INTRODUCTION Throughout the 1990s and the early 2000s, there has been a significant shift in market share amongst product categories in the low-slope commercial roofing market. According to the most recent SPRI statistics, thermoplastic membranes now comprise more than 40% of the single-ply commercial roofing industry. This growth has been the result of a number of factors, not the least of which is the reliability of the membranes’ thermally fused seams. Single-ply membranes provide no redundancy. A less-than-perfect single-ply seam almost inevitability leads to a leak. A key advantage of thermoplastic membranes is the ability to fuse their seams together with hot air. Once properly welded, seams remain watertight throughout the service life of a roof. Although there have been few changes in the fundamental technology of hot-air welding over the decades, new materials such as thermoplastic olefins (TPO) have created new challenges for contractors installing these materials. This study was conducted to compare the welding properties of a sampling of commercially available thermoplastic roofing membranes. EXPERIMENTAL PROGRAM Five different thermoplastic membranes were tested: two TPOs, one ketone ethylene ester (KEE) modified polyvinyl chloride, and two “traditional” polyvinyl chloride (PVC) membranes. The original intent was to test 1.5-mm (60-mil) thick, polyester-reinforced samples of all products. However, only 0.9- mm (36-mil) KEE could be sourced. Al – though thinner than the other samples, the material should nonetheless be tested to ensure all three categories of thermoplastic roofing membranes were represented in the study. The samples were labeled TPO 1, TPO 2, PVC E (KEE material), PVC 1, and PVC 2. Products TPO 1, PVC 1, and PVC 2 had smooth surfaces on both the top and the bottom. The scrim telegraphed through both surfaces of membranes TPO 2 (significantly) and PVC E (moderately), providing a texture to both sheets. Each material was welded as received, after being subjected to two conditioning procedures. One set of samples was loosely rolled and fully submerged in water for a period of four days at room temperature. An additional set of samples was laid outdoors and covered with a mixture of organic topsoil, stone dust, and fine sand. The soiling Table 1 – Cleaning procedures used to remove soiling. MA R C H 2009 I N T E R FA C E • 5 PRODUCT CLEANING PROCEDURES TPO 1 Water and rag Simple Green and floor brush Water rinse Acetone with cotton rags TPO 2 Water and Simple Green 3M scrub pad with manufacturer’s solvent (Xylene) Cotton rag with solvent PVC E Water and rag Cotton rag with acetone PVC 1 Water and rag Cotton rag with acetone PVC 2 Water and rag Rag with Methyl Ethyl Ketone (MEK) compound was broomed across the membrane samples and then rolled over numerous times with a weighted lawn roller. The soil was left in place on the samples for a 30-day period during the month of July. The conditioning procedures are intended to assess the degree to which products maintain their weldability when subjected to conditions common to rooftops: moisture and soiling. WELDING TRIAL All 15 sets of samples (five materials, three conditions) were welded for each combination of three welding speeds: 1.5, 2.0, and 2.5 m/min (4.9, 6.5, 8.2 ft/min); and four temperatures: 350, 400, 450, and 500°C (662, 752, 842, and 932°F). A total of 12 weld conditions were carried out per sample set. Samples were removed from the baths, and unrolled, and both the top surface and the underside of each sample were dried with a cloth until all visible moisture was removed. The samples were then welded. Typically, welding occurred within 15 minutes of drying. For products with a factorysealed edge, the sealed edge was used as the top sheet in welding the samples that had been immersed