Journal of Structural Engineering

A continuum analysis for determining the flexural‐torsional vibration of suspension bridge under vehicular movement is presented. The dynamic analysis duly considers the nonlinear bridge‐vehicle interactive force, eccentricity of vehicle path, surface irregularity of the bridge pavement, cable‐tower connection and end conditions for the stiffening girder. The surface irregularity of the bridge pavement is considered as a stationary random process characterized by a power spectral density function (PSDF) of the surface irregularity. The irregular profile of the bridge pavement is synthetically generated from the PSDF using Monte Carlo simulation technique. The analysis makes use of normal mode theory and calculates the response in time domain utilizing an iterative scheme and a few closed‐form expressions. The responses are obtained and compared for three types of vehicle models, namely, 3‐D, 2‐D, and single sprung mass system. A parametric study is also conducted to investigate the influence of some important bridge and vehicle parameters on the dynamic behavior of the bridge.

Vibrational behavior of simple‐span highway girder bridges with rough surfaces due to heavy trucks is discussed. The causes of vibration and dynamic characteristics of bridge structures are investigated in the time and frequency domains. An appropriate vehicle model for vibrational analysis of bridges is found by comparing dynamic responses from four different vehicle models, and impact factors are calculated using the suggested vehicle model with different speeds, deck roughnesses, and span lengths. The results are compared with the values specified by several standard codes, and it is found that the current design codes tend to underestimate impact factors especially in long‐span bridges with rough decks. Finally, empirical formulas for impact factors represented in terms of span length, vehicle speed, and surface roughness of a bridge deck are suggested from the multiple linear regressional using the data obtained from this study.

In the present study, the boundary element method is employed to analyze the top slabs of rectangular or skewed composite slab‐on‐girder and box‐girder bridges, while the finite strip method is used to analyze the girders. The boundary elements and the finite strips are connected by transition strips. One side of the transition strip coincides with the nodal line of the adjacent finite strip and has the same degrees of freedom; while on the opposite side, a number of nodes coincides with the nodes of the neighboring boundary elements. Within the transition strip the displacements are expressed in terms of the degrees of freedom of the finite strip nodal line and the boundary element nodes by their corresponding shape functions. In this way, a fully compatible connection between slab and girders is achieved. This combined analysis is very effective in calculating the local bending moments due to moving vehicle wheel loads.

Force transfer from steel to concrete without mechanical transfer devices is investigated in composite columns at points of load introduction. The subject is studied experimentally by means of several tests on short composite columns, and theoretically by means of a numerical model. A mechanism of force transfer and the factors influencing the transfer are identified. The important factors are quantified and incorporated into a proposed simplified design method. The most important result obtained is that shear resistance values are highly dependent on the size of the embedded steel section. Lower shear resistance values than those found in current design codes were obtained for large embedded steel sections.

Based on wind‐tunnel measurements of wind‐induced response of tall buildings, a semianalytical method of performing parametric study of tuned mass dampers for tall buildings is proposed. This method was verified through a series of aeroelastic model tests of buildings with and without mass dampers, and a close agreement between semianalytical and experimental results was found. The method can be used to provide an accurate estimate of motion reduction of a building without having to conduct aeroelastic model tests of mass damper‐building systems. Both the analytical and experimental results show that the effectiveness of tuned mass dampers depends to some extent on the type of wind excitation. The conventional parametric study method, which is based on a white‐noise excitation model, usually overestimates the performance of tuned mass dampers. The practical application of the proposed method and the selection of the optimal mass damper parameters are demonstrated. The limitation of the present study and the required further research are pointed out.

There is a damage potential for the columns of building frames due to either unexpected causes such as vehicle impact, boiler explosion, or terrorist attack, or due to design or construction deficiencies aggravated by an earthquake, severe wind, or excessive foundation movement. This paper investigates the redistribution paths of released forces resulting from a column failure and identifies the basic structural defense mechanisms developed in a damaged building frame. Extensive analytical studies revealed that the effect of a column failure is localized to the beams connected to the vertical axis of the failed column and to the adjacent columns. Although the load‐carrying capacity of the connected beams is far exceeded by the redistributed forces, the presence of even light architectural infill walls reduces the beam forces remarkably. It is observed and analytically demonstrated that a partially infilled building frame may survive a base‐column failure without any damage to the other members. Further, it is shown that the redistributed internal forces can be predicted with reasonable accuracy by employing basic structural analysis.

This paper presents an experimental study of the behavior of large‐scale high‐strength concrete columns confined by rectangular ties under concentric loading. Effects of key variables such as the concrete compressive strength, the tie yield strength, the tie configuration, the transverse reinforcement ratio, the tie spacing, the longitudinal reinforcement ratio, and the spalling of the concrete cover are studied in this research program. The behavior of high‐strength concrete columns is characterized by the sudden separation of the concrete cover, leading to a loss of axial capacity before the lateral confinement becomes effective. After the concrete is completely spalled, important gains in strength, toughness, and ductility are recorded for the concrete core of well‐confined columns.

