Creep and shrinkage of concrete subjected to variable environmental conditions

Abstract

This thesis deals with the numerical modeling of concrete creep and shrinkage at variable environmental conditions. At lower relative humidity concrete creeps more slowly than at full saturation but during drying it creeps faster. Creep is also accelerated at elevated temperature or by temperature variations.One of the physically based models for concrete creep and shrinkage that takes into account variable temperature and humidity is based on the theory of microprestress and solidification (MPS). Unlike the MPS model to the models from the design codes which use the average cross-sectional approach, the MPS model operates at the material point level, which makes it possible to capture the stress distribution more realistically. Assessment of the MPS model is the main topic of this work. Several severe deficiencies of this model have been identified and appropriate remedies have been proposed. Comparing to the experiments, the original formulation of the MSP model exhibited the opposite size-effect on drying creep, spurious sensitivity to the particular choice of relative humidity and excessive compliance during the repeated cycles of temperature and relative humidity. First, the model was reformulated, making the governing equations simpler yet equivalent. Afterwards, the model was improved and validated on typical examples from the literature.The MPS model was implemented into the open-source finite element package OOFEM developed mainly at the Department of Mechanics, Faculty of Civil Engineering, CTU in Prague.The model was calibrated on the experimental specimens, and afterwards applied in the analysis of a real-world structure - a concrete floor subjected to drying. The results indicate that even the improved model needs further improvements regarding the relationship between shrinkage and relative humidity; however, this relationship cannot be uniquely identified from the currently available experimental data. The engineering approach of shrinkage updating based on short-time measurements has also been questioned.

This thesis deals with the numerical modeling of concrete creep and shrinkage at variable environmental conditions. At lower relative humidity concrete creeps more slowly than at full saturation but during drying it creeps faster. Creep is also accelerated at elevated temperature or by temperature variations.One of the physically based models for concrete creep and shrinkage that takes into account variable temperature and humidity is based on the theory of microprestress and solidification (MPS). Unlike the MPS model to the models from the design codes which use the average cross-sectional approach, the MPS model operates at the material point level, which makes it possible to capture the stress distribution more realistically. Assessment of the MPS model is the main topic of this work. Several severe deficiencies of this model have been identified and appropriate remedies have been proposed. Comparing to the experiments, the original formulation of the MSP model exhibited the opposite size-effect on drying creep, spurious sensitivity to the particular choice of relative humidity and excessive compliance during the repeated cycles of temperature and relative humidity. First, the model was reformulated, making the governing equations simpler yet equivalent. Afterwards, the model was improved and validated on typical examples from the literature.The MPS model was implemented into the open-source finite element package OOFEM developed mainly at the Department of Mechanics, Faculty of Civil Engineering, CTU in Prague.The model was calibrated on the experimental specimens, and afterwards applied in the analysis of a real-world structure - a concrete floor subjected to drying. The results indicate that even the improved model needs further improvements regarding the relationship between shrinkage and relative humidity; however, this relationship cannot be uniquely identified from the currently available experimental data. The engineering approach of shrinkage updating based on short-time measurements has also been questioned.