1 RESIDENTIAL WINDOW INSTALLATION O PTIONS FOR HURRICANE PRONE REGIONS By CORY T HOMAS SALZANO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2009
2 2009 Cory T homas Salzano
3 To my late Uncle, Edward G. Salzano Jr.
4 ACKNOWLEDGMENTS I thank the members of my research oversight committee : Alside; American Architectural Manufacturers Association (AAMA) ; American Forest & Paper Association (AFPA) ; APA The Engineered Wood Association; Architectural Testing Inc.; Atrium Co mpanies Inc.; Cast C rete Corporation; C.B. Goldsmith and Associates Inc.; CEMEX; Certified Test Labs; Do Kim & Associates; DuPont ; Fenestration Manufacturers Association (FMA) ; Florida Building Commission ; Florida Home Builders Association (FHBA) ; General Aluminum ; Henkel; Institute for Business and Home Safety (IBHS); James Hardie; JBD Code Services; JELD WEN Windows and Doors; Lawson Industries Inc. ; Marvin Windows and Doors ; Masonry Information Technologists Inc.; MI Windows and Doors ; NuAir Windows and Doors ; Painter Masonry Inc. ; PGT Industries; PPG Industries; Protecto Wrap Company ; Silver L ine Windows and Doors; Simonton Windows ; TRACO; and WCI Group, Inc. The steadfast support of these companies was vital to this research. I thank my faculty advisor, Forrest J. Masters, Ph.D.; and my committee members, Kurtis R. Gurley, Ph.D. and David O. Prevatt, Ph.D., P.E. for the expertise they provided throughout the duration of this research. The continuing effort of these individuals to deconstr uct the near surface behavior of hurricane winds serves as the foundation for this research. This research was supported by the National Science Foundation (CMMI 0729739) and the Florida Building Commission.
5 TABLE OF CONTENTS ACKNOWLEDGMENTS .................................................................................................................... 4 Page LIST OF TABLES ................................................................................................................................ 8 LIST OF FIGURES .............................................................................................................................. 9 ABSTRACT ........................................................................................................................................ 11 CHAPTER 1 INTRODUCTION ....................................................................................................................... 13 2 BACKGROUND ON DAMAGE CAUSED BY WIND DRIVEN RAIN INGRESS ........... 25 Building Envelope Performance ................................................................................................ 25 Water Ingress and the Window Wall Interface ......................................................................... 28 Previous Research ....................................................................................................................... 31 Chapter Summary ........................................................................................................................ 35 3 WINDOW INSTALL ATION TECHNIQUES ......................................................................... 41 Water Barrier Method vs. Drainage Method ............................................................................. 41 Standard Practices of Installation ............................................................................................... 43 ASTM E 2112 Standard Practice for Installation of Exterior Windows, Doors and Skylights ........................................................................................................................... 44 FMA/AAMA 100 Standard Practice for the Installation of Windows with Flange s or Mounting Fins in Wood Frame Construction for Extreme Wind/Water Conditions ......................................................................................................................... 46 FMA/AAMA 200 Standard Practice for the Installat ion of Windows with Frontal Flanges for Surface Barrier Masonry Construction for Extreme Wind/Water Conditions (draft) ............................................................................................................. 47 Chapter Su mmary ........................................................................................................................ 49 4 EXPERIMENTAL PROCEDURE ............................................................................................ 53 Testing Apparatuses .................................................................................................................... 53 Air -Caster T ransportation Unit ........................................................................................... 54 Negative Pressure Chamber ................................................................................................ 55 Hurricane Simulator ............................................................................................................ 56 Hydraulic Leakage Pressure Apparatus ............................................................................. 58 Testing Protocols and Sequencing ............................................................................................. 59 Static Air Pressure Difference ............................................................................................ 60 Cyclic Static Air Pressure Difference ................................................................................. 61 Dynamic Pressure ................................................................................................................ 61
6 Interior Moisture/Air Barrier Testing ................................................................................. 62 Hydraulic leakage pressure test ................................................................................... 63 Adhesion strength ......................................................................................................... 64 Chapter Summary ........................................................................................................................ 66 5 SPECIMEN DESIGN AND CONSTRUCTION ...................................................................... 79 Wood Frame Wall Specimens .................................................................................................... 79 Masonry Wall Specimens ........................................................................................................... 81 Liquid Applied Flashing Application ........................................................................................ 83 Test Specimen Matrix ................................................................................................................. 84 Chapter Summary ........................................................................................................................ 84 6 RESULTS .................................................................................................................................... 95 Installation Dependence on Moisture Management Strategy ................................................... 96 Effects of Exterior Cladding on Installation Options ................................................................ 98 Wood Frame Wall Specimens ............................................................................................ 99 Concrete Masonry Unit Wall Specimens ......................................................................... 100 Performance of Liquid Applied Flashing Application Methods ............................................ 101 Ease of Installation and Repeatability ...................................................................................... 104 Performance of FMA/AAMA Installation Variants ............................................................... 105 Flashing System ................................................................................................................. 106 Interior Moisture/Air Barrier ............................................................................................ 106 Water Penetration Resistance of Sealants used for Interior Moisture/Air Seals ................... 108 Hydraulic Leakage Pressure.............................................................................................. 109 Water penetration performance of gunnable sealants vs. low expansion foams .... 110 Effect of binder type on interior seal performance .................................................. 112 Adhesion Strength ............................................................................................................. 114 Chapter Summary ...................................................................................................................... 116 7 CONCLUSIONS ....................................................................................................................... 128 Water Barrier Method vs. Drainage Method Installations ...................................................... 128 Interior Moisture/Air Barrier .................................................................................................... 130 Sealant Characteristics for Use in the Installation of Fenestration ................................. 131 Use of Low Expansion Foams in Drainage Method Installations .................................. 132 Factors Affecting the Water Penetration Resistance of Fenestration I nstallations ............... 133 Recommendations ..................................................................................................................... 134 8 SUGGESTIONS FOR FUTURE RESEARCH ...................................................................... 136 Identification of Leakage Paths ................................................................................................ 136 Environmental Aging ................................................................................................................ 138 APPENDIX : WALL SPECIMEN LEAKAGE RESULTS .......................................................... 141
7 LIST OF REFERENCES ................................................................................................................. 148 BIOGRAPHICAL SKECTH ........................................................................................................... 153
8 LIST OF TABLES Table page 1 1 Hurricane strikes from 18512006 on the mainland U.S. coastline, and for individual states, including inland areas if effects were only inland portions of the state, by Saffir -Simpson c ategory ........................................................................................................ 20 1 2 Saffir -Simpson Hurricane Scale ............................................................................................ 21 1 3 Ten most costly U.S. catastrophes ........................................................................................ 21 4 1 Sealant sample matrix ............................................................................................................ 67 5 1 Test specimen matrix ............................................................................................................. 86 6 1 Leakage results for varying window -wall moisture management combinations ............. 117 6 2 Comparison of leakage results for wood frame wall exterior finishes ............................. 118 6 3 Comparison of leakage results for CMU wall exterior finishes ........................................ 118 6 4 Leakage results for the flashing options of the FMA/AAMA 10007 standard............... 119 6 5 Leakage resu lts for the interior sealant options of the FMA/AAMA standards .............. 119
9 LIST OF FIGURES Figure page 1 1 Inflation adjusted U.S. catastrophe losses by cause of loss, 19882007 ............................ 22 1 3 Storm path for Hurricane Andrew, 1992 .............................................................................. 24 2 1 Hurricane Jeanne wind swath map with counties surveyed ................................................ 36 2 2 Surveyed homeowner opinion of reason for water intrusion .............................................. 37 2 3 Common sources of water intrusion through stucco cladding and the resulting damage .................................................................................................................................... 38 2 4 Possible leakage paths ............................................................................................................ 39 2 5 Critical barriers of the typical window -wall interface ......................................................... 40 3 1 Water barrier method installation detail ............................................................................... 51 3 2 Drainage method installation detail ...................................................................................... 52 4 1 Air -caster transportation unit. ................................................................................................ 68 4 2 Negative pressure chamber .................................................................................................... 69 4 3 Hurricane simulator ................................................................................................................ 70 4 4 Resi dential house model ........................................................................................................ 71 4 5 Sill specimen for hydraulic leakage pressure test ................................................................ 72 4 6 Hydraulic leakage pressure apparatus ................................................................................... 73 4 7 Loading function for the static air pressure difference test ................................................. 74 4 8 Loading function for the cyclic static air pressure difference test ...................................... 74 4 9 Loading function for the dynamic pressure test ................................................................... 75 4 10 Adhesionin -peel specimen construction .............................................................................. 76 4 11 Adhesionin -peel test ............................................................................................................. 77 4 12 Sealant failure modes ............................................................................................................. 78 5 1 Wa ll specimens ....................................................................................................................... 88
10 5 2 Wood frame wall specimens .................................................................................................. 88 5 3 Base of wood walls ................................................................................................................ 89 5 4 Window flashing .................................................................................................................... 90 5 5 Wire lat h application for wood frame walls ......................................................................... 91 5 6 pH test ..................................................................................................................................... 91 5 7 Construction of CMU walls ................................................................................................... 92 5 8 Prepared window rough opening in CMU wall .................................................................... 93 5 9 Application methods for LAF ............................................................................................... 94 6 1 Drainage channel constriction ............................................................................................. 120 6 2 Flashing options ................................................................................................................... 120 6 3 Gunnable sealant leakage comparison ................................................................................ 121 6 4 Low expansion foam leakage comparison .......................................................................... 121 6 5 Gunnable sealant water penetration resistance ................................................................... 122 6 6 Low expansion foam water penetration resistance ............................................................ 122 6 7 Water intrusion through low expansion foam .................................................................... 123 6 8 Test 1 adhesion-in -peel values ............................................................................................ 124 6 9 Test 2 adhesion-in -peel values ............................................................................................ 125 6 10 Adhesionin -peel values to sill flashing membrane ........................................................... 126 6 11 Comparison of water penetration resistance and adhesionin peel values for gunnable sealants .................................................................................................................................. 127
11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering RESIDENTIAL WINDOW INSTALLATION OPTIONS FOR HURRICANE PRONE REGIONS By Cory T. Salzano May 2009 Chair: Forrest J. Masters Cochair: Kurtis R. Gurley Major: Civil Engineering Water ingress to structures is one of the most critical and recurring issues during hurricane impacts. Although most homes and businesses survive structurally, a significant number experience enough rain penetration to cause damage to the building interior and contents. This issue came to the forefront after the 2004 hurricane season, when it was shown that homes built to the standards of the Florida Building Code following the 1995 revisions suffered minimal physical damage yet incurred substantial insured losses due to water intrusion. The window -wall interface has been identified as a key source of water intru sion problems in buildings, particularly in extreme exposure regions Considerable effort has gone into the development of robust practices to enhance the water management performance of window installations, such as redundant drainable methodologies with enhanced flashing and sealant products, but actual performance testing of the installations under real life extreme exposure conditions has been severely lacking. This study examines the effect of extreme wind -drive n rain exposure on a variety of window installation details, utilizing conventional fenestration testing methods ad apted to accommodate full -scale test specimens and a unique hurricane simulation apparatus.
12 Static, pulsating as well as amplitude and frequency -modulated sinusoidal pressure load sequences were applied under simulated wind driven rain conditions to 18 wall assemblies that were varied uniquely by their fenestration, installation methodology, wall structural syst em and exterior cladding system. Selected installation methods were tested, including various flashing products and sealant applications. Based on the findings of the testing, conclusions are made regarding the effectiveness of window installations i n e xtreme exposure conditions as well as the impact s the components of the surrounding wall syst em have on their performance. This paper details various insights concerning the moisture management strategy as it relates to the integration of the window into the critical barriers of the wall assembly along with insights into the performance of sealants used to create interior moisture/air seals in drainage method window installations. Suggestions are also made to impr ove the efficacy of future research based on the results of this study. .
13 CHAPTER 1 INTRODUCTION Ranking as the leading catastrophe within the United States in terms of the cost of incurred losses (Figure 1 1) ; the damages produced by hurricane impacts are significant and widespread. The adverse effects of hurricanes include but are not limited to loss of life, destruction of infrastructure, disruption of businesses and displacement of families during repairs With an average annual economic loss estimated at $5 billion ( Pielke and Landsea 1998), the devastation caused by these natural disasters has a national reach and a lasting impact. Furthermore, societal trends and the possible link between anthropogenic climate changes and elevated extreme weather activity ensure that tropical cyclon es will continue to dis tress the nation in the future (Valverde and Andrews 2006). In the past 50 years, hurricane induced economic losses have increased steadily in the U.S., with estimated annual total losses (in constant 2006 dollars) averaging $1.3 billion from 19491989, $10.1 billion from 19901995, and $35.8 billion per year during the last 5 years (NSB 2007). The trend of increasing economic loss is likely to continue in subseq uent years as populations and infrastructure on and near the shoreline grow It was estimated that in 2003, approximately 153 million people (53 percent of the nations population) lived in the nations coastal counties, an increase of 33 million people since 1980 (Crossett et al. 2004). The threat of increasing hurricane losses has been brought to the forefront in recent years by the heightened activity of the 2004 and 2005 hurricane seasons, which resulted in more than $150 billion in damage (Pielke et al. 2008) and claimed t he lives of 1,585 individuals ( Blake et al. 2007). The storms of these seasons, namely Hurricane Katrina, attracted significant media interest as the toll of the human and financial loss reverberated throughout the nation. Public attention in the nations coastal communities has since been focused on the need to better
14 understand the multifaceted impacts associated with hurricanes and to develop mitigation tools necessary to reduce losses. Tropical cyclones have long been a pervasive risk to the U.S. Atlantic and Gulf Coasts. While every state from Maine to Texas is affected by hurricane impacts, none are more frequent ly affected by landfalling storms than the S tate of Florida (as illustrated in Table 1 1). Over half of the hurricane related damage in the U.S. occurs in Florida, and will likely grow as recent construction has increased the building stock appreciably around its 1 926 km ( 1,197 mi ) coastline. The total value of insured coastal property in Florida as of 2007 was $2.5 trillion, up nearly 27 percent since 2004, and that number is expected to double by 2014 ( I. I. I. 2008b ). As the coastal exposure of insured property increases so does the looming threat of significant losses due to hurricanes impacting the coast. For example, it has been suggested that a repeat of the Great Miami Hurricane of 1926 (Figure 12) would today produce insured losses in the range of $80 billion to $100 billion ( Robert P. Hartwig, Insurance Claims Payment Processes in the Gulf Coast after the 2005 Hurricanes wr itten testimony to the United States House of Financial Services Committee on Oversight and Investigation, Washington, DC February 28, 2007). When it is considered that the Great Miami Hurricane of 1926 made landfall just south of Miami as a Category 4 h urricane on the Saffir -Simpson scale (Table 1 2) and that the 2007 hurricane season became the first season to produce two hurricanes (Felix and Dean) making North American landfalls as Category 5 storms (Gutro and Patzert 2008) the scenario becomes eeril y plausible. The possibility of such losses underscores the need to further investigate the efficacy of current design standards for construction in hurricane prone regions, a task that has been ongoing in the state of Florida since 1992.
15 On August 24, 1992 Hurricane Andrew tracked along the southern tip of Florida as a Category 4 hurricane (Figure 1 3). Ranking as the second most costly catastrophe in the U.S., surpassed only by Hurricane Katrina in 2005 (Table 13), Andrew wrought extensive damage on Fl orida with total costs estimated upwards of $30 billion dollars and more than $15 billion in insurance claims (Grier 1996) The aftermath of the storm spawned an immediate imperative to evaluate the performance of the current building stock. Extensive bu ilding assessments were performed analyzing the damage incurred mainly by residential and light commercial structures with the intent to evaluate the adequacy of current building codes and provide recommendations for reducing future losses. Assessment tea ms investigated the performance of the primary structural systems of buildings, i.e. the main wind force resisting system (MWFRS) designed to resist the lateral and vertical forces imparted by wind loading, as well as building envelope systems, such as roofing, windows and doors. It was observed that much of the damage to residential structures resulted from inadequate design, substandard workmanship, and/or misapplication of various building materials (FEMA 199 2 ). Poorly designed connections between cri tical components allowing for a proper path for load transfer were of particular concern attributing to the majority of structural failures. Moreover, extensive loss of roof cladding and the failure of windows and doors caused the interiors of buildings vu lnerable to further damage from wind and rain. The result was significant damage to building interiors and contents that rendered many buildings uninhabitable (FEMA 199 2 ). In response to the damage observed in the wake of Andrew, the state formulated the enhanced South Florida Building Code adopted in September of 1994. Over the years these provisions were repeatedly modified and ultimately, on March 1, 2002, the improved structural portions of the South Florida Building Code were enveloped into the 2001 Florida Building Code
16 (FBC) as the High Velocity Hurricane Zone (HVHZ) provisions, applicable to Dade and Broward counties (BCCO 2006 ). The newly revised FBC instituted more rigorous guidelines for structural design and required contractors to use only pr oducts that complied with the standards set for hurricane -force winds. The recent hurricane impacts of 2004 and 2005, while unfortunate, have provided a valuable chance to scrutinize these improvements and evaluate their effectiveness in actual conditions The Federal Emergency Management Agency (FEMA) deployed numerous teams to perform field observations to determine how residential buildings in Florida performed during the hurricane impacts of the 20042005 hurricane seasons. These teams consisted of en gineers, architects, policy makers, code specialists, and building officials assembled to document observations and provide recommendations. In a summary report compil ed from these observations ( FEMA 2005c 2006b ) it was noted that the current building co de appeared to adequately address the structural design of residential buildings as there w as little wind damage to the structural systems in buildings constructed to the revised code. The majority of the damage from these storms was concentrated to the building envelope systems such as roof coverings, soffits, doors, and windows. The failures of these systems allowed winddriven rain to enter building interiors causing not only loss of function, but millions of dollars of damage to building contents due to water damage from rain and subsequent mold growth ( FEMA 2005c ). With the improvements to the structural capacity of buildings following the 2001 FBC, attentions must be focused on improving the building envelope to prevent costly water intrusion. One of the major sources of water i ngress through the building enve lope is through openings in walls caused by windows and other fenestration, in particular the interface between the window and the wall (Crowder -Moore et al. 2006) While stricter requirements for shutters and the
17 growing popularity of impact resistant glass may prevent failures to the window glazing during hurricane impacts, window openings are not yet impervious to the effects of the associated winddriven rain. Leakage paths originating from various weak points in the wall assembly of the building such as through cracks or breaches in the exterior cladding (refer to Figure 1 4) exploit the window -wall interface as a pathway to the building interior. For example rainwater entering a crack in the stucco rendering surrounding a window may travel down the concealed drainage plain of a wood frame wall system and ultimately penetrate to the building interior through the window -wall interface. Water that migrates to the building interior can lead to wood rot, peeling paint, and microbial growth which if left unchecked can displace occupants from their homes. In order to preclude the deleterious effects of water intrusion observed following recent hurricane impacts t he vulnerability of the building envelope must be add ressed beginning with the window -wall interface and an evaluation of current window installation standards. Current water penetration resistance test methods for fenestration (e.g. ASTM E331 00, AS/NZS 4284:1995, JIS A 1517) evaluate products in isolation and not as a component integrated into a wall system. Test ing in this manner negates the a ffects that possible leakage paths originating in the surrounding wall assembly have on the water penetration r esistance of the window installation. Thus the efficacy of the various window installation techniques employed in the field are not well understood, and less is known about leakage paths originating from the wall or the wi ndow -wall interface. Moreover, i t is unclear how effective these water penetration control strategies work in the extreme buffeting loads associated with the turbulent wind action in a hurricane surface layer wind field In collaboration with industry professionals, researchers at the Uni versity of Florida (UF) have conducted a study focusing on the use of full -scale experimental testing methods to
18 evaluate common window installation options for hurricane -prone regions The goal of this study was to evaluate the water penetration resistan ce of fenestration installation techniques and their interaction with the various components of the completed building assembly under simulated extreme wind -driven rain conditions Eighteen 2.4 m by 2.4 m (8 ft by 8 ft) wall specimens were constructed each with a window installed using either the traditional exterior barrier interface prescribed in Method A1 of ASTM E 2112 or the drainable installation practices detailed in the FMA/AAMA 10007 and the draft FMA/AAMA 200 Stan dard Practices. Each specimen w as comprised of a unique combination of fenestration, installation methodology, and finished wall system. The wall specimens were subjected to four rounds of pressure loading under static, pulsating, and dynamic tests while the corresponding times and pre ssures were recorded at which water leakage through or around the test assembly to the interior occurred Results from the tests were then compared to evaluate the effectiveness of the varying moisture management strategies employed by the wall specimens. Scope of w ork. Water ingress is one of the most critical, recurring issues during hurricane impacts. Although most homes and businesses survive structurally, a significant number experience enough rain penetration to cause damage to the interior and a loss of building c ontents. This problem came to the forefront after the 2004 2005 hurricane seasons, when it was clearly demonstrated that homes built to the new Florida Building Code standards suffered minimal physical damage, yet incurred substantial insured losses due t o water intrusion. This paper provides an overview of experimental research conducted to investigate the performance of various window installation options for single -family homes in extreme wind driven rain conditions.
