Abdul Razak Abdul Karim, Pierre Quenneville, Norazzlina M.Sa’don and Jason Ingham
From published literature, it was found that through-bolt connections were typically applied as a retrofit technique to most New Zealand unreinforced masonry (URM) buildings following the 1931 Hawke’s Bay earthquake. As connection failure by tearing out part of the diaphragm joist was observed in past earthquakes due to lateral earthquake loading, the strength of the bolted connection in existing indigenous New Zealand timber joists needs to be assessed. The main objectives of this study were to evaluate the strength and to identify the possible failure modes of bolted connections in New Zealand hardwood. Bolted connection tests loaded parallel-to-grain were performed using recycled native New Zealand Matai and Rimu hardwoods because the timber diaphragms in URM buildings are typically constructed using such wood species. From the experimental study, it was observed that the timber bolted connection can fail in either ductile or brittle modes. The test results obtained were compared with the European Yield Model (EYM), the New Zealand timber code (NZS 3603:1993), and a proposed set of equations (Quenneville 2009) in order to evaluate the applicability of those equations in predicting bolted connection strength for New Zealand hardwood. It was found that the EYM equations provide better predictions than the NZS 3603:1993 when compared to the actual capacity. However, the EYM predictions are only good in estimating the strength of timber bolted connections that fail under ductile mode. For the connections that fail exhibiting the brittle mode, the proposed row shear equation by Quenneville was found to give better strength estimation.
U. Akguzel and Stefano Pampanin
Exterior beam-column joints in existing reinforced concrete structures built prior to the introduction of modern detailing practices have arguably been observed to constitute the weakest link under strong earthquakes. In this paper, the results of an experimental and analytical investigation, that were carried out to illustrate the effect of variation in axial load on the performance of beam-column joint seismic retrofit using glass fibre reinforced polymer (GFRP) sheets, are presented. The experimental results of four 2/3-scale exterior beam-column joint subassemblies including one as-built and three retrofitted configurations are presented. All specimens were designed and constructed according to pre-1970s construction practice. Comparison with the experimental results are performed successfully in order to validate the assessment methodology proposed recently by the authors in literature and also to emphasize the unconservative assumptions made by retrofit design assuming constant axial load.
Majid Ali, Xiaoyang Li and Nawawi Chouw
The use of natural fibres as reinforcement in concrete is an innovative technology which is developing, because natural fibres are more economical than steel and synthetic fibres. The aim of this technology is to construct affordable but safe housing in earthquake prone areas. Recently, dynamic properties of coir fibre and rope reinforced concrete (CFRRC) beams were determined experimentally for different properties of coir fibres (fibre content and length). Now the research is extended to CFRRC columns for some selected coir fibre properties based on previous optimum outcomes. In this work, 5cm long fibres with a content of 3% by the weight of cement will be used to cast CFRRC columns. A coir rope with 1cm diameter will be used as an alternative reinforcement to steel bars. The basic static properties such as compressive strength, modulus of elasticity, splitting tensile strength and modulus of rupture for coir fibre reinforced concrete will be determined. The CFRRC column will be fixed on shake table, which will then be excited by a harmonic force with preselected amplitude and period. An axial load will also be placed at the top of the column. The response of the CFRRC column under the applied loads will be recorded by accelerometers, mounted along the height of the column. The data recorded by accelerometers will be analysed to study the cyclic response of CFRRC columns. This study will serve as a reference for further investigations on CFRRC columns subjected to real earthquake loading.
Majid Ali, Anthony Liu, Hou Sou and Nawawi Chouw
The overall goal of this research is to investigate low-cost but safe buildings in earthquake prone regions. The dynamic behaviour of CFRC structural member is practically unknown. In this work, the effect of fibre content on properties of CFRC is studied. To evaluate the efficiency of coir fibres in improving the properties of the concrete the performance of plain concrete (PC) is used a reference. Compressive strength, modulus of elasticity, splitting tensile strength and modulus of rupture were determined for all CFRC and PC specimens. Dynamic properties of the CFRC beams were investigated experimentally using impact loading with a calibrated hammer for the different stages of damage in beams. The damage was produced using four points loading in an universal testing machine. It is observed that the damping is increased up to 229 % and natural frequency is decreased up to 63 %, when CFRC beam with 3 % fibre content goes from uncracked to cracked stage.
Sumit Anand, Timothy Blackbourn, Troy Hoogeveen, Bilel Ragued and Liam Wotherspoon
A team of undergraduate students from the University of Auckland competed in the 2009 Asia Pacific IDEERS (Introducing and Demonstrating Earthquake Engineering Research in Schools) Seismic Design Competition. The team designed and constructed a model structure using MDF and other basic materials to resist seismic loading, with steel blocks used to represent the mass of the structure. The aim of the competition was to create an efficient design by reducing the mass of the supporting structure, increasing the number of steel blocks it could carry, and being able to resist large earthquake accelerations. The rules of the competition were altered significantly compared to previous years, resulting in a broad range of design solutions being implemented by the participating teams. This Poster provides an overview of the competition, from conceptual design through to final testing during the event at NCREE in Taiwan.
