Abstract:
Rice husk ash (RHA) is one attractive additive that can be used for the enhancement of
engineering properties of problematic soils as a full/partial replacement of cement/lime. The
present study reviews the production of RHA and its basic characteristics as it relates to the
performance of RHA-stabilized soils. For each soft ground improvement application, a
substantial amount of time, money and effort is required for conducting extensive laboratory
tests to evaluate the improvement of geotechnical properties of stabilized soil and identify the
optimum mixture of additives. As an indirect approach, the present study explores the
development of predictive models for geotechnical properties using multiple regression
analysis (MRA) and artificial neural network (ANN) analysis. Laboratory experimental data
sets from an extensive literature review were used to develop the models. The models for the
prediction of Unconfined Compressive Strength (UCS), Soaked California Bearing Ratio (SCBR),
Maximum Dry Density (MDD), Optimum Moisture Content (OMC), and Plasticity
Index (PI) of RHA-stabilized clayey soil were proposed. It is noticed that at the prediction of
S-CBR, and MDD, MRA gives better correlations with more than 95% prediction accuracy.
Since MRA does not provide satisfactory performance for the prediction of UCS, OMC, and
PI, ANN models were developed with R2 of more than 0.95. The proposed models were
validated using the independent sets of data which is 30% of total data points. All the models
express a good prediction capability with a prediction error of less than ±7.5%. A parametric
analysis is performed to evaluate the variation of UCS of RHA-stabilized soil with the effect
of influencing input parameters. The result of PA suggests that 6-12% of RHA in
combination with a very little amount of cement (4-8%) or lime (3-6%) is the optimum mix
proportion for RHA-stabilized soil ensuring the robustness and reliability of the proposed
model. Hence, the proposed correlations may give easy access for facilitating the engineering
decisions during the pre-feasibility assessment. Further research into the production process
of RHA, the application potential of other waste materials to incorporate with RHA-cementlime
binder mixtures and evaluation of indirect approaches to assess the characteristics of
stabilized soil systems could lead to the utilization of RHA as a beneficial and productive
alternative in soil stabilization.
