Thesis (MSc)--Stellenbosch University, 2011. / ENGLISH ABSTRACT: FLAC3D is a three-dimensional explicit nite difference program for solving a variety of solid mechanics problems, both linear and non-linear. The development of the algorithm and its initial implementation were performed by Itasca Consulting Group Inc. The main idea of the algorithm is to discritise the domain of interest into a Lagrangian grid where each cell represents an element of the material. Each cell can then deform according to a prescribed stress/strain law together with the equations of motion. An in-depth study of the algorithm was performed and implemented in Java. During the implementation, it was observed that the type of boundary conditions typically used has a major in uence on the accuracy of the results, especially when boundaries are close to regions with large stress variations, such as in mining excavations. To improve the accuracy of the algorithm, a new type of boundary condition was developed where the FLAC3D domain is embedded in a linear elastic material, named the Boundary Node Shell (BNS). Using the BNS shows a signi cant improvement in results close to excavations. The FLAC algorithm is also quite amendable to paralellization and a multi-threaded version that makes use of multiple Central Processing Unit (CPU) cores was developed to optimize the speed of the algorithm. The nal outcome is new non-commercial Java source code (JFLAC) which includes the Boundary Node Shell (BNS) and shared memory parallelism over and above the basic FLAC3D algorithm. / AFRIKAANSE OPSOMMING: FLAC3D is 'n eksplisiete eindige verskil program wat 'n verskeidenheid liniêre en nieliniêre soliede meganika probleme kan oplos. Die oorspronklike algoritme en die implimentasies daarvan was deur Itasca Consulting Group Inc. toegepas. Die hoo dee van die algoritme is om 'n gebied te diskritiseer deur gebruik te maak van 'n Lagrangese rooster, waar elke sel van die rooster 'n element van die rooster materiaal beskryf. Elke sel kan dan vervorm volgens 'n sekere spannings/vervormings wet. 'n Indiepte ondersoek van die algoritme was uitgevoer en in Java geïmplimenteer. Tydens die implementering was dit waargeneem dat die grense van die rooster 'n groot invloed het op die akkuraatheid van die resultate. Dit het veral voorgekom in areas waar stress konsentrasies hoog is, gewoonlik naby areas waar myn uitgrawings gemaak is. Dit het die ontwikkelling van 'n nuwe tipe rand kondisie tot gevolg gehad, sodat die akkuraatheid van die resultate kon verbeter. Die nuwe rand kondisie, genaamd die Grens Node Omhulsel (GNO), aanvaar dat die gebied omring is deur 'n elastiese materiaal, wat veroorsaak dat die grense van die gebied 'n elastiese reaksie het op die stress binne die gebied. Die GNO het 'n aansienlike verbetering in die resultate getoon, veral in areas naby myn uitgrawings. Daar was ook waargeneem dat die FLAC algoritme parralleliseerbaar is en het gelei tot die implentering van 'n multi-SVE weergawe van die sagteware om die spoed van die algoritme te optimeer. Die nale uitkomste is 'n nuwe nie-kommersiële Java weergawe van die algoritme (JFLAC), wat die implimentering van die nuwe GNO randwaardekondisie insluit, asook toelaat vir die gebruik van multi- Sentrale Verwerkings Eenheid (SVE) as 'n verbetering op die basiese FLAC3D algoritme.

