Journal of Structural Engineering

This paper is the first part of a two part summary of a series of seismic tests on a full‐scale steel building. The design and construction of the test structure are described and the instrumentation and the test procedures are noted. The phase I tests consists of a series of tests on a concentrically braced frame. During the moderate test a connection panel zone failed, and this resulted in a reduction in the strength and stiffness of the structure. However, the connection was repaired and the test was continued. Brace buckling occurred at a number of levels in the final test and ultimately led to a rupture which concluded the phase I testing. These tests provide valuable information on the elastic and inelastic behavior of braced frames, and they further illustrate the importance of the connection details and ductility in seismic design.

This paper is the second part of a two‐part summary of a series of seismic tests on a full‐scale steel building. The damage repair required after the phase I test is described, and the Phase II, eccentrically braced frame test results are presented. The eccentrically braced frame behaved well and dissipated large amounts of energy in the eccentric link until a gusset plate failure occurred. This failure resulted in a dramatic reduction in strength and stiffness; it demonstrates the importance of connection detail on seismic behavior. The phase III, moment‐resisting frame, and phase IV, nonstructural element tests are also described. The moment frame was very ductile despite the many prior tests, but its stiffness was much smaller than the braced structures. The nonstructural elements added considerable stiffness to the structure, but it deteriorated under cyclic loading. The sequence of failure of these elements is described and correlated to the deformation of the structure.

Reinforced concrete slabs with edges restrained against lateral displacement are capable of carrying ultimate loads which are considerably higher than the loads predicted by using conventional yield‐line theory. An analytical relation for the load‐deflection behavior is derived in the paper by using a plastic‐flow theory. The compatibility condition introduced in the paper for the deformation of the centerline of the slab incorporates boundary displacements as well as the in‐plane deformations of the slab. Combining the compatibility condition, the flow rule associated with the load‐moment yield criterion, and the axial equilibrium condition, a nonlinear differential equation is obtained for the membrane force in the slab. The load on the slab is then calculated for a given deflection by balancing the external and internal energy dissipation. The predicted load‐deflection curves are compared with those obtained from two test series and a good agreement is obtained up to the peak load. The load‐deflection curve beyond the peak load is found to be predicted better by the incorporation of an analysis which approximates strain softening of the concrete.

The effect of parapets on the wind loading of flat roofs has not been examined in detail in the past. This paper refers to the experimental data obtained in a recent study of the effects of wind on a variety of flat roofs with and without parapets when exposed to simulated atmospheric boundary layer flows (open country and urban exposure). Geometrical parameters examined include the effect of building height (12, 24, 48, 96, and 145 m) and parapet height (0‐3 m) on both local and area‐averaged roof pressures for a variety of wind directions (0°, 30°, 45°, 60°, and 90°). Results show that parapets generally reduce the high suctions on roof edges and may slightly increase the suctions on the interior areas of the roof. Roof‐corner suctions, however, increase significantly for low parapet heights. This is in contrast to the current American National Standards Institute (ANSI) provisions for the effects of parapets on roof suctions. Suggestions for the incorporation of the results of the study into wind standards and codes of practice are made.

A method is presented for the inspection and repair of fatigue cracks that occur in the bottom tension flanges of crane runway girders. The date of first inspection and the interval of inspection are determined from fatigue crack initiation and propagation tests, respectively. The toes of the welds should be finished by rotary grinders to increase the crack initiation life. A wet‐type magnetic test appears to be the most promising tool for crack inspection. Methods of repair by welding are also presented.

A method for selecting the optimum design variables in the design of horizontally curved two‐way reinforced concrete floor slabs is presented. The slab panels are assumed to be uniformly loaded by characteristic dead and live load and supported on all sides by rigid boundary supports. The optimum values are selected from within the feasible region in the design space defined by the limit state requirements of CP 110 (British Standard Code). Cost of material is treated as the objective function and is optimized subject to the bending behavior and side constraints imposed by the limit‐state requirements. For the ultimate limit‐state analysis, the bending moments in the curved slab panels are determined from the linear‐elastic theory of sector plates. Nonlinear programming problems of high degree are generated by setting the effective depth and the transformed steel ratios as the design variables in both the objective function and the constraints. The sequential linear programming is introduced as an analytical method of nonlinear optimization. The optimum design variables are then determined.

