Durability of Polymer Modified Asphalt Shingles

Abstract

The Insurance Institute for Business & Home Safety (IBHS) Research Center is a multi-peril applied research and training facility in Richburg, South Carolina. The IBHS Research Center tests building components and systems against natural hazards including wind, hail, wildfire, and wind-driven rain. In 2013, IBHS began investigating the impact resistance and wind performance of asphalt shingles. Preliminary results indicated polymer modified asphalt (PMA) impact resistant (IR) shingles may perform better than basic oxidized and traditional IR oxidized shingles in both impact and wind tests. This study seeks to investigate the mechanical and physical properties of asphalt shingles that may cause increased impact and wind resistance. IBHS is collaborating with the asphalt shingle industry and other labs to understand possible relationships between binder properties and durability, performance, oxidation and modification. The project seeks to define components and characteristics of, and to develop a minimum standard for PMA shingles. Work is ongoing, and results presented here are preliminary, serving as indicators for progress in relating durability testing to analytical and materials testing.

Full text

1
Durability of Polymer Modified Asphalt Shingles
1
Heather E. Estes1, Tanya M. Brown-Giammanco, Ph.D.2, Ian M. Giammanco, Ph.D. 3
2
1–3 Insurance Institute for Business & Home Safety, 5335 Richburg Rd. Richburg, SC 29729,
3
USA, hestes@ibhs.org, tbrown@ibhs.org, igiammanco@ibhs.org
4
5
Abstract
6
The Insurance Institute for Business & Home Safety (IBHS) Research Center is a multi-peril
7
applied research and training facility in Richburg, South Carolina. The IBHS Research Center
8
tests building components and systems against natural hazards including wind, hail, wildfire, and
9
wind-driven rain. In 2013, IBHS began investigating the impact resistance and wind performance
10
of asphalt shingles. Preliminary results indicated polymer modified asphalt (PMA) impact
11
resistant (IR) shingles may perform better than basic oxidized and traditional IR oxidized
12
shingles in both impact and wind tests. This study seeks to investigate the mechanical and
13
physical properties of asphalt shingles that may cause increased impact and wind resistance.
14
IBHS is collaborating with the asphalt shingle industry and other labs to understand possible
15
relationships between binder properties and durability, performance, oxidation and modification.
16
The project seeks to define components and characteristics of, and to develop a minimum
17
standard for PMA shingles. Work is ongoing, and results presented here are preliminary, serving
18
as indicators for progress in relating durability testing to analytical and materials testing.
19
Keywords: Asphalt shingles, polymer modified, oxidized, wind performance, impact
20
performance.
21
1. Introduction
22
Polymer modified asphalt (PMA) shingles utilize synthetic polymers such as styrene butadiene
23
styrene (SBS) to increase flexibility. Typical oxidized asphalt shingles are brittle and stiff when
24
compared to PMA shingles since they are air-blown asphalts. Polymer modification allows the
25
asphalt in PMA shingles to exhibit increased elastomeric properties [1], which is thought to
26
decrease granule loss and result in greater durability. To investigate this hypothesis, IBHS has
27
undertaken a collaborative investigation to relate binder properties to the durability and
28
performance of asphalt shingles.
29
30
Some impact resistant (IR) shingles achieve impact resistance by modifying the binder with
31
polymers (PMA/IR shingles), while others obtain it by including a scrim or mesh on the back
32
side of an oxidized shingle (traditional IR shingles), which can improve tear resistance. In 2013,
33
IBHS began investigating the impact resistance of asphalt shingles using the UL 2218 steel ball
34
impact test. Initial tests of 22 off-the-shelf asphalt shingles, including three PMA shingles,
35
indicated PMA/IR shingles performed better than non-IR shingles and traditional IR oxidized
36
shingles. Additionally, IBHS conducted ASTM D3161 wind tests to assess the wind performance
37
of 26 off-the-shelf asphalt shingles, including four PMA shingles. Results from these initial tests
38
indicated the sealant strip adhesion was key in determining the wind resistance. However, if the
39

