DEVELOPMENT AND ANALYSIS OF THREE APPROACHES FOR GCL DESICCATION CYCLING

Abstract

Geosynthetic clay liners (GCLs) are hydraulic barriers that contain of a thin layer of Nabentonite clay encased by two geotextiles and are held together by needle-punching fibers or glue. Na-bentonite, mainly comprised of the smectite mineral montmorillonite, has characteristically high swell and low hydraulic conductivity (i.e 10-11 m/s), making Na-bentonite an ideal material for use in landfill covers to decrease the influx of water into a landfill. There is the potential for cation exchange of divalent cations for bound monovalent sodium during GCL permeation. When cation exchange is coupled with desiccation, the hydraulic performance of a Na-bentonite GCL can be affected over time. While previous laboratory wet-dry cycling studies have explored the importance of applied stress and environmental controls during testing, they have generally required a large amount of specimen handling that can result in non-representative deformations in the form of GCL curling. Experiments conducted in this study have been designed to examine more representative methods of GCL desiccation during wet-dry cycling (Chapter 1). Conventional Na-bentonite specimens were initially permeated with either deionized water or a pore water prepared to simulate typical landfill cover soils. Specimens were then subjected to one of three cyclic desiccation techniques that were designed for a greater degree of environmental control. Unconfined desiccation is the method previously used in literature that represents the least environmental control and consists of drying in a controlled-humidity environment without applied overburden stress. GCLs subjected to perforated plate desiccation were dried in a controlled-humidity environment under an applied overburden stress of 20 kPa. In-permeameter drying represents the greatest degree of environmental control and consists of the GCL being flushed with controlled-humidity gas while maintained under isotropic stress conditions.

Various aspects such as the testing duration, hydraulic conductivity as a function of wetdry cycles, and drying uniformity were compared between the three desiccation techniques. In addition, images of the desiccation cracking patterns were taken and quantitatively compared using a novel image analysis algorithm (Appendix C). Although the unconfined method results in an unnatural degree of GCL deformation in the form of curling, the hydraulic conductivity as a function of wet-dry cycles was unaffected when bentonite paste is added to the sidewalls. Therefore, the results of previous wet-dry cycling studies utilizing this technique are valid. The testing duration of a GCL subjected to perforated plate desiccation is significantly longer than the other methods. As a result, perforated plate desiccation is not recommended for future GCL wet-dry cycling studies. Even though in-permeameter desiccation requires additional equipment and can artificially accelerate the desiccation process, the method is recommended with a reduced air flow rate. A study was undertaken to verify the GCL image processing and analysis techniques (Chapter 2). Validation of the image analysis method is provided by observation of small error values in comparisons made between simulated crack geometries of known dimensions and those quantified using the analysis. In addition, polymer modifications of bentonite focused on activating and maintaining osmotic swell in adverse conditions. Experiments conducted in this study have focused on the effect of wet-dry cycling on the hydraulic performance of polymer-modified GCLs (Appendix A). Three GCL products were tested and include a bentonite polymer composite (BPC) GCL, a dry polymer- bentonite mix GCL, and a conventional untreated Na-bentonite GCL.