A comparison between porosities measured from specimens with regular and irregular shapes and their accuracy in predicting the strength characteristics of carbonate building stones

Abstract

Porosity is among the basic physical parameters widely used in evaluating the strength characteristics of building stones. Based on the geometric shape of the stone specimen, there are two procedures for determining the porosity: 1) measurements on specimens with regular shapes and 2) measurements on specimens with irregular shapes. The present study correlates the strength characteristics of stones and porosities measured from specimens with regular and irregular shapes. To this end, samples of carbonate stones were collected from different locations in Iran. The porosity of the samples was measured for specimens with regular (nr) and irregular (nir) shapes. Next, the strength characteristics of the samples, including uniaxial compressive strength (UCS), Brazilian tensile strength (BTS), and point load index (PLI), were determined. Results revealed that for all samples, nr has lower values than nir. Based on the scanning electron microscope (SEM) micrographs, this difference is due to the filling of some pore spaces of the specimens with regular shapes during their preparation by the coring machine. UCS, BTS, and PLI(50) were correlated with nr and nir using simple regression analyses. The accuracy of the correlation equations was compared based on their determination coefficient (R2) and diagonal line (1:1) measures. The findings indicated that nir provides higher accuracy than nr in evaluating the UCS, BTS, and PLI(50) of the samples. The effect of dry density (ρd) on the correlations between UCS, BTS, and PLI(50) with nir was investigated using multiple regression analyses. According to the results, ρd has a positive role in the prediction accuracy of the UCS, BTS, and PLI(50). The validity and accuracy of multiple regression equations were verified according to the statistics criteria and the published data by various researchers. Using these equations obviates the need to perform the UCS, BTS, and PLI(50) tests as time-consuming and laborious efforts. © 2024 Elsevier Ltd