in water. The soil was removed from the outdoor test area with a broom before the samples were taken inside to the test area. Soiled samples were cleaned according to procedures outlined in each membrane manufacturer’s product literature prior to welding (Table 1). Every effort possible was made to return the membranes to their original appearances using the prescribed procedures. Welding was conducted within the same day as the cleaning processes. WELDING TEMP (°C) WELD AS RECEIVED SPEED TPO 1 TPO 2 PVC 1 PVC E PVC 2 m/min kN/m kN/m kN/m kN/m kN/m 350 1.5 11.2 6.7 2.1 4.9 11.6 350 2 11.3 6.8 0.8 5.5 13.1 350 2.5 11.5 6.9 0.6 5.0 12.1 400 1.5 11.5 6.9 12.0 5.7 11.3 400 2 12.0 7.0 3.2 5.5 10.9 400 2.5 12.2 7.0 2.2 6.0 12.0 450 1.5 10.8 7.0 12.2 6.8 10.2 450 2 11.7 7.1 12.6 6.1 11.1 450 2.5 12.0 6.9 12.1 4.4 11.4 500 1 11.3 7.0 12.2 6.7 11.5 500 5 12.1 7.1 13.5 7.2 9.3 500 2.5 11.7 7.0 12.1 5.6 9.8 WATER CONDITIONING 350 1.5 9.3 4.8 7.0 5.2 12.0 350 2 8.9 3.4 4.8 4.9 11.3 350 2.5 7.1 2.3 2.0 4.1 8.9 400 1.5 9.0 7.2 10.5 4.5 9.0 400 2 9.2 7.1 6.1 3.8 9.0 400 2.5 9.7 3.7 5.2 4.4 11.3 450 1.5 8.7 7.3 7.0 4.6 6.6 450 2 9.0 7.8 8.7 3.9 8.0 450 2.5 8.0 6.4 7.9 4.1 8.1 500 1.5 7.7 6.9 4.7 4.5 9.4 500 2 9.3 7.9 8.1 5.2 6.5 500 2.5 8.2 7.1 8.2 4.5 6.5 SOIL CONDITIONING 350 1.5 9.1 6.7 10.5 4.0 10.8 350 2 9.7 6.0 9.8 5.0 11.5 350 2.5 7.4 4.1 7.2 4.6 11.1 400 1.5 9.1 5.9 7.8 4.1 12.4 400 2 9.5 5.8 6.6 4.4 12.1 400 2.5 7.8 5.5 7.5 4.3 12.6 450 1.5 8.1 6.8 9.8 4.8 11.2 450 2 8.5 5.6 9.1 3.8 11.6 450 2.5 7.5 4.5 8.6 4.2 12.4 500 1.5 6.5 6.2 10.2 5.3 11.8 500 2 9.0 6.6 13.2 4.3 12.1 500 2.5 9.2 6.3 13.5 4.7 13.1 Figure 1 – TPO 2, soiled, 350ºC, 2 m/min, Table 2 – Peel test results for all conditioning procedures and welding parameters. 5.9 kN/m (34.0 lbf/in). 6 • IN T E R FA C E MA R C H 2009 SAMPLE TESTING T peel tests were conducted on a United Tensile Tester according to ASTM D1876- 95. All data noted in Table 2 represent the average of five T peels. Data were recorded in pound-force and converted to SI units. DISCUSSION The peel test data must be viewed in a performance context to be of practical use. Failure modes during peel tests can be divided into two categories: adhesive and cohesive, with failure occurring within the seam and within the membrane material, respectively. Thermoplastic membrane man ufacturers call for peel tests to be done manually on seams throughout the installation phase of the roofing materials. Seams that fail adhesively are deemed to be unacceptable and typically must be patched or stripped in as a precondition to the warranty being issued. Simmons et al. found that adhesive failure was typically observed when seam strength was found to be 4.5 kN/m (kilonewtons/meter, equivalent to 26 lbf/in) or less; whereas, cohesive failure typically was observed in samples with seam strengths greater than this value. In the present study, 26 seams were found to have failed at that value or lower. Examination of the tested samples revealed that 25 of the 26 had failed adhesively, validating Simmons et al.’s observation. One must be cautious, however, in the use of this concept. Relying exclusively on simple numerical values can in some instances prove misleading. For example, TPO 2 (soiled, 350°C, 2 m/min), achieved a weld strength of 5.9 kN/m (34 lbf/in) – cohesive failure by this definition. However, examining the sample (Figure 1), one can see that the area of fused polymer holding the welds together is minuscule and that the greatest part of the seam has failed adhesively. Such a seam would not likely be able to withstand the stresses imposed within a lap-attached system. Blistering was also observed in a number of the seams after water immersion. Although high peel strengths were still measured, such seams may deteriorate prematurely due to freeze/thaw cycling