PSI - Issue 22

V.M.G. Gomes et al. / Procedia Structural Integrity 22 (2019) 401–406 Author name / Structural Integrity Procedia 00 (2018) 000 – 000

403

3

Table 1. Chemical composition (%) and mechanical properties of the S350GD steel grade, according to EN 10346 standard ( σ Y min , minimum yield strength, σ U , ultimate tensile strength and ε R , total elongation). Chemical Composition (max %) Mechanical Properties C Si Mn P S σ Y min (MPa) σ U (MPa) ε R (%) S350GD 0.20 0.60 1.70 0.10 0.045 ≥ 350 420 16 2.2. Experimental Details An experimental testing campaign was designed and performed in order to evaluate the influence of the connection geometry, the bolts preload, the reduction of plate width (net cross section), the holes manufacturing process, and the fatigue stress R-ratio in the fatigue behaviour of double-covered bolted butt joints made of S350GD steel grade. Thus, several fatigue tests were performed with M12×30 bolts (DIN933, Grade 8.8), M12 washer (DIN 125) and M12 nuts (Grade 8). Two types of connection geometries (single (1+1) and multiple (4+4) bolts), two plate widths (80mm and 70mm), two preload levels (preloaded bolts: 70% Fu and snug tight bolts: 25%×70% Fu ), two hole manufacturing processes (punching and drilling), and three stress R-ratios (R=0.1, R=0.3, R=0.5). Table 2 presents the designation of each test series and the tested load levels. Fig. 1 illustrates the test specimens with single and multiple configurations as well as the single specimen with reduced net section. Also, in Fig.1 the nominal dimensions of the central region of the specimens are represented. Fatigue tests were performed in MTS servo-hydraulic machines with maximum loads of 100kN and 120kN each.

(b)

(a)

(c)

Fig. 1. Geometries of test fatigue specimens: (a) single connection (width = 80mm); (b) multiple Connection (width = 130mm); (c) single connection (Width = 70mm). All plate thicknesses were 2mm.

2.3. Statistical Linear Model Fatigue results are normally presented in the form of S-N curves, illustrating the relationship between the stress amplitude or stress range and the corresponding fatigue life. The S-N curves presented in this work were determined from a statistical analysis of the fatigue data obtained experimentally, in accordance with the ASTM E739 (2012) standard. The statistical treatment consists of a linear regression aiming the evaluation of the mean S-N curves, and the respective confidence intervals as well as the statistical indicators, which indicates the quality of the regression. Thus, considering a linear model, the average S-N curve may be written as: = − ∙ (1)