Data Fusion to Improve the Concrete Diagnosis

Abstract

Numerous non-destructive testing (NDT) methods are used for concrete structures to obtain relevant data about material properties and damage states for reliable condition assessment. Whether the objective is to determine physical properties such as the porosity and the water saturation rate, or mechanical properties such as the elastic modulus or the compressive strength, sensitivity of NDT techniques to many characteristics of the material and its environment is a commonly encountered problem. Thus, accurate and reliable information is often difficult to extract due to the high level of uncertainty involved. Complementary use of different NDT methods for coherent combination of information obtained from each method is a sensible strategy to improve evaluation. The data fusion methodology presented in this paper makes use of the complementary data obtained from different non-destructive or destructive techniques to improve diagnosis reliability. In the case of imprecise and uncertain data, an assessment can still be made with a quantitative measure of the uncertainty involved. The methodology is based on the possibility theory and allows the selection of the best combination of data and techniques to evaluate the material. Applications of the methodology are presented and the results are discussed. Results show good agreement between estimations by data fusion and measured values. Also shown by the results is that the selection of complementary techniques is essential for a better estimation of indicators and improved diagnosis.