The effects of loading eccentricity on axial strength of cold‐formed steel lipped channels are investigated in a series of 48 individual long‐column tests. Experimental results are compared with flexural and torsional‐flexural buckling‐strength predictions based on the American Iron and Steel Institute (AISI) 1986 specification, with local buckling treated using the uniform‐effective‐width approach. Widely spaced, rectangular web perforations are modeled using an “unstiffened‐strip” method, with the web idealized as two unstiffened elements, one on either side of the perforation. A method for predicting weak‐axis flexural buckling and the load‐versus‐lateral‐deflection response of eccentrically loaded, initially imperfect columns is also examined. Estimates of the actual loading eccentricities are made by comparing the predicted and empirical load‐versus‐deflection responses. Axial strength is shown to be very sensitive to small eccentricities. Conservative strength predictions are obtained with the AISI procedures by including loading eccentricity effects and modeling the perforations using the unstiffened‐strip method.

This is the first of two papers that describe the role of in‐service inspection/repair in maintaining the reliability of concrete structures during a projected service period taking into account the randomness in existing damage and in damage detection. The degradation in strength of a component due to environmentally aggressive factors is evaluated based on a growth model for individual random damage that can be obtained from experimental data. The distribution function of damage intensity after an inspection and repair is updated using a Bayesian analysis. This methodology can be used to determine inspection/repair strategies that are necessary to keep the failure probability of a concrete structure at or below an established target. Sensitivity analysis shows that the variance in the function describing degradation in strength is small even if the variance of individual damage intensities is large. However, the degradation function is sensitive to the threshold of damage detection of the nondestructive evaluation (NDE) method selected for in‐service inspection.

This is the second of two papers that describe the role of in‐service inspection/repair in maintaining the reliability of concrete structures taking into account the randomness of existing damage and damage detection. Since inspection and maintenance are costly, there are trade‐offs between the extent and accuracy of inspection, required level of reliability, and cost. The method to evaluate degradation in strength of a component described in part 1 is combined with the time‐dependent reliability analysis to devise optimum strategies for inspection and maintenance that minimize the expected future cost of structures and components, while maintaining their limit‐state probabilities at or below an established target failure probability. Optimum inspection/repair strategies are sensitive to the relative costs of inspection, repair, and failure as well as to the threshold value of damage detection. Inspection at approximately uniform intervals leads to near minimum costs for a wide variety of cases studied.

An analytical approach to determine the need to provide transverse reinforcement at boundaries of reinforced concrete structural walls with rectangular, T‐shaped, or barbell‐shaped cross sections is presented. By relating the expected displacement demands on the building system to the local deformations imposed on the wall cross section, the magnitude and distribution of wall normal strain is determined. The primary variables affecting the wall‐strain distribution are found to be the ratio of wall cross‐sectional area to the floor‐plan area, the wall aspect ratio and configuration, the wall axial load, and the wall‐reinforcement ratios. Based on the computed wall‐strain distribution, required transverse steel for concrete confinement and length of the wall cross section requiring concrete confinement is computed. The wall‐strain distribution is also used to evaluate required transverse reinforcement to restrain buckling of longitudinal reinforcement. The validity of the proposed analytical approach is demonstrated by comparison with an experimental study of a full‐scale building system conducted in Japan. In addition, applications of the proposed analytical approach are provided for preliminary design and for the evaluation of an existing building.

The slant‐legged rigid‐frame bridge is one type of the most widely used highway bridges. However, very few dynamic characteristics of this type of bridges have been reported. The objective of this paper is to investigate the dynamic response of a slant‐legged rigid‐frame bridge to one or two trucks (side by side) passing over the rough bridge deck. The bridge was modeled as a space bar system. Trucks were modeled as a nonlinear system with 12 degrees of freedom according to the HS20‐44 truck design loading specified in the American Association of State Highway and Transportation Officials (AASHTO) specifications. The road surface roughnesses of the approach roadways and bridge decks were generated from power spectral density function. Maximum dynamic responses of each rigid frame at different sections were determined for two trucks through changing their transverse positions with different speeds and road surface roughnesses.

The first U.S. full‐scale five‐story building test under simulated seismic loads was conducted at the University of California, San Diego. The full‐scale reinforced masonry (RM) research building consisted of coupled, flanged, fully grouted concrete masonry walls and precast prestressed hollow‐core plank floors with reinforced‐concrete topping. The seismic‐simulation tests were conducted under the direction of the Technical Coordinating Committee for Masonry Research (TCCMAR) to validate new analysis and design models for masonry buildings in seismic zones. The complete seismic simulation testing of the five‐story structural wall building under recorded earthquake ground‐motion segments was possible through soft coupling between the loading system and the stiff test structure, and improvements to the on‐line actuator‐control algorithm as part of the implicit time‐integration scheme. The development and implementation of the generated sequential displacement (GSD) procedure for the experimental multidegree‐of‐freedom seismic‐load simulation is described and illustrated by detailed examples of responses.

