Journal of Structural Engineering

Special steel plate shear walls (SPSWs) are an economical seismic lateral force resisting system used throughout North America and Japan. Architectural constraints often encourage the designer to place a pair of planar SPSWs adjacently. A logical extension of the planar SPSW system is to link two SPSWs together with coupling beams to form a SPSW with coupling (SPSW-WC). SPSW-WCs have only been addressed in a limited fashion by prior research, and current code provisions contain no guidance for their design. This research studied the force transfer mechanisms, behavior, and design of SPSW-WCs. A design methodology was developed by extending the capacity design procedures that are currently used for planar SPSWs. Fourteen prototype buildings were designed using this methodology, and representative numerical models were developed. These numerical models were validated and then used to conduct static pushover analysis and earthquake response history analysis. The results of the numerical simulations demonstrate that SPSW-WC systems exhibit robust seismic performance that is equivalent to or better than similar planar SPSW systems, and the SPSW-WC systems improve architectural flexibility and material efficiency.

Concentrically braced frames (CBFs) are stiff, strong systems frequently used to resist seismic loading. Special CBF (SCBF) behavior is dominated by brace buckling, while buckling restrained braced frames (BRBFs) develop tensile and compressive yielding and avoid brace buckling. Both systems are widely used in seismic design, and both have a number of specific design issues. This paper describes a first of its kind, 2-story, 1-bay by 1-bay frame tested at the University of Minnesota Network for Earthquake Engineering Simulation facility to examine the large-displacement, bidirectional behavior of SCBFs and BRBFs with realistic boundary conditions and to verify the design approach. The SCBF had rectangular hollow steel section (HSS) braces in a single-story X configuration, and the BRBF used a single-diagonal configuration. The design of the gusset plates for the HSS braces followed a previously proposed balanced design procedure with an elliptical clearance to permit out-of-plane rotation caused by brace buckling. The single-story X-brace SCBF concentrated damaged into one-half the brace length, and the first HSS brace fractured at 2% story drift. The BRBF gusset-plate design followed current design standards, and two of the BRB cores fractured at 3.6 and 4.2% story drift prior to any instability in the BRB or system. The SCBF sustained limited damage to the beams and columns; however, the BRBF had much more significant damage to these members because of larger deformations and BRBF behavior. The results indicate that these systems have a stable response to large cyclic deformations and the impact of bidirectional loading on the measured response was minimal.

As part of the Network for Earthquake Engineering Simulation Research (NEESR)-Grand Challenge Project Simulation of the Seismic Performance of Nonstructural Systems, an experimental program was carried out to evaluate the seismic responses, failure mechanisms, and fragilities of cold-formed steel framed gypsum partition walls. Understanding the seismic behavior of building interior partition walls is important because damage to these nonstructural components can initiate at relatively low story drift levels, potentially degrading the overall functionality of the building and contributing toward earthquake economic losses. To this end, in-plane quasi-static and dynamic tests were conducted on 36 partition walls constructed using common construction details. Variables examined on the 16 configurations tested include framing thicknesses, stud connections to top and bottom tracks, wall intersection details, and partial height walls among others. In addition, new details are proposed to increase the drift demands at which damage is first observed and to minimize the propagation of damage through the wall. The failure mechanisms observed for the different wall configurations are reported and a seismic fragility database for groups of partitions dependent on the construction details is generated. Fragility functions are provided for three distinct damage states based on the level of repair required for the partition wall. The resulting fragility database partially fills a critical need to more accurately estimate nonstructural damage and consequential losses in buildings during earthquakes.

The understanding of, and ability to predict, the seismic behavior of woodframe structures has improved immensely in the last 10 years. Two notable projects occurred in succession, the Consortium of Universities for Research in Earthquake Engineering (CUREE) and California Institute of Technology (Caltech) CUREE-Caltech Woodframe Project, and the Network for Earthquake Engineering Simulation (NEES) NEESWood Project, which provided greater understanding, improved analytical modeling, and advances in force- and performance-based seismic design (PBSD) of woodframe buildings, respectively. The NEESWood project focused on the development of a PBSD philosophy (and procedures) for midrise woodframe buildings and included full-scale system-level validation. To date, the complete design procedure has not been systematically explained, which serves as the impetus for this paper. While the method was shown to be quite viable, some challenges remain, including identifying a format that would facilitate widespread adoption and use by engineers. Thus, this paper examines the progress, current state, and challenges for PBSD of midrise woodframe buildings.

