New Approach for Conditional Coring in RC Structures Using Bivariate Distributions of Nondestructive Test Results

Abstract

Nondestructive tests (NDT) and partially destructive tests (PDT) are routinely employed for the management of aging infrastructure assets. Owing to the uncertainties in such NDT/PDT data, codes and guidelines recommend the fusion of multiple test results for making judicious inferences regarding the in-situ material properties. Researchers have proposed conditional coring to address the various constraints encountered in the costlier and damaging core tests (PDT). Conditional coring is the selection of core locations using the distribution of one NDT data (ultrasonic pulse velocity: UPV; or, rebound hammer: RH). This article first elucidates the limitations of such univariate approaches, providing a scientific basis for the variety of uncertainties typically present in RH and UPV data. Thereafter, a new approach for bivariate conditional coring is proposed to address these limitations. This method employs information from both RH and UPV to identify the PDT (e.g., core test) locations. The proposed method is validated against a variety of synthetic and real test data triads (RH-UPV-Core). It is concluded from the results that the bivariate method provides efficient and consistent estimates of the in-situ strength. These estimates are superior to those obtained from univariate approaches. This happens because of a better incorporation of the spatial variability in in-situ concrete strength and the measurement accuracy in NDT. The advantage of the proposed approach is that the number of PDTs can be reduced to achieve the same target accuracy as a univariate approach. Thus, bivariate conditional coring promises excellent potential for improving the structural health assessment and auditing practices for reinforced concrete structures. The factors affecting the in-situ strength of concrete are numerous and their interplay is complex. Therefore, the exact dependency of the strength on any nondestructive test (NDT) cannot be determined for a specific structure. Estimation of concrete strength during structural audit has benefited from univariate conditional coring [e.g., core strength using rebound hammer (RH) or core strength using ultrasonic pulse velocity (UPV)] and is being implemented in practice. However, further improvements are possible by using data from two types of NDTs in guiding the coring scheme. A new bivariate conditional coring approach is proposed here that addresses the uncertainties of partially destructive test (PDT) data in a systematic manner while balancing the test costs against safety concerns. The results demonstrate that the proposed method can either (a) achieve better accuracy with the same number of PDTs or (b) reduce the number of PDTs required to achieve the same accuracy. Although the bivariate approach would require both RH and UPV data, and some additional resources for computation, it is expected that the health assessment would benefit in terms of overall economy because NDTs are significantly cheaper than PDTs. The flowcharts presented in this paper should make the proposed approach easily implementable in practice.