Behaviour of epoxy filled flat-face joints in concrete for wave energy converters

  • Noushin Khosravi

Student thesis: Doctoral ThesisDoctor of Philosophy


The need and demand for renewable energy is widely understood but there has been equally vociferous opposition to the capital and operational costs of these technologies. While wind energy has developed extensively in the offshore environment over the past decade, problems with intermittent generation means that more reliable technologies are necessary. Of these wave power is particularly attractive as it can be reliably predicted. Although wave energy convertors (WEC) have existed for decades none has achieved significant commercial adoption. Wave energy is diffuse, and this means WECs must be both large and used in arrays. Existing types have generally been constructed in steel and this has proven too expensive.

The need for cost reduction and rapid construction of WECs has stimulated interest in structural concrete. However, while marine concrete is well established, WEC will require joints and there is limited data regarding the performance and behaviour of epoxy bond lines in concrete structural elements exposed to the marine environment. In particular, there is a need to understand the macro and micro mechanical behavior of jointed systems under static and cyclic loading in seawater salt solution and under cycles of wetting and drying.

Therefore, this research was aimed at undertaking an extensive experimental programme to determine how concrete elements joined by epoxy bond lines function. The work required innovative testing rigs to be designed and constructed and novel monitoring techniques to be developed. The experimental programme was divided into three main phases.

Phase I addressed the macro scale performance of epoxy bonded concrete under shear/compression, flexure, and torsion. The failure mode and the joint strength were found to be sensitive to the experimental setup, the thickness of the joint, surface preparation, the mechanical properties of the epoxy and concrete, and the exposure to seawater. Digital Image Correlation (DIC) analysis showed that the presence of joints did not interrupt the flow of strain in specimens. Regarding epoxy bonded concrete to steel, the effects of anti-corrosion paint and applying the fresh epoxy concrete directly on the steel plates were assessed. The results indicate failure mode change and strength reduction by normal anti-corrosion paint and double shear strength by direct fresh epoxy concrete application. However, fresh epoxy concrete joint degrades faster and makes an
unfortunate better path for water and chloride penetration.

Phase II focused on the durability of epoxy bonded concrete in seawater. A prototype was made in accordance with one of the famous WECs and evaluated after two years of exposure in the marine environment. This assessment highlighted the deficiency of onsite standard monitoring methods to pick up the defects within the joints. The performance of this assembly was also dependent on the presence of the post tensioning. Furthermore, the water penetration and ion diffusion testing of laboratory samples showed that full immersion leads to higher water penetration while cycles of wetting and drying lead to higher chloride diffusion in the bond line and direct application of fresh epoxy concrete on a hard concrete can have negative durability consequences.

Phase III comprised a number of post tensioned epoxy bonded concrete samples with varying loading conditions in under cyclic wetting and drying exposure with artificial seawater solution. The focus has been on three different stress distribution zones (cracked concrete, joints, and un-cracked concrete) under no loading, sustained loading, and cyclic loading. No load and sustained load conditions showed very similar behaviour, but the damage evolved considerably in cyclically loaded samples. Reduced stiffness of the beams was observed in all cases. There was also a shift of crack propagation from the middle of the epoxy layer to the bond line under cyclic loading. Chloride diffusion and water penetration, which were originally one of the main concerns of the bonded system, did not progress at the interface of concrete and epoxy, and the diffusion rate
was not affected by the state of loading.

Overall, the results demonstrated the interconnectivity and complexity of numerous variables involved in assuring the performance of epoxy bonded concrete in the marine environment. Although the initial mechanical strength of the system may look acceptable, the problem of initial defects, residual stresses, substrate sliding, and also the hydrolysis or plasticization in presence of water should be addressed. It is also very important to consider fatigue and extreme loading conditions together with the environmental effects in designing the bonded systems. Finally, issues are identified that require further investigation to ensure acceptable performance of epoxy bonded concrete during its service life in the marine environment especially in the absence of post tensioning.
Date of Award2018
Original languageEnglish
SupervisorMoray Newlands (Supervisor) & Rod Jones (Supervisor)

Cite this