Journal of Structural Engineering

Notwithstanding powerful computational simulation methodologies available today, there remain significant challenges: the plane-section assumption of popular fiber models, the scale limitations of sophisticated microscopic methodologies such as particle-lattice models, and the difficulty in describing structural damage in actual physical terms. Here, the authors validated a structure-independent parallel platform that tackles these challenges. Nonlinearity is captured by novel microphysical mechanisms: a multidirectional smeared crack model, a tribology-inspired three-dimensional (3D) interlocking model, a topological transition-based steel bar model that captures progressive buckling, and a general confinement model exploiting nonlocal information (i.e., mesh-objective proximity to adjacent reinforcements and boundaries). These innovative features are made possible by virtue of optimized parallel algorithms. The validation and application span a variety of RC elements: columns with a hollow or solid section, rectangular walls with or without opening, and H- or T-shaped multistory walls. Importantly, all simulations embrace realistic geometry and reinforcements, but they require only two material properties and no structure-dependent calibrations. The universality and efficiency of the platform will feed more physical damage information to fragility functions, and also give rise to a powerful tool for next generation performance-based engineering, which calls for a multitude of structural analyses.

Although a few fire experiments have been carried out on load-bearing cold-formed steel (CFS) wall systems, the understanding of their fire performance is still limited, and further parametric analysis on the basis of such experiments is expensive and time-consuming. This paper presents a simplified numerical approach to predicting the thermal and mechanical responses of CFS wall systems in fire. Thermal physical property experiments were carried out to measure the essential material properties. With those material properties as inputs, a one-dimensional thermal response model was developed to predict the heat transfer across the cross section of CFS wall systems. Both heat convection and radiation were considered in the thermal boundary conditions. The governing equation was expressed in the form of implicit finite-differential equations and solved using the Gauss-Seidel method. The model predictions of temperature responses of CFS wall systems were in good agreement with the measured temperature responses in the fire experiments and also compared well with the modeling results in literature with improved efficiency and convergence in computation. In addition, a thermomechanical response model was developed to predict the time-dependent lateral deflection and fire resistance time for CFS wall systems in fire and was validated by the experimental results in literature. All these studies provide an efficient approach for the fire performance analysis of load-bearing CFS wall systems.

The seismic response of midrise light wood-frame buildings incorporating portal frames and shear walls were investigated via numerical simulation. Because portal frames and shear walls exhibit different ductility characteristics, of specific interest in this paper is the seismic force-modification factor for the combined building system. An approach to estimate the ductility-related force-modification factor, $R_d$, for the hybrid light wood-frame building containing two types of lateral load-resisting systems with different $R_d$ was proposed. One-, four-, and six-story buildings with layouts that rely on shear walls only or a combination of shear walls and portal frames were designed using the $R_d$ estimated from the proposed approach. Frequency and nonlinear time history analyses were performed on these multistory hybrid buildings using a modified macro element model to represent the lateral-resistant element in finite-element program, ABAQUS. The results show that the use of a $R_d$ value higher than the lowest $R_d$ factor of the two lateral load-resisting systems is justified.

In finite-element (FE) model updating using response surface (RS) models as surrogate, the procedure of finding an appropriate design to build the RS models requires a number of trial-and-error approaches with different designs and subset models. To address this issue, a procedure is proposed in this paper to design and fit proper RS models in FE model updating problems. Also, formulation of the problem in an iterative format in time domain is proposed to extract more information from measured signals and compensate for the error present in the regressed models. This procedure is applicable to both linear and nonlinear models under static or dynamic analysis. The proposed methodology is applied to a numerical case study of a steel frame with global nonlinearity. Appropriate design and model order are successfully established and optimization in time performs well in all the simulated scenarios. Finally, the performance of this method in presence of measurement noise is compared with a method based on sensitivity analysis in terms of required time and accuracy.

This paper presents the application of an advanced modeling strategy for the nonlinear analysis of structures with unreinforced masonry (URM) components under blast loading. This approach enables the investigation of the nonlinear dynamic response of large structures with URM walls, accounting for the mechanical and geometrical characteristics of URM components, the coupling between the in-plane and out-of-plane response as well as the interaction between URM panels and the other parts of the considered structural system. According to the utilized strategy an URM wall is described by a parent structure, which consists of super-elements representing the partitioned subdomains, allowing effective parallelization of the nonlinear structural analysis simulation. Each partition is represented by a detailed 3D mesoscale model, which uses an advanced 2D nonlinear interface element that allows the representation of crack propagation in URM elements. Furthermore, the macroscale model considers only the partition boundaries of the mesoscale descriptions and specific macro-elements are introduced to reduce the number of freedoms leading to further enhance the computational savings. Several examples are presented in the paper to validate the proposed approach and to demonstrate its major computational benefits in simulating the response of structures with URM walls subjected to blast loading.

