Reducing Weld Sizes in Circular Hollow Section Connections

In Canada, the cost of steel fabrication labour (for connections) can account for 60% of the cost of a steel structure. This % continues to rise. It is therefore increasingly important for engineers to reduce labour required to make connections to keep Canada's steel industry competitive. For welded tubular structures, this is a present challenge due to a lack of design guidance on welding to circular hollow sections (CHS). This is because the behaviour of welds in such joints, with respect to eccentric loading and non-uniform loading, is not well-understood. In lieu of guidance, designers “oversize” welds to prevent failure. This is usually a safe approach, but it leads to a lot of unnecessary labour for joint preparation and welding. The long-term objective of this research program is to develop a better (more economical, yet still safe) design approach for welds in CHS connections.

First, this research will identify the effect of eccentric loading on the strength of two types of welds: fillet welds and partial joint penetration (PJP) groove welds. Lab tests will be conducted on weld-critical eccentrically tension-loaded cruciform connections (ETLCCs) with one-sided welds (to mimic the situation in CHS connections). The effect of weld type, root penetration, eccentricity (magnitude and direction), and plate-element thickness on weld strength will be identified. Then, post-rupture tests (e.g. Vickers hardness) will be used to characterize the local material properties of the ETLCCs near the welds (i.e. the properties of the filler metal, base metal, and heat-affected zone). These properties will subsequently be used in finite element (FE) models of the ETLCCs to study approaches to simulating weld fracture. A standardized approach for extending large-scale fracture-critical connection test results (using FE methods) will be developed. Lastly, with this new approach, FE models of CHS connections (with weld fracture) will be developed to determine the combined effect of non-uniform loading and eccentric loading on the weld strength. Large-scale, fracture-critical lab tests will be conducted, in parallel, to verify these models.

This research will generate new knowledge and much-needed design guidance on welding to CHS that will help engineers reduce weld sizes and associated labour costs. It will also support the training of 2 doctoral, 3 master's and 2 undergraduate highly qualified personnel.