Journal of Structural Engineering

This specification and supporting commentary cover the design of simply supported composite joists and trusses used in one-way floor systems. Basic design procedures are based on achieving nominal flexural strength before shear failure or local instability occur. The specification considers strength, serviceability, and construction criteria. A design example is included.

A two-bay six-story moment-resisting steel structure was designed according to ATC 3-06 as an example, and its behavioral characteristics in inelastic range were observed through the inelastic analyses. The design concept of strong column and weak girder, which is usually implemented by the design rule that the sum of column plastic moments should be larger than that of girder plastic moments at the joints by some margin, cannot actually prevent the occurrence of plastic hinges in columns. The reason is clarified by the close investigation of inelastic behaviors around joints up to the point of the formation of a collapse mechanism. Based on the results of the study, a new design rule to ensure the prevention of plastic hinges in columns is suggested.

Riveted steel connections were used throughout most of the United States for many years. The connections were commonly encased in massive but lightly reinforced concrete for fire protection. These older structures were designed with little or no consideration of seismic behavior, and today these buildings sometimes require seismic rehabilitation and retrofit, but little information regarding this behavior is available. The present paper provides information regarding the seismic performance of these older steel structures. An experimental study which evaluates the seismic behavior of subassemblages with T-stub and clip angle connections is described in detail. The results are analyzed. The strength, stiffness and ductility of the connections are shown to be related to the properties of the steel and the concrete, the mode of failure of the connection, and the concrete encasement. The hysteretic behavior, energy dissipation and deformation capacity of the connections are shown. Approximate models for predicting the strength, stiffness, and deformation limits are presented as practical implications of the research.

The flexural requirements for plastic design of I-shaped steel sections in international codes of practice are inconsistent and provide misleading emphasis on local buckling as the primary strain-weakening effect. The loss of ductility due to simulated axial force occurring simultaneously with inelastic bending is also underestimated in these codes. A model is described to assess the maximum moment and available inelastic rotation prior to strain weakening due to the interactive local flange and web buckling and lateral-torsional buckling. The results of 44 tests on I-shaped steel beams under pure flexure and 14 tests on similar specimens under combined bending and simulated axial force are also described. The rotation capacities measured in these tests are examined in terms of standard local and lateral buckling parameters as well as the proposed model. The relative importance of the different parameters is assessed and the effectiveness of the theoretical model is demonstrated in the analysis of these experimental results.

This paper presents detailed results on the shear behavior of 2.44 m by 2.44 m light-gauge steel stud walls for three different shear resisting systems: framed walls with 20 gauge flat strap X-bracing on the face—type A; framed walls with 12.5 mm (1/2 in.) single-ply gypsum wallboard on the back and 12.5 mm (1/2 in.) single-ply gypsum sheathing board on the face—type B; and framed walls with 12.5 mm (1/2 in.) single-ply gypsum wallboard on the back, 12.5 mm (1/2 in.) gypsum sheathing board on the face, and 20 gauge flat strap X-bracing on face—type C. The steel framing used in these tests is typical of framing used in residential construction. The behavior of the type A walls was governed by the yield strength of the straps with practically no resistance provided by flexure in the studs. In the type B and type C tests, the measured maximum load was controlled by the breaking of the wallboard along its edges. The failure mechanism was initiated by a rotation of the screws at the edges. This was followed by the partial pull-through of the screws at the edges of the gypsum board and simultaneous breaking of the board at the edges. Each 1.22 m by 2.44 m (4 ft by 8 ft) gypsum panel was observed to behave independently during loading. A comparison of the results from the type B and type C tests showed that the use of 50.8 mm, 20 gauge tension bracing prevented cracking of the boards at the perimeter, reduced the lateral displacement, and increased the maximum load capacity of the wall by as much as 28%.

Beams with corrugated webs were tested to failure under shear; the failure was due to buckling of the web. The test specimens were modeled using finite elements and the computer program ABAQUS was used to perform nonlinear analysis; the load increment as well as the mesh size were selected based on studies to examine their effects. The comparison between the results from the finite-element anaylsis and the tests was satisfactory. The average ratio between the analytical and experimental buckling loads was calculated to be 1.15. The primary reason why the analytical results are higher is the presence of unavoidable imperfections in the webs of the test specimens. Finite-element models with seeded imperfections gave results very close to those obtained from the tests. It was noted from the experimental as well as the analytical results that buckling of the web is local and global for the coarse and dense corrugations, respectively. Buckling formulas are suggested, which are based on local buckling of the corrugation folds as isotropic flat plates or global buckling of the entire web panel as an orthotropic plate. Comparisons between the results from the formulas, the finite-element analysis, and the tests are given and shown to be satisfactory.

This paper reports high-temperature properties of stainless steel for structural use in comparison with conventional carbon steels. Because of the elements contained, stainless steel retains strength at high temperatures. Through high-temperature tensile tests, Young's modulus tests and creep tests, the constitutive equations for a fire-safe design relating to the stress-strain and creep strain-time relationships at high temperatures were formulated. It was made clear that stainless steels have relatively large work-hardening of stress-stain curves at high temperatures after the strain surpasses the elastic limit and thereby the fire resistance of stainless steel frames is considered to be excellent. Also, high-temperature thermal constants such as the thermal expansion rate and the heat transfer rate were measured, and it was made clear that these properties should be taken into account upon fire-safe design. Through these studies, it became possible to simulate thermal and structural behavior of stainless-steel frames during fire.

