Geomembrane tear in Canyon C (after EMCON Associates 1994). 

Contexts in source publication

Context 1

... EMCON Associates (1994) conducted a post-earthquake forensic investigation of the geomembrane tears. Samples of geomembrane were obtained from the two tear areas. In situ density measurements (using the sand cone method) were conducted on the soil above and below the tear areas and the soil was sampled for laboratory testing. An undrained shear strength of 62.2 kPa was reported for the low permeability soil-bentonite mixture beneath the Canyon C base geomembrane. EMCON Associates (1994) also conducted a series of interface shear tests on the geomembrane/ soil interfaces for the side slope liners in Canyon C and Canyon D. These tests were conducted in a 304.8 mm by 304.8 mm direct shear device. Geomembrane samples recovered from the landfill were used in the interface shear testing. All the testing was done using soil compacted to the in situ dry density at 2% above the in situ moisture content. Interface shear test results are summarized in Table 1. EMCON Associates (1994) repeated the direct shear tests using wetted interfaces and saturated soil with no decrease in interface shear strength. As part of the post-earthquake investigation, a study on the fracture morphology of the geomembrane was conducted by researchers at Drexel University using a scanning electron microscope to investigate the tear initiation and growth mechanism (EMCON Associates 1994). Six specimens from the Canyon C side slope geomembrane liner (S-1 to S-6) were sampled from at the locations indicated in the sketch presented in Fig. 3, i.e. adjacent to two tear faces. Based on the fracture morphology, it was concluded that the tear likely initiated from location of samples S-3 and S-4, i.e. at the fillet extrusion weld used to weld the patch at this location, and then propagated perpendicular to the dual hot wedge seam, i.e. perpendicular to the loading direction (EMCON Associates, ...

Context 2

... Table 7 presents a summary of the maximum tensile strain in the geomembrane from the analyses for cross section C1-C1. The maximum tensile strains for the two strong motion records were 4.3% to 3.8%. In both cases, the maximum tensile strain was at the anchor point at the top of the slope. The tensile strains calculated in the back analysis were well below the yield strain of the intact geomembrane. However, Giroud (2005) showed that failure in geomembranes in the field can occur in cases where the tensile strains are well below the yield strain due to strain concentration at seams and scratches in the geomembrane. Therefore, the procedure proposed by Giroud (2005) to estimate geomembrane strain concentration was followed to estimate the strain concentration and factor of safety for the geomembrane in cross section C1-C1. Giroud (2005) presented several correction factors that should be applied to the nominal yield strain of a geomembrane from uniaxial tensile testing to estimate the yield strain in the field. First, Giroud (2005) showed that the yield strain under plane strain conditions is lower than the yield strain in uniaxial tensile tests by a factor depending on Poisson’s ratio. Assuming a Poisson's ratio of 0.46 and a uniaxial yield strain of 12% for the HDPE geomembrane, the yield strain in case of plane strain will decrease to 10.9%. Giroud (2005) also showed that the yield strain will decrease due to scratches in the geomembrane according to the depth of the scratch, as illustrated in Fig. 9. Assuming that the ratio of the depth of the scratch to the geomembrane thickness is 0.2, the ratio of the yield strain of scratched geomembrane,  SY , to the intact geomembrane yield strain is 0.35 according to Fig. 9. This means the yield strain for a geomembrane with a scratch that penetrates 20% of the thickness of the geomembrane under plane strain conditions,  , is 3.8%. Giroud (2005) also provides a procedure that can be used to estimate the additional strain due to bending at a geomembrane seam perpendicular to the loading direction. Figure 10 shows a plot developed by Giroud (2005) for a 1- mm thick geomembrane and different types of seam. The fillet weld at the top of the patch where the tear initiated in cross section C1-C1, shown in Fig. 3, is perpendicular to the loading direction. Therefore, the Giroud (2005) procedure was used to estimate the incremental strain due to the seam at this location. Based upon an assumed thickness of 5 mm for the extrusion fillet weld, the additional strain due to bending at the seam was estimated to be 3.0%. Adding the calculated strain in geomembrane due to the earthquake loading (3.8%) to the bending strain due to the seam stress concentration, a total tensile strain in the geomembrane in the vicinity of the seam of 6.8% is calculated. Considering the strain ...