By Allan J. Breitenbach, P.E.
Editor’s note: Part 1 of Allan Breitenbach’s treatise recalling geomembrane testing in days of yore appeared in the February/March issue of Geosynthetics and continued with Part 2 in the April/May issue. This concluding section continues with descriptions of factors that influence liner interface test strengths. -rb-
Cohesion vs. no cohesion
Standard practice from the 1980s to today for conservative design strengths is to assume no cohesion in the slope stability analyses for liner interface strengths. The cohesion intercept decreases to zero for a clean and fully drained sand or gravel fill and increases in value for low permeability or wet of optimum high plasticity soils.
The friction strength can approach zero degrees with the cohesion intercept at its maximum value for a relatively saturated and fine- grained soil approaching unconsolidated and undrained (total stress) conditions with quick loading and fast shearing rates.
For three decades, labs have applied a normal confining stress load to the prepared test sample and immediately began the shear test. If direct shear liner tests are conducted under “quick” loading and shearing conditions, or if the sample has been remolded to wet of optimum moisture, the effective stress conditions may be unknown due to the potential for partial pore pressure conditions in the test sample at failure.
Under these conditions, the apparent cohesion should be discarded for a conservative friction strength estimate or a retest can be conducted with a longer time for pretest load consolidation and dissipation of excess pore pressures (24 hours minimum load time and longer depending on the material remolded density, gradation, plasticity, and moisture content).
The best chance for a true effective stress strength (fully drained for no excess pore water pressure) can be determined by allowing the confining stress to consolidate the sample before shear testing, in combination with a sufficiently slow strain rate to failure.
The effects of time consolidation and increased strengths is illustrated by test results for Sites B, C, D, E, and F listed in Tables 1 and 2.
Note that the friction strength increases and the apparent cohesion decreases as load consolidation time is allowed prior to shear testing.
Allowing as few as 12–24 hours of load consolidation time before testing increases the interface liner strength by several degrees. There may be some effect from microscale deformation or “dimpling” (Breitenbach and Swan, 1999) of the flexible geomembrane liner surface with time under loading (see section below discussing low vs. high fill load strengths).
There may also be some residual excess pore pressures along the failure plane due to dilation of compacted soils to a less compacted state at shear failure.
In either case, throwing out the adhesion or cohesion value will be a conservative approach.
More vs. less flexible liner
There is still debate about whether a difference in peak friction strength exists between the more flexible vs. less flexible liners.
Some confusion likely originated in early literature on published strength test results conducted in the 1980s using the small-scale 4in.-square shear box or the early direct shear tests that continue today on liner systems with wet of optimum low permeability soils.
The mining project experience gained from conducting large pilot scale liner test fills using loaded rock haul truck traffic at mine sites, observations of liner drain fill cover and heap lift construction, numerous large-scale direct shear box tests (performed at optimum or dry of optimum moisture content), and a review of known pad liner slope failures all point to only one conclusion.
There is no debate for liners subjected to landfill and heap leach pad fill loading conditions.
HDPE and MDPE liners are the most semi-crystalline and exhibit the lowest peak interface friction strength characteristics on smooth liner sheets, associated with being the least flexible of all the liner types on the market. PVC, CSPE (Hypalon), PP, TPO, EIA, and EPDM liners exhibit higher peak interface friction strength characteristics, associated with being the less semi-crystalline (most flexible) liners.
LDPE smooth sheet liners are generally a few degrees higher in interface friction strength relative to HDPE/MDPE liners and a few degrees lower in peak interface friction strength relative to the more flexible liners. Textured sheets help HDPE/MDPE liners increase in peak interface friction strength, but not to the degree of more flexible liners like PVC.
More flexible liners allow “dimples” or slight undulations to occur more readily on the interface surface, even under relatively low fill cover loads, which increases the shear strength test results. The dimpling or uneven geomembrane liner surface under consolidation loads causes the slide failure to shear through a portion of the underlying or overlying soil materials for an overall increase in shear strength with time along the geomembrane liner interface contact (Breitenbach, 2004).