19 Information on typical building en velope failures experienced duri ng hurricane impacts and the ir e ffects on the water penetration performance of installed fenestration systems are presented in Chapter 2. The installation standards for the fenestration products tested in this study may be classified as either water barrier methods or drainage method installations. The differences between these two methods occur in the provisions made in each standard to manage incidental water that may enter the window -wall interface; these differences are further explained in Chapter 3. Static, pulsating, as well as amplitude and frequency -modulated sinusoidal pressure load sequences, as described in Chapter 4, were applied under simulated wind -driven rain conditions to 18 wall assemblies that uniquely v aried by their fenestration, installation methodology, wall system construction, and exterior finish. The construction of the wall assemblies is t horoughly detailed in Chapter 5 and w ater penetration results for the wall assemblies are presented in Chapte r 6. Based on these findings, conclusions were made regarding the effectiveness of installations in extreme exposure conditions. These conclusions may be found in Chapter 7. Suggestions to improve the efficacy of future research are presented in Chapter 8.
20 Table 1 1. Hurricane strikes from 1851 2006 on the mainland U.S. coastline, and for individual states, including inland areas if effects were only inland portions of the state, by Saffir -Simpson category. [ Adapted from Blake, E. S., Rappaport, E. N ., and Landsea, C. W. (2007). The Deadliest, Costliest, and Most Intense United States Tropical Cyclones from 1851 to 2004 (and Other Frequently Requested Hurricane Facts) NOAA Technical Memorandum, NWS TPC 5, National Oceanic and Atmospheric Administra tion, Tropical Prediction Center, National Hurricane Center, Miami, Florida, 1 24. (Page 18, Table 10)] Area Category Number All Major Hurricanes 1 2 3 4 5 U.S. (Texas to Maine) 110 73 75 18 3 279 96 Texas 23 18 12 7 0 60 19 (North) 12 7 3 4 0 26 7 (Central) 7 5 2 2 0 16 4 (South) 7 7 7 1 0 22 8 Louisiana 18 14 15 4 1 52 20 Mississippi 2 5 8 0 1 16 9 Alabama 16 4 6 0 0 26 6 (Inland only) 6 0 0 0 0 6 0 Florida 43 33 29 6 2 113 37 (Northwest) 26 17 14 0 0 57 14 (Northeast) 12 8 1 0 0 21 1 (Southwest) 18 10 8 4 1 41 13 (Southeast) 13 13 11 3 1 41 15 Georgia 15 5 2 1 0 23 3 (Inland only) 9 0 0 0 0 9 0 South Carolina 18 6 4 2 0 30 6 North Carolina 24 14 11 1 0 50 12 (Inland only) 3 0 0 0 0 3 0 Virginia 7 2 1 0 0 10 1 (Inland only) 2 0 0 0 0 2 0 Maryland 1 1 0 0 0 2 0 Delaware 2 0 0 0 0 2 0 New Jersey 2 0 0 0 0 2 0 Pennsylvania (Inland) 1 0 0 0 0 1 0 New York 6 1 5 0 0 12 5 Connecticut 5 3 3 0 0 11 3 Rhode Island 3 2 4 0 0 9 4 Massachusetts 6 2 3 0 0 11 3 New Hampshire 1 1 0 0 0 2 0 Maine 5 1 0 0 0 6 0 State totals will not equal U.S. totals, and Texas or Florida totals will not necessarily equal sum of sectional totals. Regional definitions may be found in Appendix A of Blake et al. (2007). Gulf Coast state totals will likely be underrepresented because of lack of coastal population before 1900.
21 Table 1 2. Saffir Simpson H urricane Scale Category Winds (km/h) Pressure (millibars) Surge (m) Damage 1 119 153 > 980 1.2 1.5 Minimal 2 154 177 965 979 1.8 2.4 Moderate 3 178 209 945 964 2.7 3.7 Extensive 4 210 249 920 944 4.0 5.5 Extreme 5 > 250 < 920 > 5.5 Catastrophic Information for pressure, surge, and damage are characteristics typical of storms of the particular category and vary on a storm by storm basis. Table 1 3. Ten mo st costly U.S. catastrophes [Adapted from the Insurance Information Institute (2008c). Catastrophes: U.S. Facts and Statistics (Feb. 18, 2009)]. Insured loss ($ millions) Rank Date Peril Dollars at occurrence 2008 dollars 1 Aug. 2005 Hurricane Katrina 41,100 45,309 2 Aug. 1992 Hurricane Andrew 15,500 23,786 3 Sep. 2001 World Trade Center, Pentagon terrorist attacks 18,779 22,830 4 Jan. 1994 Northridge, California earthquake 12,500 18,160 5 Oct. 2005 Hurricane Wilma 10,300 11,355 6 Sep. 2008 Hurricane Ike 10,655 10,655 7 Aug. 2004 Hurricane Charley 7,475 8,520 8 Sep. 2004 Hurricane Ivan 7,110 8,104 9 Sep. 1989 Hurricane Hugo 4,195 7,284 10 Sep. 2005 Hurricane Rita 5,627 6,203 Values represent property coverage only. Does not include flood damage covered by the federally administered National Flood Insurance Program. Values in 2008 dollars were adjusted by the Insurance Information Institute. Totals for Hurricane Ike are estimated.
22 (1 ) Catastrophes are all events causing direct insurance losses to property of $25 million or more in 2007 dollars. Adjusted for inflation by ISO. (2 ) Includes hurricanes and tropical storms. (3 ) Includes snow. (4 ) Includes other geologic events such as volcanic eruptions and other e arth movement. (5 ) Does not include flood damage covered by the federally administered National Insurance Program. (6 ) Includes wildland fires. Figure 1 1. Inflationadjusted U.S. catastrophe losses by cause of loss, 198820071. [Adapted from the I nsurance I nfo rmation Institute (2008c). Catastrophes: U.S. Facts and Statistics (Feb. 18, 2009)].
23 Figure 1 2. Storm path for the Great Miami Hurricane, 1926. [Reprinted with permission from National Hurricane Center (NHC). (2007). Hurricane History. Hurricane Preparedness (March 3, 2009).]
24 Figure 1 3. Storm path for Hurricane Andrew, 1992. [Reprinted with permission from National Hurricane Center (NHC). (2007). Hurricane History. Hurricane Preparedness (March 3, 2009).]
25 CHAPTER 2 BACKGROUND ON DAMAGE CAUSED BY WIND DRIVEN RAIN INGRESS The ability of the structural system to resist the wind loads experienced during hurricane impacts is imperative to the preservation of a building and its occupants. The resilience of the structural system alone; however, does not guarantee the protection of the building interior and contents. In the damage assessment reports for the recent storm seasons following Hurricane Andrew ( FEMA 1999, 2005a, 2005b, 2005c 2006a ,2006b ), it was consistently noted that significant damage to residential and light commercial structures was a result of poor building envelope perf ormance. The insufficient wind resistance of the building envelope systems caused failures which allowed the infiltration of winddriven rain into buildings, resulting in costly content damage and at times loss of function. With noted improvement in buil ding structural capacities attributable to stronger building codes and enforcement, it is becoming increasingly appare nt that research is also needed to address the vulnerability of the building envelope Building Envelope Performance Historically, poor bu ilding envelope performance is the leading cause of damage to buildings and their contents in weak to moderate intensity hurricanes (FEMA 2005c ). The building envelope is comprised of exterior -mounted mechanical and electrical equipment, non load bearing walls, soffits, wall and roof claddings, and fenestration systems (windows and doors). The individual systems of the building envelope work to aid in the structural integrity of the building, facilitate climate control, manage interstitial air movement, a nd control moisture from infiltrating into the building interior. When the envelope is compromised, so is the separation between the exterior environment and the building interior. This is of particular concern in the wind-driven rain environment typical of tropical cyclones.
26 A summary report on building performance for the 2004 hurricane season ( FEMA 2005c ) outlined four building envelope failures that were most consistently observed in the damage following h urricanes Charley, Ivan, and Frances. The occ urrence of these typical failures are further reiterated in the preliminary damage observations for the same storms provided by the Institute for Business and Home Safety (IBHS 2004) and a progress assessment produced by the Miami Dade County Building Code Compliance Office following Hurricane Wilma ( BCCO 2006). The common failures and typical causes are listed and paraphrased in the following paragraphs. Failures in roof coverings were commonly due to poor attachment and damage from windborne debris. Tile and shingle roof coverings were the most commonly observed roofing systems in the damage reports. Tile damage was prevalent around roof ridges, hips, and windward edges where a lack of mechanical anchorage w as observed. Asphaltic shingle performance in windstorms showed improvement from previous years due to product advancements in the adhesive strips on the underside of the shingles (IBHS 2004) Failures in shingles were attri buted to the lack of or incorrect placement of nails (FEMA 2005c) The failures of shingles and tiles exposed roof underlayment and sheathing to further damage allowing water to penetrate through to the building interiors. Mechanical and electrical equipm ent attached to buildings failed at the connections under wind loading oftentimes leaving large openings in the roof or punctures in the roof membrane resulting in water ingress (FEMA 2005c) Much of the equipment that failed was essential to the proper operation of the building. Failures included electrical and heating, ventilating, and air conditioning (HVAC) equipment resulting in prolonged loss of building function compounding the economic losses of water damage.
27 Soffit failure s were commonly observe d and generally attributed to poor detailing. Soffit failures resulted in significant amounts of rain water intrusion into attic spaces and wall cavities and eventually to the building interior. Windows and doors failed in two distinctly different manner s. Windows and doors broken by impinging winds or windborne debris allowed for internal pressurization to quickly magnify the intensity of the structural loading. Broken windows and doors also exposed building interiors to wind -blown rain. However, sign ificant water intrusion was also observed in buildings with no structural damage to windows and doors. As seen from the previous list, water intrusion is a common theme in the insufficient performance of the building envelope. Controlling moisture infiltr ation is one of the primary functions of the building envelope and failures in these systems lead to water ingress to the building interior Once the building envelope is breached during a hurricane wind -driven rain is forced through gaps, cracks, and op enings, eventually reaching the building interior. The resulting water lead s to several adverse effects including the degradation of the building materials that create the wall assembly. In their summary r eport on building performance for the 2004 hurri cane season a Mitigation Assessment Team (MAT) assembled by FEMA observed that most of the wind damage to the building envelope was preventable (FEMA 2005a ). While a significant portion of the damage occurred to older structures with aged materials and dated construction techniques damage to building envelope systems of structures constructed with newer materials and according to current codes and standards were also observed. Failures in these instances were either attributed to poor craftsmanship an d/ or insufficient detailing of crucial wind resistant components of the systems. Recommendations to amend future performance issues made in the
28 report included providing additional guidance to contractors and designers on the wind resistance issues associ ated with these systems, supplementing information where code guidance is vague, and expanding methods used to test these systems. Destructive failures in the building envelope were not the only sources of water intrusion during these hurricane impacts. T he University of Central Florida (UCF) Housing Constructability Lab and the Building Science Corporation engaged in separate studies followin g the 2004 hurricane season (Mullens et al. 2006 and Lstiburek 2005) that observed significant water ingress in bui ldings showing no visible signs of wind damage to the building envelope In these cases, cracks in the stucco cladding of the wall system created entry points for wind -driven rainwater which then migrated to the building interior through the wall system or the window wall interface. These reports highlight weaknesses in the window opening and the surrounding wall assembly which underscores the need for more stringent testing protocols for window components and increasingly robust window installation stan dards to improve the window -wall interface. Water Ingress and the W indow Wall Interface In 2006, the UCF Housing Construct ability Lab conducted a study ( Mullens et al. 2006) funded by the U.S. Department of Energy, Office of Technology, State and Community Programs in response to an overwhelming number of water intrusion complaints received by the Orlando Home Builders Association following Hurricane Jeanne. The water intrusion from these complaints was unusual in that the majority arose from newly constru cted homes that exhibited little to no wind damage to the building envelope. To determine the cause of the water intrusion researchers conducted a survey of 173 residential homes built between 2001 and 2004 in six central Florida counties (Figure 2 1): B revard, Lake, Orange, Osceola, Seminole, and
29 Volusia. The goal of the study was to characterize the leakage paths associated with the water intrusion complaints and provide recommendations to limit future loss. Survey results indicated that 20 percent of all new homes built in central Florida in 2003 experienced water intrusion relat ed to the wall assembly during H urricane Jeanne ( M ullens et al. 2006). Inspection of the homes exposed several causes for the damage including poorly sealed windows, unseale d wall penetrations, and cracks of various shapes and sizes in the exterior finish of the wall assembly. Small cracks sometimes measuring less than 0.39 mm (1/64 in) wide, were the most consistent source of water intr usion observed during the study. Wal ls with cracks in the stucco rendering concentrated around window openings from various causes including movement of the underlying block wall system and shrinkage of the stucco, serve d as entry points for winddriven rain. When these cracks were not sealed during the application of paint they facilitate d the migration of water t hrough the wall resulting in damage to the building interior concentrated around windows and the base of the wall (Figure 2 4) Of the 173 homes originally surveyed, 166 homeowners responded to an additional survey requesting their opinions on the sources of the water intrusion. Although the majority of the 166 homeowners questioned were unable to provide a definitive answer, almost 20 percent concluded windows and doors to be the cause of the water ingress, as shown in Figure 2 2 (Mullens et al. 2006). Similar conclusions are found in a report produced by the B uilding Science Corporation (Lstiburek 2005). The main objective of this report was to review the performance of residential building assemblies in the central Florida area following the 2004 hurricane season. While this report mainly focused on the issues associated with stucco cladding systems, window systems and their integration into the adjacent wall assembly were also documented as weak points in the building envelope.
