Structures Congress 2008: 18th Analysis and Computation Specialty Conference

The emergence of information technology (IT) is enabling engineering educators to constantly reconsider both the content and the means of delivery of modern undergraduate curricula. Online learning environments are rapidly becoming viable options for providing students with a bridge from theoretical concepts to practical engineering applications. They can be made to represent repositories of integrated tools that provide a delivery mechanism for rich learning content, advanced assessment capabilities as well as affordable access to a wide range of other resources. Online educational environments are being used at Stevens Institute of Technology (SIT) to provide undergraduate engineering students with a comprehensive laboratory experience based on content-rich and flexible remote and virtual experiments. The concept of online laboratories (i.e. remote experiments based on actual physical devices and/or virtual experiments representing pure software simulations) is being expanded through the use of information technology to create standardized laboratory, experiment, device and enhanced simulation descriptions. This enables students to run experiments that may involve multiple devices in different laboratories at various locations, perform collaborative experiments with multiple participants, and combine experiments and simulations into one integrated experience. This paper presents the results of an effort to design, implement and evaluate an integrated laboratory system for delivering both remote and virtual experiments. The methods and software modules implemented and the pedagogical approach developed for integrating them into a comprehensive student laboratory experience have been used in a sophomore-level core undergraduate course on solid mechanics taken by all undergraduate engineering majors at SIT as well as in a junior-level course on mechanisms and machine dynamics for mechanical engineering majors. Some results of the learning outcomes assessment conducted over a period of time are presented.

Laboratory experiments help students physically apply concepts proposed in lecture toward actual applications. With the involvement of senior civil engineering students, a Construction Materials Course has been restructured at Gonzaga University. This paper will illustrate how the newly designed course will enhance student learning in the Department of Civil Engineering at Gonzaga University within the Structural/Construction field. The new course will train students in the use of construction materials through lectures, site visits, literature research, and conducting experiments in the laboratory. With the addition of experimental procedures, students will respond better to the lectures, retaining more knowledge of concepts than in a traditional course setting. The course will also service other programs and be better integrated within the engineering curriculum, while emphasizing both technical and communication skills. This course differs from traditional laboratory classes, due to the fact that the students experience the highest level of active learning based on their responsibility to conduct a portion of the laboratory with minimal supervision. This paper is of interest to those who desire introducing a new course based on traditional teaching methodologies, complemented with nontraditional components, such as site visits, library literature research, and experiment design and performance.

The Gonzaga University Center for Evolutionary Algorithms (GUCEA) is an interdisciplinary and collaborative research group which brings together faculty and undergraduate students from Civil Engineering, Computer Science, and Mathematics. GUCEA students develop genetic algorithm (GA) software which is applied to difficult problems in structural engineering and mathematics. Gonzaga University is primarily an undergraduate institution, so developing a research program involves challenges unique to smaller schools. This paper presents ways to overcome these challenges, discusses the benefits of creating a research program involving undergraduate students, and highlights GUCEA findings in optimal truss design and book embeddings of graphs.

This paper introduces a new Parallel Structural (ParaStruc) analysis framework. This framework combines a newly introduced Row-wise Parallel Finite Element Analysis Algorithm (RWPFEA) with a new set of parallel numerical and communication libraries (Trilinos). The performance of the new framework is evaluated using a shared-memory parallel environment with Silicon Graphics Altix 3700 supercomputer, and using a distributed-memory parallel environment using Apple Xserve G5 compute cluster. In both cases, a 3-dimensional simulation of 50-story/10-frame/10-bay undergoing nonlinear response is carried out and the performance results are presented.

The open source finite element software framework OpenSees, developed specifically for earthquake simulation, is extended with features that enable the framework to be used for highway bridge management. At the core of the bridge management system are scripts, written in the Tcl programming language, that simulate bridge response due to live loading of vehicle traffic. Additional Tcl scripts are developed for rating calculations in order to assess and prioritize deteriorating bridges for rehabilitation or replacement. The network programming features of Tcl give the system access to databases for conducting internet-based simulation. The open source environment of OpenSees is ideal for experimenting with and developing advanced bridge rating methodologies based on state of the art structural analysis and reliability methods.

There is little argument that Building Information Modeling is the way of the future in the AEC industry; however, few would freely admit that the future is not here yet for the type of the building projects that the current wave of architecture is pushing for. It is interesting to note that for most of the conferences and presentations widely available on this hot topic, few negative opinions are voiced. We found the article by Michael Tardif, titled "Faith-based BIM," to be one of the few such articles. While with each release of the BIM upgrade, the programs improve by leaps and bounds, as of the time of writing it is still very cumbersome, inaccurate, if not plain impossible, to model more complex geometry so prevalent in modern architecture, as so noted by the authors' experience, as well as Dr. Lachmi Khemlani's honest review on the AECbytes website. Of equal, if not more, importance is the lack of standard and clear delineation of liability in the building information model between the design and construction team. For an ideal project to be conceived in BIM, there needs to be a high level of trust between various parties involved that are unfortunately not commonly found in the current atmosphere of the industry. A series of surveys commissioned anonymously by Adobe Systems in 2004 and 2006 revealed an interesting finding that was covered in Sara Ferris' article on the Cadalyst. Amid the advancement and widespread availability of computing power, the AEC industry is relying on paper more than before: CAD files are reviewed in paper format more often in 2006 than in 2004. In 2004, 33% of the responders reviewed CAD files solely in electronic format, and that number dropped to 23% by year 2006. This apparent regression of acceptance in technology was commonly attributed to concerns about document security, with a substantial size of the responders sharing documents in noneditable digital format (e.g. TIFF or PDF) at 44%, and with another 37% of the survey participants providing paper copies only. While this particular survey did not cover the topic of BIM and the sharing of data in a central model, one gets an idea of what kind of comfort level the industry is currently operating in and thus the resistance one can expect for the implementation of BIM across not just the offices of the various design consultants, but more importantly, between the design team and the construction team, and ultimately, given to the building users for long term maintenance. For our project, the design team is comfortable and has prior experience with sharing information electronically for construction. However, the decision was made that only Rhinoceros files were released to the construction team for geometry control. The BIM files were only used to produce 2D construction drawings and were not released as part of the construction document due to unclear liability issues. At the time of writing, all eyes are on the National Institute of Building Sciences on their National Building Information Model Standard project. As part of the buildingSMART Initiative, this committee is charged with the follow standards: BIM scope, Coverage of Version, Reference Standards, Business Processes, Business Rules, Data Structures and Models, Implementation Guidance, and Maturity Model. It is the hope that with better definition of BIM, the AEC industry can have a better understanding of the risk involved in its implementation and thus be able to better manage the associated risk and make an informed decision about its adoption.

