More than a few times, the issue of a “brittle test” has come across my desk. Misuse is becoming more prevalent for both the term and the test as the concept gains momentum among folks hoping to discount the reparability of a shingle roof. A casual web search will validate the gross confusion surrounding this matter. I even saw the brittleness aspect recently applied to a wood shake roof, generating consternation and prompting development of this treatise. The procedure is a qualitative measure, presumably intended to characterize shingles with regard to carrying out localized repairs. The problem is that the procedure is touted as a bona fide determination by an installer who clearly has interest in the outcome. It is somewhat analogous to the equally useless “finger test” for evaluating the bond among sealing tab shingles. But that is another topic for another time. Meanwhile, the bigger problem is: There is no such thing as a “brittle test” for asphalt shingles (or asphalt pavement, BUR, modified bitumen, or anything else made from asphalt). That is because there is no singular identifiable temperature at which asphaltic materials become brittle. Some folks like to toss the term around, but the test simply does not exist in any recognized forum. Asphalt is neither a thermoplastic nor a thermoset; it is, instead, a viscoelastic material exhibiting rheological properties. That is, asphalt is a non-Newtonian fluid whereby every test conducted on it is temperature-dependent—and highly so. The behavior of bitumen (asphalt for our purposes here) merits further discussion. The freezing and boiling points of water are both well known, even to grade-school students. Asphalt, however, makes a transition from solid to semisolid and, eventually, to liquid over a wide range of temperatures. Such behavior can be plotted to express the glass transition, but there is no finite marker at which the material becomes liquid or solid. Yet for grading and selecting the appro- 3 6 • I n t e r f a c e S e p t e m b e r 2 0 1 6 Figure 2 – Wet soil is not a Newtonian fluid. Following inward wall collapse from hydrostatic pressure, all of the soil does not self-level; a portion of it remains in a somewhat undisturbed position. Figure 1 – Softening point testing underway by the author around 35 years ago. priate material (such as for slope limitations), a uniform index was needed. Part of this characterization is seen in the softening point test (not the “melting point,” as it is sometimes erroneously called; the melting point applies only to crystalline homogenous substances, and asphalt is neither of these). In the case of roofi ng, ASTM standard D361 is the uniform standard for such evaluation. Figure 1 depicts such testing underway by the author around 35 years ago. Roofi ng asphalt specimens can be characterized and graded when evaluated by this method, uniformity being the entire purpose of the standards organization, ASTM International. For those unfamiliar with Newtonian fl uids, consider a tall basement wall, improperly constructed and backfi lled with poorly draining soil. It may eventually collapse inward from hydrostatic pressure, but all of the wet soil does not self-level. A portion of it remains in a somewhat undisturbed position (Figure 2); therefore, wet soil is not a Newtonian fl uid. Water in a swimming pool would be Newtonian, exerting pressure against the sides in a directly linear fashion proportional to the water depth. The behavior of asphalt can be seen when the kettle man chops and opens containers in preparation for loading into the heating vessel. Some of the material may or may not deform and fl ow from the carton sides, depending on ambient t e m p e r a t u r e (Figure 3). The carton contents may indeed be brittle during the cold season, and semi-fl uid at other times. When practitioners of the brittle test are encountered, a few questions come to mind: • Is the roof brittle in December or July? • Are molasses and honey brittle in the wintertime? How about in September? • A motorcycle kickstand sometimes S e p t e m b e r 2 0 1 6 I n t e r f a c e • 3 7 RCI, Inc. 800-828-1902 Enroll today @ rci-e-learning.org • Demonstrate to your clients the value of working with you — an RCI member. • Show your commitment to education and integrity with a gifted subscription to RCI Interface, technical journal. RCI members can now give clients or colleagues an RCI Interface Gift Subscription for only $35 a year (U.S. only). The recipient’s first issue will be accompanied by a letter detailing the benefactor of the gift subscription. At the end of the year, the subscriber will be contacted to become a member or to renew the subscription at the regular annual subscription rate. Introducing the RCI Interface Gift Subscription Give the Gift of RCI Interface today! To purchase, login to your Member Account and RCI, Inc. 800.828.1902 rci-online.org search the Online Store for RCI Gift Subscription. Figure 3 – Asphalt is also non-Newtonian. Some of the material may deform and fl ow from the carton sides, depending on ambient temperature. digs into hot asphalt pavement in July, but is otherwise fi ne in December. During which month did the pavement become brittle? If the coming warm season would make the pavement convert back to being resilient, wouldn’t it do the same for an asphalt shingle roof? In order to be valid, any test must be 1) random, 2) representative, and 3) repeatable. That applies for soil, concrete, steel, aggregates, wood, plastics, fasteners, asphalt pavement, and any kind of roofi ng material that can be named. The so-called brittle test for shingles—often administered by folks of dubious qualifi cation—is in no way repeatable. Unless it’s carried out on all the directional exposures, it is not representative. When individuals press the point, I simply ask for the ASTM test number; things suddenly go quiet. If the individual doesn’t know what ASTM is, that just makes my point. There can be low-temp fl exibility testing, which can be very helpful in evaluating bituminous and pure thermoplastic membrane products (Figure 4). But, again, these results are highly infl uenced by temperature as well as by the mandrel size being used.2 Without uniform constraints for the test procedure and environment (including conditioning of the specimen beforehand), these evaluations would mean nothing. And they mean nothing when carried out on a pitched roof during any season by all manner of talent. There is simply no meaningful basis for an outdoor brittle test of shingles. Regarding the reparability of localized wind-damaged shingles, there is certainly a condition whereby old, weathered shingles are too deteriorated to be disturbed (Figures 5A, 5B, and 5C). That point is well taken, as many roofs are so badly neglected and replacement has so long been postponed that shingles may tear or pulverize— even from light contact. That is all that needs to be said. On the other hand, if shingles are resilient enough (during whatever season) that tabs can be lifted for determination of the fastening pattern, they can certainly be lifted enough to facilitate repairs. This apparently runs counter to the brittle test, which is completely subjective and should be given a decent burial. REFERENCES 1) ASTM D36, Standard Test Method for Softening Point of Bitumen (Ringand- Ball Apparatus). 2) Lyle D. Hogan. “The Province of Testing.” Western Roofi ng. September/ October 1996. pp. 66-68. 3 8 • I n t e r f a c e S e p t e m b e r 2 0 1 6 Lyle Hogan is owner and principal engineer of Fincastle Engineering, Inc., Greensboro, NC. He is a registered engineer in fi ve states, a Registered Roof Consultant, a Fellow of RCI, and an ICC structural masonry inspector. He has designed and administered roofi ng projects in half of the U.S. using a variety of systems. Hogan has received RCI’s Lifetime Achievement Award, its Michael DeFrancesco Award, its William C. Correll Award, and its Richard M. Horowitz Award. Lyle D. Hogan, RRC, FRCI, PE Figure 4 – Low-temp fl exibility testing can be very helpful in evaluating bituminous and pure thermoplastic membrane products. But these results are highly infl uenced by temperature as well as by the mandrel size being used. Figures 5A, 5B, and 5C – Attempting to repair these shingles would be futile. Yet it would be inaccurate to characterize them as brittle. They are instead highly weathered, impractical to repair, and should be in a landfi ll.
