Behaviour of fillet welds as a function of the angle of loading

Abstract

Fillet welds subjected to longitudinal shear have been recognized as being weaker than those loaded at other angles, particularly transversely, since the early 1930's. However, the lack of an analytical method to relate the strength of fillet welds to the angle of loading has led to a reluctance to utilize the increased weld strength for other loading angles. Designers have taken a lower-bound approach basing the strength of fillet welds on the strength of longitudinal fillets regardless of the loading direction. Methods are available that predict ultimate strengths reasonably well but may not correlate well with all fracture observations.A rational analytical model for determining the ultimate strength of fillet welds loaded in shear dependent on the angle of loading is developed, relating the ultimate strength to the inclination of the fracture surface and the restraint conditions at the root of the weld. Good correlations of the variation of both the fracture angles and ultimate strengths with angle of loading exist between the results of 42 fillet welded specimens tested in tension and predictions based on this model. Correlations with the test strengths of others, while showing greater scatter, are also demonstrated. The model, extended to limited tests by others on fillet welded specimens loaded in compression, provides substantial correlation for these weaker welds. While weld deformations vary considerably with the loading direction, weld strains are found to be independent of the loading direction. Key words: analytical model, behaviour, fillet welds, fracture angle, loading angle.