2
shingles became unsealed during the tests, PMA shingles were better able to reseal and self-heal
1
without creasing or cracking like the oxidized asphalt shingles.
2
3
In this project, impact and wind tests are being conducted on additional PMA shingles, for two
4
replicates of each product. IBHS is also collaborating with asphalt shingle manufacturers and
5
other labs to determine the mechanical and physical properties that may cause increased wind
6
and impact resistance of the modified shingles, and to evaluate how products on the market may
7
resist damage during hailstorms and high-wind events.
8
9
Table 1 lists the architectural PMA/IR shingles that have been tested at IBHS. Shingles were
10
purchased through a third party vendor and distributed by freight shipping. All products were
11
labelled as UL 2218 Class 4 IR and Class F, or 177 km/h, wind compliant. Manufacturing dates
12
are provided to indicate the relative age of the samples, since in-package aging and distribution is
13
thought to have an effect on the performance compared to when the products are tested for
14
certification straight from the factory. Some oxidized shingles from each manufacturer were also
15
tested in previous studies, such that performance comparisons can be made for this study.
16
17
Table 1. Architectural PMA IR shingles tested at the IBHS Research Center.
18
Shingle
Product
Manufacturer
Manufacturing Date
A**
1
5/22/2014
B**
2
7/28/2014
C**
3
8/13/2014
D**
4
2015
E*
5
2013
F*
1
5/2013
*Products tested in 2014, **Products tested in 2015.
19
20
2. Test Methods
21
2.1. Impact Tests
22
UL 2218, developed by Underwriters Laboratory [2], is a steel ball impact test used to rate the
23
impact resistance of asphalt shingles and other roofing materials. FM 4473, developed by FM
24
Approvals, is similar to UL 2218 but uses ice spheres made from distilled water propelled at
25
speeds necessary to reach appropriate impact kinetic energy (Table 2) [3]. Products rated as
26
Class 4 under these tests have the highest impact resistance.
27
For this study, asphalt shingles were installed on 0.91 m by 0.91 m panels according to the
28
manufacturers’ instructions. Panels were then conditioned, stored, and tested according to UL
29
2218 and FM 4473 requirements. Tests were conducted with two impacts at each impact location
30
on the test panels. Impact locations were areas that tended to be more susceptible to damage,
31
such as edges, corners and joints. Tests were conducted for four different classes of steel balls
32
dropped at prescribed heights (UL 2218), or ice balls propelled at speeds necessary (FM 4473) to
33
obtain appropriate impact kinetic energies as outlined in Table 2. Impacted areas were then
34

3
carefully removed, bent over a 102 mm mandrel, and inspected. Panels tested using UL 2218
1
were inspected under a microscope using 7.5x magnification. Any evidence of cracking, tearing
2
or fracturing on the back side of the shingle was counted as a failure. Panels tested using FM
3
4473 were inspected with the naked eye, and classified as a failure if there was any evidence of
4
cracking, tearing or fracturing on the front or back side of the shingle.
5
Table 2. UL 2218 and FM 4473 impact test specifications. Steel balls (UL 2218) are
6
dropped from prescribed heights to create the required kinetic energies. Ice balls (FM
7
4473) are pneumatically propelled at speeds necessary to create the required kinetic
8
energies for each class.
9
Class
Projectile Diameter (mm)
Kinetic Energy (J)
UL
FM
1
31.8
4.78
5.0
2
38.1
9.95
10.4
3
44.5
18.37
19.0
4
50.8
32.12
32.2
10
11
2.2. Wind Tests
12
Wind and uplift resistance were tested using a modified ASTM D3161 fan-induced wind test for
13
steep-slope roofing products [4]. Asphalt shingles were installed on 1.27 m by 1.68 m panels
14
according to the manufacturers’ instructions. Panels were conditioned using a modified
15
conditioning method of solar radiation for ten days, in which most shingles on the panels reached
16
71°C. The panels were then stored and tested according to ASTM D3161 requirements, using a
17
steady speed wind for two hours. A failure was defined as lifting, cracking, creasing, splitting or
18
tearing of the shingle. If the panels passed their Class F (177 km/h) rating, additional testing was
19
conducted at 16 km/h intervals from 193–274 km/h. All tests were performed outside at 24 +/-
20
3°C and stopped when a failure was observed. After a failure was documented, the ability of the
21
shingle to recover and re-seal was evaluated following an additional 10 days of solar radiation
22
exposure.
23
24
2.3. Material Tests
25
Asphalt shingles meet a variety of material property standards, as outlined in ASTM D3462, [5].
26
Portions of ASTM D3462 were conducted on bundles of PMA and oxidized asphalt shingles.
27
Several labs collaborated to conduct physical and analytical tests to quantify material differences
28
between oxidized and PMA shingles. The tests conducted included ASTM D1922 tear strength,
29
ASTM D3462 fastener pull-through resistance, ASTM D4977 granule adhesion, ASTM D4073
30
tensile strength, and Fourier Transform Infrared Spectroscopy (FTIR) scan.
31
Tear strength was performed using an Elmendorf tear tester. Fastener pull-through was
32
conducted on a universal testing machine with appropriate attachments. Resistance to granule
33
loss was conducted using a granule adhesion test machine, which uses a steel brush to scrub the
34