The software framework OpenSees makes extensive use of object-oriented techniques, namely software design patterns, to create flexible and interchangeable modules to form and solve the equations that govern a nonlinear structural analysis. Separate interfaces for linear equation solving, root finding algorithms, constraint handlers, and time integration methods are developed in order to give an analyst full control over a structural simulation. The software design also allows a developer to incorporate numerical modules from outside the structural engineering community. Example simulations demonstrate the modeling flexibility afforded by the framework.

This paper presents an approach to introductory (junior-level) structural analysis designed to achieve a broad set of learning outcomes by combining a broad set of tools and activities in an integrated and cohesive manner. The desired learning outcomes include: 1 the ability to formulate and solve fundamental problems by hand, 2 competency using MATLAB both as a simple matrix calculator and as a more sophisticated programming tool in order to demonstrate understanding of the automation aspects of contemporary analysis, 3 basic competency using structural analysis software with high-level interfaces, 4 qualitative, behavior-oriented analysis skills in addition to quantitative skills, 5 exposure to selected classical methods of analysis. This must all be achieved within a 10-week quarter, and so it is necessary to arrange material effectively and efficiently, and to use tools and activities in such a way as to leverage strengths and avoid pitfalls.

At most universities, matrix methods of structural analysis are taught in the senior year, or in the first year of graduate study. For students taking the course, the material may be very challenging because it has been several semesters since they have taken the prerequisite courses (generally linear algebra and theory of structures). More problematically the course is challenging because many of the students have had little exposure to computer programming. At Virginia Tech, the course Computer Methods of Structural Analysis I (CEE 4404) has been designed to minimize these challenges, but still provide a firm theoretical basis in matrix structural analysis. The theoretical basis of the course is rooted in the concepts of equilibrium, compatibility, and superposition (requiring linear-elastic constitutive laws), and is presented in the context of four different levels of "Scope" within a structure. A key aspect of the course is a heavy reliance on a variety of mathematical transformations that relate the levels of scope to each other. Because of the reliance on transformations, the methodology described in this paper is termed the "Transformational Approach" to teaching matrix structural analysis. The implementation of the method is facilitated through the use of the commercial mathematics program Mathcad. Mathcad is used in two ways; first as a visual matrix manipulation tool, and second, as a framework for writing complete structural analysis programs. While a variety of programming platforms could be used (e.g. C++, C#, Visual Basic, Matlab, Mathematica) Mathcad was chosen because it is highly visual, relatively easy to learn, and is widely used in the structural engineering profession. By the end of the semester, quite complex problems may be solved with Mathcad, including any two-dimensional structure incorporating frame or truss elements. Practically any type of loading may be considered; shear deformations, rigid ends, and member end-releases may be included; and a variety of constraints may be modeled. Aside from Mathcad, no commercial structural analysis software is used in the course.

Mathcad is unique among calculation and programming environments in that it permits programming constructs, mathematical expressions, and graphical visualizations to be easily and compactly integrated. In teaching advanced structural analysis topics, these unique capabilities can be leveraged to improve student comprehension and elevate the level of analysis that can be performed directly by students. Instead of focusing on low-level programming issues, Mathcad allows students to focus instead on high-level engineering concepts. As a result, students can implement analysis procedures in Mathcad that would require far greater (and often impractical) amounts of time to code using traditional programming languages. Coursework related programming of this type is not intended to train the next generation of engineering software developers. Rather, it is intended to instill in students a fundamental understanding of structural behavior and an understanding of the inner workings of commercial software. When students program numerical analysis procedures themselves, and then compare their results to output they obtain from commercial structural analysis software packages, they gain insight into the capabilities—and limitations—of the commercial software.

It is predicted that huge earthquakes such as Tokai Earthquakes will occur in the near future. It is alarmed that ground motions with long period components ranging from several to ten seconds and with duration of nearly 10 minutes will hit Tokyo, Nagoya and Osaka in the event of such earthquakes. The Special Investigation Committee on Response to Catastrophic Disasters Including the Anticipated Tokai Earthquake (chair: Hiroshi Akiyama) of the Architectural Institute of Japan (AIJ) has been carrying out research to prepare for the occurrence of these earthquakes. The committee's report clarified the need to study the cumulative energy absorbing capacity and other performance parameters of seismic isolators, as well as the maximum value of seismic isolation layer displacement and other responses by seismic-isolated buildings under long-period earthquake motions. Further, the report demanded further verification of the critical performance of these buildings under larger earthquake motions than assumed in the design phase. By Revision of the Building Standard Law in 2000, use of isolators and dampers approved by the Minister has been stipulated in the design procedure of isolation buildings. As verifications of various characteristics by testing full-scale specimens are required to obtain the approval, the manufacturers carried out a lot of experiments, which have made characteristics of various isolation devices clearer. In this paper performance of seismic-isolated buildings against the long-period ground motions and limit performance to be necessary for seismic isolation devices are considered.

Since the Kobe earthquake in 1995, Japan has constructed a number of seismically isolated buildings based on codes and governmental appraisals. NIKKEN, a Japanese architectural and engineering firm, has designed more than 130 buildings using various seismic isolation methods, including base isolation, middle-story isolation, and retrofitting. Some projects involved improving the seismic safety of historic buildings by retrofitting them with base isolation systems while preserving the buildings' original designs. Other projects involved incorporating middle-story isolation systems into high-rise vertically multifunction buildings. This paper describes the practical designs of these examples as well as the details of their systems.

The first application of seismic isolation in the world that permits the vertical expansion of an existing structure by introducing isolation bearings between the existing structure and the addition is presented. In essence, the isolated addition acts more like a tuned mass damper than an isolated structure. The existing building which serves as the base for the isolated addition is about 825 ft long and 110 ft wide with two expansion joints at approximately the third of the length. The new addition is a continuous structure that bridges over the expansion joints and utilizes concentric braced frames for its lateral strength. This is truly a unique application of the base isolation technology where the plane of isolation has been moved from the base of the building to the roof of an existing building. A 3D non linear analysis using site specific time histories was used to demonstrate the adequacy of the base structure and extract the design parameters for the new addition.