30 The effectiveness of stucco cladding as a water shedding surface is dependent on its continuity. Consistent with the inspections conducted by UCF Lstiburek notes in his review that significan t water penetrated stucco systems through micro cracks that typical paint finishes were unable to seal ( 2005). The water penetrated the stucco and became a problem for wood framed wall assemblies which rely on the effectiveness of a concealed water resi stive barrier (WRB) applied over the wall sheathing to serve as the drainage plane for the wall Normally, water reaching the drainage plane would be redirected to the exterior; however, it was observed in many cases the drainage plane was negated due to adhesion between the WRB and the stucco rendering. In the absence of a properly functioning drainage plane, water flows through the wall and enters the building interior under the extreme wall wetting conditions experienced during hurricane events. In addition to his performance review of stucco cladding systems, Lstiburek (2005) highlighted several key issues involving the water management at window openings and other service penetrations. While windows are only rated for water holdout to 15 percent of t h eir design pressure rating ( AAMA/WDMA/CSA 101/I.S.2/A44005), testing performed by Lstiburek indicated that windows leaked at conditions well below these rated values during the storms o f the 2004 hurricane season ( 2005). Furthermore, he suggests that installation instructions regarding windows and doors are inadequate with respect to the issu es involving water management Currently, the performance of the window -wall interface is not subject to an ASTM standard under the FBC The performance of window openings with respect to water management is crucial to the proper function of the building envelope. As a component system integrated into the wall assembly, windows installations must resist water penetrating the window assembly as well as
31 any water en tering through the interface with the adjacent wall system. Deficiencies in the surrounding wall assembly such as poor drainage plane performance, unnecessary perforations in wall cladding systems, and improper flashing act to compound the amount of water that must be managed at the window opening. A successful water management strategy at the window -wall interface requires the development and strict enforcement of increasingly robust window installation methodologies. Guidance on water management issues at the window interface must be provided to designers and contractors to ensure pro per design and the implementation of construction best practices. This can only be accomplished by first determining the main leakage paths associated with water ingress through windows and the window -wall interface. Previous Research Significant research has been published on the issues associated with the insufficient wind resistance of building envelope systems and the associated rainwater intrusion. Research has included field documentation following windstorm events, assessments of insurance records tying the magnification of insured losses wit h building envelope damage ( Sparks et al. 1994), and the development of damage prediction models to as sess building failure ( Pinelli et al. 2004). However, there is limited research on water penetration related to the window -wall interface. While it is vital to establish a window and a wall assembly with sufficient wind driven rain penetration resistance, the performance of the interface between these assemblies is equally important. The advancement of the wind driven rain penetration resistance of the interface must begin with a clear understanding of the leakage paths involved. The most comprehensive study of the moisture related performance problems surrounding the interface between the window and the nei ghboring wall assembly was conducted by RDH Building Engineering Limited of Canada. In response to the recurring issue of water intrusion
32 prevalent in the residential building assemblies of British Columbia, RDH conducted a two part study on the water penetration resistance of windows (RDH 2002a, 2002b) The first, was a study to determine the primary leakage paths and causes of water penetration associated with windows and the window -wall interface concentrating on the factors attributed from manufacturi ng, building design, ins tallation, and maintenance ( RDH 2002a ). The goal of the second study was to identify improvements to be made to building codes, standards, testing protocols and certification process es to better control water penetrating the interf ace between the window and the wall ( RDH 2002b ). These studies presented fundamental information on common leakage paths influencing the window -wall interface and provided a detailed understanding of the critical barriers used to prevent water migration t hrough the wall assembly. Through an exhaustive assessment of water penetration associated with various window operating types and adjoining wall systems, RDH defined six possible leakage paths through the window and the window -wall interface (2002a). The se leakage paths are detailed in Figure 2 4. Extensive water penetration testing performed on in-service windows demonstrated that leakage paths occurring through the interface between the window and the wall were the most significant in terms of both fre quency of occurrence and risk of consequential damage. Three of the six leakage paths defined directly influence the window -wall interface, L3 (through window -wall interface to the building interior), L4 (through window assembly to adjacent wall assembly) and L5 (through window -wall interface to adjacent wall assembly) Of these three, L4 and L5 were considered to pose the highest threat of consequential damage. The leakage paths resulted from a variety of casual facto rs stemming from a poor understanding of the critical barriers involved in the successful moisture management of the wall assembly.
33 The critical barrier s refer to materials and components that together perform a specific waterproofing function within a wall or window assembly ( RDH 2002a ). A fundamental understanding of these critical barriers is paramount to the water penetration resistance of the building envelope. There are four critical barriers specific to the integration of a window into a wall system (Figure 2 5) Starting from the exterior of the wall assembly and moving toward the building interior these critical barriers include the water shedding surface, exterior moisture barrier, v apor barrier, and air barrier. While these critical barriers may not always appear in the same l ocation or in the same form, the ir presence i s essential for the proper function of the building envelope. The water shedding surface is the initial barrier in preventing water intrusion. The components of the typical window and wall assembly form ing the water shedding surface consist of the glazing, sealant between the glazing and the frame, surface of the window frame, the sealant between the frame and the sill, the sill, and the exterior surface of the stucco. The water shedding surface acts to deflect the bulk of the water impact ing the faade. Water that is not deflected by the water shedding surface must be managed by the exterior moisture barrier. The exterior moisture barrier is usually provided by the glazing, sealant between the glazing and the window frame, sealant between the window frame and the sill pan flashing, the sill pan flashing, and the house wrap applied overtop the sheathing. The exterior moisture barrier represents the farthest point into a wall assembly that moisture may be accommodated. Liquid water penetrating this barrier will flow unrestricted to the building interior. The next critical barrier in the progression throug h the wall assembly is the vapor barrier. The purpose of the vapor barrier is to resist vapor diffusion and aid in the control of interstitial
34 air movement. The vapor barrier may consist of the interior glazing, the interior of the window frame, and a lo w permeability building sheet. The location of the vapor barrier in the wall assembly varies geographically with climate exposure The vapor barrier should always be placed on the high (humidity) vapor pressure side of the exterior wall (Anis et al. 2006). In colder northern climates, warm/moist interior air migrating through the wall is cooled by the cold outside environment causing it to condensate, trapping moisture inside the insulated wall space. To prevent this, the vapor barrier in northern clima tes is placed near the interior of the wall assembly. Conversely, in warmer southern climates the vapor barrier is usually located at the exterior side of the building. This is due to the tendency of the inward driven warm/humid outside air to condensate as it migrates through the wall assembly and is cooled by the air conditioned interior environment At times vapor barrier s in southern climates may be achieved with the use of a low permeability WRB for the exterior moisture air barrier. While this st udy focuses on fenestration installation practices common to the s outheast U.S., the vapor barrier in F igure 2 5 is shown near the interior of the wall assembly; its location for cold weather climates. It should be noted that t his location was chosen to distinguish it presence and function apart from the exterior mo isture barrier and that in the s outheast U.S. the vapor barrier of the wall assembly will usually overlap the exterior moisture barrier of F igure 2 5. The last barrier in the wall system is the air barrier. The purpose of the air barrier is to resist interstitial air movement between the exterior environment and the building interior. The air barrier usually consists of the interior glazing, sealant between interior glazing and interior surface of the window frame, interior surface of the window frame, seal between the window frame and the interior sill, and the inte rior sill return to the drywall (RDH 2002a).
35 The continuity and proper sequencing of the four critical barriers should be the found ation for the successful wind -driven rain penetration resistance of window installation guidelines. While the installation sector has little control over the leakage paths through the window assembly, the most frequent and damaging leakage paths occur thr ough the window and the wall interface. Recognizing that the dominant leakage paths contributing to water intrusion occur at this interface, installation methodologies in extreme wind -driven rain areas are beginning to shift focus toward water managemen t rather than prevention. In order to assist the progression of the wind driven rain resistance of installation methodologies research must be preformed to evaluate the effectiveness of traditional and proposed installation options. Chapter Summary With a significant reduction i n building structural failures observed during recent hurricane impacts, the insufficient wind resistance of the building envelope has garnered increased attention Some of the most significant damage documented following these storms resulted from rainwater breaching the building envelope and flowing to the building interior. Moreover, investigations of residential structures after hurricanes show that significan t water intrusion can occur to homes that suffer little or no visible wind damage. A survey showed that homeowners believed their window openings to be a primary source of this water leakage. A scientific study of the water penetration resistance of wind ows concluded that the window -wall interface is a dominant source of water leakage. In response, current installation methodologies are being developed to manage the leakage paths affecting the window -wall interface
36 Figure 2 1. Hurricane Jeanne win d swath map with counties surveyed [Adapted from FEMA. (2005c). Summary Report on Building Performance: 2004 Hurricane Season, FEMA 490, Federal Emergency Management Agency, Washington, DC.(Page 9, Figure 5 )].
37 Figure 2 2. Surveyed homeowner opin ion of reason for water intrusion. [ Adapted from Mullens, M., Hoekstra, B., Nahmens, I., and Martinez, F. (2006). Water Intrusion in Central Florida Homes During Hurricane Jeanne in September 2004, University of Central Florida Housing Constructability Lab, Orlando, Florida. ( Page 18, Figure 10)]. 5 18 32 1110 20 40 60 80 100 120Number of ObservationsReason of Water Intrusion Paint Defective Wall/Stucco/Foundation Doors/Windows Unknown to Homeowner
38 A B C Figure 2 3. Common sources of water intrusion through stucco cladding and the resulting damage. A) Crack in stucco at window corner. B) Small cracks in stucco rendering approximately 0.39 mm (1/64in) wide C) Water intrusion damage around window. [Reprinted with permission from Mullens, M., Hoekstra, B., Nahmens, I., and Martinez, F. (2006). Water Intrusion in Central Florida Homes During Hurricane Jeanne in September 2004, University of Ce ntral Florida Housing Constru ctability Lab, Orlando, Florida. (Pages 29,34, and 36; Figures 20,30, and 33)]
39 Figure 2 4 Possible leakage paths. [Reprinted with permission from RDH Building Engineering Limited. (2002a). Water Penetration Resistan ce of Windows: Study of Manufacturing, Building Design, Installation, and Maintenance Factors RDH Building Engineering Limited, Vancouver, B. C., Canada. (Page14, Figure 2.4 1)].
40 Figure 2 5. Critical barriers of the typical window -wall interface. Water Shedding Barrier Exterior Moisture Barrier Vapor Barrier Air Barrier Critical Barriers: Stucco Cladding Exterior Sealant Sill Flashing House wrap Sheathing Interior Gypsum Board Polyethylene Sheet Sub Sill Insulated Stud Space
41 CHAPTER 3 WINDOW INSTALLATION TECHNIQUES Water intrusion at the window -wall interface is a significant problem associated with the installed window assembly (Katsaros 2005) Six possible leakage paths common at fenestration openings have been well documented by the research of RDH (2002a) W hether entering the window -wall interface from the adjacent wall assembly or through the window frame these leakage pat hs are considered the most frequent and pose the highest risk of subsequent damage. In response, window installation techniques are altering moisture management strategies. During the installation of a window, traditionally it was common to apply a conti nuous bead of sealant around the backside of the mounting flange to prevent the intrusion of water at the exterior interface of the window and the wall However, installations of this manner are vulnerable to water penetrating the window -wall interface be cause no provisions are made within the details of the installation to prevent this water from entering the building interior N ewer installation methodologies are being developed to correct this issue seeking to manage this water and redirect it to the e xterior. The two methods may be respectively referred to as water barrier method and drainage method installations. Water Barrier Method vs. Drainage Method Water penetration performance of windows in residential construction is highly dependent on t he moisture control strategy used to integrate the window into the critical barriers of the surrounding building envelope. Window installation methods can be divided into two approaches based on these moisture control strategies. The first is the water ba rrier method, which seeks to prevent water migration through the external interface of the window and the wall system by creating a water shedding barrier that is coincident with the exterior moist ure barrier and air barrier (RDH 2002a ). The second approa ch is the drainage method, which places the water
42 shedding barrier inside the interface to redirect any leakage to the drainage plane of the wall (Katsaros and Hardman 2007, Lstiburek 2008). In installations using the traditional water barrier method, such as those outlined in ASTM E 2112 prior to the 2007 revision, the interior surface of the windows mounting flange receives a continuous bead of sealant to provide a moisture and air barrier at the external interface of the window opening. The details of a water barrier method installation are shown in Figure 3 1 on a cross -section of a typical window installed in a wood frame wall. While this method is currently considered common practice for the installation of fenestration products, several shortcomings have been identified in the literature (e.g., Katsaros and Hardman 2007). This installation technique makes no provisions to control leakage that may occur through the window -wall interface due to incorrect installation or through the window itself due to the deterioration of the window c omponents over its service life. In terms of the leakage paths defined in F igure 2 4 water barrier methods only resist water ingress through leakage paths L3 and L5, making no provisions to manage water intrusion throu gh leakage path L4, one of the most frequent as reported by RDH (2002a). The effectiveness of water barrier installations is defined by the ability of either an external wall cladding system or a concealed barrier to prevent water ingress into the building. In installations using the drainage method as a water penetration control strategy, such as thos e described in FMA/AAMA 100 ( 2007), gaps are left in the exterior seal at the sill between the mounting flange and the sill pan flashing and a continuous se al is provided around the interior perimeter of the window. In contrast to the exterior water barrier method, the moisture and air barrier for the drainage method is located at the interior interface between the fenestration product and the rough opening. Water that may leak into the window opening through the
43 window joinery or at the interface with the adjacent wall assembly is stopped at this interior barrier and is redirected to the drainage plane of the wall through the openings at the exterior seal. The details of a drainage method installation are shown in Figure 3 2 on a cross -section of a typical windo w installed in wood frame wall. While the drainage method installations may not prevent water intrusion at the interfaces interior to the window frame (i.e. leakage paths L1, L2, and L6), the location of the interior seal enables the installation to manage water through leakage paths L3, L4, and L5. The drainage method assumes that the exterior cladding of the wall assembly is not the primary mois ture barrier of the wall system and takes a practical approach to handling any water that may seep into the opening. For this reason, the most re cent version of ASTM E 2112 ( 2007) includes a recommendation for the use of pan flashings, which constitute a drainable installation, for all windows and doors. Standard P ractices of Installation Standard installation practices common to hurricane -prone regions were selected for evaluation. Windows were installed into test specimen walls representative of const ruction practices common to the s outheast U.S. Eighteen 2.4 m by 2.4 m (8 ft by 8 ft) wall specimens were constructed varying uniquely in wall structural system window operating type, window frame material, exterior cladding system, and installation meth odology. Water barrier type and drainage type installations were tested. Water barrier method installations were implemented according to the guidelines ASTM E 2112 (2007) and the barrier provisions of the draft FMA/AAMA 200. Drai nage method installatio ns were implemented according to the guidelines of FMA/AAMA 100 and the drainage provisions of the draft FMA/AAMA 200. The installation methods were tested for water intrusion in their in -service conditions and the results were used to evaluate and compar e their performance. The installation methods are briefly described in the sections herein.
44 ASTM E 2112 Standard Practice for Installation of Exterior Windows, Doors and Skylights This practice provides the minimum requirements for the installation of fen estration products in new and existing construction. The installation process is detailed from the pre installation through the post installation procedures for residential and light commercial buildings. The guidelines and recommendations of this docume nt serve as the basis for the majority of manufactures installation instructions. While the revised ASTM E 211207 document provides provisions for a drainage method installation through the recommended use of a pan/sill flashing (where applicable), only the water barrier installation guidelines were evaluated in this study. Within the standard are several installation methods unique in the manner with which the critical barriers of the fenestration product are integrated into the wall assembly. The sele ction of the appropriate method requires that the installer understand the building system employed to prevent water and vapor penetration through the building envelope. The installation methods are first divided into two broad categories based on the moi sture management strategy of the wall assembly. These two categories are installations for surface barrier wall systems and installations for membrane/drainage wall systems. Surface barrier wall systems are those systems which are intended to manage all w ater at the outermost exterior surface. Examples of these systems include exterior insulating finishing systems (EIFS) and single -wythe masonry. In general, these systems make no provisions to manage any incidental moisture that may bypass the water shed ding surface. In this sense, the water shedding barrier and the exterior moisture barrier are in the same plane. The continuity between the critical barriers of the fenestration assembly and the wall assembly is attained by
45 sealant joints. Two of the wa ter barrier installations into the concrete masonry unit (CMU) wall specimens of this study follow the procedures of this method. The second installation category is for windows being installed into membrane/drainage wall systems. Membrane/drainage wall s ystems employ a concealed weather resistant barrier such that the exterior surface of the wall components are not the sole means of preventing water intrusion. Examples include wood frame walls with an exterior cladding system such as stucco or siding. I n these systems the exterior cladding serves as the water shedding surface and the exterior moisture barrier is provided by the weather resistant house wrap applied overtop the sheathing. Unlike the surface barrier systems, the water shedding barrier and the exterior moisture barrier are not located in the same plane. The continuity of the critical barriers of the fenest ration unit and the wall system is achieved by directly sealing the window to the WRB of the wall. The water shedding barrier is kept co ntinuous by sealing the cladding return at the window frame. Installation methods for windows with mounting flanges in membrane/drainage wall systems are further subdivided into four separate methods in this standard: A, B, A1, and B1. The integration of the fenestration unit into the WRB of the wall relies on the implementation of proper flashing techniques. The four methods differ in the application sequencing of the WRB and the flashing, as summarized in Table 3 1. It is the responsibility of the contractor to verify the appropriate installation method and maintain its consistency throughout the project. The two water barrier window installations for the wood framed walls of this study follow the guidelines of Method A1. The WRB was applied prior to the installation of the window. The window unit was then installed into the rough opening and its mounting flange was integrated into the WRB with the appropriate shingle lap flashing.
46 FMA/AAMA 100 Standard Practice for the Installation of Windows with Flanges or Mounting Fins in Wood Frame Construction fo r Extreme Wind/Water Conditions This standard practice covers the installation of windows with integral, structural mounting flanges in new construction employing a membrane/drainage wall system. The i nstallation details of this standard were developed specifically to allow any incidental water entering the window -wall interface to be redirected to the drainage plane of the wall. This standard is designed particularly for window installations in buildi ngs subject to excessive wind -driven rain exposure, such as the hurricane -prone regions of the southeast U.S. The methods of this standard are water tested up to and including a water test pressure of 574.6 Pa (12 psf) according to the ASTM E 331 water te st to ensure their faculty under extreme wetting conditions. The installation procedures of FMA/AAMA 100 are based on the same basic sequencing as that of ASTM E 2112 Method A1. The characteristic differences between the two methods rely on the provisions provided to manage incidental water entering the window opening. The details of this standard qualify it as a drainage method installation. In order to ensure drainage to the building exterior, the details of the FMA/AAMA 100 require a three component d rainage assembly that employs a sill flashing with water tight end dams a discontinuous seal at the sill, and a robust interior air/water seal. A sill flashing is installed over the sill to prevent water from migrating around the sill and into the wall assembly. A discontinuous exterior seal is then formed by leaving two, 50.8 mm (2 in) voids in the sealant near each corner of the window sill on the backside of the mounting flange to allow water collected by the sill flashing membrane to exit to the building exterior. An interior air/water seal prevents liquid water from entering the building interior, redirecting it to the drainage voids at the sill and onto the drainage plane of the wall. The interior seal shall be compose d of either backer rod and a gunnable sealant or a low expansion aerosol foam, and applied continuously around the interior perimeter of the window
47 between the window frame and the rough opening. In this system, the interior perimeter seal comprises the exterior moisture barrier and the air barrier of the window interface. The adhesive and cohesive strength of this interior seal as well as its durability is crucial to the adequate performance of the installation method. The eight drainage method installa tions in the w ood framed walls of this study use the procedures of this standard practice. FMA/AAMA 200 Standard Practice for the Installation of Windows with Frontal Flanges for Surface Barrier Masonry Construction for Extreme Wind/Water Conditions (draft ) At the conception of this study the FMA/AAMA 200 was a draft document. The windows installed into the wall specimens of this study followed the provisions of the document current at that time. Revisions have since been made to the document and the standard has been accepted. The revisions to the document did not greatly affect the sequencing or the details of the installation procedures. The results of this study as they pertain to the water penetration resistance of the installations prescribed in th is practice are likewise considered valid for the newly revised document. The installation procedures discussed herein are based on the draft FMA/AAMA 200 document current at the time of wall specimen construction. This standard details the installation o f windows with integral or applied, nonstructural flanges in buildings with surface barrier wall construction. The standard provides the provision for both a water barrier installation and a drainage installation. The most distinguishable characteristic of the draft FMA/AAMA 200 standard is the application of a liquid applied flashing (LAF) in and around the window rough opening prior to installation. The LAF is applied to the CMU to aid in the water penetration resistance of the wall assembly around the widow opening. The methods of this standard are water tested up to and including a water test pressure of 574.6 Pa (12 psf) according to the ASTM E 331 water test to ensure their faculty under extreme wetting conditions.
