Segurança viária em trechos urbanos de rodovias federais
Carmo, Cássio Leandro do
MetadataShow full item record
This thesis analyzed road safety of urban stretches of Brazilian federal highways, in an evaluation that involved the verification of occurrence of victims of crashes together with geometric conditions, signaling and pavement of the highways, and characteristics of urban areas. For the development of this work, accident records by the Federal Highway Police (PRF), and data on pavement, geometry and traffic signaling conditions, according to the methodology of the National Transportation Confederation (CNT), were used. Concepts based on the Safe System approach were utilized for the delimitation of accidents, which considers that traffic accidents, when they occur, should not cause serious or fatal victims. In the states of Paraná and Santa Catarina there were, in absolute numbers per state, the highest numbers of victims between 2010 and 2014. Espírito Santo presented the worst rates of injured and fatal victims per 100,000 inhabitants and per 10,000 vehicles, while Amazonas and São Paulo had the best performances. Throughout Brazil, 3,178 federal highway segments were found in urban areas with the occurrence of at least one seriously injured or fatal victim. The 64 most critical stretches were identified. It was found that most of these critical segments are in large urban areas, and many were in very good condition. This result demonstrates that other variables, such as composition and volume of traffic, can contribute decisively to the most serious accidents. By means of a proportional stratified sample of 118 stretches, a primary relation was found between the number of victims and stretches with regular road infrastructure conditions inserted in urban centers with greater area of influence and high Municipal Human Development Index (MHDI). The Poisson regression showed higher values of the Prevalence Ratio (PR) for poor road sign, good geometry and regular pavement, in the disaggregated analysis of road factors.