published_or_final_version / Civil Engineering / Doctoral / Doctor of Philosophy

Geologists have indicated that the question is not if a catastrophic earthquake will occur in Oregon but when one will occur. Scientists estimate that there is close to 40 percent conditional probability that a Cascadia subduction zone earthquake of magnitude 8.0 or above will strike Oregon in the next 50 years. In addition, the majority of Oregon's bridge inventory was built prior to the current understanding of bridge response and prior to current understanding of the expected earthquake demands. In order to minimize potential bridge damage in the case of an earthquake, one approach is to retrofit seismically deficient bridges. However, often times the decision maker is faced with the difficulty of selecting only a few bridges within the inadequate ones. Hence, the issue of prioritizing upgrading naturally arises. The goal of this study is to assess and refine bridge prioritization methodology to be utilized for ranking Oregon's bridge inventory. CFRP retrofit has been experimentally and analytically evaluated to demonstrate the effectiveness of the technique and was found to be an efficient and economical option. A vulnerability assessment estimates that close to 30 percent of Oregon's highway bridge inventory will sustain moderate damage to collapse. However, retrofitting two most common bridge types in the inventory will reduce the number of damaged bridges by about 70 percent. A cost-benefit assessment that takes into consideration direct and indirect costs associated with damaged bridges and retrofitting of bridges shows that the benefit is up to three times the cost to retrofit. The same principle was applied to rank twelve highway segments for seismic retrofit considered important by Oregon Department of Transportation. One selected segment was considered to be retrofitted and vulnerability assessed. The benefit to cost ratios for each assessment was compared and the highway segments were ranked accordingly. The top five segments in the ranking happen to be located in the East-West corridor connecting I-5 to US-101.

A large magnitude, long duration subduction earthquake is impending in the Pacific Northwest, which lies near the Cascadia Subduction Zone (CSZ). Great subduction zone earthquakes are the largest earthquakes in the world and are the sole source zones that can produce earthquakes greater than M8.5. Additionally, the increased duration of a CSZ earthquake may result in more structural damage than expected. Given such seismic hazard, the assessment of reinforced concrete substructures has become crucial in order to prioritize the bridges that may need to be retrofitted and to maintain the highway network operable after a major seismic event. Recent long duration subduction earthquakes occurred in Maule, Chile (Mw 8.8, 2010) and Tohoku, Japan (Mw 9.0, 2011) are a reminder of the importance of studying the effect of subduction ground motions on structural performance. For this purpose, the seismic performance of substandard circular reinforced concrete bridge columns was experimentally evaluated using shake table tests by comparing the column response from crustal and subduction ground motions. Three continuous reinforced columns and three lap-spliced columns were tested using records from 1989 Loma Prieta, 2010 Maule and 2011 Tohoku. The results of the large-scale experiments and numerical studies demonstrated that the increased duration of subduction ground motions affects the displacement capacity and can influence the failure mode of bridge columns. Furthermore, more damage was recorded under the subduction ground motions as compared to similar maximum deformations under the crustal ground motion. The larger number of plastic strain cycles imposed by subduction ground motions influence occurrence of reinforcement bar buckling at lower displacement compared to crustal ground motions. Moreover, based on the experimental and numerical results, subduction zone ground motion effects are considered to have a significant effect on the performance of bridge columns. Therefore, it is recommended to consider the effects of subduction zone earthquakes in the performance assessment of substandard bridges, or when choosing ground motions for nonlinear time-history analysis, especially in regions prone to subduction zone mega earthquakes. Finally, for substandard bridges not yet retrofitted or upgraded seismically, the following performance limit recommendation is proposed: for the damage state of collapse, which is related to the ODOT's Life Safety performance level, the maximum strain in the longitudinal reinforcement should be reduced from 0.09 (in./in.) to a value of 0.032 (in./in.) for locations where subduction zone earthquakes are expected, to take into consideration the occurrence of bar buckling.

Unreinforced masonry (URM) construction, which has been widely used in the United States, presents a large threat to life safety and regional economic development because of its poor seismic resistance. In this research, the nonlinear seismic properties of URM structures were investigated via a quasi-static test of a full-scale two-story URM building and associated analytical and numerical studies. The tests of the 24ft. by 24ft. in plan 22ft. high URM building revealed that the damage was characterized by (1) the formation of large discrete cracks in the masonry walls and (2) the rocking and sliding of URM piers. Both of these results were consistent with the predictions based on individual component properties obtained in previous research. However, the tests also revealed significant global behavior phenomena, including flange effects, overturning moment effects, and the formation of different effective piers in a perforated wall. This global behavior greatly affected the response of the URM building tested. In order to understand the nonlinear behavior of the test structure, a series of analytical studies were conducted. First, at the material level, a mechanical key model was proposed to describe the failure of URM assemblages under a biaxial state of stress. Second, at the component level, an effective pier model was developed to illustrate the mixed failure modes of a URM pier and its nonlinear force-deformation relationship. Third, at the structure level, a nonlinear pushover model was built using the mechanical models at the material and component levels to describe the nonlinear properties of a URM building. This nonlinear pushover model and a three-dimensional finite element model were employed to analyze the test structure. Both gave results in good agreement with the test data. Improvements to current provisions for the evaluation of existing masonry structures were proposed.

