Journal of Structural Engineering

Dynamic response of structures subjected to earthquake-induced base excitations are often simplified by ignoring the tilt components of ground motion. However, close to the earthquake source, tilting of the ground surface may become significant. Based on strong-motion records at the Pacoima Dam–upper left abutment obtained during the 1994 Northridge Earthquake, residual tilt reached 3.1° in the N45°E direction while the dynamic tilt remained higher. This study investigates the consequences of neglecting the effects of the tilt component in ground motions on elastic and inelastic spectral ordinates. A complete equation of motion for a single-degree-of-freedom (SDOF) oscillator is developed which includes the effects of tilt (as the secondary $P–\Delta$ effect) in addition to inertial force effects due to angular and translation accelerations. The expected values of the largest response peaks are computed for the translational and tilting excitations to investigate the relative contribution of each forcing function. The coupled tilt and translational ground motion response spectrum (CTT spectrum) is generated considering elastic and inelastic response of SDOF oscillator. The CTT spectrum reflects kinematic characteristics of the ground motion that are not identifiable by the translational ground motion response spectrum alone and therefore emerges as a distinct intensity measure of translational ground motion when it is coupled with dynamic tilting of the ground surface on the order of few degrees.

A frequency-domain method is developed for evaluating the earthquake input energy to two building structures connected by viscous dampers. It is shown that the earthquake input energies to respective building structures and added viscous connecting dampers can be defined as works done by the boundary forces between the subsystems on their corresponding displacements. Fourier and inverse Fourier transformations enable one to transform the earthquake input energy in time domain to that in frequency domain. The earthquake input energy in frequency domain has an advantage that, while the input energy cannot be decomposed into the term for the structural parameters and that for the parameters of input ground motions in the conventional time-domain approach, it is possible in the proposed frequency-domain approach. It is demonstrated that the proposed energy transfer function is very useful for clearer understanding of dependence of energy consumption ratios in respective buildings and connecting viscous dampers on their properties. It is concluded that the total input energy to the overall system including both buildings and connecting viscous dampers is approximately constant regardless of the quantity and location of connecting viscous dampers. This property leads to an advantageous feature that, if the energy consumption in the connecting viscous dampers increases, the input energies to the buildings can be reduced drastically.

Although traditional mechanical dampers can help reduce cable vibrations, they are not the most effective solution due to their position restrictions. Tuned mass dampers (TMDs) were proposed to overcome this limitation and have been proven effective for cable vibration mitigation according to previous research. However, the effectiveness of TMDs is very sensitive to the vibration frequency, which implies that TMDs designed for a specific vibration frequency may not be efficient for other frequencies. A new type of mechanical damper named the TMD magnetorheological (MR) damper system is proposed in the present study to address this problem. The feature of this proposed damper is the combination of the position flexibility of TMDs and the adjustability of MR dampers. The proposed damper system is attached to a $7.16\text{-}\mathrm{m}$ -long cable to investigate its vibration reduction effectiveness and the dynamic properties of the combined cable-damper system. Experimental results show good vibration reduction effects of the TMD-MR system. The present study focuses on a conceptual exploration of the TMD-MR damper system for cable vibration mitigation through an experimental approach.

An accurate multidirectional structural testing system is developed. This testing system is used to carry out a multidirectional pseudodynamic test with the help of a unique three-dimensional hinge and a spatial truss-type transducer. With this testing system, multidirectional seismic behaviors of thin-walled circular steel columns are examined in comparison with their in-plane behavior. Further, the accuracy of the testing system is confirmed by the numerical results of nonlinear FEM shell analyses.

There is a common perception that seismic detailing can improve the collapse resistance of steel frame buildings. However, the effect on connection performance of the potentially large catenary, i.e., tensile, forces that can develop during collapse has not yet been adequately studied. The objective of this paper is to use computational structural simulation to investigate a number of key design variables that influence formation of catenary action in steel special moment resisting frame subassemblages. The numerical model used in the study employs a calibrated micromechanical fracture model and is validated using existing test data. The simulation results demonstrate the ductility of seismically designed special moment frame connections and their ability to deform in catenary mode. It is shown that connection ductility and strength are adversely influenced by an increase in beam depth and an increase in the yield to ultimate strength ratio and that the beam web-to-column detail plays an influential role in connection response. A number of conclusions with practical implications are drawn from the numerical results.

