Field Performance of Corrugated Pipe Manufactured with Recycled Polyethylene Content (2018)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

41.1 Overview of Recycled Materials for Corrugated HDPE Pipes Corrugated high-density polyethylene (HDPE) pipe was first manufactured in the early 1970s for agricultural drainage appli- cations. The product offered several advantages over traditional clay tiles, including resistance to corrosion and abrasion as well as ease of installation (6). In the early 1980s, corrugated HDPE pipes made their way into highway and railroad applications, first for side drains and later for culverts and storm drains as larger pipes were developed. The first corrugated HDPE cross- road culvert was installed in Ohio in 1981 (7). Most of the small diameter pipes used in agricultural and other land drainage applications today contain varying percent- ages of post-consumer recycled (PCR) polyethylene materials, though these materials are currently not allowed in pipes used for federal and state highway applications. PCR materials are polyethylene materials that have served some previous func- tion as a consumer good—such as laundry detergent bottles and milk jugs—and have been recycled for use in new applica- tions. The recycling process consists of sorting these materials into similar plastics, chopping them up into flakes, washing them to reduce contamination, and then homogenizing and filtering them through a pelletization process. The final pellet- ized materials are sometimes referred to as repro (reprocessed). Figure 1-1 shows examples of typical PCR materials used in corrugated HDPE pipe applications that have been chopped into unwashed flake, washed, and reprocessed. Post-industrial recycled (PIR) materials can also be used for corrugated HDPE pipe applications. PIR materials are those materials that have been diverted from a waste stream and have never made it into the hands of an end user. Sources for these materials include scrapped finished goods that did not meet the manufacturers’ specifications for whatever reason and waste material generated in the production start-up process. If a manufacturer re-uses its own scrap material, the material is referred to as “re-work” or “regrind.” These materials are allowed in corrugated HDPE pipes for state and federal high- way drainage applications according to AASHTO M 294 (2). However, if corrugated HDPE pipe manufacturers purchase these PIR materials from an outside manufacturer or other source, they are not allowed in pipes for state and federal high- way drainage applications. In other words, only PIR materials obtained from one’s own manufacturing process (also known as re-work or regrind materials) are currently allowed in AASHTO M 294. While PIR materials can sometimes be of higher quality than PCR materials, depending on the source, they are not as widely available as PCR materials. Addition- ally, though PCR materials may be of lower quality than PIR materials with regards to stress-crack resistance and other prop- erties, they are relatively consistent and thus can be blended with higher quality virgin materials to develop a final blend with predictable performance (1). Because of this, PCR materials are more prevalent for non-AASHTO corrugated HDPE pipe applications and are the primary focus of this research project. There are two primary benefits to incorporating recycled materials into corrugated HDPE pipe. First, there is an eco- nomic benefit as recycled materials can cost significantly less than virgin materials. Discussions with pipe manufacturers have shown that savings of 20% or more can be realized by incorporating recycled materials into their pipes. Some of these cost savings are offset by reductions in manufacturing efficiencies, but these reductions in efficiencies can be mini- mized by proper filtration and cleaning of recycled materi- als. The second benefit of incorporating recycled materials into corrugated HDPE pipe is from an environmental stand- point. The majority of greenhouse gas emissions associated with the life cycle of corrugated HDPE pipe occur during the manufacturing and refinement of the raw materials (8). Substituting recycled materials for virgin materials into the manufacturing of pipe can greatly reduce this environmental impact. There is also significant public interest in using more sustainable resources in our transportation infrastructure, which was the primary motivation for this research project. C H A P T E R 1 Background Information

