Each of the Universities led the development of different self-healing techniques. At the University of Cambridge, their research focused on the development and incorporation of microcapsules containing mineral healing agents, such as sodium silicate, into the concrete. These microcapsules are ruptured by the propagation of cracks, thereby releasing healing compounds into the crack plane that seal it.
This action serves to block the ingress of harmful substances, hence reducing the permeability and enhancing the durability, as well as aiding some recovery of structural strength Kanellopoulos et al.
For the field trials, the microcapsules were scaled-up in collaboration with Lambson. Cardiff University developed a technique that uses shape memory polymers SMP to close cracks in concrete structures. This followed on from previous research at the university into the use of polyethylene terephthalate PET strips to induce a compressive stress in concrete, which reduces the crack size and enhances autogenous healing Jefferson et al.
Flow networks that can be placed in concrete structures were also developed at Cardiff. These consist of a network of artificially created small diameter channels, through which healing agents can be pumped under pressure. To enable the healing agents to migrate to areas of damage the network is designed for and placed in the zone most susceptible to cracking Gardner et al. The University of Bath's research focused on bacterial self-healing Alazhari et al. Specially selected bacteria, which, in their spore form, can survive in the high alkaline environment of concrete, were used.
Upon the concrete cracking and once the conditions become favorable, the spores germinate and the bacteria break down the nutrients and precipitate calcite within the concrete cracks. This paper describes the site trial concept and design section Concept and Design , the trial panel contents and construction details section Trial Panel Contents and Construction Details , the loading configuration and monitoring undertaken section Loading Configuration, Monitoring, Measurement, and Loading Procedure and a summary and discussion on the key findings of the trial section Results and Discussion.
Costain Group Plc was the lead contractor for the project. An area within the project's site compound was used as the location for the trial. In this way, it was not interfering with the main works but would be exposed to the same conditions and require the same construction processes as the concrete structures being built for the permanent works Teall, The A HoV project includes long lengths of retaining walls of varying height and design and so the trial comprised multiple sections of mock retaining walls, referred to as panels, which were designed to contain different combinations of the self-healing techniques that have been developed.
Two panels that did not contain any self-healing mechanisms were also constructed to act as the controls. The overall structure included a reaction wall for loading the trial panels, as well as a base slab to prevent overturning of the walls during loading. A concept model of the structure is shown in Figure 1.
For all elements of this structure, the detailed structural design was completed according to the provisions of BS EN The panels were designed to crack at mm above the base slab by including 16 mm diameter starter bars at mm centers on the front face up to this point, before changing to an A mesh 10 mm diameter bars mm centers to create a weak section in the panel at this change of steel section location Teall, The nominal cover to the A mesh was 30 mm and the rear face was reinforced with an A steel mesh 6 mm diameter bars mm centers with a nominal cover of 20 mm.
This control mix was designed to have a consistency of class S3 and the measured slump when casting the trial panels was mm. The trial structure was constructed over an 8-week period. The base slab was initially cast and allowed to cure for a minimum of 28 days before casting the reaction wall and finally the trial panels. Each panel on the trial structure was used to test a particular self-healing technique or combination of techniques.
These techniques are detailed in Table 2 and their setup is shown schematically in Figure 2. Ready mix concrete was used and Panels B, D and E were cast using material supplied directly from the mixing truck. To ensure the quality and robustness of the control mix throughout the casting procedure a retarder was added to the mix. Casting, compaction and finishing were carried out in accordance with standard construction practice. For Panel A, prior to casting the concrete was transferred to a l Belle mixer, where the microcapsules were added.
Panel C included a section that comprised concrete using CEM II cement and a lightweight aggregate of bacteria infused perlite particles. This was also mixed on-site using the l Belle mixer. Spherical, polymeric microcapsules, carrying a sodium silicate emulsion, were used in Panel A. Sodium silicate was selected as the healing compound, as it forms products of a similar nature to the host cementitious matrix.
The potential of sodium silicate as a healing agent for cement-based composites was previously investigated, both in terms of crack closure and durability, by Kanellopoulos et al. Figure 3 shows an optical microscope image of the microcapsules. As the microcapsules were provided by Lambson in a preserving aqueous solution, this added an additional small quantity of water to the concrete mix, increasing its water-cement ratio from 0. Figure 4 shows the microcapsules in solution prior to and during their addition to the mix.