Experimental Investigation of the Seismic Behaviour of Slotted Reinforced Concrete Beam-Column Connections
Eu Ving Au, Des Bull and Stefano Pampanin
Judged 'best poster' for 2010
Past experimental investigations on conventional reinforced concrete frames at the University of Canterbury have highlighted extensive damage to floor slabs from the formation and elongation of beam plastic hinges. In light of this, a non-tearing floor connection in the form of a slotted reinforced concrete beam has been investigated. This beam forms a top-hinge of reinforced concrete above a slot and next to the floor. The connection differs from past, relatively complex to fabricate, non-tearing floor solutions by focusing on a simple connection that is as similar as possible to current reinforced concrete construction. The goal of this concept is to provide an immediate solution, which avoids damaged floor plates and can be easily implemented into New Zealand practice.
Quasi-static cyclic testing was carried out on three beam-column subassemblies utilising the proposed slotted-beam detail. This included 2/3rd-scale exterior and interior joint subassemblies and a ½-scale interior joint subassembly with precast floor units. A traditional beam-column joint subassembly was also tested which provided a benchmark for the 2/3rd-scale specimens.
Experimental results show that slotted-beams are a viable non-tearing floor solution, with significant reductions in beam elongation and floor damage observed compared to traditional connections. Tests also highlighted design issues unique to slotted-beams which must be addressed, such as anchorage of beam longitudinal reinforcement passing through interior columns, altered joint behaviour, low cycle fatigue, buckling of beam reinforcement and beam torsion. Design solutions to these issues were sought, a majority of which could be achieved by simple modifications to existing NZS3101:2006 provisions.
A generalised conditional intensity measure (GCIM) approach is proposed for use in the holistic selection of ground motions for any form of seismic response analysis. The essence of the method is the construction of the multivariate conditional distribution of any set of ground motion intensity measures conditioned on the occurrence of a specific ground motion intensity measure from probabilistic seismic hazard analysis. The approach therefore allows any number of ground motion intensity measures, identified as important in a particular seismic response problem, to be considered. A holistic method of ground motion selection is also proposed based on the statistical comparison, for each intensity measure, of the empirical distribution of the ground motion suite with the ‘target’ GCIM distribution. A simple procedure to estimate the magnitude of potential bias in the results of seismic response analyses when the ground motion suite does not conform to the GCIM distribution is also demonstrated. The combination of these three features of the approach make it entirely holistic in that: any level of complexity in ground motion selection for any seismic response analysis can be exercised; users explicitly understand the simplifications made in the selected suite of ground motions; and an approximate estimate of any bias associated with such simplifications is obtained.
F. Butt and Piotr Omenzetter
To improve the understanding of dynamic behaviour of structures during earthquakes, many structures across New Zealand are instrumented under the banner of the GeoNet project. GNS Science building in Avalon is amongst the structures which are permanently instrumented to record the data in real time. This three storey RC building has been instrumented with ten three-axial accelerometers in both of its wings along with a similar remote sensor to measure the ground motion. The data recorded using the strong motion array is used to determine modal properties like frequency, damping and mode shapes with the help of several system identification techniques. This study examines how these modal properties vary with several influencing factors such as ambient temperature, soil moisture, and amplitude and frequency content of seismic excitation. A finite element (FE) model of the building using structural drawings and at-site measurements is created to calculate the dynamic structural properties. Then, a comparison is made between the analytical and experimentally identified modal properties. The FE model is then updated, or calibrated, to provide an optimal agreement with the experimentally determined modal properties.
Arthur Lu, Gregory MacRae and Charles Clifton
In the non-seismic plastic design of steel structures, column residual stress effects affect the frame tangent stiffness and the frame lateral strength. These effects can be considered simply using the extended direct analysis (EDA) approach as part of design. However, for the design of frames subject to earthquake, the effect of residual stresses is generally ignored, except as part of the final column design check. Since residual stresses may reduce the frame capacity, there is a need to consider how important it is for earthquake shaking. This Poster looks at the effect of residual stresses on the seismic performance as a step in addressing this problem.
For single storey structures with high levels of column axial load, pushover, push-pull analyses and inelastic dynamic time-history analyses are undertaken both considering and not considering residual stress effects. The effects of residual stresses on the behaviour of these structures are summarized and design recommendations made.
José Chanchí, Gregory MacRae and Charles Clifton
For this reason, and many others, well designed, fabricated and constructed steel structures may be some of the most sustainable. However, severe earthquake damage may require major repair or replacement. This Poster evaluates a number of steel earthquake resistant structural forms and rates them according to their sustainability.