5 1.2 Failure Modes for Corrugated HDPE Pipes Manufactured with PCR Materials The global failure modes for HDPE have been well docu- mented [e.g., in Janson (9)]. When subjected to tensile stress, HDPE exhibits three distinct failure regimes or stages, as illustrated in Figure 1-2. At higher stresses, the primary fail- ure mode is ductile yielding, known as Stage I failures. An example of this failure mode would be a sudden burst in a pressure pipe due to a pressure surge or sustained high inter- nal pressures. Stage I failures are not typical in gravity-flow corrugated drainage pipes, as these pipes are not subject to high internal or external pressures. Additionally, since in-field deflections are limited to 5% in AASHTO highway applica- tions, corrugated HDPE pipes are not subject to high enough bending strains to result in Stage I failures. Stage II failures are more brittle in nature and occur at lower stresses and longer durations. Note the slope of the Stage II failure curve is steeper than the Stage I curve, indicating that this failure mode is less sensitive to stress. Stage II failures are caused by the slow crack growth (SCG) mechanism and are the primary service-limiting failure mode of concern for corru- gated HDPE drainage pipes. This failure mechanism will be discussed in more detail in Chapter 2 for corrugated HDPE pipes manufactured with PCR materials. Stage III failures occur when the HDPE material has essentially broken down due to oxidation, ultraviolet (UV) degradation, or other chem- ical attack. The curve for Stage III failures is virtually indepen- dent of stress, indicating the material is no longer capable of carrying any load. It is important to prevent the occurrence of Stage III failures prior to the end of the intended service life of the pipe. This can be done by ensuring that the material is sufficiently stabilized via the addition of antioxidants as well as carbon black for protection against UV degradation. While an important mode of failure for corrugated HDPE drainage pipes—particularly those exposed to sunlight for extended periods of time—Stage III failures were not the subject of this research project. Another potential mode of failure for corrugated HDPE drainage pipes is cracking due to fatigue. While there has been no known evidence of fatigue-related cracking on corrugated HDPE pipes manufactured with virgin materials in the field, it is a potential failure mode for pipes manufactured with PCR materials, as the likelihood of contaminants that act as stress risers is higher for pipes manufactured with PCR materials than virgin materials. The mechanism for fatigue-related cracking on virgin PE materials has been shown to be similar to that for cracking caused by constant tensile loads (10, 11). Fatigue testing on cracked round bar (CRB) specimens has shown that their failure curves exhibit a ductile to brittle transition, similar to that shown in Fig- ure 1-2, when testing at various stress intensity factors, and the CRB fatigue test has been used to effectively rank virgin PE materials according to their stress-crack resistance (12, 13, 14, 15, 16, 17). The mechanism for fatigue-related cracking on corrugated HDPE pipes manufactured with PCR materi- als should be similar to that for virgin materials, though the (A) (B) (C) Figure 1-1. PCR materials, from left to right: (A) unwashed flake; (B) washed flake; and (C) pelletized reprocessed materials. Figure 1-2. Illustration of HDPE failure modes.

6effects of contaminants on crack initiation are expected to play a larger role, similar to creep testing (3). 1.3 Stage II Stress Cracking Analysis and Mechanism The Stage II SCG failure mechanism has been well studied for virgin polyethylene pipe materials. The SCG mechanism is illustrated in Figure 1-3, courtesy of Peacock (18). When a load (stress) is applied perpendicular to the direction of a notch, the notch (denoted as “crack opening displacement” or COD in this illustration) will open. Over time, continued applied stress results in the creation of microscopic voids at the tip of the notch, known as “crazing.” Eventually, the stressed fibrils in this craze zone begin to fracture. As this occurs, a new craze zone is established further into the speci- men, and the craze-crack process continues until the speci- men fails via ductile yielding when the remaining ligament (a) A void or pre-existing notch acts as a stress concentration point; (b) Under constant or cyclical loads normal to the notch, the notch opening widens and micro-voids start to form in a zone at the tip of the notch; this development of micro-voids is known as crazing; (c) Over time, the fibrils start to yield and break, and a new craze zone develops further into the specimen. Eventually, the specimen fails by ductile yielding when the ligament stress becomes too high. Figure 1-3. Illustration of the SCG mechanism from Peacock (18).