4
surface of the shingles. Adhesion in this test method refers to the bond between the mineral
1
granules and the shingles surface. Failures were determined when results fell outside the
2
specified range as set by the standard. Material differentiation between polymer modified and
3
oxidized shingles were observed in the granule adhesion and tensile strength results. On average,
4
the PMA shingles had improved granule adhesion and increased tensile strength.
5
6
2.4. Analytical Tests
7
FTIR can be used to characterize and create a fingerprint spectra of an asphalt binder, which
8
could be used to differentiate between oxidized and PMA shingles [6]. FTIR does this by
9
displaying the absorbance of infrared (IR) wavelength emissions of a solid, liquid or gas. When
10
the IR radiation is passed through the sample, different molecules absorb and transmit the
11
energy. Different molecules will react by bending, stretching or vibrating, which creates peaks
12
that can be used to indicate key functional groups in the substance. Those peaks indicate
13
different bond structures that can be compared to other spectra to help identify the chemical
14
makeup.
15
16
3. Results
17
3.1. Impact Durability
18
In a previous laboratory study, IR shingles were shown to have improved UL 2218 performance
19
when compared to non-IR shingles [7]. In addition, PMA/IR shingles indicated slightly higher
20
passing rates than traditional oxidized IR shingles [7]. It should be noted that all IR shingles
21
tested were listed to pass the UL 2218 Class 4 test at the time of manufacture, but none passed
22
for 100% of the impact locations. One possible explanation is that after manufacturing, an
23
asphalt shingle’s physical properties may change due to aging, thermal conditions, pressure from
24
stacking, and handling during distribution. By the time the tests are conducted in a third-party
25
lab, results may be different as a result of property changes. Another possible explanation is a
26
lapse in time between when the shingles were originally listed as IR following the UL test and
27
when the product was manufactured; the lapse could be several years. UL has recently increased
28
the frequency and stringency of its follow up service for UL 2218, such that IR shingles are re-
29
tested every six months, as opposed to just once when the product is first listed.
30
31
For the purposes of Figure 1, the results of the UL 2218 tests from the previous study [7] were
32
combined with results for four additional architectural PMA/IR products A-D in Table 1. Figure
33
1 displays the percentage of impact locations with passing ratings by class for the UL 2218 steel
34
ball test for the shingles of each type grouped together as outlined in Table 3. This indicates IR
35
oxidized architectural and the thicker, premium oxidized architectural shingles (35%)
36
demonstrated a slightly higher passing rate for Class 4 when compared to basic oxidized
37
architectural shingles (32%). PMA/IR shingles have even higher passing rates at 48%.
38
39

