Journal of Structural Engineering

Currently, ASCE standards require that nonstructural components of a building be positively attached to the building in regions of moderate to high seismicity. However, for solar arrays that bear on low-slope roofs, friction between the array and roof surface limits or prevents sliding of the array in an earthquake. In many cases, the seismic performance of such arrays can be shown by analysis to meet the design intent of the building code without being fastened to the building structure. This paper presents a methodology for estimating the sliding displacement corresponding to ASCE standards for design-level earthquake shaking. The writers conducted nonlinear response-history analyses considering a range of different seismicity levels, roof slopes, and coefficients of friction, which lead to design values of sliding seismic displacement for solar arrays in different conditions. Important aspects of a broadly applicable analysis include performing testing to determine appropriate coefficients of friction, performing analysis with earthquake records spectrally matched to broadband response spectra, and considering the effect of vertical earthquake motion on frictional response.

This paper presents a study aimed to characterize the structural performance of laminated bamboo lumber (LBL) and bamboo glulam beams (BGBs) as a first step to evaluate their potential application as a structural material. LBL was tested to determine their flexural, tensile, and shear properties, whereas BGBs were tested for their flexural, shear, and compressive properties. The BGBs were fabricated using two different adhesives: isocyanate resin (ISO) and phenol-resorcinol formaldehyde (PRF). BGBs with ISO performed better in bending strength, whereas the stiffness of glulams with both glue types was equivalent. Irrespective of the glue type, failure modes and shear test data showed BGB bending strength was limited by interlaminar shear in the LBL used in BGB fabrication. From the experimental study, it is concluded that the LBL does possess higher allowable and average strength values in tension and bending and comparable stiffness values, with much less variability to a commonly used structural species of wood, Douglas fir. The potential of using LBL in framing applications exists. However, certain impediments need to be addressed and researched before acceptance of LBL and BGB in the construction marketplace.

This paper provides an overview of a moment connection for use in prefabricated timber portal frame buildings. The connection, known as the quick-connect joint, uses rods housed in timber sleeves that are placed on the outside of the main portal members to create a semirigid moment connection similar to those used in steel portal frames. Timber sleeves are connected to the main portal members using fully threaded, self-drilling timber screws that have high strength and stiffness characteristics. These characteristics limit connection slip and limit joint rotation. The quick-connect connection allows manufacture to take place offsite, thus reducing onsite plant and labor costs. A proposed design procedure is introduced in this paper. This procedure uses individual component deflections to calculate the overall rotation of members at the joint. Accurate rotation predictions allow for more efficient member design. Full-scale testing of a timber beam to steel column, and timber beam to timber column knee joint, is discussed. Results show that the quick-connect joint does not constrain the portal member sizes. Full-scale test results are used to verify the proposed design procedure. Test and analytical results show good correlation. The authors believe the quick-connect connection allows for more efficient and competitive design, manufacture, and erection of timber portal frame buildings by providing a joint and design procedure that is as easily used as those in the steel industry.

This paper presents the results from an experimental program that studied the effects of simulated delamination and corrosion on the structural response of six two-way slab-column specimens with details typical of older construction. The results in terms of load-deflection response, punching shear resistance, and cracking are compared. The test results demonstrate that a reduction in reinforcing bar cross-sectional area (simulated corrosion) leads to a significant drop in punching shear resistance and results in increased crack widths. In addition, the study demonstrates that a critical degree of delamination can lead to a reduction in punching shear capacity, greater crack widths, and loss of stiffness.

The performance of concrete anchors located close to a free edge and loaded in shear toward the edge of a concrete member is influenced by several geometric and structural/material parameters. In the case of anchor groups, the behavior of the system is described by highly nonlinear mechanical procedures, due to the interaction with anchors in vicinity. Apart from that, the existence of constructional tolerances (i.e., hole clearance of the bolted connection) leads to randomness in the distribution of the group load to the individual components, thus to uncertainties in the expected resistance of the anchorage. In that context, the structural effects of assembly tolerances have not yet been adequately investigated. As a result, current design guidelines are based on extreme-case concepts, thus often leading to a partial utilization of fastening systems. In the present study, it is attempted to capture the performance of the discussed structural system focusing on the influence of assembly tolerances by means of a stochastic approach.