48 In the water barrier installatio n provisions of the draft FMA/AAMA 200, a continuous seal is applied to the backside of the mounting flange around the perimeter of the window. When the window is installed into the rough opening the seal is compressed against the window bucking creating an exterior moisture barrier. The installation also requires that an air/water seal is achieved around the interior perimeter of the window between the window and the rough opening. While the barrier installation procedure provides a redundant barrier to water intrusion (both a continuous exterior and interior seal), these installation details are not recommended by the standard for buildings susceptible to increased levels of rain exposure. Incidental water that may enter the window opening through the window assembly or superficial cracks in the exterior cladding will not be drained to the exterior and will be absorbed by the wall system. Water that is absorbed by the wall system inhibits its ability to rapidly dry following extreme wetting conditions. Two of the water barrier installations in the (CMU) wall specimens of this study follow the procedures of this method. In buildings with high moisture exposure, the draft FMA/AAMA 200 recommends the implementation of the drainage installation provisions prescribed in the standard. The drainage installation method of the draft FMA/AAMA 200 follows the same procedures as that of the barrier installation method with the exception of a discontinuous seal at the sill. Two 50.8 mm (2 in) voids are left near the corners of the window sill in the sealant applied to the backside of the mounting flange to allow the drainage of water to the building exterior. The effectiveness of this installation method relies heavily on the ability of the interior seal to preve nt the migration of water to the building interior. The drainage method installations in the CMU wall specimens of this study follow the procedure s of this method.
49 Chapter Summary Water penetration performance of windows in residential construction is highly dependent on the moisture control strategy used to integrate the window into the critical barriers of the surrounding building envelope. Window installation methods can be divided into two distinct approaches based on the moisture management strategie s employed: water barrier installations and drainage installations. Water barrier installation methods seek to prevent the intrusion of water at the exterior interface of the window and the wall assembly. These methods make no provisions to manage incidental moisture entering the opening through the window or the surrounding wall assembly. Drainage method installations place the exterior moisture barrier at the interior interface of the fenestration unit and the rough opening. Water that may enter the opening during the service life of the window is prevented from infiltrating into the building by the interior seal and is redirected to the drainage plane of the wall through small voids left in the sealant at the sill. Due to the manner in which they ma nage water ingress, drainage installation methods have been recommended for use in buildings susceptible to extreme wind-driven rain exposure. Three standard practices were selected for this study in order to test the effectiveness of the water penetratio n resistance of the two installation methodologies.
50 Table 3 1. Installation procedure summary for windows with mounting flanges installed into membr ane/drainage type wall systems. [Adapted from ASTM. (2007). Standard Practice for Installation of Exterior Windows, Doors, and Skylights. ASTM E 2112-07, American Society for Testing and Materials, West Conshohocken, Pennsylvania. (Page 33, Table 8)]. Head and jamb flashing applied over the mounting flange Jamb and sill flashing applied behind the mounting flange WRB is applied after window installation Method A Method B WRB is applied before window installation Method A1 Method B1
51 Figure 3 1. Water barrier method installation detail. Sill flashing Lack of interior seal allows water penetrating to the window -wall interface to enter building interior Water that may enter the window wall interface from leaks penetrating the window or migrating from the surrounding wall assembly Continuous bead of sealant applied to backside of mounting flange Water intrusion is prevented at the external interface of the window and the wall Water is redirected to the concealed water resistive barrier of the wall Water resistive barrier
52 Figure 3 2. D rainage method installation detail. Drainage channels left in the exterior seal at the sill allow water to drain from the window -wall interface Water that may enter the window wall interface from leaks penetrating the window or migrating from the surrounding wall assembly Interior seal provided by backer rod and sealant (shown here) or low expansion foam prevents water intrusion into the building Water exiting the drainage channels in the exterior seal is redirected to the concealed water resistive barrier of the wall Water from the window wall interface is redirected by the interior seal to the drainage channels in the exterior seal Sill flashing W ater resistive barrier
53 CHAPTER 4 EXPERIMENTAL PROCEDURE The window -wall interface has been identified as a key source of water intrusion problems in buildings, particularly in extreme exposure regions. Considerable effort has gone into the development of robust practices to enhance the water management performance of window installations, including redundant drainable methodologies with enhanced flashing and sea lant products, but actual performance testing of the installation under real life extreme exposure conditions has been severely lacking. This study examines the effect of extreme wind driven rain exposure on a variety of window installation details, utilizing testing protocols derived from conventional methods and a unique hurricane simulation apparatus. The UF H urricane S imulator generates Saffir -Simpson Hurricane Scale Category 3 hurricane force wind pressures and wind-driven rain to test wall specimens utilizing both wood frame and concrete block construction, along with various faade systems that are common to the Southeastern U S region. The experimental focus of this testing was the full -scale evaluation of in -service window installations. The te sting apparatuses and procedures developed for this full -scale testing are discussed herein. Testing Apparatuses Unlike current testing standards that evaluate fenestration products in isolation (e.g. ASTM E33100, AS/NZS 4284:1995, JIS A 1517), the test methods used in this study focus on evaluating the water penetration resistance of the installed window and the wi ndow -wall interface. Windows were installed into 2.4 m by 2.4 m (8 ft by 8 ft) wall specimens using water barrier and drainage type installation approaches. To accommodate the full -scale window -wall specimens each ap paratus used for testing was custom de signed and built. A negative pressure chamber was constructed to evaluate the wall specimens under static and cyclic air pressure
54 differences and the UF Hurricane Simulator was used to subject wall specimens to dynamic pressure sequences. Test specimens were representative of completed construction and careful precaution was taken in their movement to avoid damage that may lead to unnecessary water intrusion. A specially designed air -c aster transportation unit was fabricated to transport the specimens fr om the staging area to the test locations. Air -Cast e r Transportation Unit The careful transportation of the specimens was a topic of concern. With the CMU walls weighing approximately 8,818 kg (4,000 lbs), precautions had to be taken to prevent damage to the specimens. The focus of the testing was aimed at the water penetration resistance of the window and the surrounding wall system. Cracks in the stucco or in the CMU walls incurred during the movement of the specimens could introduce incidental water i nto the window -wall interface during testing that would not be typical of a properly finished wall system. In order to facilitate the careful movement of the specimens around the testing facility a specially designed air -cast e r transportation unit was con structed. The air -cast e r transportation unit (shown in Figure 4 1) consists of two steel sleds each equipped with two 8,818 kg (4,000 lb) capacity air -cast e rs and two 17,637 kg (8,000 lb) capacity bottle jacks. An air -cast e r is a perforated rubber membran e attached to the underside of a steel frame. When supplied the specified air pressure rated for the particular cast e r, the rubber membrane fills until excess air s p ills out of the perforations. The air is forced to escape beneath the rubber membrane lif ting the cast e r on a thin film of air. To move the specimens each sled is independently positioned underneath the base of a wall at opposite ends so that the wall is centered between the jacks. The sleds are then connected to each other with steel strut s at the base as well as steel cross bracing between the jacks to prevent spreading during movement. In the assembled configuration the air -cast e r transportation unit measures 1.2 m (4 ft) in the
55 dimension perpendicular to the wall to prevent overturning. The four bottle jacks are simultaneously raised until the both sleds touch the base of the specimen. The specimen is then fastened to four steel cages housing the hydraulic jacks. To ensure the weight of the wall was properly distributed the specimens were leveled and plumbed before transport. The air -cast e rs were then slowly pressurized with an air compressor to lift the specimens on a 9.5 mm (3/8 in) film of air. With the use of the air -cast er transportation unit two technicians were able to move the specimens to and from testing sites. Negative Pressure Chamber A negative pressure chamber (Figure 4 2) was constructed to observe the water intrusion through the test specimens during the stat ic and cyclic tests. The design was based on the general arrangement for a water penetration test apparatus shown in Figure 1 of ASTM E 110500. The chamber is comprised of a 25.4 mm (1 in) thick acrylic sheet attached to a steel frame measuring 2.4 m by 2.4 m by 304.8 mm (8 ft by 8 ft by 1 ft), and is designed to a maximum deflection criterion of 1.6 mm (1/16 in) at a pressure of 2,872.8 Pa (60 psf). The face of the ac rylic sheet is trimmed with 2x4 s providing the airspace pressurized during the tests. Two centrifugal blowers provide the required differential pressures for the loading functions; as well provide sufficient airflow to overcome any leakage through and around the specimens. The blowers are capable of producing a pressure differential of 2, 394 Pa (50 psf); however, specimens were not tested at these levels in order to maintain the integrity of the windows for subsequent testing. Pressures for the loading functions were achieved with two 101.6 mm (4 in) electro pneumatic valves operated by a custom active control system created using National Instruments LabVIEW 8.5 software. The two valves work in unison to perform the tests. While one valve regulates the air movement in the suction line from the chamber to the blowers, the other valve ve nts the chamber to the atmosphere, partially equalizing the pressure difference.
56 Pressure readings inside the chamber are provided by a pressure transducer accurate to 6.2 Pa (0.13 psf). The wetting conditions for the tests were suppli ed by a custom built spray rack, which is calibrated to deliver between 223.5 mm/hr and 447.0 mm/hr (8.8 in/hr and 17.6 in/hr) of simulated rain uniformly across the wall in accordance with ASTM E 110500. The water running off the wall is collected, re -circulated, and filte red back to a 757 L (200 gal) tank. To perform the tests on the negative pressure chamber, a specimen is attached to the face of the chamber with clamps compressing a cellular urethane seal. A wetting chamber, constructed to house the spray rack, is pla ced against the face of the specimen, and the spray rack is started. The pressure loading function appropriate for the test is initialized by an engineering technician running the control system for the blowers. Observers monitor the water penetration th rough the specimens from the interior through the large viewing area provided by the acrylic sheet. Hurricane Simulator The water penetration resistance of the test specimens in response to wind -driven rain under dynamic pressure loading sequences was eval uated using the UF H urricane S imulator. The 2.09 MW (2800 hp) simulator is the largest portable version of its kind and is capable of replicating turbulent wind and rain loads on a full -size low rise structure as shown in Figure 4 3. Four 0.52 MW (700 hp ) diesel engines spin eight hydraulically actuated vaneaxial fans to produce stagnation pressures of 1,675.7 Pa (35 psf) on the windward wall. Specially designed venture inlets force the air to travel perpendicular to the fan discs for maximum efficiency. Air accelerates through the contraction and passes a series of custom designed neutral shape NACA airfoils positioned at the exit and designed to discharge water at the trailing edge to simulate wind driven rain. The airfoils are connected to a hydrauli c rotary actuator, which changes the
57 wind direction ; however, l ateral turbulence was not incorporated into the loading for this testing. To recreate hurricane conditions an active computer control system modulates wind speed by varying fan RPM, creates di rectional effects by articulating the airfoils at the exit, and injects water into the flow field to simulate rain. The control system utilizes a fast running PID -control operated in the LabVIEW 8.5 environment. For dynamic testing, the specimens were integrated into the windward wall of a full scale residential house model and subjected to a designed load history via the H urricane S imulator. The house model (Figure 4 4) measured 9.8 m (32 ft) in length and 4.9 m (16 ft) in width and was designed accor ding to the 2004 FBC based on a 225.3 km/h (140 mph) wind velocity and exposure C as defined by ASCE 7 05 (2006) The deck was framed with 2x8 joists spaced 304.8 mm (1 ft) o.c. on top of a steel box truss system to elevate the house into the wind field of the simulator. Brackets positioned every 609.6 mm (2 ft) were welded to the box truss system during fabrication to allow the deck to be attached with 12.7 mm (1/2 in) bolts through every other joist. Windward and side walls were frame d with 2x4 s spaced 406.4 mm (16 in) o.c. and attached to the deck with galvanized metal flat straps and 12.7 mm (1/2 in) lag screws spaced every 812.8 mm (32 in). The windward wall was built with a 2.44 m (8 ft) span missing at the center to accommodate the wall specimens. Specimens were raised into the windward wall with a 26,455.4 kg (12,000 lb) capacity scissor lift and attached to the trusses with hitch pins to simulate a typical roof to wall connection. The leeward side of the house model was kept open and trusses we re supported by double 241.3 mm (9 1/2 in) deep laminated veneer lumber (LVL) beams held up by 4x4 columns. Prefabricated trusses were used to form the 6/12 monoslope roof. A 1 ft closed soffit was detailed at the overhang of the walls. A deck construct ed at the base of the windward wall along with a flexible wind resistant fabric attached to the exit of the
58 simulator formed a ground barrier to ensure the proper flow field shape was achieved against the house model. Hydraulic Leakage Pressure Apparatus T he effectiveness of drainage method installations relies heavily on the performance of the interior moisture/air seal. As per the installation guidelines of FMA/AAMA 100 and the draft FMA/AAMA 200 document this interior seal may be achieved using either a backer rod with gunnable sealant or a low expansion aerosol foam. Water penetration results from the four rounds of pressure loading displayed varying performance between the two methods. The difference in the performance was attributed to the sealants respected ability to fill any voids between the window and the rough opening as well as their ability to maintain the adhesion strength necessary to hold back increasing amounts of water at high pressures. In order to test the correlation between the adhe sion properties of the sealants and their water penetration resistance a test was formulated to determine the leakage pressure of a seal subject to an increase d hydraulic head. The results were then compared with the adhesion strength of the sealants dete rmined according to ASTM C 794 01. To test the leakage pressure of the interior seals, a reservoir was gradually filled with water to form a hydraulic head that could be converted to a pressure acting on the backside of the seal. Sill specimens were creat ed from 457.2 mm (18 in ) pieces of both aluminum and vinyl extruded sill stock (Figure 4 5). The sill pieces were sealed on both sides and mounted on a mock -sill such that the gap for the interior seal was consistently 6.4 mm (1/4 in) for all specimens. The sills were then sealed to hold water within the cavity of the sill frame. Two valves were installed into the sills. One valve was used to allow water to enter the sill cavity while the other valve was used to evacuate air from the sill cavity as it f illed with water. The sill specimens were then connected to a reservoir constructed from a 37.9 L (10 gal) aquarium with vinyl tubing (Figure
59 4 6). Water was siphoned into the reservoir from two 18.9 L (5 gal) buckets through 6.4 mm (1/4 in) vinyl tubing As the water level in the reservoir increase s the sill cavity fills and a hydraulic pressure is applied to the backside of the interior seal. The height of the water column in the reservoir and the height of the seal above a specified datum were measur ed and recorded to the nearest 1.6mm (1/16 in). The hydraulic head was determined by subtracting the elevation of the seal above the datum from the elevation of the water column inside of the reservoir. The leakage pressure for the seal was calculated by multiplying the hydraulic head at the time of leakage by the density of water. Testing Protocols and Sequencing In order to test the effectiveness of the water barrier method and the drainage method approaches to the installation of fenestration products, each window -wall specimen was cycled through four rounds of pressure loading and water testing. In order, specimens were evaluated under a static, a cyclic, as well as an amplitude and frequency -modulated pressure load sequence, followed by a repeat of the initial static test. A minimum dry time of 2 days was allotted between consecutive tests. The static test was performed at the beginning and the end of the testing sequence to determine whether the occurrence of a leakage path in a particular test pr edisposed the window to water intrusion through the same path in subsequent tests. Such behavior would comprise the validity of comparing the water penetration behavior of the specimens between testing methods. Similar leakage times and pressures during the initial and final static tests confirmed the integrity of the specimens. To further prevent permanent damage to the test specimens the pressure loading functions for the tests, described herein, were tailored such that the maximum test pressure did not exceed 50 percent of the design pressure (DP) of the installed window.
60 Static Air Pressure Difference The static test was derived from Procedure A in ASTM E 110500, which outlines the procedures for a static uniform pressure difference test. Following the details of Procedure A, specimens are subjected to a specified static air pressure difference, along with a specified rate of water spray, for 15 minutes. Points of water penetration are observed and documented. The static pressure loading of this te st method makes it difficult to determine the water penetration performance limits of a fenestration product. To determine the water penetration thr eshold of a particular specimen an iterative process would be required where subsequently larger pressures would be applied until the onset of leakage occurred. Moreover, because a static pressure must be specified and held constant through the duration of the test, evaluating the water penetration performance of the assembly in response to hurricane wind loading would be difficult. Pressure loading in these extreme weather events is far from constant and can rapidly fluctuate making the task of defining a specific pressure at which to evaluate the specimen complicated In the static pressure test designed for this study, specimens are loaded with an initial pressure of 137 Pa (2.86 psf) for 5 minutes before the pressure is linearly increased over 15 minutes to half of the windows DP rating (Figure 47). The initial pre ssure difference was chosen because it is the standard test pressure difference at which water penetration is to be determined per ASTM E 33100, and the duration of linear pressure ramp was selected to maintain consistency with ASTM E 110500 Procedure A. Loading in this manner allows the leakage paths in the specimen to be primed during the initial static portion so that water intrusion can be documented at higher pressures. The benefit inherent in linearly increasing the loading is that the maximum pr essure difference at which the assemblies resist the infiltration of water may be recorded.
