Earthquake Engineering and Structural Dynamics in Memory of Ragnar Sigbjörnsson

Structural engineering for earthquake resistance is undergoing a major revision in its approach toward the fulfillment of seismic safety and utilitarian serviceability in design. Rather than sticking to the established precepts of prescriptive design rules, design has turned toward the achievement of specific results through procedures that are tailored for different buildings and uses. These procedures represent notable research contributions, but they are complicated conceptually for implementation in structural engineering practice, and nonlinear building response estimates, frequently assumed to be performance, can vary within broad limits even for simple applications.

In this text we relate the history of code developments. We focus on the two main requirements of earthquake-resistant design of building structures: (1) Life Safety and (2) Protection of the Investment and relate the two demands to current concepts of Performance-Based Building Design. While we provide a personalized vision for the way in which the PBSD framework developed and matured during the last half century, a thorough historiography is not within the scope of the text. We nominate drift to serve as the prime metric for performance judgment.

Fire, blast, impact, earthquake beyond design and other extreme events are neglected in structural design but account for a good part of the actual rate of failures. In view of the uncertainty about such events, structures should be designed to be resilient to their foreseeable consequences. The paper presents an overview and sample results of an investigation of critical components, subassemblies thereof and whole concrete structures under simulated blast or impact, loss of columns due to extreme events, or under exposure to fire or earthquake loading. Design measures which promote resilience to earthquake, such as base isolation, footing uplift and dry joints, are evaluated for blast or progressive collapse. The role of masonry infills against progressive collapse is discussed. On the basis of the outcomes, design features which enhance resilience to multiple hazards are elaborated and contrasted to those which are efficient against one type of hazard but adverse for others.

The design acceleration spectrum requires site investigations and site-response analyses in accordance with the local seismic hazard. The variability in earthquake source and path effects may be considered using a large number of acceleration records compatible with the earthquake hazard. An important step is the selection and scaling of input acceleration records. Likewise, a large number of soil profiles need to be considered to account for the variability of site conditions. One option is to use Monte Carlo simulations with respect to layer thickness and shear wave velocity profiles to account for the variability of the site factors. The local seismic hazard analysis yields a uniform hazard acceleration spectrum on the bedrock outcrop. Site-specific response analyses also need to produce a uniform hazard acceleration spectrum on the ground surface. A simplified approach is proposed to define acceleration design spectrum on the ground surface that may be considered a uniform hazard spectrum.

In a nonlinear dynamic analysis, the most popular way of representing the inherent damping exhibited by the structure is by adopting the classical Rayleigh damping model. Although a large number of studies have identified issues with this model, it remains the most popular choice in the currently available non-linear time-history analysis software. A new paradigm for modelling damping was proposed by the authors in Puthanpurayil AM, Lavan O, Carr AJ, Dhakal RP (Bull Earthq Eng 14:2405–2434, 2016) and is summarised in this chapter. This new model provides for the damping effects at an elemental level. The chapter outlines the performance of the elemental damping models in comparison to the classical global damping models by conducting Incremental Dynamic Analyses (IDA) on a four-storey RC frame designed to Eurocodes. The IDA study presented in the chapter illustrates the fact that the elemental damping models give a more reliable estimate for the structural responses in comparison to all other models.

Recent earthquakes have clearly demonstrated that many key mechanisms of the seismic response of RC structural walls are still not adequately understood. To improve the knowledge about RC walls, adequate numerical models are needed. A robust macro MVLEM-FD-SFI element has been developed at the University of Ljubljana. It can be efficiently used to model very complex axial-flexural-shear interactions which are typical of the seismic response of many RC walls. The element was evaluated by several experiments that were selected for case studies within the International Wall Institute. The element MVLEM-FD-SFI can capture different aspects of the global response very well, e.g., it can describe the buckling of the longitudinal bars as well as the significant degradation of different shear mechanisms.

The earthquake response of an elastic-plastic structure to near-fault ground motions can be well represented by the response to a double impulse or a triple impulse with the equivalent magnitude. This is based on the fact that the main part of most near-fault ground motions consists of a few-cycle, nearly sinusoidal waves, transformable into a double impulse or a triple impulse, and such waves govern the maximum response. It is shown that, when a double impulse or a triple impulse is adopted as an input, the earthquake response is composed of a free vibration component only. This enables the closed-form expression of the critical elastic-plastic response maximizing the response with respect to the impulse timing. In order to demonstrate the practical applicability of the proposed theory, a base-isolated building is employed, and an approximate and simple expression of the maximum plastic deformation at the isolation story is derived.

