Modelagem numérica para o estudo do subsolo em regiões urbanas através do método da interferometria sísmica

• Ano de defesa: 2018
• Autor(a) principal: Calvet, Jahiro Damian Miranda
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Rio de Janeiro; Brasil; Instituto Alberto Luiz Coimbra de Pós-Graduação e Pesquisa de Engenharia; Programa de Pós-Graduação em Engenharia Civil; UFRJ
• Palavras-chave em Português: Interferometria Sísmica; Sísmica Passiva; Função de correlação; CNPQ::ENGENHARIAS::ENGENHARIA CIVIL
• Link de acesso: http://hdl.handle.net/11422/11424

In the many urban areas exist diverse seismic passive sources, originated from railways, road, civil constructions in the surface and subsurface and traffic noise in general. Geophones located in the surface areas can record all these signals. Processing these signals for one specific area, it makes possible to construct images with essential information of the geological stratum. Creating many images in each t period observes some stratum geology variation along the time. Defining in this form a potential risk index can be determined if the variation is compared with computational modelings for the area. In this work the Seismic Interferometry (SI) method was applied to model the passive sources presented around of the tunnel “Prefeito Marcello Alencar“ in Rio de Janeiro, Brazil. Firstly, three simple geological models were created with multi-sources positioned randomly and with different time recorded. Those multi-sources models were compared with the same models, but with one source positioned at the middle of the surface space. Secondly, the geological stratum of the tunnel was reproduced with four different multi-sources positions and compared with one source at the middle of the surface space. All of these models were applied in two cases for the acoustic and elastic physical properties. The Finite-Difference method of the order four order was used for the computational modeling results. Applying the cross-correlation functions and some transform like FFT and Gabor functions, the results are satisfactory. All modeling compared have precisions around to 95% and show great potential that the method offers for the analysis of urban areas.