Abstract

This thesis focuses on fatigue problem on riveted structures and the correlated experimental tests. The term “fatigue” denotes the cracking of metals under repeated loading. The technique of riveting structures is obsolete due to the low level of standardization in the construction process. Knowledge concerning riveted structure’s ability to withstand fatigue has not been investigated to the same extent as for modern structures assembled by welding. Nevertheless, many riveted structures are still in service after over 100 years. Clamping force originates when the hot rivet is placed into the hole of the plates and the rivet shorten in length due to cooling. In Europe, a large number of railway bridges are riveted. Moreover, all over the world, the rivet assembly technique has been largely used for different types of buildings. The riveted structures are subjected, evidently, to fatigue degradation as all the other steel structures. The riveted joint complexity and the non-uniform realization lead to a structure-specific consideration. In Europe, current regulations recognise only one fatigue class for riveted structures neglecting the clamping force effect. This approach is characterized by an elevated standard deviation for the assessed results. This research aims to improve the riveted structures fatigue comprehension and to propose appropriate tests. Some elements have been extracted from a dismantled railway bridge located near San Stino di Livenza (Venice). An innovative clamping test (TCT) has been conceived: applying a torsional moment to the rivets and evaluating the sliding friction, the clamping force has been estimated. Contrary to traditional clamping tests, this setup, at an affordable price, provides good precision. Moreover, the test is in-site and does not provoke damages to the examined structure. Fem models of rail bearer have been realized taking into account rivets, holes and multilayer plates section inertia. A specific laboratory set up has been designed with the intention to calibrate the fem models. Frictions and clamping force have been taken into account in the rail bearer models. There is a close correspondence between the non-linear models and the experimental tests. The clamping assumptions, derived from the TCT, have been verified. A full scale fatigue test has been prepared and the first cycles have been monitored. Full scale fatigue tests are sophisticated and involve many parameters. For this test, a specific metallic frame has been designed and realized using advanced fatigue models. An advanced analysis (hot spot method) has been carried out, for the frame, in order to evaluate the stress into the welds. This frame will be used by the DICEA laboratory to test, principally, specimens extracted from bridges.