Liner thickness appears to only influence less flexible HDPE/MDPE liners with lower interface friction strengths associated with thicker sheets.
An example of the relative difference in interface test strengths between liner types (HDPE to LDPE to PVC) tested with the same underliner and overliner fill materials with the same test conditions is shown by test results for Site A listed in Tables 1 and 2.
Smooth vs. textured liner surface
Textured geomembrane liner sheets (roughened surface with variable protrusion patterns) for the HDPE, MDPE, and LDPE liners were first used in the early 1990s for increasing the smooth surface friction strength along the more critical downhill toe limits of high-fill structures.
A textured sheet of HDPE liner improves its interface strength compared to a smooth liner sheet, but remains less than the more flexible liners such as PVC, CSPE, and PP smooth liner sheets, as shown by test results for Sites A, C, and E in Tables 1 and 2. Other design considerations should be noted in selecting textured liners for increased interface strength purposes.
Textured liner sheets have some installation difficulties in tacking, grinding, and welding of the seams across the textured rough surface for a watertight seam. Most textured PE sheets have a 4in.- (100mm-) wide smooth surface edge for hot shoe wedge welding, but impoundment corners and occasional angled sheet overlaps require crosscutting and seaming through the textured surface area. Liner installers can deploy the textured sheets in a parallel pattern as much as practical to avoid the crosscut extrusion welded seams.
Texturing decreases the parent liner integrity and thickness, therefore textured sheets in mining operations are typically designed for placement only in heavy fill traffic areas or along the critical perimeter and downhill toe locations of the lined facilities for fill slope stability. Landfill liner designs have migrated toward texturing the entire bottom areas, even where toe stability or interior slippage is not a concern. The logic for using high semi-crystallinity textured liners under landfills is not something I can explain. It appears to be a regulatory preference rather than an engineering preference.
Textured liner sheets can be single or double sided, with a variety of textured patterns and protrusion shapes to select for design, depending on the liner manufacturing process. The deeper and more sharp-edged texture patterns may have the best interface strength, but can weaken the crosscut seams and the liner sheet material itself.
Direct shear test results to date indicate minimal texturing is required for adequate liner interface strengths at moderate to high fill loads. More testing is required to judge the effects of texturing on the clean and coarse-grained drain fills placed above or below the geomembrane liner.
There may be clean drain cover fills that exhibit lower—rather than higher—peak friction strength characteristics on textured sheets, with the larger rock particles essentially riding over the textured protrusion points for less surface contact to resist sliding. This phenomena may occur for rounded rock particle shapes and for poorly graded rock size distributions. Finer rock particles, such as sand sizes, can nestle within the textured protrusions for greater interlocking contact strength.
In one case, composite liner test results on smooth and textured HDPE sheets showed an increase of only 2° peak friction strength difference with shear movement occurring along the underliner fine grained soil contact for the smooth sheet on the first test, followed by shear movement along the overliner gravel drain fill contact (rather than at the underliner soil contact) for the second double- sided textured liner test.
The main question clients ask about the single side textured sheet is whether the textured side should be facing up or down. The answer depends on which side has the lower strength soil or geotextile contact, since shear movement occurs only on one side of the liner interface contact at shear failure.
Numerous direct shear tests have been performed on a wetted sand or gravel overliner drain cover fill and a composite fine-grained clayey or silty underliner bedding fill placed generally within 1% wet or dry of optimum moisture content in contact with all types of geomembrane liners and liner thicknesses.
The observed liner surface after shear failure consistently showed that shear movement generally occurs at the geomembrane to underliner soil contact. Therefore, the single side textured sheet for composite liners should be placed facing downward in contact with the fine grained and more plastic soils for the best contact friction strength.
Low vs. high fill load strengths
An interesting aspect of high load consolidation and “dimpling” of the geomembrane liner surface with time is that the real-world interface strengths during operations are likely to be higher than the laboratory test strengths (Breitenbach, 2004).
