No abstract

The effect of steel grade on the weight and cost of Warren-type welded tubular roof trusses was investigated. For an unbiased comparison, best truss designs were obtained through optimisation. The steel strength investigated ranged from S355 to S960. Costs were calculated based on the features of the trusses. The starting point was the exact geometry of the truss, from which a finite-element model was derived. This approach allowed the resistance and other requirements of design standards of both members and joints to be included as constraints in the optimisation problem. The design variables were the truss height, the locations of joints, the gap width at the joints and the member sections, from a catalogue of cold-formed square tubes. Eurocode 3 requirements were applied. The results of the comparison imply a significant saving in weight when using high-strength steel (HSS), by as much as 50% for S960. The cost reduction was smaller, but still about 20%, for a hybrid solution of S700/S355 for the higher of two investigated load values. It is hoped the results will motivate the use and further investigations of HSSs in building products.

A considerable number of studies have been conducted worldwide on fires that act on all four sides of a column (symmetrical fire). These cases are used for the validation of the analysis models developed in this study. In real buildings the columns are often embedded.If the fire does not act similarly on all surfaces of the column (non-symmetrical fire), it is extremely difficult to predict how the column will behave. The key research questions are: Is resistance stronger in non-symmetrical than in symmetrical fires? What is the final buckling mode, towards the fire or in the opposite direction?Results of numerical analyses for reinforced concrete filled square steel tube columns in nonsymmetrical fires are presented for a total of 150 cases. An ISO 834 fire acts constant along the column on one, two adjacent or three sides. Three embedding systems are considered for the remaining sides: adiabatic, concrete wall and sandwich panel. The material models are done using the Eurocodes and an initial bow imperfection is considered. The reference cases are symmetrical cases.When fire acted on one, two adjacent or three sides, the fire resistance times were on average about 3.4, 2 and 1.3 times longer than in a symmetrical fire. A concrete wall is a good thermal sink for columns. Slender columns typically buckle towards the fire. The final failure mode and corresponding resistance time depend on the direction of the initial bow imperfection. Experimental tests are needed to verify the results.

No abstract

No abstract