Recent earthquakes have shown that even moderate ground shaking can produce large economic losses and major societal disruptions due to the widespread structural and nonstructural damage, and it is increasingly recognized that the traditional goal to protect against loss of life, may be insufficient. Seismic isolation offers a simple and direct opportunity to control or even eliminate damage by simultaneously reducing deformations and accelerations. While applications of seismic isolation in the U.S. have been limited to specialized facilities requiring high performance, the pace of innovation and application of isolation technology outside of the U.S. is growing at an exponential rate. A new NEES project, Tools to Facilitate Widespread Use of Isolation and Protective Systems, or TIPS, will explore innovative ideas to economically and reliably apply seismic isolation and supplemental damping to a much broader range of structures. The NEESTIPS project will fill critical knowledge gaps concerning device and system response, develop strategies and tools to reduce unnecessary costs of design and construction, and develop analysis and design tools to achieve targeted performance goals. This paper highlights the results of our strategic assessment of the knowledge gaps, economic barriers, and procedural barriers that limit the use of isolation systems in the United States. A research plan that addresses the above areas will be summarized. Through the NEESTIPS project, hybrid and shake table tests of seismically isolated building systems will be conducted at UC Berkeley and Univ. at Buffalo NEES facilities and also the E-Defense shaking table. Among the goals of the analytical and experimental program are: (1) obtain a better understanding of the behavior of full scale devices at realistic rates, (2) develop strategies for reducing design and construction costs such as relocating the level of the isolation plane or allowing the superstructure to yield, (3) identify performance limit state behavior of various isolated systems, (4) develop emerging isolation systems that can effectively achieve specific multi-level performance goals, and (5) extend performance-based design tools so that conventional solutions and isolation systems can be evaluated consistently.

Hybrid simulation is an effective structural test technique that takes advantage of numerical simulation of substructures with well-identified behavior, and experimental testing of complex and nonlinear components. As a result of combining numerical and experimental simulations, hybrid simulation is prone to both numerical and experimental errors. In this study, the dominant sources of numerical and experimental errors of hybrid simulation are examined. It is shown that analytical stability and accuracy limits of the utilized test procedures may fail to adequately predict the outcome of hybrid simulation due to the experimental errors and nonlinearities. As a result, these criteria may not be suitable for the assessment of the accuracy and stability of hybrid simulations. An alternative approach using the energy balance of the system as an overall error indicator is proposed. First, an online error monitor for experimental errors is studied that evaluates the difference between the actual experimental energy dissipation and the energy dissipation apparent from the final states at each integration step. Next, this energy error is extended to evaluate the overall energy balance of the system in order to capture both numerical and experimental errors. The effectiveness of this energy error indicator in predicting unacceptable levels of error is demonstrated through numerical and experimental simulations.

A hybrid simulation test setup is developed to investigate and validate the application of hybrid simulation to the gravity load collapse of reinforced concrete frames. The OpenFresco software framework for hybrid simulation is used in combination with an event-driven real-time predictor/corrector ensuring continuous hybrid testing. A shear-critical reinforced concrete column loaded through three dynamic actuators constitutes the physical substructure while a nonlinear ductile reinforced concrete frame makes up the numerical substructure within the OpenSees environment. This paper presents a nonlinear transformation method designed to allow for an accurate application of the loading on the specimen and a new predictor corrector scheme to eliminate force feedback oscillations. Validation of this hybrid simulation setup is achieved through a comparison with a shaking table test of the same reinforced concrete frame by performing two hybrid tests. Based on the results from the hybrid tests, procedures which are expected to improve the results of the hybrid test are presented.

The hybrid simulation test method is a versatile technique for evaluating the seismic performance of structures by seamlessly integrating both physical and numerical simulations of substructures into a single test model. Hybrid simulation results have been shown to be reliable by comparison to shake-table simulations when propagation of experimental errors is successfully mitigated. In this paper, an on-line error monitoring method for errors in the experimental setup is presented to predict the reliability of the test results in real-time. The proposed method can provide valuable decision-making information by confirming the accuracy of simulation results and generating early warnings of unacceptable levels of experimental errors. A recently completed hybrid simulation of seismic response of an innovative lateral load resisting system, a suspended zipper braced frame, is used to demonstrate the effectiveness of the proposed on-line error monitoring indicators. It is shown that a well-calibrated simulation conducted using state-of-the-art equipment can deliver accurate results.

The Peer-to-Peer Internet online hybrid test system, used in this study, is featured with equally treated substructures and geographical distribution of the substructures. Substructures are connected through the Internet. Each substructure is well encapsulated by a standard input and output interface so that it is possible to handle nonlinear substructures using both commercial finite element programs and experiments. A series of verification experiments with the P2P hybrid test system have been conducted in the past. To further explore its capability of hosting multiple tested substructures and the stability with significant nonlinearity, a one-bay, four-story steel moment frame was tested. Two column bases were taken as the experimental substructures, while the remaining superstructure was analyzed using ABAQUS. The three substructures were distributed to three locations and connected through the Internet. The system performed well without any problems, demonstrating the feasibility and stability of the proposed system to implement multiple tested substructures with significant deterioration.

Managers of civil infrastructure aim at reaching the best balance between life-cycle performance of structures and the necessary maintenance, repair and replacement costs. However, long time prediction of performance, considering the effects of deterioration and maintenance actions, can not be very accurate, and decisions must be made considering the uncertainties in initial performance, deterioration, and time of application and effects of maintenance actions. The uncertainties in performance can be reduced through inspection, non-destructive testing or load tests. However, information from these sources is not completely accurate and must be combined with existing information, in order to obtain reliable posterior information and predict more accurately future deterioration. In this paper, the application of maintenance actions on existing bridges is optimized in order to obtain maximum performance and minimum life-cycle maintenance cost. Performance is defined in terms of the condition index, which describes the effects of deterioration as can be seen by an inspector, and the safety index, which measures the safety margin of the structure. Constraints are considered for both the condition and the safety index over the entire lifetime. Several maintenance actions are considered, each defined by probabilistic effects on performance, times of application and costs of applications. The reduction in uncertainty associated with periodical inspections is considered through updating of performance profiles. Considering multiple objective functions as well as the complexity of the performance profiles under maintenance, genetic algorithms are combined with Monte-Carlo simulation, in order to obtain a set of Pareto solutions. Examples of application are presented, showing the effects of inspection and updating on the optimization of maintenance strategies.

