Abstract
Glass-fibre reinforced polymer (GFRP) composites provide superior corrosion resistance, enhanced strength-to-weight ratios, in addition to an improved modulus-to-weight ratio in contrast to traditional materials like stainless steel or wood. They are often used in the automotive, wind energy, and energy (e.g., petroleum, natural gas) industry sectors. To develop an effective durability assessment methodology, understanding synergistic effects is necessary. This review paper examines the condition of GFRP materials today, with an emphasis on their longevity and highlighting the importance that a synergistic approach to material assessment may have in a way that improves our knowledge of material durability and functional lifetime. Specific degradation stressors discussed include water absorption, temperature, ultraviolet (UV) radiation, creep, and relaxation.
The literature shows strong agreement that moisture, temperature, and sustained loading progressively reduce the glass transition temperature (Tg), stiffness, strength, and interfacial integrity in GFRP composites, with interlaminar shear strength (ILSS) and creep particularly sensitive to ageing. Key mechanisms remain disputed, including the degree of non-Fickian diffusion, the roles of matrix hydrolysis versus interfacial degradation, and the legitimacy of the time–temperature superposition principle (TTSP) in submerged environments. Major gaps persist, including limited long-term field data, a poor understanding of multi-stressor interactions, an inconsistent reporting of degradation kinetics, and a lack of standardised ageing and lifetime-prediction protocols. Emerging machine-learning (ML) methods show promise but require larger, standardised, multi-stressor datasets and closer integration with physics-based models.
To address these limitations, the paper recommends: (i) the development of a harmonised, multi‑stressor durability testing methodology; (ii) an expansion of long‑duration marine field trials to validate accelerated ageing assumptions; (iii) the creation of shared, standardised datasets capturing environmental, mechanical, and chemical interactions; and (iv) the integration of physics‑informed models with machine‑learning approaches to improve material lifetime prediction.