Journal of Structural Engineering

Hybrid masonry is a new structural concept for buildings that incorporates the in-plane strength and stiffness of reinforced concrete masonry with the ease of erecting conventional steel framing. Because the masonry structural panels can also serve as architectural elements, hybrid masonry has the promise to be highly competitive with conventional lateral force resisting systems including reinforced masonry or concrete shear walls, steel braced frames, or concrete or steel moment-resisting frames. As part of an integrated program of multi-institutional research, seismic response and behavior of this construction type have been studied at the University of Illinois at Urbana-Champaign with a series of full-scale tests that are described in this paper. This paper provides an introduction to this innovative seismic structural system, summarizes the experimental research, and outlines how this research can be used to transform current seismic design practice, and by so doing make structural masonry more competitive in regions of moderate and high seismicity.

Thin steel plate shear walls (SPSWs) are an effective lateral force resisting system and are increasingly being used in structures. Engineers require the ability to predict both elastic and inelastic structural responses of SPSWs accurately by a simplified model that is convenient to use. Based on the strip model, this paper proposes a new simplified model for SPSWs, called the three-strip model, in which the infill plates of SPSWs are replaced by three tension-only strips. Comparisons of experimentally obtained pushover responses of SPSWs and that predicted by the proposed model are given and reasonable agreement is observed. A parametric study examines the effect of aspect ratio and column flexibility on the accuracy of the predicted inelastic behavior from the three-strip model. In addition, the capacity of the proposed model in predicting frame forces is verified based on test and strip model results. It was found that the three-strip model can provide generally accurate results for SPSWs and meanwhile reduce modeling workload and computational expense.

Vibration has become a serious serviceability problem due to the large openings of girder, lightweight floor systems, and high-strength materials used in practice. The results from both heel-drop and jumping floor vibration tests of a large-span prestressed cable RC truss (PCT) floor system are presented. System parameters such as peak acceleration, fundamental frequency, and damping ratios were considered and studied. The results and comparisons were taken based on the on-site test data. The comparison results indicate inconsistencies with the heel-drop or jumping impact models proposed by other researchers. The results from heel-drop and jumping tests were comprehensively compared against published codes. Based on the classic plate theory, a derived accurate formula for a PCT floor system is proposed. Lastly, a simplified method is suggested, which uses the constant moment of inertia for the cross section of a PCT girder.

This paper reports on an experimental study on four unbonded posttensioned masonry walls (PT-MWs). All walls had identical thickness, height, and length of 190, 2,000, and 1,400 mm, respectively, and were constructed using concrete masonry units (CMUs) and mortar type N, and were fully grouted. Different horizontal spacing values of 400, 600, and 1,200 mm were used between the posttensioning bars in the walls. Only Wall W4 had horizontal bonded reinforcement, located in the fourth and seventh courses. Two different levels of posttensioning force corresponding to an average posttensioning compressive axial stress on the masonry of 1.35 and 2.7 MPa were applied to the walls. Different initial posttension stresses in the bars ranging from 0.32 to 0.63 of the yield stress of each bar were applied to the walls. The walls were subjected to incrementally increasing in-plane lateral displacement cyclic load applied to the top of each wall. The experimental results including damage pattern, force displacement response, residual drift ratio, masonry compressive strain, wall rotation, damping, stiffness, stiffness degradation, and displacement ductility of the tested walls are presented and discussed in detail. The accuracy of ignoring the elongation of post-tensioning bars in predicting the strength of the tested walls is investigated based on the experimental results. The experimental tests showed that PT-MWs having the same total initial PT force but different stress levels in the PT bars displayed different lateral strengths. Doubling the total initial PT forces in the bars resulted in insignificant increases in the wall lateral strength. Finally, using bonded horizontal steel improved the lateral displacement capacity of the test specimen. Specimens that had horizontal reinforcement did not suffer postpeak shear cracking and hence were able to reach higher level of lateral displacement.