Static analysis of space truss structures including inelastic material and large deformation geometric nonlinearities is considered. Member material behavior modes include: buckling, yielding, inelastic postbuckling, unloading, and reloading. Large deformation behavior is based on the nonlinear updated Lagrangian space truss displacement equations. Due to the interaction of the nonlinear member behavior modes, force redistributions in a space truss structure can lead to progressive structural failure. The developed model traces the sequence of member buckling, yielding, and inelastic member response. Results are presented to demonstrate model capabilities for representing the inelastic, progressive failure response of space truss structures.

Tie plates are integral components of pin connectors, lifting modules, and bridge restrainers. As a bridge restrainer, they are supposed to bear the dynamic seismic loads. In a previous work (Obata et al. 1991), the writers elucidated the ultimate behaviors of tie plates under quasi-static loading by the experiments and the numerical analysis. The objective of this work is to investigate the ultimate behaviors of tie plates at high-speed tension both experimentally and numerically. The high-speed tension experiments showed that the maximum strength increases with the increase of the tensile velocity. On the other hand, the deformation at the load maximum point is relatively insensitive to the tensile velocity. Therefore, the quasi-static test results always give the conservative values for the design. The numerical analysis is based on the J₂ power law strain-rate-dependent plasticity theory. Despite the simplification of the complex phenomena, the numerical results give useful insights into the experimental results and in the meantime serve as reliable tools for a rational design of these tie plates.

This paper proposes a simplified time-dependent analysis for continuous steel-concrete composite beams with flexible shear connection, subjected to static actions, settlements of the supports, shrinkage, and prestressing of the concrete slab. By approximating the integral terms with particular quadrature rules (algebraic methods), the integral-differential system, which governs the problem, reduces to a differential system. This provides approximate solutions of the viscoelastic problem directly at the final time, without having to consider the real stress history. Some of the algebraic methods known in the literature (e.g., effective modulus, mean stress, and age adjusted effective modulus methods) are applied to continuous composite beams subjected to the four different actions examined separately. Subsequently, the numerical results obtained are compared with the results of the accurate numerical analysis (general method), and the validity of each simplified method is discussed. Finally, some information on the practical application of the effective modulus and the mean stress methods by means of routine computer programs, are given.

This paper presents a theoretical model that has been developed to represent and include architectural walls in structural analyses. The model has been incorporated into a nonlinear, dynamic, finite-element analysis program. While analytical models have been developed previously to represent architectural components in buildings, few models are available to include them in structural analyses. The model is intended to be used by researchers and practitioners to study the effects of architectural walls on building response. The level of complexity and the level of accuracy of the model are balanced so that the model may be incorporated into large finite element models without an undue increase in computational effort, yet is representative of actual wall characteristics. Comparisons between theoretical and experimental results indicate that the model effectively represents the behavior of interior partition walls; the ratio of energy dissipated by the analytical model versus experimental results ranges from 0.9 to 1.1. The analytical response of walls in a two-story frame subjected to input ground motions is included to illustrate the incorporation of architectural walls in structural analyses.

The influence of floor flexibility on the seismic response of building structures is discussed through comparison of the computed seismic response for structures with flexible diaphragms and counterpart structures with rigid diaphragms. Case studies of three existing buildings with flexible diaphragms and analogous systems with rigid diaphragms are presented to illustrate these differences. Each building was subjected to the 1989 Loma Prieta Earthquake. The structures were: (1) A two-story firehouse in Gilroy with unreinforced masonry walls; (2) a two-story timber office building in Palo Alto with grouted and reinforced clay-unit masonry shear walls; and (3) an eight-story hotel in Oakland with unreinforced clay-unit masonry and reinforced-concrete shear walls. The analytical studies show that, in some cases, diaphragm and shear-wall accelerations can increase with the flexibility of the diaphragm. Torsional forces can reduce considerably as diaphragm flexibility increases. Further, approximate expressions prescribed in current seismic codes can underestimate the period of vibration of systems with flexible diaphragms.

Preliminary results from a project studying the behavior of a hidden timber tension connections are presented. The connection consists of a threaded steel rod inserted in a longitudinal hole along the central axis of a timber member, with a bearing plate and nut fitted on to the end of the rod in a specially cut mortise. This assembly was tested under pure axial load statically applied to the steel rod and the timber member. Observed failure modes are described and photographic documentation of failed connections is presented. Failure mode distribution is shown as a function of two physical parameters, the size of the bearing plate and the distance from the bearing plate to the end of the member, and a comparison is made with theoretical failure-mode domains derived from National Design Specification allowable stresses. The correlation of variations in the physical parameters with ultimate strength and ductility is also discussed. Initial results show the connection to be predictable, ductile, and potentially economical.

The classic Janssen theory takes no consideration of the interaction between the stored material pressure and the wall deformation, and thus may give poor prediction of pressures in flexible bins. Based on the following two assumptions: (1) The vertical pressure in the stored material is uniformly distributed at the same depth; and (2) the stresses in bin walls can be modelled by the membrane theory, a modified Janssen theory has been developed, which can not only take account of the wall flexibility, but also preserve the simplicity of the original theory. It is shown that the wall flexibility can be represented by a nondimensional parameter α=E _s R/E _w t. In addition, the lateral pressure ratio, κ, may be computed based on Poisson's ratio of stored material, ν_s. It is also illustrated that the modified Janssen formula could give a very close estimation with the finite-element method for different levels of α. Therefore, the present theory may at least provide a reference for the more rational determination of some critical parameters which affect the wall pressure in future designs of flexible wall bins.