PSI - Issue 19

Jennifer Hrabowski et al. / Procedia Structural Integrity 19 (2019) 267–274 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

272

6

from the weld toe on the chord to the weld toe at the brace, as the Mises-stress plot in Fig. 7 (left) shows. The crack initiation on the brace starts later compared to that the crack initiation on the chord without an over-welded gap. The crack then initially moves along the weld seam transition of the brace. After about 25 degrees around the brace circumference the maximum stresses in the chord occurs. At this point, the crack migrates through the weld seam into the chord, see Fig. 7 (right). As a result, the criterion "through-thickness crack " is fulfilled much later and the lifetime of the joint is increased.

Fig. 7. Stress distribution (left) and crack (right) in an CHS K-joints with over-welded gap

This shows that the fatigue strength of welded joints can be significantly increased by appropriate designs. In an earlier research project (Puthli et al. 2006) investigated the fatigue behavior of rectangular hollow sections with longitudinal stiffeners. Therein, it has been shown that the fatigue resistance of the respective samples could be considerably increased by shaping the attachment and a specific run of the weld seam. With so-called inward running welds, where the start point was chosen not in the area of the stiffness change at the begin of the attachment, but a few centimeters before, the fatigue strength of the samples could be increased by two detail classes. 3. Design recommendations The first step when planning a structure is the assessment of the state of strain of the framework. Zhao et al. (2001) give the following possibilities: • Sophisticated three-dimensional finite element modelling • Simplified structural analysis using frame analysis for triangulated trusses or lattice girders. • Rigid frame analysis for two- or three-dimensional Vierendeel type girders. When using a simplified frame analysis with pinned joints, the nominal stress range must be multiplied with a magnification factor MF, which accounts for secondary bending moments. With the knowledge that the geometric design of the joints has a significant influence on the fatigue strength, it is advisable to consider this already in the planning. The experimental fatigue test results demonstrate that with smaller wall thickness ratio  the fatigue resistance increases. That is why for fatigue design of hollow section joints not only dimension values, but also wall thickness ratio  =t i /t 0 , width ratio  =b i /b 0 as well as the chord slenderness 2  = b 0 /t 0 plays an important role and should already be considered during planning. In the research project FOSTA P1132 / CIDECT 7AB an extensive parameter study via FEA is carried out, in which also the load cases in-plane-bending (IPB) and out-of-plane bending (OPB) are considered. Table 3 gives an overview of recommended geometric parameters, which aim to minimize the SCFs at the chords, which is decisive in most cases. The maximum SCFs for RHS K-joints result for mean  -values from  = 0.50 - 0.60 for the chord and the brace. It is therefore advisable to aim for larger or smaller  -values, with large  -values being more advantageous for the SCF on the brace. Therefore, the small  -values are in parenthesis. Outliers to significantly smaller SCFs are obtained