Journal of Structural Engineering

Timber‐concrete floors are widely used in the Persian Gulf region because of their resistance to the hot and aggressive environment of the area. Because no shear connector is provided, the timber joists and concrete slab work independently. In this study, it is suggested that relatively inexpensive high‐strength nails be used as shear connectors so that the timber joists and concrete slab resist the loads as a composite section. Experimental investigations have shown that it is possible to attain full composite action through the use of these high‐strength nails. The proposed composite timber‐concrete floor system was subjected to static short‐term loading, repeated loading, and long‐term sustained loading. It is observed that in terms of strength and serviceability the behavior of the proposed system is within the limits set by standard building code. This behavior is also verified by a finite‐element analysis. The experimental and numerical results clearly reveal that it is possible to achieve a stiffer floor system, longer spans, and a significant reduction in the number of timber joists, while attaining the same receptivity to a hot and aggressive environment.

This paper reports on the results of experimental studies on the high‐temperature mechanical properties of high‐strength fire‐resistant (FR) bolts for fire‐resistant steel (FR steel), and the shear strength of the friction joints using FR bolts at high temperatures. Since the FR bolts have been adopted for joints made of FR steels to be used at temperatures as high as 600°C, this study placed emphasis on the shear strength after slip occurs in addition to the conventional slip load. The results led to the following conclusions. The tensile strength of FR bolts at high temperatures is superior to that of conventional high‐strength bolts. The slip load of the joints can be roughly estimated on the basis of the results of relaxation tests for bolt materials. The slip load of the joints using FR bolts decreases rapidly in temperatures over 400°C, and the shear strength is still greater than allowable shear force at temperatures up to 650°C.

Iterative one‐step, iterative step‐by‐step, and noniterative step‐by‐step approaches for calculating prestress losses due to shrinkage, creep, and steel relaxation for partially prestressed, continuous concrete/steel composite beams are presented. A simplified method that accounts for the aging effect on losses is also formulated. The method is based on exploiting the noniterative step‐by‐step approach to accumulate the losses through an infinite number of time intervals. It provides a simple yet accurate explicit formula for determining shrinkage and creep losses. This simplified method can be used for estimating the prestress losses of any prestressed concrete member. An example is included to show the application of the proposed simplified method for partially prestressed continuous composite beams and to demonstrate its accuracy for design purposes. A parametric study is conducted on the given example, taking into account geometrical and material properties, to illustrate the validity of the proposed method relative to the iterative one‐step and the iterative step‐by‐step methods. It is concluded that the proposed method is in close agreement with the more rigorous iterative procedures.

Tetrafluoroethylene (TFE) is commonly used as a slide surface mating with polished stainless steel in bridge bearings that accommodate translation and rotation of the superstructure relative to the substructure. The primary requirement of the slide surface is a low coefficient of friction that is maintained throughout the service life of the bearing. The paper reports on a laboratory investigation comprising 14 tests undertaken at room temperature under simulated field conditions that was performed to determine the effect of contamination on the coefficient of friction in a lubricated TFE/stainless steel interface. It is shown that the friction of a lubricated surface increases significantly with a relatively low amount of contamination in the form of unhydrated portland cement in the interface, emphasizing the need for assembling bearings in a dust‐free environment and avoiding disassembling of bearings on site. The influences of the degree of contamination on the coefficient of friction at initial movement and after 200 cycles of movement, as well as on the peak value of the coefficient of friction recorded during a test, are discussed.

In this paper two equivalent systems are presented to replace in the analysis nonprismatic members with or without standard end conditions. While for a symmetrical nonprismatic member only a single prismatic beam with shear effect included is adequate, in the case of asymmetrical members, two or three equivalent prismatic beams are required. The characteristics of these beams are determined in terms of the real member characteristics by equating the strain energy stored in the real member and in the equivalent system. Improved versions of the stiffness matrix of the real beams are considered into the strain‐energy expression. The equivalent system characteristics are derived for both the general case and some particular cases. Two numerical examples are given to illustrate the procedure: One demonstrates that in the deck‐grillage analogy applied to the analysis of bridge decks the equivalent system presented in the paper can lead to very good results; the second shows that in the analysis of skeletal structures, the results obtained with the equivalent systems are practically the same as the “exact” solutions.

This paper discusses the issues involved in using simplified second‐order inelastic analysis for design of steel‐frame structures. In particular the elastic‐plastic hinge method is examined and the assumptions inherent in its use are outlined. It is shown that this method does not always lead to an accurate representation of inelastic behavior, and that it does not furnish enough information for assessment of the ultimate strength of individual members. An alternative approach called the refined plastic‐hinge method is introduced, and its design implications are discussed. The refined plastic‐hinge approach accounts for inelastic stiffness degradation, and a column tangent‐modulus expression is utilized to represent the effective stiffness of the element when it is loaded with large axial loads. This paper highlights and illustrates the qualities and limitations of these two inelastic analysis methods with regard to their acceptability for use in predicting the ultimate strengths of steel frames and their members. Case studies and recommendations for use of inelastic analysis for direct frame design are presented in a companion paper.