Forms studied include traditional moment steel frames, moment frames with post-tensioned steel beams, steel concentrically braced frames (with traditional and buckling-restrained braces), eccentrically braced frames and steel frames with plate shear walls. Yielding elements, replaceable ductile elements, traditional friction connections, sliding hinge type connections, structures with lead dissipaters, and improved sliding hinge type connections are considered. The frames are assessed in terms of the likelihood of structural damage, damage to the floors, damage to non-structural elements, and the likelihood permanent displacements.
Finally, the on-going studies of José Chanchí in order to develop robust improved connections so that these structural forms can be realized are described.
Gregory Cole, Rajesh Dhakal, Athol Carr and Des Bull
Building pounding describes the phenomenon of two adjacent buildings colliding under seismic excitation. Pounding has caused extensive building damage in many earthquakes, including earthquakes in Mexico City (1985), Turkey (1999) and (Italy 2009). Non-linear time history modelling is an important analytical tool that can be used to predict likely building pounding damage. However, any model must appropriately approximate the properties of the actual structures they represent. This poster outlines the consequences that may arise from two different mass approximation methods used when modelling building pounding. The governing assumptions of each approximation are presented and their effects on the response of two buildings are tested. The lumped mass approximation is found to significantly differ to the distributed mass approximation for the considered building configuration.
Jim Cousins, Nick Perrin, Graeme Hancox, Biljana Lukovic, Andrew King, Warwick Smith, Alastair McCarthy and Tony Shaw
Urban Wellington Region is uniquely vulnerable to large earthquakes. Not only is it bisected by large active faults but it is extremely isolated, with all supplies being transported along a small number of lifelines. All are vulnerable to earthquake damage. The potential for total loss of water and food supplies is real and could render large areas uninhabitable for weeks to months.
Greater Wellington Regional Council is evaluating new sources of potable water for the four cities of the Wellington Region. One important consideration is security of supply following major earthquakes near Wellington, in particular an earthquake involving rupture of the Wellington Fault because current bulk-supply pipelines cross the fault at several places.
We describe the pipeline damage expected from a Wellington Fault Earthquake and estimate the times needed for restoration of the bulk supply to main delivery points throughout the region. Priority is given to Wellington City because of its high population and lack of access to alternative supplies. One key finding is that the shortest time needed to get a limited supply back into Wellington is five to eight weeks. A second is that constructing a new supply dam west of the Wellington Fault could lower the restoration time by about one third, to three to five weeks. Even the lowered times are too long and it is clear that alternative solutions must be found.
Effects of Unreinforced Masonry Wall Slenderness Ratio on Out-of-plane Post-cracking Dynamic Stability
Hossein Derakhshan, Jason Ingham and Michael Griffith
A large number of time-history analyses were performed on several unreinforced masonry (URM) walls that had different slenderness ratios, and the viability of adopting wall slenderness ratio as a criterion for seismic assessment was investigated. Several combinations of three wall properties were assumed to cover most walls found in New Zealand URM buildings, and 30 representative time-history records were used to perform analyses. Walls were either two-leaf thick with no overburden load applied or three-leaf thick having an overburden load applied equal to the weight of a typical second-storey two-leaf URM wall. Wall behavioural data was obtained based on a previous laboratory based study, and each wall was subjected to ground motion scenarios with increasing peak ground acceleration (PGA). The ground motion record PGA that caused the wall to undergo a displacement limit equal to 60% of wall instability displacement was identified, and the sensitivity of the obtained PGA to wall slenderness ratio was studied for all the used records. It was shown that increasing wall slenderness ratio resulted in the wall being more vulnerable.
Liam Edwards, Daniel Paul, Charles Clifton and John Butterworth
Moment Resisting Frames (MRF) in multi-storey buildings are designed to resist varying earthquake actions. A conventional MRF and two uniform strength MRFs were designed according to current New Zealand structural standards, and modelled as 2D perimeter frames using the structural analysis software, SAP 2000. The uniform stiffness frames have the same beam size throughout their heights and the Non Uniform frame was designed to have four changes in beam size. One of the Uniform frames used lower size columns as entitled by the New Zealand standards, the other used the same size columns as the conventional design. The three frames were analysed using numerical integration non-linear time history analysis, considering P-Δ effects using a suite of 7 earthquake records. The Uniform frames experienced greater average displacements, post earthquake residual deflections, inter storey drifts, and plastic hinge rotations in comparison with Non Uniform frame. The floor accelerations between the three models were found to be insignificant. The New Zealand code (NZS 1170.5) does not conservatively predict the inter storey drifts of uniform frames.