Micromechanics-based models that capture interactions of stress and strain provide accurate criteria to predict ductile fracture in finite-element simulations of structural steel components. Two such models—the void growth model and the stress modified critical strain model are applied to a series of twelve pull-plate tests that represent reduced (or net) section conditions in bolted and reduced beam section connections. Two steel varieties, A572 Grade 50 and a high-performance Grade 70 bridge steel are investigated. The models are observed to predict fracture much more accurately than basic longitudinal strain criteria, by capturing stress–strain interactions that lead to fracture. The flat stress and strain gradients in these pull plates allow the use of relatively coarse finite-element meshes providing economy of computation while capturing fundamental material behavior and offering insights into localized ductile fracture effects.

Unstiffened, unbraced steel beams used in bridge falsework are subjected to patch loads from timber or steel posts. This paper investigates the critical web limit states in these types of beams experimentally and through finite-element analysis. Existing equations were appropriate for calculating the web yielding and crippling capacity, as observed in experiments, except that a 1:1 stress gradient through the flange and fillet of the web is found to be more appropriate and conservative for web yielding than the 2.5:1 gradient assumed in the current AISC specifications. Lateral deformation of a beam flange, due to buckling of the web, was also observed experimentally. The capacity of the web for resisting lateral buckling can be calculated assuming that the web of the beam acts like a column with an appropriate effective area and length factor. The effect of accidental eccentricity between the flange and post is not significant if limited to three times the web thickness. Blocking is approximately 50% effective in increasing the web yielding and crippling capacity. Bracing or stiffeners should be used to prevent lateral buckling of a web where necessary.

The lateral–torsional buckling behavior of general prismatic and tapered steel members with doubly and singly symmetric sections are investigated based on the general forms of differential equilibrium equations for minor-axis bending and torsion. Fundamental lateral–torsional buckling moments are derived for general tapered steel members that are in forms similar to those for prismatic members. The factor accounting for the effects of level loading application is derived and incorporated into the moment diagram modification factor that is generalized for the prismatic and tapered steel members. American Institute of Steel Construction-type equations for calculating the inelastic lateral–torsional capacity are proposed by using straight-line transition, and limiting unbraced member lengths are explicitly derived. The analytical procedure presented can give reasonable results with consistent accuracy for general design consideration and follows the same philosophy as for prismatic doubly symmetric wide-flange design rules. The proposed procedure is user friendly for structural design.

Tests on steel beams with adhesively bonded elastic reinforcing strips are used to deduce the mechanics of stress build up in the steel-to-strip interface up to failure. The solely elastic nature of the strips is a key concept in this study, as this elasticity leads to high interface stresses due either to nonlinear stiffness disparities between the steel and the reinforcing material when the steel yields, or to nonlinear geometric effects which are parasitic on this elasticity, including buckling of the strips and sharp section changes along the reinforced beam. Four features have been used in the tests to study these nonlinear effects, namely, multi-layer multi-length strips, imperfections in the adhesive, strips in both compression and tension, and tapering of the strip. The test results suggest that interface stresses may be sensitive to through-thickness variations of axial strain in the strips and to local variations of interface quality. Data from a nonlinear finite-element analysis, based on a novel two-layer interface element, are given for further insight into interface stress profiles. Conclusions are drawn on interface performance in the tests. Improved measurement strategies are suggested for shedding further light on the stress-transfer mechanics of this interface. The term connection is used synonymously with interface in this paper.

The transfer of forces across the interface between steel bars and concrete through bond is one of the most important aspects that control the behavior of reinforced concrete members. This paper summarizes the results of a series of experimental and analytical studies undertaken by the writer at the American University of Beirut for evaluating the bond characteristics of developed or spliced bars in tension under different design and strength variables. Particular emphasis is placed on the evaluation of the average bond strength that develops in steel bars at bond failure by integrating experimentally based local bond law(s) into a numerical analysis technique of the bond problem. A general experimentally derived local bond stress–slip relationship corresponding to the splitting mode of bond failure is presented. With simple modification of the characteristic parameters, the relationship is applicable for plain normal-strength concrete, high-strength concrete, concrete confined with transverse steel ties, fiber-reinforced concrete (FRC), and concrete confined externally with fiber-reinforced polymer (FRP) laminates. The accuracy of the numerical analysis results and the local bond law(s) was verified by comparing them to a large body of test data. The mechanism by which the concrete compressive strength and concrete confinement influence the bond strength is discussed, and generalized expressions of the bond strength of steel bars embedded in FRC- or FRP-confined concrete are presented.

