Page images
PDF
EPUB

hailstones on the solidly supported areas. Similarly, three other shingles, all Class A, based on glass mat felts showed no felt damage on their obverse sides with hailstones below 2 in (5.1 cm) in diameter; one of these had a damage threshold at the 21/2-in (6.4 cm) hailstone on all three portions of its surface. Some of the conventionally made heavy shingles, usually with Number 9 granules or with high concentrations of mineral additives, performed equally well. One Type 290 Class C shingle actually had a damage threshold at the 234-in (7.0 cm) hailstone. While it is outside the province of this report to identify these heavy Class C and Class A shingles more specifically, the manufacturers have been informed of how their individual products performed and the basic principles required to make more hail-resistant products. All of these shingles were mounted on 12-in (1.3 cm) plywood.

Because the vast majority of hailstorms occur in warm weather, the roofs are above ambient temperatures when the hailstorm starts. Hailstorms that produce large hailstones are always of short duration and are preceded by a cloud cover, which drops the roof temperatures below their daily highs. Therefore, the hail resistance evaluation was conducted at 75 to 80°F (24 to 27°C). However, one Type 235 and one Type 315 Class C shingle and one Type 240 Class A shingle were tested at 120°F (49°C). on a 1/2-in (1.3 cm) plywood deck with a 15 lb saturated felt underlayment. The hail resistance of the Type 235 shingle was increased to the 212-in (6.4 cm) hailstones from the 12-in (3.8 cm) hailstones on all three surfaces. That of the Type 315 shingle was improved only on the unsupported areas and that of the Type 240 was not changed. Thus, the results on these three shingles indicate that shingles tend to be more resistant to hail damage at higher temperatures. It is fortunate that hailstorms occur in warm weather.

4.2. Built-Up Roofs

Occasionally in residential construction and much more frequently in commercial and industrial construction relatively flat roofs are used. These roofs are not "factory manufactured," but "built up" on the site from alternate layers of bitumen and reinforcing membranes. Some of these roofs are surfaced with a smooth layer of bitumen and others are surfaced with a layer of pebbles, crushed stone or light weight aggregate particles in addition to a layer of bitumen. There are many variations of this type of roof system; only a few representative ones were tested. The construction of these roofs and the results of the hail-resistance tests are summarized in table 3.

The conventional smooth-surface built-up roof [la and le in table 3] on a dense deck showed visible signs of damage; i.e., cracking of the surface, when 2-in (5.1 cm) hailstones were used. Smaller hailstones usually indented the flood coat, but did not crack it. When fiberboard (1b) or glass fiber (1g) insulation was installed between the deck and the roof membrane the indentations were larger and coating cracks appeared with 13/4-in (4.5 cm) hailstones. The roofing on Foamboard A insulation (1c) performed better than on the dense decks when 2-in (5.1 cm) hailstones were used, but 211⁄2-in (6.4 cm) hailstones penetrated through the roofing into the insulation. Foamboard B (1d) delaminated; i.e., the insulation broke away from its protective asphalt coated felts, when impacted with 2-in (5.1 cm) hailstones. The roofing on glass fiber insulation (1g) on steel decking was penetrated by 21/2-in (6.4 cm) hailstones.

The flood coat of the built-up roof made with asbestos felts on a plywood deck (2a) did not crack or become indented by 212-in (6.4 cm) hailstones; however, the flood coat was indented and cracked by 2-in (5.1 cm) hailstones when fiberboard insulation (2c) was used between the membrane and the deck. The asbestos-felt roofs had better hail resistance than the rag felt built-up roofs on comparable decks.

The built-up roofs made with coal tar pitch [3], referred to as tar in table 3, did not indent, but developed concentric cracks with all sizes of hailstones. The 21/2-in (6.4 cm) hailstones caused some of the flood coat to spall from the top felts [3a]. Coal tar pitch generally tends to be more brittle than asphalt and would be expected to respond to the hail impact as a brittle material.

The roofs built up with glass fiber felts on the dense decks (4a and 4b) (plywood and asbestos cement) did not experience flood coat cracking with hailstones 212-in (6.4 cm) in diameter and smaller, but when insulation was present (4c. 4d, 4e, and 4f), cracks were produced with 21/2-in (6.4 cm) hailstones. The glass felt roofs fell in between the organic felt builtup roofs and asbestos felt built-up roofs in hail resistance.

The roofs constructed of two base sheets (5) performed much better on the asbestos cement deck (5b) than on plywood (5a); their performance on plywood or insulation were about the same as conventional asphalt-organic-felt built-up roofs on the same substrates. Where these roofs were covered with 300 lbs/sq of slag (14.7 kg/m2) (6), no damage was done to the roof membrane by any of the hailstones. (The slag was retained by cheese cloth when the decks were tested in vertical positions). The hailstone energy was dissipated in scattering

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][ocr errors][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][ocr errors][merged small][merged small][merged small][merged small]

(4a)-in (1.8 cm) Plywood..

on

(4b) 1-in (2.5 cm) Asbestos cement

(4c) 1-in (2.5 cm) Fiberboard on in (1.3 cm) plywood.......... (4d) 1-in (2.5 cm) Fiberboard on 1-in (2.5 cm) asbestos cement. (4e) 4-in (1.9 cm) Glass fiber insulation on in (1.3 cm) plywood....

