Building pathologies caused by failure of Fundão Tailing Dam: A principal component analysis aproach

Abstract

Abstract This article presents a study of the impacts caused by Fundão dam failure in buildings from Gesteira district, Barra Longa city, Brazil. The analyzed dataset was built using technical reports from surveys carried out in 152 buildings. Principal component analysis was capable of explain the interdependence of data variable and allowed conditions of understand the consequences and evidenced pathologies. Heavy vehicle traffic caused more damage (57% of buildings) to the studied buildings than the contact with the tailing mud (43% of buildings).

Keywords

• Heavy vehicle traffic
• impacts
• pathologies
• tailing mud
• unsupervised machine learning