Numerical and experimental reassessment of Eurocode 3 design rules for cellular beam end-posts

This paper investigates the end-post behaviour of cellular beams focusing on the failure mechanisms subject to shear adjacent to the connections. Three cellular beams were tested to failure in a companion paper and their behaviours were modelled and validated herein using Abaqus. The cellular beams were connected to the columns through a bolted end-plate at one end and a bolted fin-plate at the other end. Finite element (FE) analyses were carried out for end-posts with and without notches, and web openings with half infill plates for cases where the minimum end-post width requirements were not satisfied. It was found that the FE results were in good agreement with the test results while both gave significantly higher failure loads than the design predictions provided by BS EN 1993-1-13. A parametric study was carried out on each connection side and in total 160 end-post models were examined. It was concluded that the connection type not only affected the end-post shear resistance but also influenced the failure mode. It was shown that using a half infill plate at the first opening can greatly increase the end-post's shear resistance. • A comprehensive FE parametric study of 160 cellular beam end-posts, validated against an experimental study of 6 tests. • The end-plate connection made the end-post stronger than the fin-plate, guiding the failure to the first web opening. • The web opening diameter is crucial to the end-post shear resistance. • Half infill plates welded to the web openings next to the end-posts failed by local buckling due to their strut action. • The combination of a narrow end-post and notched flanges caused failure by buckling of the end-post, near the notch corner.