Comparison of different approaches for self-healing concrete in a large-scale lab test

• Both approaches have potential to be applied in real-scale concrete structures.
• Use of encapsulated PU requires more preparation compared to the addition of SAPs.
• One approach is triggered through crack appearance the other by water ingress.
• SAPs resulted in the highest healing efficiency based on crack width measurements.
• Release of PU from the capsules and crack closure was clearly noticed from CT.

After several years of research in the Magnel Laboratory for Concrete Research (Belgium) to obtain concrete with self-healing properties, two of the most promising mechanisms were tested on a larger scale. One mechanism is based upon the encapsulation of polyurethane which is embedded in the matrix. Self-repair is obtained when crack creation causes capsule breakage, release and subsequent hardening of the polyurethane inside the crack. The second approach relies upon the addition of superabsorbent polymers (SAPs) to the concrete. These SAPs take up water entering via the crack, swell and block the crack. In addition, when they release their water content later on, they induce continued hydration and calcium carbonate precipitation. Real-scale concrete beams (150 mm × 250 mm × 3000 mm), with and without self-healing properties, were made and the self-healing efficiency was evaluated after crack creation by means of four-point bending. Based on the measured crack width reduction over time, it was shown that improved autogenous crack healing was obtained when superabsorbent polymers were added to the mixture. From the acoustic emission analysis, the proof of glass capsule breakage upon crack formation was obtained. X-ray tomography, fluorescent light microscopy and thin section analysis demonstrated that cracks were indeed partially filled with hydration products, calcium carbonate crystals and/or polyurethane which leached from the broken embedded capsules. Although it would be expected from both findings that this would result in a decrease of water ingress into the healed cracks, this could not be proven within this study.