Under the American design standard, ACI 318, the strength of slab-column connections is assessed using an interaction equation that includes contributions from both shear and unbalanced moment. Based on a reexamination of tests reported by Hanson and Hanson, the unbalanced moment at interior connections is shown to contribute far less to transverse shear than is assumed in design. A simple limit analysis is presented. This analysis is more consistent with observed behavior and accepted material limits and better predicts the test results.

Tests have established that punching shear in slabs can be effectively resisted by reinforcement consisting of vertical members mechanically anchored at the top and bottom of slabs. ACI 318 sets out the principles of design for slab shear reinforcement and makes specific reference to stirrups and shear heads. This report reviews other available devices and makes recommendations for their design. The application of these recommendations is illustrated with a numerical example.

A test program designed to evaluate the straight-line strut and tie model for predicting punching shear capacity is described. Eight isolated column-flat plate connections were loaded to failure. The primary variables were clear cover, spacing of reinforcement through the column, and boundary restraint. Measurements of strain were made along reinforcement passing through and immediately adjacent to the column. These measurements do not support the concept of an inclined straight-line compression strut as assumed in the strut-and-tie model, but indicate that the compression strut is more in the form of an arch. The beneficial effects of increased cover in improving the shear capacity and connection ductility is demonstrated. Increasing the amount of slab reinforcement passing through the column can lead to anchorage failures. An anchorage failure of this reinforcement is not distinguishable from a punching failure on the basis of external appearances during or after a test and it is concluded that many punching failures reported in the literature were limited by anchorage capacity.

Six specimens simulating isolated slab-column connections were loaded to failure to determine the effect on the shear capacity of reinforced concrete slab-column connections from adding corrugated steel fibers to the concrete. The primary variables were the concrete cover and the density of fiber reinforcement. The addition of fibers increased the shear capacity from 20 to 30 percent, increased the initial stiffness, and greatly increased the ductility of the connection at failure. The means by which the fibers enhance the shear capacity are described. It is concluded that, in many instances, adding corrugated steel fibers offers a viable alternative to increasing the shear capacity of slabs.

This paper presents a new model that describes the behavior of concentrically loaded flat plate-column connections at failure. The model combines radial arching action with the concept of a critical shear stress on a critical section (beam action shear). For brittle punching failure, bond strength of the reinforcement is seen as the significant factor limiting beam action shear. Based on this model, a simple lower bound estimate of the ultimate strength of a plate-column connection is derived. The model is compared to those found in building codes and to 115 test results from the literature.

The accidents at Chernobyl and Three Mile Island focused world attention on the behavior of nuclear facilities during severe accidents and stimulated interest in the role of containments in mitigating the consequences of such accidents. It is well known that an intact containment is highly effective in limiting fission product releases to the atmosphere, but it is less appreciated that concrete containments can perform well even if their walls are severely cracked. Through-wall cracks will provide pressure relief that can reduce the risk of exceeding the containment's structural limits. They also act as filters and thus present a significant barrier to the release of fission products. It is probable that the postulated consequences of a severe nuclear accident will be greatly reduced when these effects are considered. To take credit for the benefits of cracking in accidents well beyond the design basis, it is necessary to be able to predict the frequency and characteristics of any through-wall cracking, the nature of the resulting relief flow, and the efficacy of the many mechanisms for fission product retention within these cracks. The research required for the formulation and verification of detailed predictive response models is examined in this paper.

Knowledge-based expert systems for proportioning and detailing reinforced concrete members must incorporate both the mathematical reasoning wherein each decision step follows directly from the previous calculated information and the intuitive reasoning based on acquired knowledge that is required to arrive at a plausible design. The general requirements for such systems in the domain of structural engineering are examined and some solution strategies and heuristics used in the design of reinforced concrete columns are presented. The article discusses the use of KBES in structural engineering, and its use in member design. COLUMN, a computer program that will proportion reinforced concrete sections that are comparable to those produced by an experienced structural engineer, is described.

