A previous study, looking at the creep-rupture behaviour of mixed reinforcement GRP when immersed in water, had discovered that low loads, behaviour became temperature sensitive. Since the recorded time to failure of a sample was reduced at elevated temperatures, from that predicted by a linear extrapolation of the short term creep-rupture results, this deviation caused problems in the accurate prediction of long-term design stresses. In order to improve the accuracy of long term design predictions, it was decided to study the mechanisms of creep in GRP that initiates time dependent failure. From this, it was hoped that accurate design criteria suitable for predicting GRP response over a 30 year design life from short term creep tests, could be developed. This thesis reports the results obtained from such a study. A series of creep tests were performed on mixed reinforcement GRP samples at several stress levels, both in air, and in room temperature distilled water, using a microcomputer based data collection system. In conjunction with this work, damage development in samples, due to combinations of water uptake and creep loading, was followed, using both scanning electron, and optical, microscopy. Moisture uptake measurements were undertaken under a series of load/temperature regimes, and fibre/matrix debonding followed using photographic techniques. In this way, water absorption, both in terms of uptake rate, and location within a sample, could be characterised. Tensile tests were also performed to determine the standard mechanical properties of the mixed reinforcement GRP used. It was found that a critical damage state was created at loads in excess of 50% of ultimate, but not below. This took the form of between 2 and 8 neighbouring filament breaks in the longitudinal woven rovings at weave crossover points, producing microcracks in the reinforcement. The creation of this multifilament fracture damage during primary creep, was considered to be necessary for time dependent failure to occur in air. Secondary greep strain was found to increase in discrete steps, both in air and water. This was attributed to the formation of transverse grasks in the longitudinal woven rovings, propogating from the above critical damage. In water, diffusion was found to be non-Fickian. Moisture uptake increased with increases in applied load and temperature. Water was seen to accumulate at weave cross-over points when immersed under load. This led to stress-enhanced fibre corrosion in these regions, weakening the reinforcement, and reducing the failure time from that expected at the same load level in air. The localised nature of moisture degradation was thought to result in the formation of critical fibre damage at loads below 50% of ultimate, when immersed in water. Two design criteria based on the observed creep mechanisms, have been developed for GRP that predict response when loaded in either air, or water. Both predict the existence of creep-rupture limits at low loads.
|Date of Award||1985|