Abstract (EN):
The characteristics of structural materials deteriorate as a result of the application of repeated loads, such as railway traffic loads. The application of these loads originates lower stresses than the failure stress of the materials, which are experimented repeatedly throughout the design life of the structure, thus potentially resulting on the occurrence and propagation of cracks in certain elements or joints. This phenomenon is defined as fatigue. Such situation may require complex reinforcement operations, which might lead to traffic in-eruptions, particularly relevant in the event of occurrence of fatigue phenomena in the main structural elements of the bridge. The vibrations induced by the passage of trains can contribute to the magnification of the fatigue phenomena. Such contribution may be especially important in cases of resonance of the structure, which tend to occur predominantly for speeds greater than 200 km/h. In this paper, a methodology for the fatigue analysis of metallic railway bridges, known as the method of damage accumulation, is presented. The application of this methodology involves the stress cycles count by an algorithm generally used for this purpose, the rainflow counting method which was implemented within the scope of this work. The method of damage accumulation was applied to the fatigue analysis of structural elements of the main beams of the deck of the Alcâcer do Sal bridge, a bowstring metallic bridge, located in the Southern Line of the Portuguese Railways, for fhe passage of real high-speed trains and fatigue trains, as established in EN1991-2 (2003). The damages were obtained for a hypothetic traffic scenario and for the traffic mixes established in the regulation, that is, normal, heavy and light traffic. Concerning the real high speed trains the analyses were performed for the passage of the Alfa Pendular train and the high speed trains ICE2, EUROSTAR, TGV, TALGO, THALYS, ETR-Y and VIRGIN. The time records of the axial forces, from which the normal stresses were obtained, resulted from dynamic analyses for the passage of the referred trains at speeds between 145 km/h (¿40 m/s) and 420 km/h (1.2 x 350 km/h), except for the Alfa Pendular train in which case speeds ranging from 145 km/h and 265 km/h (1.2 x 220 km/h) were considered. The fatigue strength curve used in these analyses corresponds to detail category 80, as defined in prEN1993-1-9 (2003). The calculation curve has been obtained from the characteristic curve considering a partial safety factor equal to 1.35, corresponding to a cut-off limit (¿¿L) of 24 MPa. The results regarding high speed trains have shown that the damage is only significant near the resonance speeds. As an example, it is presented the stress histories in bar UC for the passage of the TALGO train at two speeds: 185 km/h, at which there is no resonance of the deck and 265 km/h, to which resonance occurs in the deck. The results show that for a speed of 185 km/h, there are no stress range cycles higher than 24 MPa, therefore the induced damage is null. For a speed of 265 km/h, a significant number of cycles at stress ranges higher than 24 MPa has been detected, hence there is a contribution to damage. The assumption of a hypothetic traffic scenario of real high speed trains, corresponding to the passage of 156 daily trains and a design life of 100 years, has revealed that the damage is greater than unity in the speed interval of [240,300] (km/h) for the TALGO train, in the speed interval of [300,360] (km/h) for the EUROSTAR train, and in the interval [360,420] (km/h) for the EUROSTAR, TGV and THALYS articulated trains. As for the fatigue trains, the greatest damages were obtained for the CF5 to CF8, CF11 and CF12 freight trains. In terms of individual passages, it was observed that the damage induced by real high speed trains in case of resonance is greater than that caused by the most adverse freight trains. In what concerns the traffic mixes (heavy, normal and light), the damages were lower than unity for the three scenarios and for a design life of 100 years. © 2006 Taylor & Francis Group.
Language:
English
Type (Professor's evaluation):
Scientific
No. of pages:
2