Journal of Structural Engineering

Nonlinear static and dynamic analyses are utilized by engineers for performance-based seismic risk evaluation of new and existing structures. In this context, nonlinear component modeling criteria are typically based on ASCE 41 guidelines. Experiments on wide-flange steel columns suggest that the ASCE 41-13 nonlinear component models do not adequately reflect the expected steel column behavior under cyclic loading. To help bridge the gap between state-of-the-art research and engineering practice, this paper proposes new modeling criteria for the first-cycle envelope and monotonic backbone curves of steel wide-flange columns for use in nonlinear static and dynamic frame analyses. The proposed nonlinear provisions include new parameters for concentrated hinge models to facilitate modeling of strength and stiffness deterioration of steel columns under seismic loading. The associated variability in the model parameters is also quantified to facilitate reliability analyses and development of probabilistic acceptance criteria for design. Recommendations are made to account for the influence of bidirectional lateral loading and varying axial load demands on the steel column’s hysteretic behavior. Also proposed is an increase in the compression axial force limit for characterizing columns as force-controlled versus deformation-controlled in line with the new ASCE 41 provisions. The proposed modeling parameters are validated against test data and continuum finite-element analyses, and they are proposed for consideration in future updates to ASCE 41 requirements for nonlinear static and dynamic analyses of steel frame buildings with wide-flange columns.

Steel–concrete composite bridges, which combine the excellent tensile properties of steel beams and the compressive properties of concrete slabs, are widely used worldwide because of their superior mechanical properties, convenience of construction, and economic efficiency. Although various analysis tools have been presented for the design and evaluation of steel–concrete composite beams, there is still no effective tool that considers the effect of interlayer bonds on the flexural behavior of composite beams. This study developed an engineer-friendly mechanical model to analyze the flexural behavior of composite beams with consideration of the interlayer slip between the steel and concrete. In the model, the shear studs in the composite beams were simulated using nonlinear springs. The mechanical model was validated using experimental results. The model was used to predict the interlayer slip and deflection of composite beams under different loading levels. The analysis results indicated that neglecting interlayer slip leads to underestimation of the deflection of composite beams by as much as 27%–35%.

This paper studied experimentally the behavior of circular fiber-reinforced polymer (FRP)–steel-confined concrete columns subjected to reversed cyclic loads. The influence of main structural factors on the cyclic behavior of the columns is discussed. Test results showed the outstanding seismic performance of FRP–steel-confined RC and steel-reinforced concrete (SRC) columns. The lateral confinement effectiveness of glass fiber–reinforced polymer (GFRP) tubes and GFRP–steel tubes was verified and a simplified OpenSees-based finite-element method (FEM) model was developed to simulate the experimental results of the test columns. Based on the proposed FEM model, a parametric analysis was conducted to investigate the effects of main factors on the reversed cyclic behavior of GFRP–steel-confined RC columns. Based on the test and numerical analyses, the study discussed the influence of variables such as the lateral confinement on the plastic hinge region (PHR) height and peak drift ratio of the columns under reversed cyclic loads. Results indicate that lateral confinement significantly affects the PHR height of circular confined RC columns. Based on the analyses of the data from this study and literature, a simple model was suggested to predict the peak drift ratio of confined RC columns.

In this paper, a total of 12 circular tubed steel-reinforced concrete (CTSRC) short columns with high-strength concrete are constructed and tested under axial compressive load. The main variables of the test include diameter-to-thickness ratio and yield strength of the steel tube, concrete strength, and the steel ratio of the profile steel. The failure modes, axial load-displacement curves, and ultimate axial load of the CTSRC columns are analyzed systematically. It can be determined from the test results that all specimens exhibited shear failure under concentric loading. However, the shear cracks around the concrete circumference were not connected throughout the entire section due to the presence of internal profile steel. Moreover, as a result of double confinement provided by the steel tube and profile steel, the ultimate concrete strength and ductility of the CTSRC columns were enhanced significantly, which led to a considerable improvement to the axial capacity. Finite-element (FE) analysis is conducted on the CTSRC columns using ABAQUS software, and the results show that good agreement and satisfactory accuracy between the test observations and FE methods can be achieved. Therefore, the numerical simulation used in this study is relatively reasonable and accurate. Based on the effective FE model mentioned previously, extensive parametric analysis is performed, and a more accurate axial strength model is proposed. It is evident that good agreement of theoretical predictions and the test data from this paper and other relevant literature can be achieved using the proposed model.

The inelastic response of a class of nonlinear oscillators, characterized by a relatively rigid superstructure on a soft ground story under base seismic excitation, is numerically investigated. Both lateral translation and rocking due to overturning of the rigid superstructure are considered in the formulation. Initially, the governing equations of motion are established in a nondimensional form in order to identify the performance parameters of the system, namely initial elastic characteristics (i.e., periods and damping), normalized resistance, and hysteretic response shape. Then, the inelastic response of the oscillator is numerically evaluated under seismic excitations at the base over a range of periods and resistance parameters (lateral and rocking) encountered in practice. Inelastic ductility demands and reduction factors are obtained and are compared to those from conventional single-degree-of-freedom lateral response used in seismic design. The results show that the presence of the rocking component increases the transverse ductility and internal force demands, when compared against the conventional spectral estimates using the single-degree-of-freedom approach.