Three theorems of the flexural theory for RC members are derived to identify two critical points on a moment–curvature curve: (1) the onset of flexural strength and (2) the onset of the so-called true ultimate curvature. The true ultimate curvature is reached at the exact moment when a RC member loses its integrity. The first two theorems concern the first point and the third relates to the second. Exact analytical solutions of the extreme concrete strain for these two critical points are derived for general RC members, including underreinforced and overreinforced beams and RC columns. The solutions can be used to calculate the exact strength and the corresponding deformation and to evaluate concrete damage at the onset of the two critical points. The analytical results can also be used in practice for the design of RC members, which is particularly useful when precise evaluations of flexural strength and damage conditions at the onset of maximum moment or ultimate displacement are required.

The use of structural glass in buildings is a growing trend, especially with glass beams, which are increasingly used to support roofs, floors, and glazed facades. However, beams consisting of glass laminates only expose a brittle behavior. To overcome this, research has been performed during the last decade to create structural glass beams with enhanced failure behavior. Among a variety of developments, the application of concepts obtained from concrete construction is very promising. Reinforced and posttensioned glass beams were developed and investigated, yielding good results with respect to failure behavior. This paper gives an overview of the research performed on both types up till now. For reinforced glass beams, categorization is made according to the different reinforcement materials. Although remarkable differences in structural performance exist between the different approaches, the majority of the concepts illustrated an enhanced failure behavior, characterized by significant postfracture strength and sufficient ductility. This enhanced robustness on element level is a huge step forward for the application of structural glass beams in practice. However, additional research and optimization is still required to arrive at practical application.

This paper presents the experimental investigation on the compressive behavior of cold-formed high-strength steel tubular stub columns. In this paper, the nominal 0.2% proof stress of the high-strength steel was 700, 900, and 1,100 MPa. A total of 25 stub column tests on circular, rectangular, and square hollow structural sections were conducted. Geometric imperfections and load-deformation histories were reported and assessed. The experimental results were compared against the design values calculated from the Australian, European, and North American codes, and the corresponding compactness criteria were assessed. A finite-element model, which incorporated the varying material properties, was described, and the influence of geometric and material imperfections was also evaluated. The finite-element results successfully captured the experimental observations and can be adopted for future parametric studies.

This paper presents a mixed finite element (FE) model for the nonlinear time-dependent analysis of composite beams with partial shear connection. The key idea is to consider, as a first approach, a viscoelastic/plastic model for the concrete slab in order to simulate the interaction between the time effects of concrete, such as creep and shrinkage, and the concrete cracking. Creep is taken into account via linear aging viscoelasticity, while cracking is modeled using an elastoplastic model with softening. A nonlinear isotropic/kinematic hardening model is adopted for steel behavior and an appropriate nonlinear constitutive relationship is utilized for the shear stud. A consistent time integration is performed by adopting the Euler backward scheme. Finally, comparisons between the numerical results and experimental data available in the literature are undertaken to validate the accuracy of the model. It is shown that the interaction between cracking and time effects (creep and shrinkage) significantly increases the deflection.

To enhance the structural capacity and constructability of composite columns, a prefabricated steel-reinforced concrete (PSRC) column was developed. By using the prefabricated steel cage, field rebar work is unnecessary, and the self-erectable steel cage can provide sufficient strength and rigidity to support the construction loads of beams and slabs. In the present study, various steel angle ratios, transverse bar spacings, and prefabrication details were considered for better structural capacity and constructability. To investigate the performance of the proposed method, axial loading tests and cyclic lateral loading tests were performed. In the axial loading test, the structural performance of the PSRC columns was comparable to, or even better than, that of the conventional composite column using the wide flange steel section, when early spalling of the cover concrete was restrained by closely spaced hoops. Under cyclic lateral loading, on the other hand, the PSRC columns were more susceptible to early concrete spalling and steel angle buckling. Thus, when high ductility is required under cyclic lateral loading, the steel angles need to be embedded in the core concrete using thicker cover concrete and hoops.

An equivalent frame is developed to simulate the effect of a large-size truck colliding with a reinforced-concrete column. The surrogate frame is made of simple steel elements and blocks, representing the major features of a reference truck. Equivalence between the frame and actual truck is achieved by comparing their deformation versus internal energy relationships. The equivalent truck frame is scaled down and used in laboratory tests to study the collision process between trucks and bridge piers, thereby significantly reducing the difficulty and cost of conducting full-scale collision tests while maintaining high fidelity. Using the scaled surrogate frames, three drop-weight impact tests are conducted on circular reinforced-concrete model columns. Test results show that the structural responses of the columns are not only influenced by the impact energy itself but also the impact speed and impactor mass. Finite-element simulations of the collision process are conducted using both the equivalent truck frame and the reference truck, confirming the overall similarity between them.

