PSI - Issue 75

J. Havia et al. / Procedia Structural Integrity 75 (2025) 43–52 Havia et al./ Structural Integrity Procedia (2025)

50

8

Table 5. Structural and effective notch stress concentrations for studied cases solved from FEA and using Eq. (1). k s,m k s,b k f,m k f,b

Case1 1.05 0.26 2.07 2.54 Case2 1.05 0.26 2.07 2.84 Case3 1.04 0.24 2.02 3.04

The results obtained using the ENS concept and 4R method are presented in Fig. 8. The ENS concept with r ref = 1.0 mm seems to provide conservative assessments for the test results. Some scatter exist, but the datapoints are significantly above the characteristic design curve with fixed slope parameter m = 3. Including local cyclic behavior and residual stresses within the 4R method, reduced the scatter of the test results. The mean S - N curve of 4R method Havia et al. (2025) is slightly lower than that of obtained for the data points from his study. Moreover, mean fatigue curve based on large dataset was 119 MPa m = 5.7 which close to characteristic value based on the finding of this study. The difference could be explained by more detailed analysis withing this study in comparison to assumptions used to conduct data analysis.

a

Δσ c 4R combined: 50% 133 MPa m = 5.7 97.7% 116 m = 5.7 T σ = 1.18

500

500

b

Δσ c ENS combined: 50% 108 MPa m = 3 97.7% 81 MPa m = 3 T σ = 1.42

Two points

4R stress range [MPa]

ENS stress range [MPa]

DED on Base plate DED welded on base plate DED on top DED surface

DED on Base plate DED welded on base plate DED on top DED surface

50

50

20000

200000

2000000

20000

200000

2000000

Fatigue life [cycles]

Fatigue life [cycles]

Fig. 8. Fatigue test results (a) effective notch stress and (b) 4R reference stresses.

All S-N curves evaluated with nominal, ENS and multiparametric 4R method are shown in Table 6. The fatigue strength assessment considering local properties and residual stresses with 4R method was found to be able to parametrically evaluate the factors influencing on fatigue strength. The effect of residual stresses on fatigue performance is considered via R local . Table 6. S-N curves of each case and combined data, with the nominal stress, ENS and 4R methods. Nominal stress method ENS method 4R method Case n m Δ σ c ,50% Δ σ c ,97.7% T σ m Δ σ c ,50% Δ σ c ,97.7% T σ m Δ σ c ,50% Δ σ c ,97.7% T σ FD1 5 3 32.3 23.2 1.43 3 107.4 79.0 1.39 5.7 130.5 114.7 1.15 FD2 6 3 29.9 24 1.27 3 99.7 76.8 1.34 5.7 130.9 113.1 1.18 FD3 4 3 30.5 22.1 1.40 3 120.3 86.4 1.41 5.7 139 115.5 1.21 All 15 3 30.8 24.2 1.35 3 107.5 80.7 1.43 5.7 132.9 116.1 1.18 In this study the additional reinforcement and smooth transition to the plate was not studied. However, the optimization of the initial layers of the DED feature under fatigue loading may enhance the competitiveness of DED hybrid structures compared to traditionally fabricated structural joints. Ideally, during the design stage, stress levels in fatigue-critical areas can be adjusted, and welding-induced residual stresses may be significantly influenced by modifying the geometry and stiffness of the DED feature.