		2007 NZSEE Conference
	Abstracts

Keynote Address 1 Defining Acceptable Performance Criteria Planning for Expected Performance / Improving Existing Performance I Improving Existing Performance II Keynote Address 2 Developments Within Design I Developments Within Design II Developments Within Design III / Understanding the Inputs Assessing Existing Performance Poster Papers

Design of R. C. Frames using the Displacement Focussed Force Based Design Procedure

Barry Davidson

The design procedure using Displacement Focussed Force Based Design is described for multi level reinforced concrete frames. A design example is provided and the designed frames are analysed using nonlinear time history analyses. The results of these analyses demonstrate that the design procedure assists the designer to provide a structure with a more predictable performance than can be achieved by current methods.

Paper P26: [Read]

Structural Performance of Horizontal Slanting Beam-column Joints under Seismic Loading

Yuji Ishikawa and Hideki Kimura

Thanks to the advances in testing and computation in recent years, it has been possible to design complex space frames of high-rise RC buildings in Japan. This paper discusses the structural performance of horizontal slanting beam-column joints under seismic loading through a series of structural tests and FEM analyses. Analytical / numerical investigations using FEM modelling were carried out to evaluate the influence of the angle of horizontal slanting on the joint shear strength and thus support the definition of the structural testing protocol and overall set-up. On the experimental side, four specimens were tested to investigate three variables: beam-column joint shear stress, concrete strength and beam main bars anchor type. The column section shape of specimens was six-sided. The main results of the tests and FEM analyses are presented and design recommendations for horizontal slanting beam-column joints under seismic loading are proposed.

Paper P27: [Read]

The Little Red Hill Field Experiment: Seismic Response of an Edifice

F. Beuch, T.R. Davies, J.R. Pettinga, M. Finnemore and John Berrill

A field experiment was conducted near Lake Coleridge in the Southern Alps of New Zealand, focusing on the kinematic response of bedrock-dominated mountain slopes to strong seismic shaking. To study site effects of topography on seismic ground motion in a field situation, a small, elongated, and bedrock-dominated mountain ridge was chosen and equipped with a seismic array. In total seven EARSS instruments (Mark L4C-3D seismometers) were located on the crest, the flank and at the base of the 180 m high, 500 m wide, and 800 m long mountain edifice from February to July 2006. Seismic records of local and regional earthquakes as well as seismic signals generated by a small explosive source nearby were recorded and are used to provide information on amplification and deamplification effects within different parts of the edifice. Preliminary time and frequency domain analyses confirm site amplification is concentrated along the elongated crest of the edifice where amplifications up to 1100 % were measured relative to the motion at the base, and that the main response frequency of the edifice is around 5 Hz.

Paper P48: [Read]

Why Rock Motions are So Low in New Zealand Earthquakes - Site Classification Effects

Graeme McVerry, Jian Zhang and John Zhao

Rock motions for crustal earthquakes predicted by a recently-published response spectrum attenuation model for New Zealand earthquakes (McVerry et. al., 2006) are generally considerably less than those for the Abrahamson & Silva model on which it is based. The Abrahamson & Silva rock motions are closer to New Zealand values for shallow soil rather than rock. Much of the difference can be explained by different site classification schemes for the two models. The New Zealand model uses three site classes: rock, shallow soil and deep soil. The Abrahamson & Silva model uses only two classes, “rock” and deep soil. Its rock class is really a combined rock and shallow soil class, including sites with up to 20m of soil. The differences between the predicted motions largely disappeared when the New Zealand model was re-derived using the Abrahamson & Silva site classification. However, the additional class of the New Zealand model is important. Differences in motions for its three site classes are statistically significant. These results were the basis for the redefinition of the rock class to exclude the very stiff soil sites (up to 25m depth in some cases) that were combined with them in the 1992 Loadings Standard.

Paper P29: [Read]

GeoNet Project Update:Strong Ground Motion and Structural Monitoring in New Zealand

Ken Gledhill

The GeoNet project, which is now in its sixth year of operation, provides national-scale coverage for hazard detection, emergency response and data collection to increase the quality, applicability and confidence limits of hazards and engineering research. GeoNet is funded by the Earthquake Commission, operated by GNS Science and consists of national broadband seismograph, accelerograph and continuous GPS networks, and dense networks in areas of higher hazard, transmitting data to two independent data centres. Two important features of the GeoNet project are the integration of a wide range of monitoring networks and technologies into a national system, and the concentration on societal good through both hazards alerts and the availability of high quality research data. Strong ground motion is monitored by a national network of wide dynamic-range accelerographs, supplemented by dense regional networks in special areas. The building response programme is installing instruments in a variety of structural types to develop enhanced structural health monitoring (SHM) techniques and to provide data for the validation of engineering designs. The data recorded by the project are made available via the GeoNet web site (www.geonet.org.nz).

