Self-Healing Concrete Encapsulated polymer precursors as healing agents for active cracks

Self-healing concrete aims at the autonomous healing of small cracks, with widths in the order of a few hundred micrometers. This way, the service life of reinforced concrete structures can be extended and their watertightness guaranteed for longer periods. Self-healing technologies studied so far are based either on promoting the autogenous self-healing that occurs naturally in concrete or on the addition of encapsulated healing agents. With self-healing concrete based on encapsulated polymer precursors, capsules rupture upon crack formation and the precursors flow to fill the crack and subsequently cure to form a cohesive polymer bonding the opposite sides of the crack. Most importantly, polymers cover a wide range of mechanical properties that may allow overcoming the high stiffness and brittleness of mineral materials filling the cracks when other self-healing technologies are employed. Such mechanical properties of the healing material are thought to result in the reopening of healed cracks or the formation of new ones if the loading on the concrete element is increased or has a cyclic nature. The major motivations behind the research presented here were the assessment of the fitness of commercial precursors for self-healing of active cracks and the upscaling of the technology, whose bottleneck was the inability of the capsules to survive when mixed into concrete. As with other emerging technologies, self-healing concrete lacks straightforward test methods to assess its performance, especially those that simulate field conditions, so the development of test methods was also a major motivation. Several moisture-curing precursors of polyurethane were selected for screening based on their low viscosity (required for good dispersion inside a crack) and high elongation and adhesion to concrete (required for healing of active cracks). The precursors tested resulted in 8 series, which covered a wide range of properties in terms of viscosity, foaming potential and mechanical properties after hardening. The screening tests were performed on mortar prisms with a single crack and containing glass capsules with thin walls of 0.18 mm. Series SLV (polyurethane of super low viscosity) stood out as the best performing due to its better dispersion upon release from the capsules and a reduction of water uptake through the healed cracks, to the level of sound specimens. This polymer was able to withstand a strain level of at least 50%, after which it detached from the crack faces, which may be below the crack movement magnitude found in some field structures. A study was performed to assess the fitness of continuous monitoring methods to detect failure due to excessive strain on polymers bridging active cracks and to highlight the consequences of using stiff polymers for this application. Detection of failure through acoustic emission analysis coupled with DIC was possible only in case of failure due to brittle fracture of a rigid foam after 9% strain, which generated high-energy acoustic events. Direct observation of interfaces with SEM in-situ loading allowed determination of failure of a rigid foam due to cracking of the polymer matrix and detachment at the interface with the cementitious matrix, with an onset at 5% strain and complete detachment at 16% strain. For a flexible, continuous film of polymer based on a nonfoaming precursor, detachment occurred before 50% strain. Assuming adequate adhesion, polymers with high elongation (>100%) and modulus of elasticity much lower than 10 MPa are suggested if cracks subjected to a realistic amplitude of movement are targeted. Healed mortar prisms were also tested for their resistance against high hydrostatic pressure and fatigue due to cyclic crack movement, with the latter involving parallel monitoring with transmitted ultrasound pulses. The healed cracks remained sealed under all pressures tested, up to 2 bar (or 20 m of water column). Strain limits were overall reduced to below 20% of the original crack size due to fatigue under hundreds of bending cycles, while under a single cycle the strain limit of the SLV polymer had been determined to be around 50%. Performance assessment proceeded at a large scale with concrete specimens healed with the best performing SLV precursor, but still using pre-placed glass capsules. With ultrasound transmission analysis it was possible to detect and monitor healing, as an increase of maximum amplitude of ultrasonic pulses occurred mostly during the first 10 hours after cracking and stabilized after 48 hours. The concrete specimens containing a single healed crack were loaded under bending and a considerable 47% or 74% regain of strength, depending on the dimensions of the specimens, was achieved due to healing of a large part of the cracked plane, while stiffness increased by either 19% or 75%. Trials were performed with additional non-destructive techniques. Vibration analysis was easy to implement and able to detect both a decrease and an increase of elastic modulus of the concrete specimens respectively after cracking and after healing. Localization of acoustic events made acoustic emission analysis particularly useful, as it allowed guaranteeing that specific emissions were due to damage occurring at the healed section. Several configurations of capsules were tested in order to assess their potential for upscaling of self-healing concrete. While spherical polymeric microcapsules did not rupture when crossed by a crack, cylindrical polymeric capsules were not able to keep a moisture-curing polyurethane stable. Cylindrical glass capsules were then determined to be the best encapsulating technique when moisture- sensitive healing agents are used. They also ruptured when crossed by very small cracks in concrete of ~25 μm and survived concrete mixing if their wall was at least 0.5 mm thick. A mechanism for pressurizing the capsules was also developed, leading to a considerable improvement of the release of precursor. To test a final upscaling solution, glass capsules 30 mm long and with a wall thickness of 0.5 mm were mixed into concrete at a dosage of 13 and 36 capsules per litre and moulded into specimens, on which multiple cracks were created. Partial success was obtained, as 2 out of 6 cracked planes were well sealed when tested for water flow under high pressure. A higher dosage of capsules, estimated to be about 50 capsules per litre would be required to achieve a consistent sealing effect. Finally, a cost assessment of this solution was performed and compared to alternative options. Addition of capsules to a whole concrete slab would be too expensive compared to a typical repair action. If self-healing concrete is applied only to the outer 5 cm layer of the concrete element, the cost of this solution becomes 19% cheaper than conventional repair, but still more expensive than the cheapest alternative solution to achieve sealed cracks, which is using additional reinforcement to keep the crack size below 0.05 mm.

http://ift.tt/2p4N8cF