or other mechanisms in the field. Therefore, it is important to remember that acceptable peel strength and weld continuity are equally important in assessing a seam on a roof. This value does, nonetheless, provide us with a basis for calculating a weld safety factor to evaluate and assess seam quality at various conditions – or to compare the Table 3 – Safety factors evaluated for all conditioning procedures and welding parameters. MA R C H 2009 I N T E R FA C E • 7 WELDING TEMP T (°C) WELD AS RECEIVED SPEED TPO 1 TPO 2 PVC 1 PVC E PVC 2 m/min Safety Safety Safety Safety Safety Factor Factor Factor Factor Factor 350 1.5 1.5 0.5 -0.5 0.1 1.6 350 2 1.5 0.5 -0.8 0.2 1.9 350 2.5 1.5 0.5 -0.9 0.1 1.7 400 1.5 1.5 0.5 1.6 0.3 1.5 400 2 1.6 0.5 -0.3 0.2 1.4 400 2.5 1.7 0.5 -0.5 0.3 1.6 450 1.5 1.4 0.5 1.7 0.5 1.2 450 2 1.6 0.6 1.8 0.3 1.4 450 2.5 1.6 0.5 1.7 0.0 1.5 500 1.5 1.5 0.5 1.7 0.5 1.5 500 2 1.7 0.6 2.0 0.6 1.0 500 2.5 1.6 0.5 1.7 0.2 1.2 WATER CONDITIONING 350 1.5 1.0 0.0 0.5 0.1 1.6 350 2 1.0 -0.3 0.1 0.1 1.5 350 2.5 0.6 -0.5 -0.6 -0.1 1.0 400 1.5 1.0 0.6 1.3 0.0 1.0 400 2 1.0 0.6 0.3 -0.2 1.0 400 2.5 1.1 -0.2 0.1 0.0 1.5 450 1.5 0.9 0.6 0.5 0.0 0.4 450 2 1.0 0.7 0.9 -0.1 0.8 450 2.5 0.8 0.4 0.7 -0.1 0.8 500 1.5 0.7 0.5 0.0 0.0 1.1 500 2 1.1 0.7 0.8 0.1 0.4 500 2.5 0.8 0.6 0.8 0.0 0.4 SOIL CONDITIONING 350 1.5 1.0 0.5 1.3 -0.1 1.4 350 2 1.1 0.3 1.1 0.1 1.5 350 2.5 0.6 -0.1 0.6 0.0 1.5 400 1.5 1.0 0.3 0.7 -0.1 1.7 400 2 1.1 0.3 0.4 0.0 1.7 400 2.5 0.7 0.2 0.7 0.0 1.8 450 1.5 0.8 0.5 1.1 0.1 1.5 450 2 0.9 0.2 1.0 -0.2 1.5 450 2.5 0.7 0.0 0.9 -0.1 1.7 500 1.5 0.4 0.4 1.3 0.2 1.6 500 2 1.0 0.4 1.9 -0.1 1.7 500 2.5 1.0 0.4 2.0 0.0 1.9 welds of various materials, at least under experimental conditions. The safety factor (SF) will be defined as shown in Formula 1. Safety factor data is compiled in Table 3. As received, TPO 1 and PVC 2 provide safety factors in excess of 0.4 in all cases, with most conditions yielding safety factors well in excess of 1.0 (i.e., 100%, or double the defined threshold value). TPO 2 and PVC E represent the other end of the spectrum, providing little room for error at any set of parameters. The telegraphing of the scrim through the surfaces of these materials is apparently a detriment to achieving high weld strengths. In the case of PVC E, the lesser thickness of the available polymer to achieve a weld very likely compounds the challenge of trying to achieve a strong weld. For materials that had been immersed in water, TPO 1 and PVC 2 appear to provide the greatest overall degree of safety, albeit in inverse fashion. PVC 2 allows for a higher level of safety at the more moderate temperatures, whereas TPO 1 provides it at the higher welding temperatures. TPO 2 achieves nominal safety factors at higher temperatures, as does PVC 1, except at the lower welding speeds. Based on the data generated in this test program, PVC E provides no room for error. Although four days of immersion may be severe, the results are troubling when one considers the material was thoroughly surface dried and that welding was completed under highly controlled conditions. After soiling and cleaning, PVC 2 provides the greatest margin of safety at all conditions, with safety factors ranging from 1.4 to 1.9. PVC 2 achieves safety factors equal to or better than the “as-received condition.” Although TPO 1 provides for good safety factors despite an aggressive, multistep cleaning process, there is clearly a reduction from the values achieved “asreceived.” TPO 2 suffers from a modest degree of loss in safety factor, although at such low levels for new material, any reduction in safety margin could be critical in practice. The results for PVC E are quite similar to those evaluated after water immersion. Once again, there is no margin of safety in welding this product. WELD