|
Title |
Determination of Degree of Hydration, Temperature and Moisture Distributions in Early-age Concrete
|
|
Keywords |
초기재령 콘크리트; 수화도; 온도 및 습도분포; 3차원 유한요소해석 프로그램; 재료 모델; 고강도 및 고성능 콘크리트; degree of hydration; temperature and moisture distribution; finite element program; age dependency desorption isotherm; self-desiccation; early age cracking |
|
Abstract |
The purpose of the present study is first to refine the mathematical material models for moisture and temperature distributions in early-age concrete and then to incorporate those models into finite element procedure. The three dimensional finite element program developed in the present study can determine the degree of hydration, temperature and moisture distribution in hardening concrete. It is assumed that temperature and humidity fields are fully uncoupled and only the degree of hydration is coupled with two state variables. Mathematical formulation of degree of hydration Is based on the combination of three rate functions of reaction. The effect of moisture condition as well as temperature on the rate of reaction is considered in the degree of hydration model. In moisture transfer, diffusion coefficient is strongly dependent on the moisture content in pore system. Many existing models describe this phenomenon according to the composition of mixture, especially water to cement ratio, but do not consider the age dependency. Microstructure is changing with the hydration and thus transport coefficients at early ages are significantly higher because the pore structure in the cement matrix is more open. The moisture capacity and sink are derived from age-dependent desorption isotherm. Prediction of a moisture sink due to the hydration process, i.e. self-desiccation, is related to autogenous shrinkage, which may cause early-age cracking in high strength and high performance concrete. The realistic models and finite element program developed in this study provide fairly good results on the temperature and moisture distribution for early-age concrete and correlate very well with actual test data.
|