5
1
Figure 1. UL 2218 steel ball impact test passing rates for the different types of architectural
2
asphalt shingles tested at IBHS.
3
4
Table 3. Number of architectural shingles tested at the IBHS Research Center.
5
Type of Shingle
Number of Products Tested
Basic Oxidized Architectural
5
IR Oxidized Architectural
3
Premium Oxidized Architectural
5
IR Polymer Modified Architectural (PMA/IR)
6 (as listed in Table 1)
6
FM 4473 ice ball impact tests were conducted for most of the PMA/IR shingles. Figure 2 shows
7
the Class 4 steel ball (UL) passing percentage for individual products compared to the ice ball
8
(FM) passing percentage for individual products. The two methods use different performance
9
rating systems; UL 2218 uses a microscope to identify if there is a crack or tear on the back of
10
the shingle, while FM 4473 does not use a microscope, and quantifies a failure as a crack or tear
11
anywhere on the shingle. Results showed that UL 2218 (steel ball) passing results were higher
12
than the FM 4473 (ice ball) passing results by as much as 43%.The rating of cracks or tears
13
anywhere on the shingle as test failures under the FM 4473 method may explain why the passing
14
rates were lower than the UL 2218 results. A large range of passing rates for both UL 2218 and
15
FM 4473 were observed between different products, even though they are all PMA/IR shingles,
16
which is consistent with individual UL 2218 product performance data collected in the 2014
17
IBHS study (individual results not published), which also indicated a large range of UL 2218
18
passing rates between oxidized products, particularly for Class 1 and 2 tests. These observations
19
support the need for understanding the relationship between product characteristics and
20
performance. Furthermore, this highlights the importance of defining a standard, or
21
characteristics, of PMA shingles that will lead to better prediction of real-world performance.
22
23
55%
69%
60%
83%
42%
61%
43%
76%
33%
56%
38%
64%
32% 34% 35%
48%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Basic oxidized
architectural
IR oxidized
architectural
Premium oxidized
architectural
IR polymer modified
architectural
Passing Percentage
UL 2218 Impact Testing Passing Rates by Shingle Type
Class 1 impacts
Class 2 impacts
Class 3 impacts
Class 4 impacts

6
1
Figure 2. UL 2218 and FM 4473 Class 4 impact testing results for PMA/IR shingles tested
2
at IBHS as outlined in Table 1.
3
4
3.2. Wind and Uplift Durability
5
Wind resistance of asphalt shingles can generally be attributed to the bond strength of the sealant
6
strip. Dixon et al. highlighted the placement of the sealant strip relative to a shingle’s leading
7
edge varies, and will affect the wind uplift force on the surface of the shingle [8]. Typical sealant
8
strips are 12–20 mm wide and are applied in different patterns along the shingle edge when they
9
are manufactured at the factory. If the sealant strip fails to cure properly (Figure 3A), bonding
10
between the shingle and the sealant strip may not fully occur, leading to a wind performance
11
vulnerability. Initial tests indicated that PMA shingles were not more likely to pass ASTM
12
D3161 test when compared to oxidized shingles. However, if shingles sealed (Figure 3B) but
13
later failed, oxidized asphalt shingles were more likely to tear or crack, likely because they are
14
more brittle than PMA shingles (Figures 3E and 3F). After a sealant failure, PMA shingles
15
would bend, but not crack or tear (Figures 3C and 3D). Therefore, PMA shingles may be able to
16
re-seal with little or no surface damage following the wind event.
17
18
19
Figure 3. (A) Sealant strip fails to cure properly. (B) Thermally activated seal strip. (C) and
20
(D) Minimal cracking on a PMA/IR architectural shingle after lifting during ASTM D3161.
21
42%
21%
45%
75%
50% 50%
4% 8%
21%
32%
14%
0%
20%
40%
60%
80%
100%
A B C D E F
Passing Percentage
Class 4 Passing Percentage for the Impact Performance of
Polymer Modified Asphalt Shingles
Class 4 UL
Class 4 FM