In recent years there has been a dramatic increase of interest in sandwich cladding panels consisting of metal faces and a rigid foam core. Previous studies have concentrated on the behavior of these panels when subject to bending action due to wind and snow load and under temperature gradient. Comparatively little is known about their behavior under axial compressive load when acting as load‐bearing wall panels. This paper first considers appropriate methods of analysis for sandwich panels subject to combined axial compressive load and bending moment and an exact finite element is derived. Some early and previously unpublished tests are described and the results compared with the theoretical values. It is concluded that great care must be taken in the design of load bearing wall panels because of the interaction between local and global buckling and their sensitivity to eccentricity of load.

An effective procedure for limit analysis of plane frames under proportional loading is developed, within the framework of simple plastic theory. The proposed method is sequential in the sense that a strength (static) analysis precedes deformation analysis. The static analysis starts from the independent mechanisms of the structure, which are generated in a systematic and efficient way. Then, through an energetic theorem, deformation analysis is cast as a linear programming (LP) problem, in which extensive use is made of the previous information from the static analysis. Passive hinges are detected by a posteriori check of moment and plastic deformations distributions, and additional analyses are performed until safety and parity conditions are satisfied. In comparison with other methods, the proposed one uses fewer variables and constraints. Two examples (one of them worked in detail) are included.

A method for predicting the torsional strength of a reinforced concrete beam containing a rectangular opening is developed using the principles of limit analysis. A simple yield criterion for yielding under combined bending and torsion is employed. The applicability of the method is limited to openings in which the chords are subjected to predominant bending and torsion, while the shear force at collapse is less than 2/3 of the ultimate shear force. Such openings are characteristically long, and a suitable length parameter to use is the ratio of ultimate bending moment to ultimate shear force of chords. The proposed method is easier to apply than other available methods because it requires only some hand calculations beyond the computation of ultimate values of bending moment, torsion, and shear force of the chords using standard methods. Comparison with available test results shows that the proposed method provides consistently good predictions.

An accurate line model based on the finite‐element method is developed for analyzing the inelastic lateral buckling of I‐section beams and beam‐columns. The prebuckling in‐plane bending is analyzed using a geometrically nonlinear finite‐element method that accounts for the effects of prebuckling displacements and residual stresses on yielding. The results of the prebuckling analysis allow the distributions of yielding and strain‐hardening throughout the beam to be determined. The out‐of‐plane flexural‐torsional buckling of the inelastic beam is analyzed by adapting an elastic tapered monosymmetric finite element. For this element, the deflections and twists are referred to an arbitrary straightline axis along the midheight of the web, instead of the shear center or centroidal axes. The elastic element is adapted for inelastic buckling by using reduced out‐of‐plane stiffnesses that allow for yielding and strainhardening. The method is used to give some indications of the accuracy of earlier studies based on less accurate assumptions.

Experimental studies were conducted on the local bond stress‐slip relationships of deformed bars embedded in confined concrete under monotonic and cyclic loading in the temperature range between −10 °C and −70 °C (+14 °F and −94 °F). The bond behavior between steel reinforcement and concrete under low temperature appears quite different from the bond behavior under normal temperature $[+20°C(+68°F\left)\right]\text{.}$ The main differences observed are in the values of maximum bond stress and in the relationship between maximum bond stress and concrete compressive strength. Under monotonic loading, the ratio of maximum bond strength at low temperature to that at normal temperature is almost identical to the ratio of the respective concrete compressive strength values.

The concept of an active structure is presented in this paper. An active structure is defined as one consisting of two types of load‐resisting members: the static or passive members, and the dynamic or active members. Their integration in an optimal fashion produces a structure that is adaptive to changing environmental loads and changing usages. That the active structure concept is sound is first demonstrated through examples. The mathematics of the basic problem, which can be viewed as an application of the general theory of optimal control of parametric systems, are then carefully investigated, and a solution methodology is outlined.