Case Study

Background information [51]

A landing gear shock-strut cylinder was subjected to fatigue tests with an applied internal pressure of 41.4 MPa (6 ksi) and an R-value of zero in the laboratory air environment. The cylinder (wall thickness 't' = 5.59 mm or 0.22 in. and inner radius = 44.5 mm or 1.75 in.) was made from the die forging of 7075-T73 aluminum alloy material. The KIc value for 7075-T73 is about 32.6 MPa m0.5. After 30,000 cycles, fracture occurred along the parting plane as shown in Figure 3. Subsequent failure analysis revealed numerous pits on the internal surface of the cylinder. Figure 4 shows a transverse section through one of those pits.

The depth of the pit was quantified at about 6 mils or 0.15 mm. Also, as shown in Figure 5 the fracture surface revealed a semi-circular fatigue crack that originated from a pit on the internal surface of the cylinder. The crack depth (a) and crack width (2c) were found to be 4.32 mm (0.17 in.) and 9.65 mm (0.38 in.) respectively [52]. Nominal hoop stress at the fracture location was calculated at about 331.2 MPa (48 ksi). The calculated hoop stress was about 68% of the parent material ultimate strength and 80% of the nominal design stress (414 MPa or 60 ksi) for the component.

As there is no data available with regard to the pit growth rate for 7075-T73, the Phase 1 life, that is, the time or cycles to nucleate the pit, from which the crack formed resulting in fracture of the cylinder, can not be estimated. In addition, fatigue crack growth rate data for 7075-T73 in a realistic environment is not available to estimate the number of fatigue cycles for fracture of the cylinder. However, fatigue crack growth rate data for 7075-T73 in a laboratory environment will be used in the estimation. Therefore, with the available information from the failure analysis, the applicability of the PCF models proposed by Hoeppner [41] and Kawai & Kasai [48] to the shock-strut cylinder is demonstrated below.

From Equation 2, the critical stress intensity factor at fracture can be calculated. Considering the measured crack depth value from the failure analysis as 'a', the calculated shoop as s and Q is assumed as 2.48, the critical stress intensity factor at fracture is found to be 26.95 MPam0.5. However, as mentioned before, the KIc value for 7075-T73 was about 32 MPam0.5. The lower KIc value at fracture could be attributed to numerous pits that were found on the surfaces of the cylinder.

Using Equation 2 from Hoeppner's PCF model, the pit-to-crack transition can be estimated. For s = shoop = 331.2 MPa, Ksf = DKscth = 0.75 MPa m0.5 (for 7075-T6 from [50]), the pit-to-crack-transition length is determined to be 0.0035 mm.

Considering the pit-to-crack-transition length (0.0035 mm) as the initial crack size and the measured crack depth from the failure analysis (4.32 mm) as the critical crack size, the number of fatigue cycles to failure once the pit is transitioned to a crack is determined as shown in Table 3. The procedure outlined in the previous section is used in estimating the number of cycles to failure. The fatigue crack growth rate data for 7075-T73 is used in determining da/dN for each calculated DK. As determined from Table 3, the estimated number of cycles to failure is 20,793. When it is compared to the actual cycles to fracture from testing, that is, 30,000 cycles, it is a reasonable estimate.

Using the model proposed by Kawai and Kasai, the allowable stress at which the shock-strut cylinder can be operated is determined using equation (1). In this equation, DKall (allowable stress intensity threshold value) is considered equal to the "long" crack threshold stress intensity value for 7075-T73, that is, 5 MPa m0.5 (from [51]). The geometric factor 'F' is assumed as 1. The quantified depth of pit (6 mils or 1.5e-4 m) on the inner surface of the cylinder from failure analysis is considered as hmax, that is, the maximum pit depth. Substituting these values in Equation 1, the estimated allowable stress at which the shock-strut cylinder can be operated is determined to be 230.3 MPa. It is about 30% lower when compared to the calculated hoop stress (331.2 MPa) at the fracture location of the shock-strut cylinder.

Review of Pitting Corrosion Fatigue Models, D.W. Hoeppner, V. Chandrasekaran, and A.M.H. Taylor, University of Utah and FASIDE International Inc.