The goal of this study is to show advantageous use of superelastic shape memory alloy (SMA) damping devices for amelioration of earthquake response in a tall steel frame structure. Damping systems discussed in this work are performance-based in that the three-story structure and its SMA bracing system are designed to minimize quantities calculated for response metrics due to a special time history of earthquake excitation. That is, a unique acceleration time-history is developed using RSPMatch2005 which adjusts an actual time-history record to fit a given design response spectrum and level of damping by means of a wavelet-based technique. In order for the frame to remain within allowable bounds of displacement and acceleration for the adjusted earthquake record, a number of SMA elements are optimized with respect to their area and location within the structure. Multiple-objective numerical optimization that simultaneously minimizes both structural displacements and accelerations is carried out using a genetic algorithm that employs NSGAII-CE. After design of an optimal SMA damping system is complete, full-scale experimental shake table tests are conducted on a large-scale steel frame that is braced with SMA elements at the National Center for Research on Earthquake Engineering (NCREE) in Taiwan. Then, a fuzzy inference system is developed to simulate the nonlinear dynamic material response of the SMA wires. Experimental and numerical results are compared in order to verify a subset of the results predicted by extensive numerical simulations. Finally, nonlinear dynamic analysis of a chevron-like braced frame is carried out to compare the performance of SMA and steel braces. It is shown that residual lateral displacement of the columns in a structure can be minimized or eliminated through a judicious selection of location and physical characteristics of the SMA devices.

A computational algorithm based on nonlinear dynamic method of analysis is derived to evaluate the local seismic energy response of structures with rigid-end offsets and deformable panel zones. This algorithm is based on the force analogy method, which uses a change in displacement field to represent the nonlinear behavior of the structure instead of the traditional method of changing stiffness. The element stiffness matrices is enhanced to incorporate the effects due to rigid-end offsets of each structural member, and a rotational spring model is tailored to include panel zone deformation at each joint. To demonstrate the applicability of the proposed numerical procedure, a six-story moment-resisting frame with active control based on the optimal linear control algorithm is used as the illustrative example. Comparison of energy response is presented for cases: (1) with and without considerations of rigid-end offsets in the derivation of the stiffness matrices; (2) with and without accounting for panel zone deformation in the model; and (3) with and without the use of active control.

The current research effort is geared towards development of an automated performance-based design environment to optimize structural and non-structural system performance and gaining better understanding of system design for minimum life-cycle costs. The automated design methodology starts with the definition of a performance-based optimization statement which can be formulated as: (1) Minimize the initial capital investment in the structural system; (2) Minimize the expected annual losses (EAL). A time-based performance assessment methodology, according to the next-generation of performance-based earthquake engineering, is used to compute the EAL of a given design. The approach considers three seismic hazard levels (2%, 10% and 50% probability of exceedence in 50 years). Inelastic time history analysis is used to evaluate structural response and to obtain engineering demand parameters (inter-story drift, and floor acceleration). Damage to the structural system, non-structural displacement-sensitive components, and non-structural acceleration-sensitive components is characterized using HAZUS fragility functions. From damage measures an estimation of the direct economic expected loss (in percentage of the building replacement cost) is obtained. The approach could be modified to consider other performance measures as casualties or downtime. A genetic algorithm (GA) is used to find optimum designs that solve the formulated optimization problem where EAL is incorporated into a GA fitness function along with initial construction cost for an example building. Sample results provide optimum designs on the Pareto-front that can be used as decision-making tool for guidance as to how a design should be adjusted to either reduce the potential for damage, or to avoid excessively costly construction practices, while maintaining desired earthquake protection.

This paper presents a robust direct adaptive control scheme for the active control of nonlinear base isolated buildings subjected to near-fault earthquakes. The control architecture is based on the premise of direct adaptive control, where the system is made to follow a desired trajectory without the need of an identifier. The control force is calculated using a single hidden layer nonlinearly parameterized neural network in conjunction with a Proportional-Derivative (PD) type controller. Stable tuning laws for the free parameters of the nonlinearly parameterized network are derived based on Lyapunov theory. To achieve good performance and to ensure that the network parameters remain bounded, initialization of the weights is required. A perturbed model is used for the initialization purposes in order to simulate the uncertainty typical of the mathematical models of civil engineering structures. The initialized parameters provide a starting point for the subsequent online adaptation of the controller under earthquake excitations. Set in the framework of adaptive control, the proposed control architecture addresses important issues related to the stability of the closed loop system and parameter bounds, issues that have previously not received the attention they deserve in a majority of the neural network based structural control approaches available in the literature. The robustness of the controller is investigated under actuator failure conditions. Simulations are performed on a full-scale nonlinear three-dimensional base isolated benchmark structure incorporating lateral-torsion superstructure behavior and bi-axial interaction of the nonlinear bearings in the isolation layer. Results are presented in terms of a comprehensive set of performance indices to reflect the tradeoffs in performance commonly associated with structural control methods.

This study looks at the influence of superelastic Shape Memory Alloys (SMAs) on the response of a single degree of freedom (SDOF) oscillator. SMAs are a metallic class of materials which possess unique properties that include stress plateaus, hysteretic damping, and large ductility with small residual strains. Superelastic SMAs have the ability to return to their original shape from strains of up to 8%. This study investigates the effects of the unique stress-strain behavior of SMAs on displacement demand, absolute acceleration, and residual deformation of a SDOF system under code level earthquakes. First, the paper investigates the differences in the response of an elastoplastic (EP) and a superelastic system. A double trigger-line model is used to capture the SMA's hysteretic properties. Previous research has noted that SMAs have varying mechanical properties depending on thermomechanical processing. Therefore, the effects of several different SMA properties on the structural response are investigated. This paper investigates the importance of damping on the response of the SDOF for a suite of ground motions. The results show that shape memory alloys are much less effective in reducing the peak response of a SDOF system in the short period range, as compared to an EP system. As the strength reduction factor increases, peak displacements increase for both the SMA system and the EP system. However, the EP system, in some cases, has large residuals, compared with perfect recentering for the SMA system. Finally, the paper takes an initial look into the effects of using a parallel system consisting of a SMA and an EP element. The parallel system has over three times the energy dissipation in the hysteresis as the SMA system while maintaining significant recentering capabilities compared to the EP system.

Recently, structural problems in tall buildings and power plants have become evident as a result of long-period vibrations due to earthquakes. In order to ameliorate these problems a number of damping devices have been developed by many researchers. One of the authors also has developed the Mechatro damper which has a damping force that is imposed on the structure by a power generator. The force of the damper is proportional to its velocity. On the other hand, the mechanical snubber utilizes an inertial force derived from a flywheel that is proportional to acceleration. That is, the device has a so-called series inertia mass. The authors developed a vibration cut-off system using water or a functional fluid that acts as a series inertia mass, and the effects of vibration cut-off have been confirmed. In a previous paper the authors proposed a new vibration control device (VCD) that also generates power. The VCD has an inertial force created by an inertial disk and damping force provided by energy dissipation of the power generator. This device acts as a semi-active damper, and functions as a fail-safe mechanism under controller malfunction. A prototype model was fabricated and vibration tests were carried out in a laboratory. The focus of this paper is application of the prototype large-scale VCD to a real structure. Two new VCDs that have a force capacity in the range of 15 kN have been manufactured and shipped to the National Center for Research on Earthquake Engineering (NCREE) in Taiwan for testing on a three-story structure that is excited by a large shake table. In order to investigate dynamic properties of the VCD, performance tests are carried out and the resisting force characteristics of the device are measured. Next, vibration tests are conducted on the structure by a shake table with the VCDs installed. Seismic responses at each story level are measured for the Imperial Valley, El Centro north-south component of motion. A control law that is based on minimizing the Lyapunov function is used along with bang-bang operation of the VCD. The effects of vibration suppression using the VCD are shown to be confirmed.