Transfer length is a significant parameter in the design of pretensioned concrete members. The estimation of transfer length is typically achieved by measuring concrete strain or strand slip at the member ends. The end slip method is simple and requires little effort, but relies on an empirical formula with an undetermined coefficient ($\alpha$), which relates to the bond stress distribution. Many studies have proposed appropriate $\alpha$ coefficients; however, the variability in estimating the $\alpha$ coefficient exists between studies depending on the chosen empirical fit. This study gives an estimation of the $\alpha$ coefficient using a probabilistic approach called the maximum likelihood (ML) method. Approximately 600 data points of transfer lengths and strand end slips were measured for 25 pretensioned concrete beams. Strand end slips were measured at release and at 3, 5, 7, 14, and 28 days. Concrete strains were measured at the same period of time using a detachable mechanical strain (DEMEC) gauge to determine the transfer lengths. An estimated $\alpha$ coefficient of 2.61 maximized the likelihood of the observed sample of the measured transfer lengths and end slips.

Column stability during seismic events is a critical consideration for all steel-framed structures, but is of particular importance for a unique class of steel frames called multitiered braced frames (MT-BFs). A MT-BF is created by subdividing a tall single-story braced frame into two or more braced tiers with intermediate horizontal struts in the plane of the frame. Since out-of-plane bracing is not provided at the tier levels, under pure axial compression, a MT-BF column may buckle in plane over one tier or out of plane over the full story height. Typically, wide flange shapes with weak axis oriented in the plane of the frame are used for MT-BF columns. During a seismic event, brace inelasticity may result in differential tier drifts that induce weak-axis flexure in the wide flange columns. However, this flexural demand is not considered in current braced frame seismic design procedures, which only require column design for axial force. To facilitate development of seismic design guidelines for MT-BF columns, their inelastic stability behavior was explored using three-dimensional finite-element analysis. This paper describes studies of isolated columns subjected to compressive loads in combination with varying magnitudes of imposed weak-axis rotation. The location of the imposed rotation and the axial force distribution were also varied. In addition, the influence of torsional bracing at the tier levels on buckling strength and buckling limit states was examined. The results from the computational simulations indicate that strong-axis buckling strength can decrease significantly due to the presence of weak-axis flexure, if the imposed rotation is large and torsional restraint is not provided at the tier levels. Furthermore, the performance of many columns was enhanced by providing torsional bracing at the tier levels. Finally, weak-axis flexure at column midheight was found to be more detrimental to strong-axis buckling strength than weak axis flexure closer to the column end. The results from this study provide guidance for design provisions and will be used to inform the development of three-dimensional finite-element models of complete MT-BFs.

The concept of apparent weakening by adding true negative stiffness to a structure has been previously introduced by the authors in order to reduce simultaneous drifts, accelerations, and displacements in a structure without yielding or permanent deformation in the main system. A novel negative stiffness device (NSD) that generates true negative stiffness has been developed, built, and tested and has been previously described by the authors in terms of operation and analytical and numerical modeling. This paper presents results that represent proof-of-concept for weakening with the use of the NSD based on the shake table testing of a 3-story seismically isolated structure, equipped with these devices complemented by viscous dampers. The NSD is shown to have a significant effect on the superstructure response by reducing floor accelerations, story drift, and the base shear and upon the addition of dampers, also results in a reduction in isolator displacements. Moreover, this paper provides validation of the analytical models that were previously developed by comparison of analytical and experimental results.

With the numerous catastrophic incidences of progressive collapse in the past, there has been increasing interest from both engineering and academic communities toward evaluating the ability of reinforced concrete (RC) structures to resist progressive collapse. Of the numerous studies conducted in the past decade, the majority has focused on scenarios in which a single vertical loading bearing member, such as a column, is removed. However, accidental events (such as terrorist attacks or vehicular impact) often cause severe initial damage with multicolumn removal. Recognizing this fact, this study aims to assess the effects of multicolumn removal on the overall performance and load resisting mechanisms of RC structures. Two multipanel beam-column-slab substructures were tested, differing only in the number of columns to be removed. The test results collected included failure modes, load resisting capacities, deformation capacities, strain gauge results, horizontal movement of supports, and torsional rotations of edge beams. Based on test observations, a simple approach is given to evaluate progressive collapse vulnerability of RC buildings under single or multiple column removal scenarios.