Modelling the site response for earthquake scenarios is largely dependent on knowledge of the local shear wave velocity (Vs) structure and basin geometry. In order to ultimately understand the response of the Wellington region to large and proximal earthquakes, we refine our loosely constrained 3D model of the region. As part of the It’s Our Fault project, we implement methodology for the application of ambient noise based cross-correlation techniques to better constrain Vs and depth to bedrock in the Hutt Valley. By cross-correlating time windows of ambient seismic noise, we reproduce a station-station response that approximates the surface wave component of the inter-station Green’s Function. We apply frequency-time analysis to measure surface wave dispersion. We calculate synthetic dispersion from ~50,000 hypothetical models for each station pair. We then asses the likelihood of each model by assigning a fit value based on comparing the synthetic and measured dispersion curves. In this paper, we present one of the first instances, globally, of applying this methodology to basin imaging. We effectively image velocity stratigraphy of the basin with Vs ranging from 300-500 m/s in the shallowest layers that we resolve, trending to bedrock velocities averaging between 1300 and 1600 m/s. As there is a trade-off between interstation distance and depth resolution, our array is designed to provide Vs estimates of the middle to lower depths in the basin. Shallowest Vs estimates are provided by complementary seismic and geotechnical methods. However, these methods and results are outside the scope of the current paper.
Experimental design and analytical modeling of 1/2.5 scale under-designed reinforced concrete frame subassemblies with masonry infills
Patricio Gallo, Stefano Pampanin and Athol Carr
In the past years, extensive experimental work on innovative feasible retrofitting solutions has been carried out at the University of Canterbury, as part of the FRST Project ‘Seismic Retrofit Solutions for NZ Multi-storey Buildings’ on 2/3 scale beam column joints subassemblies representative of under-designed reinforced concrete buildings. As a dynamic validation of the seismic vulnerability of such structures and the feasibility of the developed retrofit solutions, a series of 4 1/2.5 scale frame subassemblies (two 3 storey - 2 bay asymmetric frames jointed together) will be tested on the shake table.
In this contribution, the experimental modelling and design of the benchmark experimental model is presented as well as the preliminary analytical predictions. Theory of similitude is used to develop an adequate distorted model, considering the impossibility of achieving a ‘true replica’ model. The theoretical basis needed to derive the appropriate scaling parameters for the physical quantities considered in the problem is derived from Buckingham’s PI-theorem. It is shown that a distributed artificial mass simulation leads, in this case, to an acceptable distortion when placing relatively thin led and steel plates on top of the floor slabs, so that the additional inertial and gravitational mass required by Cauchy principle coincide, avoiding the use of driving masses. The analytical expected response of the model subjected to different earthquake records is also presented, comparing the predicted response of the structure when considering or not the inelastic behaviour of the panel zone and/or the masonry infills.
Debra Gardiner, Des Bull and Athol Carr
The floor diaphragm of a structure provides a primary role in the performance of a structure. It connects the elements (frames and walls) of the structure together to form a structural system and it also transfers the forces which develop within the floor, from dynamic affects such as wind or earthquake action, out to the vertical lateral forces resisting elements of the structure. Structures designed today are required to meet various functional, geometrical and architectural requirements. These requirements have resulted in the design of more complex floor diaphragm layouts than what has been previously designed in the past. Simplistic design methods, such as the ‘beam analogy’ method, have become obsolete due to these changes of the types of floor diaphragms and also due to changes in construction practises of floor diaphragms over the past few decades. These changes have resulted in an increase of the flexibility of the floor and further the development of significant flexural and shear stresses within the floor. Limited guidance is currently available for designers on the methodology required to design complex floor for dynamic affects. This research investigates how the sizes and distributions of forces which develop in floor diaphragms from seismic loading are affected by various floor layouts including: different geometries, different locations of voids and penetrations within the floor and different layouts of walls and frames around the floor. Comment is given on a selection of the trends of the distributions of floor forces obtained from this research and recommendations on suggested floor diaphragm design procedures have been provided.
Anas Ibrahim, Rolly Orense, Mick Pender and Naotaka Kikkawa
This Poster focuses on the laboratory measurements of small strain shear modulus Gmax) of undisturbed Auckland residual clays using bender elements. Two elements were installed, one at the top and another at the bottom of the soil specimen. A function generator created a change of voltage in the transmitter to induce bending and transmission of shear wave through the specimen. The arrival of the shear wave at the other end of the specimen was recorded by an oscilloscope. The shear wave velocity, Vs, was calculated from the travel time of shear wave, taken as "start-to-start" between two instants at generation and at reception of shear wave, and the “tip-to-tip” distance between the elements. Gmax was then calculated from the known soil density and Vs. Input pulse using various kinds of wave form over a wide range of frequency was used. In the tests, the shear modulus was obtained at different levels of consolidation pressure.
Based on the results, a comparison was made on the small strain shear modulus obtained under dynamic loading using bender element tests, and that under monotonic (static) loading using submersible miniature linear variable differential transducers (LVDTs) clamped on soil specimen, which was conducted earlier by the authors.
