PSI - Issue 75

Florian Kalkowsky et al. / Procedia Structural Integrity 75 (2025) 581–592 Florian Kalkowsky et al. / Structural Integrity Procedia 00 (2019) 000 – 000

586

6

3.2. Fatigue tests and statistical evaluation according to Eurocode 3 The statistical evaluation of the fatigue tests was carried out for the fixed slope parameter (m = 5) of the S-N curve acc. to the EC 3 background information by ECCS (2018) with the probability of survival S = 50 % and S = 95 %. The diagrams in Fig. 2 show the test results (fractures and run outs) as well as the S-N curves with a constant slope parameter of = 5 for S = 50 % and S = 95 % for test series with at least ≥ 12 fracture results. The variable slope parameters of all test series were between = 4.2…10.8 which justifies a standardized evaluation with = 5 and at the same time enables better comparability of the results. Furthermore, this is in line with the new definition of the slope parameter for non-welded constructional details with light notch effect, which also includes the components in connections with non-preloaded bolts. The current slope parameter of = 3 according to EN 1993-1-9 (2010) can therefore be considered inapplicable. This adjustment in the 2 nd Eurocode generation is therefore welcomed by the authors. In a first step the effect of material strength on the fatigue resistance was subject of investigation. In Fig. 2 a) the S-N curves for the bearing type connections with components made of S235JR up to S700MC are shown. All samples failed with a crack initiation at approx. 90° to the load direction, see Fig. 2 d). The known correlation of increasing fatigue strength with a simultaneous increase in material strength could only be observed between test series I and II. With a fatigue strength of Δσ C,PS=50% = 124 N/mm² in series I compared to Δσ C,PS=50% = 169 N/mm² in test series II for a fixed slope parameter, the increase was significant and the fatigue strength has increased by approximately the same as the tensile strength. However, the use of the high-strength structural steel S700MC resulted in no further increase. The fatigue strength for a fixed slope parameter was only Δσ C,PS=50% = 165 N/mm² and thus below the series II with components made of S460MC. It should be emphasized at this point that the fatigue strength of test series I with a factor of 2 – 3 is significantly above the current value of Δσ C = 50 N/mm² according to Eurocode 3 or of Δσ C = 67 N/mm² for the 2 nd Eurocode generation, although only the structural steel S235JR was used in the connection. Moreover, a pronounced mean stress dependency is not to be expected for this steel grade and would not sufficiently justify this deviation. There is obviously not only a correlation between the material strength and the fatigue resistance of components in bearing type connection, but also other influences, e.g. through the stress concentration in the connection.

300 400 500

300 400 500

300 400 500

(a)

(c)

(b)

200

200

200

124 169 165

149 168

a 140 128 150 136

50 Direct stress range Ds [N/mm²] 100

50 Direct stress range Ds [N/mm²] 100

50 Direct stress range Ds [N/mm²] 100

105 | 106 131 | 132

Fracture series VIII | K t,zd = 3.16 | S235JR Run out series VIII P S = 50% m = 5 P S = 95% m = 5 Fracture series IX | K t,zd = 3.16 | S235JR P S = 50% m = 5 P S = 95% m = 5 Fracture series X | K t,zd = 3.16 | S500MC

45 67 50 70 77

Fracture series I | K t,zd = 3.19 | S235JR Run out series I P S = 50% m = 5 Fracture series II | K t,zd = 3.19 | S460MC P S = 50% m = 5 Fracture series III | K t,zd = 3.19 | S700MC P S = 50% m = 5 EN 1993-1-9 FprEN 1993-1-9

50 67

67 50

Fracture series IV | K t,zd = 9.04 | S355J2+N P S = 50% m = 5 Fracture series V | Kt ,zd = 3.19 | S355J2+N P S = 50% m = 5 P S = 95% m = 5 Fracture series VI | K t,zd = 6.83 | S355J2+N Run out series VI P S = 50% m = 5 P S = 95% m = 5 Fracture series VII | K t,zd = 2.82 | S355J2+N

35

Run out series X P S = 50% m = 5 P S = 95% m = 5 EN 1993-1-9 FprEN 1993-1-9

EN 1993-1-9 FprEN 1993-1-9

P S = 50% m = 5 P S = 95% m = 5

20

20

20

10 4

10 5 Fatigue life N [-]

10 6 2×10 6

10 7

10 4

10 5 Fatigue life N [-]

10 6 2×10 6

10 7

10 4

10 5 Fatigue life N [-]

10 6 2×10 6

10 7

Fig. 2. Fatigue diagram (a) Series I to III (b) Series IV to VII (c) Series VIII to X (d) Fracture pattern centre plates of test specimens Based on the previous observations, the geometric parameters ( 1 , 2 , 2 ) of the connection were also investigated. The results from fatigue tests in test series IV to VI are shown in Fig. 2 b). Particularly noteworthy are the test series IV and VI with clearly adapted geometric parameters in comparison to the other test series. In both test series the point of crack initiation was moved in load direction on the hole edge and the lowest fatigue strengths of