ELSEVIER zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA PII: 81359-835x(%)00131-5 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO Composites Part A zyxwvutsrqponmlkjihgfedcb 28A (1997) 327-331 c 1997 Elsevier Science Limited Printed in Great Britain. All rights reserved 1359-835X/97/$17.00 Fatigue of carbon-fibre-reinforced plastics under block-loading conditions B. Harris*, N. Gathercole, H. Reiter and T. Adam School of Materials Science, University of Bath, Bath BA2 7AY, UK (Received 77 June 1996; revised 79 September 1996) Following an earlier study of block-loading fatigue on a high-performance T800/5245 carbon-fibre composite laminate of [(* 45,0z)2]s construction, further experiments of a similar nature have now been carried out on other CFRP laminates with the same lay-up but with different fibre/resin characteristics. In this more recent work, comparable T800/924 and IM7/977 materials and a lower-performance HTA/913 laminate appear to behave in a similar manner to that previously found for the T800/5245 composite. The most significant common feature of the behaviour is that combinations of tension and compression loading are more damaging than purely tensile or purely compressive stress cycling rCgimes. 0 1997Elsevier Science Limited. All rights reserved. (Keywords: carbon-fibre-reinforced plastics; fatigue; block loading) INTRODUCTION The majority of studies of the fatigue of high-performance resin-based composites have been concerned with inves- tigating the physical mechanisms of degradation during cyclic loading, and establishing the effects of material variables or environmental variables on the basic fatigue response, mainly under conditions of constant stress amplitude. There have been several studies of the accumulation of damage in a variety of composites, and arising out of these a number of models have been proposed that permit fatigue life predictions to be made, with varying degrees of success. For variable-amplitude cycling, however, the situa- tion is less satisfactory. Attempts to predict behaviour following simple or complex block-loading patterns have often given non-conservative results’ and have estab- lished that linear cumulative damage rules are inap- propriate in most circumstances2A. In particular, Schiitz and Gerharz observed that for certain types of laminate the Miner linear-damage rule’ predicted lives up to three times those actually measured. They suggest that this could be due to the very detrimental damage contributed by low-load cycles in the compression region, which is not accounted for adequately in a Miner estimate. We have recently published the results of a pro- gramme of block-loading fatigue experiments carried out on a T800/5245 carbon-fibre-reinforced plastic (CFRP) laminate6. This is a modern high-performance aerospace *To whom correspondence should be addressed composite consisting of a high-strain fibre in a toughened resin matrix. Two types of experiment were discussed in the paper. In the first type, a range of four-unit blocks containing different combinations of tension and com- pression stresses were applied and the effects of sequence on the Miner sum were examined with reference to the established constant-stress conditions. In the second group, an attempt was made to break down the fully reversed stress cycling condition, characterized by a value of the stress ratio R(= g,in/amax) of - 1, into separate tension and compression blocks, and the effect of applying each of these first was examined. When the results for the four-unit block-loading experiments were analysed in terms of the Miner number it was found that in sequences of all-tension blocks, a linear-damage law was reasonably satisfactory, but in all-compression and mixed sequences the linear law overestimated the actual composite life. Since this earlier work related only to a single CFRP composite, we have repeated some of the experiments on three other carbon/epoxy systems, one of which was a lower- performance material based on HTA fibres, and the other two alternative modern high-performance materials, T800/924 and IM7/977. MATERIALS AND TESTING PROCEDURES Experimental materials Details of the materials investigated, together with relevant information for the older T800/5245 discussed 327