Najif Ismail, Jason Ingham and Peter Laursen
A performance based design procedure was developed based on the out-of-plane flexural testing of seismically retrofitted unreinforced masonry (URM) walls using posttensioning. A macro level single degree of freedom (SDOF) dynamic model for the retrofit design of URM walls was developed. The test walls were dynamically tested by exciting the walls with a hammer and the developed model was updated to match the actual dynamic response of the tested walls. The developed SDOF model was used to find the pushover capacity curve for the posttensioned walls. The New Zealand Loading Standard’s (NZS 1170) defined elastic site spectrum was used to develop the demand spectra. Consequently, a simplified demand-capacity phase diagram was developed to study the seismic behaviour of posttensioned walls and was used to analyse the tested posttensioned URM wall. Using the graphically aided analysis, seismic performance of the wall was investigated and it was inferred that the simplified analysis procedure can be used for performance based posttensioning seismic retrofit design of New Zealand URM buildings.
B. Li, M. Jamil, N. Chouw and J.W. Butterworth
A bridge model with two identical steel segments but adjustable fundamental frequencies was tested on two shake tables providing various simulated non-uniform earthquake ground motions. The ratio of the fundamental frequencies of the adjacent segments; non-uniform versus uniform ground motion; soil-structure interaction (SSI) for two soil types versus fixed based response; ground motions corresponding to soft, medium and hard soils and pounding for various inter-segment gaps versus no pounding were studied. The investigation concluded that SSI tends to reduce the relative movement of adjacent segments and column bending moment. Conversely, non-uniformity of ground motion tends to increase relative response and bending moment.
Amir Khanlou and Serhan Sensoy
An accurate estimation of the seismic demand and capacities of structures is an important concern in Performance-Based Earthquake Engineering (PBEE). To fulfill the needs and objectives of PBEE several methods have been proposed, requiring nonlinear static or dynamic analyses. Although nonlinear static procedures have been widely used in seismic design and evaluation, compared to nonlinear dynamic analysis results may be inconsistent.
In present study, a novel analysis known as Incremental Dynamic Analysis (IDA) is employed to evaluate and assess the seismic performance of a seven story RC frame building. Nonlinear time history analyses of the structural model have been performed for a suite of eleven ground motion records, each scaled to several intensity levels. IDA curves have been generated and limit–states (e.g. immediate occupancy, life safety and collapse prevention) have been defined and summarized into their 16%, 50% and 84% fractile IDA curves. Since selected structure represents common mid-rise existing buildings in Turkey, fragility curves are also obtained in terms of (PGA) for each considered performance level to develop probabilistic structural damage estimation. In addition, phase portraits are also depicted to give better intuition of structural behavior at each defined performance level.
Based on obtained results, it could be inferred that given criteria in FEMA356, especially for collapse prevention may yield inconsistent results, pointing inadequate safety margin and insufficient confidence of estimating structural capacity before collapse.
Adane Gebreyohaness, Charles Clifton and John Butterworth
Dual-system buildings comprise a portion of buildings of historical significance in NZ. These buildings were constructed before and during the early 20th century, following a different design philosophy than today with little or no consideration for seismic loading. Therefore, assessment and, if necessary, retrofitting of such structures with no loss of heritage attributes is crucial in order to safeguard their occupants and the public in general in the event of severe earthquake. In light of this need, a reasonably accurate assessment of their seismic performance is indispensable. However, this has been difficult since, until recently, sufficient data on the inelastic cyclic behaviour of components of these buildings and a numerical tool which could model these properties have not been available.
Focusing on reducing the earthquake damage to non-structural components in buildings: research needs and future internationally coordinated plans
Alessandro Palermo, Stefano Pampanin, Andrew Baird and Pasquale Riccio
Earthquake Engineering is facing an extraordinarily challenging era. These challenges are driven by the increasing expectations of modern society to provide low-cost, architecturally appealing, high seismic performance structures. Modern structures need to be able to withstand a design level earthquake with limited or negligible damage such that disruption to business be minimised because of the economic consequences of such downtime.
Patel Chetan and Minesh Lal
Eccentrically braced frames (EBF’s) are seismic resistance structures with excellent energy dissipation properties. The active links which give EBF’s their unique characteristic allow for this by undergoing large plastic deformations. Traditionally, the active links on an EBF tend to vary throughout the building. However new research from the University of Canterbury given under the Poster titled, Evaluation of Vertical Mass Irregularity on the Seismic Performance of Nine Storey Buildings gives rise to the question of whether or not uniform strength active links throughout an EBF are a viable option. This research Poster has analysed uniform strength EBF’s and has compared their performance against the traditional complying designs. The first step in doing this required a preliminary design of both the uniform strength and complying designs. These were then modelled under SAP 2000 and integration time history analysis of four earthquake records were performed on the designs. The results obtained were surprising in that although average lateral displacement was well within drift limits; both sets of designs had very large active link rotations. These peak values of rotations however occurred at different locations showing that the size of the active link member was very influential in determining the concentration of demand. Although both designs would have been earthquake resistant buildings, they both would still suffer considerable damage. This thus led to the conclusion that neither design was better than the other.