7
(E) and (F) Cracking on an oxidized architectural shingle after lifting during ASTM
1
D3161.
2
3.3. Material Properties
3
Table 4 displays results from mechanical tests that were performed on three PMA/IR shingles
4
from Table 1, and one oxidized IR architectural shingle for comparison. Tear strength was higher
5
in the oxidized IR architectural shingle, which was most likely due to the scrim on the back.
6
Fastener pull-through did not appear to show consistent results that distinguished between
7
oxidized IR and PMA/IR shingles. Granule adhesion is important for the durability of asphalt
8
shingles because the granule coatings protect the binder from UV degradation. This in turn will
9
protect the binder from weathering and further embrittlement [9]. It is hypothesized that PMA
10
shingles have reduced granule loss due to their elastomeric properties. In this small sample,
11
PMA/IR shingles showed improved granule bonding when compared to product D-OxIR, which
12
failed to pass the ASTM D3462 maximum allowable amount of 1.0 g of granule loss. Additional
13
granule adhesion tests were performed on eight oxidized asphalt shingles (0.88 g to 2.72 g of
14
granules loss). This range was higher than eight PMA shingles (0.10 g to 1.40 g of granule loss).
15
The tensile results also showed increased results for PMA/IR shingles when compared to the
16
oxidized IR shingle. More samples will need to be tested to see if these trends are observed
17
across a variety of shingles.
18
19
Table 4. Mechanical properties of architectural PMA and oxidized IR shingles tested at the
20
IBHS Research Center and PRI Asphalt.
21
Shingle
Product
Manufacturer
Type
Tear
Strength
(g)
Fastener
Pull-
Through @
23°C (N)
Fastener
Pull-
Through @
0°C (N)
Weight of
Displaced
Granules (g)
Tensile
(N)
ASTM
Test:
D 1922
(1,700)
D 3462
(133)
D 3462
(178)
D 4977
(1.0 max)
D 4073
D
4
Architectural
PMA/IR
2,118
165
231
0.1
698
D-OxIR
4
Architectural
Oxidized IR
4,710
156
205
1.2
339
E
5
Architectural
PMA/IR
1,802
169
205
0.4
527
F
1
Architectural
PMA/IR
1,984
222
285
0.8
574
22
3.4 Analytical Testing
23
FTIR analysis can be used to assess chemical differences in asphalt binders. Roofing oxidized
24
coatings are typically air blown fluxes that can show significant sulfoxide and carbonyl
25
intensities. After aging, these functional groups can be measured to determine the correlation to
26
oxidation increases [6]. Measurement of the carbonyl region can be conducted using ASTM
27
E1252, which compares the area in the 1650–1850 cm-1 spectral region of an oxidized asphalt to
28
its flux or non-oxidized asphalt binder. Typically, this will result in higher values for the
29

8
oxidized asphalt sample. In Figure 4, an increased carbonyl region is observed in the 1650–1850
1
cm-1 region.
2
FTIR can also be used to determine SBS content by comparing the ratio of styrene to butadiene
3
peaks using test methods such as AASHTO T302. Figure 4 points at 910 cm-1, which would be
4
where a butadiene group, or the covalent double bond absorbance would be present if the sample
5
was modified. A sharp peak at 699 cm-1 can also indicate the absorbance of the aromatic ring
6
structure of styrene. Styrene and butadiene are co-block polymers that make up the polymer
7
SBS. Although other polymers can be mixed into roofing asphalts, SBS is the primary one
8
indicated on the packaging for materials tested during this study. Other polymers can include
9
Styrene Ethylene Butylene Styrene (SEBS) and Atactic Polypropylene (APP). FTIR can be
10
coupled with other analytical methods such as oxygen content and size exclusion
11
chromatography (SEC) to get a better understanding of the properties of the binder.
12
13
Figure 4. FTIR scan comparing an oxidized IR architectural shingle (black). Blue text
14
indicates common peaks that may be present in polymer modified asphalts.
15
16
4. Discussion
17
4.1. Impact Test Method Improvements
18
A field survey in 2011, following a hail event in Dallas-Fort Worth, found IR shingles performed
19
better than non-IR shingles [10]. However, there is some concern the repeatable laboratory tests
20
do not fully and accurately predict real-world performance and damages [11]. The steel ball
21
impacts in UL 2218 crush the granules on the surface of the shingle, as shown in Figure 5A,
22
which is an uncommon damage mode in the field. This was verified by Crenshaw and Koontz,
23
who found that when an ice stone impacts a solid surface it will compress or fracture upon
24
impact, whereas steel is harder and will not compress, so the observed damage mode is more
25
mechanical and less realistic [12].
26
27
Ice spheres are more similar to hailstones than steel balls, but they do not perfectly replicate
28
hailstones in the field [13]. Mass, diameter, and propulsion speed of lab ice spheres are measured
29
during impact testing, but the hardness qualities are variable, and the masses required by FM
30
4473 result in density values that are equivalent to pure ice rather than those of real hailstones.
31
Figure 5 demonstrates damage modes created by the standard test methods (5A, 5B) compared to
32
more realistic test methods which better match field observations (5C). The impacts in Figure 5C
33
were created using 20% tap water mixed with 80% seltzer water, which caused the density to
34

