Roof-System Performance II

The last issue of this column began exploring roof-system performance; this issue continues that discussion.

Performance of Roofing Components and Systems by Charles Hedlin was documentation for a series of seminars conducted by National Research Council Canada – Institute for Research in Construction (NRC-CNRC) during 1989 called “Roofs that Work.” This paper provided a comprehensive overview of this very complex issue.

Topics included:

Roof systems and assemblies, and their components.
Performance of roof components.
Main functions and properties.
Securing the components.
Decks and structure.
Vapor barrier functions.
Thermal insulation.
Additional factors affecting roof performance.

In any mathematical analysis, the number of possible variables greatly affects the results. If variable a has two possibilities, variable b has six, and variable c has 10, then we express this as V_total = a • b • c = 2 • 6 • 10 (or 120 possible results). Looking at Hedlin’s paper from this perspective, variables include:

Roof systems: (2) Conventional and Protected Membrane Systems
Roof membranes, top covers, and types of thermal insulation

Bituminous (BUR): (8) (Asphalt types = 4; Coal-tar types =1; Hot applied, mop, torch, cold adhesive = 3)

Modified bitumens: (14, at least)

APP, SBS, APO, asphalt = 4
Reinforcements woven, non-woven, combinations of both, thickness or weight per square meter in grams = 10

Single-ply membranes: (9 at least)

Polymer type (PVC, TPO, EPDM, CSPE, KEE, CPE) = 6

Thickness of sheets (45, 60, 90 mil) = 3

Reinforcements: (8, at least)

Glass fiber, polyester, woven, scrim, non-woven = 4

Number of layers: one, two, three, more = 4

Thermal Insulations possible

Wood fiber, perlite, cellular glass, glass fiber, urethane/isocyanurate, phenolic foam, polystyrene (extruded or expanded), mineral wool fiber, high density urethane cover board = 11
Differing densities, facers, thicknesses, cover boards = 12 (at least)

Methods of attachment = 5 (at least!)

Thermal insulation attached, membrane adhered
Thermal insulation attached, membrane mechanically attached
Thermal insulation nominally attached to stay in place, membrane loose laid and ballasted
Protected configuration, insulation on top of membrane
Self-adhered

Use of air and vapor barriers (yes or no) = 2
Structural support

Continuity, strength, moisture content, rigidity (deflection), vibration

Surfacing options = 4

Coatings
Factory-applied granules
Ballast
None

As Hedlin points out, each component will have a full range of other properties, such as coefficient of thermal expansion; tensile, compressive, and shear strengths; vapor permeability and moisture accumulation rates; aging characteristics (including embrittlement and reduced thermal resistance); ductility; softening point for asphalt; and flammability. According to Hedlin, “Depending on the circumstances, all of the properties may come into play, anticipated or not. Their effects may be unimportant or significant, beneficial or damaging.”

By now, the reader should get the point. By the time you add the variables for workmanship, abuse, desired aesthetic appeal, and, of course, life and cost, it’s easy to see why we have taken the prescriptive approach.

Let’s fast forward. Recognizing the oil embargo of 1973-1974 was yet another variable not fully anticipated. In 1979, National Bureau of Standards (now National Institute of Standards and Technology) published Building Science Series (BSS) 167: Interim Criteria for Polymer Modified Bituminous Roofing Membrane Materials. Quoting from the abstract of this document:

“… [it] presents the results of a study to develop interim criteria for the selection of polymer modified bituminous roofing membrane materials. The criteria are based on a review of existing standard specifications and related documents. They are intended for use by the construction agencies of the Department of Defense in specifying polymer modified bituminous roofing membrane materials until voluntary consensus standards are developed in the U.S. The suggested interim criteria generally are presented using a performance criteria format. The membrane characteristics for which performance criteria are suggested are dimensional stability, fire, flow resistance, hail impact, moisture content and absorption, pliability, strain energy, uplift resistance, and weathering resistance (heat exposure). Prescriptive criteria for five member characteristics are used to complement the suggested criteria. The approach of using complementary prescriptive criteria is taken to incorporate the performance criteria test methods, which can be relatively rapidly performed for characterization or identification of the membrane material. Other membrane requirements are listed for future development of criteria, but performance criteria for these requirements are not suggested at present. Lack of a consistent database in the existing standards and related documents precludes suggesting criteria at this time. It’s considered beneficial to present the needed criteria as a first step toward directing future research for standards development for polymer modified bituminous roofing membrane materials.”

