Inhomogeneity Study and Prediction Model of Carbonation Depth of Recycled Aggregate Concrete


Student thesis: Doctoral Thesis

View graph of relations



Awarding Institution
  • Ganghua Pan (External person) (External Supervisor)
  • Kim Meow LIEW (Supervisor)
Award date3 Aug 2021


With the fast growing of sustainable and green construction materials, the carbonation of recycled aggregate concrete (RAC) and the steel corrosion in RAC structures induced by carbonation have attracted intensive interests. However, it is of great difficulty to evaluate the carbonation degree of RAC precisely due to its structural inhomogeneity. To achieve remarkable economic, environmental and social merits, it is important to reveal the effect of inhomogeneity on the carbonation resistance of RAC and predict the carbonation degree accurately.

The carbonation degree inhomogeneity of RAC due to the addition of old mortar (OM) can cause potential errors during the carbonation evaluation of RAC structures. Therefore, the carbonation depth inhomogeneities of natural aggregate concrete (NAC) and the RAC samples prepared with different recycled coarse aggregates (RCAs) are examined. The effects of different strengthening methods on the carbonation depth inhomogeneity are then explored. Moreover, calculation methods for the carbonation zone widths and steel corrosion zone widths of RAC are developed. The results showed that the carbonation depth inhomogeneity of RAC was greater than that of NAC. Additionally, although using nano-slurry or CO2 curing methods can reduce carbonation depth, these methods cannot reduce the carbonation depth inhomogeneity of RAC, and altering the CO2 concentration and carbonation period cannot improve this effect. Fortunately, selecting RCA produced by proper-strength original concrete to produce the RAC can reduce its carbonation depth inhomogeneity. Moreover, selecting a larger carbonation zone width or steel corrosion zone width can avoid overestimation of the carbonation service life of RAC structures. Furthermore, polynomial functions are suggested for the calculations of carbonation zone widths and steel corrosion zone widths obtained using new mortar (NM) and OM as samples.

The respective effect of aggregate and interfacial transition zone (ITZ) on the carbonation resistance of cement-based materials cannot be ignored. A CO2 diffusion model, based on Fick's second law, that considers the effects of aggregate and ITZ is proposed. A CO2 reaction model based on the law of conservation of mass is also proposed. Subsequently, a calculating methodology for the carbonation service life of reinforced concrete beams considering the distribution of carbonation zones is proffered. The results showed that the carbonation prediction method considering the effects of aggregate and ITZ was capable of handling the diffusivity of CO2 in cement-based materials, and this theoretical model can thus predict the carbonation depth values with reasonable accuracy. Additionally, the service life results that account for carbonation zone distribution were lower than those that do not account for the same. It was also observed that using the width sum of the complete carbonation zone and the semi-carbonation zone with the pH value between 9.0 and 11.5 in tensile zone was the most conservative method to calculate the carbonation service life of reinforced concrete beams under load effects.

There should be a better understanding regarding the inhomogeneity of CO2 diffusion coefficients of RAC. Therefore, the pore structures of NMs and OMs in RACs with the RCAs strengthened by different methods are examined. The CO2 effective diffusion coefficients in these NMs and OMs are then modelled based on Fick’s second law and mass conservation law, and the effect of strengthening methods on the diffusion coefficients is explored. The findings of the study revealed that compared with a nano-strengthening method, CO2 curing can more effectively reduce the porosity of the matrix, and the effect of high-concentration CO2 gas on the porosity reduction was more obvious than low-concentration. In addition, selecting a high-strength original concrete to prepare RCA can more effectively decrease the porosity of OM than using other methods, and RCA type showed no effect on the constrictivity. Clear differences for the CO2 effective diffusion coefficients of OMs in different RCAs were observed. Therefore, selecting appropriate RCA for improving the carbonation resistance of OM was more effectual than using some complex and uneconomic strengthening methods.

Carbonation aggravates the inhomogeneity of ITZs in RAC. Two kinds of modelled RACs before and after accelerated carbonation are studied using phenolphthalein spray test, pore structure test, backscattered electron test and nanoindentation test. The results showed that ITZ effect primarily concentrated on the weak mortar. Additionally, the ITZs before carbonation were more porous compared with those after carbonation, and accelerated carbonation decreased the porosity results of NM and OM. Besides, there existed much unhydrated cement in ITZ compared with OM; however, a different trend was observed between ITZ and NM. Furthermore, clear differences were observed among the elasticity modulus and ITZ widths before carbonation, and carbonation increased the elasticity modulus and decreased the widths.

The effects of mortar and ITZ inhomogeneities on the carbonation resistance of RAC cannot be ignored. The CO2 diffusion models in different paths considering mortar and ITZ inhomogeneities, based on series-parallel theories in the literature and multi-path modelling theory proposed by the author, are first proposed. Certain crucial parameters (RCA type, RCA replacement ratio, mixing method and loading effect) are then reflected in the diffusion models. Subsequently, CO2 reaction models in different paths are proposed. The reliability of theoretical modelling was further verified by the comparisons with experimental data. The results showed that multi-path modelling method and series-parallel theories were capable of handling the diffusivity of CO2 in RAC reflecting its inhomogeneity. Comparisons of modelling results and experimental data from this study and the literature revealed that this developed theoretical model can predict the characteristic values of carbonation (maximum and average values) with reasonable accuracy. In addition, using weighted average values to evaluate the carbonation degree of RAC can underestimate the results and using maximum carbonation depth can avoid the underestimation of RAC carbonation degree.