Testing different incremental load consolidation times (using various soil and liner sample materials on seven mine projects, 1995–98) verified the trend of increasing laboratory liner test strengths with respect to pre-test load consolidation time. However, this is dependent on the amount of normal force load and duration of the load on the geomembrane liner surface.
Geomembrane liners on slopes such as landfill caps have a significant reduction in the normal load force vs. flatter grades at the top, and geomembrane liners subjected to high normal force fill loads such as on leach pad liner systems gain additional long-term strength from a mircoscale change in the liner surface conditions.
The microscale “dimpling” effect of the liner interface under loading, as discussed earlier, is related to several factors, including the plastic deformation of the soil (contact soil density, gradation, plasticity, moisture content), the potential for any localized differential settlement of the subgrade under loading, the flexibility of the geomembrane liner, and the fill load conditions (incremental vs. ultimate load and rate of loading).
The dimpled surface allows the liner interface strength to approach the strength of the underlying or overlying soils by becoming non-planar in nature, similar to the effects of a textured vs. smooth geomembrane liner sheet. Under low fill load conditions of typically less than 5ft (1.5m), such as a liner cap on a landfill, thicker and less flexible HDPE/MDPE liners do not have enough normal loading on the liner surface to cause dimpling of the larger rock particle sizes into the liner sheet surface.
Therefore, less flexible liners such as HDPE will show the lowest peak interface friction strengths compared to thinner sheets of the same HDPE liner type or more flexible liners like PVC.
I first studied the dimpling effect on several leach pad projects in the late 1980s through 1992 using PVC liner glued to relatively rigid wood blocks. The purpose of the glued block tests was to find the lowest planar effective stress strength of the flexible liner without the added strength of dimpling from sand and gravel rock particles in a drain fill cover.
The peak planar test strengths of the PVC liner with the wood block substrate were found to be in the range of 14–19° friction strength compared to about 22–26° friction strength with a composite liner and drain cover fill. Direct shear tests with wood blocks are shown for Sites G and H on Tables 1 and 2.
High plasticity clays for underliner bedding fill were being tested by the mid-1990s at extremely low peak and residual interface friction strengths at optimum moisture content. The normal confining stresses were applied for simulated high fill loads and allowed to consolidate 0, 12, and 24 hours before testing.
The delay in testing to allow consolidation of the sample increased the peak friction strength typically by 3–5° friction with a reduction in apparent cohesion. In this case, the gain in strength occurred from significant dimpling in the liner surface combined with pretesting load consolidation time (some potential pore pressure reduction, as well).
The individual fill lift slopes for landfills and ore heaps are generally placed at the natural angle of repose, and subsequent lifts are set back with benches to develop an overall flatter slope for enhanced stability or for ease in slope regrading at closure. In many cases the landfill lift slopes can be near vertical with compactor dozing. The first lift of fill in a multiple lift landfill or leach pad operation creates the greatest incremental change in load stress on the liner system and underlying subgrade soils.
As the fill height increases with the same lift thickness, the incremental change in load stress decreases. If the foundation liner system and subgrade soils contain materials with low strength characteristics (i.e., planar geosynthetic liner surface in contact with fine grained, low permeability, low density, wet of optimum, and/or high plasticity soils), a large incremental change in load stress may create temporary unstable conditions until the load consolidation has stabilized to effective stress conditions.
The impact of load stress consolidation on a general range of crushed gold and copper ore heap fill materials is illustrated in Figure 2, with a large initial fill lift change in stress at 0-50ft (0-15m) causing significant consolidation compared to minimal fill lift change in stress consolidation at 200-300ft (60-90m).
Therefore, early start-up fill lift operations are generally the most critical for the underlying geomembrane liner system.