This paper discusses contemporary design procedures for fillet welds to hollow structural sections (HSS) in several prominent design codes. The structural reliability associated with the directional strength-enhancement factor contained in North American Specifications is examined, based on a set of laboratory tests on fillet-welded connections between HSS and rigid end-plates. A total of 33 connections, in which the welds had been designed to be the critical elements, were tested to failure by axial tension loading applied to the HSS members. The experimentally obtained weld strengths were compared to the predicted nominal strengths. The directional strength-enhancement factor was found to lead to unsafe strength predictions, particularly for large weld sizes. Hence, a restriction on the use of this factor for fillet welds to HSS members, in North American steel design specifications, needs to be considered. The analysis also shows that Eurocode 3 fillet weld design provisions give conservative strength predictions.

This paper presents an experimental investigation of the racking performance of light-frame wood shear walls to evaluate the effect of Type X gypsum wall board (GWB). Twelve shear walls sheathed with oriented strand board (OSB) alone, GWB alone, or in combination were tested under static monotonic or reversed cyclic lateral load. The racking performance of these wall specimens in terms of stiffness, strength, ultimate displacement, ductility, energy dissipation, and failure modes was investigated. Specifically the effects of panel orientation, GWB panel joint taping, and double-layer GWB were examined. It was found that shear walls sheathed with GWB provide ductility ratios similar to or higher than those of OSB sheathed walls. The direct superposition rule can be applied to estimate the overall strength of a shear wall sheathed with single-layer OSB and GWB on opposite faces, provided there is no panel joint taping and the panel orientation is the same for both materials. The superposition rule proposed in current United States standards provides a conservative estimate for other cases considered in this study. In addition, shear walls with two GWB panels jointed by taping and double-layer GWB panels behaved like shear walls sheathed with a single large-size panel, leading to a reduction in the system ductility ratio and corresponding seismic force modification factor. This test series have provided essential test data for future modeling work to investigate the seismic response analysis of light wood-frame buildings containing Type X GWB.

A new and advanced beam-column element, namely the curved tapered-three-hinges (TTH) beam-column element, is proposed in this paper. The present element can perform large deformation analysis and explicitly simulate the initial member curvature, which is essential for the second-order direct analysis using one-element-per-member models. Another distinct feature of the element is to analytically express the flexural rigidity of tapered I-sections in the stiffness matrix through a series of tapered stiffness factors such as the ${\alpha }_i$ and ${\beta }_i$ factors. Unlike the conventional models using the approximated distributions (e.g., linear, parabolic, or cubic) or stepped-elements modeling approaches in an analysis, the present study gives an accurate simulation solution on nonprismatic beam-column elements. Herein, the element derivations and formulations are given in detail. To consider the large deflection effect in the analysis, the incremental tangent stiffness method is adopted and the kinematic descriptions are presented. Finally, several examples are employed to validate and verify the reliability and accuracy of the proposed element.

Concrete gravity dams are essential infrastructure components and because of earlier design practices, they may be vulnerable to future earthquake events. Therefore, their vulnerability is important to characterize, particularly given the large societal consequences of dam failure. Seismic fragility curves allow a rational safety and vulnerability assessment of existing structures under earthquake hazards offering statements of the conditional probability of reaching a limit state. This paper evaluates the seismic fragility of a case study concrete dam located in northeastern Quebec, improving on existing approaches by including spatial variation of the angle of friction and offering insights regarding the significance of modeling parameters, importance of spatial variation, and overall vulnerability of the dam to two failure modes. Finite-element techniques are used to model the dam system and to fully account for dam-reservoir-foundation interactions. Uncertainties in the ground motions and modeling parameters are included and propagated using a sampling technique. The fragility curves are developed using nonlinear time-history analysis, and the uncertainty related to the spatial variation of the angle of friction is included in the fragility analysis through the incorporation of random fields modeling. The study reveals that this additional source of uncertainty has a slight effect on the fragility of the dam and is mainly critical when severe damage levels are considered.