9
drop below 0.9 g/cm3, thus creating an ice ball that more closely matches a hailstone. Research is
1
ongoing to further develop methods to produce lab ice spheres that more closely replicate real
2
hailstones and result in more realistic damage modes as illustrated in Figure 5C. The kinetic
3
energies used for UL 2218 and FM 4473 impact tests are based on Laurie [14], and recent
4
research suggests revisions may be necessary not only because of the lower density of hailstones,
5
but also because the terminal velocities assumed by the standard test methods may not be
6
achieved by real hailstones, which are lighter and aerodynamically different than spheres [15].
7
8
9
Figure 5. (A) Crushed granules on an asphalt shingle impacted with UL 2218 Class 2 steel
10
ball. (B) Indented surface of an asphalt shingle impacted with FM 4473 Class 2 pure ice
11
ball. (C) Damage on a naturally aged two-year old asphalt shingle impacted with multiple
12
dissolved CO2 ice spheres. Figures courtesy of Giammanco et al. [15]
13
14
4.2. Wind Test Method Improvements
15
The standard test method for wind performance raises some concerns regarding wind loads,
16
conditioning methods, and installation. In 1989, Smith and McDonald conducted roof surveys
17
after Hurricane Hugo hit South Carolina, and commented on the deficiencies of ASTM D3161 to
18
predict wind performance of asphalt shingles [16]. In 2012, Dixon et al. surveyed 27 roofs
19
throughout Florida to investigate the sealing of asphalt shingles, and discussed how the constant
20
wind speed tested in ASTM D3161 was not representative of the fluctuating wind patterns or
21
loads seen in the field [8]. Continued research is needed to determine effective conditioning
22
temperatures and equipment, and test methods appropriate to better predict real-world and long-
23
term performance.
24
25
5. Summary and Future Work
26
The durability of PMA shingles can be observed through a variety of testing. Ultimately, the
27
higher flexibility of the modified asphalts used in PMA shingles did result in increased passing
28
rates for impact tests. However, products were still observed not to pass the Class 4 rating at
29
100% as illustrated in Figures 1 and 2, confirming results from a previous IBHS study. Product-
30
by-product UL 2218 and FM 4473 testing of PMA shingles illustrated performance differences
31
among products, which could be due to the type of polymer that is used in the modification or
32
the process by which the polymer is mixed into the binder. If polymers are not sheared and
33
mixed at a high temperature, there is the potential for separation and incomplete mixing [1]. This
34
performance variability highlights the need to define characteristics which result in a good
35
quality PMA shingle. Future work will also include improvements to ice ball testing to more
36
accurately replicate damaging hail impacts.
37
38

