Clay Mineralogy Effects On Long-term Performance Ofchemically Treated Expansive Clays

Stabilization of expansive soils using chemical additives such as cement and lime had been in use for several decades and these treatments provide riding comforts to travelers. However, many state Departments of Transportation (DOTs) in the United States have had problems with subgrade failure even after stabilization with chemical additives due to a loss of stabilizer over time, or a stabilizer being ineffective in some soils while other soils with the same index properties respond well to that stabilizer. These problems are attributed to the limitations of the current procedures of stabilization for pavement subsoil layers. One such limitation is the lack of understanding of the complex interactions between the mineralogy of the soil and the additives used for soil stabilization. Also, the design procedure is time consuming and hence the specifications are bypassed and the design is conducted based on the local experience. Hence, in this research study an attempt is made to address some of the limitations in the soil stabilization area. Incorporation of the mineralogy aspects of the soils into the stabilization design process, durability studies to address the long-term effectiveness of the stabilization and leachate studies are the main focus of this dissertation investigation.The first task was to develop a simple procedure to identify the dominating clay mineral in a given soil as the current procedures of mineral quantification are expensive and highly skill oriented. Hence, properties such as Cation Exchange Capacity, Specific Surface Area and Total Potassium were used and statistical regression equations were developed to predict the dominating clay mineral in a given soil. A total of twenty natural soils from different regions of the state of Texas and 6 artificially mixed soils with known mineral percentages were obtained Prediction models were developed using the database. Tools such as regression analysis and artificial neural networks were utilized to develop the prediction models. These clay mineralogy prediction models were validated using the validation artificial soil data. Predictions by both methods were compared and it was observed that ANN based prediction model showed better prediction capabilities than regression model based equations. However, the differences between predictions are small and practically negligible. Hence any of these models could provide realistic prediction of clay mineralogy in a given soil.The second task was to assess the effects of clay minerals on the long-term durability of stabilized expansive clays by conducting wetting/drying (W/D) studies replicating moisture fluctuations expected during summer and winter seasons in the field. A total of eight soils were selected for studying the long-term performance of stabilized expansive soils by conducting wetting/drying studies. Stabilizer design was carried out as per the TxDOT methods Tex 120-E (Lime as additive) and Tex 121-E (Cement as additive) and the results are presented. An accelerated curing method was developed and followed in this study for curing and moisture conditioning of the treated soil specimens. The effect of curing methods is studied on four select soils and it is observed that both the curing methods including old and longer curing (Tex 121-E) for 17 days and the present accelerated curing methods of 3 days yielded similar UCS and volume change test results. It has been interpreted from these results that there was no considerable effect of curing on the long-term performance of these treated soils except for an initial strength since soil was partially saturated.Soils containing montmorillonite as a dominant mineral are more susceptible to premature failures after chemical stabilization when they are exposed to volume changes caused by swell and shrink related volume changes. Also, it is understood that low amount of additive dosage can cause premature failures in the pavement structure. The third task was to assess the performance of the stabilization under severe rainfall conditions where heavy amounts of rain water infiltrates into the soil and causes leaching of the stabilizer and thereby reducing the life time of the stabilization. To understand this behavior leachate studies were conducted on all the eight soils selected to replicate moisture ingress and digress in the field during rainfalls and to study the effect of these moisture infiltrations on the long-term performance of stabilized soils. Leachate samples were collected after 3, 5, 7, and 14 cycles of leaching to address the chemical changes occurring due to leaching of the additive from the soil specimen. Also, unconfined compressive strength tests were conducted on soil specimens after 3, 7 and 14 cycles of leaching to address the strength changes from leaching. Finally, an attempt was made to highlight the effect of the loss of strength in the treated soils due to the above mentioned climatic changes on the performance of a flexible pavement. Four different flexible pavement sections with varying asphalt concrete layer and base course layer thicknesses were altogether analyzed. The effect of treated base modulus deterioration on the pavement performance was assessed by obtaining the compressive strains on the subgrade top. These strains were used with the Asphalt Concrete Institute (2006) formulation to predict the ESALs required to cause rutting failure. It is expected that the present research findings will be helpful in the future modifications of current stabilizer design practices by implementing clay mineral identification methods in the initial screening of soils and also selecting the dosages based on both durability and clay mineral information.