Reliability of various soil characteristics relationships for Kirkuk soil using consistent undisturbed field soils samples with linear regression models

Abstract

Due to the new construction and development of massive geotechnical superstructures, as well as a limited budget and time frame, it is critical to assess the reliability of numerous correlations between different soil characteristics. Various field soil data from 40 boreholes across Kirkuk City (Iraq) were obtained and they include water content (w%), liquid and plastic limits (LL and PL), plasticity and liquidity indices (PI and LI), unconfined compressive strength (UCS), wet and dry densities (γ_t and γ_d), and consolidation characteristics (initial void ratio “e_o”, compression index “C_c” and recompression index “C_r”). Field correlations with linear regression models were used to examine the reliability of relationships between different soil characteristics. The acquired soil data were evaluated using distinctive previous published models. It was demonstrated that the PI and LI are correlated positively whereas the w% and γ_d are correlated negatively and their degree of correlations are 0.878 and -0.573 respectively. It was proved that positive linear regression models can be obtained for the relations between the values of the LL, PL, PI, LI, and γ_t with the increase of the water content. Finally, it is demonstrated that the soils of Kirkuk City (Iraq) have Cc values less than the prediction of the previous models.