61 Cyclic Static Air Pressure Difference The cyclic pressure test was derived from a combination of ASTM E 226804 and AAMA draft of the Voluntary Specification for R ating the Severe Wind Driven Rain Resistance of Windows, Doors and Unit Skylights The cyclic test utilize s the same testing chamber used for the static test. ASTM E 2268 defines a median pressure of 137 Pa (2.86 psf) in which specimens are to be pulsate d at a pressure difference equal to 50% and 150% of the median test pressure. The AAMA XXX-XX draft specification provides a range of increasing performance levels from 1 to 10, where level one pulsates from 239 to 718 Pa (5 to 15 psf) and level ten pulsa tes from 670 to 2011 Pa (14 to 42 psf). For the specimens tested in this study, a pulsating pressure schedule was developed with ranges lower than specified in AAMA draft specification (to accommodate for ASTM E 226804) up to the highest performance level whose upper pressure was half the manufacturers DP rating The pressure loading function for the test as shown in Figure 4 8, begins by loading the specimen without water for 1 minute at an air pressure difference equal to 50 percent of the DP rating for the window followed by a rest period of 1 minute where the specimen is equalized to atmosphere. Water is then incorporated for the remaining duration of the test consisting of 60 cycles over 3 minute periods at each of the varying pressure levels such that 50 percent of the windows DP rating is not exceeded. For example, a wall system with a DP60 window was loaded to the performance level that pulsed from 479 to 1436 Pa (10 to 30 psf). In this respect, the total duration of the cyclic static air pre ssure difference test is a function of the DP rating of the window and varies among specimens. Dynamic Pressure An amplitude and frequency -modulated sinusoidal pressure loading sequence was applied to the test specimens using the hurricane simulator for the dynamic test (shown in Figure 4 9).
62 The loads were designed using 10 minute wind speed observations collected by the Florida Coastal Monitoring Program that were converted to velocity pressures. It was conservatively assumed that there was perfect a erodynamic admittance between the free stream velocity pressure and the stagnation pressure on the windward wall. Records with a mean velocity greater than 20 m/s (44.7 mph) were extracted and linearly detrended. The longitudinal velocity component was c alculated and passed through nine band -pass filters in 0.1 Hz pass band increments. The lowest three frequencies were used (0.15, 0.25 and 0.35 Hz). The peak amplitude for each pass band was recorded and divided by the mean velocity over a 10 minute recor d, to determine the peak amplitude to mean velocity ratios. Data was stratified into three turbulence intensity regimes, of which the middle turbulence range ( / = 0.15 0.20) was chosen since it represents open exposure (terrain) conditions. The 50th percentile peak values were employed to construct a sinusoidal loading pattern at three different velocity thresholds that correspond to 239, 479, and 718 Pa (5, 10 and 15 psf). Inter ior Moisture/Air Barrier Testing During the initial four rounds of pressu re and water testing, varying performance of the different sealing methods used to create the interior moisture/air barrier was observed. The installation procedures detailed in both the FMA/AAMA 100 and the draft FMA/AAMA 200 document allow the interior seal to be formed using either a backer rod with a gunnable sealant or a low expansion foam. It was noticed that frequently under even median pressure differentials that a water column developed inside the jamb of the window frame. At increased pressures th is water bypassed the interior seal at the interface with the sill flashing membrane thus penetrating the exterior moisture barrier of the wall The water penetration was attributed to the inability of the interior sealant to maintain adequate adhesio n with the sill flashing membrane at increased pressures. To test this hypothesis, a selection of gunnable sealants and low expansion foams
63 were subjected to a hydraulic leakage pressure test to determine their water penetration thresholds when installed as an interior seal. The hydraulic leakage pressures for the gunnable sealants were further compared to their adhesionin -peel strengths according to ASTM C 79401 to determine if a correlation exist ed between the two performance criterions. Sealant sampl es were selected for testing based on composition, price, availability, and frequency of use in practical construction. Seven gunnable sealants conforming to either ASTM C 920 or AAMA 808.3 and five low expansion foams conforming to AAMA 812 were selected for testing. The anonymity of the products used was maintained using a four character alpha numeric indexing system. The first two characters indicate whether the specimen classifies as a gunnable sealant (S) or a low expansion foam (F) and indexes the order number within this classification. The last two characters refer to the sill material, aluminum (A) or vinyl (V), and the test number (1,2, or 3). Table 4 1 lists the sealant samples and their compositions used in the interior moisture/air barrier testing. Hydraulic leakage pressure t est Sealants were applied to the sill specimens to replicate an interior moisture/air barrier as it would occur in a drainage method installation. Three sill specimens were created per sealant, from each of vinyl and a luminum sills, totaling to 72 specimens. All sealants were applied in the same manner as the original window -wall test specimens so that the results could be compared. Gunnable sealants were installed with a 9.5 mm (3/8 in) backer rod to prevent threepoint adhesion and hand tooled to ensure proper sealant shape Every seal was tooled by the same installer with the utmost attention to maintain consistency among specimens Low expansion foam seals were not trimmed prior to testing. Following a pplication, the sill specimens were allowed to cure for a minimum of 72 hours before testing.
64 For each test run, a sill specimen was connected to the hydraulic leakage pressure apparatus and the reservoir was gradually filled with water. Once water was observed to permeate the outer plane of the seal the height of the water level in the reservoir, the water temperature, the location along the sill, along with a brief description of the leak were recorded. The first two leaks for each test were documented. In the case that the interior seal showed no signs of water intrusion when the reservoir was completely filled specimens were slowly lowered below the testing table to gradually increase the pressure on the seal. Due to the limiting factors of the t esting table height and the vinyl tub ing lengths the maximum achievable pressure for the testing apparatus was 4788 Pa (100 psf). Using the testing table as a datum the hydraulic head was determined by subtracting the elevation of the seal from the elevati on of the water level in the reservoir. The hydraulic head was then converted to a pressure applied to the backside of the interior seal by multiplying the head by the density of water at the recorded water temperature. The resulting pressure was determi ned to be the hydraulic leakage pressure. Adhesion strength The adhesion strength of the gunnable sealant samples was determined based on the procedures of ASTM C 79401. This standard specifies adhesion testing to anodized aluminum, mortar slabs, and pla te glass material samples However, none of these materials are common adhesive substrates for the interior moisture/air seal. For this reason adhesion of the sealants w as also tested to SPF grade wood, painted aluminum coil stock extruded vinyl window stock and the sill flashing membrane used in the original testing. The adhesion of the sealants to the sill flashing membrane was of utmost concern as the majority of the water intrusion through the interior seal observed during the initial testing occur red at this interface.
65 The peel strength of the sealants was tested to seven different material substrates including unfinished aluminum, brick, plate glass, painted rolled aluminum, vinyl, sill flashing membrane (same used as in wall specimen testing), an d wood. Two specimens were prepared per sealant for each substrate. To create the specimens (Figure 4 10), sealant was extruded onto the material over an area sufficient to create two specimens 25.4 mm (1 in) wide by 76 mm (3 in) long and then screeded t o create a layer of sealant 1.6 mm (1/16 in) thick. Two pieces of metal screen 25.4 mm (1 in) wide were placed lengthwise on top of the sealant spaced 9.5 mm (3/8 in) apart. A second layer of sealant was then extruded over top of the metal screens, embed ding them in the sealant. The second layer of sealant was then screeded to create a sealant sample 3.2 mm (1/8 in) thick. Specimens were then cured for 21 days as follows: 7 days at 23 2C (73 3.6F), 50 5 % relative humidity; 7 days at 37.8 2C (100 3.6F) and 95 5 % relative humidity; and 7 days at 23 2C (73 3.6F) and 50 5 % relative humidity. After curing, two 25.4 mm (1 in) wide specimens were marked from each 3.2 mm (1/8 in) thick sealant layer. Specimens were carefully cut clea nly away from the surrounding sealant down to the substrate. The initial portions of the specimens were undercut to release them from the substrate to make certain the sealant would properly peel. An ATS 1101 testing machine was used to perform the adhes ion in -peel tests. A digital reading displayed the peel force in lbs as the specimens were peeled from the substrates at a rate of 51 mm/min (2 in/min). For each test the peak and average peel strengths as well as the failure mode were recorded. Failure modes were classified as adhesive, cohesive, or mixed mode. Examples of each are shown in Figure 4 12. The average value of the peel force was derived by the observer through careful monitoring of both the failure mode and the digital force reading. In cases where the sealant failed in
66 cohesion, the sealant was under cut to reinitiate peeling from the substrate. This enabled the observer to confirm the failure mode and develop a better sense of the average peel force. Chapter Summary Considerable effort has gone into the development of robust practices to enhance the water management performance of window installations, including redundant drainable methodologies with enhanced flashing and sealant products, but actual performance testing of the installat ion under real -life extreme exposure conditions has been severely lacking. This study focuses on the full -scale water penetration resistance testing of window -wall assemblies utilizing specially designed pressure loading sequences and a unique hurricane s imulator. In order specimens were subjected to a static test, a cyclic static test, a dynamic test, and then reevaluated under the initial static test to ensure specimens were not permanently damaged during testing. Given the large size of the full -scal e test specimens, custom designed testing apparatuses were fabricated. The performance of the drainage installations interior moisture/air seal prompted further sealant testing to determine a possible correlation between the hydraulic leakage pressure and the adhesion strength of the sealant.
67 Table 4 1. Sealant s ample m atrix. Index Sealant Type Composition Color S1 Gunnable sealant Siliconized acrylic latex White S2 Gunnable sealant Siliconized acrylic latex White S3 Gunnable sealant Siliconized acrylic latex Clear S4 Gunnable sealant One component polyurethane Grey S5 Gunnable sealant One component polyurethane White S6 Gunnable sealant One component polyurethane White S7 Gunnable sealant One component polyurethane White F1 Low expansion foam One component polyurethane Yellow F2 Low expansion foam One component polyurethane White F3 Low expansion foam One component polyurethane Yellow F4 Low expansion foam One component polyurethane Light green F5 Low expansion foam One component polyurethane Tan
68 A B C Figure 4 1. Air -c aste r transportation unit A) T est specimen loaded on air -caste r transportation unit B) Air -cast e r transportation device equipped with two hydraulic jacks enclos ed in steel cages. C) Air -caste r pad supporting test specimen weight during movement.
69 A B C D E Figure 4 2. Negative p ressure c hamber. A) Test specimen attached to chamber. B) View from interior of Negative Pressure Chamber. C) Wetting chamber containing the calibrated spray rack. D) Wetting chamber is sealed to the front of the test specimen to collect water. E) Observers monitor water penetration behavior through specimens during test.
70 A B C Figure 4 3. Hurricane s imulator. A) Test specimen loaded into house model for dynamic testing. B) Specially tuned inlets connected to vaneaxial fans. C) House model designed to accept 2.4 m x 2.4 m (8 ft x 8 ft ) test specimens into windward wall.
71 A B C D E Figure 4 4. Residential h ouse m odel. A) House model measuring 9.8 m x 4.9 m ( 32 ft x 16 ft ). B) Leeward side of House Model. C) Inside House Model. D) Steel box truss system support model. E) Wall specimen being inserted into windward wall.
72 A B C D Figure 4 5. Sill s pecimen for h ydraulic l eakage p ressure t est. A) Sill specimen mounted to maintain 6.4 mm (1/4 in ) gap for seal. B) Sill extrusion affixed to mock -sill with screws through mounting flange. C) Ends of specimens sealed to hold water inside sill cavity. D) Sealant applied to mimic interior moisture/air seal.
73 Figure 4 6. Hydraulic l eakage p ressure a pparatus.
74 Figure 4 7. Loading function for the static air pressure difference test Figure 4 8. Loading function for the c yclic s tatic air p ressure difference test
75 Figure 4 9. Loading function for the d ynamic p ressure test. 0 200 400 600 800 1000 1200 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00Pressure (Pa)Time (min)
76 A B C D E F Figure 4 10. Adhesion-in -p eel s pecimen c onstruction. A) Sealant is extruded onto substrate. B) Sealant is screeded to 1.6 mm (1/16 in). C) Metal screen in placed on top of the sealant layer. D) Second layer of sealant is extruded to embed screen. E) Second layer is screeded to create a 3.2 mm (1/8 in) specimen. F) Prepared specimen.
77 A B C D E F Figure 4 11. Adhesion-in -p eel t est. A) Individual specimens are marked. B) Two 25.4 mm (1 in) wide specimens are cut from the sealant layer. C) Initial portion of sealant is undercut to release it from the substrate. D) Specimens are loaded into the testing machine. E) Specimens are peeled at a rate of 51 mm/min ( 2 in/min). During testing specimens are cut to ensure failure type and to develop a sense of the average peel strength. F) The average and peak peel strengths along with the failure mode are recorded for each specimen.
78 A B C Figure 4 12. Sealant f ailure m odes. A) Adhesive failure. B) Cohesive failure. C) Mixed Mode (combination of adhesive and cohesive failures).
79 CHAPTER 5 SPECIMEN DESIGN AND CONSTRUCTION The test specimens consisted of a representative sample of common residential wall assemblies from the coastal southeastern United States. A total of ten wood frame and eight concrete masonry unit (CMU) wall sections were designed and constructed by indus try partners (Figure 5 1). Wood frame wall specimens were designed in accordance with Section R602 of the 2004 FBC. CMU walls were designed in accordance with Section 2104 of the 2004 FBC. Each wall measured 2.4 m by 2.4 m (8 ft by 8 ft) and featured a unique combination of wall type, exterior finish and window interface. The window interface was determined by the installation approach, which followed either an exterior water barrier or drainage method. Wood Frame Wall Specimens The wood frame wall spe cimens were constructed off -site by a certified home builder and delivered to the testing facility (Figure 5 2). The wood walls were framed using No. 2 Hem Fir 2x4 stud and plate members at 406 mm (16 in) o.c. Sheathing consisted of 11.1 mm (7/16 in ) ori ented strand board (OSB) installed in the vertical direction and attached with 8d common wire nails spaced 152.4 mm ( 6 in ) along vertical edges, 304.8 mm ( 12 in ) in the field, and 101.6 mm (4 in ) staggered along horizontal edges. The headers of the window openings were fra med using two No. 2 Hem Fir 2x6 s connected with two 16d common nails per 304.8 mm (1 ft) minimum (7/16 in OSB was sandwiched between header plies to create a header thickness similar to wall thickness). To transfer loads 31.8 mm ( 1 1/4 in ) x 20 gage ASTM A653 grade 33 steel straps with four 10d common nails in each end were installed under the sheathing and wrapped around the plates and the header of the wall specimen The bottom plates of the wall specimens were affixed to a steel C4x7.5 channel to facilitate the transport of the specimens and the attachment to the testing apparatuses (Figure 5 3). Steel channels were fastened with 12.7 mm ( 1/2 in ) bolts
80 with 76.2 m m x 76.2 mm x 6.4 mm ( 3 in x 3 in x 1/4 in ) washers through the channel at 6 09.6 mm (24 in) o.c. in order to simulate a typical slab -on -grade base plate connection. A polymer based house wrap material was installed on all wood wall specimens to serve as the WRB for the wall assembly. The windows were then installed by certified window installers and flashed (Figure 5 4). For installations using the barrier method as a water penetration control strategy, the flashing sequence of ASTM E 2112 Method A1 (2007) was followed, which is the basis of most window manufacturers installation instructions. A s elf -adhered, flexible sill flashing membrane was also installed in these wall openings, which although not required by the ASTM standard for exterior barrier installations, provides additional protection to the wood sill area. The s tandard p ractice described in FMA/AAMA 100 (2007) was used as the basis for installations e mploying a drainage method as a water penetration control strategy This standard allows the option to use either a 101.6 mm (4 in ) self adhered flashing or a 228.6 mm (9 in) mechanically attached flashing to flash the exterior jambs and head of the window as well as the option to use either a low expansion foam or backer rod and gunnable sealant to provide the interior perimeter seal. Different combinations of these installation options were tested. All sealants used in these installations met the re quirements of either or both ASTM C 92005 and AAMA 808.305 for gunnable sealants and AAMA 812 04 for low expansion foams. After the windows were installed, the wall specimens exterior cladding system, consisting of either fiber cement board or stucco, w as applied. The fiber cement board was attached by a professional installer in accordance with the manufacturers installation instructions. The fiber cement siding planks were blind nailed to the structural framing of the wall using siding nails (2.28 m m shank x 5.61 mm HD x 51 mm long) installed 10 mm ( 3/8 in) from the edge of the
81 planks and 25 mm (1 in) from the top. A minimum 32 mm (1 1/4 in) overlap was maintained between planks. The windows, as well as the top and the sides of the walls were fram ed with fiber cement trim pieces to provide a clean finish on all the fiber cement siding planks. Horizontal flashing was provided at the head of each window. The stucco was applied over lath in a three coat application consisting of a scratch coat, brow n coat, and finish coat in compliance with ASTM C 926 06 and ASTM C 106306. Expansion joints were placed at the four corners of the window on all wood wall specimens receiving stucco to allow for the proper movement of the stucco while curing. It was necessary to provide 10 1.6 mm (4 in) decorative banding around the windows to accommodate the large projection from the integral mounting flange. These bands are purely decorative and were installed onto a scratched surface around the window after the prop er stucco application had been applied to the rest of the wall. The stucco was allowed to cure a minimum of 7 days before being tested with a pH pen to ensure that a pH level below 13 was obtained prior to the application of paint (Figure 5 6). The requi red pH of the stucco is critical in the painting process to reduce the risk of the paint being subjected to alkali burn. All wall specimens, fiber cement board and stucco, received a three coat paint finish. The first coat was an alkali resistant prime r applied in a 2 mil thickness. The second and third coat was a construction grade acrylic paint with a slight tint applied to a thickness of 4 mil wet. A minimum cure time of three hours was given between coats and the walls were not tested until a minimum of 10 days after the finish coat was applied to the wall cladding. Masonry Wall Specimens The CMU walls were constructed in accordance the requirements of Sections 2104.1.1 through 2104.5 of the 2004 Florida Building Code and ACI 530.105/ASCE 605/TMS 60205.
