4. Discussion
The results of studies on the regularities governing the propagation of fatigue cracks in the structural components of RRJ-type aircraft wing indicate that cracks propagated for a long period under variable amplitude loading conditions. This fact allows to organize effective monitoring for timely in-service crack detection.
At the same time, the question of how to reproduce the duration of fatigue crack growth in the RRJ-95 aircraft (No. 89051) bracket remains unresolved, since both MBM blocks (see crack growth in section No. 1) and fatigue striations (other sections) were identified simultaneously. The analysed regularity of the fatigue striation formation with different spacing under variable amplitude loading conditions did not allow to demonstrate convincingly the block of variable loads responsible for the flight of the aircraft. Thus, the number of flights in the indicated bracket was estimated based on the data of the aircraft wing bench tests using the blocks that reproduced the schematized flight loading cycle.
The program of bench tests consisted of sequence of loading blocks. Each block was equivalent to the set of variable loads acting on the aircraft wing for the full flight-cycle and simulated the entire sequence of the aircraft flight regimes and operation on the ground – taxiing, warming, etc. It was taken into account that the loading of the aircraft wing resulted in the realization of a biaxial stress state. In this regard, in-phase loading was carried out along two axes (Fig. 11). The preliminary results of the strain gauge measurement for stress-state estimation in stiffened panels of the RRJ-95 aircraft wing had shown that for all the stages of flight the in-phase biaxial loading takes place. Therefore, the bench-test program consisted of only the in-phase loading.
The resulting crack growth duration was about 50% of the wing operating time on the device at the moment of test termination. As shown by the results of the fractographic analysis, the origins of the cracks are located near the hole edges at a distance of about 1 mm similarly in the bracket of the RRJ-95 aircraft (No. 95075) wing airframe and in the wing of the aircraft (Fig. 12). This fact allows to suppose that the stress concentration is close in the compared cases.
The tensile strength of both aluminium alloys is as close as the chemical composition and microstructure of the deformable heat-treated alloy.
All of the above-mentioned facts allow to assume that at the durability near the low-cycle fatigue regime, the ratio between the crack growth period and durability under comparable conditions is also similar, i.e. about 50%.
The validity of such an analysis is consistent with the results of the RRJ-95 aircraft (No. 4862) wing previous tests, in which fragments of the lower panel were fractured during the operating time of 28180 laboratory flights. The development of the crack was accompanied by the MBM formation (Fig. 13a), and an assessment of the crack growth duration in the blocks of test flights showed that it was approximately 16000 blocks (Fig. 13b). The obtained assessment indicates the fact that the crack growth duration is about 16000/28180=57%.
Therefore, it can be assumed that for the fractured bracket of the RRJ-95 aircraft (No. 89051) the duration of crack growth at the operating time of 2345 flights is about 2345×0.5 = 1172 flights.
As a result of the study, it was found that the longest period of crack growth was in section No. 3 for about 97000 unit loading cycles. This duration is slightly less than that corresponding to the full period of crack growth because the first section was No. 1, in which the crack began to propagate. However, it could not reach the maximum length, because due to the operating other sections, the load was redistributed from section No. 1 to the remaining sections of the bracket lug.
It is worth noting that with the performed assessment of the block duration per flight, the error will not exceed 15%.
So, if we use the duration of 1172 flights and the number of striations equal to 97000, it appears that for one flight the bracket was loaded approximately 97000/1172 = 83 times. The obtained result of estimating the duration and the number of unit loading cycles per flight does not contradict the fractographic data (see, for example, Fig. 5b). On the fracture surfaces of the bracket in different sections, the extended areas with fatigue striations were revealed, which do not have clearly defined block boundaries due to the fact that the number of single cycles in the blocks reached many numbers. Moreover, the fatigue striation spacing varies irregularly due to the random loading of the bracket.
A comparison of the operating time according to the schematized flight cycle and the bracket in-service duration indicates that the loading on the device does not fully reproduce the structural component loading intensity for the flight.
Nevertheless, the duration of the crack propagation in the structural components of the RRJ-type aircraft wing is significant, which allows cracks to be reliably detected with an organized inspection frequency exceeding 500 flights. The start of monitoring can be arranged after the operating time of 1000 flights. By reducing the stress concentration on the holes in which cracks occurred, it is possible to significantly increase the duration up to the moment of crack initiation.