A new triaxial constitutive model [Tsinghua University Concrete Three-dimensional (THUC3)] of concrete was developed for high-fidelity finite-element (FE) simulation of composite plate shear wall–concrete encased (C-PSW/CE), which has been adopted as the lateral force-resisting system in numerous ultrahigh-rise buildings. In this formulation, a uniaxial constitutive model of concrete was introduced. The following five key features of concrete were included in the proposed model: compressive softening of unconfined and confined concrete, tension softening, pinching effect, shear softening, and strength degradation due to principal tensile strain. A triaxial constitutive law was also illustrated based on a fixed crack assumption. By assuming that the stress of the concrete can be decoupled in the crack coordinate system upon initial cracking, the triaxial constitutive law was assembled from the aforementioned uniaxial stress–strain relations. A new formulation is proposed to achieve strain decomposition and stress assembly in three-dimensional stress space. The numerical implementation of the stress update algorithm in an ABAQUS user material (UMAT) subroutine was illustrated. The modeling scheme and material constitutive laws of reinforcing bars, steel plates, and studs were also introduced. Subsequently, the developed triaxial constitutive model was validated at the material level by simulating the cyclic behavior of concrete and reinforced concrete (RC) specimens. The developed FE model showed good accuracy in simulating the hysteretic behavior, ultimate capacity, and residual strain in cyclic tensile loading. Two flexural critical C-PSW/CE tests, Specimens C30-5 and C40-5, were also completed and reported in this research, and 18 shear critical C-PSW/CE tests were collected from the existing literature. These tests were simulated to further validate the developed model at the structural level. Comparisons indicated that the proposed model showed a reasonable level of accuracy in terms of ultimate capacity, energy consumption, and failure mode. In addition, the modeling of studs by a nonlinear connector element is recommended based on comparisons between the test and FE results. The current AISC provisions for shear capacity of C-PSW/CE specimens were also evaluated relative to the computational and experimental results.

Accurate modeling of the damage caused by hurricanes making landfall to physical infrastructure relies on the accurate modeling of temporally and spatially varying wind fields. However, wind field measurements from past events only include the wind field data at discrete time points separated by hours. In community-scale modeling, a hurricane wind field of a desired strength for the purpose of resilience planning (including investigation of mitigation alternatives) may be needed for a community of interest that has not experienced such a prior event, thereby necessitating simulation of a hurricane wind field of a specified strength with an arbitrary landfall location. In this context, this paper proposes a novel data-driven simulation technique to simulate temporally and spatially varying hurricane wind fields for the purposes of hindcasting and synthetic scenario analysis based on integrated asymmetric Holland models. First, the backward Holland model (i.e., Lambert $W$ function) is used to derive the time-varying model parameters from measurement data of historical events. Then the data-driven model parameters are adopted in the forward Holland model to simulate snapshots of missing times for historical events or realistic synthetic events with a synthetic track passing close to the community of interest (e.g., for the purpose of community resilience planning). To achieve high simulation accuracy, the wind fields for inner and outer core regions are modeled separately by two sets of asymmetric Holland models, whose parameters are estimated using two different branches of the Lambert $W$ function and in the end are integrated to represent the entire wind field. In addition, the sudden change of the surface wind speed due to the roughness change from water to land is explicitly modeled using a speed conversion process. In this way, the proposed technique successfully overcomes two shortcomings of the existing Holland-type models, that is, poor representation of the wind field in the inner core region and the inability to model surface wind speed change due to roughness changes. The performance of the proposed data-driven simulation technique is illustrated in examples of simulations for both historical and synthetic hurricanes.

This paper focuses on the performance-based structural fire engineering of steel buildings subjected to realistic design-basis fire scenarios. Medium-rise office buildings were designed in accordance with U.S. building codes and standards. These buildings were designed with steel gravity frames and different lateral force resisting systems: (1) interior rigid core walls, and (2) perimeter moment resisting frames. Code-based prescriptive approaches were used to design the passive fire protection for the structural members of the steel buildings. The structural performance of these buildings, when subjected to realistic design-basis compartment fire scenarios (one-hour fire with cooling), was evaluated by conducting detailed nonlinear inelastic finite-element analysis of the complete three-dimensional (3D) building system. The analysis results indicated that steel gravity columns were the most vulnerable component, susceptible to inelastic buckling during the fire events. This could further precipitate the partial or overall collapse of the building during the fire event. This fundamental understanding of structural performance and vulnerabilities of the steel buildings (designed according to prescriptive approaches in codes) was leveraged to optimize the layout and distribution of passive fire protection in the steel buildings, leading to performance-based structural fire engineering. Detailed analysis results indicated that moving the fire protection from intermediate filler beams in the floor systems to the gravity columns improved structural performance, fire resistance, and collapse resistance of the steel buildings. This exemplifies the potential of performance-based fire protection engineering linked with structural engineering analyses to optimize design and economy.

Lattice models are well suited for crack simulations; however, their use has been mostly limited to the fracture of plain concrete. In this study, a practical two-dimensional mesoscale lattice composed of overlapping elements was employed to simulate the monotonic response of reinforced concrete elements. The force-deformation response of each element is calibrated from direct tension tests. An explicit time integration technique with novel proportional-integral-derivative control is used to efficiently simulate the response under monotonic loading. Six different reinforced concrete member simulations were conducted to validate the proposed approach. It was found that the proposed approach was capable of reproducing the load-deformation characteristics of elements failing in shear or flexure with a reasonable accuracy. A deterministic sensitivity analysis was conducted to uncover the response parameters with the most influence on the response estimations. Concrete tensile strength and steel yield strength were found to be the most influential parameters affecting strength and energy absorption capacities. Interestingly, the variation in fracture energy and tensile-softening parameters appeared to exhibit insignificant differences for strength and energy absorption estimations in the reinforced concrete simulations.