This paper presents a computational assessment of the performance of steel gravity framing systems with single-plate shear connections and composite floor slabs under column loss scenarios. The computational assessment uses a reduced modeling approach, while comparisons with detailed model results are presented to establish confidence in the reduced models. The reduced modeling approach enables large multibay systems to be analyzed much more efficiently than the detailed modeling approaches used in previous studies. Both quasi-static and sudden column loss scenarios are considered, and an energy-based approximate procedure for analysis of sudden column loss is adopted after verification through comparisons with direct dynamic analyses, further enhancing the efficiency of the reduced modeling approach. Reduced models are used to investigate the influence of factors such as span length, slab continuity, and the mode of connection failure on the collapse resistance of gravity frame systems. The adequacy of current structural integrity requirements is also assessed, and based on the computational results a new relationship is proposed between the uniform load intensity and the tie forces required for collapse prevention.

Most industrial and nuclear facilities rely on reinforced concrete structural walls as their primary seismic lateral-force-resisting components. In nuclear facility structures, squat walls commonly have an aspect ratio lower than 0.5 and are designed to be thick for radiation shielding and blast and fire protection. The combination of a squat and thick wall geometry causes very high wall stiffness and strength. However, there is significant uncertainty about expected strengths, deformation capacities, and failure modes of these walls in earthquake load sequences, such as main-shock/aftershock combinations. Hybrid simulation is an effective experimental method to examine these issues: it enables simulation of the seismic response of squat and thick shear walls without the need to recreate the often very large mass associated with the remainder of the prototype structure. A hybrid simulation program utilized two shear wall specimens to investigate the variations of squat wall responses caused by different earthquake magnitude sequences. The results of these simulations indicated that different earthquake magnitude sequences do not have a significant effect on the force-deformation response and failure mode sequence of squat reinforced concrete shear walls.

A full-scale earthquake simulation of a five-story steel moment frame building at E-Defense, the world’s largest single shake table facility in Miki, Japan, in August 2011 provided realistic data to improve and validate current modeling approaches for steel structures with composite floor systems. In this paper, computational models of the tested fixed-base specimen and the specimen isolated with triple pendulum bearings were developed and validated using the test data, and the influence of various modeling assumptions on the accuracy of the numerical prediction was investigated. The results demonstrate that the displacement of the isolation system can be predicted accurately without a well calibrated superstructure model. However, the response of the isolated structure, especially when subjected to 3-dimensional motions, is quite sensitive to the frame modeling assumptions and the superstructure damping model. The modeling assumptions that most accurately predict the response of the tested specimen are described in this paper.

This paper presents a three-dimensional (3D) numerical model for capturing the time-dependent behavior of layered wood-concrete composite beams with notched shear key interlayer connections, subject to long-term static load and simultaneous exposure to naturally changing ambient environmental conditions, including temperature and relative humidity. For the wood material, the research extends the formulation of a previous rheological model, originally proposed for uniaxial stress state of wood exposed to moisture variations, to 3D finite-element modeling with a multiaxial stress state. For the concrete material, a standard creep model for uniaxial stress states was approximated using a generalized Maxwell’s model and then expanded to 3D formulations. The 3D model described in this paper was implemented via external user subroutines in the commercial software ABAQUS. Finite-element models for a wood-concrete composite beam with notched shear key were developed to model the diffusion problem of moisture content and the heat transfer when the beam is exposed to the environment. The moisture content, temperature profiles, and long-term static loading are used as input parameters for the time-dependent analysis problem involving the expanded material models. The results of numerical models are compared and verified with available experimental results as well as with numerical results obtained using a more simplified one-dimensional (1D) beam model previously developed. Although the improvement in accuracy compared to the 1D model is insignificant, the 3D model has the advantage of not requiring mechanical and rheological characterization of the connection, which can only be carried out via expensive short- and long-term experimental tests.

