Investigation into the mechanism of shrinkage is essential for the engineering purposes to mitigate cracking of concrete structures. Because comprehensive analysis of unit processes of self deformations is inevitable for systematic understanding of interaction between various self deformations, such as drying shrinkage, autogenous shrinkage and thermal deformation etc.,or their influences on stress generation and crack extension or crack propagation. In this paper one example of this trial is presented based on analysis of relationship between weight loss and drying shrinkage measured by ASTM method. By this analysis, it is shown that unit processes involved in these physical phenomena are not so simple as can be stated by Fick's diffusion theory. Depending upon duration of underwater curing or degree of hydration, unit processes are drastically varied suggesting that variation in microstructure gives rise to quite different situations for desiccation and deformation. Based on these models, relation and interaction between drying shrinkage and autogenous shrinkage is discussed.

In this paper, measurements of non-evaporable water content, chemical shrinkage, autogenous deformation, internal relative humidity (RH), pore solution composition, and early-age elastic modulus are presented and discussed. All experiments were performed on Portland cement and blast-furnace slag (BFS) cement pastes. Self-desiccation shrinkage of the BFS cement paste was modeled based on the RH measurements, following the capillary-tension approach. The main findings of this study are: 1) self-desiccation shrinkage can be related to self-desiccation both for Portland and for BFS cement pastes, taking into account the influence of the dissolved salts in the pore solution, 2) the BFS cement paste studied shows pronounced self-desiccation and self-desiccation shrinkage, mainly caused by its very fine pore structure.

A micro-macro experimental study has been performed, from the end of mixing up to several months, on a set of plain cement pastes prepared with the same type I ordinary Portland cement (OPC) and various water-to-cement ratios (W/C), and cured at various constant temperatures. Chemical shrinkage, volumetric and one-dimensional autogenous deformations have been measured and analyzed in relation to the hydration process (degree of hydration of the cement a, Ca(OH)7 content, ...) and to the microstructural characteristics of the material. The effects of the curing temperature at early age (< 24 hours) in the range 10-50°C, and of W/C in the range 0.25-0.60, have been investigated. The temperature-induced changes recorded on both the magnitude and the kinetics of volumetric autogenous shrinkage clearly show the irrelevance of using the usual maturity concept to describe such phenomena within the whole early-age period. In addition, a threshold is pointed out at about a = 7%, both defining the range where autogenous shrinkage is linearly related to a and corresponding to the precipitation of Ca(OH)2. Moreover, a W/C threshold is pointed out both at the macro-level (autogenous deformations, ...) and at the micro-level (characteristics of the hydration products, MIP porosity and pore size distribution, ...). The chemical and (micro)structural basic effects of calcium hydroxide are in particular distinguished.

An experimental investigation of HPSCC, is outlined. Optimizations were performed on a laboratory scale according to an ideal grading of the particles in the fresh concrete for SCC, with high strength, high durability in marine environment or with fire spoiling safety. SCC was introduced in the full-scale production of beams and piles. The results showed high slump flow and robustness that allowed for a reasonable variation of the water-cement ratio, w/c, keeping the fresh concrete properties within the limits of the full-scale production even at elevated temperature. Creep, shrinkage, salt frost scaling and sulphate resistance did not differ much from the corresponding properties of vibrated concrete, NC. Internal frost resistance was improved for SCC compared with NC but the chloride migration was larger in SCC with limestone powder than in NC. Spoiling of the concrete during fire, especially in low-w/c concrete, was avoided by use of polypropylene fibers.

Recent experimental results (creep tests and indentation tests at a nanometer scale) on Ductal®, a non-brittle (fiber-reinforced) ultrahigh-performance concrete (UHPC), show that only one constituent of this composite, the C-S-H phase, exhibits creep. Former creep tests on hydrated cement paste have shown a very high creep rate of the cement gel which decelerates very slowly (much more slowly than concrete creep). Furthermore, these results provide a clear explanation for the observations of a strong correlation between shrinkage and creep values. The reason is, when hydration rate becomes negligible (typically after a few weeks), the dominant part of shrinkage is nothing but the viscoplastic response of the cement gel to the internal stress which is applied by the liquid phase on the pore surface. This statement makes wrong the last argument against the explanation of shrinkage by capillary tension, the so-called argument of reversibility. Creep aging, as well as the very low creep of high-strength concretes can be explained by the consumption of creep potential by the hygral stress. Several coupling effects between creep and shrinkage can be explained, as for example the so-called PICKETT effect.