PSI - Issue 19

Helen Bartsch et al. / Procedia Structural Integrity 19 (2019) 395–404 Helen Bartsch, Benno Hoffmeister, Markus Feldmann / Structural Integrity Procedia 00 (2019) 000 – 000

401

7

relevant details (the weld toes and roots at the inside of the section and the outside at the overhang of the end plate) plotted over the nominal stress ranges of the respective details. Additionally, the stress plot of the submodel shows stress concentrations at all four details. The most critical detail proves to be the inner weld root with a notch factor of 3.35, however the differences to other details are small. The picture of the cracked surface on the right hand side of the figure also shows cracks growing from the weld root on the left hand side of the web and cracked surface at the outer area of the flange at the weld toes.

400

weld toe outside weld toe inside weld root outside weld root inside

0 Notch stress range ∆σ ENS [MPa] 100 200 300

0

50

100

Nominal stress range ∆σ n,i [MPa]

Fig. 7. Comparison of experimentally and numerically obtained failure location in specimen B2

By comparing experimental observations to numerical results, the FE models could be verified and further utilised to investigate different influencing parameters on the fatigue strength of the different details in the endplate connection, as presented in the next section. 5. Investigation of influencing parameters on the fatigue of endplate connections with prestressed bolts With the help of the verified numerical models, the influences of end plate thickness, bolt diameter, vertical and horizontal bolt distances and flange thickness in terms of different HEA-sections have been investigated for a connection with fillet welds and overhanging end plate (Type B). As far as possible, only one parameter of the actual connection has been changed in order to eliminate multiple influencing factors. Additionally, also various standardized joints have been analyzed. In all simulations, a stress range of 100 MPa at the outer weld toe has been applied, resulting in different forces acting on the connections. Fig. 8 presents the influence of thickness of the end plate on the bolt bending stress ranges and the fatigue strength of the weld details. Ten different values for the thickness of the plate from 15 to 80 mm have been investigated. Since realistic end plate thicknesses for the investigated connection are in the range of 20 mm to 30 mm, untypical dimensions are also investigated in order to evaluate the exact influence of the thickness. The bolts in the overhang ("outer bolt") and the bolts within the section ("inner bolt") have approximately same stresses ranges. With increasing end plate thickness the deformations are reduced due to the higher stiffness of the plate. For thicknesses greater than 45 mm, the bolt bending stresses only change marginally due to outer loading. The right hand side of Fig. 8 presents the FAT class in terms of nominal reference fatigue strength Δσ c at 2 ∙ 10 6 cycles of each configuration calculated with the help of effective notch stresses (see section 2.1). By relating nominal stresses to notch stresses, the notch factor is obtained. The notch stress FAT class 225 can now be divided by this notch factor to obtain the corresponding nominal stress Δσ c . For small end plate thicknesses (15 and 20 mm), the outer weld toe is decisive due to large bending stresses in the plate. From the thickness 25 mm upwards, the inner weld root is always most critical. All calculated fatigue strengths lie above the FAT class 36* or 40 according to (Eurocode 3-1-9, 2005).