PSI - Issue 18

Benjamin Möller et al. / Procedia Structural Integrity 18 (2019) 556–569 Author name / Structural Integrity Procedia 00 (2019) 000–000

562

7

2.5. Fatigue test results under constant amplitude loading On the basis of optimised welding process parameters resulting from the investigations of Lahdo et al. (2016) and Möller et al. (2017), according to Fig. 4, fatigue testing under constant amplitude loading (CAL) has been performed on lap joints with a single weld seam and on LBW adapters ‘V2’ both joined by laser beam welding with beam oscillation and without or with control of the penetration depth, as well as explosive welded adapter specimens. The crack propagation and final fracture is observed by camera tracking on the side of each specimen and photographs. In the case of single welded lap joints, the test set up is shown in Fig. 5(a). The free length between the upper and lower clamping is 120 mm. For the purpose of observing the fatigue failure at the sides of the lap joint, two mirrors are used to show an actual image of both sides in addition to the top side of the weld, as shown in Fig. 5(b). The final rupture of the specimen occurs due to crack propagation through the aluminium base material and crack initiation in the transition from the fusion zone to the aluminium sheet.

(b)

(a)

Fig. 5. (a) Test set up for fatigue testing; (b) observation of the fatigue failure by camera tracking; F a = 1 kN

Load-dependent Wöhler curves are evaluated from the results of the fatigue tests according to the Maximum Likelihood Method, introduced by Spindel and Haibach (1978). In addition to the scatter T F , the Wöhler curves are defined by the load amplitude F a at cycles to failure of N f = 2ꞏ10 6 , a knee point defined by N k and F a ( N k ) as well as the slope k before (in the high cycle fatigue regime) and k * after (in the long life fatigue regime) the knee point. The slope after the knee point is set to k * = 22, corresponding to the recommendations for welded joints, made by Hobbacher (2016) and Sonsino (2007). A Wöhler curve for the failure criterion of total rupture of the specimen and a combined evaluation of the test results of laser beam welded joints without and with control of the penetration depth is shown in Fig. 6. Compared to the first test series ‘1A’, where the fracture occurred due to shearing through the weld seam (see Möller et al. (2017)), the failure of all of the following test series ‘1B’ to ‘1E’ started at the root of the weld seam and propagated through the aluminium sheet. Furthermore, the weld seams of both without control of penetration depth and controlled welded lap joints are characterised by a high averaged width-to-depth ratio of w w / d p = 2.2, which can be cited as the reason for the increased fatigue strength of F a ( N r = 2ꞏ10 6 ) = 1.03 kN compared to single welded lap joints of the previous investigations described in Möller et al. (2017). The slope of the Wöhler curve of k = 3.4 is in line with the slopes of these previous investigations, which were in the range 3.1 ≤ k ≤ 3.4. Therefore, the Wöhler curve is almost a parallel shift. The scatter T F = 1:1.14 is comparably low for welds, which justifies the combined evaluation of welded lap joints without and with control of the penetration depth.