This paper presents a new two-dimensional shear-wall and diaphragm model developed as part of a Network for Earthquake Engineering Simulation (NEES) Project entitled NEES-Soft: Seismic Risk Reduction for Soft-Story Woodframe Buildings. A large portion of the older multistory buildings in the California region were constructed with a deficiency that makes them vulnerable to collapse in the first story during earthquakes. This deficiency is referred to as soft-story. The new model presented in this paper was developed using a corotational formulation, which makes it suitable for modeling the side-sway collapse of wood shear walls under large displacement as well as estimating the in-plane stiffness of floor diaphragms. To achieve high computational efficiency, a nodal condensation technique is used to eliminate the degrees of freedom (DOFs) associated with the nail connections from the global DOFs of the model. To verify the validity of the new model, the model was coded into a computer program and was used to analyze selected shear walls and diaphragms tested by various institutions and research programs. Good agreement was observed between the test and model-predicted backbone and cyclic curves for shear walls with various gravity loads and different anchorage conditions. The model is highly flexible and has been shown to be able to model older shear wall construction with horizontal sheathing boards and diagonal bracings.

The performance of full-scale light-frame wood walls subjected to wave loading was examined using the Large Wave Flume of the Network for Earthquake Engineering (NEES) Tsunami Facility at Oregon State University. The hydrodynamic conditions (water level and bore speed) and structural response (horizontal force, pressure, and deflection) were observed for a range of incident tsunami heights and for several wood wall framing configurations. The walls were tested at the same cross-shore location with a dry-bed condition. For each tsunami wave height tested, the force and pressure profiles showed a transient peak force followed by a period of sustained quasi-static force. The ratio of the transient force to quasi-static force was 2.2. These experimental values were compared with the predicted values using the linear momentum equation, and it was found that the equation predicted the measured forces on the vertical wall within an accuracy of approximately 20% without using a momentum correction coefficient. The experiments also showed that the more flexible $2\times 4$ wall resulted in lower peak forces when compared with $2\times 6$ walls subjected to similar tsunami heights. However, the $2\times 6$ walls were able to withstand larger waves before failure.

This paper describes a coordinated experimental and analytical study that investigated the seismic performance of wood-stud construction with clay masonry veneer and of RC masonry construction with clay masonry veneer designed according to current U.S. building codes. As part of this study, two 1-story buildings were tested on a shaking table. Both had clay veneer on the outside. One had wood-frame walls for the backing and load-resisting system, whereas the other had RC masonry walls. With one exception, the behavior of the veneer in the wood-framed building was satisfactory up to levels of shaking consistent with the design level earthquake. That exception was related to the relatively low pullout strength in wet wood of the conventional nails used to attach the veneer connectors to the wood-stud wall. On the basis of this study, code changes have been implemented to require the use of higher-capacity fasteners for such applications. The wood-frame building specimen itself did not collapse under levels of shaking consistent with the maximum considered earthquake. Behavior of the RC masonry building and the veneer on that building was satisfactory up to levels of shaking well in excess of the maximum considered earthquake.

This paper presents a study that investigated the effectiveness of retrofitting unreinforced masonry infill walls with composite materials to enhance the seismic performance of infilled nonductile RC frames. The primary retrofit scheme considered was the use of engineered cementitious composite overlays. Shake-table tests were conducted on a $2/3\text{-scale}$, 3-story, 2-bay, masonry infilled RC frame that had one bottom-story wall retrofitted with engineered cementitious composites. The influence of this retrofit on the performance of the structure was investigated using the experimental observations and results of nonlinear finite element analyses. Furthermore, after walls in the second story of the structure were damaged, they were repaired by injecting epoxy into cracked mortar joints, and strengthened with a glass-fiber reinforced polymer overlay. It has been shown that both retrofit schemes are effective in enhancing the seismic performance of the structure and preventing diagonal shear failures of the RC columns.

