1.0       INTRODUCTION

This edited report was prepared to help determine whether ripples along the edges of textured HDPE geomembrane being installed on long slopes of a landfill had adversely affected the long-term performance of welded seams made along those edges.

The client was concerned that the forces required to bring the two surfaces into contact for welding would result in there being significant residual stresses in the weld that might prematurely initiate failure by stress cracking while the liner is in service.

2.0       BACKGROUND

In the majority of cases the edges of geomembrane rolls used to fabricate lining systems for waste containment are quite flat.   However, even when flat, due to shading of the bottom sheet by the top sheet, and due to differential thermal expansion and contraction, welding leads to occasional wrinkles (“fishmouths”) in the top sheet.   Care is taken to cut the wrinkle and to make it flat for welding.  Therefore, the client was concerned when textured material with smooth edges (for easier welding) was deployed with continuous rippling (equivalent to many fishmouths) as shown in Figure 1.  The ripples were limited to the edges of the sheet and were up to approximately 9 in. in amplitude with a wavelength of about 12 to 18 in., as determined form the photographs.   Obviously welding would occur between two rippled surfaces, further increasing the concern.

Figure 1.         Ripples on edges of geomembrane rolls

Five samples were cut from affected seams selected by the client and submitted to the TRI/Environmental (TRI) geosynthetics testing laboratory in Austin, TX., for testing.  TRI was the first laboratory to be accredited under the Geosynthetic Accreditation Institute’s laboratory accreditation program (GAI-LAP).

The five samples were cut from areas of the seam shown in Figures 2 through 6 and were identified IRT-S1080101, IRT-S2080101, through IRT-S5080101 respectively.   They will be referenced as S1 through S5 in this report.

Figure 2.         Location of S1

Figure 3.         Location of S2

Figure 4.         Location of S3

Figure 5          Location of S4

Figure 6.         Location of S5

3.0       TESTING AND OBSERVATIONS

The samples, shown in Figures 7 through 11, were received by TRI on 2 August 2001.   It was noted that the outer track of the weld (closer to the edge of the top sheet) is generally not quite as well defined or as wide as the inner weld track.  However, it is not unusual for the two tracks to be different.   Except for the small hump on the inside track of S3, there were no significant creases or overheated areas evident on the tops or bottoms of the welds.  

Figure 7.         Sample S1

Figure 8.         Sample S2

Figure 9.         Sample S3

Figure 10.       Sample S4

Figure 11.       Sample S5

However, there were transverse curved striation marks (Figure 12) within the smooth area at the edge of the geomembrane that are indicative of differential flow of the geomembrane as, or after, it exited the extrusion die.

Figure 12.   Transverse curved striations on smooth surface
Textured region at bottom, seam at top.

Thin slice microsections were prepared at the following features for transmitted light microscopy:

These locations are marked A, B, and C, in Figures 8 and 9.   Specimens were examined with and without crossed polarizing filters to identify locations of residual stress.    The significant feature of each specimen is as follows:

Figure 13 shows a cross section of one edge of the seam of specimen 3A – the reference seam.    The squeeze-out bead is quite symmetrical, the sides are in good uniform contact with the two geomembranes, and the root of the bead flows smoothly towards the weld area.    Figure 14 shows the same area using crossed polarizing filters – the squeeze-out bead and weld zone are further highlighted, and there are no colored highlights indicative of residual stress.   This is further confirmed in Figure15 (under partial polarized light) showing the root of the squeeze-out bead where residual stress is often encountered.   

Figure 13        Edge of 3A. (x10)

Figure 14.       Weld zone between arrows (x10)

Figure 15.  Edge of 3A - weld zone between arrows (x40)

Figures 16 and 17 show equivalent views of specimens 3B and 3C respectively.   The squeeze-out beads are a little less symmetrical at the outside edge, and 3B shows some asymmetry towards the root of the bead resulting from the hump profile, but there are no residual stresses.   There is a little residual stress within the weld zone of specimen 3C but at this location it is of little significance.    The damaging residual stresses occur outside the weld zone within the geomembrane itself.

Figure 16.       Specimen 3B. - Weld zone between arrows.

(top x10, middle x25, bottom x40)

Figure 17.       Specimen 3C.  Minor residual stress arrowed. (x10)

Photomicrographs of the three specimens from Sample 2, at different locations on the wrinkles, are shown in Figures 18 to 20.   They are very similar to Figures 13 to 15 of the reference sample.   There is a little asymmetry of the squeeze-out beads and there is a little color generated by some residual stress.    However, these features are again in the weld zone and are not the brilliant reds and blues of potentially damaging residual stress.

Figure 18.       Specimen 2A.  (top x10, bottom x40)

Figure 19.       Specimen 2B. (top x10, bottom x40)

Figure 20.       Specimen 2C. (x10)

DISCUSSION

The presence of the ripples could affect the quality of the weld bond strength itself or may induce residual stresses within the weld or in the parent material at or close to the roots of the squeeze-out beads.    In a good weld, examined under the microscope, it will be difficult to differentiate between the weld zone, the adjacent heat affected zone, and the original parent material.    Inferior welds may show a distinct line along the interface at the center of the weld, voids, particulates, and periodic coloration of small regions of residual stress.   In a good weld the weld zone will be symmetrical across the seam, as will the squeeze-out beads at each edge.     Residual stresses are of concern since, in conjunction with generally applied stresses on the liner, they can be the initiation site of stress cracks; brittle cracks that occur in ductile material under a constant stress lower than the short term yield or break strengths of the material.   In addition, the presence of chemicals such as oxidizing acids, chlorinated solvents, and detergents, can accelerate the initiation and propagation of stress cracking – a phenomenon known as environmental stress cracking (ESCR).   However, although these chemicals are present in landfill leachates, their concentrations in leachate that collects on the primary geomembrane liner are so low as to be of no practical concentration

There are two ways to minimize the potential for stress cracking.    The first is to avoid stressing the seam, which cannot totally be avoided, and the second is to select an HDPE geomembrane resin with a high resistance to stress cracking.   Typical industry specifications require an HDPE geomembrane with a stress cracking resistance (SCR) of greater than 200 hr in the ASTM D5379 single point notched constant tensile load test.   However, production HDPE geomembranes have SCR values between 210 hr and as high as 10,000 hr, the latter being for more mechanically durable than the former.   Field failures have been seen in materials with SCR values up to 240 hr. 

It is understood that the HDPE resin from which the textured geomembrane is made is Chevron Phillips 9642.   This has an ASTM D5397 notched constant tensile load stress cracking break time of over 1000 hr, according to Chevron Phillips data sheets attached as Appendix A.   This compares favorably with the industry specification of >200 hr found in Geosynthetic Research Institute standard GM13.  With such a high SCR material and no significant residual stresses in the seams there is no reason for concern about the rippling compromising the durability of the seams.

4.0       CONCLUSIONS 

The rippling has not adversely affected the microstructure nor, probably, the performance of the resulting seams in the five samples examined.

* * *

Technical Data Sheet for Chevron Phillips 9642 HDPE Resin