Condition Assessment of In-situ Concrete in Aging Residential Buildings in Mumbai ﾖ A Probabilistic Approach

Abstract

This paper presents condition assessment of in-situ concrete in existing residential buildings in Anushaktinagar, Mumbai region. For this purpose, various non-destructive and partially destructive tests have been carried out on 29 existing residential buildings which are more than 50 years old since their completion of construction. These residential buildings are mainly reinforced concrete framed structures. The typical plan dimensions of these buildings vary from 34.4 m x 4.85 m to 151.5 m x 7.4 m; the height above ground level varies from 10.55 m to 16.5 m. The tests included Rebound Hammer Test, Ultrasonic Pulse Velocity Test, Core Extraction Test, Half Cell Potential Test, Carbonation Depth Test, and Chemical Analysis of Concrete for Chlorides and Sulphates contents, and pH of pore solution in in-situ concrete. Locations and number of non-destructive and partially destructive tests on these buildings had been decided after having an initial visual inspection survey of these buildings. Various structural members, such as, beams, columns, and slabs were tested. The tests were performed at most of locations on the external members which are exposed to environment.

The paper presents both qualitative and statistical analyses of test result data. In statistical analyses, firstly statistical descriptors were evaluated for the tests results of non-destructive and partially destructive tests, and then by using these statistical descriptors fitting of different probability distributions have been attempted for these test results. Checks for goodness of fit for different probability distributions have been carried out using standard Chi-Square Test and K-S Test. Finally, best estimated values and 95 % confidence values have been presented for different non-destructive and partially destructive tests. These findings help in ascertaining the quality of in-situ concrete in existing aged residential buildings in a probabilistic manner.