Experimental Investigations And Evaluations On Early-Age Volume Changes And Mechanical Properties Of High Performance Concrete

Due to its high durability, high workability and excellent mechanical properties, high performance concrete (HPC) is more and more utilized in modern infrastructures. However, the performance of HPC has been undermined since it is susceptible to premature cracking which is induced by non-load causes. The cracking is primarily attributed to much more considerable and complicated early-age volume changes of HPC than those of normal concrete. This thesis presents experimental investigations on fundamental components of early-age volume changes of HPC, i.e., thermal deformation, autogenous and drying shrinkage, along with early-age mechanical properties. Based on the experimental data obtained, previous models and formulas, early-age volume changes and mechanical properties are evalutaed.In this thesis, firstly, physical mechanisms and driving forces of early-age volume changes of concrete are reviewed and extended. Secondly, mechanical properties and the triple components of early-age volume changes of HPC are experimentally investigated based on renewed testing methods. In which, evolutions, characteristics, influencing factors, interdependences and potential mechanisms are discussed. Modifications of estimations are also performed on these volume change components and mechanical properties. Finally, by combining the estimations of respective components, early-age volume changes of HPC under a couple of factors are evaluated. So far as the thesis is concerned, main conclusions can be drawn as follows:(1) Characteristics, evolutions and influencing factors of early-age mechanical properties of HPC are definite and interpretable. The parameters of mechanical properties are strongly correlated with each other. The correlations can be expressed with functions that are deduced from normal concrete. Compressive strength is linearly correlated with the peak value of hydration-induced temperature rise. This makes it possible to predict strength from early-age temperature evolution. The estimation equations of compressive strength and Young's modulus originated from CEB-FIP Model Code are modified by introducing setting time factors. Compared to several commonly used equations, the modified ones are more adaptable to early ages.(2) Temperature evolution curves of HPC under semi-adiabatic and isothermal conditions can be characterized with the same exponential function. Thermal expansion coefficients of hardening concrete can be exactly measured with the method based on a lower temperature change zone. Time-dependent thermal expansion coefficients sequentially undergo sharply decreasing, slightly increasing and continuously stable stages. Calculated thermal strain based on time-dependent thermal expansion coefficients are generally higher than their counterparts based on a constant value of thermal expansion coefficient. The both calculated results under semi-adiabatic condition are linearly correlated with each other.(3) Higher curing temperature results in greater magnitude and rate of autogenous shrinkage. The influences of curing temperature are more significant at earlier ages and with lower water-binder ratios. Although the temperature dependence of autogenous shrinkage cannot be properly evaluated by the original equivalent age equation, a modification on activation energy of hydration in the equation improves evaluations to some extent. Autogenous shrinkage curve under semi-adiabatic condition is quite different from the one under isothermal condition. The former successively undergoes initial rapid shrinkage, abrupt expansion and slow shrinkage stages. Compared to HPC mixtures containing fly ash or blast furnace slag, pure cement mixtures show higher abrupt expansion.(4) The relationship between autogenous shrinkage and compressive strength under isothermal condition could be expressed as a quadratic polynomial function. Based on theories of surface tension of adsorption layer by Breugel and Koenders, and with the aid of HYMOSTRUC software package supplied by Microlab, early-age autogenous shrinkage of HPC is predicted. The results match well with the experimental data.(5) The rates of drying shrinkage are much smaller than those of autogenous shrinkage for HPC with lower water-binder ratios. Meanwhile, the percentages of drying shrinkage in total shrinkage (excluding thermal deformation) are relatively smaller. The magnitude, rate and percentage of drying shrinkage increase with an increase of water-binder ratio. When the drying initiation age is delayed, the magnitude and rate of drying shrinkage reduces accordingly. Drying shrinkage is approximately linearly correlated with autogenous shrinkage. The ratio of drying to autogenous shrinkage tends to increase with a delay of drying initiation age. Drying shrinkage strain is positively related to moisture loss ratio of the specimen. The modified equation in which a series of influencing factors are taken into account is compatible with the test data of drying shrinkage.(6) Early-age autogenous and drying shrinkage are non-independent phenomena. It causes large errors when the both components are simply separated with the superposition principle. A modified version of the superposition principle is proposed. In the modification, a reduction coefficient of autogenous shrinkage is introduced to consider the effect of external drying. It is more reasonable that evaluates early-age volume changes of HPC under multiple factors with the modified superposition principle.