WIDTH During the testing program, significant discussion about weld width ensued. Many manufacturers require a minimum of a 37- mm- (1.5-in-) wide weld. Factory Mutual’s system approvals typically note a minimum of 37-mm- (1.5-in-) wide welds. Many inspectors – both manufacturers’ representatives and third-party consultants – consider anything narrower to be unacceptable, and they will typically insist that such seams be stripped in. Those involved decided to conduct a small test program by preparing two sets of samples (one glass-mat-reinforced and one polyester-reinforced) with precisely controlled weld widths for T Peel testing. This was followed by a full-scale 366-cm x 732- cm (12-ft x 24-ft) uplift test at Factory Mutual Engineering with 13-mm- (0.5-in-) wide welded seams in the test panel. The results of the small-scale peel tests at various weld widths are shown in Table 4. As can be seen, the difference in strength between the narrowest and the widest seam was only 2% for the 1.5-mm (60-mil) glassmat- supported membrane. The 13-mm- (0.5-in-) wide seam was stronger than the 25-mm- (1-in-) wide seam. For the 1.2-mm (48-mil) polyester sheet, the gap was a little greater at 8%, but the trends were similar. Tripling the width of the weld appears to have only minimal effect on the seam strength of these products. The second test involved assembling a test panel using the same polyester-reinforced PVC membrane used for the lab peel tests for simulated uplift testing at Factory Mutual Engineering. The manufacturer’s mechanically attached listings were, at the SF = T Peel (lbf/in) – 26 = T Peel (kN/m) – 4.5 26 4.5 Formula 1 Where T Peel is the ultimate separation resistance of the seam, measured in lbf/in or kN/m, respectively. www.rci-online.org 8 • IN T E R FA C E MA R C H 2009 Table 4 – Small-scale peel tests at various weld widths. MEMBRANE WELD WELD PEEL PEEL WIDTH (in) WIDTH (cm) (lbf) kN/m 60-mil, PVC, glass mat 0.5 1.3 49 8.5 60-mil, PVC, glass mat 1.0 2.5 46 9.0 60-mil, PVC, glass mat 1.5 3.8 50 8.7 48-mil PVC, polyester 0.5 1.3 50 8.7 48-mil PVC, polyester 1.25 3.1 49 8.5 48-mil PVC, polyester 1.5 3.8 54 9.3 time, all based on a 38-mm (1.5-in) minimum seam width. The 3-m (10-ft) membrane, fastened 152 mm (6 in) o.c., failed at 120 lbf/ft2 as a result of fastener pullout, a result comparable to that achieved with a 38-mm (1.5-in) weld in previous approval testing. As a result of this test, FM modified the manufacturer’s listings to include the following: All currently approved singlewelded, {Manufacturer’s name} mechanically fastened roof cover constructions with a maximum Class 1-105 rating are approved with a minimum 0.5-in (13-mm) wide heat weld placed on the outside edge of the lap. Within the context of the minimal amount of data generated, it appears that for practical purposes, there is little, if any, difference in performance between a properly constructed 13-mm (0.5-in) and a properly constructed 38-mm (1.5-in) wide weld. CONCLUSIONS The study demonstrated that the welding properties of thermoplastic membranes vary significantly, even within a given generic group. Ideal conditions for achieving the strongest weld are very different for every product. In many instances, increasing or decreasing weld speed or temperature even one level can have a dramatic impact on seam strength. Even under ideal conditions, some products provide for little, if any, margin of safety as defined in this paper. Further work done in parallel with field surveys would be required to establish a minimum safety factor that would allow for the various field variables in defining welding parameters for each product under different conditions. However, in evaluating the data generated and considering the practical implications, one must keep in mind the relative contexts of the