Unsaturated behaviour of geosynthetic clay liners

2016-12-13T23:35:38Z (GMT) by Asli Senem Acikel
Geosynthetic clay liners (GCLs) are widely used as hydraulic and gas barriers. However, insufficient initial GCL hydration from a surrounding soil can lead to poor performance in field applications. The extent to which a GCL can hydrate from the surrounding soil will be dependent on the unsaturated behaviour of the GCLs. This thesis seeks to advance the current knowledge of the unsaturated behaviour of needle punched GCLs. Three main areas are examined: (1) the limitations encountered when characterizing the unsaturated behaviour of GCLs to identify suction measurement/control techniques capable of providing a water retention curve (WRC) most applicable to geotechnical requirements; (2) three experimental water retention techniques (filter paper, chilled mirror and osmotic) were modified and improved to allow more accurate evaluation of the WRC of GCLs including consideration of their time dependency; and (3) experimental implementation of the outcomes from the WRC tests into a composite liner application where the GCLs are in contact with different subgrade types.
   The WRC of GCLs is related to the properties of both the bentonite component (very large suction range, bi-modal structure, swelling behaviour) and the composite structure of GCLs, (tri-modal structure composed of bentonite and geotextile). Potential for capillary breaks due to the substantially different nature of the GCL components (hydrophilic bentonite and hydrophobic geotextiles) as well as their different porous structure (tri-modal GCL and bi-modal bentonite) and the time dependent changes in suction values of the bentonite all complicate testing.
   The filter paper technique was used to measure GCL suction on the wetting path following three procedures: (1) noncontact filter paper test (NCFPT) for total suction; (2) initially dry contact filter paper test (ID-CFPT), and (3) initially wet contact filter paper test (IW-CFPT) for matric suction. Testing time dependency was examined. The contact filter paper tests gave accurate matric suction limits up to ~ 66 kPa for the ID-CFPT and ~ 146 kPa for the IW-CFPT due to the pore structure and wetting characteristics of the filter paper. These limits corresponded to the water
  entry value of the wetting curve (~66 kPa) and the residual water zone of the drying curve (~146 kPa) of the filter paper (Whatman no. 42).
   Different conditioning and testing times were used with the chilled mirror hygrometer technique to evaluate time dependent behaviour and measurement accuracy on WRCs of the GCLs in terms of total suction (drying and wetting). The GCL with granular bentonite and woven carrier geotextile showed a greater effect of conditioning time (both in wetting and drying) than the GCL with powdered bentonite and thermally treated woven carrier geotextile. The higher confinement of scrim-reinforcement and thermal treatment significantly minimized conditioning time effects for the powdered bentonite. A conceptual model was developed to explain the observed differences in time dependent behaviour and estimate the impact of suction measurement on accuracy of GCL WRCs.
   The osmotic method with a dynamic gravimetry modification used to monitor the water retention process during both wetting and drying provided data supportive of the proposed conceptual model of time dependent behaviour. The data also indicated that the scrim-reinforcement and thermal treatment minimized the wetting/drying hysteresis. The largest hysteresis had been observed on the GCL with granular bentonite and woven carrier geotextile.
   Based on the studies conducted, the ID-CFPT is recommended for drying path “total suction” measurements for suctions > 400 kPa. The ID-CFPT and NCFPT methods are not recommended for wetting paths. For GCLs, the osmotic technique controls total suction rather than matric suction as had been reported previously.
   It is shown that the unsaturated behaviour of GCLs and subgrade explained why insufficient hydration of GCLs may occur in barrier systems. The high suctions in subgrade containing smectite limited the water uptake of the GCLs. When there was a capillary break, attributed to the pore size gaps between subgrade pores and GCL, GCL hydration from the subgrade soil was limited to ~25% gravimetric water content (GWC), corresponding to the residual zone of GCL WRCs and also to the volume of micro - pores in the GCLs. WRC and hydration tests indicated that the thermally treated scrim reinforced geotextile surfaces are most likely to promote capillary breaks while the woven geotextile surfaces are least likely.