John B. Berrill

LIQUEFACTION AND PILED FOUNDATIONS: SOME ISSUES

Selected case histories recording the behaviour of piled foundations in lateral-spreading fields are reviewed. From these observations and from laboratory experiments in the centrifuge and on the shaking table, it is clear that the resistance of liquefied soil is very small and that the critical points for a pile are at the bottom as well as near the top of the liquefied layer. It is also apparent that liquefaction does not-occur until a significant threshold level of shaking is exceeded, and that even then it does not occur instantaneously. Thus damaging motions can be felt by the pile and transmitted to the superstructure before liquefaction and lateral spreading occurs. There is a critical period early in the shaking when the soil has softened enough for ground motion to be amplified but is still stiff enough to induce large bending moments in the pile, near the top of the liquefiable layer. Liquefaction is certainly not a dependable base-isolation mechanism. A controversial issue concerns the continuity or localisation of displacement within the liquefied layer. There is evidence that, in the majority of cases, displacement within the liquefied layer is continuous with depth; although when a sufficiently impermeable layer overlies the liquefiable one, a layer of free water can accumulate and localise displacement. Another old controversy, which now seems to be resolved, concerns the generation of passive earth pressures in non-liquefiable surface layers. There is clear field and laboratory evidence for the development of the passive state in superficial layers. For use in everyrday design, two simple methods of analysis have recently emerged. They are the earth pressure method and the seismic deformation method. They have been calibrated against the Kobe data and appear to work well in the few cases where they have been checked on other soils. It remains for them to be verified against a broader range of soils and for further soil-spring models.

Concerning Baseline Errors in the Form of Acceleration Transients When Recovering Displacements from Strong Motion Records Using the Undecimated Wavelet Transform

Abstract This paper discusses the progression of a novel algorithm that uses a wavelet‐transform approach. The transform is a generalization of the decimated, discrete wavelet transform (DWT) that is the undecimated DWT or stationary wavelet transform (SWT) also known as the undecimated a trous algorithm. It forms the basis for recovering displacements from acceleration time histories. The approach recovers a low‐frequency fling that is usually an almost sinusoidal or cosinusoidal pulse responsible for the big ground motions in strong motion events. The algorithm implements a well known and non‐linear, denoising scheme and is applied to the low‐frequency sub‐band and, in particular, succeeds in recovering the acceleration‐fling pulse. The progression is that in order to obtain estimates of displacements, the algorithmic baseline‐correction scheme can now locate an acceleration transient (i.e., a spike), which creates the DC shift in velocity and the linear trend in displacement, and is therefore the baseline error. Once this acceleration transient is corrected for or eliminated, double‐time reintegration recovers the velocity‐fling pulse and residual displacement. The paper infers that these acceleration transients may be due to ground rotation, embedded in the translational data. The scheme provides for easier integration once the low‐ and higher‐frequency accelerations are extracted. Online Material: Additional results for the Chi‐Chi TCU068 (1999) station, the New Zealand Darfield Station (2010), and the Olfus Earthquake (2008) in Iceland.

Rational approximation of stress and strain based on downhole acceleration measurements

Recordings from downhole accelerometer arrays offer unique insight into soil behavior and ground response during earthquakes. In this paper we present a scheme for interpolating displacement and acceleration measurements to provide approximations for subsurface shear strain and stress as continuous functions of time. Our suggested interpolating functions are constructed in such a way that the free surface boundary condition will always be satisfied and the interpolated displacement and acceleration remain finite for all depths. We also show how the functions can be adapted to represent layered soil profiles. Depending on the number of instruments in the downhole array, a truncated series of functions can be derived so that each represents a modal shape for the layered soil profile. The resulting approximations for strain and stress are considered more accurate and robust than previous approximations.

A pattern recognition approach to evaluation of soil liquefaction potential using piezocone data

Abstract A pattern recognition approach to liquefacation evaluation is propoesed. The state of any soil layer at a level ground site subject to seismic loads is represented by a pattern in a seven-dimensional feature space and can be classified into one of three classes: liquefiable cohesive soil, and non-liquefiable cohesionless soil. The liquefaction potential of the soil layer can be assessed according to the probabilities of the pattern belonging to the three classes. Training patterns derived from field data (piezocone (CPTU) data and maximum ground acceleration) from sites which liquefied or did not liquefy during earthquakes in New Zealand are randomly chosen to design a pattern recognition system to provide an optimal estimation of the liquefaction potential of any soil stratum of interest. Two recognition systems have been set up to estimate the state-conditional probability density function. One is based on a Parzen window approach in which no knowledge of the probabilistic structure of the training patterns is assumed; the other is based on a parameter estimation approach assuming a multivariate normal distribution. The error rate of recognition by the Parzen window approach is 6·9% when taking the window size as 1·5, and the error rate by the parameter estimation approach, which can be easily, is 7·7%. implemented without reference to our training patterns