Abstract
56 4. SUMMARY and CONCLUSIONS A literature survey on fatigue crack closure and its implications on mean stress effect, have been conducted. Mechanisms of the closure phenomena are discussed, revealing the dominant role of plasticity, roughness and crack filling induced closure at different Kmax levels. Methods and related problems about the measurement of closure level are discussed. Various sources are compared and no standard method was found, but some was reported to have higher accuracy. This give rise to difficulties in verifying closure measurements. Fatigue crack growth tests were conducted under constant amplitude loading, on M(T) and C(T) specimens made from five mm thick 7075-T6 Alclad aluminum alloy. Load ratio of the constant amplitude loading was varied between Rc= - 0.60 to 0.80. Data base extracted from experimentation, was processed numerically to set crack growth equations. An estimation of fatigue crack closure level was done, by adopting calibration with high load ratio method, to the current data set, using various selections of the cut-off ratio Rc. On the basis of the experimental results and subsequent discussion, the following conclusions may be drawn. (1) Fatigue crack growth behaviour of the aluminum alloy under various load ratios, share the same mechanism, except for R=0.80 which possibly undergoes static modes of crack growth, with tears due to high Kmax level. This is also pointed out with optical microphotographs of the crack surface patterns. (2) Negative load ratios do not have remarkable effect on the growth behaviour, for the load levels employed in this study. Thus, the stress intensity factor range may be used with positive portion of it, as well. (3) For positive load ratios, mean stress effect clearly exists, accelerating crack growth with higher R. (4) Fatigue crack growth data may be represented by tiie Paris law, with mean exponent w and V as a function of the load ratio. Crack growm data for R=0.60 and 0.70 are similar, pointing out that the cut-off ratio is realized; in other words, stress intensity factor variation is in effective values.57 (5) By selecting a value for the cut-off ratio, fatigue crack closure functions U(R) and y(R) may be derived for the employed load ratio range. (6) Crack closure functions U(R) and y(R) increases with load ratio R. Selection of higher cut-off ratios results in an increase with the closure function determination, however the differences are small. (7) For the negative portion of the load ratio range, closure functions collect data towards the positive portion of the loading range, rather than accounting for closure. Crack growth rate for negative load ratios is unaffected by the mean stress. (8) Fatigue crack growth data are in quite good correlation with effective portion of the stress intensity variation. For R= - 0.20, appreciable scatter exists. Behaviour of the fatigue crack is not affected by the selection of the cut-off ratio, for a range of Rc=0.60 to 0.80. From an engineering point of view, results promise applicability, as the closure level is estimated by simple instrumentation and experimental procedure, in the expense of test material and time consumption. Comparing the closure functions derived from crack growth data with closure measurements may lead information of higher reliability. However it will not be realistic to talk about any certain verification, as there exists no standard for the measurement of closure. (9) Investigation of the crack surfaces with optical microscope, gives some information about crack growth mechanisms. Scratches observed for various load ratios remark the touch of crack faces during cycling. Existence of these scratches even at positive load ratios, indicate fatigue crack closure. (10) Further improvement may be achieved by closure measurements during fatigue crack growth at various load ratios and by examining crack surfaces with SEM technique which may reveal the whole fractographic structure of the metal. Load ratio range also, may be extended to lower values to reveal a possible effect of negative stress magnitude.