Denis Pino, Stefano Pampanin, Andy Buchanan and Bruce Deam
This Poster describes the results of shake-table testing on one-quarter scale of LVL frames with post-tensioned beams using the Pres-Lam system. The main objective of this research is to validate previous results obtained from static testing conducted as part of the extensive on-going research program at the University of Canterbury, New Zealand.
Additionally, given that these are the first dynamic tests performed for Pres-Lam systems, an investigation of the dynamic factors that define the behaviour is carried out. The main factor to determine is the elastic damping, which until now has been assumed as 2% and is known to have a significant influence in the analysis and design of buildings subject to seismic loading.
The experimental results show that Pres-Lam systems have the ability to resist strong ground motions and undergoing large deformations without failure, the level of damaged is minimal keeping beams and columns close to the elastic range, and the self-centring action provided ensures negligible residual deformations. The elastic damping varied between 2% and 6% and it was found to be drift proportional.
Bill Robinson and Chris Gannon
Initially the RoGlider™, designed for vertical loads of 20 to 40 tonnes, consisted of a double acting sliding bearing which included an elastic restoring force provided by two rubber membranes. The development of the RoGlider family of Seismic Isolators has continued with the extension to lower loads, <2 tonnes, (LoGlider) by replacing the rubber membranes with elastic cords, enabling the elastic restoring stiffness to be reduced, and to higher loads, ~100 tonnes, (HiGlider) by extending the 'puck' outwards beyond the top and bottom plates, increasing the elastic restoring stiffness by both thickening and shortening the rubber membrane.
Advanced Model Development and Validation of Damage-free Precast Structural Connections with Unbonded Post-Tensioned Prestress
Geoff Rodgers, Geoff Chase and John Mander
The use of jointed precast concrete and steel connections with unbonded, post-tensioned prestress has been the focus of a significant amount of recent research. These systems provide a controlled inelastic response through gap-opening at the beam-column interface instead of through yielding and damage of the structural elements. The development of a model that captures all of the associated characteristics and provides an accurate prediction of connection response provides significant added confidence in response simulations. A model is developed that utilises a time-incremental model of the connection behaviour that accounts for yielding of the prestress tendons, the reduction or elimination of the prestressing force, friction between the post-tensioning tendons and the containing ducts, and asymmetry from non-centrally located tendons. The model is formulated using incremental versions of the Menegotto-Pinto and Ramberg-Osgood type, providing a smooth, continuous loading and unloading approximation to the piecewise linear behaviour. The model is validated against experimental results for an 80% full-scale jointed precast concrete connection tested with inputs drifts to a maximum of 4%. Results show very good agreement between the model and the experimental results, with errors generally less than ±5%. Overall, the model is generalisable to other connections using steel and concrete rocking connections that utilise this damage-free design approach and is a useful tool for evaluation of connection designs.
Vinod Sadashiva, Gregory MacRae and Bruce Deam
No real structure is perfectly regular. While some structures are planned to be architecturally irregular, other structures may be irregular due to unplanned effects. The present NZ seismic standard, NZS 1170.5 defines structural irregularity in terms of the size and shape of the building, the arrangement of the structural and non-structural elements within the structure, the distribution of mass in the building etc. Regularity limits set in NZS 1170.5 are based on heuristic thinking. Such limits determine the analysis method permitted in current worldwide codes. In this paper, a simple method is developed to quantify appropriate structural irregularity limits considering coupled vertical stiffness-strength irregularities.
Stiffness irregularity is almost always associated with a change in vertical strength over the building height. Various realistic combinations of stiffness-strength irregularity were investigated for shear-type structures of 3, 5, 9, and 15 storeys. Both regular and irregular structures were defined to have a constant mass at every floor level, and were designed for various ductilities in accordance with the NZ Equivalent Static method. Regular structures were either designed to produce a constant target interstorey drift ratio at all the floors simultaneously, or to cause a uniform stiffness distribution over the building height, with the target interstorey drift ratio at the first floor level. Irregular structures were created by modifying the lateral stiffness separately at the first level, mid-height and at the roof by amounts between 0.5 and 2 times the stiffness at the corresponding level in the regular structure. Strengths at these levels were also modified according to coupled stiffness-strength relationships at the chosen floor level for irregularity. The modified structures were then redesigned until the target interstorey drift ratio was achieved at the critical floor level. Inelastic dynamic time-history analyses were carried out by subjecting both the regular and irregular structures to a suite of code design level earthquake records. The increase in median peak drift demands were used to describe the effects of coupled stiffness-strength irregularity. Simple conservative equations allowing designers to estimate the likely increase in the response due to coupled stiffness-strength irregularity were then developed.
