International Concrete Abstracts Portal

The present paper presents in chapter 1 a model for the characterization of concrete creep and shrinkage in design of concrete structures (Model B3), which is simpler, agrees better with the experimental data and is better theo-retically justified than the previous models. The model complies with the gen-eral guidelines recently formulated by RILEM TC- 107. Justifications of various aspects of the model and diverse refinements are given in Chapter 2, and many simple explanations are appended in the commentary at the end of Chap-ter 1 (these parts are not to be read by those who merely apply the model). The prediction model B3 is calibrated by a computerized databank comprising practically all the relevant test data obtained in various laboratories throughout the world. The coefficients of variation of the deviations of the model from the data are distinctly smaller than those of the latest CEB model (1990), and much smaller than those for the previous model in ACI209 (which was devel-oped in the mid- 1960s). The model is simpler than the previous models (BP and BP-KX) developed at Northwestern University, yet it has comparable accuracy and is more rational. The effect of concrete composition and design strength on the model parameters is the main source of error of the model. A method to reduce this error by updating one or two model parameters on the basis of short-time creep tests is given. The updating of model parameters is particularly important for high-strength concretes and other special concretes containing various admixtures, superplasticizers, water-reducing agents and pozzolanic materials. For the updating of shrinkage prediction, a new method in which the shrinkage half-time is calibrated by simultaneous measurements of water loss is presented. This approach circumvents the large sensitivity of the shrinkage extrapolation problem to small changes in the material param-eters. The new model allows a more realistic assessment of the creep and shrinkage effects in concrete structures, which significantly affect the durability and long-time serviceability of civil engineering infrastructure.

A simple model for the characterization of concrete creep and shrinkage in design of concrete structures is proposed. It represents a shortened form of model B3 which was presented in [2] (as and improvement of the original version [3]) and appears in this volume, and an update of a previous short form [4]. The main simplification compared to model B3 comes from the use of the log-double-power law as the basic creep compliance function. The B3 formulae for predicting material parameters in the model are simplified by dropping the dependence of these parameters on the composition of concrete mix, leaving only dependence on the strength and the specific water content of the concrete mix. The model is justified by statistical comparisons with all the data in the internationally accepted RILEM data bank. The differences be-tween the present short-form and model B3 are discussed and limitations of the short form are compared to model B3 are noted. The model is suitable for design of concrete structures with the exception of highly creep-sensitive struc-tures for which the full model B3 is necessary

This paper presents a simple design-office procedure for calculating the shrinkage and creep of concrete using the information available at design; namely the 28 day specified concrete strength, the concrete strength at end of curing or loading, element size and the relative humidity. The method includes strength development with age, relationship between modulus of elasticity and strength, and equations for predicting shrinkage and creep. The only arbitrary information are the factors appropriate to the cementitious material, which can be improved from measured strength age data. At the most basic level the proposed method requires only the information available to the design engi-neer. The prediction values can be improved by simply measuring concrete strength development with time and modulus of elasticity. Aggregate stiffness can be taken into account by back calculating a concrete pseudo strength from the measured modulus of elasticity. Measured short term shrinkage and creep values can be extrapolated to obtain long duration predictions for simi-lar sized elements. The predictions are compared with experimental results for seventy nine data sets for compliance and sixty three data sets for shrink-age. The comparisons indicate shrinkage and creep can be calculated within +/- 25%. The method can be used regardless of what chemical admixtures or mineral by-products are in the concrete, casting temperature or curing regime.

This paper discusses possible mechanisms of basic creep and autogenous shrinkage and their couplings. The starting point is a kinetics analysis of the basic creep of different types of concrete, a normal strength concrete and a high performance concrete. This approach reveals two domains: short term creep kinetics, active for some days after loading, and long term creep kinet-ics, characterized by a pronounced and non-asymptotic aging. Then, by ex-ploring the creep-shrinkage interaction under sealed conditions, we confirm that the long term autogenous skrinkage, which cannot be explained by pure hydration effects, can be associated with a matrix creep under internal pore pressure. This pressure seems to depend mainly on the water:cement ratio. Finally, we present some preliminary experimental results on the creep dila-tancy behavior of concrete. The results indicate that the short term creep is characterized by a viscous dilatant behavior (i.e., positive volume increase rate), while the long term creep is of rather non-dilatant nature occurring at constant volume.

Variation of statistical scheme of reinforced and prestressed concrete struc-tures is frequent in modern construction techniques. Construction sequences may include application of permanent loads and of prestressing in one or more steps, and connection of different portions of the structure, or introduction of additional restraints (sometime forcedly applied to correct the internal stress conditions), at different ages during or after the construction process. The resulting stress distribution is largely influenced by the time-dependent deformability of concrete. The paper presents a unified approach for its evalu-ation based on the linear theory of viscoelasticity for aging materials, which is normally adopted for modeling concrete creep, evidencing the important role played by the non-dimensional redistribution function (t, to, t1) describing the creep induced stress variation at time t for loading at t, and variation of re-straint conditions at t,. The obtained solutions are exact for all problems of variation of restraint conditions in homogeneous structures with rigid restraints, and normally suffkiently accurate for problems concerning structures charac-terized by heterogeneities in the properties of concrete along their structural configuration. Redistribution function may be computed from the creep function J characterizing the creep prediction model under consideration and made available in terms of design aids (graphs or tables). The computational procedure is illustrated and an example of application to a typical structural problem is presented.