The factors that influence the shear–moment interaction of slab–column connections are examined. The Canadian design code, CAN3-A23.3-M84, is evaluated on the basis of its agreement with the observed connection behavior. Other approaches to modelling slab–column connections are examined. It is concluded that, although the code procedure appears to give satisfactory results in practice, it is not based on a rational model of the connection behavior. Of all the alternative methods examined, the truss model provides the best description of the shear–moment interaction behavior of any given connection. Key words: column–slab connections, shear–moment interaction, punching shear, reinforced concrete, analytical models.

Large span elliptical culverts stiffened over the upper portion with a concrete cap can be used economically in highway construction at locations where only a small soil cover is possible. The behavior of such structures based on analyses obtained using the finite element program ADINA is presented. Primary variables are the thickness of the concrete cap and the depth of the soil. The effects of both the construction sequence and highway loading are considered.The load-carrying mechanism for the composite structure is dependent primarily on the amount of lateral deformation permitted by the adjacent soil and the thickness of the concrete cap. It is concluded that the behavior of the structure is that of a flat arch; but, with the lateral deformations likely to occur, the concrete section should be proportioned as a simple beam. Key words: concrete cap, culverts, construction loads, deformations, finite element analysis, stresses.

For buried structures, where the structural action is the result of an interaction between the surrounding soil mass and the structure, the analysis must consider the effects of the construction sequence. The usual method of adding elements, when using a finite element program that has the capability of adding and removing elements to simulate the progress of construction, may lead to either numerical instabilities or results that are unacceptable. Both problems are demonstrated in the analysis of a stiffened culvert using the program ADINA. A technique which combines application of a preload with element birth to overcome these problems is described and illustrated.

The construction and testing of a model of a prestressed concrete containment structure is described. The test structure consisted of a reinforced concrete base, cylindrical wall, ring beam and dome built of prestressed concrete with construction details patterned after the Canadian CANDU reactor containment. The overall height above the base was 12 ft‐6 in. (3,810 mm), and the outer diameter was 10 ft‐6 in. (3,200 mm). Internal pressure was obtained using water, and leakage was prevented by using a flexible plastic liner. Mea‐, surements made during the test included internal pressure, steel and concrete strains, crack widths and spacing, and curvatures at the base of the cylindrical wall. The test structure began to exhibit cracking at a pressure of. 30 psi (0.28 MPa), and yielding of the reinforcement at approximately 110 psi (0.76 MPa). The structure displayed considerable ductility before failing at internal pressure of 159 psi (1.10 MPa) by rupture of three horizontal tendons at midheight of the wall. Outward deflection of the walls damaged the anchorage zones of some of the tendons. The cracking behavior and failure mechanism are described.

Eight reinforced concrete specimens were fabricated and subjected to tensile membrane forces and air pressure to study the air leakage characteristics in cracked reinforced concrete members. A mathematical expression for the rate of pressurized air flowing through an idealized crack is presented. The mathematical expression is refined by using the experimental data to describe the air flow rate through any given crack pattern. Graphical charts are also presented for the calculation of the air leakage rate through concrete cracks. The concept of equivalent crack width for a given crack pattern is introduced. The mathematical expression and graphical charts are modified to include this equivalent crack width concept. The proposed technique is applicable for the prediction of the leakage from concrete containment structures or any similar structures due to high internal pressure sufficient to initiate cracking.

Rules for determining the spacing and widths of both through‐thewall and surface cracks in post‐tensioned concrete containment structures under internal pressure are presented. These involve the construction details and the average strain obtained from an analysis that accounts for concrete properties in the post‐cracking range. To evaluate these rules and the concrete constitutive relationship used in the analysis, twelve quarter‐scale segments were designed to simulate construction and stress conditions at various locations of a containment structure. Major variables included ratio of prestressing, concrete cover, reinforcement spacing, lap splices and combined axial load and moment. It was observed that the crack spacing depended on the spacing of the reinforcement and prestressing tendons parallel to the cracks and that the final crack pattern was fully developed at the yield strain of the reinforcing steel. Concrete cover was not found to have a significant influence on crack spacing. The procedures developed to determine crack spacing and widths are used to predict the cracking of a one‐fourteenth scale model of a containment structure tested later in the research project.