Figure 4. A Microcapsules in solution prior to mixing and B addition of microcapsules into the mix. Panel B contained a mat of SMP tendons and flow networks that were set up within the formwork prior to the control mix concrete being poured. These were tied onto the reinforcement in the concrete cover zone and were activated manually after casting. Ten SMP tendon units were placed in the panel, which were designed, upon activation, to generate a stress of 0. Each tendon contained PET filaments, surrounded by a heating system and injection molded sleeves, as shown in Figure 5.
The PET filaments were manufactured by Bradford University specifically for this project, and were found to be capable of generating a restrained shrinkage stress of 30 MPa in the laboratory. These tendons were mm in length and placed at an eccentricity of 40 mm from the center of the panel's cross-section in a staggered layout, as shown in Figures 6 , 7.
Figure 5. Shape memory PET tendon crack closure system Teall et al. Flow networks were included in Panel B to allow the introduction of healing agents into the concrete. A 2D network of 4 mm diameter channels was created using polyurethane tubes, which were removed from the concrete after the formwork had been struck. The channels were connected using 3D printed joints made from polylactic acid PLA , which were tied to the outermost reinforcement, allowing the networks to pass in front of the SMP tendons.
At either side of the panel, the flow networks were terminated with lockable steel injection packers, which allowed each channel to be sealed individually to aid the loading and pressurizing of the network.
The final layout of the tendons and flow networks within the panel prior to casting is shown in Figure 8. Following laboratory experiments in which different potential strains were investigated, the bacteria concrete mix developed by the University of Bath for use in the site trial contained spores of Bacillus pseudofirmus DSM , infused into lightweight perlite aggregate particles. An organic mineral precursor, which included yeast extract and calcium acetate, was also included in separate aggregate particles as a food source for the bacteria.
Due to the challenge of producing a sufficient quantity of spores for an entire panel, it was decided that Panel C would contain three lifts. The first was a mm layer of structural concrete using the control mix, the second a mm layer of bacteria concrete in the zone in which the panels had been designed to crack, and the third a layer of the control mix to complete the panel.
Panel C also contained flow networks as a potential feeding system for the bacteria in the later stages of testing. These networks were formed in the same way as in Panel B. Panels D and E were cast as controls. Panel D was cast using the control mix without any additions, while Panel E used the control mix together with flow networks as in Panels B and C. This was to investigate any impact on the structural properties because of the incorporation of these networks.
The cracks that were to be investigated for healing were created by damaging the panels using controlled loading. A threaded bar and a hollow ram hydraulic jack system was adopted to apply the load.
This system had a bar running through the center each panel and reaction wall at 1. A hollow ram, hydraulic jack connected to a hand pump was then attached to the bar beyond the cradle to enable a load to be applied to the panels.
The general arrangement of this loading system is shown in Figure 9. The reaction wall was designed to be of sufficient strength and stiffness to allow the loading to damage the panels whilst experiencing minimal damage and displacement itself.
Figure 9. Loading arrangement A Front face of trial panel B Rear face of reaction wall. Throughout the site trial, crack widths, deflections, strains, permeability, and applied loading were all monitored. Panel B also contained temperature monitoring equipment and an electrical activation system for the SMP tendons. The surface of each panel was painted with white then black emulsion paint to create a speckled pattern that could be picked up by the dual-camera DIC system.
For Panels A, C and D the pattern was only applied to half of the panel width to allow a comparison to be made between the permeability measurements obtained on the painted and unpainted surfaces.
The surfaces of Panels B and E were completely covered in the speckled pattern to allow monitoring of the strain development over the whole of the panel to determine the performance of the SMP tendons. The two LVDTs were placed on the front face to monitor cracks opening and four LVDTs were located between each panel and the reaction wall to monitor the displacement of the panel and reaction wall during loading. The LVDTs were attached to a RHS steel column, which was in turn bolted to the base slab using chemical anchor bolts, to provide displacement readings of the panel independent of the reaction wall.
All LVDTs were covered with aluminum sheet boxes to protect them from the weather. Five crack width measurements were taken at the change of section CoS location across the width of each panel using a hand held microscope. A notched gauge was used as a scale for each image and the crack width was measured perpendicular to the crack direction using ImageJ software Schneider et al. Three measurements were made from each image, at approximately equal spacing across the field of view and then averaged to give a single crack width value for that location.
Figure A non-destructive air permeability measurement device Torrent device was used to measure the permeability of Panels A, B, C, and D prior to cracking and just after unloading. These values provided a base line for comparison with permeability measurements taken over the entire monitoring period.
For all panels three permeability measurements were taken in the expected cracking location prior to loading. For Panels A and C, a further 3 measurements were taken over the height of the panel to monitor any permeability variations due to the addition of the self-healing techniques.
