Estimation of compressive strength of various stones based on non-destructive techniques

This work examines the compressive strength of natural stones, as estimated by Uniaxial compressive strength tests (UCS) and respective non-destructive tests results (Schmidt Hammer Rebound test, SHR, or an Ultrasonic Pulse Velocity transmission, UPV). In practice, aiming to correlate the values of the UCS test with the results of the NDT techniques, three empirical regression models are usually applied, a linear, an exponential, and a power mathematical model. In most cases, the regression analysis is referred to a limited dataset, for a specific type of stone, and thus the investigation is rather limited and cannot be used for the evaluation of a whole category of stones (such as limestone, marble etc.). For this reason, in the present work these three models are examined for a more expanded dataset, including different types of stones, and measured by different researchers. An optimized linear model is proposed for the prediction of the compressive strength based on the NDT results. The proposed models are extremely useful for various applications and different construction works, concerning either modern or historic restoration, especially when stone samples cannot be obtained and information about the properties of the building stone, or calibration curves, are not available.