Shake-table response of a 33.6-m-long (110-ft-long), four-span, RC bridge model with a continuous posttensioned superstructure supported on three, two-column bents was studied under horizontal bidirectional coherent simulated earthquakes. The pier heights varied, introducing a slight asymmetry about the centerline of the bridge model. The superstructure ends were supported on seat-type abutments that were connected to hydraulic actuators simulating abutment movements. The skew angle was zero. The bridge model was subjected to successive motions simulating a modified version of a record from the 1994 Northridge Earthquake. Results showed that damage was concentrated in plastic hinge regions of the columns and that the pier caps and the superstructure remained essentially elastic. The shortest pier failed, but this pier continued to carry vertical loads, and the bridge did not collapse. Despite the lack of skew, the superstructure-abutment interaction led to large in-plane rotations that caused significant residual displacements in the piers.

The seismic behavior of coupling beams and walls constructed with tensile strain-hardening, high-performance fiber-reinforced concrete (HPFRC) was studied through tests of large-scale precast coupling beams and coupled walls. A precast coupling beam design was developed to speed up construction and minimize interference with wall reinforcement. Three isolated precast coupling beam specimens with a span/depth of 1.75 were tested under large displacement reversals. Test results indicate the use of HPFRC allows a reduction of the reinforcement required to achieve a stable coupling beam response by providing confinement and contributing to beam shear strength. A concrete design shear stress capacity of $0.41\mathrm{\surd }{f}_c^{\prime } \text{MPa} (5\mathrm{\surd }{f}_c^{\prime } \text{psi})$, where $f_c^{\prime }$ is the compressive strength of the concrete, was found to be appropriate. In addition to the coupling beam tests, two 4-story coupled wall specimens with precast HPFRC and regular concrete coupling beams were tested under lateral displacement reversals. Besides allowing the evaluation of seismic behavior of coupled walls with HPFRC coupling beams, the use of HPFRC in the plastic hinge regions of the walls as a means of relaxing transverse wall reinforcement was evaluated. The two coupled wall specimens exhibited drift capacities of at least 2.5%. The HPFRC coupling beams were more ductile and damage-tolerant than the regular concrete beams. The incorporation of a HPFRC material in the wall allowed the use of a concrete design shear stress capacity of $0.33\mathrm{\surd }{f}_c^{\prime } \text{MPa} (4\mathrm{\surd }{f}_c^{\prime } \text{psi})$ and a wider spacing of transverse reinforcement confining the wall boundary regions.

RC core walls are used commonly in modern building construction as the primary lateral load–resisting system. To meet architectural constraints, including elevator, stair, and doorway openings, common configurations include walls coupled together with heavily reinforced, low-aspect-ratio coupling beams. Numerous studies have focused on coupling beams and improved seismic performance and design of coupling beams. However, far fewer research programs have studied the seismic behavior of the coupled wall system. Coupled walls are typically used in mid- to high-rise construction, and understanding their seismic response requires simulation beyond the coupling beam and must include all important yielding components of the system. A research project was undertaken to investigate the response of a midrise coupled wall designed to meet current codes. The advanced testing capabilities of the University of Illinois at Urbana-Champaign George E. Brown, Jr. Network for Earthquake Engineering Simulation (NEES) experimental facility were used to conduct a large-scale test of a coupled wall. The test specimen modeled the lower 3 stories of a 10-story building. The demands from the upper stories were imposed on each pier using an innovative loading device capable of imposing shear, moment, and axial loads or displacements. Damage progression included yielding of coupling beams, yielding of wall piers, spalling in the wall piers, and the coupling beams. Specimen failure occurred almost without warning at 2.27% drift, resulting in significant loss of lateral and axial load–carrying capacity. The damage sustained included concrete core crushing and longitudinal bar buckling in the compressive wall pier. These experimental results were combined with prior test data to develop improved design models for coupled wall systems.