The many factors influencing the direct shear test strengths and the differing site conditions compared to simulated laboratory test conditions demonstrate the importance of good engineering judgment in selecting the appropriate interface test strengths for design in combination with careful placement of the initial fill lifts during the construction and operation of the lined facilities.
Several “good news” items should be pointed out for young liner engineers:
- Loaded liner systems improve in strength over time due to both “microdimpling” and reduction in pore pressures.
- Higher fill loads are less of a slope stability concern compared to the start-up fill lifts (believe it because it is true).
- Liner quality from the manufacturers has significantly improved since my younger days.
- There will be many more debates among the best minds in the liner industry regarding the accepted landfill practice of wetting clayey soils above optimum moisture, and the related arguments to clamp and restrain liners in the direct shear box to overcome the wet clay soil deformations without consideration for the scale effect of the very small direct shear box and clamped liner size compared to the more massive size of the slope failure at the downhill toe of the fill.
Old-timers tend to observe the actual construction and operation performance, then go back to the laboratory test work and slope stability theory to explain what happened.
If I was able to get a word or two into the excellent discussions on floating vs. restrained liners (direct shear testing discussions at the end of the 9ICG Brazil 2010 conference), I would have mentioned not to forget both the “gnat on elephant” and “over optimum clay is trouble” Breitenbach theories (see Part 2).
I hope that other “old timers” will be encouraged to come forward and continue these excellent discussions at every conference and ongoing in Geosynthetics magazine.
Allan Breitenbach is Principal Geotechnical Engineer at Ausenco Vector, based in Denver, Colo., USA.
Breitenbach, A.J. (1997); “Geomembrane Pad Liner Failures Under High Heap Fill Loads,” Geosynthetics ‘97 Conference, Industrial Fabrics Association International (IFAI), Long Beach, Calif., Mining Session, Vol. 2, pp. 1045–1062.
Breitenbach, A.J., & Swan Jr., R.H. (1999); “Influence of High Load Deformations on Geomembrane Liner Interface Strengths,” Geosynthetics ’99 Conference Proceedings, IFAI, Boston, Mass., Vol. 1, pp. 517–529.
Breitenbach, A.J. (2004); “Improvement in Slope Stability Performance of Lined Heap Leach Pads from Operation to Closure,” Geotechnical Fabrics Report (GFR), IFAI, Roseville, Minn., Vol. 22, No. 1, pp. 18–25.
Daniel, D.E. and Wu, Y. (1993); “Compacted Clay Liners and Covers for Arid Sites,” Geotechnical Engineering Journal, ASCE, 119(2), pp. 223–237.
Estornell, P. and Daniel, D.E. (1992); “Hydraulic Conductivity of Three Geosynthetic Clay Liners,” Geotechnical Engineering Journal, ASCE, 118(10), pp. 1592–1606.
Hermann, J.G., and Elsbury, B.R. (1987); “Influential Factors in Soil Liner Construction for Waste Disposal Facilities,” Geotechnical Practice for Waste Disposal ’87, R. Woods, ed., ASCE, pp. 522–536.
Lambe, T.W., and Whitman, R.V. (1969); “Tests to Measure Stress-Strain Properties,” Soil Mechanics, Ch. 9, John Wiley & Sons.
Matasovic, N., Kavazanjian Jr., E., Augello, A.J., Bray, J.D., and Seed, R.B. (1995); “Soil Waste Landfill Damage Caused by 17 January 1994 Northridge Earthquake,” Woods and Seiple, eds., California Department of Conservation, Division of Mines and Geology.
Mitchell, J.K., Seed, R. B., and Seed, H.B. (1990); “Kettleman Hills Waste Landfill Slope Failure, Volume I: Linear Systems Properties,” Geotechnical Engineering Journal, ASCE, 116(4), pp. 647–668.
Koerner, R.M., Soong, T.Y. (1999); “Stability Analyses of Ten Landfill Failures,” Proceedings 2nd Austrian Geotechnical Congress, Austrian Engineering and Architects Society, Eschenbachgasse, Vienna, pp. 9–50.