Liquid storage tanks constitute an important portion of the critical infrastructure whose failure in case of an earthquake would lead to significant economic losses. Seismic base isolation is an emerging technology for assuring seismic safety of these critical structures. It reduces the effective seismic forces by shifting the fundamental period of the structural system out of the resonance range via use of laterally flexible isolation system elements. Despite the success of these structures under frequently occurring typical far-fault earthquakes, their behavior under near-fault earthquakes are being questioned recently. Near-fault earthquake records may contain long-period velocity pulses with high amplitudes, which may be close to or even coincident with the periods of isolation systems and/or the period of the sloshing fluid inside the tanks. This may result in unacceptably large isolation system and/or sloshing fluid displacements which would threaten the safety of the isolation system and the tank superstructure. Thus, both engineers and researchers turn to numerical investigations of the behavior of seismically isolated liquid storage tanks under near-fault earthquakes. However, there is a scarcity of the number of recorded near-fault ground motions as of today, and thus, artificially developed near-fault earthquakes, which are also known as pulse models, are being used as an alternative to recorded near-fault earthquakes in evaluating the near-fault behavior of these structures. There is no question that the reliability of the results obtained from such investigations which make use of pulse models would be strongly dependent on how realistic these models are. Therefore, it is deemed necessary to assess the capability of popular, current pulse models in representing the effects of near-fault ground motions on the responses of seismically isolated liquid storage tanks. For this purpose, we compare the structural response parameters, including isolator and sloshing displacements and isolation system and fluid-tank shear forces of a prototype, seismically isolated liquid storage tank under recorded near-fault earthquakes, and their approximate counterpart synthetic pulse models.

The conventional opinion of scientists is that tsunamis are usually caused by earthquakes, but the discovery of the huge submarine landslides on the continental shelf of the North Atlantic Ocean has changed this view considerably. Most tsunamis are triggered by earthquakes, but if it starts a landslide, the tsunami can become enormous, even though the earthquake is small. The energy balance in such tsunamis was studied for the first time using the translatory wave theory by Eliasson and Sigbjörnsson. The theory was originally proposed by Eliasson. The mechanical wave energy generated in tsunamis of such origin is estimated for landslides on mountain slopes and submarine landslides. The initial tsunami wave height is estimated and compared to the results of the block slide estimation method. The energy transmission of the water wave can be translatory or by a solitary group of oscillatory waves where the length of the group varies according to water depth. In a case study, the Tohoku tsunami and earthquake in 2011 in Japan is found to be caused by a coseismic slip and a landslide in combination. A discussion of applying the theory to find a hazard curve for a tsunami wave from different sources at a given location is discussed.

The theme of this paper is the transformation of velocities from a GPS network into a surface strain rate field that illustrates recent and underlying movements within the tectonic system of Iceland. The surface strain rates presented are derived using GPS measurements at the base stations of the National Land Survey of Iceland reference network. The GPS observations are discussed, and the derived average velocity field for an 11-year period, from 1993 to 2004, is presented. The measurements at the westernmost and the easternmost points, which are located on the oldest parts of the country, i.e., the North America plate and the Eurasia plate, respectively, show the rigid movement of the two crustal plates. The average northwards velocity of the two plates is 25 mm/year, and the differential movement is 22 mm/year in the direction 281° from north. The definitions of strain rate tensor and vorticity tensor are outlined and applied in a numerical analysis to derive tensor strain rates from the GPS data. The strain rate tensor field is displayed on maps showing normal and shear strain rate fields, along with principal and dilatation rate fields, as well as the vorticity field. The derived tensor fields are discussed and interpreted in relation to the present-day view on the tectonism of Iceland and recent tectonic activity. The results indicate that most of the significant strain is concentrated on a tongue zone that goes from the North American plate into a slit in the Eurasian plate. The results clearly show the strained volcanic zone of Iceland and the rigid zones with zero strain on either side of it.