82 A licensed professional masonry contractor built the walls directly on top of a 203 .2 mm (8 in) steel MC8 X 21.4 channel (Figure 57). Number 5 rebar was welded to the base channel prior to construction to provide the vertical reinforcing steel for the walls as wel l as to prevent the wall sections from overturning while moving them between testing sites. The wall specimens were constructed from normal weight 203.2 mm (8 in) CMUs (ASTM C 9001a) in a typical running bond with a Type S mortar (ASTM C 27001a). Down-cells on both sides of the window, and at both sides of the wall were filled with a coarse grout (ASTM C 47601) in each wall specimen to provide flexural rigidity. Bond beams were poured at the top and the base of the wall. To enable the viewing of any water that may leak into the cavities of the wall system without compromising the window -wall interface, the block faces on the interior side of the wall above the poured base beam were removed, leaving open gaps on the interior block surface. The windows were sized to accommodate 19 .1 mm ( 3/4 in) pressure treated wood bucking in the rough opening (Figure 5 8). Due to the slight variations in the construction and window sizes, some leniency was required; a maximum of 25 .4 mm (1 in) bucking was accepted. Bucking was not required at the sill of the opening because the CMU walls were constructed with a precast concrete sill. Two precast sill designs were utilized ; some sills were flush with the exterior block surface (flush sill) and others featured a protr uding edge (face sill). The windows were installed into the CMU walls following the guidelines of the traditional exterior barrier installation described in ASTM E 2112 (2007) or the installation details of either the modified exterior barrier method or t he drainage method given in the draft standard FMA/AAMA 200 (200x). Under the barrier installation provided by the draft FMA/AAMA 200 (200x) standard, a continuous exterior seal between the window and the wall system is required as well as an interior se al around the entire window perimeter. The interior seal may be
83 achieve d using either a backer rod and gunnable sealant or a low expansion foam compliant with the specifications listed above for the wood wall systems. Both of these options were tested. T he drainage method described in this standard follows the same installation procedures as the barrier method with the exception that gap s are left in the exterior seal located 38 mm (1 1/2 in) from each corner of the window sill. After the windows were installed into the rough openings of the CMU walls, the exterior cladding finishes were applied. All the wall sections received either a 13 mm (1/2 in) three coat stucco application or a decorative cementious coating (DCC) consisting of a stucco skim coa t applied directly over the block with a thickness that varied from a paint thickness to approximately 6 .4 mm ( 1/4 in). The stucco was allowed to cure under moist ambient conditions until the appropriate pH level was achieved to accept paint. The paint w as applied in the same three coat application as the wood walls and allowed to dry for a minimum of ten days before the wall specimens were tested. Liquid Applied Flashing Application The most distinguishable characteristic in the draft FMA/AAMA 200 (200x) standard is the application of a liquid applied flashing (LAF) on the CMU in the pre -installation procedures. The LAF is a vapor permeable waterproofing material that is applied onto the CMU wall 22 8.6 mm (9 in) around the perimeter and in the return of the rough opening to enhance the water resistance of the masonry wall system. The adhesion of a stucco application applied directly to the LAF was of some concern because the LAF dries with a smooth surface leaving a questionable interface for the mechani cal bond required to adhere the stucco to the wall. For this reason, three different application methods for the LAF were tested on the six specimens that follow the draft FMA/AAMA 200 (200x) installation (Figure 59).
84 To promote the adhesion of the stucco to the LAF, a non -emulsifiable bonding agent was applied to the treated area on two of the six CMU wall specimens. In two other specimens, a wire lath was applied over the LAF treated area before the application of stucco. The remaining two FMA/AAMA 200 (200x) walls were treated with the LAF on the interior surface of the wall surrounding the window rough opening, including the rough opening return, and down to the base of the wall. Test Specimen Matrix The test specimen matrix demarcates the wall co nstruction, window size, window operator type, window frame material, installation methodology, flashing and interior sealant option (where applicable), as well as the sill type unique to each of the 18 test specimens. In order to isolate the leakage path s of the window interface, the test matrix initially consisted of fixed windows only. It was anticipated that the operable components of the window specimens would begin to leak at 15% of their design pressure rating, which is the current requirement stip ulated in most building codes and standards (e.g., AAMA/WDMA/CSA 101/I.S.2/A440 2005), making leakage detection through the window interface difficult. Logistical issues later necessitated the need to allow the incorporation of two awning windows and a si ngle hung window into this project. However, these substitutions did not interfere with any of the testing procedures or results derived from tests on the wall assemblies. The test specimen matrix is shown in Table 5 1. Chapter Summary Eighteen 2.4 m x 2.4 m (8 ft x 8 ft ) window -wall assemblies were constructed to represent residential construction practices common to the coastal regions of the southeastern U.S. The specimens varied uniquely in wall construction, window type, window material, exterior fi nishing system, and installation methodology. Wood frame and CMU wall specimens were
85 constructed according to the guidelines of the 2001 FBC. With the exception of the paint application, all phases of the construction were completed by industry professio nals. In order to test the application of the LAF detailed in pre installation procedures of the draft FMA/AAMA 200 document three different approaches were taken. A test specimen matrix was created to simplify the multitude of variations among the speci mens.
86 Table 5 1. Test s pecimen m atrix. Specimen Wall Window Type Ext. Finish Sill Operator Type Dimensions (cm) Material Design Pressure (Pa) 017 Wood Stucco Flush Fixed 110.5 x 158.8 Aluminum 1915 017B Wood Stucco Flush Fixed 110.5 x 158.8 Aluminum 1915 017C Wood Stucco Flush Hung 110.5 x 158.8 Aluminum 2633 017D Wood Stucco Flush Fixed 110.5 x 158.8 Aluminum 1915 017E Wood Stucco Flush Fixed 110.5 x 158.8 Aluminum 1915 018 Wood FCB Flush Fixed 110.5 x 158.8 Aluminum 1915 018B Wood FCB Flush Fixed 110.5 x 158.8 Aluminum 1915 018C Wood FCB Flush Fixed 110.5 x 158.8 Aluminum 1915 018D Wood FCB Flush Fixed 110.5 x 158.8 Aluminum 1915 018E Wood FCB Flush Fixed 110.5 x 158.8 Aluminum 1915 019 CMU DCC Flush Fixed 110.5 x 158.8 Aluminum 1915 019B CMU DCC Flush Fixed 110.5 x 158.8 Aluminum 1915 019C CMU DCC Face Fixed 110.5 x 158.8 Aluminum 1915 035 CMU DCC Flush Awning 120.7 x 74.9 Vinyl 2523 020 CMU Stucco Face Fixed 110.5 x 158.8 Aluminum 1915 016 CMU Stucco Flush Awning 120.7 x 74.9 Aluminum 2523 020C CMU Stucco Flush Fixed 110.5 x 158.8 Aluminum 1915 020D CMU Stucco Flush Fixed 110.5 x 158.8 Aluminum 1915
87 Table 5 1. Continued. Specimen Installation Standard Method Flashing Interior Seal 017 ASTM E 2112 Barrier 017B FMA/AAMA 100 Drainage 101.6 mm Self adhered Backer rod and gunnable sealant 017C FMA/AAMA 100 Drainage 101.6 mm Self adhered Low expansion foam 017D FMA/AAMA 100 Drainage 228.6 mm Mechanically attached Backer rod and gunnable sealant 017E FMA/AAMA 100 Drainage 228.6 mm Mechanically attached Low expansion foam 018 ASTM E 2112 Barrier 018B FMA/AAMA 100 Drainage 101.6 mm Self adhered Backer rod and gunnable sealant 018C FMA/AAMA 100 Drainage 101.6 mm Self adhered Low expansion foam 018D FMA/AAMA 100 Drainage 228.6 mm Mechanically attached Backer rod and gunnable sealant 018E FMA/AAMA 100 Drainage 228.6 mm Mechanically attached Low expansion foam 019 ASTM E 2112 Barrier 019B Draft FMA / AAMA 200 Barrier Backer rod and gunnable sealant 019C Draft FMA / AAMA 200 Drainage Backer rod and gunnable sealant 035 Draft FMA / AAMA 200 Drainage Low expansion foam 020 ASTM E 2112 Barrier 016 Draft FMA / AAMA 200 Barrier Backer rod and gunnable sealant 020C Draft FMA / AAMA 200 Drainage Backer rod and gunnable sealant 020D Draft FMA / AAMA 200 Drainage Low expansion foam
88 A B Figure 5 1. Wall s pecimens. A) Wood frame wall specimens staged during construction. B) Concrete wall specimens staged during construction. Figure 5 2. Wood frame wall specimens. Wood walls constructed off -site and delivered to the testing facility where the windows were installed and the cladding system s applied.
89 Figure 5 3. Base of wood walls. Steel channels were attached to wood walls to facilitate specimen transport.
90 A B C D Figure 5 4. Window flashing. A) Window installed and p repared for flashing. B) Mechanically attached flashing attached over right jamb. C) Mechanically attached flashing attached over left jamb. D) Mechanically attached flashing placed over header such that the ends overlap the flashing at the jamb and t he WRB header flap may be folded down in a shingle lap fashion.
91 A B C D Figure 5 5. Wire lath application for wood frame walls. A) Wall prepared for j -stops at window frame and casing bead around wall edges. B) Expansion joints were installed at the four corners of the window. C) Wire lath is applied from the bottom of the wall to the top. D) Wall specimen ready for stucco application. Figure 5 6. pH t est.
92 A B C D Figure 5 7. Construction of CMU w alls. A) Vertical reinforcing steel we lded to the steel base channel of the wall. B) First course of CMU set on channel. C) Half constructed CMU wall. D) Completed CMU wall.
93 Figure 5 8. Prepared window rough opening in CMU wall The rough openings for the windows in the CMU walls were framed with pressure t reated wood bucking at the heads and jamb s along with a precast sill.
94 A B C Figure 5 9. Application methods for LAF A) LAF applied to exterior surface with bonding agent applied over the treated area. B) LAF applied to the exterior surface with wire lath over the treated area. C) LAF applied to the interior surface of the wall.
95 CHAPTER 6 RESULTS The 18 wall specimens were individually subjected to the four rounds of pressure and water testing while a team of o bservers documented points of water intrusion, or leaks. For the purposes of this study a leak was defined as any liquid water observed from the interior of the test specimen to have bypassed the exterior moisture barrier of the window wall assembly. When a leak was observed during testing the time into the test, pressure, location, and a brief description of the leak was recorded. The complete leakage results for the 18 wall specimens from the four rounds of testing are given in the Appendix All tests were documented with pictures as well as video footage. Data sheets for the four rounds were compared for each wall and conclusions were drawn from these leakage pressure result s No provis ions were made to determine the origins of these leakage paths in order to preserve specimens for later rounds of testing. In some instances the origins were obvious, for example through the operable portion of the window. However, in other cases the sources of water intrusions were less apparent and could have resulted due to deficiencies in the wall system, window assembly, or installation method. R egardless of the origins of the leakage paths, it is the responsibility of the installation to manage any incidental water entering the interface between the window and the surrounding wall system. While pressure loading during the tests never exceeded 50 percent of the DP rating for the window these pressures are beyond the window s designed limitations W hen reviewing the results it is instructive to realize that windows are only rated for water holdout to 15 percent of their DP rating according to AAMA/WDMA/CSA 101/I.S.2/A440 (2005) or to 12 psf according to the extreme exposure conditions of the FMA/AA MA standards. The majority of the test specim ens passed their designed limits ; however, testing was continued to the thresholds stated
96 in Chapter 3 in order to determine the limit s of the water penetration resistance of the installation methods. Installation Dependence on Moisture Management Strategy The water penetration resistance of windows in residential construction is highly dependent on the moisture management strategy employed to integrate the window into the water -shedding surface of the surrounding wall assembly. Window installation options are generally separated into two distinct approaches based on these moisture management strategies. The first method is the water barrier approach, which intends to prevent water ingress at the exter nal interface of the window and the wall system (RDH 2002a ). The second method is the drainage method approach, which controls the migration of water at the interior perimeter of the window by redirecting any leakage to the drainage plane of the wall (Kat saros and Hardman 2007, Lstiburek 2008). This study evaluated the performance of water barrier and drainage installation approaches in wood frame and CMU wall systems. Similar to the installation methods, the wall systems used in this study differ in the manner in which they prevent water intrusion. Wood frame wall systems employ a polymer -based house wrap as a concealed WRB to prevent the infiltration of water to the building interior. Conversely, CMU wall construction is considered a surface barrier sy stem where the outermost surface of the wall assembly is the sole restraint to water intrusion (ASTM E 211207). Comparison of the leakage results from the CMU and the wood frame wall specimen testing indicates that the highest water penetration resistance occurs when the installation option and the wall system share a consistent moisture management strategy. Table 6 1 summarizes the installation methods and the moisture management strategies utilized by the test specimens along with the corresponding leak age pressures. Two considerations should be taken into account
97 when reviewing the information presented in this figure. First, test specimen 017C was constructed with a single hung window; and although several leaks were observed throughout testing, this leakage occurred through the meeting rail of the window and not through the window -wall interface. This project is specifically concerned with the water penetration resistance of window installation options; therefore, only leakage occurring through the window wall interface is displayed for test specimen 017C. Second, test specimen 035 incorporated a drainage installation, but the window was installed so tightly against the bucking that the drainage channels at the sill were reduced to almost crack -size d openings (shown in Figure 6 1). The application of paint most likely sealed these drainage channels marginalizing their draining effectiveness. In Table 6 1, test specimen 035 is listed as a water barrier installation to reflect its inability to allow proper drainage through the gaps in the sealant. The drainage approach performed well in the wood frame wall systems. The only specimens that did not leak through the four rounds of pressure loading (018B, 018C, 018D, 018E) incorporated drainage method installations with fiber cement board exterior finishes. The voids in the sealant behind the integral mounting fin at the sill are coincident with the concealed barrier of the wall system and are not directly exposed to the wind-driven rain simulated enviro nment. This allows the drainage voids to properly function and redirect any leakage down the WRB of the wall. The barrier approach also performed well on the wood frame wall specimens. Only two test specimens employed a barrier installation in the wood frame wall system, 017 and 018. While both walls leaked, Table 6 1 shows that they only leaked during a single test and at relatively high pressures.
98 In the CMU wall specimens, the drainage method installations did not perform as well as the water barri er method counterparts. A total of three CMU walls (including 035) did not leak through all four rounds of pressure loading, and all incorporated barrier method installations. The exterior surface of the wall cladding serves as both the water shedding bar rier and exterior moisture barrier for the CMU wall system s In drainage installation approaches, when the wall is not subjected to positive lateral loads, the gaps left in the perimeter sealant open a pathway for gravity to drive rainwater from the windo w -wall interface to the drainage plane of the wall. Under wind loading, this migration pattern reverses if the pressure gradient between the exterior and the window -wall interface (which is dependent on the air permeability of the window -wall interface) o vercomes gravity forces and the frictional losses associated with the water passing through the orifice created by the gaps. The net result is the accumulation of rain inside of the interface until the forces balance. The barrier method; however, forms a water -resistive barrier coincident with the drainage plane at the surface of the wall. Water sheds on this barrier and cannot directly enter into the window -wall interface. The sacrifice is the redundancy offered by the interior seal of the drainage method. The critical consideration is that the water penetration resistance of the perimeter seal will likely degrade with time. Thus it is unclear if the water barrier methods would perform any better than the drainage method, when lifecyc le of the assembly is considered. Effects of Exterior Cladding on Installation Options Variations in the exterior wall finishes were evaluated to determine if the water penetration resistance of a particular window installation option depended on the exter ior cladding system of the building envelope. Two of the most common finishes used in residential construction were considered for each wall system. The wood frame walls received either a paper back lath and
99 three coat stucco application or a fiber cemen t siding. The CMU walls received either a three coat stucco application or a decorative cementious coating (DCC). Per ASTM C 92606, DCC is defined as a stucco skim coat applied directly over the block with a thickness that varies from a paint thickness to approximately 6.4 mm (1/4 in). Wood Frame Wall Specimens The results indicate that the efficacy of the window installation may be strongly influenced by the choice of exterior cladding. Table 6 2 lists the observed time and pressure of initial leakage for each of the four rounds of pressure loading along with the combinations of installation methods and exterior finishes for the wood frame walls. The fiber cement siding walls installed using the drainage method found in the FMA/AAMA 10007 standard did not leak through the four rounds of testing while the stucco walls employing the same installation method leaked on average at 598.5 Pa (12.5 psf). The ASTM E 2112 walls showed comparable leakage results between the two exterior cladding systems. It is d ifficult to make definitive statements regarding the performance of the two cladding systems in reference to the drainage installation guidelines without using destructive methods to determine the origins of the water intrusion. However, it may be possibl e that the fiber cement siding performed better with the FMA/AAMA 100 07 installation because relative to the stucco rendering, it is less intrusive to the drainage channels left in the sill of the window. In normal stucco applications over wood frame wa ll systems, wire lath attached to a felt paper backing is fastened to the wall with stub nails. The stub nails penetrate the flashing of the window and the house wrap affixing the lath to the wall. A 22.2 mm (7/8 in) thick continuous application of stucc o is then applied over top of the lath. It may be possible that the felt paper in combination with the continuous nature of the stucco application creates a closed cladding system that blocks the drainage channels located in the sill. The effectiveness of the drainage
100 installation method requires that any water finding its way into the window wall interface will drain out through the gaps in the exterior seal at the sill and migrate through the stucco to the exterior. If the water cannot drain at a rat e faster than it accumulates in the window -wall interface, it will increase the installations susceptibility to leakage. Unlike the stucco finish, the fiber cement board may be considered a vented cladding system. The laps of the fiber cement board ar e not sealed and thus the cladding system is permitted to breathe. The open drainage plane of the wall system imposes less constriction on the voids provided in the sill sealant allowing them to freely redirect water to the drainage plane of the wall. The re was no evident difference in the performance of the two exterior cladding systems for the ASTM E 2112 installation option. Although the two barrier method walls, 017 and 018, leaked during different tests, they leaked at similar pressures. The install ation procedures of ASTM E 2112 require a continuous bead of sealant around the backside of the windows integral mounting fin. When the window is installed, the moisture/air barrier is created at the interface of the mounting fin and the concealed water barrier. Leakage through the water barrier installations is dependent on the continuity of the perimeter seal and its interaction with the WRB not the choice of cladding system. Although a particular cladding system may allow more water to reach the inte rface of the window and the WRB increasing the possibility of leakage, it was not observed from the results of this testing. C oncrete M asonry U nit Wall Specimens The choice between the two exterior cladding options for the CMU walls did not seem to noticea bly affect the performance of the window installation approaches as it did in the wood walls. Table 6 3 lists the combinations of installation methods and exterior finishes for the CMU walls along with the observed time and pressure of initial leakage for each of the four rounds of
101 pressure loading. The thickness of the stucco application marginally affected the water penetration resistance. The average leakage pressure for specimens with a DCC exterior finish was 421.3 Pa (8.8 psf) while the average lea kage pressure for the specimens with a normal stucco application was 430.9 Pa (9.0 psf). An assessment of the leakage pressures for the two cladding options showed comparable results, consistent with expectations for the post cure performance. Both fini shing options have the same material properties and interact with the other components of the wall assembly in similar ways. However, the DCC is applied in a thin single coat causing it to be more susceptible to the effects of plastic shrinkage than the t raditional three -coat stucco application. When the DCC is applied to the dry surface of the CMU, moisture is drawn from the application before it reaches its full cure, which causes hairline cracks to form beneath the exterior surface. Over time these cr acks may propagate to the face of the wall system resulting in a pathway for moisture intrusion. It is unclear how longterm aging would affect comparative performance, and this outstanding issue should be prioritized for further investigation. Performance of Liquid Appli ed Flashing Application Methods The draft FMA/AAMA 200 standard current at the time of specimen construction required the application of a LAF on the CMU in the pre installation procedures. The LAF is a vapor permeable waterproofing materi al that is applied onto the CMU around the perimeter and in the rough opening of the window to enhance the water resistance of the masonry wall system. The adhesion of the stucco application applied directly onto the LAF was of some concern; therefore, th ree different application methods were tested on the six specimens that follow the draft FMA/AAMA 200 installation procedures. Since the conclusion of testing, the draft document has been accepted. The only significant difference between the accepted sta ndard and the draft do cument pertinent to this study concerns the LAF application. In the pre installation procedures
102 of the accepted standard, the LAF is only applied to the return of the rough opening. The water penetration results for the test specime ns as they pertain to the different LAF application methods evaluated in this study are discussed herein. The first application method, tested on walls 019C and 020D, utilized a bonding agent applied directly over the LAF treated area to promote adhesio n of the stucco (refer to Figure 5 9A). The bonding agent serves to provide a chemical bond for the stucco to adhere to the smooth surface of the LAF Although a particular bonding agent is not stated in the draft FMA/AAMA 200 standard, a non-emulsifiabl e bonding agent is recommended for this application. Non -emulsifiable bonding agents are more commonly used for exterior purposes because, and unlike re -emulsifiable bonding agents, they do not re -wet when exposed to moist conditions. The re -wetting of r e -emulsifiable bonding agents often causes loss of bond strength and can lead to cracking and spa l ling of the stucco. This application was expected to serve as a benchmark for the comparison of the three methods. In the second application method, used in the window installations of specimens 019B and 020C, wire lath was applied over the LAF treated area (refer to Figure 5 9B). The intent of the wire lath was to provide a means of mechanical bond for the stucco over the area where the adhesion was in quest ion. The paper backing was removed from the wire lath and it was attached with stub nails directly to the face of the CMU. Prior to testing, an examination of these two walls revealed that where the stucco surrounding the window perimeter was tapped, a f aint hollow noise could be heard. Presumably there was no direct bond of the stucco to the LAF The stucco is being supported solely by the lath which may not be held tight enough to the CMU by the s tub nails. This was carefully monitored during the pro gression of testing.
