Fatigue behaviour of riveted steel lap joints

• Tipo: Artigo em Livro de Atas de Conferência Internacional
• Data: 2006

Autores

Figueiredo, MAV (Autor) FEUP; Jesus, AMP (Autor) Outra; Ribeiro, AS (Autor) Outra; De Castro, PMST (Autor) FEUP; Fernandes, AA (Autor) FEUP

Ata de Conferência Internacional

Título: Proceedings of the 3rd International Conference on Bridge Maintenance, Safety and Management - Bridge Maintenance, Safety, Management, Life-Cycle Performance and Cost

Páginas: 479-481

3rd International Conference on Bridge Maintenance, Safety and Management - Bridge Maintenance, Safety, Management, Life-Cycle Performance and Cost

Porto, 16 July 2006 through 19 July 2006

Indexação

ISI Web of Knowledge

Scopus - 1 Citação

Classificação Científica

CORDIS: Ciências Tecnológicas

Outras Informações

ID Authenticus: P-007-P1E

DOI: 10.1201/b18175-194

Abstract (EN): The maintenance and safety of existing bridges is a major concern of governmental agencies. In particular, the safety of old riveted highway bridges fabricated and placed into service at the end of the 19th/ beginning of 20th centuries deserve a particular attention, since they were designed taking into account traffic conditions, both in terms of vehicle gross weight and frequency, completely different from those observed currently. In order to assure high safety levels in old riveted steel bridges, highway authorities have to invest heavily in their maintenance and retrofitting. In this context, knowledge of the fatigue behaviour of riveted joints is of paramount importance. The present paper reports research work carried out to characterize the fatigue behaviour of the Portuguese Pinhão riveted highway bridge, designed by Eiffel at the end of 19th century and built between 1903 and 1906. The Pinhão bridge crosses the Douro river and links Pinhão to São João da Pesqueira and Peso da Régua. The bridge has three spans of 68.8 meters each and one span of 10 meters; there is only one deck with 6 meters width, divided in one traffic lane with 4.60 meters width and two sidewalks with 0.675 meters width each. The goal of the paper is to characterize the residual fatigue strength of the bridge. Both traditional S-N approaches and Fracture Mechanics approaches were used. The study is supported by an experimental program for the evaluation of material properties such as tensile strength, toughness and crack growth properties. Also, fatigue tests of riveted joints were carried out. The material and riveted joints were extracted from bridge members. The members removed from the bridge were a diagonal and a bracing from the first span, from Pinhão side. While the bracing is composed by two equal-leg angles, riveted to each other, the diagonal has a rectangular cross section. Several types of specimens were prepared using the material samples removed from the bridge. These specimens were used in chemical and metallographic analyses, hardness measurements, tensile tests, notch toughness tests, fatigue crack propagation tests and fatigue tests of riveted lap joints. The chemical and metallographic analyses revealed that the material of the bridge is a carbon steel with small content of Mn, Si and C. The microstructure is composed, essentially, by ferrite with a small content of perlite. The tensile tests demonstrated a high ductility of the materials with almost null strain hardening. Elongations of 70%, associated to reductions in area of 33%, were observed. The measured mean yield stress and the ultimate tensile strength were 306 and 361 MPa, respectively. The hardness measurements revealed a small scatter, which confirms a good microstructure homogenization. Mean hardnesses of 108 and 116 HV40 were found for the diagonal and bracing materials, respectively. Two types of notch toughness tests were carried out. The Charpy V-notch and the COD tests. Both tests revealed that the materials exhibit very acceptable toughness properties even for the current design requirements, which allows a high tolerance to the presence of defects. Crack growth tests using Compact Tension (CT) specimens were also carried out to characterize the crack propagation law for the materials of the bridge. Two stress ratios were investigated, namely, R = 0.01 and R = 0.5. It was observed that Paris law gives a good description of the crack propagation data (Paris & Erdogan 1973). The global mean crack growth law derived was: da/dN = 3.1961×10 -15·¿K3.6117 (1) where da/dN = crack propagation rate expressed in mm/cycle; and ¿K = stress intensity factor range expressed in N.mm-1.5. The influence of the stress ratio is small and the materials from the two members present very similar fatigue crack propagation behaviours. Finally, fatigue tests of riveted lap joints were performed which allowed the evaluation of a S-N curve with the following form: log¿¿ = 3.3108 - 0.2226 log N (2) where ¿¿ = stress range in MPa; and N = failure cycles. The observation of the facture surfaces revealed, for some specimens, the existence of initial cracks that nucleated and propagated during the last century. The comparison of the experimental data with the AASHTO (1995) class D S-N curve revealed that the later is conservative. The residual life of the bridge was evaluated using a S-N approach, based on the class D S-N curve, proposed in the AASHTO procedures for riveted joints. Considering a standard vehicle with a gross weight of 300 kN and three axles (RSAEP, 1983) and using information about the daily average traffic flow supplied by the Portuguese highway authorities, the stress spectra at each member was evaluated and the critical one was identified. This analysis demonstrated the safety of the bridge, against fatigue, for a period of 30 years after its rehabilitation. Subsequent analyses, with more precise stress spectra, based on the actual weight density distribution of trucks, confirmed again the safety of the bridge. In this case, the damage accumulation rule proposed by Miner (1945) was applied. Finally, Fracture Mechanics was applied to assess the number of cycles required to propagate an initial crack until critical dimensions, leading to unstable propagation. An initial semielliptic crack was assumed and its propagation was modelled using the Paris law identified for the materials with the experimental work. The stress intensity factor range was evaluated using formulations available in the literature (Cheng & Li 2003). The results from the application of the Fracture Mechanics confirmed the results of the simulations based on the S-N approach. The main conclusions of this study can be summarized as follows: - The material of the Pinhão bridge presents mechanical strength properties similar to values obtained with materials of other European bridges built at same time. - The toughness values of the material are much higher than values demanded by current design codes of practice, which allows a high tolerance to the presence of cracks. Cracks hided by the riveted heads are not critical unless if they become visible. Inspection routines for crack detection are required. - The fatigue resistances obtained with the fatigue tests of the riveted joints are compatible with the recommendations of actual international codes of practice such as the AASHTO. - The present study demonstrates the bridge safety against fatigue, after rehabilitation, for a period of 30 years. This analysis was supported by important assumptions related with the stress spectra at the critical locations. More accurate analysis can be adopted if the stresses/loads are experimentally monitored, during a representative period of time. © 2006 Taylor & Francis Group.

Idioma: Inglês

Tipo (Avaliação Docente): Científica