Analysis of piled bridges at sites prone to liquefaction and lateral spreading in New Zealand – NZ Transport Agency Report

INTRODUCTION

Early this year, New Zealand Transport Agency (NZTA) has published the NZTA Stage 2 Research Report titled “Analysis of piled bridges at sites prone to liquefaction and lateral spreading in New Zealand”. The report is available on NZTA’s website at the following address:
http://www.nzta.govt.nz/resources/analysis-of-piled-bridges-at-sites-prone-to-liquefaction-and-lateral-spreading-in-new-zealand/

It follows the first stage of the project summarised in the Stage 1 Research Report titled “The development of design guidance for bridges in New Zealand for liquefaction and lateral spreading effects”, NZTA Report 553, which is also available on NZTA’s website:
http://www.nzta.govt.nz/resources/research/reports/553

The NZTA research project included a review of seismic behaviour of bridges on sites prone to liquefaction and lateral spreading in New Zealand and overseas, and reviewing available design methods for bridges against liquefaction and lateral spreading effects. The latter involved detailed consideration of the bridge design framework in New Zealand and developing an appropriate design methodology for New Zealand conditions.
The overall purpose of the project was to identify a clear set of available procedures for analysis and design that are based on observed seismic behaviour of bridges (reviewed case studies) and the most recent research findings from across the world. The outcomes of the NZTA project were presented in two reports:

• Stage 1 Research Report that summarises these methods and provides references to supporting materials (where such references are available).
• Stage 2 Research Report that presents a summary of an example application of pseudo-static analysis in the design of bridges on sites prone to liquefaction. The report gives detailed procedures for geotechnical field and laboratory testing, liquefaction evaluation methods and examples for two bridge sites, detailed description of recommended procedures for pseudo-static analysis (including flow charts for the design process), an overview of dynamic analysis methods, and presents detailed design examples for two bridges.

The design requirements and guidelines given in the Stage 2 Research Report are to be incorporated in the NZTA Bridge Manual and disseminated to the wider New Zealand engineering community.

The project team was led by WSP Opus and comprised WSP Opus researchers (Dr Alexei Murashev, Messrs Campbell Keepa, Gopal Adhikari, Dejan Novakov and Donald Kirkcaldie), the University of Canterbury (Prof Misko Cubrinovski and Dr Jennifer Haskell) and the University of Auckland (Prof Rolando Orense).

A brief summary of the Stage 2 report is given below.

STAGE 2 RESEARCH REPORT

In the Stage 2 of the NZTA research project, methods for the evaluation of bridges on liquefiable sites were assessed and a detailed pseudo-static procedure was developed for the analysis and seismic assessment of piled bridges in New Zealand. The Stage 2 report summarises this procedure and presents two examples of its application (Tauherenikau Bridge and Belfast Road Underpass). Also included in the Stage 2 report are two examples of the evaluation of liquefaction and lateral spreading, where guidance on the scoping of site investigations at bridge sites with liquefaction susceptible geologies is presented (Anzac Bridge and Belfast Road Underpass).

Liquefaction evaluation examples

Example 1: Tauherenikau Bridge

The first liquefaction site evaluation example is for the Tauherenikau River Bridge near Featherston. The 127 m long eight-span concrete bridge is located on deep gravelly alluvium located about 10 km from both the Wellington Fault and the Wairarapa Fault that have recurrence intervals of 715 to 1,575 years and 1,200 years, respectively. These faults are capable of producing Mw7.5 and Mw8.2 earthquakes. A simple deterministic assessment of ground motions proved useful in the assessment of seismicity at this site where there are known active faults nearby. Comparison of the Bridge manual’s 1,000-year return ground motions (peak ground acceleration (PGA) = 0.45 g, Mw7.0) to ground motion predictions from rupture of the Wairarapa and Wellington Faults suggests the seismic hazard could be under-represented by the Bridge manual at this location.

The importance of understanding a) whether the gravels are clast supported and b) the characteristics of the supporting soil matrix in the liquefaction evaluation of gravelly soils is emphasised in the Tauherenikau Bridge example. Challenges with assessing the density of gravelly soils, and therefore their liquefaction potential using penetration testing techniques, are addressed with the supplementation of a triggering evaluation using shear wave velocities and the adjustment of standard penetration tests using blow counts measured in 25 mm increments.

Example 2: Belfast Road Underpass

The second example of a liquefaction site evaluation is for a site just north of Christchurch where a new underpass is to be constructed. The two-lane, two-span reinforced concrete bridge will take Belfast Road over the new at grade Christchurch Northern Motorway. The bridge has approach embankments approximately 8 m high at the abutments with spill through slopes and a central pier.

The site is underlain by variable thinly bedded silty sands, sands and soft organic silts that overlie dense sands and gravels from a depth of about 9.5 m. Little surface manifestation of liquefaction was observed at this site in the Darfield 2010 or Christchurch 2011 earthquakes, yet analysis solely using cone penetrometer tests (CPTs) to assess liquefaction susceptibility suggested extensive liquefaction near the surface. This example demonstrates the importance of assessing the susceptibility of silty soils using measurement and observations of their plasticity and comparing this with the soil behaviour index calculated from the CPT data.