A numerical and large-scale experimental evaluation of super-elastic, semi-active damping for structures subjected to seismic loadings is offered in the current study. An array of shape memory alloy (SMA) wires is installed in concurrence with multiple magnetorheological (MR) dampers to provide recoverable hysteretic damping in a three-story benchmark structure. Optimal distribution of SMA wire is determined through a non-dominated sorting genetic algorithm with controlled elitism (NSGA2CE). For management of MR damper resistance levels a fuzzy logic controller (FLC) is also generated using the NSGA2CE algorithm. Numerical and large-scale experimental tests reveal favorable hysteretic behavior that results from use of the combined SMA and MR devices. The total response of the structure shows significant reductions in peak inter-story drifts and peak acceleration responses, when compared to the instance where SMA braces are installed in the structure and MR dampers are operated in a passive mode.

Current activities within the NEESWood project focus on the development of a performance-based seismic design (PBSD) philosophy for woodframed structures. One approach to improving the performance of such structures is to utilize an advanced seismic protection system (e.g., a seismic isolation or damping system). The feasibility of implementing a seismic damping system within a woodframed building has been demonstrated by the authors via shake table testing of a full-scale woodframed benchmark structure at the University at Buffalo NEES site. Testing of prefabricated modular damper walls was conducted as Phase 2 of the benchmark testing program within the NEESWood project. The research presented herein extends the Phase 2 testing by assessing the seismic performance of the benchmark structure based on PBSD procedures that are under development within the NEESWood project. This includes assessing the performance based on results from nonlinear response-history analyses and comparing to acceptable performance levels/metrics being articulated as part of the NEESWood project.

The design performance objectives implicit in U.S. building codes currently differ for fixed-base and base-isolated buildings. The imposed standard for a fixed-base building is comparable to "life safety", while the imposed standard for an isolated building is comparable to "immediate occupancy" or "operational". For example, fixed-base buildings are permitted a force reduction factor R of up to 8, while isolated buildings are limited to R no larger than 2. The base shear demands in fixed-base buildings are reduced considerably by allowing superstructure inelasticity, whereas the superstructure of an isolated building remains essentially elastic due to overstrength. Consequently, the superstructure design forces in an isolated building are often larger than in a comparable fixed-base building. Factoring in the added design, material, and testing costs; base isolation in the U.S. has become an expensive technology that is considered only when the owner is willing to pay a cost premium for very high performance. Structural systems should be evaluated or compared relative to a consistent performance objective, such as life safety or continued occupancy. In particular, relaxing the design standards for isolated buildings may lead to improved cost-competitiveness, while such systems still potentially allow for a substantial performance advantage. In recent years, a performance-based design approach has been under development in the U.S. Performance-based earthquake engineering (PBEE) encourages owners to select appropriate performance objectives for the structural and non structural building components and systems in different events or considering the composite probabilistic seismic hazard. The new approach, developed by PEER and being adapted for practice by ATC-58, specifies performance in terms of probabilistic losses (casualties, repair costs, downtime). When performance-based engineering matures, designers will employ the latest design and analysis techniques to create efficient designs that meet specified performance objectives, and building owners can comparatively evaluate base isolation and fixed-base design with reference to a quantitative performance objective. In a previous related study, Ryan et. al. analyzed fixed-base and base-isolated structures with identical fixed-base periods and responding with identical deformation ductility. A comparative performance measure (CPM) was developed to assess relative performance — quantified by structural drift and acceleration — of the comparable isolated and fixed-base buildings. This approach restricted comparison to structures with identical ductility demands, and did not allow identification of the best design considering several systems, performance objectives and economy. The present study presents a methodology to systematically evaluate the relative performance of fixed-base and base-isolated buildings. To compare the relative performance of multiple systems, including fixed-base and base-isolated buildings, a performance index (PI) is developed. The methodology can be used as desired; e.g., to identify the best performing system, to identify the lowest cost system that meets the performance objective, or to identify a desirable combination of performance and cost. In this study, evaluation of performance is restricted to engineering demand parameters such as story drift and floor acceleration. If the vision of PBEE is realized, performance will ultimately be described in terms of damage and expected losses. As a short term goal, we plan to extend this study by applying current damage and loss estimation techniques to the structures examined here.

The performance of restrainer cables made of Nitinol shape memory alloys (SMA) is evaluated using a large-scale test setup. The performance was evaluated on both a ¼-scale in-span hinge model, and a ¼-scale continuous bridge model, using a range of ground motion levels. The results indicate that the SMA restrainer cable was extremely effective in limiting the hinge displacement, while simultaneously limiting the acceleration due to impact. Using the results from the experimental tests, a probabilistic seismic risk assessment is performed. Vulnerability curves are developed for a bridge model with traditional steel restrainer cables and SMA restrainer cables. The results show that for the higher damage states, the SMA restrainer cables become exceedingly more effective in limiting damage or collapse when compared with traditional steel restrainer cables. Finally, a benefit-cost study is performed, and a comparison between SMA cables and steel cables is presented. The results show, despite the increase in cost for SMA restrainer cables, the enhanced performance results in a higher benefit-cost ratio.

An objective in the design of seismically isolated structures is the selection of bearing properties so that optimal performance is achieved over a range of excitations and performance metrics. One dilemma in the design of isolation systems is that, to withstand very severe or near-fault motions, bearings often become so large, stiff and strong that they provide little isolation during moderate events. Experimental and numerical investigations are presented to characterize a new multi-stage isolation bearing, capable of progressively exhibiting different hysteretic properties at different stages of response and the feasibility of targeting these properties to achieve specific performance goals for a range of ground motion intensities and structural dynamic characteristics. This newly-developed triple pendulum isolator incorporates four concave surfaces and three independent pendulum mechanisms. Pendulum stages can be set to address specific response criteria for moderate, severe and very severe events. In particular, the tradeoff between limiting very rare isolator displacement demands and inducing high-frequency floor accelerations is examined for a range of levels of seismic hazard. Nonlinear dynamic analyses of realistic building systems are presented, including a thorough description of key structural demand parameters such as inter-story drift and floor acceleration spectra. Recommendations for design based on achieving target performance objectives under multiple seismic hazard levels are given.

