Methods
Determination of the chemical composition of the rice husk ash and soil sample were determined by means of gravimetric method.
The soil sample was subjected to preliminary tests, such as the particle size distribution, specific gravity and Atterberg limits. The natural soil samples were also subjected to compaction tests and the California bearing ratio (CBR).
The soil samples were later treated with cement at values of 0, 2, 4, 6, 8, 10 and 12% which were separately mixed with rice husk ash (RHA) at varying proportions of 2, 4, 6, 8, 10, 12, 14 and 16%. These samples were each subjected to Atterberg limits, compaction and California bearing ratio. The entire tests were carried out according to BS 1377[12] for natural soil samples and British Standard 1924 [24] for stabilized samples.
Results from the laboratory tests especially that of the liquid limit, plasticity index, maximum dry density, optimum moisture content, the percentages of cement and rice husk ash served as inputs which were employed in developing the Artificial neural network (ANN) model for predicting soaked and unsoaked CBR values of the stabilized soils.