This paper discusses several issues pertinent to the use of advanced analysis for design of two‐dimensional steel‐frame structures. In particular, the accuracy of the elastic‐plastic‐hinge method is examined, and its range of application and limitations are identified. An approach called the refined plastic‐hinge method, the formulation of which has been presented in a companion paper, is recommended and its design implications are discussed. Case studies are presented to demonstrate the validity of the refined plastic‐hinge method in predicting the strength and stability of individual member and overall system strength. Also, the performance of this analysis model is evaluated by comparing the results with those from elastic‐plastic‐hinge and refined plastic‐zone analyses. The goal is to illustrate the qualities and limitations of the plastic‐hinge‐based analysis methods with regard to their acceptability for use in predicting the limiting strength and behavior of steel members and frames.

This paper is based on the results of a comprehensive field investigation of the behavior of a double‐cell cast‐in‐place box culvert. The investigation was conducted by the University of Nebraska and was completed in November of 1990. The culvert was instrumented for soil pressure, moment, and deflection measurements. Moment measurements were accomplished indirectly by measuring strains in outer and inner steel rebars embedded in the concrete walls. Vibrating‐wire strain gages were mounted on the rebars in the laboratory prior to placement in the forms. All of the instruments were read periodically during the construction and backfilling processes. Live‐load measurements were taken at 2 ft fill increments over the top slab, starting with the exposed top slab. This paper describes the full‐scale testing of the functional box culvert and provides a detailed discussion of its measured structural response to soil and truck loading.

A technique of calculating inelastic deformation of low‐rise shear walls having height‐width ratios of 0.5 and 0.75 without boundary elements is presented with consideration of the coupling effect for bending and shear deformations as well as the deformation due to base rotation. An interaction surface of moment, shear, and curvature or moment, shear, and shear strain is developed. The deflections at crack, yield, and ultimate loadings can be calculated separately from bending and shear deformations, which are compared favorably with experimental results. The shear deformation is significant for the low‐rise walls studied because the deformation due to bending deformation is about 40–60% of the total deformation after the walls have reached 20% of ultimate deformation. The hysteresis rules are developed for both bending and shear deformations on the basis of theoretical and experimental studies. Favorable comparisons between the calculated and experimental responses were observed for individual walls and a low‐rise two‐story building on a shaking‐table test. A computer program was developed for structural system analysis subjected to seismic excitations.

Classical reliability has significantly enhanced the ability of engineers to assess the safety of constructed projects. It has been shown that Bayes' theorem is an effective tool in updating prior probabilities when the value of a random parameter is known. However, the preceding theorem usually fails to adequately address the uncertainties of the subjective parameters (such as “the connections are good”) that are associated with structural evaluation. It has been demonstrated that these parameters can be significant to the overall safety assessment. With the introduction of fuzzy‐set theory, it is possible to quantify the qualitative evaluation and incorporate it into the safety assessment. This paper presents an algorithm to compute the posterior probability based on visual inspection of structural components by incorporating fuzzy‐set theory into Bayes' theorem. The results based on two examples—a reinforced concrete beam and a structural frame—showed that the fuzzy‐Bayesian approach is a viable enhancement to the safety assessment of existing structures.

The cyclic response of steel coupling beams embedded into reinforced concrete boundary elements was studied. Three half‐scale subassemblies representing a portion of a prototype structure were designed, constructed, and tested. The main test variables were the amount of vertical tension/compression stresses in the boundary element, and presence of auxiliary bars attached to the beam flanges for a smoother transfer of bearing stresses. The steel coupling beams exhibited very stable hysteresis characteristics, and could develop the theoretical plastic moment when the boundary and, hence, the connection region were subjected to normal compressive stresses. Tensile stresses in the boundary element reduced the stiffness, and smaller moments could be developed. A significant amount of dissipated energy could be accounted by that energy dissipated in “plastic hinges” formed in the exposed portion of the coupling beam. The stiffness was found to be different depending on whether the boundary element was under compressive or tensile normal stresses. The initial stiffness of the coupling beams in all the specimens was smaller than the value computed assuming full fixity at the face of the wall. The “effective fixed point” of the coupling beam was found to be inside the wall at approximately one‐third of the embedment length from the face of the wall. Analysis and design of composite coupled walls should account for the actual location at which the coupling beams are fixed in the walls.

An analytical approximate constitutive relation is derived for a form of fluid damper, which exhibits viscoelastic behavior. The damper is used for vibration isolation of piping systems and industrial equipment, as well as for vibration and seismic isolation of building structures. The damper consists of an open pot, filled with highly viscous fluid, and a piston that moves within the fluid. The analytical solution relates the damping constant of the unit at vanishingly small frequencies with the material constants of the fluid used in the damper and the geometric characteristics of the damper. With the determination of the damping constant, a macroscopic model may be constructed that describes the damper behavior over a large frequency range. The other parameters of the macroscopic model are identical to those in the corresponding constitutive relation of the damper fluid. The results of the analytical solution are in very good agreement with experimental data and with numerical solutions based on boundary‐element analysis.