study and the field in which welding is carried out in practice. All welding was done in a controlled environment, on a uniform substrate, by an experienced, skilled technician. The welding equipment was in excellent condition and fed by a clean, uninterrupted source of power. Ideally, contractors should be working with products that provide the greatest level of safety over the broadest range of weld parameters and material conditions. The results for TPO 2 and PVC E highlight the negative influence surface texture has on weld strength. It is not clear whether the lower results achieved with PVC E are the result of the thin sheet tested in this program, the physical properties imparted by the KEE component of the product formulation, a combination of both, or other factors. As was observed in ancillary testing done in parallel with this program, focusing on weld width alone can be misleading. Ensuring the inner edge of a weld is continuous and straight is more important than the absolute width. Using the proper welding equipment (for example, a welder equipped with a spring-loaded air trap along the inside of the weld to compensate for surface irregularities) and a dedicated power source to minimize energy fluctuations go a long way to achieving the desired weld quality. Wider, inconsistent welds will not compensate for the uneven loads imposed upon irregular inner-weld edges, which can result in pinholing under wind load. The weld width topic merits further work. Further study could also be envisaged to assess the effects of different atmospheric conditions (e.g., high and low tempera- …When You Install The Environmentally-Responsible Duro-Last® Cool Zone® Roofing System. The highly-reflective Cool Zone membrane helps reduce energy consumption across North America. Strategically-located manufacturing facilities and the membrane’s light weight mean less fuel is needed to get it to your job site. The Cool Zone system can often be installed over your existing roof, eliminating a costly tear-off and reducing landfill waste. Our tightly-controlled manufacturing process recycles scrap back into roofing membrane and other products. When your time is right for a photovoltaic system, the Cool Zone system is solar-ready. And after its useful life on your rooftop, the membrane is recyclable. A Cool Zone roof can help in obtaining credits toward LEED and LEED-EB certification. Another plus: Cool Zone roofs can help lessen the urban heat island effect. The Cool Zone roofing system: The best sustainable roofing choice – by leaps and bounds. Sustainable Roofing Is Just A Short Hop Away… To find out more, call us or visit www.white-equals-green.com and request our free brochure. 800-248-0280 “Duro-Last”, “Cool Zone”, and the “World’s Best Roof” are registered marks owned by Duro-Last Roofing, Inc. MA R C H 2009 I N T E R FA C E • 9 tures) on the welding process for different thermoplastic materials. A sufficiently comprehensive study of weld parameters that would allow for a correlation between both laboratory and field measurements might serve as the basis for the development of an ASTM standard. Such a standard, which should cover seam strength, surface preparation, and retention of seam strength as a percentage of original, would no doubt result in more consistent welding in all conditions in the field and, ultimately, better performance. In the interim, contractors can increase their chances of successful installations by working with as few products as possible in order to build up their own experience and knowledge base of a product’s welding behavior under all conditions. ACKNOWLEDGEMENTS The author wishes to express his appreciation to Paul Peterson and John Algird for their preparation and testing of the close-to- 1,000 samples generated in this test program, and their