Issue 45

D. Yang et alii, Frattura ed Integrità Strutturale, 45 (2018) 45-52; DOI: 10.3221/IGF-ESIS.45.04 46 overloading and industrialized construction [1, 2]. As a result, the CFST has been extensively applied to build complex structures like industrial plants, high-rise buildings and large-span bridges. The superiority of the CFST is partially attributable to its welded joints. Compared to other types of joints, the CFST’s welded joints have a simple and neat appearance, with no protruding node, a clear force transmission line, a simple and easy-to-maintain structure. The static strength, stiffness, seismic property and fatigue performance of the welded joints are critical to the overall state of the CFST. For instance, the fatigue performance of the joints directly bears on the structural safety of bridges and offshore platforms, which are subjected to long-term alternating loads, and the service life of high- rises and other structures under long-term wind loads. Over the years, various methods have been developed to evaluate the fatigue performance of steel tube joints. Among them, hot spot stress stands out as the most popular and successful strategy for fatigue calculated of welded tube joints [3- 15]. Nevertheless, there is not yet a consensus on the effects of weld size on hot spot stress in the calculation of fatigue performance of the joints. To solve the problem, this paper explores the hot spot stress of CHS-CFSHS T-joints, which consists of circular hollow section (CHS) braces and concrete-filled square hollow section (CFSHS) chords. The T-joint is commonly used in engineering practice. It is the building block of complex 2D and 3D joints. After reviewing the previous studies and the relevant specifications on weld size, the author probes into the effects of weld size on the stress concentration factors (SCF) of CHS-CFSHS joints via finite-element analysis. L ITERATURE REVIEW o far, there is no agreement on the impacts of weld size on the SCF of welded joints. Some hold that the extrapolated SCF is not affected by weld size, because the extrapolation region is far enough from the weld toe. This view is reflected in the SCF formula for pure CHS joints in Reference [6], which does not consider the impact of weld size. Some believe that the weld size has a certain effect on the SCF of the welded joint, as evidenced by the SCF formula in Reference [3]. Below are two representative studies concerning the effect of weld size on the SCF of welded joints. Wingerde conducted a finite-element modeling of fillet weld and full penetration butt weld of different sizes, aiming to disclose the effects of weld size on the SCF of SHS joints (Fig. 1). The modeling reveals that the brace-side SCF of fillet weld was 1.4 times that of the butt weld, while the chord-side SCF of the two types of weld was almost the same. (a) Fillet weld (b) Full penetration weld Figure 1: Finite-element model of Wingerde. Through finite-element analysis, Zheng Hongzhi examined the SCF of CHS-SHS joints under the different sizes of fillet weld, partial penetration butt weld, and full penetration butt weld [16]. To obtain the brace-side SCF of the joints with full or partial penetration weld, the weld size-brace wall thickness ratio was taken as a dimensionless factor in the finite- element model, while brace-side weld size was maintained equal to the chord-side weld and subjected to four changes on four levels (Fig. 2(a)). To acquire the chord-side SCF of the joints with full or partial penetration weld, the weld size was set to the same level as that in the model of Wingerde for full penetration butt weld (Fig. 1(b)). The SCF of the joints with fillet weld was computed by the finite-element model in Fig. 2(b), where both brace-side and chord-side sizes of the fillet weld were 1.2tl. The theoretical SCFs obtained by Zheng were the upper limits of the corresponding joint. Nevertheless, the weld size-brace wall thickness ratio must be considered before computing the brace-side SCF of the joints with full or partial penetration weld. This complicates the computing process and adds to the difficulty of engineering application. S