· · · Aerospace Engineering reports

Paper presented at the 16th ICAP Conference, Brussels, 14 and 15 May 1979.

The current status of flight simulation fatigue crack growth concepts is reviewed. Crack growth models and methods of crack growth prediction are compared with respect to applicability and limitations. Research areas for improvement of crack growth prediction are specified.

Results are presented from crack propagation tests on aliuninium alloy sheet specimens tested by flight-simulation loading. Five design stress levels were adopted. The usefulness of the stress intensity factor for correlating the data was evaluated. The discussion includes various approaches for predicting crack rates unaer service loading, damage calculations and the significance of flight-simulation testing procedures. The paper is summarized in some more details in the conclusions

The present review consists of three contributions, prepared by (l) The National Aerospace Laboratory NLR, (2) The Aerospace Department of the Delft University of Technology and (3) The Fokker Aircraft Industries. A detailed list of subjects is given in the contents of the review.

Full-scale fatigue tests were previously carried out on wing tension skins under random flight loading, program loading and constant-amplitude loading. Tension skins left from these tests were partly stripped. Tensile test revealed fatigue crack nuclei at various locations. Results have some meaning with respect to the frequency of occurrence of cracks that had not been detected previously, the possibility of indicating part-through cracks by X-raying, the influence of a tensile stress on the sensitivity of X-ray inspections and the occurrence of certain cracks as dependent of the maximum load in a full-scale fatigue test.

This paper deals with crack propagation under programme-fatigue loading; the programme being derived from a gust spectrum. The validity of the Palmgren-Miner rule for crack propagation has been studied and its usefulness for design purposes is discussed.

Constant-amplitude tests and programme tests were carried out on 2024-T3 lug specimens with standard holes, slotted holes and expanded holes. The results give information on (1) programme loading versus constant-amplitude loading, (2) cumulative damage and (3) the improvement of the fatigue endurance obtained by the application of slotted and of expanded holes.

Constant-amplitude tests and simplified flight-simulation tests with 10 or 50 gust cycles per flight were performed. The effect of the ground-to-air cycle is compared with the Palmgren-Miner prediction. Optical and electron microscope observations of the fracture are presented.

The size effect on crack propagation was studied with sheet specimens of 2024-T3 Alclad. Values of the width were 80, 160, 300 and 600 mm, sheet thickness 3 mm, mean stress 8 kg/mm2 and stress amplitudes 2.5, 4 and 6.5 kg/mm2. Results are compared with the theory of the stress intensity factor and the Weibull theory. Practical implications of the results are indicated

Sheet specimens of 2024-T3 Alclad and 7075-T6 Clad were tested in a test hall and outside this hall. Each specimen contained five cracks. The crack growth data allow comparisons to be made (I) between indoors and outdoors (2) between random and program loading (3) between these loadings with and without ground-to-air cycles and (4) between the two alloys. Damage calculations are made.

Fatigue tests have been carried out on light alloy lugs of 2024 and 7075 material in order to investigate the effects of some geometrical parameters (lug size, geometry of lug) and various means to reduce fretting, viz. lubrication of the pin, flattened pins, slotted holes and expanded holes. Complete S-N curves have been determined at a positive mean stress. A discussion of the results includes a comparison with data from the literature. Special attention is focussed on fretting corrosion in the hole of the lug.

In a recent test program, fatigue crack propagation in 2024-T3 and 7075-T6 sheet material was studied at three test frequencies, viz.10, 1 and 0.1 cycles per second. The flight-simulation loading was based on a gust spectrum. The design stress level was adopted as a second variable of the investigation. Differences between the crack propagation rates at the three test frequencies were small and unsystematic. The propagation was much slower than predicted from constant-amplitude test data. Moreover, the macro-cracking behaviour appeared to be different. In the discussion attention is paid to interaction with environmental effects and to implications for practical problems of aircraft fatigue. This paper was prepared for presentation at the AGARD Symposium on Random Load Fatigue, Lyngby, Denmark, 13th April 1972.

The following text is given in the format required for presentation at the Conference and Exhibition "Fatigue Testing and Design", to be held in the City University, London, 5-9 April 1976. The report gives a comparison between the fatigue performances of several sheet and forging materials under gust spectrum loading representing the lower wing surfaces of commercial transport aircraft. The materials are 2024-T3 and 7075-T6 aluminum alloy sheet, annealed Ti-6A1-4V sheet, and 705O and AZ 74»61 aluminum alloy forgings. The significance of the results for the design of panel concepts and forged fittings is discussed, and proposals for further testing are made.

Tests were carried out on center section wing panels of 7075 material. The loading was based on strain gage records obtained by flying in turbulent air. Load sequences applied include random loading and programmed loading, with and without ground-to-air cycles. A testing machine was developed for this purpose. Data are presented on critical components, fatigue life, scatter and inspection techniques. Tests with the same load sequences were also carried out on long sheet specimens with five central notches as crack starters. These tests were carried out partly indoors and partly outdoors to study the environmental effect for 2024-T3 Alclad and 7075-T6 Clad sheet material. Test results are analysed and compared with data from the literature. The equivalence of random loading and program loading is discussed.