Span length effect on alternate load path capacity of welded unreinforced flange-bolted web connections

It is intuitive that for a building of set dimensions, a shorter span length will increase the ultimate load carrying capacity, subjected to an arbitrary column loss. However, the capability of beam-to-column connections to develop and maintain the catenary action of the beams in this situation is not considered directly in the design stage. Therefore, to determine the optimum span length, which provides the required load carrying capacity and cost-effectiveness, sensitivity analyses are needed. In order to save the computational efforts, a component-based model for improved Welded Unreinforced Flange-Bolted web (WUF-B) connections is developed in this study. The capability of this model for predicting the failure and ultimate load carrying capacities was validated. Using this model, analyses results of three case study structures of set dimensions showed that by decreasing the span length by 25%, and 40% the ultimate load carrying capacity increased by 33%, and 72%, respectively. However, this increase was not as much as what was expected by using concentrated plasticity model recommended by UFC. Therefore, the concentrated plasticity model does not always conservatively predict the ultimate load carrying capacity of the studied connection. Moreover, although the distributed plasticity model could predict ultimate load capacity of the connections reliably in the longest span it overestimated the load carrying capacity for the shortest span by 13%. This explains the significance of component-based modelling approach in order to simulate the structural behaviour subjected to a column loss.