Faisal Shabbir and Piotr Omenzetter
Finite element models are very useful tools for simulating the behaviour of structures, but creation of a model which closely replicates the behaviour of the original structure is not easy. Model updating is an optimization technique where error residuals between the experimental and analytical responses are minimized. In this paper, global optimization techniques such as Genetic algorithms (GA), Particle swarm optimization (PSO), and a combination of local and global techniques were explored for model updating of civil engineering structures.
The experimental data used in this study comes from a two storey laboratory structure and a full scale composite floor of a concrete building. The aluminium, bookshelf-type laboratory structure was subjected to snap-back tests. To assess the dynamic performance and characteristics of the composite floor, dynamic tests were conducted on the floor of a University of Auckland building. The floor comprises precast concrete panels, covered with a layer of cast in situ concrete and supported by steel I beams. The floor was tested using two electrodynamic shakers within the elastic range. A dense array of sensors was employed for measuring responses.
Based on the experimental modal properties, stiffness properties of the analytical models of the structures were updated using GA and PSO to investigate the effectiveness of the methods. The advantages and disadvantages of these methods were examined and a combination of local and global optimization methods is proposed to harness the benefits of both of them. This Poster emphasises the importance of the use of suitable optimization method to obtain a reference finite element model which correctly represents the as-built original structure.
Dejan Novakov, P. Brabhaharan and Gavin Gregg
The Pelorus Bridge on state highway 6 between Blenheim and Nelson was assessed for seismic performance, after large earthquake events. Detailed assessment of the bridge was completed using a displacement-based method. This is an intuitive method which allows assessment of seismic performance of in-elastically responding structures and overcomes many of the shortcomings of the more traditional force-based method of assessment. It also enables a cost effective strengthening design to be developed.
The assessment indicated that the bridge’s capacity to resist seismic shaking is limited by the sliding capacity of the abutments, shear capacity of the bearing pins and passive resistance capacity of the soils behind the abutments.
The retrofit works on the bridge comprised installation of new reinforced concrete facing and rock anchoring of the unreinforced concrete abutment walls. The aim of this work was to stabilise the abutments and reduce seismic load on the supporting soils/rock and on the bridge superstructure and its bearings. The retrofit also included cross-tying of the wing walls to confine the abutments.
The northern abutment was located on a rock buff immediately above the Pelorus River, which was vulnerable to defect controlled failure in earthquake events. The rock bluff below the abutment was strengthened using rock anchors to tie the potentially unstable rock blocks back into stable rock.
The retrofit work was successfully completed in April 2009.
G.R. Walker and R. Musulin
The analysis described in this Poster demonstrates the importance of including insurance transactions in analysing the benefits of mitigation and to whom they accrue if incentives are to play a significant role. From the simplified analysis presented it can be deduced that either the policyholder, or the insurer, or the reinsurer, can be the primary beneficiary, depending on the detailed structure of the insurance system. In New Zealand home owners probably stand to lose from mitigation with the EQC being the main beneficiary because of the way its insurance is structured. The basic tools exist to undertake this analysis in the more detailed manner required to fully model all the complexities, but effectively utilising them will be very dependent on significant advances in earthquake engineering research directed at the fragility of critical elements of older construction and the proposed structural details for improving the seismic performance of this construction.
During a seismic event, loads are continually transferred between a structure and its foundation, with the two elements responding as an integrated system. This Poster presents the development and analysis of three-dimensional analytical models that provide a detailed representation of the non-linear seismic response of this integrated structure-foundation system. The aim of this analysis was to determine the effect of the inclusion of a shallow foundation model on the actions and displacements throughout a structure. A three-dimensional model was used rather than any simplification as it provides insight into the variation in response across the structure and foundation.
Using a range of earthquake records, integrated models were analysed with non-linear foundations models, elastic foundation models, and fixed base models without any foundation representation. The fixed base response was used as a baseline to quantify the effect of the foundation. Limited ductility and nominally elastic reinforced concrete moment resisting frame structures were used in the analysis to show the effect of integrated modelling on different structural design approaches.
Via collaboration between the U.S. Geological Survey and the Building Seismic Safety Council, the probabilistic basis for the earthquake ground motion intensities used to design new building in the U.S. has recently undergone a significant conceptual shift. In the 2005, 2002, and 1998 editions of the American Society of Civil Engineers (ASCE) standard entitled “Minimum Design Loads for Buildings and Other Structures,” the probabilistic ground motion for design is defined as a uniform-hazard spectral response acceleration (SRA) for a ground motion exceedance probability of 2% in 50 years. For most U.S. locations, this probabilistic SRA is the ground motion used for design purposes, governing over a parallel deterministic SRA. This is also the case in the 2010 update of the ASCE standard. However, now the probabilistic ground motion is defined as a “risk-targeted” SRA that is expected to result in buildings (designed according to the ASCE standard) having a collapse probability of 1% in 50 years.
