Sprayed concrete - effect of accelerator, cracks and composition on compaction, porosity, strength and water transport properties

Abstract

The research described in this PhD thesis investigated key aspects of sprayed concrete material, with an intention of gaining knowledge to enable greater use of sprayed concrete for permanent tunnel linings. The two unique attributes of sprayed concrete are the addition of set accelerator at the nozzle, which is essential to achieve adhesion to the substrate and high early strength development, and the relatively large, irregular macro pores from the spray application. Both these topics were investigated to increase understanding of sprayed concrete. The effect of cracks on water transport was also investigated. Improving sustainability of sprayed concrete linings by reducing cement content and improving durability by reducing water transport properties of the sprayed concrete material was investigated. These investigations are described in the six papers included in this thesis. The first paper was a systematic literature review covering all aspects of sprayed concrete application, including proportioning, pumping, addition of set accelerator at the nozzle, propulsion of the concrete towards the substrate and the hardening and hardened properties. Proportioning principles based on the particle-matrix model were discussed and the term placed concrete composition was introduced.The effect of set accelerator and irregularly shaped macro pores were discussed. A definition of “sprayability” linking the production to the obtained quality of the placed sprayed concrete was offered. The research in the second and third papers studies the effect of aluminium sulphate based set accelerators on the hardening and hardened properties of sprayed concrete. Investigations were on specimens produced by full scale spraying experiments (paper two) and laboratory cast specimens with the water/binder ratio kept constant (paper three). Higher doses of set accelerator were measured to give higher early strength development but reduce density, long term strength and increase suction porosity. Strength was measured to correlate well to both density and total porosity (suction porosity macro porosity). The Powers’ equation quantifying the relationship between water/cement mass ratio, degree of hydration and porosity was demonstrated to be inapplicable for sprayed concrete with set accelerator added, due to the hydration products being different to those for conventional concrete. The research in the fourth paper investigated the anisotropy of sprayed concrete due to the shape and orientation of macro pores, orientation of fibres and the presence of laminations. A trend of increasing macro porosity with increased set accelerator dose was found. Measurements of the macro pores confirmed them to be non-spherical, with 50 % of length/width ratios greater than 1.5. The orientation of both macro pores and steel fibres were measured to tend towards orientation parallel to the substrate. Three methods to measure macro porosity were compared: image analysis, PF (pore fraction) method and CT scanning. Macro porosity measurements by image analysis were higher than by PF method with only one exception among the data set. The measurements by CT scanning were closer to the measurements by image analysis. 80–94 % of the macro pores were measured to be closed by capillary suction and PF method. Furthermore, water penetration tests on specimens with the water pressure applied parallel and perpendicular to the direction of sprayed concrete application are described. Water penetration was measured to be very low parallel to the direction of application, but much higher perpendicular to the direction of application, due to macro porosity and laminations in this direction for specimens sprayed with 10 % set accelerator in a single, continuous application. The detection of laminations at higher accelerator doses and the effect on permeability demonstrated the importance of careful execution for wet spraying of concrete. The fifth paper describes experiments undertaken on sprayed concrete discs to investigate the effect of cracking on capillary suction and permeation through sprayed concrete for tunnel linings. The specimens were cracked by tensile splitting, with the crack widths controlled and measured by digital image correlation. The use of CT scanning to measure cracks through the full thickness of specimens, compared with measurements at one surface of the disc only, was also explored. Both a standard sprayed concrete mix and a mix containing an EVA (ethylene-vinyl-acetate) based co-polymer powder were tested, to determine the effect of this co-polymer on water transport. Cracks increased the rate of capillary suction compared to uncracked samples, due to rapid rise of water in the crack compared to the intact concrete and absorption occurring over the surface area of the crack in addition to the area of the base of the disc. Inclusion of the co-polymer reduced the rate of capillary suction – both the rise of water in the crack and the rate of absorption in the bulk concrete. Wider crack widths exhibited a higher rate of water permeation per area of crack. The theoretical rate of permeation in a perfectly smooth, ideal crack is proportional to the crack width cubed, but the flow rate coefficient (measured flow / theoretical viscous laminar flow) was found to also increase with crack width. This is likely due to the decreasing effect of crack surface roughness with increasing crack width, though narrower cracks are more likely to be less continuous through the thickness of the disc. Inclusion of the co-polymer reduced the rate of permeation for a given crack width compared to the standard mix. Self-healing both during water permeation and water storage of the cracked specimens was measured. The sixth paper describes research to improve the sustainability of sprayed concrete tunnel linings by reducing the cement content and improving the water transport properties of the material. Reference mixes and innovative mixes were proportioned to investigate reducing the cement content in sprayed concrete tunnel linings, by either replacement of cement with supplementary cementitious materials or by reducing the matrix content of the concrete. Shrinkage and thermal contraction were studied, with the aim of reducing tensile strains, cracking and hence water transport. The investigations described in paper 5, where inclusion of the EVA (ethylene-vinyl-acetate) based co-polymer powder in the sprayed concrete was measured to reduce water transport both in cracks and in the capillary pore network, were applicable for this study. By reducing the water transport properties of the sprayed concrete material, improved durability over single shell linings or reduced material consumption over linings including a waterproofing membrane can be achieved. While inclusion of the ethylene-vinyl-acetate (EVA) based dispersible co-polymer powder in the concrete significantly reduced water transport, it also reduced the rate of early strength development. Inclusion of the EVA based co-polymer would be beneficial for a secondary lining with reduced water transport properties, where sufficient early age strength development is needed only to achieve sufficient adhesion for the concrete to remain in place. But, depending on early age strength and thickness requirements, inclusion of the EVA based co-polymer may not be appropriate for primary linings applied for immediate ground support. The experiments described in the sixth paper also included investigations of shrinkage and the risk of early age cracking, given the significance of cracks for water transport. The measured volume changes were due to both shrinkage and thermal contraction. Sprayed concrete experiences high early temperature development, due to the addition of set accelerator to give high early strength development, that may contribute to high risk of early age cracking during the subsequent cooling. It was determined that avoiding drying shrinkage in this period should be sufficient to reduce/eliminate the risk of cracking. This could be achieved by keeping the concrete wet at early ages, use of a curing membrane or inclusion of a shrinkage reducing agent in the mix. Reducing the cement content by replacement with fly ash or limestone powder reduced the rate of early strength development due to the slower reaction rate of these supplementary cementitious materials, though they did increase long term strength and reduced suction porosity. The benefits of this are arguable – early strength is considered more critical than long term strength for sprayed concrete. On the other hand reducing the cement content by reducing the matrix volume gave promising results – the mix was successfully sprayed and the early age strength development was comparable with the reference mixes. Within the scope of the study, reduction of the matrix volume gives a more effective reduction of cement content than replacement of cement by supplementary cementitious materials.