International Concrete Abstracts Portal

With the aid of modern digital computers and sophisticated computational techniques such as the finite element method, it is now possible to simulate the structural behavior of an arbitrary reinforced concrete shell structure under general loading through its elastic, cracking, inelastic, and ultimate load ranges, taking into account nonlinear material, nonlinear geometry and time-dependent effects of creep and shrinkage. In this paper, a method of analysis and a computer program based upon a composite layered finite element displacement model are briefly described. The analysis recognizes the nonlinearities due to cracking, nonlinear stress-strain behavior in concrete, yielding of the steel reinforcement and the tension stiffening between cracks. The effects of the countinuously changing structural geometry are taken into account by an updated Lagrangian formulation. The time dependent effects of creep and shrinkage are also included by an initial strain procedure. Numerical results for reinforced concrete shells obtained with the computer program are presented which indicate that in some cases an increase and in other cases a large reduction in the calculated ultimate load occurs as each of the nonlinear factors is included in the computer analysis.

A finite element procedure is presented for the analysis of reinforced concrete shearwalls. The wall is idealized as a two-dimensional structure, and the global behavior of the wall under static loading conditions is emphasized. A combination of a new family of higher-order quadrilateral elements and beam elements is employed in the finite element discretization of the wall. Constitutive models of material behavior are based on the nonlinear elasticity. The main material nonlinear effects accounted for in the analysis are the tensile cracking, the biaxial compressive response of concrete, and the yielding of steel reinforcement. A smeared approach is used in the representation of concrete cracking and steel bars. Simplified uniaxial and biaxial material models for reinforced concrete are developed and presented in detail. The incremental-iterative nonlinear solution techniques employ both constant and variable stiffness with the option of selective updating of the stiffness matrix in the load increment. Numerical examples are presented and compared with other existing solutions.

A large amount of specialized factual and heuristic knowledge on the relations between the design of concrete mixtures, including the constituents, and the durability of concrete has been gained through research and field experience. Effective dissemination of this knowledge should result in fewer incidents of premature deterioration of concrete. Expert systems appear to be an effective means for transferring the knowledge on the durability of concrete obtained through laboratory and field studies and experiences to engineers and designers responsible for the design, construction, and maintenance of concrete structures. Durcon is a prototype expert system being developed to give recommendations on the selection of constituents for durable concrete. The purpose of developing Durcon is to demonstrate the application of expert systems to improve the process of selecting construction materials. Four major deterioration problems are covered by Durcon: freezing and thawing, corrosion of reinforcing steel, sulfate attack, and alkali-aggregate reactions. This report discusses the approach being followed and the progress being made in developing Durcon. In addition, model systems for recommendations for concrete exposed to corrosive environments and for preventing alkali-aggregate reactions are presented.

Because of the presence of lateral loads and high-end eccentricities, the ACI 318-83 empirical design method cannot be used for design of tilt-up walls. Analysis must be performed during design to account for the P-{delta} effects. To confirm various design concepts and to evaluate the slenderness limitations, a series of tests on concrete wall panels was conducted. Several simplified design procedures were used to compare analytical results to test observations of slender load-bearing walls. Results of computer program TILT for IBM-PC (or compatible) computers were compared with the results of simplified design procedure calculations and test observations. Investigations show that the actual strength of load-bearing tilt-up walls are considerably higher than predicted by simplified analysis. The paper briefly discusses the theory of geometrical and material nonlinearities and presents methods for solutions that are incorporated in the program TILT. Conclusions and effectiveness of the TILT computer program for analyses of tilt-up load-bearing walls are shown.

A computational procedure is described for determining nonlinear response of a building system subjected to earthquake motion. The method is sufficiently simple for use with a microcomputer because system response is expressed in terms of a single generalized coordinate. Deflected shapes for the systems are assumed to be invariant for all amplitudes of motion. The equation of motion is integrated for each instant of response on the basis of a normalized relation between base shear and top-level deflection. The hysteretic relation is constructed for each new cycle using cubic segments to express a path from initial unloading through force and deflection reversals to the point of maximum deflection. The base motion is selected from a menu of earthquake records stored on diskette. Results displayed on screen consist of histories of acceleration at the top-level and maximum interstory drift, and the computed force-deflection relation.