This paper describes laboratory tests on nine simply supported one-way reinforced concrete members subjected to immediate live load and sustained load. The specimens are 12 ft (3.66 m) long supported on an 11 ft. (3.35 m) span, 12 in. (304.8 mm) wide, and 5 in. (127 mm) deep with two #3 bars at an effective depth of 4 in. (101.6 mm) providing a reinforcement ratio of 0.0046. The specimens were moist cured for up to seven days. Three specimens each were removed from the forms and subjected to immediate live load at three days, seven days, and twenty eight days followed by sustained load due to self weight. Each specimen was subjected to immediate full live load again after six months. Applied load and mid-span deflections were recorded under immediate live load and sustained load. The test results demonstrate the effect of shrinkage restraint provided by embedded bars on the flexural cracking of the specimens under applied load, as well as effects of early age loading on time-dependent response. Calculated deflections based on effective moment of inertia for immediate deflection and long time multipliers for time-dependent deflection are compared with measured deflections.

Concrete cracks when the tensile stress exceeds the tensile strength. Tensile stress arises from restrained contraction. The main sources of contraction are early age effects, temperature drop and drying shrinkage. Creep offers significant relief to early-age contractions, but only a little to shrinkage. Tensile strength reduces under sustained stress, so the duration of the contraction is important. A new graph is presented to show that although cracking can occur at any age, it is most likely either during the first 3-10 days or after some years. This is because the creep relief provides a margin for temperature variations and some shrinkage to occur without exceeding the critical tensile stress. However this margin diminishes with time. A new flow chart is then presented which shows what controls are necessary at each of these two key stages. This differentiates between avoiding cracking and controlling cracking, an important distinction that is often confused in the approach to design. The case study of a feature wall where this was not appreciated is described.

This paper shows the importance of autogenous shrinkage on serviceability performance of reinforced high-strength concrete (HSC) flexural beams, and also the effectiveness of low-shrinkage HSCs (LS-HSC) that is made by using an expansive additive and/or shrinkage-reducing chemical agent and/or Belite-rich low heat Portland cement with regards to the improvement of flexural serviceability performances of the beams. In addition, this paper, from a simple design equation point of view, proposes a new concept for evaluating flexural crack widths and deformation of RC beams, considering the early age deformation of concrete before loading. The experimental results show that autogenous shrinkage of HSC increase crack width and deformation of the RC beams significantly, while LS-HSCs markedly improve its serviceability performances. The present concept, taking into account strain change in tension reinforcement and curvature change at cracked section before and after loading, is effective in explaining the effects of shrinkage and expansion of concrete before loading on maximum crack width and flexural deformation. JSCE (Japan Society of Civil Engineers) code equations for predicting maximum crack width and flexural deformation into which the present concept is incorporated improve the prediction accuracy compared with conventional ones and show fairly good agreement with experimental results.

An extensive experimental study has been carried out to extend the Belgian technique of steel-concrete prebent beams to Very High Performance Concrete (VHPC: self-compacting C80/95 with silica fume). In comparison to the present constructions in Belgium, the main advantage of using VHPC instead of C50 concrete is to decrease the prestressing losses of the system thanks to a significant decrease of the creep deformations, together with the possibility to optimize the beam weight and its serviceability domain. As one research step was the casting of concrete around the bottom flange of two 13m-span steel girders, it was necessary to use a self compacting concrete with a suitable maximum aggregate size due to the very dense network of wires, sensors, ribbed stirrups and steel square ribs. In addition, the VHPC had to develop a very high compressive strength at early age to remove the prebending loads applied on the steel girders 48 hours after casting. The autogenous shrinkage development was examined both under standard isothermal (20°C) and realistic temperature (same as within the beams) conditions. Mechanical characterization, Young’s modulus, creep and shrinkage tests in standard (20°C, 50% RH) and variable (same as within the beams) ambient conditions were performed, so that a correct analysis of the structural behavior of the beams could be done. Due to the efforts of mixture proportions optimization, it could be demonstrated that VHPC delayed deformations were reduced by about a factor 2 in comparison with currently used C50 concrete, which made the use of prebending significantly more efficient. The purpose of this paper is to report on these experimental investigations and to present the computation method which was used to predict the time-dependent evolution of VHPC creep and shrinkage based on European model codes for the creep and shrinkage and on the recovery model proposed by Yue & Taerwe for the creep recovery.

Previously, a significant research effort was made to analyse the long-term behaviour of a kind of composite steel-concrete, prebent and prestressed beam used extensively in Belgium as simply supported railway bridge decks. They are trough type with U shaped cross section. Parallel with that, an extensive experimental research has been carried out to extend the Belgian technique of steel-concrete prebent beams to Very High Performance Concrete (VHPC). The main advantage of using VHPC instead of the C50 concrete currently used is to decrease the prestressing losses of the system thanks to a significant decrease of the creep deformations. One of the motivations behind this research was to develop realistic models to assess the feasability of constructing continuous railway bridges by connecting such simple decks over intermediate supports. For railway bridge decks, the condition of no cracking under load is paramount. This implies to devise a form of in situ prestressing of the connection between the simply supported decks. A structural analysis program was developed that could cope with multiple phases of construction and loading for such a highly heterogeneous structure. The structural analysis is performed within the framework of beam-type displacement-based finite element analysis. A step-by-step time-dependent analysis is based on the algorithm of superposition. The creep and shrinkage model is the CEB-FIP MC 90 model, which was found to best reproduce the strains recorded in the laboratory creep and shrinkage tests for the C50/60 concrete of the singly supported decks and for the C80/95 VHPC. Creep recovery effects occurring at unloading have been accommodated in the step-by-step time-dependent analysis with the two-function method proposed by Yue and Taerwe. The purpose of this paper is to present a design optimization study of the construction phases of a continuous bridge with prebent and prestressed composite decks. The most influential parameters on the long-term behavior of the continuous bridge are analyzed.