103 On the remaining two draft FMA/AAMA 200 walls, 035 and 016, the LFA was applied to the interior surface of the wall surrounding the window rough opening, including the rough opening return, and down to the base of the wall (Figure 5 9C ). Test specimens containing awning windows were used to provide ample wall area beneath the window opening for the application of the LAF In this method the LAF augments the water resistance of the wall without interfering with stucco adhesion. All six walls were subjected to the four rounds of pressure loading. While some of the walls leaked and some did not, without the use of more invasive leakage detection techniques it was difficult to make a decisive conclusion on the optimal application method f or the LAF in terms of its water resistance. Although some of the walls specified above exhibited leakage at the window opening, without the ability to pinpoint the origin of the leaks it cannot justly be attributed to a failure in the application method of the LAF Conversely, the fact that some of the wall specimens did not leak through the four rounds of pressure loading should not be used to validate a particular application. Moreover, the water penetration resistance of these application methods can not be fully tested without aging the specimens to the extent that cracks develop in the stucco renderings of the walls. For these reason s, conclusions on the application methods for the LAF will be held to comments concerning the installation procedures and effectiveness of construction. Further investigation, particular through destructive testing, should be made into the leakage performance of these applications. While all of the walls su ccessfully accepted the application of the stucco, the use of the non -emulsifiable bonding agent was considered to be the preferred method due to the facility of the application and the location of the treated wall area. When comparing application times w ith the use of a bonding agent, the installation of the wire lath in the second application method
104 prove d to be time -consuming. Also, because the wire lath is applied with stub nails, unnecessary penetrations are created in the LAF that may be avoided by using the non-emulsifiable bonding agent. In the third application method, the functionality of the LAF is questionable because of its location. As discussed previously, CMU walls are a surface barrier system with the drainage plane of the wall located a t the exterior surface. In order for the LAF to be most effective in enhancing the water resistance of the wall, it should be applied as close to the drainage plane as possible. The shortcoming of the third application method is that the LAF treated area is located at the interior surface of the wall assembly. The LAF will prevent any water penetrating the drainage plane on the interior side of the wall and as a result, trap moisture inside of the wall cavity From these observations, it may be conclude d that that use of a bonding agent would serve as the most consistent and reliable application method. Ease of Installation and Repeatability When comparing the current ASTM E 2112 and the proposed FMA/AAMA installation methods, the complexity of the installation procedures should be considered. The effective water penetration resistance of fenestration products in residential construction is contingent on the proper implementa tion of the installation procedures. As the procedures of an installation method become more intricate the likelihood that they will be executed successfully in the field decreases. For water penetration resistance strategies without redundant measures ( e.g. barrier methods), this is a critical issue. A single installation or fabrication defect can lead to water ingress. The drainage installation approaches outlined in the FMA/AAMA standards rely heavily on the proper coordination of the trades involved in the installation of the window. Even under the careful supervision provided for the construction of the test specimens in this study several mistakes that may prove commonplace in the field were revealed. During the installation of the
105 windows careful supervision was required to ensure that drainage channels, of the proper size and location, were left in the sealant behind the sill of the window. Once the windows were installed these drainage channels had to be protected from obstruction during the ap plication of the exterior cladding system. The drainage channels of the CMU walls must be left open to the exterior surface; however, because these voids were located on the backside of the mounting flange of the window it was difficult to pinpoint their location once the window was installed against the bucking. During the application of stucco, these channels were nearly filled because there was no way to tell that a particular wall contained voids in the sealant. For similar reasons these drainage cha nnels were nearly sealed during the preparation for paint when the exterior faces of the windows were being sealed. It should be made clear that these observations are not presented to deter the use of the drainage approaches described in the FMA/AAMA stan dards. The effective water penetration resistances of these methods were discussed previously. Rather, these observations are meant to bring attention to possible areas in which improper execution could hinder the efficacy of the installation method. It is important to note that while simplification and repeatability play an important role in guaranteeing the performance of an installation method, the moisture management strategy of a window should not be chosen based on the level of craftsmanship possessed by the installer. Performance of FMA/AAMA Installation Variants The FMA/AAMA 100 07 and the draft FMA/AAMA 200 standards contain options for the use of either a 101.6 mm ( 4 in ) self adhered (Figure 6 2A) or a 228.6 mm ( 9 in ) mechanically attached (Figu re 6 2B) flashing to integrate the window into the drainage plane of the wall, as well as either a low expansion foam or a backer rod and sealant to create the interior moisture/air barrier. Different combinations of these installation options were paired with the varying wall
106 types and exterior finishes according to the details of the test specimen matrix. Leakage results were compared to determine the relative performance of these installation alternatives and their interaction with the components of the surrounding wall assembly. Flashing Syste m The results suggest that the 101.6 mm (4 in ) self adhered and the 228.6 mm ( 9 in ) mechanically attached flashing are comparable alternatives. Table 6 4 shows the initial leakage times and pressures for the test specimens following the installation guidelines of the FMA/AAMA 100 07 standard sorted by the chosen flashing system. Each flashing system passed the same number of pressure loading sequences without water infiltration. In tests where the walls did leak the pressures and times of the initial leakage are comparable. Specimens 017B and 017D exhibited nearly identical performance through all four rounds of pressure loading, leaking at similar times and pressures during each test. The average initial leak age pressure of the 101.6 mm (4 in) self adhered flashing is 646.4 Pa (13.5 psf). The average initial leakage pressure of the 228.6 mm (9 in) mechanically attached flashing is 550.6 Pa (11.5 psf). While these values are similar, the 228.6 mm (9 in) mecha nically attached flashing leaked at the lowest pressure of any test represented in Table 6 4 (during the Cyclic Test). Interior Moisture/Air Barrier The leakage results for the specimens employing the installation methods of the FMA/AAMA standards indicat e the backer rod and the gunnable sealant as the superior option for the interior moisture/air seal. The initial leakage times and pressures observed for these walls sorted by choice of interior seal are displayed below in Table 6 5. There were a total o f four walls using backer rod and gunnable sealant as the choice of interior seal that completed all four rounds of pressure loading without leaking, as opposed to only three for the low expansion foam walls. The average initial leakage pressure for the b acker rod and gunnable sealant installations
107 was 555.4 Pa (11.6 psf). The average initial leakage pressure for the low expansion foam installations was 392.6 Pa (8.2 psf). Therefore, on average test specimens with interior seals formed using the low expa nsion foam leaked at pressures 162.8 Pa (3.4 psf) lower than those employing the backer rod and gunnable sealant. Although this pressure difference may seem rather insignificant, the dynamic pressure developed from impinging winds is proportional to the v elocity squared and this pressure conservatively represents a difference in wind speed of almost 40.2 km/h (25 mph). If the CMU specimens demonstrating leakage are isolated for comparison from Table 6 5 it would appear that contrary to the previous conclus ion, the low expansion foam outperformed the backer rod and gunnable sealant. Over the four rounds of testing, specimens 019C and 020C using backer rod and gunnable sealant leaked at lower pressures than specimen 020D using the low expansion foam. A clos er investigation into the location and type of leakage; however, shows that these low pressure leaks were not a result of the interior seals performance. Specimen 019C leaked through the glazing and specimen 020C leaked through a skip in the interior sea l caused by improper tooling of the joint (i.e. improper installation). When the location and type of leakage is considered in evaluating the performance of the interior moisture/air barriers of these three walls, analogous leakage results are observed. This demonstrates that either option employed for the interior seal preformed adequately on the CMU walls in this study when properly installed. From observations made during testing it appears that the low expansion foam is more apt to allow water ingress because it does not provide the continuity and adhesion achieved by the backer rod and gunnable sealant. Frequently leaks were observed at the interface between the foam seal and the flexible sill flashing membrane used on the sills of the wood frame wal ls. The
108 flexibility in the flashing product is provided by small crepes in the surface that spread as it is stretched. When the low expansion foam is applied the chemical expansion of the product is left to fill any voids between the rough opening and th e window providing a continuous interior seal. If an insufficient amount of sealant is applied during installation gaps will remain after the foam expands and these areas will be subject to leakage. The majority of the low expansion foam walls that leak ed during testing did so because the expanding foam failed to seal the grooves in the crepe flashing. Water that accumulated within the window opening eventually began to seep past the low expansion foam through these grooves. This was not such a problem with the backer rod and gunnable sealant alternative because in tooling the sealant the installer is able to apply adequate pressure to force the sealant into the striations of the flashing. Water Penetration Resistance of Sealant s used for Interior Moist ure/Air Seals The varying water penetration performance of the interior seal options observed in this study raised concerns with the current requirements for sealants used in the installation of fenestration products. Currently, window installation stand ards (ASTM E 211207, FMA/AAMA 100 07, FMA/AAMA 200 08) simply require gunnable sealants to meet the specifications of ASTM C 920 or AAMA 808 for elastomeric sealants and AAMA 812 for low expansion foams. The adhesion to common substrates used in the inst allation of fenestration products is not detailed in these specifications, nor is any test that determines how well a particular sealant may prevent water penetration at increased pressures. Moreover, the proper selection of a sealant in these standards is left to the interpretation of vague guidelines. For example in ATSM E 211207 the appropriate sealant is to be selected from Tables A4.1 and A4.2 which simply provide a listing of applicable specifications and general adhesion information to common bui lding products. The water intrusion performance of an installation
109 rests largely on the quality of sealant used; therefore, more stringent requirements and definitive guidance should be developed. Observed leakage through the interior moisture/air seal during the original four tests indicated that a correlation may exist between the adhesion of a sealant and its water penetration performance. Testing was performed to determine the hydraulic leakage pressure of typical elastomeric sealants and low expans ion foams. Selected sealants and there compositions are listed in Table 4 1. The elastomeric sealants were further tested to evaluate their adhesion strengths to the materials used in the window installations of this study. The results were analyzed and compared for apparent trends in the water penetration resistance of the sealants. Hydraulic Leakage Pressure A total of 72 sill specimens, were tested according to the procedures described in Chapter 4 for the Hydraulic Leakage Pressure Test. The interio r seals in 42 of the specimens were created using backer rod and a gunnable sealant while the remaining 30 specimens employed low expansion foam for the interior seals. Each sealant was tested three times with both the vinyl and aluminum sill materials. The pressures of the first and second leaks were documented for each test. The leakage values for the three tests of a particular sill material were then averaged for each sealant. The average leakage pressure results are displayed in Figure 6 3 for the gunnable sealants and Figure 6 4 for the low expansion foams. G ood agreement between the first and second leakage pressures suggests consistency in the application of the sealant. Multiple leaks initiating at various locations along the sill specimen within a close pressure range was considered to be an indication of the sealants diminishing ability to prevent water ingress under the increasing hydraulic pressure. If the initial leak in the sills would have occurred at pressures far different from subsequent leaks then the cause of the initial water penetration would most likely have been the fault of a lapse in the sealant due poor
110 application and tooling. The average difference between the pressures of the first and second leaks for all specimens is less than 10 percent which proposes consistent sealant application for the test ing. Since the pressures for the first and second leaks are so similar and the initial leak is the firs t sign of water bypassing the exterior moisture barrier of the installation, comparisons of the hydraulic leakage pressure results are based on the ini tial leakage values only. Resultant leakage data was further reduced for each sealant by determining the maximum, minimum, and average initial leakage pressure for each sill material. These pressures are displayed in Figure 6 5 for the gunnable sealants a nd Figure 6 6 for the low expansion foam sealants. The average pressure values allow for easy comparison among specimens, while the maximum and minimum leakage pressures give a sense of the variability associated with the sealants performance. Data prese nted in these figures serve as the foundation for the results discussed herein. Water p enetration performance of g unnable sealants vs. low expansion foams In c omparing the performance of the two sealant types based on Figures 6 6 and 6 7 it is implied that the low expansion foam is the superior alternative for the interior moisture/air seal in the installation of fenestration products. The low expansion foam products showed a greater degree of variability than the gunnable sealants; however, the leakage pr essures were much greater. Many of the low expansion foam seals reached the maximum attainable pressure for the testing apparatus of 4788 Pa (100 psf) without leaking. None of the gunnable sealants ever reached this maximum test pressure. The weakest lo w expansion foam performer was actually on par with the second best gunnable sealant performer. However, the data for the water penetration resistance of the low expansion foams are suspect due to the application used during testing.
111 The drainage method i nstallations outlined in the FMA/AAMA 100 and the draft FMA/AAMA 200 document allow the use of a low expansion foam to create the interior moisture/air seal. These installation methods expect that liquid water will enter the window -wall interface and be r edirected by the interior seal to the building exterior through the drainage channels left in the exterior seal. However, if a low expansion foam is installed according to the product manufacturers application instructions the foam seal works against the theory behind drainage installations. Low expansion foams are intended to seal and fill gaps between the rough opening and the window. To apply, foams are to be injected into the deepest portion of the gap filling anywhere from 1/3 to 1/2 of its volume (based on the expansion properties of the foam). The low expansion properties of the foam then allow it to expand out of the opening filling any voids along the way as it cures. Foam applied in this manner would prevent any expected water from entering t he window -wall interface and may even prevent the proper functioning of the drainage channels at the sill. During the original window -wall specimen testing it was noticed that in some instances a water column developed in the jambs of the windows. This wa s made apparent in windows using translucent sealants for interior moisture/air barriers as well as in windows with persistent leaks from markedly submerged fasteners through the jamb. In order to replicate the observed conditions sealants were applied i n a manner that maintained a cavity in the sill specimens which was to be filled with water in the development of a hydraulic head. For low expansion foams this meant that the sealants were extruded only partly into the seal gap to ensure that a sill cavi ty remained after the foam expanded. An application of this manner negates the performance criteria established for the product because it contradicts the manufactures instructions.
