Physico-Mechanical and Petrographic Analysis of Mansehra Granite, Hazara Basin, Pakistan

Abstract

The Mansehra Granitic Complex of the Hazara Basin, northern Pakistan, represents a critical lithological unit whose geotechnical properties have direct implications for infrastructure development, slope stability assessment, and rock engineering across the tectonically active Himalayan foreland. This study presents an integrated petrographic and geotechnical characterization of six granite specimens (S1– S6) collected from the complex, correlating modal mineralogy derived from visual estimation of petrography with two primary rock strength parameters; Uniaxial Compressive Strength (UCS) and Schmidt Rebound Hardness (SRH). UCS values across the six samples ranged from 124 MPa (S2) to 150 MPa (S1), while SRH values ranged from 52 MPa to 71 MPa, placing all specimens within the Strong to Very strong and High rock categories as per (GS, 1977, IAEG, 1977 and ISRM, 1978) classifications. Modal analysis revealed quartz contents of 45–66%, cumulative felsic indices (quartz + total feldspar) of 72–88%, and biotite contents of 4–18% across the sample set. Regression analysis demonstrated a strong positive correlation between SRH and UCS (R² = 0.9328), confirming Schmidt hammer reliability as a field strength proxy for this lithology. UCS correlated strongly with the cumulative petrographic index (R² = 0.8276), with higher felsic proportions consistently yielding greater compressive strength. The SRH vs Petrography regression produced a moderate R² of 0.6462, attributable to surface scale effects inherent to the Schmidt hammer test. Specimen dominated by K-feldspar and quartz (notably S1: 88% cumulative felsic, 29% K-feldspar) exhibited the greatest strength, while biotite rich, feldspar depleted specimen (S2: 72% cumulative felsic, 18% biotite) showed the lowest. These findings establish that Mansehra Granite is a very strong rock but due to the sample pool being limited to six (6) samples the correlation and regression analysis must be further validated by increasing the number of samples.