PSI - Issue 19

344 Sandro Citarelli et al. / Procedia Structural Integrity 19 (2019) 336–345 Sandro Citarelli, Markus Feldmann / Structural Integrity Procedia 00 (2019) 000 – 000 9 Fig. 12 shows the results of the statistically evaluated fatigue strength based on the nominal stress concept. One can note that with ∆ = 116 N/mm² (for P = 95%) a clear discrepancy to the current reference value ∆ = 71 N/mm² in (EN 1993-1-9, 2010) is evident. 4.2. Recommendations for different geometric weld executions Using extended parametric effective notch stress investigations the fatigue class evaluated in 4.1 could be further differentiated as a function of the weld execution in terms of weld geometry. The fatigue strengths determined at notch stress level then could be retransferred to nominal stresses according to Eq. (6) by means of notch factors, which were calculated using two-dimensional FE-models validated with three-dimensional FE-analysis, see Fig. 13. = ℎ (6)

Fig. 13. Derivation of 2D model and computation of notch factors

The numerical investigations essentially revealed the following influencing parameters on fatigue failure: • Weld type (K weld, HV weld) • Weld angle • Plate thicknesses • Weld thickness The results of this extended study are presented in Table 3 and Table 4 as recommendations for a supplementary notch detail classification for detail 2 in Table 8.10 in (EN 1993-1-9, 2010), considering the above mentioned influence parameters of weld geometry and connection type.

Table 3. Recommendation for general fatigue classification considering the weld execution – K welds Detail category Constructional detail

Description

t w ≤ 30

140 125

K weld - roots removed - weld toes blended smoothly

30 < t w ≤ 50

 ≤ 15°

50 < t w ≤ 100

100

t w ≤ 30

112 100

30 < t w ≤ 50

15° <  ≤ 45°

50 < t w ≤ 100

80