DEVELOPMENT OF QUASI-EMPIRICAL MODEL FOR PREDICTING DEGREE OF SOIL COMPACTION
This study developed a model, ?1 = a?2b for predicting the degree of soil compaction using compaction operational variables such as compactive energy (E), area (A) of the soil on which E is exerted, the thickness of the compaction slice (z), and acceleration due to gravity (g); and soil index properties such as soil moisture content (mc), antecedent soil bulk density (?bc), soil particle density (?s), mean soil particle size (?50), soil's uniformity coefficient (Uc = ?60/?10), the density of water (?w) in the soil and coarseness of the soil (C/F), where C is a coarse fraction of the soil and F is a fine fraction, using three soil classes namely, laterite, non-cohesive and cohesive soils obtained from Iwontu community, near Ilorin, Kwara State. A function of the type C_d=f[(E/A),t,z,g,m_c,?_bc,?_s,?_50,U_c,?_w,C/F] that relates all the thirteen independent dimensional variables believed to be influencing Cd have not been successfully determined so far. The common practice has been to rely on case-specific model studies for a given project and apply the results to that project type only. That approach lacks the capacity for general application. The aim of the study therefore was to use all the variables influencing compaction to develop a quasi-empirical model that can be used to predict the degree of soil compaction. Seven levels of compactive efforts, using a drop-weight type compactor varying from 147.15 to 1030.05 Nm were used to compact five pairs of embankment and slice thicknesses, Z and z respectively, with Z varying from 140 to 700 mm and z from 20 to 100 mm. Dimensional analysis was carried out on all the governing variables to make them dimensionless and expressed in functional forms. A functional relationship, f, of the dimensionless terms, ?1 = Y and ?2 = X was established for the soils of the known values of ?2. Five (5) type functions namely linear, power, exponential, polynomial and logarithmic were considered to establish the best-fit function; and the power function gave the highest value of the coefficient of determination, R2, varying from 98.8% to 98.9%. Validation tests were carried out and the model estimated the soil degree of compaction with more than 75% accuracy. It is recommended that further studies of different sub-soils under different conditions will be required to strengthen the applicability of the model. The model is important especially when there are financial restraints, short timeframes, or unavailability of compaction test equipment, particularly in early design stages and preliminary studies, before appropriate soil sampling and field investigation can be conducted; thus saving substantial time and financial resources.
Authors : Audu, D., Balogun, O.S., Fadeyibi, A., Busari, R.A. and Adebayo, K.R.
Category : Open Access Volume (Issue) : 10(2) Date Uploaded : 29th October 2024