valuable input in analyzing the data. REFERENCES S. Graveline, “Welding of Thermoplastic Roofing Membranes Subjected to Different Conditioning Procedures,” Journal of ASTM International, Vol. 4, No. 8, Paper ID JAI101018, December 2007. K. Liu, A. Baskaran, “Wind Fatigue Ef – fects on the Seam Strength of TPO Roofs,” RSI, “Ther mo plastic Sup ple – ment,” January 2001. M. Russo, “Thermoplastics: Revolu – tionizing the One-Ply Market,” RSI, pp. 17-33, January 2000. T.R. Simmons, D. Runyan, K.K.Y. Liu, R.M. Paroli, A.H. Delgado, J.D. Irwin, “Effects of Welding Para – meters on Seam Strength of Ther – mo plastic Polyolefin (TPO) Roof ing Membranes,” Proceedings of the North American Conference on Roof – ing Technology, pp. 56-65, 1999. Editor’s Note: This article is an abridged version of “Welding of Thermoplastic Roofing Membranes Subjected to Different Con – ditioning Procedures,” by Stanley Graveline, published first in the Journal of ASTM International, Vol. 4, No. 8, Paper ID JAI101018, December 2007. 10 • I N T E R FA C E MA R C H 2009 Stanley P. Graveline is vice president of technical services for Sika Sarnafil, a division of Sika Corp., Canton, MA. He has worked in the roofing industry for more than 20 years in various technical, sales, and management capacities in Canada, Switzerland, and the U.S. Graveline has participated in numerous technical committees and standard-writing bodies in North America and Europe. He is a frequent speaker at national and international symposia and other industry events. Stan is a member of the Professional Engineers Ontario and of RCI. He is currently active on technical committees within the Cool Roof Rating Council, the Chemical Fabric and Film Association, and the National Roofing Contractors Association. He has a bachelor of applied science in chemical engineering from the University of Ottawa and a masters of business administration from the International Institute for Management Development, Lausanne, Switzerland. Stanley P. Graveline The Center for the Polyurethanes Industry (CPI) of the American Chemistry Council issues a call for papers and posters to be presented at the Polyurethanes 2009 Technical Conference on October 5 – 7, 2009, at the Gaylord National in Fort Washington, MD. Paper abstracts are due March 2, 2009. In order to be included in the conference program, poster abstracts must be submitted by April 1, 2009, but abstracts will be accepted until September 1, 2009. Submission instructions, paper guidelines, and templates can be found on CPI’s Web site, www.americanchemistry.com/polyurethane. The Polyurethanes 2009 Technical Conference represents an opportunity to reach an audience of more than 1,500 manufacturers and processors of polyurethane-based products across all industrial sectors. Papers and posters may address innovations and developments in any area of polyurethane, including flexible slabstock, flexible molded, rigid foam, soy-based polyols, polyurethane waterborne dispersions, equipment/machinery, cast elastomers, thermoplastic polyurethanes, integral skin, RIM, binders/fillers, coatings, adhesives, sealants, and more. CPI also encourages paper and poster submissions on breakthrough technology and innovations in polyurethane markets or chemistry, developing markets such as medical and fiberboard, and product areas such as rigid foam insulation or coatings. Presentation submissions also may focus on issues such as combustibility, transportation, recreation, security, as well as environmental and stewardship issues, including emissions, regulatory developments, waste management, and related topics. The highly successful Polyurethanes 2008 Technical Conference saw 22 authors receive awards at the conference for their papers and posters, as judged by their peers. CPI ISSUES CALL FOR PAPERS AND POSTERS