Historical seismic events such as the San Fernando earthquake of 1971 and the Loma Prieta earthquake of 1989 did much to highlight the vulnerabilities in many existing highway bridges. However, it was not until 1990 that this awareness extended to the moderate seismic regions such as the Central and Southeastern United States (CSUS). This relatively long neglect of seismic issues pertaining to bridges in these moderate seismic zones has resulted in a portfolio of existing bridges with seismic deficiencies which must be assessed and addressed. An emerging decision tool, whose use is becoming ever increasingly popular in the assessment of this seismic risk, is that of seismic fragility curves. Fragility curves are conditional probability statements which give the probability of a bridge reaching or exceeding a particular damage level for an earthquake of a given intensity level. As much research has been devoted to the implementation of fragility curves in risk assessment packages, a great need has arisen for bridge fragility curves which are reliable, particularly for those in moderate seismic zones. The purpose of this study is to use analytical methods to generate fragility curves for nine bridge classes which are most common to the CSUS. This is accomplished by first considering the existing bridge inventory and assessing typical characteristics and details from which detailed 3-D analytical models are created. The bridges are subjected to a suite of synthetic ground motions which were developed explicitly for the region. Probabilistic seismic demand models (PSDM) are then generated using these analyses. From these PSD models, fragility curves are generated by considering specific levels of damage which may be of interest. The fragility curves show that the most vulnerable of all the bridge nine bridge classes considered are those utilizing steel girders. Concrete girder bridges appear to be the next most vulnerable followed by single span bridges of all types. Various sources of uncertainty are considered and tracked throughout this study, which allows for their direct implementation into existing seismic risk assessment packages.

The goal of this study is to design and evaluate economic and rapid seismic retrofit strategies for relatively small rehabilitation projects for steel structures consistent with the tenets of sustainable design. The need to retrofit existing structures in earthquake prone regions may arise directly from the problem of aging and deteriorating conditions, recognition of the vulnerability of existing infrastructure, from updates in seismic code requirements, or changes in building performance objectives. Traditional approaches to seismic hazard mitigation have focused reducing the failure probabilities, consequences from failures, and time to recovery. Such paradigms had been established with little regard to the impact of their rehabilitation measures on the environment and disruptions to occupants. The rapid rehabilitation strategies proposed here have sustainability benefits in terms of providing a more resilient building stock for our communities as well as minimizing environmental and economical impacts and social consequences during the rehabilitation project. To achieve these goals, a unique approach to design supplemental systems using tension-only elements is proposed. In this design approach undesirable global and local buckling are eliminated. Two rapid rehabilitation strategies are presented. The first is a bracing system consisting of cables and a central energy dissipating device (CORE Damper). The second is a shear wall system with the combined use of thin steel plate and tension-only bracing. Analytical studies using both advanced and simplified models and proof-of-concept testing were carried out for the two devices. The results demonstrated stable, highly efficient performance of the devices under seismic load. Preliminary applications of the CORE damper to the retrofitting of a braced steel frame showed the ability of the system to minimize soft story failures. Both techniques can be implemented within a sustainability framework, as these interventions reduce the seismic vulnerability of infrastructure, are low cost, utilize materials and fabrication processes widely available throughout the world, can be handled by unskilled labor and carried out with minimal disruptions to the environment. The approach taken in this study can provide a road map for future development of sustainability-based rehabilitation strategies.

