The Research On Carbonation Depth And Life Prediction Of Concrete Structures

Concrete is one of the most important materials of civil engineering. It has being widely used in many areas, such as industrial and buildings, transportation facilities, water utilities construction and infrastructure projects ect... Study results show that the performance of concrete would be deteriorated with time and the external load changes, and the safety and reliability of concreate structure would be decreased. The strudy of concrete structure's durability has been become a hot topic of many areas such as structural engineering, material engineering and other disciplines of research.Based on previous studies, the damage causes of concrete structures, the mechanism of reinforced carbonation and corrosion and their factors have been studied systemly. A variety of research nonlinear methods has been used to analyse the carbonation depth and life prediction method of concrete in this thesis.The innovations of this these are listed as follows:1. Concrete carbonation is a complex physical and chemical process. The carbon dioxide in the atmosphere gradually spreads and reacts in the concrete from the outside to the inside. It is affected by many factors and is random and unascertained process. This is the first time, that the unascertained measure theory is applied to the study of concrete carbonation depth prediction and an unascertained means clustering analysis model of carbonation depth prediction is established.2. The key of concrete carbonation life prediction is the rate coefficient of concrete carbonation. Based on BP neural network, the prediction model for concrete carbonation rate coefficient is established in this thesis.3. Set the protective thickness of concrete, concrete carbonation depth, corrosion start time and corrosion depth of steel when the protective layer cracks as evaluation indexes, a model based on the mean cluster analysis for steel corrosion depth prediction is established, and example shows that the newly established prediction model has strong predictive power.