Suggested criteria for modified bitumens included:

Dimensional stability – maximum 1 percent.
Fire resistance – meet FM Global, UL, code.
Flow resistance – no slippage.
Hail impact – no leakage after impact 30J falling dart or 1.5-inch diameter hailstone
Moisture content – 0.5 percent maximum.
Water absorption – 1 gram for specified size of specimen.
Low-temperature flex – no cracking at expected low temperatures.
Strain energy – not less than 3 pound force (lbf) • inch/inch² at 0 degrees F. (This is also called work-to-break, the area under the curve of stress and strain. Just as with the criterion of low-temperature strength for BUR, strain energy has been the most quoted criterion of the BSS 167 document.) At this time, this property is rarely cited, and most MB manufacturers are at a loss to explain why they have so many products in the marketplace with differing weights, amount, and type of reinforcement, or even whether combinations of polymer-modified cap sheets, for example with non-modified base sheets, makes sense or not.

The good news is that, since BSS 167 was published, ASTM has managed to publish a number of prescriptive specifications for modified bitumens:

D 5602: Static Puncture Resistance of Roofing Membrane Specimens.
D 5635: Dynamic Puncture Resistance of Roofing Membrane Specimens.
D 5849: Evaluating the Resistance of Modified Bituminous Roofing membranes to Cyclic Joint Displacement.
D 6950 Application of APP Systems.
D 6135 Application of Self-Adhering MB Systems.
D 6509 APP Modified Base Sheet.
D 6222 APP with Polyester Reinforcement.
D 6223 APP Using Combinations of Glass & Polyester Fibers.
D 5849 Test Method for Cyclic Fatigue.
D 6298 SBS with Metal Foil Surfacing.
D 7051 Test for Thermal Cycle Shock of Foil-Faced Sheets.

(In future columns in this series, we will look at existing standards for SPF, TPO, PVC, metal roofing, etc.)

In 1979, Centre Scientifique et Technique du Bâtiment published FIT Classification for Roofing Systems. According to this document, “FIT is a performance-based classification for roofing systems. It’s mainly aimed at architects, engineers, and users and is, therefore, simplified in its design to be easy to understand and use . . . [it] should enable suitable roofing systems to be chosen for a given use based on major performance criteria regarding the stresses to which the systems are subjected."

This is where the first significant strides were made on the performance concept: a group of experts, based on their experience and knowledge, have classified performance of roofs into levels of F, I, and T: F for fatigue, I for indentation (puncture), and T for temperature. For example, if a roof membrane is fully exposed to service traffic on the roof, it requires a higher degree of impact resistance than a protected membrane roof system would. Looking at indentation, the FIT systems looks at two conditions: static puncture, such as a concentrated load due to heavy A/C units placed on sleepers, and dynamic puncture, such as a dropped toolbox. (ASTM test methods for these conditions are now available.)

For Static Load L:

Classification	Failure Load, N	Examples of systems
L1	<5	Marginal roofing
L2	>5, <15	Glass-reinforced MB
L3	>15, <25	Polyester-reinforced MB
L4	>25	Heavy polyester reinforced PVC or EPDM

For Dynamic Impact:

For Fatigue Resistance: (Level 5 is the highest level of resistance to damage)

Classification	Examples of systems
D1	<10	Marginal systems
D2	10 to 20	BUR & MB min 4mm thickness PVC and EPDM min 1.2 mm thickness
D3	>20	Special systems for car parks, roof gardens

Bitumen type	Reinforcement of Base Layer	Cap Layer	Fatigue Resistance Level
Oxidized	180 g polyester	same	2
Oxidized	60 g glass mat	APP/180 g polyester	2
Oxidized	200 g woven glass	SBS/ 250 g polyester	4
SBS	60 g glass mat	SBS/180 g polyester	4
SBS	180 g polyester	SBS/180 g polyester	5

Note in this FIT classification method: Any roof system that meets these requirements is permitted. It also provides a logical way to differentiate levels of performance based upon anticipated project conditions – something that individual modified bitumen manufacturers have been unable to do.

The next column in this series will update you on the industry’s progress since the issuance of these documents.

Resources: