Keynote Address Session 1 Session 2 Session 3 Session 4 Session 5 Session 6 Poster Session Session 8 

A Ground Shaking Amplification Map for New Zealand

Umut Destegul, Grant Dellow and David Heron

A ground shaking amplification map of New Zealand has been compiled from data held by GNS Science. The resulting map is being used in RiskScape, a tool for comparing risks at a given site from a variety of hazards by estimating potential losses.

A GIS-based geological map with national coverage has been compiled from several sources, and is used as the base data. Geological maps at a scale of 1:250,000 from the QMAP project which is the geological mapping of New Zealand have been used where available supplemented with detailed geological maps at scales ranging from 1:25,000 to 1:50,000 for the larger urban areas. The gaps have been filled by the 1:1,000,000 ‘Geological Map of New Zealand’.

Every geological polygon in the composite geological map has been assigned one of the ground shaking amplification classes from the New Zealand Standard for Structural Design Actions NZS1170.5 to produce the product map. The classes conform to the site class definitions in NZS1170.5, which describes five classes with respect to ground shaking amplification. Assignment of these classes was straightforward for rock sites but more involved for soils where, for example, at boundaries between weak rock and deep soil sites a buffer zone of shallow soil was applied.

The product map was checked by comparing it against 687 sites in a database of accelerograph locations where the ground class had been determined from site-specific information. Currently, for 72% of the sites the NZS1170.5 site class is the same for both the site specific data and the site class assigned to the geological polygon.

Paper P41: [Read] [Presentation]

Design of Retaining Walls for Outward Displacement in Earthquakes

John Wood

It is often not practical to design major retaining walls and bridge abutment walls to resist the peak ground accelerations expected in the design level earthquake in regions of high seismicity. The current Transit NZ Bridge Manual requires structures on important routes to be designed for a 2,500 year return period event and this leads to high design level accelerations even in areas of moderate seismicity. A satisfactory design approach is to design for a resistance level less than the peak ground acceleration and accept some outward movement. Outward movements can be computed using the Newmark sliding block method. However, published procedures for applying the Newmark method to walls have been numerous and there are variations in analysis complexities including whether or not to allow for vertical acceleration effects. It is not clear which approach is best for design applications and what values of peak ground velocity, required as one of the input parameters, should be used to be consistent with the provisions of NZS 1170.5 and the Transit NZ Bridge Manual.

The paper compares the most widely used of the numerical and empirical methods of applying the Newmark method including Matthewson et al (1980), Ambraseys and Menu (1988), Cai and Bathurst (1995), Bray and Travasarou (2007), and Jibson(2007). Recommendations are made as to the best approach for wall design applications and charts are presented that enable outward displacements to be rapidly estimated when using the Transit Bridge Manual Provisions and NZS 1170.5.

Paper P12: [Read] [Presentation]

Nonlinear Foundation Response of Liquid Storage Tanks under Seismic Loading

Michael Chung and Tam Larkin

This paper presents a brief review of liquid storage tank performance under earthquake loading. The dynamic properties of a tank system are also addressed. Emphasis is placed on the development of a nonlinear analytical model in the time domain. The model captures the important aspects of foundation compliance, including both foundation uplift and soil yielding beneath the tank base. Experimental and analytical work performed by Bartlett (1976) provides the framework for this work. Results predict that the response of tanks under lateral loading may be significantly influenced by soil structure interaction effects. These effects may increase or decrease seismic loading depending on the individual circumstances of the tank and foundation.

Paper P43: [Read]

Revised NZSEE Recommendations for Seismic Design of Storage Tanks

David Whittaker and Dean Saunders

The 1986 NZSEE document Recommendations for Seismic Design of Storage Tanks has been updated and will be republished during 2008. The original publication has been widely used and acknowledged internationally. Since 1986 there have been substantial changes to legislation and applicable standards in New Zealand and internationally. The design basis used in the 1986 document assumed no yield or damage being permitted to tanks under the design earthquake loading, and therefore led to some conservatism in the design of large steel storage tanks. The 2008 document bases design seismic loads for tanks on the recently issued national Standard for loads for buildings in New Zealand, NZS 1170.5. Design seismic loads are based on the ductility and damping applicable to tank behaviour. Modest levels of ductility (or force reduction) are permitted for steel tanks on grade, which generally reduces the load demands from those given previously. Benchmarking comparisons of the revised approach against the previous document and other relevant codes suggest that the proposed approach is reasonable. NZSEE is aiming to have this document recognised as a code of practice in New Zealand. The document has been structured in a way that it could also be used with other international codes.

Paper P04: [Read] [Presentation]

Earthquake performance and permanent displacements of shallow foundations

Jeremy Toh and Michael Pender

The standard approach to seismic design of shallow foundations is equivalent to ensuring that the bearing strength factor of safety does not fall below a certain value. However, brief instances of bearing failure (yielding) during an earthquake may not necessarily be serious. A more important consideration will be the residual foundation displacements accumulated at the end of the earthquake. Macro-elements provide a simple way of capturing the essential features of soil-foundation interaction, including the residual foundation displacements, and are readily amenable to routine structural analysis.

The shallow foundation macro-element examined in this study is based on existing macro-elements. It accounts for both the elastic (small displacement) and plastic (yielding) phases of dynamic behaviour. Modelling with the macro-element produces important insights into shallow foundation performance. Permanent vertical displacement (settlement) of the foundation is predicted to accumulate only during prolonged yielding, with the magnitude of residual settlement being dependent on the earthquake magnitude and duration, and the static vertical factor of safety against bearing failure. Residual rotation and translation of the foundation is inferred to be dependent on the characteristics of an individual earthquake.

Paper P40: [Read] [Presentation]

Preliminary analysis on critical factors for restoration of water distribution pipelines in the Hutt City after a magnitude 7.5 earthquake from the Wellington fault

John Zhao, Jim Cousins, Biljana Lukovic and Warwick Smith

The water distribution network for Hutt City consists of nearly 700 km of pipes, 6500 valves, 4000 hydrants, and 21 reservoirs located on the surrounding hills. The bulk water is pumped to hillside reservoirs from trunk lines and some water for the city is also taken from artesian wells located on the floor of the Lower Hutt Valley. The Wellington fault, capable of generating an MW 7.5 earthquake with a return period about 600 years, lies at the western side of the Hutt Valley and passes through both residential and commercial areas. The shaking intensity from this earthquake is expected to be MMI 9 (where MMI is Modified Mercalli Intensity) or greater over the entire City. We simulate the damage to the pipelines using fragility curves derived from overseas data, which are functions of MMI, material type and soil condition. We use a Poisson process to generate damage locations along a pipe, given an expected break rate, then repair the network using an algorithm based on a likely restoration sequence. First all valves are closed. Then starting from a reservoir, a valve is opened and pressure is established. We account for pipe repair time (a function of pipe diameter), opening and closing of valves, checking hydrants and travelling from one location to another. Multiple crew repair parties are used. In the present study, we have excluded the repair time required for damages to the bulk supply pipes (trunk lines) that carry water from water treatment plants to the reservoirs of the surrounding hills in the Hutt City, the reservoirs, wells and the customers own pipes that link individual properties to the council-owned distribution system.

Paper P51: [Read] [Presentation]

Keynote Address Session 1 Session 2 Session 3 Session 4 Session 5 Session 6 Poster Session Session 8