Application of stochastic modal identification for damage evaluation in reinforced concrete bridges

• Grant number: 12/00431-9
• Support Opportunities: Regular Research Grants
• Start date: May 01, 2012
• End date: December 31, 2014
• Field of knowledge: Engineering - Civil Engineering - Structural Engineering
• Principal Investigator: Ricardo Carrazedo
• Grantee: Ricardo Carrazedo
• Host Institution: Escola de Engenharia de São Carlos (EESC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated researchers: Tiago Marrara Juliani

Abstract

Reinforced concrete bridges are essential to transportation infrastructure of Brazil. However, a considerable amount of bridges have suffered excessive deterioration after several years in service. Consequently an accurate diagnosis of structural integrity is necessary to decide if a strengthening process is necessary and, when necessary, to develop an economic design of retrofit. Several tests may be applied in field to obtain important information on structural integrity. Non destructive tests have a special role because they avoid additional damage with the extraction of samples, allow that measurements be carried out in several positions and be easily repeated when necessary after a specific time interval. Among these techniques, the experimental dynamic analysis has shown to be promising, as natural frequencies, modal shapes and damping offer important information on the global response of the structure and allow the detection of localized damages that affect this global response. This technique is specially desired from an operational point of view when environmental loads are applied such as wind, earthquakes or vehicle traffic. However these forces are unknown, uncontrolled and as a consequence only the response of the structure to these actions is obtained. The stochastic modal identification is recommended in this situation and is based on the hypothesis that the forces are idealized by a white noise with constant spectral density. Damage identification is possible by comparison of the dynamic properties of an original state to the deteriorated state after some time has passed or through calibration of an analytical model which is modified by the introduction of damage until a good comparison to the experimental model is obtained. In this research the described methodology for damage identification is intended to be applied in reinforced concrete bridges which will be represented by small scale models in laboratory conditions. (AU)