Twelve specimens of reinforced concrete slab-column connections were tested to investigate fire resistance of punching shear. Half of the specimens were tested at room temperature to determine punching shear strength. Half were tested for fire resistance. The parameters of the specimens included steel reinforcement ratio, concrete compressive strength, and fire loading type. Fire resistance tests were carried out under constant maximum service loads at elevated temperatures according to ASTM E119 time-temperature curve. Temperature distribution, time-deflection relationship, and fire resistance were measured to determine the mechanical behavior of the slab-column connections subjected to fire loading on either the tension side or the compression side of the slabs. Experimental results showed that slabs heated on the compression side did not fail for up to 8 h, whereas those heated on the tension side failed at around 4 h. Specimens using normal-strength concrete have better fire resistance than those using high-strength concrete. The normal-strength concrete specimen also showed no spalling, whereas the high-strength-concrete specimen spalled under fire loading on the compression side.

This paper presents a computational investigation of two reinforced concrete beam-column assemblies, each comprising three columns and two beams, subjected to monotonically increasing vertical displacement of the unsupported center column simulating a column removal scenario. One assembly was part of an intermediate moment frame and the other was part of a special moment frame. Two types of models were developed: (1) detailed models with highly refined solid and beam elements to represent the nonlinear material behavior of concrete and reinforcement, and (2) reduced-order models with significantly fewer beam and spring elements to represent the nonlinear behavior of structural components. Reduced models are desirable for analysis of complete three-dimensional (3D) structural systems. Modeling approaches for the detailed and reduced models are described, and the computational results are compared with experimental data from full-scale tests. Good agreement is observed, which demonstrates the capability of the detailed and reduced models to capture the primary response characteristics and failure modes, including the successive development of compressive arching action and catenary action in the beams and the fracture of reinforcing bars at the beam-column interface.

This paper is concerned with the problem that structural joints in whole-frame models cannot, at present, be replicated in sufficiently minute detail to realistically represent their behavior. It is well recognized that the structural joints represent the weakest link in building frames; therefore, frame models are potentially inaccurate in a critical area. The impact of this research is in the development of an accurate frame modeling approach that achieves a realistic treatment of joint response without significantly increasing the computational requirements. The method utilizes simplified connection models using rate-dependent nonlinear springs which, when assembled, allow a realistic representation of the connection behavior. The method is found to be capable of modeling strain-rate dependent material property effects with a high degree of accuracy and coping adequately with the force and rotation combinations which develop during blast response. Increased rotation rate, which occurs as a response to blast loading, is shown to modify the rotational stiffness in joints which can in turn lead to increased dynamic shear forces. Structural models which oversimplify joint stiffness and which ignore strain-rate effects are shown to lead to potentially unsafe solutions.

The elastic and inelastic nonlinear behavior of structural members use constitutive relations in which restoring forces and deformations are not proportional and often follow different paths in loading and unloading using hysteretic functions. There are numerous available models that can trace the stiffness and strength changes through yielding, softening, and hardening; however, models that can address more complex behavior such as degradations, large deformation, bond slip, and joint gap, do so by complex polygonal rules or smooth continuous functions describing momentary (tangent) behavior. There is a need for a unified model on the basis of a combination of mechanical springs that can trace the instantaneous combined stiffness used in system analyses. In this study, a one-dimensional smooth hysteretic model using series and parallel springs, designed for nonlinear structural analysis, is enhanced to incorporate (1) time-independent properties, (2) nonlinear elastic and postelastic softening and hardening, (3) sudden or continuous variation of strength, (4) degradation of elastic and inelastic stiffness, (5) a modified bond-slip model, and (6) an alternative joint-gap model with a variable gap closing length. Using spring analogues or reciprocal structures, the instantaneous force-displacement incremental relations are formulated, which can be further used to determine the instantaneous tangent-stiffness matrices of the elements used in the analyses of complex structures.