Lateral spread displacements are evaluated using empirical methods and adjusted to account for the limited width of the embankment and the stabilising effect of the continuous crust between the two approach embankments. A simplified method for the prediction of seismic horizontal ground displacements of non-liquefied soils is implemented in the calculation of displacement profiles for each phase of the earthquake.

Examples of the pseudo-static analysis of piled bridges

Example 1: ANZAC Bridge, Christchurch

The first pseudo-static bridge analysis example is of the existing four-span ANZAC Bridge that crosses the Avon River in Christchurch. This bridge is situated on loose to medium dense sands and was damaged by liquefaction and lateral spreading in both the 2010 Darfield and the 2011 Christchurch earthquakes. The response of the bridge in the lateral spreading phase following the February 2011 Mw6.2 Christchurch earthquake has been analysed with surface ground displacements measured at the site that were used to develop ground displacement profiles with depth at the pile locations. Ground motions were derived from nearby strong motion stations.

Comparison of the model predictions of displacement and bending to observations following the earthquake, demonstrates that the method captures the key effects such as:

• strutting of bridge deck constraining horizontal ground displacement at the end of the deck;
• backward rotation of the abutment back-walls about the deck;
• bending of the abutment piles as the river banks spread toward the middle of the river.

Kinematic demands on the pier piles and rotation of the pier columns, consistent with the observed deformation modes, is also captured in the whole bridge analysis.
The predicted back-wall rotation was greater in the analysis than the observed amount (approximately double) and the level of bending predicted at the top of the piers is inconsistent with the degree of cracking and spalling observed at the top of the pier columns. This suggests that while the method predicts the deformation mechanism, it may overestimate the level of damage. Although in this case some of this apparent overestimation could be attributed to simplifications made in the model, particularly the degree of fixity at the connections and possibly with transforming the 3D bridge into a 2D model.

Example 2: Belfast Road Underpass, Christchurch

The second pseudo-static bridge analysis example was also undertaken for the Belfast Road Underpass bridge. In this example, a method was developed to implement the pseudo-static analysis procedure in the design of a new bridge.

Being a transient problem, the peak ground displacements at different supports of a bridge may or may not occur concurrently and may not be in the same direction as the inertial demands. Different combinations of inertia and kinematic demands for each of the three phases of the ground response are evaluated in the design of the bridge. For the specimen bridge, the lateral spreading phase proved to be more critical for the design of the abutment piles, whereas the cyclic liquefaction phase was more critical for the design of the pier piles.

The relative merits of a single pile versus a whole bridge analysis was assessed. The major downfall of a single pile analysis is knowing the conditions (load and displacement) applied at the head of the pile by the superstructure. For the specimen bridge in this example, where ground conditions are similar across the site and there is a relatively symmetric distribution of stiffness and mass, analysis of a single pile gave similar results to the global analysis for the lateral spreading phase, which greatly reduced the computational effort required for this type of analysis.

Inertia demands from the superstructure have been applied either as a force or as displacement. It is difficult to confirm which approach is more realistic. However, for the single pile model, imposing displacement makes more sense, as it provides the response that is comparable with the whole bridge model. Where inertial demands are applied as a force, an appropriate boundary condition should be assigned at the superstructure level.

CONCLUSIONS

The Stage 1 and Stage 2 Reports provide a useful design framework for bridges on liquefiable ground.

The Stage 2 Report provides practical examples of the use of the recommended pseudo-static analysis framework for the design of typical bridges on sites susceptible to liquefaction and lateral spreading in New Zealand.

The performance predictions of the pseudo-static analysis method are generally sensitive to the predicted extent of liquefaction and the horizontal ground displacements applied to the soil springs in the analysis. As shown in the examples, ground displacement predictions have a high degree of uncertainty. This is one area where further research is needed to improve the accuracy of the performance predictions using the pseudo-static method.

Other aspects of the analysis procedure requiring further refinement are:

• the evaluation of liquefaction resistance of pumice soils, gravelly soils and sites with thinly interbedded layers;
• the assessment of ground displacement profiles at bridge piles for sites with improved ground
• whether the pinning effect of the piles is sufficiently accounted for within the pseudo-static method;
• for bridges with approach embankments, how the stiffness of the bridge affects the direction of spreading;
• assessing bridge response with kinematic and inertia loading in the transverse and longitudinal directions together.

The examples highlight some of the challenges when applying the recommended methods. There is a high degree of uncertainty associated with the assessment of liquefaction hazards and the response of bridges to lateral spreading. Built-in conservatism at each step of the analysis can be applied to manage this uncertainty, but this is not always transparent. Furthermore, the compounding effect of such conservatism can result in overly conservative assessments and designs.

One key item emphasised in this study is the need for parametric studies and sound engineering judgement to envelope the response and gain a good understanding of the likely seismic performance of the bridge from which
to make appropriate design decisions.

ACKNOWLEDGEMENTS

The funding for the project was provided by the New Zealand Transport Agency (NZTA). Dr John Wood of John Wood Consulting is thanked for his detailed review of the Stage 2 Research Report. Messrs Nigel Lloyd, John Reynolds and Barry Wright of NZTA are gratefully acknowledged for their valuable comments, advice and support provided to the research team through the course of the project.