(4f) 4-in (1.9 cm) Glass fiber insulation on 1-in (2.5 cm) asbestos cement..

(5) 2 Base sheets +60 lbs/sq (2.9 kg/m3) asphalt flood coat (20-25 lbs/sq (1.0–1.2 kg/m2) interply asphalt]

[blocks in formation]
[ocr errors]
[ocr errors][ocr errors][ocr errors][ocr errors][merged small]

• Mean diameter of indentation. b C, Surface cracked. D, Foamboard delaminated. F, Felts cracked. shattered. -, Not tested.

the slag; "nests" of various sizes were left in the slag.

In summary, each roofing membrane performed better on the denser substrates than on the lighter substrates, the roofings made with inorganic felts performed better than those made with organic felts and the slag surfaced roofing was not damaged by hailstones up to and including 212-in (6.4 cm) in diameter.

[ocr errors]

CS

C

(1.3)

720

2 C

[ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][subsumed][merged small][merged small][merged small][ocr errors][ocr errors][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors]

N, No visible indentation. P, penetrated roofing. S, Coating

4.3. Nonbituminous Roofing

A number of nonbituminous roofings were tested for comparison purposes. These were applied in accordance with their supplier's recommendations. The levels of failure used in these evaluations were cracking for brittle roofings and objectionable indentations for metal roofing. Table 4 is a summary of the results of these tests.

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

• Dents proportional to hail size-visible for all hailstone sizes. The plywood deck cracked below the dents with hailstones larger than 21⁄2 in (6.4 cm). Note tested. All roofings tested were vulnerable to hail damage. As with the asphalt shingles, these other products contained areas of different vulnerability.

5. Conclusions

(a) All roofing materials have some resistance to hail damage, but as the size of the hail increases, a level of impact energy is reached at which damage occurs. This level lies in the area of 12 to 2-in (3.8-5.1 cm) stones for most prepared roofings.

(b) Because of the ways in which prepared roofings are applied, most products have areas of different vulnerability.

(c) Heavier shingles tend to be more hailresistance than Type 235 shingles.

(d) Weathering tends to lower the hail resistance of asphalt shingles.

(e) The solidly supported areas of roofing tend to be the most resistant to hail damage.

(f) Built-up roofs on hard substrates tend to resist hail better than those on soft substrates.

(g) Built-up roofs made with inorganic felts tend to be more hail resistant than those made with organic felts.

(h) Coarse aggregate surfacing tends to increase the hail resistance of roofing.

(i) The slate, asbestos cement and tile roofings tested contained areas of different vulnerability and cracked under the impact of 112- to 2-in (3.8-5.1 cm) hailstones.

(j) The sheet metal roofing was dented by all sizes of hailstones used. Deck cracking occurred when 212-in (6.4 cm) hailstones were used.

This work was sponsored by the Research Committee of the Asphalt Roofing Manufacturers Association, under the Research Associate Plan, at the National Bureau of Standards. The author wishes to thank Royce Stine and Thomas Crowe for their help in constructing the apparatus and collecting the data upon which this article is based. He also thanks the Materials Durability and Analvsis Section, and particularly its Chief, W. C. Cullen, for their help and encouragement that made this work possible.

6. References

[1] Storm Data, Environmental Science Services Administration, U.S. Department of Commerce, Asheville, N.C.

[2] Douglas, R. H., Recent hail research: A review, Meteorol. Monographs 5, 157-167 (1963). [3] Werckmann, H., The Language of Hailstorms and Hailstones, ÚSASRDL Rep. 2277, U.S. Army Research & Development Laboratory, Ft. Monmouth, N.J. (1962).

[4] Schleusener, R. A., Hailstorm characterization and the crystal structure of hail, Meteorol. Monographs 5, 173-176 (1963).

[5] Texas Hailstorms, Texas State Board of Insurance, Austin, Texas.

[6] Schumann, T. E., The theory of hailstone formation, Quart. J. Roy. Meteorol. Soc. 64, 3-4 (1938). [7] Flora, S. D., The Serious Toll of Hail Storms, Rough Notes 82, 14, 1939.

[8] Fuchs, E. R., Hailstorm, American Roofer 10 (Nov.
1955).
[9]

Laurie, J. A. P., Hail and Its Effects on Buildings,
C.S.I.R. Research Report No. 176, Bull. 21, 1–12,
National Building Research Institute, Pretoria,
South Africa (1960).

[10] Browning, K. A., F. H. Ludlam, and W. C. Lacklin, The density and structure of hailstones, Quart. J. R. Meteorol. Soc. 89, 75-84 (1963) Roy. meteorol.

[11] Bilham, E. G., and E. F. Relf, The dynamics of large hailstones, Quart. J. Roy Meteorol. Soc. 63, 149-162 (1937).

[12] Manufacture, Selection and Application of Asphalt Roofing and Siding Products, Ninth edition, 1966, Asphalt Roofing Industry Bureau, New York, N.Ŷ.

✩U.S. Government Printing Office: 1969 0-348-822

(cut here)

Announcement of New Publications in

Building Science Series

Superintendent of Documents,

Government Printing Office,

Washington, D.C., 20402

Dear Sir:

Please add my name to the announcement list of new publications to be issued in the series: National Bureau of Standards Building Science Series.

[blocks in formation]
« PreviousContinue »