The strength design of reinforced concrete (RC) rectangular sections for combined compression and in-plane bending with two levels of reinforcement is indeterminate: three unknowns are to be solved, but with only two equilibrium equations; an additional condition is necessary to solve the problem. The additional condition leads to the finding of a minimum reinforcement-concrete ratio. This paper proposes a new approach based on the equivalent eccentricity of the applied compressive load. Different domains are reported, each of which is associated with given values of eccentricity and axial load. Analytical expressions for the domain boundaries are established, and a simple procedure is described to outline the conditions corresponding to the optimal reinforcement. The main advantage of this procedure is its simplicity, which allows manual computations. Some examples employing reinforcement sizing diagrams illustrate the validity of this approach.

Many departments of transportation use four-chord aluminum tubular truss structures to support dynamic message signs (DMS). Inspections have often revealed that these structures contain cracks that likely occurred during or shortly after fabrication or were caused by fatigue while in service. Nearly all of these cracks are located at welded tube-to-tube connections, such as the diagonal-to-main chord fillet welded connections. Both weld toe cracks and surface-breaking root cracks have been observed. Destructive testing was conducted to evaluate the after-fracture reserve strength (AFRS) of 21.3-m-span and 16.8-m-span aluminum four-chord trusses. The dead weight of a DMS was simulated by attaching steel beams of the same weight eccentric to one exterior side of the truss near midspan. Wind load was simulated by hydraulic jacks and a whiffle-tree loading configuration. Simple steel frames were fabricated to mimic the sign truss end supports. Realistic boundary conditions, similar to the truss-to-end frame connections were provided by U-bolt connections and saddles found in the field. Fractures in up to two chords at the same cross section were simulated by cutting the members with a reciprocating saw. The testing revealed that both 21.3-m and 16.8-m span trusses have a considerable amount of reserve strength in the severely faulted (i.e., several members severed) condition. Testing of a main chord retrofit was conducted on the 16.8-m truss. Simple frame three-dimensional (3D) finite-element analysis was also conducted, and the predictions compared to the experimental data. Based on the results, recommendations are presented regarding how to respond if cracked members are discovered during inspection. The results are believed to be applicable to other similar trusses with span lengths up to 21.3 m.

Limited experimental results have been reported in the literature on the fatigue and ultimate static loading behavior of composite beams subjected to negative bending moment. This paper examines experimentally the behavior of composite steel-concrete beams. Eight composite specimens were tested to study the various aspects of composite beams under a negative bending moment, including the effects of repeated loading, shear connectors (studs and PBLs), rubber-latex mortar coating, and steel fiber-reinforced concrete (SFRC) on their structural performance. Load versus midspan deflection, crack formation and its developing process, slip distribution on the steel-slab interface, and the flexural strain results of shear connectors were measured and studied. The test results show that the initial cracking-level repeated load did not show obvious effects, while the stabilized cracking-level repeated load can reduce the rigidity and loading capacity of composite beams. The experimental results indicate that the plastic bending moment criteria in the current AASHTO load and resistance factor design specifications are typically conservative for composite beams under a negative bending moment. In addition, the effects of SFRC on crack width control and adhesion bonding effects of rubber latex on the steel-slab interface were confirmed in the testing.

An efficient material model was developed for reinforcing steel in reinforced concrete (RC) columns that implicitly incorporates the degrading effects of bar-buckling. Utilizing the beam-on-springs model developed by the writers to characterize a buckling bar in a RC column, a comprehensive parametric study was performed to identify the effects of several important column parameters on the buckling behavior of the longitudinal reinforcement in RC columns. Features of average stress-strain curves of compressively loaded reinforcing bars are summarized. Constitutive relations as functions of critical column parameters were developed for direct use in fiber-section model-based nonlinear analysis of RC structures. Comparison of the numerically simulated global response of experimentally tested RC columns confirm the validity and utility of the proposed material model as a simple and effective way to include bar-buckling effects in inelastic analysis of RC frame structures.