Aspects relating to codification of the in-plane shear strength of concrete masonry walls when subjected to seismic loading are presented in this paper. Particular emphasis is placed on a model that is capable of representing the interaction between flexural ductility and masonry shear strength to account for the reduction in shear strength as ductility level increases. The simple method proposed here allows the strength enhancement provided by axial compression load to be separated from the masonry component of shear strength and is considered to result from strut action. In addition, minor modifications are made to facilitate adoption of the method in the updated version of the New Zealand masonry design standard, NZS 4230:2004. Prediction of shear strength from NZS 4230:2004 and alternative methods are compared with results from a wide range of masonry walls tests failing in shear. It was established that the shear equation in the former version of the New Zealand masonry standard (NZS 4230:1990) was overly conservative in its prediction of masonry shear strength. The current National Earthquake Hazards Reduction Program (NEHRP) shear expression was found to be commendable, but it does not address masonry shear strength within plastic hinge regions, therefore limiting its use when designing masonry structures in seismic regions. Finally, the new shear equation implemented in NZS 4230:2004 was found to provide significantly improved shear strength prediction with respect to its predecessor, with accuracy close to that resulted from NEHRP.

The focus is on understanding the behavior of walls with regard to the propensity of studs to warp at different moisture content. Experimental tests on whole wall structures were conducted. The walls were built up using studs of Norway spruce (Picea abies) and cladding made of gypsum wallboard. The tests were performed simulating in-service conditions, i.e., about 35% relative humidity and a temperature of $22°\mathrm{C}$ . Tests were conducted with double- and single-sided cladding. Twist and crook were measured on the studs continuously during the tests. The warp of the studs was also measured before and after each test. The double-sided cladding reduced the twist in the studs significantly, by about 75% when compared with twist developed in the same studs when vertically suspended without restraint. The single-sided cladding only slightly reduced the twist, by about 13%, but increased the crook substantially. The value for crook for single-sided cladding was more than double and had reached more than half its final value in less than a week.

This paper analyzes the load-carrying capacity of timber–concrete joints made with dowel-type fasteners. The scope of the work was to obtain information to be used in the design of timber–concrete composite structures. In order to achieve that, two different approaches were used: experiments with timber–concrete joints loaded in shear, and the use of nonlinear analytical models. Shear tests were performed using various timber species, concrete mixtures, and fasteners. The results from these tests are presented and compared with the results determined with three different analytical models. These models differ in the approach used to simulate concrete: linear–elastic, linear–elastic with crushing, and elastic–perfectly plastic. The analysis of the test results shows that the load-carrying capacity of this type of joint is significantly influenced not only by the strength of the materials but also by the shape of the fastener. From the three models proposed, the one assuming elastic–perfectly plastic behavior for concrete leads to the results closer from the test results, however, the best statistical correlations between model results and test results were obtained with the model assuming linear–elastic behavior with crushing.

The finite-element method for simulating the nonlinear behavior of reinforced concrete structures has progressed to the point where it is close to being a practical everyday tool for design engineers. Further advancements have made the analysis of arbitrary loading conditions, including reverse cyclic loading or earthquake-type loading, feasible. Recent criticism has questioned the practicality, reliability, and robustness of the finite-element method due to perceived complexities involved in developing the model and interpreting the results. A series of analyses are presented on reinforced concrete structural walls of varying height-to-width ratio, varying wall cross section, and varying levels of reverse cyclic loading to demonstrate that the finite-element procedure is capable of providing quick and reliable simulations, while employing simple modeling techniques. The modeling herein utilizes low-powered rectangular membrane elements, and material properties are smeared within the elements. Behavioral aspects such as ultimate strength, displacements, postpeak ductility, energy dissipation, and failure mechanisms are well simulated.

This paper presents a “rotational spring” analogy which allows the geometric stiffness for a large class of structural problems to be formulated using common notions from linear structural analysis. The proposed analogy offers an intuitive framework which (1) sheds significant light on geometric nonlinearity in general, and buckling analysis in particular; (2) explains previous buckling analysis procedures, and (3) is directly applicable to the simplified buckling analysis of discrete as well as continuum structures. This paper discusses the application of the proposed analogy to single degree-of-freedom, multidegree-of-freedom, as well as continuous structural systems, providing in the process several illustrative examples. These demonstrate the simplified framework offered by the proposed rotational spring analogy for a wide range of buckling problems, and highlight its conceptual power in furthering the understanding of important issues in geometrically nonlinear analysis.