A comparative study of Schmidt hammer test methods for estimating the uniaxial compressive strength of rocks

Abstract

The Schmidt hammer is being widely used for estimating the uniaxial compressive strength (UCS) of rocks because of its simplicity, rapidity and portability. However, determination of the Schmidt hardness (R) in laboratory conditions is sometimes very difficult for weak rocks due to the fact that samples can be broken during the test, as well as sample scarcity. Additionally, some Schmidt hammer test procedures necessitate more readings for obtaining the average R values than others. For these reasons, this study aims to explore more practical and useful Schmidt hammer tests by reducing the rebound readings, especially for the UCS estimation of rock materials. Accordingly, three different trial methods (T 1, T 2 and T 3) were studied on tested rock samples. T 1 is obtained by recording six single impacts and averaging all the values. T 2 is obtained by recording eight single impacts and discarding the lowest and highest value to obtain a mean rebound number. T 3 is obtained by recording ten single impacts and discarding the lowest and highest two values to obtain a mean rebound number. For comparison purposes, Schmidt hardness values were also calculated from four other test procedures (R 1–R 4) recommended in the literature. Forty-seven rock samples were tested in this study, including igneous, sedimentary and metamorphic rock origins. Statistical equations were determined for estimating the UCS of rocks by using trial Schmidt hammer test methods and other test procedures. Correlation, ANOVA and percentage error analyses were performed between the measured and estimated UCS values. The UCS of rock materials can be reliably estimated from any of the Schmidt hammer methods (T 1–T 3, R 1–R 4), taking into account the correlation and ANOVA analyses results. This study, however, demonstrated that T 1 is slightly more reliable and simpler to use than the other tested methods, giving a better representation of overall rock hardness, and hence a better prediction of UCS based on the percentage error analysis.