Stress concentration factors for truss-end hollow structural section connections




Ziaeinejad, Ali

Journal Title

Journal ISSN

Volume Title



Design of end connections in a welded hollow structural section (HSS) truss has been a practical problem encountered by structural engineers since current design standards do not include definitive recommendations. For fatigue design, formulae in current design guidelines [for calculation of stress concentration factors (SCFs)] cater to: (i) unreinforced connections, with (ii) sufficient chord continuity beyond the connection on both sides. Existing formulae in CIDECT Design Guide 8 (DG8) for regular connections were shown to be inaccurate for calculation of SCFs for truss-end HSS T- and X-connections. In Chapter 1, a numerical finite element (FE) investigation to determine SCFs for circular hollow section (CHS)-to-CHS X-connections near an open chord end under branch axial loading is presented. Previous large-scale experiments were used to validate FE models, and a parametric study was performed. The parametric study consists of 240 models with variations in chord slenderness (2γ), branch-to-chord diameter ratio (β), branch-to-chord thickness ratio (τ), and chord end distance (e) on one side of the of the connection. Extrapolating existing formulae to predict “end-distance effects” on SCFs at these locations in CHS-to-CHS X-connections, from CIDECT Design Guide 8, were shown to be inaccurate. In Chapter 2, by reinforcing the previous 240 CHS models using a chord-end cap plate and repeating the numerical FE analysis, the SCFs for CHS-to-CHS X-connections near an open chord end was examined. Existing SCF formulae in CIDECT Design Guide 8 were shown to be inaccurate if applied to cap plate-reinforced end connections. Chapter 3 and 4 is a continuation of development of design formulae for truss-end rectangular hollow section (RHS)-to-RHS and CHS-to-CHS connections under in plane bending loading, respectively. Upon validation using the experimental data from previous chapters, a comprehensive parametric study was performed. For this purpose, a total of 448 finite element RHS models including 336 stepped T-connections and 112 matched T-connections (Chapter 3), and 1120 finite element CHS models including 896 stepped T- and X-connections and 224 matched T- and X-connections were created (Chapter 4). Similar to the previous chapters, it was shown that existing formulae in CIDECT Design Guide 8 were unsuitable for the calculation of SCFs for truss-end RHS-to-RHS and CHS-to-CHS moment connections. Finally, SCF correction coefficients (ψ) and parametric formulae to estimate ψ were derived for each chapters.



End-distance effects, Stress concentration factors, Fatigue design, Cap plates, Circular hollow sections, Rectangular hollow sections, moment connections