PSI - Issue 39

J.C. Ehrström et al. / Procedia Structural Integrity 39 (2022) 98–103 Author name / Structural Integrity Procedia 00 (2019) 000–000

101

4

50% higher than the 1g stress and assume a peak to peak representation per flight. The TWIST spectrum has 1 occurrence per flight at 1.6 σ 1g , see Schive et al. (1984). This factor reflects the complex interactions between tearing during overload and post-overload retardation and is therefore material dependent; the above factors are coming from studies on bulk 2024 T3. Beumler (2004) finds SF ranging from 1.17 to 1.28 in fuselage GLARE®. So, and taking into account internal data developed at Constellium for 0.8 mm sheet based FML, the SF of 1.5 is probably conservative. 4. Test results The sequence of the events of the test is given in Table 1 and Figure 3. From the initial defect, 49000 cycles are needed to reach 39 mm, a detectable crack size by visual inspection. After 82589 cycles in total, the crack has reached the top of the stringer: the upper or free flange is fully cracked. After 181744 cycles, a first residual strength test is performed, up to 257 MPa gross stress, corresponding to the maximum force of the testing machine. No significant propagation is observed. The crack propagation is then continued up to 220169 cycles, where a second residual strength ttest is performed, up to 257 MPa gross stress again. No visible crack growth is observed.

Table 1. Sequence of events during the test. Event (corresponding arrows in Fig. 3)

# elapsed cycles

Comment

Notches on the sides of a rivet hole Detectable crack in the skin First residual strength test up to 257 MPa Second residual strength test up to 257 MPa Second detection event

Initial crack 15 mm in total (blue)

0

39 mm long crack (orange)

48000 82589 181744

Broken stringer free flange (green) 98 mm long crack on stringer side and 62 mm long crack on bay side (red) 102 mm long crack on stringer side and 70 mm long crack on bay side (red)

220169

initial detectable

RS (2 nd and 1 st )

RS (1 st and 2 nd )

Fig. 3. Position of the crack tips in the panel at the different times listed in Table 1.

The inspection interval is derived from the number of cycles between a detectable crack size and a critical crack size. The latter is defined as the crack size for which the limit load can be sustained. The limit load is usually taken as equal to the 1g stress multiplied by a factor that is typically 2.5, Hunt (1999), Torenbeek (2013). Therefore, a minimum stress of 225 MPa should have been reached during the residual strength test. This stress level is exceeded by far since 257 MPa are reached with no significant growth. The inspection interval is based on the number of cycles elapsed before the last residual strength test (220169) and the number of cycles at the chosen detection event (82589). One cycle represents one flight, so that the inspection interval, assuming a safety factor of 3, is (220169 – 82589)/3 = 45860 flights. For comparison, the typical inspection interval for recent short/medium range aircraft (Airbus A220) is 15000 flights, Duprat (2019). In a safe – life approach, which is not the main purpose of the test, 220169 cycles are obtained from a very severe initial defect (2x3.5 mm water jet cut notch on both sides of an 8 mm rivet). This exceeds the 90000 flights Design