TY - JOUR
T1 - Nondestructive Corrosion Monitoring in Reinforced Concrete by Boundary Element Inverse Analysis
AU - Suga, Kazuhiro
AU - Yoshida, Masato
AU - Ridha, M.
AU - Aoki, Shigeru
JF - Advanced Materials Research
VL - 33-37
SP - 1293
EP - 1298
SN - 1662-8985
PY - 2008
PB - Trans Tech Publications
DO - 10.4028/www.scientific.net/AMR.33-37.1293
UR - https://www.scientific.net/AMR.33-37.1293
KW - Boundary Element Method (BEM)
KW - Corrosion Monitoring
KW - Inverse Problem
KW - Kalman Filter (KF)
KW - Optimization of Observation Condition
AB - This study proposes a monitoring method for corrosion on reinforced concrete structure
using inverse analysis approach. At first, we define an inverse problem to identify the real and
imaginary parts of the concrete conductivity and the impedance between concrete and steel. The
observation of the inverse problem is the electric potential on the concrete structure surface when the
AC impedance measurement is performed. The observation condition, such as layout of observation
point and type of observation, of the inverse is optimized. The optimization is achieved by
minimizing the average of eigen values of a posteriori estimate error covariance matrix based on the
Kalman Filter estimation algorithm. We show a numerical simulation to solve the inverse problem on
the optimized observation condition to evaluate the effectiveness of the condition. The simulation
shows the real parts of the concrete conductivity and the impedance are well identified but imaginary
parts of them are not. To overcome this difficulty, we evaluate the sensitivity of the imaginary part of
the impedance to the real part and the imaginary part of the electric potential. We find the possibility
that the observation of the imaginary part of the electric potential improves the estimation. Finally, a
numerical simulation is performed under the optimized observation condition considering the above
discussion. In the numerical simulation shows that considering the sensitivity of the parameters to the
potential improves the solution on the inverse problem.
ER -