The parameters for a point source model are estimated using strong ground motion recorded in the largest earthquakes in south Iceland since a strong motion network established (from 1986 to 2008). The model is found to provide a good fit to the ground motion records. Parameters were estimated for the earthquakes of 17th June 2000 M _w 6.5, 21st June 2000 M _w 6.4, 29 May 2008 M _w 6.3 and 25 May 1987 M _w 5.8. Great variability in parameters was observed. A mean kappa, κ = 0.04 s, was observed. The stress drop for the three larger events had an estimated average value of 80 × 10 ⁵ Pa and an average dislocation of 90 cm. Closed form relations for root-mean-squared (rms) acceleration are presented for the ground motion in the near and far field. A peak factor is then applied to rms acceleration in order to obtain peak ground acceleration. The model can be applied instead of empirical ground motion prediction equations to estimate ground motion and potential seismic hazard in Iceland. Its estimated parameters can be used in the framework of stochastic modelling to simulate artificial ground acceleration for the dynamic analysis of structures.

The seismicity in Iceland is related to the Mid-Atlantic plate boundary which crosses the country from north to south. Since 1700, there have been 25 earthquakes of magnitude six or greater in the two major seismic zones in Iceland. For many of the historical earthquakes (older than 100 years), detailed written descriptions are available that describe the damage from farm to farm. For the most recent earthquakes, like the two south Iceland earthquakes of June 2000 and the south Iceland earthquake of May 2008, comprehensive building-by-building loss data exist. The loss data in these cases are split into a number of subcategories of structural and non structural damage. The building stock in south Iceland has changed significantly from 1700 to the present. For ages, vulnerable turf and stone houses dominated, but in the twentieth century, concrete buildings and timber buildings took over. Seismic codes were implemented in 1976 and have gradually improved the seismic capacity of present building stock. In this book chapter, an overview of the seismic performance of old and new Icelandic buildings is given. Observed and reported damage caused by three earthquake sequences and a single event are discussed; first, the damage caused by two earthquakes in August 1784 in south Iceland; then, by the 1896 earthquake sequence in south Iceland; then, the damage after a single event in 1934 in north Iceland; and finally, the loss data from the 2000 and 2008 earthquakes in south Iceland. The main focus is on the last three events for which most of the data exist.

The recent earthquakes that occurred in Central Italy, in L’Aquila in 2009 and in Amatrice in 2016, have shown, once again, how infill panels and partitions, widely used as non-structural elements in reinforced concrete frames, are significantly vulnerable when subjected to seismic actions. Moreover, they highly affect the structural behavior and seismic response of typical, multi-story buildings, strongly influencing their repair costs due to the inevitable damage they usually undergo during seismic events. Therefore, they should be designed to withstand earthquakes forces, both in-plane and out-of-plane, so to ensure the safety of the occupants. In addition to seismic performance, infill panels play the primary role of providing comfort to occupants through adequate thermal and acoustic insulation. In this paper, an innovative solution of earthquake-safe and eco-friendly infill panels is presented, which are, at the same time, thermally and acoustically efficient, thus avoiding waste of energy and respecting the environment. The proposed innovative technology is tested with different surface finishes, boundary conditions and temperatures. A comparison of the proposed technology with respect to conventional infill typologies is provided, highlighting pros and cons.

Long-term monitoring data of bridge deck accelerations and wind velocities from the Hardanger Bridge were used here to investigate the relationship between the wind characteristics and the bridge response. The bridge, as well as the extensive monitoring system installed on it, is introduced. The wind velocities and bridge deck accelerations were recorded on several locations along the bridge span using 20 accelerometers and 9 anemometers. The lateral, vertical and torsional accelerations of the bridge deck are presented in the form of buffeting curves. Response surface methodology (RSM) was employed to elaborate on the significant variability observed in response components. Using wind field statistics as predictor variables in the analyses, the variability in response was attributed to the variability in the wind field. The effect of wind-related variables on the response and their interactions are presented using surface plots.

Firstly, this paper presents a review of the main steps for simulating floating bridge behaviour. Both time-domain and frequency-domain approaches are presented. An exemplified model setup and simulation results from the selected case study are presented. Secondly, data obtained from extensive structural and environmental monitoring of the studied bridge are presented. Data analyses attempting to visualize the correlations between excitation sources and response quantities are discussed. Finally, an operational modal analysis is carried out, to attempt to identify the modal parameters from response measurements only.