Ponding loads coupled with creep effects may contribute significantly to the nonperformance (ultimate failure) of wood roof members. Proposed load and resistance factor design (LRFD) specification provisions for the flexural design of wood beams in roof systems that may be subject to ponding loads include: (1) A check to determine if the member is subject to ponding loads as designed; and (2) a modified design equation to include a ponding moment amplification factor. These equations, however, include largely empirical treatments of creep effects using creep factors taken directly from the previous National Design Specification for Wood Construction. These creep factors were originally developed for use in deflection‐checking equations. This study investigates the effects of creep on the safety of wood roof beams subject to ponding loads. The results of the analysis are used to evaluate a creep factor for design on a more rational basis that includes time effects (i.e., the coupling of creep and moment magnification). Finally, the validity of proposed LRFD specifications is examined and suggestions are made for their modification.

Flexural design for concrete members reinforced with fiber‐rein‐forced‐plastic (FRP) composites may be conducted using both the ultimate‐strength method and the working‐stress method. Since FRP reinforcement does not yield, there should be the explicit provision that failure be controlled by concrete crushing as opposed to reinforcement rupture. Deflection control may become as important as flexural strength for the design of FRP‐reinforced concrete structures. The paper contends that, at this stage of development, the working‐stress method is better suited to FRP‐reinforced concrete. The primary reasons are that the predicted ultimate moment capacity represents a highly variable state only attainable at a high level of deformation and crack opening and that it depends strictly on concrete ultimate strain. FRP reinforcement is better suited to pre‐ and posttensioned‐type concrete members; however, it cannot be ruled out that FRP should become the reinforcement of choice in special nonprestressed applications where durability or magnetic permeability are the controlling parameters.

The ultimate strength of the commonly used anchorage connection for x‐bracing rods and cables is not documented in the literature from either analytical or experimental viewpoint. A comprehensive experimental laboratory testing program was executed to investigate the major parameters influencing the ultimate load‐carrying capacity of the anchorage connection detail. The tests covered four rod sizes and corresponding hillside washers, and three web thicknesses. Two methods of stiffening the web plate were used: a patch plate around the anchorage slotted hole and transverse web stiffener plates. Five failure modes were isolated and identified. A yield line, limit state design approach was used for formulating the analytical expressions for predicting the load‐carrying capacity and the failure mode. The derived expressions were found to predict very well the experimental test results. Repetitive tensile loadings beyond yielding of the anchorage connection were also used to test the anchorage connection later on in the experimental testing program. It was concluded that repetitive cyclic tension loading up to 20 cycles causes no serious deterioration to the ultimate strength of the x‐bracing connection.

The vibrations of a cable‐stayed pedestrian bridge are measured and compared with a numerical model, and used to produce improved estimates of the structural properties. Fourteen deck modes with frequencies under 16 Hz were identified using the impact hammer technique. The damping of all but one of the modes was measured at 0.75%, in spite of a wood deck. The numerical model consisted of a three‐dimensional finite element analysis which incorporated nonlinear cable effects. The experimental and numerical results were shown to be highly correlated by visual inspection, the MAC, and the CoMAC. The primary differences were: (1) Two of the experimental mode shapes showed a slight asymmetry that was not present in the numerical modes; and (2) the measured natural frequencies were consistently higher than the numerical frequencies. Parameter estimation using eigenvalue sensitivities with changes of less than 12% in the nominal values helps reconcile the experimental and numerical differences.

Moment‐magnification factors for columns in braced frames are dependent upon column slenderness and the flexural stiffness of restraining beams. A column bent in single curvature has a flexural stiffness that can be as little as one‐third that of the same column bent in reverse curvature. Column slenderness ratios $(L/R)$ above which reverse curvature moments must be magnified are significantly higher than those for which single‐curvature moments would require magnification. This paper shows that when loading conditions are considered together with framing restraints, magnified moments from the single‐curvature condition rarely generate moment values as high as moments unmagnified from the reverse‐curvature loading condition. Thus, without need for moment magnification, column slenderness ratios can be significantly higher than the limit ratios traditionally derived for single‐curvature bending conditions of columns.

The buckling of elliptical plates under uniform compression has so far been studied for clamped edges. Prompted by the lack of studies on such plate shape, this paper attempts to provide further new buckling results for elliptical plates without and with internal line/curved supports. The recently developed pb‐2 Rayleigh‐Ritz method was used for the analysis. The method views the entire elliptical plate as a single supercontinuum element, even with the presence of complicated internal curved supports, by incorporating the support equations into the Ritz function. As such, there are no discretization or boundary and internal support losses, as are usually associated with most discretization methods. Buckling loads are graphically presented for simply supported and clamped plates. Simple buckling formulas for simply supported and clamped plates are also proposed.

The stress‐strain diagram of a short concrete specimen under compressive uniaxial load cycles is enveloped by a curve that matches the stress‐strain curve obtained under constantly increasing strain. At least eight published studies have shown that this envelope curve is unique with respect to a variety of excitation histories in the range of strain rates generally met in seismic conditions. In the present work we have attempted to extend the envelope curve concept to columns in plain or reinforced concrete, with reference to compressive stresses again. The tests performed have allowed us to ascertain that this curve is unique and coincides with the monotonic curve.