Concrete is the most commonly used building material, however, is not always crack free. When cracks occur, a preferable path for corrosion promoting substances (CO2, Cl- ions, etc.) is created and corrosion might take place. For uncracked concrete the corrosion mechanism is clear and corrosion model is relatively mature. However, research on reinforcement corrosion in cracked concrete is rare. The motivation of this research is to reach a better understanding of the physical mechanisms underlying the deterioration process of reinforcement in cracked concrete subjected to cyclic chloride and to propose a reliable method for estimating these processes.
Laboratory tests were carried out on central cracked concrete beams, in which corrosion systems with single anode and multiple cathodes with different cover depth were arranged. The beams are subjected to both chloride/water wetting-drying cycle and natural exposures. Results indicate that the corrosion mechanism in cracked concrete is macro-cell corrosion which is mainly under concrete resistance control. The time development of corrosion rate is characterized as a three-phase process: the ascending phase, the descending phase and the equilibrium phase. The equilibrium corrosion state enables the prediction of the residual service life. An unconventional effect that corrosion intensity increased with concrete cover depth was observed in the laboratory tests. In order to confirm and explore this effect, numerical modelling on both the process of chloride/moisture penetration into cracked concrete and the simulation of corrosion state in propagation period was conducted.
The mechanisms of chloride transportation into non-saturated cracked concrete is considered as diffusion coupled with convection with flowing moisture, associated with chloride binding which is described by Freundlich and Langmuir binding isotherms. Basic equations were built based on modified Fick’s second law and solved by Alternating-Direction Implicit finite difference method, with required boundary and initial conditions. To figure out the moisture evaporation rate in the crack, laboratory tests were carried out and a half logarithm relationship between the evaporation rate and exposure duration was found. Main parameters for the numerical simulation (diffusion coefficients of chloride and relative humidity, chloride binding capacity and moisture capacity) were estimated by factorial approach. The chloride/moisture penetration process was successfully simulated by the proposed model. The unconventional effect of concrete cover that observed in experimental study was well explained.
The numerical simulation of corrosion state after depassivation was carried out with a self- developed FEM program ‘MCRC’ (Macro-cell Corrosion of Reinforcement in Concrete). Measured polarization curves were used as boundary conditions for the numerical modelling; a three-dimensional-location-dependent resistivity field was built to describe the electrolytic properties of the cracked beam. Corrosion potential and current density of both concrete and steel were obtained. Corrosion in cracked concrete was confirmed as under concrete resistance control.
The accordance between the numerical and experimental results proved the validity of experiment results, meanwhile confirmed the efficiency of the numerical tools.
Based on the chloride/moisture penetration model and the corrosion state model, both the initiation period and propagation period could be estimated.
«
Concrete is the most commonly used building material, however, is not always crack free. When cracks occur, a preferable path for corrosion promoting substances (CO2, Cl- ions, etc.) is created and corrosion might take place. For uncracked concrete the corrosion mechanism is clear and corrosion model is relatively mature. However, research on reinforcement corrosion in cracked concrete is rare. The motivation of this research is to reach a better understanding of the physical mechanisms underlyi...
»