International Concrete Abstracts Portal

In modern computerized structural analysis, realistic material laws should be used. This research will present a constitutive law and a numerical procedure based on the finite element method for the analysis of a prestressed concrete structure including the time-dependent effects due to the load history, creep, shrinkage, aging of concrete, and relaxation of prestress. A 32.1 meter (105 ft) long U-shaped railway bridge, composed of two precast post-tensioned concrete girders and an in situ cast prestressed young concrete slab, was instrumented to observe its long-term structural behavior and used for the comparisons with numerical analysis. To evaluate and predict the structural behavior of this concrete structure, the related experiments were designed and performed both in the field and laboratory. Some material properties needed for the analysis were obtained through the extensive program carried out in the laboratory with controlled environments. This paper will describe the details of structure, test program, and experimental results.

Tests results from long-term experiments on prestressed and partially prestressed concrete beams are reported. Tests were carried out on 10 rectangular beams spanning 2 m and undergoing sustained loading for five years. After that time, there was no evidence of stabilization of the time-dependent behavior of concrete. Numerical modeling of the deformation of the midspan section explains experimental observations and confirms that the presence of ordinary reinforcing steel in a prestressed concrete section leads to a redistribution of stresses between concrete and steel which should be taken into account in serviceability limit-state computations. Tests to failure of the beams at 5 years yield no significant differences in carrying capacity with tests executed at an early age. It is suggested that the deflection limit state is a major consideration in design and that the degree of prestressing should be chosen in function of ratio of permanent load to total design load (permanent and live).

Continuing the study of long-term behavior of a U-type composite prestressed concrete bridge, this paper presents a constitutive law for structural analysis and a nonlinear diffusion theory for the understanding of internal humidity distribution in a structure. The constitutive law was formulated based on a new rheological concrete element, which is a series coupling of a generalized Kalvin chain unit, a cracking unit, and a unit representing thermal strain or shrinkage. The previously obtained exponential algorithm for cracking, treated as strain softening, is combined with the exponential algorithm for generalized Kelvin chain element. The nonlinear diffusion theory, which considers both the pore humidity and aging effects on the diffusivity, was found to give very good fit to the distribution of internal humidity of concrete. The material parameters needed for the constitutive law and diffusion theory were identified. The numerical results using finite element method show the history of the prestress transfer between a young bridge deck and more matured girders. The calculated strains based on the presented method were found to be in good agreement with field measured data.

The first stage of an experimental investigation into the instantaneous and time-dependent behavior of reinforced concrete columns under sustained load is reported. The experimental work described in the paper involves the testing of 15 large-scale columns in compression and uniaxial bending to obtain comprehensive creep deformation data for the prediction of long-term lateral deflection and instability. Information on the range of slenderness ratios and load levels which cause creep instability for rectangular symmetrically reinforced concrete columns is of particular interest. The experimental data is also for use in the development and calibration of a theoretical model for the prediction of creep deflection and buckling under sustained load. The experimental setup used in this investigation and described in the paper is designed for the simultaneous testing of five slender columns, each up to 6 m in length. The loading frame is such that each column may have different length, different cross-sectional dimensions and reinforcement details, and be subjected to different combinations of axial force and bending moment. In the three series of tests presented here, column length, axial force, and initial eccentricity are the major variables. The loading on each column is monitored independently and maintained automatically at a constant preset value throughout each test. A direct comparison of creep effects on column behavior is therefore possible as different loading parameters are varied.

The paper presents a summary of treatment of deflection control in the Australian Standard, AS3600-1988. The top-tier method recommended by the standard requires consideration of creep and shrinkage effects on deflections. Simple expressions for calculating creep and shrinkage deflections are given and illustrated by examples.