To verify new design guidelines for reinforced‐masonry (RM) buildings in seismic zones, a full‐scale five‐story RM research building was tested at the University of California, San Diego, under simulated seismic loads. These new design guidelines, developed by the Technical Coordinating Committee for Masonry Research (TCCMAR), combine limit state and capacity design principles to ensure ductile seismic response. Design and analysis models developed under TCCMAR for individual components and subassemblages were extended to the complete system to predict the seismic response of the five‐story research building, and full‐scale seismic‐simulation test data were used to provide detailed experimental response data for model verification. This first U.S. full‐scale five‐story seismic‐simulation test under laboratory conditions showed that even stiff‐structural‐wall‐type masonry buildings can be designed to exhibit large ductilities under extreme seismic loads by controlled inelastic flexural behavior in predetermined locations, and that analysis and design models exist that can predict the overall seismic response of the complete structural system.

System identification, three‐dimensional linear dynamic analysis, and code‐specified conventional equivalent static techniques are used to evaluate the response of a nine‐story flat‐slab building during the Loma Prieta, Calif., earthquake of 1989. Four different models are used in the finite‐element‐based dynamic and static analyses. The system identification results on natural frequencies, mode shapes, and interstory drifts are used to validate the four analytical models. It is found that the model with a 30% live load included in the mass computation, a $2/3$ reduction in the stiffness of slabs, and 30% reduction in stiffness of columns and shearwalls appears to give a reasonable overall approximation of the measured response. The eccentricity of the central shearwall core in the east‐west direction promoted torsional response. Both the dynamic analysis and system identification results indicate that the shearwall kept the drift level at the central core within the code limit. At the east and west ends of the building, where frame action dominates, the drift level reached the code limit. The horizontal forces were mostly resisted by the core shearwalls. The base shear coefficient in the north‐south direction determined from the dynamic analysis was close to that of the equivalent static procedure. In the east‐west direction, however, the dynamic base shear coefficient was twice as much.

The implications that bar softening has on the design of anchorages in earthquake‐resistant concrete structures are considered. A relatively simple one‐dimensional finite‐element formulation is developed to model the response of anchorages under monotonic displacement control to failure. Resistance mechanisms for the cases with and without bar yield are compared and a rationale is presented to determine the terminal depth of yield penetration. Critical anchorage conditions are examined to establish anticipated demand levels in earthquake‐resisting structures. A parametric study of the sensitivity of anchorage performance to confinement conditions, postyield characteristics of reinforcement, bar size, and available development length was made using the analytical formulation. Trends in performance of yielding anchorages with the variables considered provide information useful for extending design standards.

Tests of anchorages long enough to develop postyield steel strain under monotonic loading are presented. The influence of varied postyield properties of steel on load‐slip response is considered. The sensitivity of anchorage performance to variable steel properties is illustrated by the relationship of slip to bar strain at the loaded face. Behavior of the specimens is shown to be very sensitive to particular confinement and internal force conditions of the basic specimen form. It is recommended that tests of longer anchorages be made with specimens modeling specific structural conditions of interest.

A composite masonry wall consists of a concrete‐block wythe, a clay‐brick wythe and a cavity (i.e., collar joint) between the two wythes that is filled with mortar or grout. The in‐plane loads on a composite wall due to gravity and/ or wind are generally applied on the block wythe and can produce large interface shear stresses in the collar joint. The composite action of the wall is ensured only if these shear stresses are less than the shear strength of the interfaces. A refined finite element procedure based on interface shear stiffness and considering stress singularity, which can accurately predict collar‐joint interface shear stresses in composite masonry walls, is proposed in this paper. This paper shows that there is no need to consider reduced shear stiffness for the interface elements. Additionally, accurate shear stresses at a collar‐joint interface can be obtained for discontinuously applied loads by disregarding stresses in the two elements closest to the point‐of‐stress singularity. Use of the proposed model permits accurate assessment of the interface shear stress in the collar joints of composite masonry walls.

Strenghtening of concrete structures with fiber‐reinforced plastic (FRP) composites epoxy‐bonded to their tension faces appears to be a method with a high potential for increasing the strength and stiffness of existing members. The writers aim at developing a fundamental understanding of the time‐dependent (creep and shrinkage) behavior of reinforced‐concrete beams strengthened with FRP laminates. An analytical procedure is given first for the deformation of cross sections based on the age‐adjusted effective‐modulus method for concrete, Findley's model for composite materials, and a stress‐relaxation procedure to yield the cross‐sectional stresses and strains as a function of time. Parametric studies assessing the effect of the type and the area fraction of composite material and of the area fraction of steel reinforcement on the long‐term response of strengthened elements are also presented. Finally, the analytical model is used to predict the deflections of concrete beams strengthened with FRP laminates of different thicknesses, and an experimental program is described confirming the analysis.