Fully overlapped rolled hollow section welded joints in trusses

• Main Author: Philiastides, Antony
• Published: University of Nottingham 1988
• Subjects: 621.88; TA 630 Structural engineering (General)
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.381062

The designer of lattice trusses has been traditionally encouraged to avoid extra joint bending moments by ensuring 'noding' of member centre lines. This can however cause problems in the design of RHS trusses with small (economical) branch members by causing large gaps at joints with undesirable flexibility for slender chord walls. 100% overlap joints overcome this problem while still maintaining economic single cut branch ends. The research programme set out first to highlight the difference in behaviour of trusses with large gap noding and completely overlapping joints. Two similar trusses - one gap, the other lap - with matched sections were tested to failure. It was concluded that the gap joint truss (branch/chord width ratio = 0.4) was much less efficient than the corresponding 100% overlap truss despite the large eccentricities. The collapse load of the latter was some 35% greater, while the stiffness properties were better, and remained linear for a substantial proportion of the loading. On the other hand the gap joint truss soon became non-linear, with large overall deflections. Local connecting chord wall deflections were quite small in the lap joints while appreciable deflections occurred at gap joints under service loads. Elastic frame analyses were carried out for all the six test trusses (one gap and five lap). For the overlap trusses, axial forces and bending moments could be predicted fairly accurately but a non-linear analysis was required for the gap jointed truss even at fairly modest loads. The effects of ß ratio, chord slenderness and branch angle were all examined within the parameter range tested. The advantages of truss continuity moments as well as plastic redistribution of moments have been observed to reduce the occurrence of the local chord buckling mode of failure (L7), compared with previous isolated joint tests. Results obtained from tests on isolated joints can give good agreement with those obtained from truss tests, both with respect to strength and failure mode. However, as the isolated joint testing cannot always reproduce the support conditions in a truss, the failure modes (and hence strengths) can differ. The current CIDECT design strength equations and recommendations for gap and overlap joints are largely based on the results of isolated joint testing. The suitability of the CIDECT strength equations and recommendations for designing RHS lattice trusses has been reviewed. Consequently, for the 100% overlap joint trusses a simple design method has been presented in conjunction with practical design recommendations. The problems associated with the analysis and design of the gap joint truss are described in detail.