PSI - Issue 43

Matthias Oberreiterr et al. / Procedia Structural Integrity 43 (2023) 240–245 Matthias Oberreiter/ Structural Integrity Procedia 00 (2022) 000 – 000

243

4

as proposed by Leitner (2001). Vickers hardness tests were conducted for the application of Murakami’s approach from samples in all positions, resulting a mean value of HV 1 =120 for ‘P osition A ’ , HV 1 =100 for ‘P osition B ’ and HV 1 =104 for ‘P osition C ’ . Fracture mechanical test results of alloy EN AC 46200 are given in detail in Oberreiter et al. (2021), for samples of alloy EN AC 42100 the mean long-crack threshold value ℎ, , =−1 = 6.02 √ was additionally determined within this study. 3.2. Fractography After fatigue testing, extensive fracture mechanical analyses using a digital optical microscope (DOM) and scanning electron microscope (SEM) were conducted for all tested specimens. Several failure mechanisms were identified, see Fig. 3 representing characteristic failures. ‘ Position A ’ exhibited predominantly porosity induced failure, see Fig. 3a. In some cases, slip-band surfaces occurred at high peak test loads within the fine microstructure, see Fig. 3c.

Fig. 3 Failure mechanisms at representative fatigue tested samples, analysed by DOM and SEM

‘ Position B ’ features a quite coarse microstructure, see Fig. 1b, leading only to porosity induced failure and a decreased fatigue strength level. Whereat ‘P osition A ’ and ‘Position B ’ cover EN AC 46200-T6 with fine and coarse microstructure, ‘P osition C ’ represents alloy EN AC 42100-T6 with fine dendrite arm spacing. Within this position, other failure mechanisms occurred beside porosity induced failure. Quite large oxide films were found at crack initiation site at about sixty percent of the tested specimens. Fig. 3d represents an oxide film, or bifilm, induced fatigue failure as well as the pronounced oxygen-content by EDX-mapping in Fig.3e within the bifilm section. In many cases, interactions between different failure mechanisms (mainly bifilm, porosity and slip band) occurred, exemplariliy depicted in Fig. 3f illustrating the interaction between slip-band (red-solid line) and oxide film (yellow dashed line). Finally, all fracure initiating defects were evaluated in regard to defect interaction according to the The modified fatigue assessment methodology of Tiryakioğlu (2008), see Eq. 1 and 2, is applied to assess the fatigue strength of the investigated samples using the CT defect distribution results. Furthermore, a modified methodology (Noguchi et al. (2007)) of Murakami ’s original model was applied for the aluminum alloys, see Eq. 3. σ LLF = C 1 HV+C 2 ⋅ E E A St l √area 1/6 (3) The statistical assessment methodology of Tiryakioğlu was applied using the parameters λ and δ of the Gumbel distribution, the long crack threshold value ℎ, , =−1 and the parameters offset B and slope m , which were determined using Eq. 1 by least-squares error minimization, see Fig. 4. This methodology matches the experimental data of the alloy EN AC 46200 best with deviations Δ between 5.6% and -1.9%, see Tab. 3 and Fig.4. The estimation of the fatigue strength for the alloy EN AC 42100 by Eq. 2 leads to a sound conservative correlation by 4.6% if only pores and inclusions (P+I) are considered. If oxide films are included in the defect distribution (P+I+B), this methodology leads to a slightly increased conservative fatigue design of 17.8%. proposal of Åman et al. (2017). 3.3. Fatigue assessment model