This paper elucidates some of the controlling factors governing soil desiccation. The desiccation tests were conducted on three materials -clay, potato starch and milled quartz sand -all three featuring similar fracture energy. Two controlling factors were identified in desiccation cracking, regardless of the material. The first is the tensile stress and strain energy development within the material when the material is restrained against shrinkage. The distribution of the tensile stress will depend on the boundary conditions and material stiffness, and will dictate where cracks are likely to originate. The second factor is that the exact positions of crack initiations will be controlled by the flaws and/or pores within the material. For materials such as clay, with very fine particles, the cracking mechanism is governed by flaws, since the desaturation of fine pores would require very high suction stress, and this requirement leads to sequential cracking and orthogonal crack patterns. If the material has particles giving relatively large and uniform pore sizes with high moisture diffusivity leading to high shrinkage energy prior to cracking, then the fracture energy balance indicates that cracking can occur in near hexagonal patterns with 1208 crack initiations, which occur predominantly simultaneously. However, even for materials with lower moisture diffusivity, such as for clay, high desiccation rates can give rise to an 'effective layer' over which high suctions and strain energy develop, leading to almost simultaneous dense cracking.

No abstract

Leakage is undesirable in water distribution networks, as leaky pipes are financially costly both to water utilities and consumers. The ability to detect, locate, and quantify leaks can significantly improve the service delivered. Optical fibre sensors (OFS) have previously demonstrated their capabilities in performing real-time and continuous monitoring of pipe strength leak detection. However, the challenge remains due to the high labour cost and time-consuming process for the installation of optical fibre sensors to existing buried pipelines. The aim of this paper is to evaluate the feasibility of a submersible optical fibre-based pressure sensor that can be deployed without rigid bonding to the pipeline. This paper presents a set of experiments conducted using the proposed sensing strategy for leak detection. The calibrated optical fibre device was used to monitor the internal water pressure in a pipe with simultaneous verification from a pressure gauge. Two different pressure-based leak detection methods were explored. These leak detection methods were based on hydrostatic and pressure transient responses of the optical fibre pressure sensor. Experimental results aided in evaluating the functionality, reliability, and robustness of the submersible optical fibre pressure sensor.

No abstract

Pitting/patch corrosion is a major and common cause of leaks and (or) bursts in cast iron (CI) pipes that consist of over 50% of global pipelines. The determination of the remaining life of a CI pipe is a major challenge facing water utilities requiring an estimate of the impact of pitting corrosion on the degradation of structural pipe capacity. This paper uses an efficient finite element analysis (FEA) to model the behaviour of large diameter CI pipes with natural or simulated corrosion pits and patches. The test results were obtained through laboratory pressure testing. Laser scanning was used to develop three dimensional geometric models of pipe specimens for direct use in the numerical modelling. The CI material was simulated by a non-linear hyperbolic elastic model (termed MHM-CI) recently developed by the authors for CI pipe modelling. The numerical results showed that the proposed FEAs with the MHM-CI model are reasonably capable to predict the measured responses with increasing water pressure, importantly at the critical pipe corrosion patches, such as hoop strains, initiation of leak and burst failures. The initiation of fracture was explained by material failure purely by tension, which can form a crack that could lead to water leakage. Final burst possibility was modelled by using a simplified fracture mechanics approach to determine the critical crack length for spontaneous fast fracture as in a burst. Our numerical findings suggest that the proposed simplified numerical approach may be used to determine whether a corroded cast iron pipe would leak before break provided that the corrosion condition of the pipe and the relevant material properties are available. However, the window of time for leak before break would require further testing since this would be governed by sub critical fracture growth subject to repetitive external and internal pipe loadings.