7 stress exceeds the yield strength of the material. SCG failures resulting from a constant applied stress condition are referred to as creep crack growth failures in this document, while SCG failures resulting from continued application of small cyclical loads are referred to as fatigue crack growth failures. Broadly speaking, the SCG mechanism is composed of two phases: the crack initiation phase and the crack propagation phase. The crack initiation phase is accelerated by the pres- ence of a stress riser such as a void, notch, or contaminant. For virgin PE materials, the SCG mechanism can be effec- tively analyzed by testing and observing artificially notched specimens held under a constant load. Tests are typically con- ducted at elevated temperatures to shorten the failure times. The purpose of the artificial notch is to provide a stress con- centration site to accelerate crack initiation. The time to crack initiation and the rate of crack growth or propagation can be determined by monitoring the COD throughout the loading period. Additionally, observation of the notch tip with a trav- eling optical microscope throughout the loading process can identify when fracture first occurs, aiding in the distinction between the crack initiation and propagation phases. Linear elastic fracture mechanics have been used to describe the rate of crack propagation for virgin PE materials, as shown in Equation 1.1 (16). i (1.1) da dt A KI m = where da dt = Rate of crack propagation A, m = Material-dependent constants KI = Stress intensity factor for load normal to crack plane For corrugated HDPE pipes containing PCR content, these traditional methods of assessing the SCG mechanism are insuf- ficient because they do not properly address the effect of con- taminants on the overall SCG performance of the material. This is because the traditional methods involve testing and evaluating artificially notched specimens, and the stress inten- sity factor resulting from a sharp notch greatly overwhelms and masks any contribution to the SCG mechanism that may be associated with irregular contaminants. Therefore, the effect of the contaminant on crack initiation and propagation will not be effectively known unless a contaminant happens to be located directly in front of the notch tip along the path of crack propagation. These contaminants are of significant importance in accurately assessing the service life relative to SCG for pipes containing PCR materials, because they are crack initiation sites that are much larger and more variable than the stress risers typically found in virgin materials. For example, a typical stress riser in a pipe manufactured with virgin materials may be a spherical carbon black particle on the order of 60–70 nm (∼2.5 × 10–6 in.) in size, while a contaminant in a pipe containing PCR content may be up to 0.18 mm (0.007 in.) in size and very irregular, based on an 80-mesh filtration screen commonly used in the production process. So while tests on notched specimens may be useful for assessing the stress-crack resistance (SCR) and the rate of crack propagation of the base polymer in a PCR material blend, they do not give a good indication of the overall ser- vice life of PCR materials because the crack initiation and propagation rates emanating from a sharp artificial notch will be different from those emanating from randomly oriented irregular contaminants. To develop an accurate service life model for the perfor- mance of corrugated HDPE pipes containing PCR content, the effect of these contaminants must be taken into account. While it may be an interesting academic exercise to develop crack ini- tiation and propagation models for various sizes and shapes of contaminants (i.e., stress intensity factors) in a typical PCR material blend, it is not practical as these stress intensity factors will be infinitely variable for the various sizes and shapes of contaminants that are common in PCR material streams. For example, it is possible that certain shapes of contaminants will have larger stress intensity factors that may shorten the time to crack initiation than contaminants of equal size but different shapes or material makeup. Additionally, any test method that would be used to validate the model would have to be based on unnotched test specimens, making it virtually impossible to distinguish between the crack initiation and crack propagation phases of the SCG mechanism. For these reasons, a more macroscopic approach to assess- ing the service life of corrugated HDPE pipes manufactured with PCR materials relative to Stage II brittle cracking is prudent. This approach must take into account both the crack initiation and propagation phases of the SCG mecha- nism, as both phases will influence the overall time to brit- tle cracking of the pipe wall. Rather than try to distinguish between the two phases or eliminate either of them, a new test method was proposed to assess both phases together as an overall failure mechanism, as shown in Equation 1.2 and illustrated in Figure 1-4 (3). Intuitively, the greater the number and size of contaminants, the shorter the time frame associated with crack initiation (tCI), as there will be more crack initiation sites present. On the other hand, the time frame associated with crack propagation (tCP) should be primarily a function of the SCR of the base polymer and somewhat independent of the effect of contaminants. Thus a desired service life can be achieved by (1) reducing the amount and/or size of contaminants (through better filtra- tion, cleaner materials, etc.), thereby increasing tCI; or (2) using base polymer with greater SCR, thereby increasing tCP; or (3) a combination thereof.