This paper aims to use the performance-based evaluation approach to investigate the performance of highway bridges with isolation devices. Fragility functions of un-isolated and isolated bridges are derived and compared to study the efficiency of various isolation devices with different combination of strength and stiffness. A simplified 2D bridge model representing the transverse response of a typical highway bridge is adopted where the fiber section element is used for bridge columns while bilinear element is used for isolation devices. The fragility curves are derived by either PSDA or IDA method based on simulation results from the nonlinear time history analysis using 250 earthquake motion records. Damage criteria for both piers and isolation devices are established to relate the component response quantities to global damage states of bridges. Extensive parametric study is also conducted to evaluate the effect of mechanical properties of isolation devices on the damage probability of isolated bridges. The study identifies the optimum combination of mechanical parameters of isolation devices, which can serve as practical guide for isolation device designs.

This paper presents a displacement-based seismic design methodology for steel concentrically braced frame (CBF) systems with a special type of bracing element termed self-centering friction damping brace (SFDB). The SFDB is a passive energy dissipation device with its core re-centering component made of stranded superelastic Nitinol wires while enhanced energy dissipation mechanism of the SFDB is achieved through friction. Compared with conventional braces for steel frame buildings, SFDB frame has a few desirable performance characteristics such as minimal residual drifts and its ability to withstand several design basis earthquakes without the need for brace replacement. A displacement-based design procedure for proportioning SFDB is proposed in this paper. A 3-story design example is presented to illustrate the design method. Nonlinear pushover and time history analysis of the 3-story CBF building is performed to validate the effectiveness of the proposed design method. The results of the nonlinear time history and pushover analysis show that with careful design the SFDB frame can achieve a seismic response level comparable to that of the BRB frame while having significantly reduced residual drifts. The SFDB thus has a potential to establish a new type of CBF systems with self-centering capability.

In this study, a multi-degree-of-freedom numerical model is created to represent the elastic behavior of a typical Chilean two-family, low-cost, confined masonry house. Ambient vibration testing was performed to determine the fundamental natural frequencies of the structure in the longitudinal and transverse directions through a non-parametric system identification technique. Based on the ambient vibration data, an estimation of equivalent viscous damping of the structure is made using the bandwidth method. The stiffness values are estimated using the natural frequencies measured from the ambient vibration testing. A number of passive isolation devices are considered as potential isolation systems for the structure. Each device is numerically modeled using fuzzy logic. The fuzzy models are created using experimental and analytical data to predict the force exerted on the structure by the device. The passive isolation systems considered include friction pendulum systems, high-damping rubber bearings, and two hybrid systems involving use of high-damping rubber bearings and shape memory alloy wire. Seismic performance of the structure in the longitudinal direction, augmented with each of the isolation systems mentioned above is compared with the performance of the traditionally-constructed structure using several performance indices.

This paper deals with the cyclic elastoplastic analysis and stability evaluation of steel braces of hollow sections subjected to axial tension and compression. The inelastic cyclic performance of cold-formed steel braces of circular and box hollow sections is examined through finite element analysis using the commercial computer program ABAQUS. First some of the most important parameters considered in the practical design and ductility evaluation of steel braces of tubular hollow sections are presented. Then the details of finite element modeling and numerical analysis are described. Later the accuracy of the analytical model employed in the analysis is substantiated by comparing the analytical results with the available test data in the literature. Finally the effects of some important structural and material parameters on cyclic inelastic behavior of steel braces are discussed and evaluated.

Methodologies for solving nonlinear equilibrium equations are reviewed in this article. Two basic numerical procedures, the pure incremental and the direct iteration methods, are briefly discussed. Then the most frequently used increment-iterative methods are presented and their limitations are discussed. These techniques are the Newton-Raphson and the displacement control method. One of the advanced nonlinear solution procedures, generalized displacement control method, is outlined, and its algorithm is also presented. The generalized displacement method shows that it is a robust numerical technique for solving nonlinear structural problems which may include softening, stiffening behavior and in the bifurcation vicinity of critical points. The elstica problem which is a classic highly geometrically nonlinear problem is presented to show the versatility of the generalized displacement control method for the solution of highly nonlinear problems.

This paper investigates in depth the dynamic responses of soil-foundation-structure systems using dimensional analysis in order to evaluate the significance of soil-structure interaction effects when subjected to near-fault ground motions. A soil-foundation-structure interacting (SFSI) system is idealized into a simplified 2DOF system where linear structures with flexible foundations are considered. Through rigorous dimensional analysis, the response of superstructure is presented in terms of the dimensionless Π-terms. This approach brings forward the self-similarity, the invariance with respect to changes in scale or size, which decisively describe the interactive behavior of a foundation-structure system. Extensive simulations have been conducted on various SFSI systems subjected to pulse-type ground motions to numerically describe the structural response, i.e. dimensionless relative displacement, as a function of distinctive properties of superstructure; foundation and input motion. Results show that soil-structure interaction is insignificant for a variety of structure-foundation systems in practice. However, there are conditions under which the response of foundation-structure system is significantly higher than that of the fixed-base structure.

This paper presents a nonlinear finite element structural model — "PB3D" to simulate the seismic response of Japanese post and beam (P&B) wooden buildings. A versatile mechanics-based "pseudo nail" wall model is used to represent the nonlinear load-drift characteristics of the shear walls under lateral loads. The roof/floor diaphragms are modeled as structural frames with equivalent diagonal bracings to consider the influence of the roof/floor in-plane stiffness on the lateral force distribution among the walls. This approach significantly reduced the overall system degrees-of-freedom while allow the model to represent the key characteristics of the system under lateral loads. Experimental studies on shear walls and floor diaphragm were conducted to calibrate the wall models and the equivalent floor diaphragms. The model prediction agrees well with the shake table test results of a single-storey P&B building. The "PB3D" model provides an efficient tool to evaluate the seismic performance of general Japanese P&B wooden buildings.

Damping is dependent on many variables and complex mechanisms that are not yet fully quantifiable in the design stage. This has led to generic assignments of viscous damping usually based on material type and at best based on height with reference to existing databases. This study will employ recent full-scale observations to demonstrate the role of the structural system's dominant deformation mechanism — frame racking vs. cantilever action — in energy dissipation capability. Specifically, it is shown that frame racking and other shear deformations dissipate more energy than the axial shortening associated with cantilever action. As such, the ratio of frame racking to cantilever action may offer a more robust and intuitive parameter better suited for use in predictive viscous damping models. This study serves as the first step in the development of such models, with a specific focus on serviceability design.

