The UniGrow model is an analytical procedure to assess the fatigue crack growth based on elastic–plastic crack tip stresses and strains. The assumption is that fatigue crack growth (FCG) can be considered as a process of successive crack re-initiations resulting from material damage in the crack tip zone. The main parameters of this model are the crack tip radius and the elementary block size. Experimental FCG data obtained for S355 carbon steel showed that assuming the elementary block size with the same value of the crack tip radius to collapse FCG data using UniGrow model is non-coherent with experimental evidence. In this sense, a new approach is proposed by establish a clear distinction between crack tip radius and elementary block size. The value of the crack tip radius, ρ, was defined by correlation with experimental and numerical values of residual compressive stress field ahead of the crack tip while for the elementary block size, ρ∗, a new expression was proposed which relies on effective stress intensity factor range and cyclic yield strength. This research intends to be a valuable contribution for the implementation of UniGrow model to assess the fatigue crack growth of a material.@en

The maintenance and safety of ancient bridges is a major concern of governmental authorities. In particular, the safety of old riveted bridges fabricated and placed into service at the end of the 19th century deserves particular attention. These structures are susceptible to exhibit high fatigue damage levels due to their long operational period with increasing traffic intensity associated to an original design not covering the fatigue phenomenon. This paper reviews recent fatigue behaviour investigations on single and double shear riveted joints performed by Universities of Porto (Portugal), Trás-os-Montes e Alto Douro (Portugal), and Wrocław (Poland), in particular concerning the fatigue characterization of riveted joints extracted from representative Portuguese riveted bridges, namely the Eiffel, Luiz I, Fão, Pinhão and Trezói bridges. In order to overcome the influence of scatter and establish a reliable assessment for the obtained experimental data, two statistical approaches were used: implement linearized boundaries following the recommendation in ASTM E739 standard and defining probabilistic S–N fields using the Castillo & Fernández-Canteli model. This statistical analysis allows to propose design S–N curves for single and double riveted joints and evaluate the applicability (safety) of using the design curves suggested in Eurocode 3 as well as design curves proposed by Taras and Greiner.@en