In this thesis an investigation into the buckling behaviour of steel trusses with eccentric joints is performed.
In recent years, Ampelmann motion compensated gangway systems have been pivotal to the advancement of offshore personnel transfers. These systems for functionality reasons require the use of eccentric joints. The behaviour of such joints is not a well-studied topic, and little is known of their effect on the buckling behaviour of structures. Buckling of trusses using conventional centric (non-eccentric) joints has been examined extensively in the literature. This work aims to answer what the effects of using eccentric joints compared to centric are, on the stability of steel truss structures. Additionally, a reduction of the computational and design time is attempted, by extending a simplified approach developed for centric joints, to eccentric joints.
The main investigation is performed on a simply supported truss comprised from eccentric K-joints and square hollow sections (SHS). Detailed models are used, with the joints being modelled by shell elements. The results are compared to equivalent trusses with centric joints. The main conclusions are that due to the eccentric joints, a better buckling behaviour can be realised for the truss members in the case of in-plane buckling. This improvement is more evident for the buckling of the braces and for combinations of chords and braces with γ>8 and 0.30<β<0.70 respectively. In those cases, a decrease of the buckling length of at least 5% is observed and reaching nearly 20% for γ=15.9 and β=0.4. In the case of out of plane buckling, the observed behaviour is slightly worse. The difference is considered negligible in practice, with all configurations checked giving a difference of less than 2% between the buckling lengths.
The simplified design approach, which uses beam elements and rotational springs, is validated, by comparing literature results with detailed shell element models. It is, also, further improved by performing new research on individual centric K-joint models. Having validated the approach for centric joints, three different modifications are proposed to extend it to eccentric joints. None of the proposed modifications give the validation required but insight is gained regarding the behaviour of eccentric K-joint trusses. It is concluded that eccentric joints require non-symmetric stiffness matrices for their behaviour to be captured.