CFST T-joints consisting of a concrete-filled circular chord and a circular hollow section brace have been used in CFST trussed arch bridges. The stress concentration factors (SCFs) of CFST T-joints have been found to be much lower than those of circular hollow section (CHS) T-joints in the existing researches. At present, no parametric formulae have been proposed for SCFs determination for fatigue design of CFST T-joints. In this study, three-dimensional finite element (FE) models of the existing experiments for CFST T-joints were developed to determine the SCFs distribution at the chord-brace intersection under axial force in the brace. After confirming the validity of the FE models by the comparison of calculated SCFs with existing experimental results, they were provided for the parametric analysis to reveal the influence of four non-dimensional parameters, i.e. diameter ratio (β), diameter to thickness ratio of chord (2γ), thickness ratio (τ) and relative chord length (α), on SCFs of CFST T-joints. In total, 212 FE models with different parameters were analyzed under tensile and compressive axial forces. Based on the results of parametric analysis, a series of parametric formulae to calculate the SCFs was proposed for CFST T-joints referring to those for CHS T-joints. The SCFs determined by the formulae showed good agreements with FE analysis results.

A B S T R A C T Using a limited set of residual stress measurements acquired by neutron diffraction and an equilibrium-based, weighted least square algorithm to reconstruct the complete residual stress tensor field from the measured residual stress data, the effect of weld residual stress on fatigue crack propagation is investigated for 2024-T351 aluminium alloy plate joined by friction stir welding. Through incorporation of the least squares, complete equilibrated residual stress field into a finite element model of the Friction Stir Weld (FSW) region, progressive crack growth along a direction perpendicular to the welding line is simulated as part of the analysis. Both the residual stress redistribution and the stress intensity factor due to the residual stress field, K res , are calculated during the crack extension process.Results show that (a) incorporation of the complete, self-equilibrated residual stress field into a finite element (FE) model of the specimen provides a robust, hybrid approach for assessing the importance of residual stress on fatigue crack propagation, (b) the calculated stress-intensity factor due to the residual stress field, K res , has the same trend as measured experimentally by the 'cut-compliance method' and (c) the da/dN results are readily explained with reference to the effect of the residual stress field on the applied stress intensity factor.

A B S T R A C T An equilibrium-based, weighted least-squares algorithm is developed to reconstruct the complete residual stress field tensor throughout the specimen volume using a limited set of residual stress measurements and a priori boundary conditions. Using a set of residual stress data acquired by neutron diffraction over a limited region of an AA2024-T351 friction-stir-welded specimen, the method is used to reconstruct the complete residual stress tensor field. Results indicate that (1) the algorithm is robust and effective in quantifying the complete residual stress field throughout the region of interest;(2) appropriate smoothing parameters are readily obtained so that the reconstructed, complete residual stress field accurately reflects the experimental measurements and (3) the resulting residual stress field is in self-equilibrium. I N T R O D U C T I O NThe origins of residual stresses, as well as the effect of residual stress on a variety of mechanical properties, are well documented. 1,2 In a recent review article, Nitschke-Pagel and Wohlfahrt 3 noted that residual stresses within weld-affected zones can promote (1) increased crack growth rate and reduced fatigue strength, (2) a reduction in fracture toughness and fatigue strength and (3) cold cracking without external load. These observations are consistent with the fact that investigators have shown that the maximum residual stresses can approach the tensile yield strength for the weld bead; 4 for friction stir welds (FSW), tensile residual stresses approaching 35% of the tensile yield stress have been measured using neutron diffraction (Ref.[5]). 1 Many investigators have shown that residual stresses affect the fatigue life of materials. 6,7 With regard to the effect of residual stresses on fatigue crack growth in weldments, the overview provided by Wohlfahrt and Lieurade 8 shows that the change in R-ratio (σ min /σ max ) introduced by welding residual stresses is directly correlated with a change in fatigue crack growth rate, suggesting that incorporation of crack closure into models may allow investigators to account for R-ratio effects. As an example, Correspondence: M. A. Sutton.

No abstract

In previous research, the authors numerically investigated 212 finite element (FE) models of concrete-filled steel tubular (CFST) T-joints under axial force in the brace to derive formulae for stress concentration factors (SCFs). The formulations involve four non-dimensional parameters: diameter ratio, β; diameter to thickness ratio of chord, 2γ; thickness ratio, τ; and relative chord length, α. In the current study, the earlier formulation is extended to include four additional loading conditions: in-plane bending (IPB) in the brace, out-of-plane bending (OPB) in the brace, axial compression in the chord, and IPB in the chord. The validity of the new SCF formulae is demonstrated by comparing the SCFs obtained using the formulae with the results of numerical analysis.