| Currently, the workability and physical-mechanical properties of concrete have been greatly improved. Most present research interests turn to focus on improving the durability of concrete. One of the most common problems of durability is the cracking caused by deformations. The cracking as a result of the drying shrinkage of concrete is still hard to control efficiently, which is directly because the knowledge of drying shrinkage of concrete has not reached a high enough level yet.Concrete mainly consists of hardened cement paste and aggregate. The aggregate is stiff enough and deforms little when the relative humidity of environment is changed, while the hardened cement paste is very easy to deform. Therefore, the study on the drying shrinkage of hardened cement paste of concrete is of great importance.An improved method to test the drying shrinkage of hardened cement paste was developed and adopted in this research. By this means, the effect of water to cement ratio (w/c), age, relative humidity of drying, and curing regime on drying shrinkage of cement pastes was investigated and analyzed. The orthogonal analysis technology is used to evaluate the weight of influence of the above factors. Three kinds of curing regimes are used as follows: cured at 20℃all the time; cured at 60℃for the first day, and then at 20℃for the rest; cured at 60℃from beginning to the end. TG-DSC thermal analysis was employed to measure the free water, adsorbed water and bound water of cement paste. And pore structure of cement paste was observed by N2 adsorption method. These results could give further insights on the interpretations of the drying shrinkage mechanism of hydrated cement pastes.On the basis of the above, the following conclusions could be reached:①The weight loss and shrinkage as well as its irreversible component of the cement pastes caused by drying are increased with the increasing of water to cement ratios from 0.3 to 0.5. This can be explained by the experimental results of pore structure analysis and TG-DSC thermal analysis. The results of pore structure analysis show that the pore volume at all levels of pore size and the total volume of pores increase with the increasing of w/c. The results of TG-DSC thermal analysis show that the TG mass loss is increased with the increasing of w/c, which means the porosity of the cement paste with higher w/c is higher. However, the above results may not be observed if the cement paste is too young, for example, less than 3 days old, or the cement paste is cured in the environment with the temperature much higher than 20 cecius degree because of the influence of age and heat curing on the pore structure of hardened cement pastes.②The weight loss and the irreversible shrinkage of the cement pastes with different ages from 3 days to 1 year caused by drying in environment of 42% relative humidity are decreased with the growth of cement paste, while its reversible component of drying shrinkage is increased with the growth of cement paste. As for the total drying shrinkage, it is increased with the age when the paste is only 3 to 28 days old, otherwise, it is decreased with the age when the paste is older. The results of TG-DSC thermal analysis show that the gel pores develpe quickly at the earlier age from 3 days to 28 days due to the more rapid hydration, more gel pores means more shrinkage when dried even though the weight loss is lower. After the age of 28 days, the hydration becomes more and more slowly, the majority of the later hydrated product is the inner and much denser one which can be concluded by the pore structure analysis. Higher density C-S-H gel is hard to shrink, which cause the less weight loss and less shrinkage. When hydrated with time, the degree of the silicate polymerization along with hydration is increased. The cement paste with higher degree of polymerization is more difficult to polymerize, so the irreversible shrinkage is decreased and thus the reversible part is increased. However, the shrinkage behavior of the cement pastes cured in the environment with temperature much higher than 20 cecius degree may be different because heat curing can accelerate the hydration, and change the pore structure of hydrated products.③The cement pastes cured in different temperature conditions perform different shrinkage behavior. In detail, provided the age is same, the drying shrinkage of cement paste cured at 20℃all the time is higher than that cured at 60℃for the first day, and then at 20℃for the rest, and the latter is higher than that cured at 60℃from beginning to the end. The results of TG-DSC thermal analysis and pore structure analysis show the same thing that heat cure not only accelerate the hydration but also change the hydrated products as well as its pore structure and thus make things more complicated.④With the relative humidity used to dry cement pastes is lower, the total drying shrinkage, the weight loss, the reversible and irreversible components of the total shrinkage are all increased at different levels, the increase of irreversible shrinkage is the least and only a little.⑤During wet and dry cycles, the drying shrinkage and its irreversible part increase with the cycles going on until reach a higher constant magnitude, at the same time the reversible part of the drying shrinkage, homever, decrease until reach a lower constant magnitude.⑥Based on the orthogonal analysis results of the above, judged by the range of the drying shrinkage at the same factor, within the range of the factors and levels given in this dissertation, the relative humidity used to dry cement pastes influence the drying shrinkage most, the curing regimes occupy the second, and the water to cement ratio come to the last.⑦Most of the completely hydrated cement paste is the C-S-H gel and crystal of calcium hydroxide, the latter is likely a kind of aggregate which shrink little with the change of the environmental humidity. The C-S-H gel shrinks when the environmental humidity is changed. Therefore, the knowledge of the drying shrinkage of C-S-H gel benefits very much the knowledge of the drying shrinkage of hydrated cement pastes. Part even all of the crystal of calcium hydroxide can be removed from the hydrated cement paste and then relatively pure C-S-H gel can be obtained if the hydrated cement paste is immersed in the saturated ammonium nitrate solution long enough. After all of the calcium hydroxide is removed, if the immerse going on, the calcium can be removed from the C-S-H gel and then the Ca to Si ratio is reduced. The results of the experimental research on the subsequent cement paste or the C-S-H gel from the cement paste decalcified in the saturated ammonium nitrate solution show that in the decalcification process, the relationship between Ca to Si ratio and weight loss of the paste appears to show two linear curves, which intersect at the point where Ca(OH)2 is completely extracted. When the water to cement ratio is 0.5, the crystal of calcium hydroxide in the cement paste cured in 20 cecius degree for 180 days is about 24%. When the cement paste is given, with the degree of decalcification, the subsequent drying shrinkage of the decalcified cement paste is increased, especially the irreversible drying shrinkage increases evidently, while the reversible drying shrinkage is kept almost the same. When the water to cement ratio of cement paste is increased, the effect of decalcification on the drying shrinkage as well as its irreversible component is increased. The percentage of irreversible component to the total drying shrinkage is related with the degree of decalcification, and almost linearly increased with the weight loss during decalcification, while the effect of curing regimes on the percentage of irreversible component to the total drying shrinkage is not so much. When the decalcification of cement paste is heavily enough, the effect of curing regimes on the total subsequent drying shrinkage of decalcified cement paste is much effective, in detail, the drying shrinkage of cement paste cured at 20℃is the most, the one cured at 60℃for the first day, and then at 20℃for the rest occupy the second, and the one cured at 60℃is the least.⑧TG-DSC thermal analysis is chosen to test the free water, adsorbed water and bound water of cement paste. And N2 adsorption method is adopted to measure the pore structure of cement paste. By this means, further insights on the interpretations of the drying shrinkage mechanism of hydrated cement pastes are obtained. The results show that the different water analyzed by TG-DSC thermal analysis and the pore structure measured by N2 adsorption method is very consistent with the macroscopical properties of drying shrinkage. Therefore, these two methods could be effective means to get the micro mechanism of drying shrinkage. |