This paper first discusses the amplitude dependency of damping ratio. It is emphasized that there is no evidence of increasing damping ratio in the very high amplitude range within the elastic range of main frames, unless there is damage to secondary members or architectural finishing. The damping ratio rather decreases with amplitude from a certain tip drift ratio defined as "critical tip drift ratio". Next, the feasibility and efficiency of two simple and user-friendly but accurate damping estimation techniques are discussed: the Frequency Domain Decomposition technique; and the Multi-mode Random Decrement technique. Some full-scale examples demonstrating the damping estimation efficiency of both techniques are also shown.

A database of 4000 vibrations measurements from 1000 buildings was used to derive models for building period and damping. Herein, we focus on the dominant trends in the median and standard deviation of the damping with respect to basic building parameters such as height, material and frame type. We also outline the statistical techniques used to identify the most important building parameters and separate the sources of variability, while avoiding model over-fitting.

Wall-type or shell-type prestressed concrete structures, such as prestressed concrete I-girders, box girders, nuclear containment vessels, offshore structures, shear walls, etc can be visualized as assemblies of membrane elements. Their behavior can be predicted if the behavior of the membrane elements is thoroughly understood. Since prestressed concrete structures are now widely used, a research project was conducted to investigate the behavior of prestressed concrete elements subjected to shear action using the Universal Panel Tester available at the University of Houston. This paper reports the Softened Membrane Model for Prestressed Concrete (SMM-PC) (Wang, 2006) developed from this study. The SMM-PC generalized the previously developed Softened Membrane Model (SMM) (Zhu, 2000; Hsu and Zhu, 2002) for reinforced concrete and can be used for prestressed as well as reinforced concrete. It is also a rational model like the SMM as both of them satisfy Navier's principles of mechanics of materials (stress equilibrium, strain compatibility and constitutive laws of materials). The new SMM-PC includes the following three new constitutive laws: (1) A constitutive law of concrete in tension that includes the decompression stage. (2) A new prestress factor W_p proposed for incorporation into the softening coefficient of the constitutive laws of concrete in compression. (3) A smeared (average) stress-strain relationships of prestressing strands embedded in concrete. In this paper the SMM-PC has also been extended to include cyclic behavior, thereby creating a Cyclic Softened Membrane Model for Prestressed Concrete (CSMM-PC). This has been accomplished by implementing the cyclic behavior of reinforced concrete previously developed at the University of Houston through the Cyclic Softened Membrane Model (CSSM) (Mansour and Hsu, 2005a, b). The CSMM-PC is implemented into a non-linear finite element program based on the framework of Opensees (Fenves, 2005) to predict the behavior of prestressed concrete structures under cyclic loading. The developed program is validated by analyzing a prestressed concrete beam tested under montonic loading, and comparing the analytical results with test data.

In this paper, a recently developed meso-scale model, called the Lattice Discrete Particle Model (LDPM), is extended in order to include the effect of the loading rate on concrete strength and fracturing behavior. As verified experimentally, the rate dependence of concrete behavior is caused by two different physical mechanisms. The first is a dependence of the fracture process on the rate of crack opening, and the second is the viscoelastic deformation of the intact (unfractured) cement paste. For concrete both mechanisms are important but the former dominates at extreme strain rates under impact. In this study, the first mechanism is described by the activation energy theory applied to the ruptures that occur along the crack surfaces. The developed model will be calibrated and validated on the basis of experimental data available in the literature. In particular numerical simulations of 1) impact tests on prismatic specimens in compression, and 2) dynamic Hopkinson bar tests in tension, are carried out. The numerical results show a very good agreement with the experimental results from both qualitative and quantitative points of view.

Impact loads in structural design are often related to vehicular impact, partial structural collapse, construction accidents or blast fragmentation. For design purposes, impact loads are often treated as static loads multiplied by a dynamic increase factor; however, this method can overly simplify the event. More sophisticated analyses use energy methods, contact elements and explicit finite element method to determine the impact behavior; however, the complexity of these methods is unattractive for most structural engineers. These analytical methods are either overly simplified or excessive complex, therefore, simplified methods must be developed that account for the complexity of the phenomenon while still providing an easy tool for structural engineers. The purpose of this paper is to identify parameters in an impact scenario associated with free falling particles impacting a stationary object. These parameters are used to determine a time-dependent load and a time-dependent mass for simplification to a single degree of freedom system.

A computer based iterative numerical procedure has been developed to analyse reinforced high strength concrete columns subjected to horizontal wave loads and eccentric compression by taking the material, geometrical and wave load non-linearity into account. The behaviour of the column has been assumed, to be represented by Moment-Thrust-Curvature relationship of the column cross-section. The formulated computer program predicts horizontal load versus deflection behaviour of a column up to failure. The developed numerical model has been validated with the available experimental results of 40 concrete columns of various slenderness, structural properties and tested under various compressive thrust levels by other researchers. The predicted values are having a better agreement with experimental results. A simplified user friendly hydrodynamic wave load model has been developed by the authors based on Morison equation supplemented with a wave slap term to predict the high frequency non-linear impulsive hydrodynamic loads arising from steep waves, known as ringing loads and enable practical applicability without involving tedious computational efforts. A computer program has been formulated based on the model to obtain the wave loads at all discretised nodes, along the length of column from instantaneous free water surface to bottom of the column at mud level and to calculate the non-dimensional load coefficient for all nodes. The lateral strength and deflection behaviour of the given columns till failure can be predicted. This paper describes the procedure of numerical analysis of high strength concrete columns, developed from continued research on assessment of wave impact loads and strength of reinforced concrete columns subjected to lateral loads.

This paper studies the effects of random properties on the strength and system reliability of steel frames controlled by frame stability. Probabilistic simulation is used to analyze the behavior of three structures—a low-rise industrial frame; a grain storage bin; and a two-story, two-bay frame. OpenSees is used to perform the non-linear structural analyses within a probabilistic simulation code. The probabilistic analyses consider randomness in yield strength, elastic modulus, residual stresses and geometric imperfections. The simulation results provide probability distributions of frame strength for various random frame properties. The simulation results are used to calculate system reliabilities and compare them to typical target reliabilities. The results of the analyses provide insight into the relationship between specific failure modes and frame properties. The paper also discusses the difficulties of developing probabilistic codes for system-based design criteria and the need for continued probabilistic simulation of non-linear behavior in steel frames.