Floating bridge and submerged floating tunnel concepts have been developed for the purpose of crossing wide straits with a span length of many kilometers. The frequencies of the first dynamic modes of these structures may typically be of the order of 1 min or more. This implies that amplified dynamic response effects need to be assessed both for the wave-frequency and the low-frequency regimes. The low inherent damping levels for the low-frequency response imply that high dynamic amplification may occur. Computational methods for evaluation of the low-frequency hydrodynamic loading and the associated dynamic response are addressed in the present paper, with a focus on developing simplified procedures. These are also compared to more refined calculation methods. A specific case study is considered for the purpose of illustrating the relative effects of the respective dynamic response components. The example structure is a tunnel with a double cross-section which was developed for crossing of the Sognefjord located at the west coast of Norway.

Earthquakes and tsunamis continue to be some of the most disastrous natural events. In this review paper, we present the work developed by the Seismic Risks Group of Instituto Superior Técnico (IST) in the last decade, with emphasis on the performance indicators proposed to deal with the effects of natural disasters. Starting with an appraisal of the world earthquake impacts since 1900, the text summarizes the most relevant impact indicators, namely, SIRIUS and disruption indexes (DI) for the urban and industrial fabrics, which includes cascading effects. Also, the early stage of the analysis of multi-hazard, namely for shaking and tsunami events, is reviewed. Finally, we use these impact indicators in conjunction with performance indicators (RRW – risk reduction worth and RAW – risk achievement worth) to communicate the risk to the population and mitigate the action of future events. Several conclusions can be taken from the work developed. The most important is that earthquake impacts cannot be measured only by human losses and losses to property, but indirect and cascading effects also play a huge role in the global impact, especially in moderate to large events.

The UPStrat-MAFA (Urban Disaster Prevention Strategies using Macroseismic Fields and FAult Sources) project, funded by the European Commission, had a multi-disciplinary approach to disaster prevention that encompassed strategies based on the analysis of the level of risk and information. Use of macroseismic intensity data and its probabilistic treatment was one of the successful innovations of the project. The method allowed incorporation of knowledge of seismic source, faults and their properties to estimate hazard and provide valuable insights on the level of expected shaking during future earthquakes. A holistic approach to risk assessment was one of the most relevant contributions of this project. It was implemented with a new concept of global damage (the disruption index, DI) that provides a systemic way to measure earthquake impact on urban areas and helps in prevention strategies, as well as in decision-making for emergency planning and post-disaster activities. The project strongly relied on prevention strategies based on education and communication of risk. Analysis on the levels of education was performed and weaknesses identified. Various educational tools were prepared: video games for children and audio-video products for the general public. The results and achievements of the project were widely distributed to both the general public and experts.

This paper addresses the question of whether or not the perception of earthquake effects is gender dependent. The case considered is the south Iceland earthquakes of June 2000. This includes two moderate-sized, shallow, strike-slip earthquakes with high peak ground acceleration in the epicentral area. After the earthquakes, a survey on earthquake intensities was carried out by the Earthquake Engineering Research Centre of the University of Iceland. The survey also included questions addressing safety issues and demographic information. The data dealt with herein cover 249 respondents in the epicentral area. In the analysis presented herein, emphasis is placed on the following questions: Did you manage to seek shelter inside a house during the earthquake? Did you manage to keep your balance? How long did it take you to recover? The main finding is that the data indicate a tendency towards gender-dependent perception of earthquake effects, which in some cases appears to be statistically significant. In particular, the time taken to recover seems to be very different between male and female respondents. The results also indicate that female respondents are biased towards higher estimation of felt earthquake intensity, while the opposite is true for male respondents.

This chapter pulls together insights about post-disaster resilience and recovery from a comparison of 10 recent earthquake disasters. Recovery is a complex process that starts immediately after a disaster. In simple terms, it involves a return to normality. But recovery is not only about speed; the quality of reconstruction and the idea of building back better are also important.

To better understand which factors may affect the speed of recovery, data from the 10 earthquake events are analysed in terms of 3 exogenous factors that are given, and 5 sets of endogenous factors that are within the control of decision-makers and planners – authority, decision-making, planning, finance and science.

The somewhat surprising finding is that there appears to be little relation between speed of recovery and the exogenous factors of size of impact, population demographics and economic factors. However, there is a clear relationship between the standard of post-disaster management decision-making and both the speed (R² = 0.56) and quality of recovery (R² = 0.90). The relationship between post-disaster decision-making and the quality of recovery in terms of whether crucial aspects of the society and economy are built back better is striking.