Past research has shown that blast-loaded steel stud walls exhibit a range of different failure mechanisms depending on the stud and track section properties, connection details, and sheathing characteristics. To date, few studies have addressed the computational modeling of these systems. Because large-scale blast tests are expensive and logistically difficult, computational models are needed to evaluate different design options prior to carrying out full-scale experiments. In this paper, the authors present finite-element models that capture the peak load and deformation capacity of steel stud wall systems, accounting for the failure modes observed in past testing. The proposed models strike a balance between level of refinement and computational efficiency. These models were validated against data collected from an extensive laboratory testing program. Based on observations from both the lab tests and computational simulations, recommendations are given for improving the large-deformation response of these systems. Through simple and inexpensive design modifications, the capacity of steel stud walls can be made to far exceed existing response limits.

This paper proposes a reliability-based framework for quantifying structural robustness considering the occurrence of a major earthquake (mainshock) and subsequent cascading hazard events, such as aftershocks that are triggered by the mainshock. These events can significantly increase the probability of failure of buildings, especially for structures that are damaged during the mainshock. The application of the proposed framework is exemplified through three numerical case studies. The case studies correspond to three SAC steel moment frame buildings of three, nine, and 20 stories, which were designed to pre-Northridge codes and standards. Two-dimensional nonlinear finite-element models of the buildings are developed with the Open System for Earthquake Engineering Simulation framework (OpenSees), using a finite length plastic hinge beam model and a bilinear constitutive law with deterioration, and are subjected to multiple mainshock-aftershock seismic sequences. For the three buildings analyzed herein, it is shown that the structural reliability under a single seismic event can be significantly different from that under a sequence of seismic events. The reliability based robustness indicator shows that the structural robustness is influenced by the extent to which a structure can distribute damage.

In this paper, a method is proposed for the robust design of tuned mass damper (TMD) systems for seismic protection of multistory buildings. The seismic excitation is a random ground motion acceleration modeled by a stationary filtered white noise process. The protected building consists of a generic multi-degree-of-freedom (MDOF) structure, represented by its modes of vibration and linear mass dampers. The design properties of the TMD system are mass, frequency and damping ratio of the TMD units, along with their location within the structure, considered as fixed at its base. Uncertainties in the properties of both the building and the input seismic excitation are explicitly accounted for in the robust design of the TMD system. In particular, the uncertain parameters considered are stiffness and damping of the structure, and frequency and damping properties of the Kanai-Tajimi model used for representing the surface ground filter of the white noise process acting at the bedrock. The response quantity chosen to be representative of the seismic demand in the building is the interstory drift ratio. Its variation to the uncertainties is treated with the direct perturbation method, by applying a mixed-order approach. Robustness in the design of the TMD properties is formulated as a multiobjective optimization problem, in which both mean and standard deviation of the building response, produced by the considered uncertain parameters, are minimized. The weighted sum method is applied for transforming the multiple objective into an aggregated scalar objective function and then solving the minimization problem. The proposed design procedure is implemented on an illustrative example, consisting of a multistory building protected with a TMD system made from two units that have to be tuned with the first- and second-mode period of the structure, respectively. Parametric analyses on protected systems characterized by different properties are carried out, and the significance of the effects produced by the variation of such properties on the optimum design of the TMD system is shown. Differences between a robust design with the proposed procedure and a more conventional one that does not account for uncertainties in the system properties are finally evaluated.

A hybrid simulation approach is used to evaluate the traffic performance of a typical California highway overpass bridge after a major earthquake. The prototype bridge used in this study is a five-span single-column-bent reinforced concrete overpass. A hybrid model, composed of the physical column specimen and the OpenSees model of the remainder of the bridge, was assembled using the OpenFresco framework. Two hybrid simulation tests were performed on the same bridge for the same recorded ground motion scaled to represent two levels of seismic intensity: moderate and high. After the earthquake, loads corresponding to a heavy truck placed in the most critical positions on the bridge were applied on the hybrid model. The data obtained from these tests show that a typical California overpass bridge has the capacity to carry heavy truck loads immediately after a major earthquake event.