Whereas the earlier, uniform-hazard definition of probabilistic ground motion for design used only a single point from a seismic hazard curve for the location of interest( namely that corresponding to the subjectively chosen 2%-in-50-year ground motion exceedance probability), the new risk-targeted definition utilizes the entire hazard curve. As a result, risk-targeted ground motions for design account for differences between shapes of hazard curves in, for example, the central and eastern versus western U.S. The derivation of risk-targeted ground motions from seismic hazard curves will be demonstrated in this presentation using two New Zealand locations.
Jenni Tipler, Margaret Worth, Hugh Morris and Quincy Ma
One third scale models of adobe earth walls, loaded out-of-plane, performed well when subjected to static and uni-directional dynamic testing at the University of Auckland. Two walls were built to comply with the requirements of NZS 4299 :1998 Earth Buildings Not Requiring Specific Design. One wall was nominally reinforced with only vertical steel at the corners and one was fully reinforced with steel vertical reinforcement for high seismic hazard and polysynthetic geogrid horizontal reinforcement every third course in the mortar layer.
The dynamic tests subjected the walls to a series of sine sweeps and scaled earthquake records. Both walls were undamaged under serviceability earthquakes and only suffered moderate cracking with the initial Ultimate Limit State scaled earthquakes records from 1985 Llolleo Chile, 1994 Northridge and 1940 El Centro with maximum accelerations of 0.5g. Partial collapse occurred in the nominally reinforced wall when the Chile record was amplified to 0.7g. Subsequent to the dynamic tests the walls were tilt tested to 50 degrees, this gave a lateral static load equivalent 0.8g and no adobe bricks dislodged even when an additional impact load was applied.
Scale dynamic testing is difficult to interpret quantitatively but clearly demonstrated the contribution of the vertical reinforcement to structural integrity of the unreinforced wall. The geogrid horizontal reinforcement is required for in-plane strength, it contributed little to the out-of-plane performance until load levels were near maximum. The geogrid significantly reduced the size of falling debris and prevented wall collapse.
In New Zealand around a dozen adobe houses are constructed annually and only in New Zealand is geogrid used within the wall mortar layer. These tests have confirmed that the geogrid makes a positive contribution to out-of -plane performance although full scale tests would be required to quantify these benefits.
Developing Collapse Fragilities and Other Vulnerability Models for Selected Building Types from the PAGER Project
Hyeuk Ryu, Kishor Jaiswal, Nicolas Luco and David Wald
We have developed collapse fragilities, as well as models for other damage states, for selected non-U.S. generic building types by performing incremental dynamic analysis (IDA). Each building type is modeled using a single-degree-of-freedom (SDOF) damped nonlinear oscillator with force-deformation behavior represented by a multi-linear capacity/pushover curve and moderate pinching hysteretic behavior. The capacity curves and other vulnerability-related parameters are obtained through contributions from international earthquake engineers to the WHE-PAGER project. This project is an ongoing collaborative effort between Earthquake Engineering Research Institute’s World Housing Encyclopedia (WHE) experts and the U.S. Geological Survey’s PAGER (Prompt Assessment of Global Earthquakes for Response) project to compile analytical seismic vulnerability parameters for structures around the world. The IDA’s involve seismic response-history analysis of each SDOF oscillator subjected to a number of selected ground motion records that are incrementally scaled in amplitude until collapse capacities are reached. A cumulative distribution function derived from these collapse capacities defines a collapse fragility for each building type. The IDA’s also provide seismic response results that are used to construct fragilities for damage states other than collapse (e.g., moderate damage). Estimated collapse capacities from an empirical and other analytical approaches being considered for PAGER are compared, and relative opportunities and challenges associated with the IDA approach are discussed.
D. Dizhur, H. Derakhshan, R. Lumantarna, M.C. Griffith and J.M. Ingham
Most research considering seismic assessment of URM walls has been conducted using laboratory-based studies with well defined but artificial conditions. Thus, in-situ testing is required to provide data with which to validate the accuracy of laboratory-based studies of URM walls. Seismic strengthening of the Wintec Block F building in Hamilton allowed an opportunity for a team of researchers from the University of Auckland to conduct in-situ testing of a wall segment in the building. This allowed comparison with companion experiments that had previously been undertaken in a laboratory setting. This field testing involved the extraction of clay brick and mortar samples, flexural bond tests, and out-of-plane testing of a wall both in the as-built condition and after the installation of a near-surface mounted (NSM) carbon fibre reinforced polymer (CFRP) retrofit solution. Testing confirmed that the boundary conditions in real buildings can significantly affect experimental response, and also confirmed that the near-surface mounted FRP solution is an excellent low-invasive option for seismic strengthening of unreinforced masonry buildings. Details of the history of the building and the methods used to undertake the field testing are reported, and experimental results are presented.