After unloading, permeability measurements were taken only in the cracked region. The testing and monitoring schedule adopted is shown in Table 3. The panels were then further loaded until a 0.
Panels B and E were loaded to 20 kN post-cracking to ensure repeatability following activation of the polymer tendons. Prior to activating the SMPs tendon, Panel E was unloaded from the locked off state and together with Panel B was again loaded and unloaded to 20 kN, to remove the contribution of short term autogenous healing.
The crack width was then measured by taking photos of the crack at the five locations across the width of each panel, and measuring the distance between the DEMEC pips as described in section Results and Discussion.
Once the crack width had been measured, the load was reduced back to zero in a controlled manner over a period of a few minutes. At zero load, the crack widths were measured again. Throughout the loading, sustained loading and the unloading cycle the DIC camera system was used to take sequential images for post-processing. Measurements from all of the LVDTs were taken continuously at a sample rate of 4 Hz during the loading and unloading stages.
One of the aims of the M4L project was to demonstrate that the self-healing techniques that were being investigated could be employed in large-scale applications and this was successfully achieved as evidenced in Figure Although it was originally intended to cast six panels, the central panel wasn't used, being kept as a reserve in case of unforeseen problems during construction. The following sections describe some of the many valuable lessons learnt from the construction of these panels.
The site trials were an opportunity to take the healing techniques out of the laboratory and to apply them at a larger scale in a construction environment. The M4L self-healing concrete trials achieved this primary aim as all four individual healing technologies were successfully deployed. The physical implementation was shown to be a relatively straightforward process with many positive indicators. The microcapsules were manufactured in bulk, by Lambson, and were readily mixed into the concrete on site.
The bacteria infused concrete preparation took significantly longer than expected, however the development of an automated manufacturing capacity, capable of producing a sufficient volume for commercial use should be relatively straightforward.
The crack closure forces generated by the SMP polymers are very much dependent on the shrinkage stress generated in the individual tendons. The compromise between the shrinkage stress generated and hence the number of tendons embedded into the concrete showed that this technique is feasible at this larger scale. The installation of the flow networks in these full scale panels was straightforward and demonstrated that it was feasible to repeatedly flush healing agent through the cracks in the panels.
The target characteristic cube strength of the concrete for the panels was 40 MPa and when measured at 28 days in accordance with BS EN was found to be The bacteria infused concrete was measured at This was the first time the bacteria mix had been tried in this quantity and outside the lab environment. The retention of workability of this mix was significantly less than expected, which made it extremely difficult to manufacture a reliable cube specimen after the wall had been cast.
Likewise, although previous work Giannaros et al. The reason for this was that difficulties were experienced in hand compacting the cube specimen as a consequence of casting them at the very end of the casting sequence having double handled the concrete to enable the microcapsules to be added to the mix.
This meant that the workability of the concrete used for the cubes had deteriorated significantly by the time they were being cast, which resulted in some honeycombing and the lower than anticipated strength. Similar workability issues were not experienced while placing the concrete in the panel itself and therefore it is reasonable to conclude that the strength of the panel was not compromised by the inclusion of the microcapsules.
Such a short lifetime makes the real-world application of the coatings impractical, which has been a foundational challenge in mechanical and materials sciences for about eight decades. Thicker coatings can be more durable, but they reduce heat transfer and erode the associated benefit of the coating.
Previous studies have shown that most ultrathin coatings develop tiny pinhole defects once they cure onto a surface. Steam penetrates through these defects, leading to the gradual delamination of the coating, the researchers said, so their goal was to develop a pinhole-free, water-resistant thin-film and enhance the overall energy efficiency of steam power plants by several percent.
Called dyn-PDMS, the material can be easily dip-coated onto materials in nanoscale layers on various surfaces like silicon, aluminum, copper or steel. We do not see this behavior in large, bulk samples of the material -- only in the thin-film, and that is a question we are trying to answer now.
The researchers posit that the ultrathin coatings developed in this study offer a solution for sustainable water-resistant materials and raise open scientific questions within materials science and fluid mechanics that remain unanswered. Miljkovic and Evans are both affiliated with the Materials Research Laboratory.
Miljkovic also is affiliated with electrical and computer engineering. Evans also is affiliated with the Beckman Institute for Advanced Science and Technology and chemical and biomolecular engineering.
Original written by Lois Yoksoulian. Note: Content may be edited for style and length. Science News. Ultra-thin self-healing vitrimer coatings for durable hydrophobicity. Nature Communications , ; 12 1 DOI: ScienceDaily, 16 September
0コメント