (Figure 2). Each laminar layer can move up to 175 mm horizontally. The sand inside the box can undergo a maximum shear strain of approximately 9%, which is useful to simulate the anticipated displacement in a large earthquake event. Figure 3 (top to bottom) shows the pipe system at the base of the box, the waterproof silicone rubber membrane inside the box and the sand raining procedure aimed at achieving a uniform sand distribution. The box is waterproof such that saturated sand can be contained without leakage while undergoing large shear strain.
Figure 2. A large laminar box at UACEER
To study the subsurface movements smart particles utilised in an existing EQC project will be further developed and improved by reducing the size and hence the differential inertial loading. These particles have the capability to detect three dimensional movements of subsurface soil in both the pre-liquefaction and liquefied states. The goal is to have an even smaller smart particle to enable a more undifferentiated unification of the sensor with the surrounding sand.
Figure 3. Pipe system, box interior and sand raining
Simulation of multiple SFSI systems
The laminar box is placed on the shake table and the structural models are placed on the sand surface. Adjacent structures can have different slenderness and fundamental frequencies. Figure 4 shows one possible set up in the laminar box with a simple frame structure with a surface footing.
Figure 4. SFSI under earthquake loadings
Summary
A much improved understanding of the subsurface liquefaction process, including the influence of superstructure dynamics on liquefiable soil in conjunction with ground excitation characteristics can be achieved using the new facilities. This will enable a more realistic estimate of where and how liquefaction risk can be avoided or mitigated.
The understanding of subsurface soil and foundation movement will enable the development of suitable ground improve-ment techniques and improved holistic design of shallow foundations and pile foundations.
The understanding of shallow foundation response, incorporating nonlinear soil properties and the group effects of the dynamics of superstructures, will enable a better assessment of foundation displacement and an estimate of the displacement threshold when significant soil yielding commences.
Acknowledgements
The authors would like to thank the Natural Hazard Research Platform for the support of this research entitled “Impact of liquefiable soil on the behaviour of coupled soil-foundation-structure systems in strong earthquakes” under the Award 3708936.
References
Orense, R.P., Kiyota, T., Yamada, S., Cubrinovski, M., Hosono, Y., Okamura, M., Yasuda, S. (2011) “Comparative study of liquefaction features observed during the 2010 and 2011 Canterbury earthquakes,” Seismological Research Letters 82(6): 905-918
Orense, R., Pender, M., Wotherspoon, L. (2012) “Analysis of soil liquefaction during the recent Canterbury (New Zealand) earthquakes,” Geotechnical Engineering Journal of the SEAGS & AGSSEA 43(2): 8-17
Orense, R., Yamada, S., Otsubo, M. (2012) “Soil liquefaction in Tokyo Bay area due to the 2011 Tohoku (Japan) Earthquake,” Bulletin of the New Zealand Society for Earthquake Engineering 45(1): 15-22
Orense, R., Towhata, I., Chouw, N. (Eds.) Proceedings of the New Zealand – Japan Workshop on Soil Liquefaction during Recent Large-Scale Earthquakes, December 2-3, 2013, the University of Auckland
Royal Commission, Summary and Recommendations, ISBN: 978-0-478-39558-7 http://canterbury.royalcommission.govt.nz/Commission-Reports
Cheung, W.M., Qin, X., Chouw, N., Larkin, T., Orense, R. (2013) “Experimental and numerical study of soil response in a laminar box,” New Zealand Society for Earthquake Engineering Conference, Wellington, 26-28 April, p16
Qin, X., W.M. Cheung, Chouw, N., Larkin, T. (2013) “Study of soil-structure interaction effect on ground movement using a laminar box,” New Zealand Society for Earthquake Engineering Conference, 26-28 April, p65