Performance-based earthquake engineering necessitates the development of reliable nonlinear analysis models that are able to simulate the behavior of structures from the onset of damage through collapse. These models provide engineering demand parameters that are then related with damage measures and describe the damage of a building and its components. To accurately simulate dynamic response up to collapse of structures, it is important to model strength and stiffness deterioration of structural components in addition to $P\text{-}\mathrm{\Delta }$ effects. These models require the use of large sets of experimental data for calibration of their deterioration parameters. This paper discusses the development of three databases on experimental data of steel W-beams, tubular hollow square steel columns, and RC beams. These databases are used for quantification of important parameters that affect the cyclic moment-rotation relationship at plastic hinge regions in steel and RC components. Emphasis is placed on the prediction of collapse of buildings caused by earthquakes. The utilization and importance of the three databases in the context of performance-based earthquake engineering is demonstrated through a case study of a 4-story steel building. Its seismic performance is successfully assessed through collapse.

Seismic performance evaluation of wharf structures constitutes the core of the vulnerability assessment of seaport infrastructure exposed to seismic events. Because this performance will depend on complex interactions between the surrounding potentially liquefiable soils and wharf foundation and structure, detailed and careful implementation of advanced modeling techniques is needed. These models are required to capture such highly nonlinear phenomena as permanent seaward deformation of embankment soils, soil-structure interaction in liquefiable soils, spread of plasticity in prestressed piles, and force-deformation of pile-deck connections. This study utilizes such modeling approaches to investigate the three-dimensional (3D) nonlinear response of a typical pile-supported container wharf structure in liquefiable embankment soils. Input excitations for this analysis consist of embankment soil deformations for a far-field and an impulsive near-field ground motion. These excitations are derived using two-dimensional (2D) plane strain free-filed analyses for transverse (seaward-landward) and vertical directions and a spectral matching technique for the longitudinal (parallel to shoreline) direction. Results of these 3D analyses show that the oscillating component of embankment deformations is the primary contributor to the maximum response of the piles sections and pile-deck connections, and that effects of permanent deformations of the embankment soil are much smaller. The analysis results also demonstrate the importance of the structure’s 3D response characteristics, including longitudinal and torsional responses of the structure that are comparable to the transverse wharf responses during large amplitude oscillating embankment deformations owing to impulsive near-field motions. These important 3D response characteristics are not captured by more typical 2D wharf response analyses. Detailed comparisons of the responses of the 3D wharf model to corresponding responses from the 2D model of the wharf are provided.

Although the nonlinear behavior of rocking shallow foundations has been experimentally and numerically demonstrated as an effective tool to dissipate vibrational energy during seismic loading, the engineering community has yet to uniformly accept it as a targeted design mechanism for diffusing seismic energy in a structure. This paper presents results of a centrifuge test program that incorporated inelastic behavior into model building systems via yielding of both structural and foundation components. Three 2-story-1-bay building models were designed with similar layouts but different combinations of foundation and structural component yield strengths and were shaken with a similar suite of earthquake motions. Measurements of behavior of each of the model buildings are presented and cross-compared in terms of time history responses, hysteretic responses of the structural and foundation fuses, and maximum response parameters. A balanced design configuration, wherein the rocking foundation and structural fuse are intended to yield at approximately the same load, is demonstrated to be a well-controlled seismic-resisting system, with greatly reduced seismic ductility demand on the structural components. Moreover, seismic energy is well distributed among the targeted yielding components. In contrast, if the footing is restrained from rocking, the structural component ductility demand is significantly greater than that compared to its demand when the foundation is allowed to rock. In essence, the foundation rocking dominated model demonstrates its ability to protect the superstructure from seismic demands. In contrast, when the rocking foundation capacity is more than twice that of the structural fuse, rotations at the foundations are reduced significantly, at the price of much larger demands to the superstructure.