Analysis and design of piles in liquefying soils

Abstract

During the intense shaking that accompanies large earthquakes groundwater pressures can raise causing soil to lose some of its strength and capacity to support structures resting on it. In the extreme case, the soil may lose its strength completely and the resulting “soil liquefaction” may cause severe damage to buildings, bridges and other engineering structures. Pile foundations are especially vulnerable to liquefaction since they are used to support structures near (in) river beds, in reclaimed land and coastal areas that are susceptible to liquefaction. This issue is very relevant for New Zealand since liquefaction is recognized as one of the principal seismic hazards affecting urban centres as well as critical lifelines and infrastructure across the country.

The processes of soil liquefaction and “soil-pile interaction” during earthquakes are extremely complex. Hence, the design of piles against earthquake loads and soil liquefaction is a very difficult task. There are several methods available to engineers for seismic design and analysis of piles. The most attractive approach for the profession is the “pseudo-static analysis”, because it is relatively easy to understand and use despite the complexity of the processes that are considered in the analysis. The pseudo static method of analysis is routinely used in the engineering practice and is commonly stipulated in modern seismic design codes.

One of the key issues in the application of the pseudo-static analysis arises from the uncertainties associated with liquefaction and unknowns in the analysis. This is not surprising in view of the fact that a very simple (static) method is used as a basis for modelling very complex (dynamic) problem. Hence, it is difficult for the designer to figure out how to model the complex processes with the simple analysis, and to know whether the adopted assumptions are on the safe side or not. Very little guidance exists in the profession in this regard.

The research presented in this report aims at providing clear guidance how to use the pseudo-static method for analysis of piles in liquefying soils. It shows which model parameters are the most important in the analysis and provides analysis strategy to the designer. The study is based on observations of the performance of pile foundations in recent strong earthquakes, benchmark experimental studies and comprehensive analytical studies conducted over the past ten years. It aims at developing simple yet effective procedure for analysis and design of piles in liquefying soils, considering specific New Zealand conditions.

---

Technical Abstract

The pseudo-static method of analysis (PSA) is a simplified design-oriented approach for analysis of seismic problems based on routine computations and conventional engineering models. The application of the method to analysis of piles in liquefying soils is burdened by significant uncertainties associated with soil liquefaction, soil-pile interaction in liquefying soils and the need to reduce a very complex dynamic problem to a simple equivalent static analogy. Hence, despite its simplicity, the application of the pseudo-static analysis is not straightforward and requires careful consideration of the uncertainties in the analysis. This study addresses some of the key issues that arise in the application of the pseudo-static analysis to piles in liquefying soils, and makes progress towards the development of a clear modelling (analysis) strategy that permits a consistent and reliable use of the simplified pseudo-static analysis.

Characteristics of liquefying soils and loads on piles are significantly different during the cyclic phase (strong ground shaking) and in the subsequent lateral spreading phase, and therefore, it is necessary to separately consider these two phases in the simplified analysis of piles. This paper describes a practical PSA procedure for preliminary assessment and design of piles, and addresses key parameters and uncertainties in the analysis, both for the cyclic phase and the lateral spreading phase of the pile response.

A comprehensive parametric study was conducted in which a wide range of soil-pile systems, loading conditions and values for model parameters in the PSA were considered, for piles in laterally spreading soils. Results from the analyses were used to examine and quantify the sensitivity of the pile response to various model parameters, and to establish a fundamental link between the sensitivity of the pile response and the mechanism of soil-pile interaction. On this basis, some general principles for conducting pseudo-static analysis of piles in liquefying soils could be established that apply across-board to different soil-pile systems and loading conditions.

In the simplified pseudo-static analysis of piles, the ultimate lateral pressure from the liquefied soil is commonly approximated based on the residual strength of liquefied soils. This strength does not have sound theoretical basis, but rather is estimated from one of several empirical relationships between the residual strength and penetration resistance. The two empirical relationships adopted in this study, even though originating from the same database, result in substantially different strength profiles (ultimate lateral pressures on the pile) throughout the depth of the liquefied layer. Series of analyses were conducted to investigate the effects of strength normalisation on the pile response predicted by the pseudo-static analysis. It is shown that effects of strength normalisation can be quite significant and that they depend on the relative stiffness of the pile and the thickness of a non-liquefiable crust at the ground surface.

The cyclic study comprises two distinct phases, the first considering the response of piles when subjected only to cyclic soil displacements, and the second considering the pile response when both cyclic soil displacements and superstructure inertial forces are present and acting simultaneously. In this comprehensive series of analyses emphasis was placed on understanding the governing mechanisms and controlling-parameters when simultaneously considering the combined inertial loads from the superstructure and kinematic loads due to lateral ground movements in PSA.

Finally, three different approaches for assessment of seismic performance of piles in liquefying soils comparatively examined. These approaches use different models, analysis procedures and are of vastly different complexity. All three methods are consistent with the performance-based design philosophy according to which the seismic performance is assessed using deformational criteria and associated damage. Even though the methods nominally have the same objective, it is shown that they focus on different aspects in the assessment and provide alternative performance measures. In this context, key features of the PSA and its unique contribution in the assessment of pile foundations in liquefying soils is discussed.