8= + (1.2)t t tSCG CI CP where tSCG = Total time to failure via the slow crack growth mechanism tCI = Portion of SCG time related to crack initiation tCP = Portion of SCG time related to crack propagation To accurately assess the overall SCG performance of corru- gated HDPE pipes manufactured with PCR materials, including the effects of contaminants with regard to both crack initia- tion and propagation, an unnotched tensile test on coupons of materials taken directly from the pipe wall was developed. The history and development of this test will be discussed further in Section 1.3 and Chapter 2. 1.4 History of Unnotched Tensile Tests The idea of a tensile stress-crack test without a notch is not new. In fact, one early test was reported in 1960 (19) and resulted in ASTM D2552, “Standard Test Method for Envi- ronmental Stress Rupture of Type III Polyethylenes Under Constant Tensile Load.” The test used a small dumbbell- shaped specimen with a reduced section, similar to that used today in ASTM D5397 and F2136. The test was performed under various loads and temperatures with different wetting agents. Unfortunately, the test suffered from very long testing times and poor reproducibility, and the method was finally withdrawn in 1986. Ultimately, a notch was placed in the face, which led to the notched, constant tensile load test that was standardized in ASTM D5397 (20). ASTM D5397 was the direct precursor to the notched, constant ligament stress (NCLS) test, standardized in F2136 (4). An unnotched tensile stress-crack test on a larger speci- men was reported by Thomas and Woods-DeSchepper in 1993 (21). Thomas and Woods-DeSchepper described a test performed on a test bar about 12.25 mm (0.5 in.) wide and 152.4 mm (6 in.) long. The test was performed at 80°C in a 5% aqueous solution of Igepal CA-720. Because the scientists at the Federal Institute of Materials Researching and Testing (BAM) in Germany described the test to the authors, it was called the BAM stress-crack test. The test proved to be valu- able for evaluating geomembranes with textured surfaces because the texturing process often left stress-crack initiation sites on the surface. This test was also useful to evaluate the quality of geomembrane seams (22). One of the drawbacks of the test was that some cracks started at the grips and at the edges of the specimens. Cracks at the edges were reduced by shaving the edges with a microtome knife. The Florida Department of Transportation (FDOT) devel- oped a semi-unnotched tensile test known as the “junction test” in 2005 (23). This test was done on test specimens taken from corrugated pipe that included the junction between the pipe liner and corrugation. In this case, the junction served as a stress riser and in essence acted like a notch. The test was performed in deionized water under three separate conditions of stress and temperature. Testing under three conditions of stress and elevated temperature allowed the application of bi-directional shifting to create a master curve estimating the stress rupture behavior at room temperature. The next test was designed to incorporate the best features of the BAM test and the FDOT junction test. It was developed during NCHRP Project 4-32 by Thomas and Cuttino and was called the BAM–FDOT–Fathead (BFF) Test (1, 24). The test specimen was an ASTM Type I dumbbell where the end tabs were at least twice as thick as the reduced section. The use of this “fathead” specimen virtually eliminated cracks at the grips. When tests were run under three conditions of stress and temperature, the results could be bi-directionally shifted, producing a master curve estimating the service lifetime of the pipe or resin blend. The test was further refined and standardized during the research project that produced this report, and it was published as ASTM F3181, “Standard Test Method for the Unnotched, Constant Ligament Stress Crack Test (UCLS) for HDPE Materials Containing Post-Consumer Recycled HDPE” in 2016 (5). Figure 1-4. Illustration of contributions of the crack initiation (tCI) and crack propagation (tCP) phases of the SCG mechanism for a given stress (3).