10
ASTM D3161 wind test results illustrated there was not a large performance difference between
1
PMA and oxidized shingles in terms of maximum wind speeds that could be achieved, but rather
2
illustrated that damage after a test failure may be reduced in PMA shingles such that they could
3
reseal after a high-wind event. Continued work is being conducted to determine better
4
conditioning methods.
5
6
Age, distribution and storage can have an effect on the performance of asphalt shingles, which
7
may result in lower passing rates. After a product has been manufactured, it is stacked and stored
8
in a variety of conditions. Increased pressure and temperatures can negatively affect the
9
properties of the shingles. Over-stacking can flatten the wind sealant strip. Additionally, shingles
10
are distributed in a variety of packing and can arrive damaged or wet.
11
12
Through testing, IBHS has found bundle-to-bundle variability of shingles may lead to different
13
results between test series, and inconsistencies in field performance. Thus, additional research is
14
underway to understand the material and physical characteristics of asphalt shingles so testing
15
and damage assessment tools can be improved. Analytical and mechanical tests may be able to
16
predict long-term durability of asphalt shingles. Certain tests may be used to distinguish between
17
oxidized and PMA shingles, as well as predict the aging and oxidation of the binder. Aged
18
samples are being collected to evaluate the change in the FTIR spectra which, over time, can be
19
useful to determine modifications and changes in asphalt roofing binders. By collaborating with
20
various labs, IBHS will continue to compare analytical and mechanical test results to asphalt
21
shingle impact and wind performance, to understand if improved results are observed in PMA
22
shingles.
23
24
IBHS is collaborating with asphalt shingle manufacturers and other labs to determine the
25
mechanical and physical properties that may cause increased wind and impact resistance of
26
modified shingles, and to evaluate how products available in the market may resist damage
27
during hailstorms and high-wind events. The preliminary results presented in this paper serve as
28
indicators, and highlight inefficiencies in impact and wind test methods in predicting real-world
29
long-term performance. Moving forward, work will continue to relate durability testing to
30
analytical and materials testing, as well as an evaluation of the performance of aged products.
31
32
6. References
33
[1] J-P. Planche. Special Features of Polymer Modified Binders, 2004 Symposium On Additives
34
In Roadway Asphalts, Organized by Western Research Institute, Laramie, Wyoming, U.S. 2004.
35
36
[2] Underwriter’s Laboratories Inc. UL 2218: Standard for Impact Resistance of Prepared Roof
37
Covering Materials, Second Edition, pp. 6. 2010.
38
39
[3] FM Approvals. Test Standards for Impact Resistance Testing of Rigid Roofing Materials by
40
Impacting with Freezer Ice Balls, pp. 4. 2011.
41
42
[4] ASTM International. ASTM D3161-15, Standard Test Method for Wind-Resistance of Steep
43
Slope Roofing Products (Fan-Induced Method), ASTM International, West Conshohocken,
44
Pennsylvania, U.S. 2015.
45

11
1
[5] ASTM International. ASTM D3462-10a, Standard Specification for Asphalt Shingles Made
2
from Glass Felt and Surfaced with Mineral Granules, ASTM International, West Conshohocken,
3
Pennsylvania, U.S. 2015.
4
5
[6] University of North Texas. Practical Applications of FTIR to Characterizing Paving
6
Materials, Technical Report 0-5608-1. 2009.
7
8
[7] Insurance Institute for Business & Home Safety. Relative Impact Resistance of Asphalt
9
Shingles. 2014. [14]
10
11
[8] Dixon, et al. The influence of Unsealing on the Wind Resistance of Asphalt Shingles, Journal
12
of Wind Engineering and Industrial Aerodynamics, pp. 30–40. 2014. [12]
13
14
[9] Wright, J.R. Weathering: Theoretical and Practical Aspects of Asphalt Durability,
15
Bituminous Materials: Asphalts, Tars and Pitches, pp. 249–306. 1979.
16
17
[10] Herzog, R.F. Hailstorm Investigation Program Report, Roofing Industry Committee on
18
Weather Issues (RICOWI), pp. 33–40. 2012.
19
20
[11] Graham, Mark S. Concerns with Impact Testing, Professional Roofing, pp. 24. 2008.
21
22
[12] Crenshaw, V.A. and Koontz, J.D. Simulated Hail Damage and Impact Resistance Test
23
Procedures for Roof Coverings and Membranes, RCI Interface, pp 4–10. 2001.
24
25
[13] Insurance Institute for Business & Home Safety. IBHS Characteristics of Severe Hail Field
26
Program. 2014.
27
28
[14] Laurie, J.A.P. Hail and its effects on building, Bulletin 21, Pretoria, South Africa: National
29
Building Research Institute. 1960.
30
31
[15] Giammanco, et al. Evaluating the Hardness Characteristics of Hail Through Compressive
32
Strength Measurements, Journal of Atmospheric and Oceanic Technology. 2015.
33
34
[16] Smith, T.L. and McDonald, J. Roof wind damage mitigation: lessons from Hugo,
35
Professional Roofing Magazine. 1990.
36
37