Nonlinear response history analysis is an important tool for accurately determining the performance of tall buildings under severe earthquake loading. When a standard genetic algorithm is used in conjunction with nonlinear response history analysis, it is desirable to use smaller generation sizes because of the computational effort to analyze individual designs. A study was conducted to evaluate how different genetic algorithm techniques influence the reliability and efficiency of the algorithm when used with nonlinear response history analysis and small generation sizes. The system used in the study was a nine-story buckling restrained braced frame that was optimized to minimize brace areas under individual earthquake records. A baseline study showed that a typical genetic algorithm did not converge to the same best design for different random number sequences (seed numbers). Forced diversity improved the reliability of the algorithm such that it converged to the same optimum, regardless of initial seed number. Adaptive mutation decreased the required number of generations when coupled with a noncrossover constraint. Consecutive identical generations were found to predict convergence and provide a basis for an exit criterion.

This paper presents an integrated, interdisciplinary methodology incorporating multiphysics, multiscale numerical modeling and simulation from tsunami generation, propagation, and inundation to subsequent coupled structural response and associated fluid loads. This novel, cohesive approach performs these simulations across a large spectrum of scales, enabling structural engineers to take full advantage of the detail available in recent advances in tsunami modeling, geospatial data collection, and computational structural mechanics. Extensive seismic networks, geodetic instruments, and water-level stations provide unprecedented data sets, enabling one to model, simulate and reconstruct tsunami events with high fidelity. A number of coordinated, ground-based surveys also collect valuable, time-sensitive quantitative information to improve understanding of structural response to tsunami loading following events. Recently, some of these surveys include high-resolution LIDAR measurements, which provide critical geospatial information to link field observations, topographic mapping, and structural performance to create and validate numerical models, enabling quantification and understanding of structural response and failure modes resulting from tsunami forces. The presented methodology is demonstrated through a case study of a building component which structurally survived the 2011 Tohoku tsunami.

To aid in the evaluation of the collapse-prediction capability of competing methodologies, a case study of a water tank subjected to the Takatori near-source record from the 1995 Kobe earthquake, scaled down by a factor of 0.32, is presented. The water tank, supported by a five-segment steel lattice tower, is so configured as to have a characteristic collapse mechanism that is triggered due to catastrophic column and brace buckling at the bottommost segment of the lattice under all forms of ground motion. A FRAME3D model of the tank reveals severe buckling in the bottom megacolumns and one of the two braces on the west face of the tower when the structure is impacted by the Takatori near-source pulse, resulting a tilt in the structure. This is followed by sequential compression buckling of braces on the south and north faces leading to $P-\mathrm{\Delta }$ instability and complete collapse of the tank. In order to verify the predictions of the FRAME3D model, a comparable PERFORM-3D model of the tank, using fiber elements and constitutive material models that are suitably calibrated against experimental data, is developed. The response of this model to the scaled Takatori ground motion compares very well against that of the FRAME3D model; the smallest scaling factor needed to collapse the PERFORM-3D model is 0.323, whereas the corresponding factor needed to collapse the FRAME3D model is 0.315. The sequence of column- and brace-buckling failures and the collapse mechanisms are quite similar in the two models.

Predicting the response of elastomeric seismic isolation bearings when subjected to severe ground motions is challenging due to the highly nonlinear behavior associated with the bearings under a combination of large displacements and axial loads. In particular, the horizontal stiffness of the bearings is a function of both horizontal displacement as well as axial load that varies due to overturning moments. Previous analytical models or formulations to model these bearings were mainly developed to estimate critical loads at the stability limit. Only few of these models are capable of estimating the correct nonlinear behavior of bearings observed at horizontal displacements in excess of the bearing width. In this study, a nonlinear analytical model is presented that is capable of modeling the dynamic response of bearings more accurately at all displacement ranges, especially beyond the stability limit and is verified with experimental data from an earlier experimental study. It was observed in the dynamic experiments that the bearings have a far larger capability to sustain horizontal loads at displacements exceeding their stability limit than predicted by earlier models and more importantly the bearings re-centered after these large displacement excursions. This behavior is captured using the analytical model developed in this study.

An algorithm is proposed for the rapid estimation of the influence of model parameter uncertainties on the seismic performance of structures using incremental dynamic analysis (IDA) and Monte Carlo simulation with Latin hypercube sampling. It builds upon existing methods that quantify the uncertainty for structural models with nondeterministic parameters by performing IDA with multiple ground motion records on each model realization out of a predetermined sample. However, their practical application is restricted due to (1) the inability to determine a priori the required number of samples and (2) the disproportionate increase of the number of analyses in realistic multiparameter models. To address these issues, two fundamental changes are incorporated. First, Latin hypercube sampling is applied progressively by starting with a small sample that is doubled successively until the desired accuracy is achieved. Second, parameter sampling is performed on a record-by-record basis rather than maintaining the same model over an entire record suite, thus expanding the model sample size without increasing the number of nonlinear dynamic analyses. Using strong-column and weak-column models of a steel moment-resisting frame, the algorithm is shown to possess excellent scalability, extending the original methodology to be applicable to large-scale models with hundreds of random variables.