Paper P30: [Read]

Sensitivity Analysis of SHAKE Based Soil Site Response Modeling

Umut Destegul, Cees Van Weston and Siefko Slob

Determining and verifying the limitations of commercially available programs for analyzing ground response are important, because inaccurate results will affect the inputs to the design response spectra of structures and thus inputs to building codes and the disaster mitigation process.

Shake2000 is a Windows based computer program that is a variant of SHAKE which is one of the earliest and most successful computer codes that calculates the ground response. Although Shake2000 has been widely used for many years, there are not enough studies that reveal the usage limitations of SHAKE

Shake2000 was used for the ground response analysis of Lalitpur, Nepal. A synthetic strong motion was used in the analysis for the worst case scenario; M=8, D=48km (hypocentral), a_max=0.48g and resulted in maximum values of PGA=1.54g, and SA=2.799g for 3Hz frequency. For M=6.7, D=6.4km (epicentral), a_max =0.5g record, maximum values were PGA=3.44g and SA=8.64g for 2Hz frequency.

Although the reason for such high SA values could have been the use of stiff, thick soils and very severe earthquakes, further research revealed that when large magnitude earthquakes with high accelerations and stiff soils were simulated, SA values obtained were not reliability, such as 8.64g. In this study, it is demonstrated that the certainty of Shake2000 decreases when used with stiff soils and records with a_max of 0.48 g and above.

Paper P31: [Read]

Damage and Casualties in the South Java Tsunami of 17th July 2006

Jim Cousins, William Power, Neville Palmer, Stefan Reese, Iwan Tejakusuma and Saleh Nugrahadi

A team of scientists from GNS, NIWA and the Indonesian Agency for the Assessment and Application of Disaster Mitigation Technology (BPPT) undertook a reconnaissance mission to the South Java area affected by the tsunami of 17^th July 2006. With a shoreline wave height of about 5 m above sea level the tsunami was of a size highly relevant to New Zealand. In the worst-affected area there were 430 deaths, 360 injuries, and 3600 people displaced from their homes.

The team used GPS-based surveying equipment to measure ground profiles and inundation depths along 17 transects across affected areas near the port city of Cilacap and the resort town of Pangandaran. The purpose of the work was to acquire data for calibration of models used to estimate inundations, casualty rates and damage levels.

Water depths were typically 2 to 4 m where housing was seriously damaged. Damage levels ranged from total for older brick houses, to about 50% for newer buildings with reinforced concrete beams and columns, and to 5-20% for multi-storey hotels with heavier RC columns. “Punchout” of weak brick walls was widespread. Death and injury rates were both about 10% of the population exposed.

Paper P32: [Read]

Reducing the Subjectivity in Modified Mercalli Intensity Analysis: Drawing Isoseismals

Warwick Smith

MM intensities are anathema to most engineers, who would naturally prefer instrumental measures of ground motion, but the fact remains that most of our large historical earthquakes were not recorded by modern equipment. The only clue we have to the severity of their ground motion is the intensity data, so we need to make the best use of it that we can.

The traditional method of analysis is to plot the spot estimates of intensity on maps, and draw contours (isoseismals) to enclose, as far as possible, all the observations of the same intensity, while excluding lower values. A new development offers a more objective way of drawing isoseismals. This is important because the dimensions of the isoseismal contours are used to derive attenuation functions, which are in turn used in the modelling of hazard and risk.

The technique uses an algorithm for fitting elliptical contours to the raw intensity observations. It is necessary to constrain the solution by (a) identifying a priori the centre of the pattern, and (b) requiring that the major axes of all the individual elliptical isoseismals for any one earthquake have the same orientation. With these constraints, the solution is usually stable.

Paper P33: [Read]

A Probabilistic Relationship between Ground Shaking Parameters and MMI Based on Felt Report Data

M. Gerstenberger, B. Worden and D. Wald

Since 1999, both GNS Science and the USGS have been collecting "Felt Report" data from moderate and large earthquakes in the form of Modified Mercalli Intensity (MMI). This has resulted in a database of several hundred thousand MMI values that we have correlated with ground shaking amplitude data from the same events. We have developed separate probabilistic relationships for NZ and the US. Additionally, we have used the data to develop a distance-dependent intra-event amplitude sigma for station spacings of less than 10km. The basic steps of our methodology are: 1) associate all amplitudes and MMIs from each event; 2) find peak ground velocity (PGV) amplitude-MMI pairs that are within 2km of each other; 3) apply relative distance and site corrections; 4) derive a probability density function (PDF) for the amplitude measure at the MMI location based on intra-event sigma and the closest true amplitude recording; 5) stack the PDFs from the previous step, which will associate each MMI with a distribution of PGV amplitudes; 6) normalize the PDFs. We then use the resulting grid to obtain a probability of MMI given a PGV amplitude or to select a particular exceedance curve (e.g., 50% probability of exceedance of MMI given PGV).

Paper P34: [Read]