In the central United States, undefined earthquake sources, long earthquake recurrence intervals and uncertain ground motion attenuation models have contributed to an overstatement of regional seismic hazard for the New Madrid Seismic Zone on the National Seismic Hazard Maps. This study examined concerns regarding scientific uncertainties, overly stringent seismic mitigation policies and depressed local economy in western Kentucky through a series of informal interviews with local businessmen, public officials, and other professionals in occupations associated with seismic mitigation. Scientific and relative economic analyses were then performed using scenario earthquake models developed with FEMA’s Hazus-MH software. Effects of the 2008 Wenchuan earthquake in central China and seismic mitigation policies in use there were considered for potential parallels and learning opportunities. Finally, suggestions for continued scientific research, additional educational opportunities for laymen and engineering professionals, and changes in the application of current earthquake science to public policy in the central United States were outlined with the goal of easing western Kentucky economic issues while maintaining acceptable public safety conditions.

The functionality of the transportation network following an earthquake event is critical for post-earthquake response and long-term recovery. The likely performance of a transportation network can be evaluated through a detailed seismic risk assessment. This paper presents an assessment of the seismic risk to the transportation network in the City of Charleston and the surrounding counties to support emergency response and the development of mitigation strategies and emergency planning efforts (such as lifeline selections). This study includes an inventory analysis of the 375 bridges in the Charleston area, and convolution of the seismic hazard with fragility curves analytically derived for classes of bridges common to this part of the country, damage-functionality relationships, and replacement cost estimates using relevant region-specific data. Using state-of-the-art tools, the distribution of potential bridge damage and functionality is evaluated for several scenario events, in order to aid in the identification of emergency routes and assess areas for investment in retrofit. Additionally, a sensitivity study is conducted to determine the criticality of a few of the different input models. Initial estimates of economic losses are assessed and preliminary recommendations for prioritizing retrofit are presented.

[Truncated abstract] Perth is the capital city of Western Australia (WA) and the home of more than three quarters of the population in the state. It is located in the southwest WA (SWWA), a low to moderate seismic region but the seismically most active region in Australia. The 1968 ML6.9 Meckering earthquake, which was about 130 km from the Perth Metropolitan Area (PMA), caused only minor to moderate damage in PMA. With the rapid increase in population in PMA, compared to 1968, many new structures including some high-rise buildings have been constructed in PMA. Moreover, increased seismic activities and a few strong ground motions have been recorded in the SWWA. Therefore it is necessary to evaluate the seismic risk of PMA under the current conditions. This thesis presents results from a comprehensive study of seismic risk of PMA. This includes development of ground motion attenuation relations, ground motion time history simulation, site characterization and response analysis, and structural response analysis. As only a very limited number of earthquake strong ground motion records are available in SWWA, it is difficult to derive a reliable and unbiased strong ground motion attenuation model based on these data. To overcome this, in this study a combined approach is used to simulate ground motions. First, the stochastic approach is used to simulate ground motion time histories at various epicentral distances from small earthquake events. Then, the Green's function method, with the stochastically simulated time histories as input, is used to generate large event ground motion time histories. Comparing the Fourier spectra of the simulated motions with the recorded motions of a ML6.2 event in Cadoux in June 1979 and a ML5.5 event in Meckering in January 1990, provides good evidence in support of this method. This approach is then used to simulate a series of ground motion time histories from earthquakes of varying magnitudes and distances. ... The responses of three typical Perth structures, namely a masonry house, a middle-rise reinforced concrete frame structure, and a high-rise building of reinforced concrete frame with core wall on various soil sites subjected to the predicted earthquake ground motions of different return periods are calculated. Numerical results indicate that the one-storey unreinforced masonry wall (UMW) building is unlikely to be damaged when subjected to the 475-year return period earthquake ground motion. However, it will suffer slight damage during the 2475-return period earthquake ground motion at some sites. The six-storey RC frame with masonry infill wall is also safe under the 475-year return period ground motion. However, the infill masonry wall will suffer severe damage under the 2475-year return period earthquake ground motion at some sites. The 34-storey RC frame with core wall will not experience any damage to the 475-year return period ground motion. The building will, however, suffer light to moderate damage during the 2475-year return period ground motion, but it might not be life threatening.