This paper presents a novel algorithm to simulate rebar-concrete interaction when concrete is modeled using the Lattice Discrete Particle Model (LDPM), a recently developed three-dimensional meso-mechanical model. In the LDPM formulation, the mesostructure of concrete is simulated by an assemblage of particles interacting through nonlinear springs. Each particle represents a coarse aggregate piece with its surrounding mortar. The rebar-concrete interaction algorithm consists of a constraint element that treats the interaction of discrete particles close to the rebar with adjacent rebar finite elements. Bond constitutive equations provide relationships for computing interface forces given the relative displacements between particles and rebars. These equations implement the complex physical mechanisms that take place in the thin concrete layer surrounding steel rebars, including the formation of oblique cracks, dilation due to slippage, friction, etc. The complete formulation (LDPM, rebars, and bond interaction) is implemented in the framework of the object oriented dynamic finite element code MARS. Calibration and validation activities are being performed using a series of pull-out experiments recently conducted at the US Army Engineer Research and Development Center (ERDC) in both quasi-static and dynamic regimes. Three examples consisting of highly dynamic `impact' pull-out tests are presented.

Stability analyses of Lagrangian and semi-Lagrangian reproducing particle methods using various domain integration methods are performed. The von Neumann stability analysis shows that both Lagrangian and semi-Lagrangian reproducing kernel discretizations of equation of motion are stable when they are integrated using stabilized conforming nodal integration in the weak forms. On the other hand, integrating the weak form of semi-Lagrangian equation of motion with a direct nodal integration yields an unstable discrete system which resembles the tensile instability in SPH. Stable time step estimation for Lagrangian reproducing kernel discretization shows enhanced stability when weak form is integrated by stabilized conforming nodal integration compared to that using direct nodal integration or 1-point Gauss integration. Penetration simulation is performed to demonstrate the applicability of the proposed method to large deformation and fragment impact problems.

The coupled 2D Godunov-variational difference approach is presented for investigation of soil-structure interaction due to a nearby explosion. It is based on the relationships of the shock and rarefaction waves with finite difference equations of the shell motion using an iterative method. It allows the reduction of the contact problem to the self-similar symmetrical Riemann problem. A numerical example of a lined tunnel buried in soil that is subjected to a buried nearby explosion is provided.

Accurate estimation of wind forces on large highway sign structures is important due to possible structural failure of these sign structures under strong winds. The use of large panels for the traffic signs is increasingly more common as we attempt to better manage the highway traffic flow and automate the highway systems. In order to be able to predict behavior of these structures, accurate knowledge of the forces on these structures must be known. An important aspect of wind loads on highway structures that is generally not appreciated is the effect of interactions among the panel structures on these forces. The main objective of the proposed study is to use Computational Fluid Dynamics (CFD) tools to determine the wind loads and pressure distributions by accurate numerical simulations of the air flow characteristics around large highway sign structures under severe wind speeds conditions. The pressure distributions are needed in case a detailed structural dynamics analysis will be performed. In particular, the present study investigates the effect induced by the presence of back-to-back signs.

Full-scale studies, wind tunnel experiments and finite element modeling were used for the study of a wooden low-rise building subjected to wind loads. The conclusions of this study can be summarized as follows: 1) The pressure distribution comparison between the wind tunnel and the full-scale results shows good agreement. Some discrepancies can be justified by the high fluctuations of the wind direction in the full-scale records. The peak pressure coefficient comparison is characterized by higher discrepancies. 2) The wind tunnel / full-scale general agreement allows the use of wind tunnel data for numerical simulation. 3) The comparison between the full-scale load cell readings and the base reactions computed by the finite element analysis made in the form of force coefficients shows good agreement, as far as mean values are concerned. Higher values of forces have been found by using the measured pressure coefficients on the building envelope in comparison with those recorded directly by load cells placed on building foundation. 4) The experimental critical local pressures (suctions) tend to be higher (lower) in comparison to the code suggested values. Topography and surrounding structures effects can justify these discrepancies. 5) Field pressure data are, in some cases, significantly higher than the corresponding ASCE 7 values for components and cladding. The study confirms that full-scale structural monitoring is very difficult, time-consuming, yet necessary task. The ongoing collection of data will lead to more complete outcomes. Furthermore, roof load cells have been installed on the base of roof trusses and a more detailed picture of the structural response will be available in the near future.

Hollywood has been tantalizing the public and tormenting wind engineers with stunning visual effects and the idea of computational wind tunnels. The reality is the physical atmospheric boundary layer wind tunnel has nearly infinite resolution of fluctuating turbulent properties. The equivalent resolution with current turbulence modeling techniques will require billions and perhaps trillions of cells and computing resources at the limits of our imagination. If the true time varying information obtained from an atmospheric boundary layer wind tunnel is out of our reach within the near future with computational methods, what is within our reach? Qualitative analysis (i.e., design A is better than design B) seems the nearest to reality. Additionally, the indoor environment, (i.e., laboratories, offices, auditoriums) while still of large scale are often driven by mean flow properties with secondary turbulent effects. For the indoor environment CFD is being used (and validated!) to evaluate occupant comfort and safety in the pre-build phase. These modern models of indoor spaces are still requiring millions of cells and mid-scale parallel computing. It is now evident that the questions architects and HVAC engineers are asking of CFD cannot be resolved by single desktop computers. The trend is toward integrated industrial applied CFD comparable to that used by the automotive and defense industry. This operation mode enables the smooth transition of computational models from CAD though solvers to the post-processing and client presentation.

Numerical (CFD) and physical (laboratory) experiments were conducted in parallel to simulate downburst and tornado-like flow fields in order to determine the effects of these thunderstorm winds on buildings and structures. The physical experiments served to: (i) benchmark the CFD results and (ii) to help in designing the next generation of downburst and tornado simulators in order to physically test scaled structural models. Downburst jet-like simulations showed the complex vortex structure of these winds for which the maximum velocity happens very close to the surface. Above a certain critical Reynolds number the downburst flow is rather independent of length or velocity scaling and only dependent on the terrain roughness. This allows the scaling of numerical or laboratory downburst-like experiments to full scale phenomena. The flow field was then applied to estimate steady-state responses of tall buildings and it was determined that in certain conditions the downburst winds may become dominant when compared to synoptic, boundary layer winds. Tornado-like simulations showed that the wind field is highly dependent on the swirl ratio. Several swirl ratios have been investigated both numerically and experimentally and results compared well. Moreover, by matching high swirl ratio numerical results with full scale Doppler radar measurements a preliminary relation has been established between the swirl and the Fujita scale.