DEVELOPMENT OF MULTI-LINEARREGRESSION MODEL FOR THE CORRELATION OF PHYSICO-MECHANICAL PROPERTIES OF OKABA COAL DEPOSIT, KOGI STATE,NIGERIA

Abstract

software for the correlation of mine roadway parameters for Okaba coal deposit. For uniaxial compressive strength (UCS), fifteen specimens of 50 mm × 50 mm × 50 mm blocks of coal were examined by loading the (cube) specimen under a compression machine with a constant stress rate between 0.5-1.0 MPa. The UCS value for coal sample is 4.35 MPa which depicts that the material has a low strength. The point load is 2.49 MPa and Aggregate Crushing Value is 13.67 % respectively. For sieve analysis, the uniformity coefficient is calculated as 8.2%. The average porosity value of the coal sample is 22.97%, average bulk density value is 1.26 kg/m3, average moisture content of the coal sample is 10.38%, average specific gravity gives 1.33 while average ash content is 12.067%. Ash content and moisture content shows a direct and indirect relationship against the three mechanical properties of Uniaxial Compressive Strength, Point Load and Aggregate Crushing Value respectively. The sensitivity analysis also established a strong relationship between mechanical properties and moisture content. Linear multiple regression was used to model the relationship between the mechanical and physical properties of Okaba coal deposit. The coefficient of determination (R2) of 0.80, 0.95 and 0.94 was obtained for UCS, Point Load and Aggregate Crushing Value respectively. The resulting models were validated using both physical and statistical approaches. Positive and strong correlation coefficients of 0.85, 0.95 and 0.95 were obtained for UCS, Point Load and Aggregate Crushing Value respectively (p =0.000). This implies a positive and strong correlation between the laboratory data and the model estimates of the coal properties. The t-test statistics also revealed a non-significant difference between the two set of data at 5% level (p>0.05). Mine-ware 2018 software is necessary because it predicts unknown variables from known variables which enable quick calculation of the mines roadway strength parameters. The software will be useful for mineral industries