On moisture induced ageing and life assessment for glass fibre composites

Sammanfattning: Polymer composites have low specific weight and high specific strength and stiffness. They are therefore a competitive material when designing mechanical applications. The mechanical advantages and the fact that they also resist corrosion well are reasons why polymer composites have a long history of use in marine vessels, piping, corrosion equipment, and underground storage tanks. Industrial experience show that they can be designed to have long service lives in contact with moisture and aqueous solutions. Moisture diffuses into all organic polymers, including polymer composites. Absorbed water or solutions can lead to changes in the thermomechanical, physical and chemical characteristics of the material. For a composite most effects of moisture and solutions are related to degradation of strength. The changes in modulus are in most cases small. Based on bad experiences with stress rupture in the early days of composite applications, industry now use factors of safety between 4 to 10 for boats and storage tanks. Since these factors seldom are chosen based on tests and modelling of the actual in-service conditions the designs commonly are very conservative, but in some cases they may also be too weak which can give catastrophic consequences. Paper A presents an attempt for service life assessment of a composite spring subject to multiple mechanical and environmental loads. Based on preliminary tests the polymer composite solution promised the best performance for the device, but to select that solution it first had to be made probable that the composite spring would survive the design life of 10 years. Due to the time-frame of the overall development project the life assessment had to be made in very short time. The spring is pre-stressed and the main functional requirement is that the reaction load may not decline by more than 10% over 10 years. A rapid, but possibly unconservative, methodology was used for the life assessment. To account for degradation of the material due to the environmental, static fatigue and cyclic fatigue, we defined partial material safety coefficients. These coefficients are then used to derive design allowables that should be low enough to ensure that macroscopic damage do not occur during the design life of the component. Linear visco-elastic material properties are then determined by short-term testing of the neat resin and micromechanics. The visco-elastic stress relaxation for the spring is then predicted by FE-analysis. Both relaxation over the design life and relaxation at accelerated conditions are predicted. To validate the analysis the predicted relaxation at accelerated conditions is compared to results from short-term component tests. Paper B presents experiments to study the effect of the reinforcement structure and manufacturing method on the sorption kinetics and ageing of E- glass/vinyl ester composites. The results show that the microstructure, in particular the void content, has a major effect on the equilibrium value for moisture as well as on the diffusion coefficient and ageing characteristics. The results imply that the effect of the voids can be modeled by rather simple means by accounting for the possibility for humidity to condense in the voids. How suitable the micromechanical methods are to determine service live for composite materials are studied in Paper C. Paper C presents a micromechanical approach for accelerated testing based on the single fibre fragmentation and bundle tests which we believe will overcome several of the shortcomings of today's methods. Conditioning of very thin polymer specimens with single fibres embedded will shorten the timescale for diffusion from years to days to enable isolated studies of ageing. The micromechanical tests will enable us to monitor both fibre degradation and interface degradation separately.