Background and goals of this thesis The concrete repair, rehabilitation and retrofitting industry grows rapidly, driven by deterioration of, damage to and defects in concrete structures. However, it is well known that to achieve durable concrete repairs is very difficult. The failure of concrete repairs causes high economical, social and environmental impacts. The main reason of failures is that most repair materials and the repair-substrate interface cannot withstand the high stresses induced by differential shrinkage. Failures, therefore, manifest themselves by cracking in the repair material and debonding of the repair material from the substrate. The first goal of this thesis is to develop a “green” Engineered Cementitious Composites (ECC) and to demonstrate the good performance of this “green” ECC. The second goal is to develop a numerical tool, which can simulate the bond strength between the repair material and the concrete substrate, to assist engineers in designing a durable concrete repair with good bond. Research methodology The stresses induced by differential shrinkage in repair systems were investigated analytically. The analytical study has shown that differential shrinkage induces lower stresses in the ECC repair system, compared with conventional repair materials. For this reason, ECC was chosen as a repair material in this research. Conventional ECC contains a larger amount of cement than concrete, typically two or three times larger. The high cement content leads a high shrinkage, high costs and poor sustainability performance. A “green” ECC with low cement content was developed by partially substituting Portland cement with limestone powder and blast furnace slag. The surface cracking and interface delamination of the “green” ECC repair system were investigated experimentally. In order to demonstrate the potential of ECC to prolong the service life of the repair structures, the rapid chloride migration (RCM) test was used to investigate the chloride penetration profile in the cracked ECC repair system. In order to investigate the microstructure development and the bond mechanism in concrete repairs, the experimental techniques, including non-evaporable water test, MIP, ESEM and bond strength test, were applied. In order to quantitatively study the microstructure development and the bond strength in the repair system, the cement hydration model HYMOSTRUC was extended. The moisture transport between the two materials (the repair material and the concrete substrate) and the cement hydration process of the repair material was taken into account in this numerical tool. Summary of the results of this thesis The analytical study has revealed that the shrinkage of the repair material, the size of the repair system, the Young’s modulus and the roughness of the substrate influence the distribution and magnitude of the differential shrinkage-induced stresses. It was also found that differential shrinkage induces lower stresses in the ECC repair system compared with conventional repair materials. This thesis demonstrates the feasibility of designing a “green” ECC with limestone powder and BFS. This mixture has a Portland cement content as low as 15% (by weight), which is about half of the standard ECC. At 28 days, the “green” ECC shows high tensile strain capacity of 3.3% and a moderate compressive strength of 38. Subjected to a differential shrinkage, the “green” ECC repair system shows a larger number of cracks and smaller crack width (40 ?m) than a conventional repair material, e.g. a fiber-reinforced polymer-modified repair mortar. Even though ECC cracks, it can carry more tensile load and accommodate larger tensile strain. Due to the smaller crack width, the maximum chloride penetration depth in ECC is much smaller than that in a conventional repair material. The service life prediction using the DuraCrete model revealed that, in order to achieve the same service life, the ECC repair system needs smaller cover thickness than the conventional repair material. It is also found that the bond strength is a crucial factor influencing the performance of the ECC repair system. To enhance the bond strength, therefore, is very important to realize durable ECC repairs. The moisture transport shows a significant influence on the cement hydration and microstructure development of the repair material. Before setting of the repair material, the concrete substrate absorbs water from the repair material. This causes a reduction of the w/c ratio in the repair material. The reduction of the w/c ratio affects the degree of hydration and the porosity of the repair material as well as the bond strength. After setting of the repair material, the cement hydration works as a “motor” and generates the driving force for the moisture exchange in the two materials, while water acts as “fuel”, which is consumed by the “motor” and influences the efficiency of the “motor”. The numerical study has shown that the microstructure development of the repair material and the bond strength are influenced by many parameters, i.e. the porosity and water content of the concrete substrate, the w/c ratio of the repair material, and the thickness of the repair material and the concrete substrate. The correlation between the bond strength and the microstructure of the repair material was observed numerically. Due to the “wall effect”, the cement particles have a loose packing at the interface, and the w/c ratio locally increases. The increased w/c ratio results in a porous interfacial zone. The tensile strength of the interface is, in case of a smooth surface of the substrate, lower than that of the repair material. As a result, under a tensile load, the specimen with a smooth surface of the substrate fails at the interface. The surface roughness does not influence the moisture exchange and the cement hydration process. However, it has a significant influence on the bond strength. As the surface roughness increases, the contact area between the repair material and the concrete substrate increases as well. The increased contact area contributes to the bonding of the interface. As a result, the bond strength increases, and the failure changes from debonding at the interface to cracking of the repair material. Based on the above fact, ECC, with its high ductility and tight crack width, is a good choice.