New reinforced concrete (RC) walls are popular in seismic rehabilitation of RC buildings as they reduce seismic-deformation demands. Their effectiveness was experimentally investigated by pseudodynamically testing $3:4$ scaled models of three four-story frames infilled with RC. The specimens were considered part of a frame building, the rest of which was substructured in the test as elastic. Unlike in past tests, the wall web was not thinner than the frame members and the specimens were tall enough to be governed by flexure. Two code-conforming designs were tried for the connection of the web to the surrounding frame members, one of them being less labor-intensive. Behavior and failure were dominated by flexure, but open, u-shaped FRP jackets at the two edges of the composite wall were essential to prevent premature failure near the base because of poor detailing of the columns. Slippage/separation at interfaces between the web and surrounding frame members was minor for both connection details. In one specimen the critical plastic hinge did not form at the base but instead in a horizontal band defined by the first-story beam of the frame, in which the vertical beam stirrups were much weaker than the vertical wall reinforcement; however, this did not adversely affect the global energy-dissipation and deformation capacity. The values of the secant-to-yield-point stiffness, yield moment, and ultimate chord rotation of the composite walls inferred from the tests are in good overall agreement with expressions given in international codes for monolithic walls with the same geometry, reinforcement, and detailing.

This paper employs a methodology for probabilistic response analysis based on the first-order second moment (FOSM) method in conjunction with response sensitivity computation through the direct differentiation method (DDM), to study the variability of the structural response of steel-concrete composite (SCC) beams. This methodology is applied to compute the first-order and second-order statistical moments of the response of two actual structural systems for which experimental data are available. The results of the DDM-based FOSM method are compared with the experimental measurements and with the results of the computationally more expensive Monte Carlo-Simulation (MCS) method. Different modeling hypotheses for the material parameter uncertainty are considered. The DDM-based FOSM method agrees very well with the MCS results for low-to-moderate levels of response nonlinearity under low-to-moderate material parameter uncertainty and up to high level of response nonlinearity under low material parameter uncertainty. The DDM-based FOSM method is shown to correctly describe the effects of random spatial variability of material parameters.

Experimental tests and subsequent finite-element (FE) calibration studies were conducted to study the static strength of axially loaded X joints between elliptical hollow sections (EHS) with braces welded to the wide sides of the chord. Good agreement between the experimental and numerical results was achieved. The behavior of the EHS joints was found to be similar to that of rectangular hollow section (RHS) joints. For the EHS X joints with a brace-to-chord diameter ratio ($\beta$) up to 0.80, the failure mode is chord plastification. For those with equal brace and chord diameters ($\beta =1$), the failure mode is chord sidewall failure. Various aspects of FE modeling, such as weld size and shape, chord length, and chord-end boundary conditions were investigated through sensitivity studies. These results constitute a solid basis for the subsequent numerical parametric study.

A research program consisting of experimental tests and finite-element (FE) analyses has been carried out to study the behavior and static strength of axially loaded elliptical hollow section (EHS) X joints of two configurations, for which the EHS braces are welded to the wide sides of the EHS chord. The experiments with the numerical calibration are presented in Paper I. This paper deals with the numerical parametric study and the determination of strength functions for the two types of joints studied. The strength equations are formulated in a similar way to those currently adopted for rectangular hollow section (RHS) joints. Limit state design equations can be derived according to current design rules in the IIW and CIDECT recommendations.