Improved analysis methods and guidelines are presented to simulate the seismic collapse of nonductile concrete frame buildings with masonry infills. The analysis tools include an inelastic dual-strut model that captures the post-peak behavior of the masonry infill and its interaction with the surrounding frame. The dual compression struts capture the column-infill interaction that can cause shear failure of the columns and loss of their vertical load carrying capacity. A rigid softening shear degradation model is implemented in the beam-column elements to capture the shear failure of nonductile RC columns. Guidelines are presented to determine the strut model parameters based on data from 14 experimental tests on infill frames. The models are applied in three-dimensional nonlinear dynamic analyses of a three-story nonductile concrete frame prototype building with infills. The incremental dynamic analyses technique is utilized to understand the effect of the infill-column interaction and the rocking of shallow foundations on collapse performance, including a parameter study to examine the sensitivity of the results to the assumed strength and deformation parameters of the infill walls. Collapse assessment of the prototype building indicates that incorporating infill strut-column interaction and the shear degradation of columns is critical to the prediction of the collapse capacity of nonductile infill frames. Otherwise, the omission of this deterioration mechanism leads to unconservative collapse capacity predictions. The analyses further demonstrate that rocking of shallow foundations has a favorable effect on the collapse performance of infill frames and that the infill strut strength has considerably greater influence on collapse performance than the infill strut deformation parameters.

This study investigates the robustness of a seismically designed steel moment frame building using 3-D nonlinear models. Computational simulations are carried out using validated models to investigate building response to loss of internal and external columns as well as columns in lower, middle and upper floors. Studies are conducted to gain insight into the sources of collapse resistance of the prototype building, focusing in particular on the role of composite action between the slab and underlying steel beams and slab membrane action. The simulation results suggest that the prototype building can be more vulnerable to loss of columns in the upper stories than in the lower ones and is particularly vulnerable to loss of interior gravity columns at all floor levels. The role of the slab is quantified and it is shown that it can contribute significantly to the robustness of the structure, but it is a double edge sword that can also be detrimental, especially in the final stages of collapse.

Posttensioned (PT) bridges are used on major thoroughfares because they are economical structures for traversing long spans. Special inspections of these bridges have revealed corrosion of the strands at void locations in the tendons. The integrity of these strands has significant influence on the safety of these bridges. Inspections indicate that many ducts (parts of the tendon that is supposed to protect strands from exposure to aggressive environments) are cracked and many grout holes and vents are opened, allowing direct ingress of moisture and chlorides. This paper presents a framework for assessing the flexural reliability of PT bridges exposed to various environmental conditions. The moment capacity of a PT girder is formulated using probabilistic models for the tension capacity of corroding PT strands exposed to various void and environmental conditions. Using Monte Carlo simulations, the flexural reliability of an example PT bridge is assessed. The research indicates that the flexural reliability index reaches a value below the recommended value within a relatively short period of time when moisture and chlorides infiltrate the tendons. These findings emphasize the critical need for new inspection, assessment, and repair methods for these bridge types.

This paper proposes a computational approach for the collapse assessment of concentrically braced frames (CBFs) subjected to earthquakes. Empirical formulations for modeling the postbuckling behavior and fracture of three main steel brace shapes that are commonly used in CBFs are developed. These formulations are based on extensive calibrations of a fiber-based steel brace model with available information from a recently developed steel brace database. As part of the same computational approach, the representation of strength and stiffness deterioration associated with plastic hinging in steel columns and gusset-plate beam-to-column connections is considered. Through a case study of a 12-story Special Concentrically Braced Frame (SCBF), the influence of classical damping on the collapse capacity of CBFs is investigated. It is demonstrated that when SCBFs attain a negative stiffness during an earthquake, their collapse capacity can be significantly overestimated, if viscous damping is based on a commonly employed Rayleigh assumption with initial stiffness approximation. It is shown that sidesway collapse of CBFs should be traced based on a combination of criteria that associate large story drift ratios and the story shear resistance of a CBF at the corresponding story drift ratios.