Mean recurrence intervals (MRIs) of wind effects for the strength design (SD) of flexible buildings by the wind-tunnel method are based on estimates of total uncertainties that do not account for uncertainties in the dynamic parameters. This paper presents a procedure for assessing this practice. The procedure accounts for the probability distributions of total uncertainties in basic wind effects corresponding to MRIs of 50 or 100 years. The total uncertainties are estimated for two cases. For Case 1, only uncertainties in the wind loading are taken into account. For Case 2, uncertainties in both the wind loading and the dynamic parameters are considered. Cumulative distribution functions (CDFs) of the total uncertainties in the basic wind effects are determined by Monte Carlo simulation. To assure risk consistency, the MRIs of wind effects considered for SD correspond in both cases to the same upper confidence bound of the basic wind effects. For a 305-m-tall symmetrical steel building, the requisite MRIs are larger for Case 2 than for Case 1 by at most 80%. These increases corresponded to wind effects considered for SD larger for Case 2 than for Case 1 by at most 6%.

A full-scale experimental program consisting of testing four as-built diaphragms and four retrofitted diaphragms in both principal loading directions is presented. As-built configurations were typical of those found in historic unreinforced masonry buildings in North America and Australasia, whereas retrofitted diaphragms consisted of plywood panel overlays with stapled sheet metal blocking systems (SMBS). Test results were characterized using bilinear representations to establish recognizable performance parameters such as shear strength, shear stiffness, and ductility capacity, which were then used for comparative analysis. The nonlinear and low stiffness behavior of as-built diaphragms was confirmed in each principal loading direction. The plywood overlay and SMBS dramatically improved as-built diaphragm shear strength and shear stiffness and were shown to perform satisfactorily from a serviceability perspective. The orthotropic nature of as-built diaphragms was proven, with perpendicular-to-joist shear stiffness being as low as 68% of the corresponding orthogonal value. A typical duly framed stairwell penetration and discontinuous joists with two-bolt lapped connections were shown to have no detrimental impact on tested diaphragm performance. Predicted diaphragm performance using state-of-art assessment documents was shown to be inconsistent with corresponding values established from testing. It is recommended that these assessment procedures be updated with revised performance parameters and provisions to address diaphragm orthotropic behavior.

A fiber beam-column model considering the slab spatial composite effect for nonlinear analysis of composite frame systems is proposed and implemented in a general commercial finite element (FE) package. A conventional fiber beam-column model of composite beams is first developed and the computational procedure of this model is presented. A hysteretic law which can consider the strength degradation attributable to repeated reversals of unloading and reloading is proposed to more reasonably and accurately trace the actual complex nonlinear behavior of the concrete material. Furthermore, the formulas for calculating the effective flange widths at two critical limit states are combined with the conventional fiber model so that an improved fiber model, which can reasonably consider the slab spatial composite effect, is realized through the modification of the uniaxial constitutive laws of slab reinforcement and concrete. Finally, the simulations of several tested composite frame structures show that the proposed model can reasonably trace the complex slab spatial composite effect of composite frame systems.

Linear and nonlinear response history analyses (RHA) are often conducted for predicting the seismic demand and capacity of structures. The proper selection of ground motions is critical to obtaining reliable results from RHA. The purpose of this study was to develop a computationally efficient and accurate procedure for selecting and scaling ground motions matching a target response spectrum with zero variance, such as a uniform hazard response spectrum, and a design response spectrum that can consider variance in nature. In the proposed procedure, one ground motion at a time is selected and scaled from a ground motion library to make the mean response spectrum of the selected ground motions match a target response spectrum. The process is repeated until a desired number of ground motions is collected. The accuracy and efficiency of the proposed procedure is verified by the selection and scaling of seven ground motions from libraries containing 40 and 124 ground motions.

The continuum model constitutes a simple and efficient tool to analyze high-rise wall-frame buildings. It has been commonly used in recent decades to analyze the behavior of these structures. The related equations are revisited to study the effect of the calculation precision on the determination of the optimum level of wall curtailment. The results obtained illustrate the influence of calculation precision in specifying the optimum level of wall curtailment. The relationship between the curtailment level and the resulting internal forces is investigated. The level of curtailment which results in the minimum top deflection of the structure eliminates, at the same time, the negative moments and negative shear forces in the wall.