112 It should also be noted that the foam was not trimmed prior to testing. After the foam is applied in construction practice the excess is carefully cut away so the window may be finished with trim pieces. Excess foam was not removed from the low expansion foam seals of the original wall specimens because the specimens did not warrant a finished interior. Low expansion foam seals for the hydraulic leakage pressure testing were not trimmed until the conclusion of the t est to maintain consistency. While the low expansion foam seals demonstrated excellent water penetration resist ance at high pressures, when the seals were cut immediately following a test almost all of the products began to leak (Figure 6 7). Liquid water was permeating the core of the foam seal but was not breaking the plane of the seal due to the closed cell nat ure of the foam surface. The extent of the deteriorating effects this water will have on the seals performance over time in unknown. Although the gunnable sealants leaked at lower pressures than the low expansion foam seals their application was consiste nt with the manufacturers instructions. Due to accumulation of water observed within the foam sealants as well as the inherent contradiction between the principles behind the application of the low expansion foam and the theories behind drainage installa tion methodologies, the use of low expansion foams in drainage method installations may be ill advised. Interior seals formed using gunnable sealants were further analyzed for their water penetration performance. Effect of binder ty pe on interior seal pe rformance The two most common sealants used in the installation of fenestration products are acrylic latex and polyurethane. Each sealant composition has it s positive and negative attributes. Acrylic latex sealants are extremely versatile. They demonstr ate good adhesion to many different materials, they are paintable, they are low in toxins, and they offer easy cleanup with water. Perhaps the two biggest shortcomings of these types of sealants in regards to window
113 installations are their lack of durability and susceptibility to shrinkage. Durability is somewhat compensated for with the addition of silicone polymers to create siliconized acrylic lat ex sealants; however, because acrylic latex sealants are water -based shrinkage remains detrimental to the water penetration resistance of installed windows. Siliconized acrylic latex sealants c ure as the water in the sealant evaporates into the air. As the water evaporates the sealant shrinks and what remains is dependent on the sealants percentage of solid s by weight. Polyurethanes, on the other hand, are known for their durability and movement capability. Polyurethanes are moisture curing sealants and cure by drawing moisture from the air to serve as a catalyst for the chemical reaction between the compo nents in the sealant making them less likely to shrink. They demonstrate good adhesion to metals and porous substrates such as mortar or cement and, like acrylic latex, may also be painted. The disadvantages associated with polyurethanes is that they are hard to tool, they require a petroleum based solvent for clean up and oftentimes they may be more expensive than their acrylic latex counterparts Both polyurethanes and siliconized acrylic latex sealants were selected for testing. Table 4 1 lists the sealants and their composition used for this testing. In the original wall specimen testing windows were installed using both polyurethane and siliconized acrylic latex sealants; however, due to the number of variables included into the Test Specimen Matri x in Table 5 1 distinctions between the water penetration resistances of these sealant types were difficult to observe. For the Hydraulic Leakage Test, three latex acrylic sealants (S1 -S3) and four polyurethane sealants (S4 S7) were selected for testing. The leakage results from this testing (Figure 65) illustrate that the ability of polyurethane sealants to resist water penetration at increased pressures exceeds that of siliconized acrylic latex sealants. The
114 average leakage pressures for all polyureth ane sealants are higher than those of the siliconized acrylic latex sealants. The performance of the polyurethane sealants in this testing is most like ly the result of their inherent curing mechanism. Since, polyurethanes are chemical curing sealants they are not susceptible to shrinkage. In tooling the sealant during application it was forced into the striations of the flexible flashing. The low shrinkage properties associated with the curing process of the sealant ensured that as the sealant cured it d id not shrink and pull out of the gro o ves in the flashing. The siliconized acrylic latex sealants on the other hand, most likely shrank out of these grooves as the water in the sealant evaporated during the curing process. Adhesion Strength Except for two occasions, every leak recorded for the sill specimens of the Hydraulic Leakage Test occurred at the interface between the sealant and the sill flashing membrane Consistent with the water intrusion observed through the window -wall specimens these leakage results underscore the possible correlation between the water penetration resistance of the sealants and their adhesion strength to the surround ing materials. One of the tests for elastomeric joints sealants required by ASTM C 920 is ASTM C 794 Standard Test Method for Adhesion-in Peel of Elastomeric Joint Sealants In this standard sealant specimens are prepared on material substrates and then peeled while the required force is recorded. The only materials substrates specified in this st andard are anodized aluminum, mortar, and plate glass; none of which are commonly adhered to in the installation of fenestration products. With the adhesion strength to the flexible sill flashing membrane of particular concern the peel strengths of the gu nnable sealants to the material substrates specified in the standard along with other materials commonly sealed to in window installations (wood, vinyl, painted aluminum, and the sill flashing membrane used in the testing) was determined.
115 Two sealant spe cimens were peeled per substrate. The average and peak peel force was recorded as well as a description of the failure mode for each specimen. Average peel strengths for the sealants to the seven material substrates are displayed in Figure 6 8 and Figure 6 9. Figure 6 8 displays the peel strength values recorded from the first test run and Figure 6 9 displays the peel strength values recorded from second test run. The two test values do not constitute enough data for a significant average; therefore, t h e values for the two tests were compared to determine repeatability in the test values and to expose possible errors that may have occurred when the values were transposed. The values for the two tests are fairly consistent ; hence further comparisons are based solely on the peel strength values from the first test run. The adhesion in peel values for the sealants exhibit a severe lack of adhesion to the sill flashing membrane used in this study. The lack of adhesion to the sill flashing membrane is a res ult of the low surface energy associated with the h igh density polyethylene ( HDPE ) laminate used to create the film of the flashing. Since the adhesionin -peel values of the sealants to the sill flashing membrane are so much less than adhesionin peel val ues to the other substrates th e values are isolated from Figure 6 8 and reproduced according to sealant index in Figure 6 10 for comparison purposes. The main concern of the adhesion testing was to determine whether the leakage observed at the interface o f the sealant and the flashing could be attributed to poor adhesion strength. Figure 6 11 overlays Figure 6 5 on Figure 6 10 to compare the water penetration resistance of the gunnable sealants with their corresponding adhesion strength to the sill flashi ng membrane It would appear that no definitive correlation exists between the two performance criterions The sealant with the largest peel force was also one of the worst performing sealant s in regards to
116 water penetration. Although its e ffects cannot be directly attributed to the leakage observed in this testing, good adhesion to the materials being sealed remains an important factor in the performance of a sealant. Chapter Summary Observed leakage for the eighteen window -wall specimens were r ecord ed for the four rounds of pressure and water testing. For the purposes of this study a leak was defined as any liquid water observed from the interior of the test specimen to have bypassed the exterior moisture barrier of the window -wall assembly. L eakage results were used to evaluate how the variables included into the Test Specimen Matrix (Figure 5 1) affected the water penetration resistance of the wall specimens. An increased water penetration resistance was observed in test specimens with windo ws installed using a moisture management strategy consistent with that of the wall system. Water barrier method window installations performed better than drainage method installations in the CMU wall system, and drainage method window installations perfo rmed better than the water barrier method installations in the wood frame walls. Furthermore, the choice of exterior cladding was found to have a significant impact on the effectiveness of the drainage method installations in the wood frame wall systems. The wood frame walls with windows installed according to FMA/AAMA 10007 and FC B exterior finishes were the only wood frame wall that showed no visible signs of leakage. Significant leakage was observed bypassing the interior moisture air/barrier of drainage method installations at the interface of the sealant and the sill flashing membrane. While the majority of the leakage observed occurred in walls using low expansion foam sealant s for the interior moisture/air barrier, the varied performance of b oth interior seal options sp urr ed follow up testing to determine if a correlation exist ed between the ability of a sealant to resist water penetration and its adhesion strength.
117 Table 6 1. Leakage r esults for v arying w indow w all moisture m anagement c ombin ations Specimen Moisture Management Strategy Wall Ext. Initial Leakage Time / Pressure (Pa) Wall System Installation Finish Static Test Cyclic Test Dynamic Test Static Test 2 017 Concealed Barrier Stucco Stucco DNL DNL 4:25 / 459.7 1024.6 DNL 018 Concealed Barrier Barrier Barrier 18:17 / 866.6 DNL DNL DNL 019 Surface Barrier Barrier Barrier 15:00 / 957.6 11:25 / 191.5 574.6 3:15 / 565.0 890.6 8:55 / 354.3 019B Surface Barrier Barrier Barrier DNL DNL DNL DNL 020 Surface Barrier Barrier Barrier DNL 18:00 / 335.2 1005.5 DNL DNL 016 Surface Barrier Barrier Barrier DNL DNL DNL DNL 035 a Surface Barrier Barrier Barrier DNL DNL DNL DNL 017B Concealed Barrier Drainage Drainage 17:45 / 842.7 16:30 / 287.3 61.8 DNL 20:00 / 957.6 017C b Concealed Barrier Drainage Drainage DNL DNL DNL DNL 017D Concealed Barrier Drainage Drainage 19:30 / 943.2 16:00 / 287.3 61.8 DNL 17:27 / 818.8 017E Concealed Barrier Drainage Drainage 6:20 / 215.5 1:46 / 67.0 205.9 2:53 / 306.4 679.9 3:18 / 138.9 018B Concealed Barrier Drainage Drainage DNL DNL DNL DNL 018C Concealed Barrier Drainage Drainage DNL DNL DNL DNL 018D Concealed Barrier Drainage Drainage DNL DNL DNL DNL 018E Concealed Barrier Drainage Drainage DNL DNL DNL DNL 019C Surface Barrier Drainage Drainage 7:20 / 272.9 6:20 / 143.6 430.9 0:40 / 181.9 296.9 6:00 / 191.5 020C Surface Barrier Drainage Drainage 6:38 / 239.4 2:30 / 67.0 205.9 2:47 / 306.4 679.9 4:15 / 138.9 020D Surface Barrier Drainage Drainage 10:45 / 502.7 6:30 / 143.6 430.9 1:45 / 373.5 593.7 11.09 / 521.9 The abbreviation DNL designates walls that did not leak during a particular test. a Specimen 035 was changed from a drainage method to a water barrier method type installation due to constricted drainage channels. b The leakage times and pressures shown for specimen 017C are for leaks through the window -wall interface only.
118 Table 6 2. Comparison of l eakage r esults for w ood f rame w all e xterior f inishes. Specimen Wall Ext. Finish Installation Initial Leakage Time / Pressure (Pa) Method Flashing Static Test Cyclic Test Dynamic Test Static Test 2 017 Stucco Barrier Mechanically attached DNL DNL 4:25 / 459.7 1024.6 DNL 017B Stucco Drainage Self adhered 17:45 / 842.7 16:30 / 287.3 61.8 DNL 20:00 / 957.6 017C Stucco Drainage Self adhered DNL DNL DNL DNL 017D Stucco Drainage Mechanically attached 19:30 / 943.2 16:00 / 287.3 61.8 DNL 17:27 / 818.8 017E Stucco Drainage Mechanically attached 6:20 / 215.5 1:46 / 67.0 205.9 2:53 / 306.4 679.9 3:18 / 138.9 018 FCB Barrier Mechanically attached 18:17 / 866.6 DNL DNL DNL 018B FCB Drainage Self adhered DNL DNL DNL DNL 018C FCB Drainage Self adhered DNL DNL DNL DNL 018D FCB Drainage Mechanically attached DNL DNL DNL DNL 018E FCB Drainage Mechanically attached DNL DNL DNL DNL Table 6 3. Comparison of l eakage r esults for CMU w all e xterior f inishes. Specimen Wall Ext. Finish Installation Method Initial Leakage Time / Pressure (Pa) Static Test Cyclic Test Dynamic Test Static Test 2 019 DCC Barrier DNL DNL 4:25 / 459.7 1024.6 DNL 019B DCC Barrier 17:45 / 842.7 16:30 / 287.3 61.8 DNL 20:00 / 957.6 019C DCC Drainage DNL DNL DNL DNL 035 DCC Drainage 19:30 / 943.2 16:00 / 287.3 61.8 DNL 17:27 / 818.8 020 Stucco Barrier 6:20 / 215.5 1:46 / 67.0 205.9 2:53 / 306.4 679.9 3:18 / 138.9 016 Stucco Barrier 18:17 / 866.6 DNL DNL DNL 020C Stucco Drainage DNL DNL DNL DNL 020D Stucco Drainage DNL DNL DNL DNL
119 Table 6 4. Leakage r esults for the f lashing o ptions of the FMA/AAMA 100 07 s tandard. Specimen Flashing Wall Ext. Initial Leakage Time / Pressure (Pa) Finish Static Test Cyclic Test Dynamic Test Static Test 2 017B Self adhered Stucco 17:45 / 842.7 16:30 / 287.3 61.8 DNL 20:00 / 957.6 017C Self adhered Stucco DNL DNL DNL DNL 018B Self adhered FCB DNL DNL DNL DNL 018C Self adhered FCB DNL DNL DNL DNL 018D Mechanically attached FCB DNL DNL DNL DNL 018E Mechanically attached FCB DNL DNL DNL DNL 017D Mechanically attached Stucco 19:30 / 943.2 16:00 / 287.3 61.8 DNL 17:27 / 818.8 017E Mechanically attached Stucco 6:20 / 215.5 1:46 / 67.0 205.9 2:53 / 306.4 679.9 3:18 / 138.9 Table 6 5. Leakage r esults for the i nterior s ealant o ptions of the FMA/AAMA s tandards. Specimen Interior Seal Wall Initial Leakage Time / Pressure (Pa) Type Ext. Finish Static Test Cyclic Test Dynamic Test Static Test 2 017B Backer rod/sealant Wood Stucco 17:45 / 842.7 16:30 / 287.3 61.8 DNL 20:00 / 957.6 017D Backer rod/sealant Wood Stucco 19:30 / 943.2 16:00 / 287.3 61.8 DNL 17:27 / 818.8 018B Backer rod/sealant Wood FCB DNL DNL DNL DNL 018D Backer rod/sealant Wood FCB DNL DNL DNL DNL 016 Backer rod/sealant CMU Stucco DNL DNL DNL DNL 019B Backer rod/sealant CMU DCC DNL DNL DNL DNL 019C Backer rod/sealant CMU DCC 7:20 / 272.9 6:20 / 143.6 430.9 0:40 / 181.9 296.9 6:00 / 191.5 020C Backer rod/sealant CMU Stucco 6:38 / 239.4 2:30 / 67.0 205.9 2:47 / 306.4 679.9 4:15 / 138.9 017C Low expansion foam Wood Stucco DNL DNL DNL DNL 017E Low expansion foam Wood Stucco 6:20 / 215.5 1:46 / 67.0 205.9 2:53 / 306.4 679.9 3:18 / 138.9 018C Low expansion foam Wood FCB DNL DNL DNL DNL 018E Low expansion foam Wood FCB DNL DNL DNL DNL 020D Low expansion foam CMU Stucco 10:45 / 502.7 6:30 / 143.6 430.9 1:45 / 373.5 593.7 11.09 / 521.9 035 Low expansion foam CMU DCC DNL DNL DNL DNL
120 Figure 6 1. Drainage c hannel c onstriction. The window of test specimen 035 was installed so tightly against the bucking that the drainage channels were constricted. A B Figure 6 2. Flashing o ptions. A) Installation of 228.6 mm ( 9 in ) mechanically attached flashing with staples through WRB. B) Installation of 101.6 mm ( 4 in ) self adhered flashing.
121 Figure 6 3. Gunnable sealant leakage comparison. Figure 6 4. Low expansion foam leakage comparison. The dashed line in the figure represents the maximum achievable pressure for the test, equivalent to 4788 Pa (100 psf). 0 200 400 600 800 1000 1200 1400 1600 1800 S1 S2 S3 S4 S5 S6 S7Average Pressure (Pa)Specimen Index Vinyl Leak 1 Vinyl Leak 2 Aluminum Leak 1 Aluminum Leak 2 0 1000 2000 3000 4000 5000 6000 F1 F2 F3 F4 F5Average Pressure (Pa)Specimen Index Vinyl Leak 1 Vinyl Leak 2 Aluminum Leak 1 Aluminum Leak 2
122 Figure 6 5. Gunnable s ealant w ater p enetration r esistance. Figure 6 6. Low e xpansion f oam w ater p enetration r esistance. The dashed line in the figure represents the maximum achievable pressure for the test, equivalent to 4788 Pa (100 psf).
123 A B C Figure 6 7. Water i ntrusion through l ow e xpansion foam. A) Water expelling from foam seal core when trimmed immediately following hydraulic leakage pressure test. B) Water accumulated inside foam seal core during testing. C) Leakage through trimmed low expansion foam seal immediately following testing.
124 Figure 6 8. Test 1 a dhesionin -p eel v alues. 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 200.0 Unfinished Aluminum Brick Plate Glass Painted Rolled Aluminum Vinyl Sill Flashing Membrane WoodAverage Peel Strength (Newtons)Material Substrate S1 S2 S3 S4 S5 S6 S7
125 Figure 6 9. Test 2 a dhesionin -p eel v alues. 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0 200.0 Unfinished Aluminum Brick Plate Glass Painted Rolled Aluminum Vinyl Sill Flashing Membrane WoodAverage Peel Strength (Newtons)Material Substrate S1 S2 S3 S4 S5 S6 S7
126 Figure 6 10. Adhesion-in -p eel v alues to sill flashing membrane 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 S1 S2 S3 S4 S5 S6 S7Average Peel Strength (Newtons)Sealant Index
127 Figure 6 11. Comparison of w ater p enetration r esistance and a dhesionin -p eel v alues for g unnable s ealants.
128 CHAPTER 7 CONCLUSIONS This study tested the wind-driven rain penetration resistance of selected installation methodologies for windows in their in -service condition. Eighteen full -scale wall specimens were constructed varying uniquely in their wall construction, window operato r type, window dimensions, window material, and exterior finish. The wall specimens were then subjected to four rounds of pressure testing and winddriven rain simulations specifically designed for this study. The leakage results were recorded and used t o draw conclusions on the water penetration resistance of the installation method and the window -wall system. Significant conclusions are discussed herein. Water Barrier Method vs. Drainage Method Installations The main difference between the water barrie r method and drainage method installations is the location at which the exterior moisture barrier is located. In water barrier installations the exterior moisture barrier of the window and the wall assembly are in the same plane. All water is prevented from permeating the building interior at either the concealed barrier of the wood frame walls or the exterior surface of the CMU walls. No provisions are made for the management of incidental watering entering the window opening from the surrounding wall assembly. In drainage method installations, the exterior moisture barrier of the window is achieved by the application of a continuous seal around the interior perimeter of the window. The interior seal functions as a barrier to any liquid water trying t o enter the building interior redirecting it to the drainage plane of the wall through voids left in the exterior seal at the sill. The results of this testing show that the best water penetration performance of the window wall assembly is achieved when th e fenestration unit is integrated into the surrounding wall system in a manner that maintains the continuity of the four critical barriers required for a
129 successful moisture management strategy. The four critical barriers, as described in Chapter 2, inclu de the water shedding barrier, exterior moisture barrier, vapor barrier, and air barrier. In order for the wall system to properly perform the key functions of the building envelope these barriers must maintain consistency with those of the installed wind ow assembly. The drainage method installations did not perform well on the CMU wall specimens tested because an inconsistency exists in both the water shedding barrier and the exterior moisture barrier of the window -wall system. The effectiveness of the drainage method installations relies on the ability to redirect infiltrating water to the drainage plane of the wall. To accomplish this, drainage channels are provided in the exterior seal of the window at the sill as close to the drainage plane of the wall as possible. In the case of the CMU walls, the drainage plane of the wall system is located at the exterior surface of the assembly; therefore, drainage channels are located at the surface of the wall creating a breach in the water shedding barrier. These drainage channels, which function to drain water to the building exterior in the absence of wind loading, allow increased amounts of water to accumulate inside of the window frame when positive pressures are applied. The exterior moisture barrier of the window -wall assembly consists of the interior seal and the surface of the wall system; however, since the drainage channels are located at the exterior surface a proper termination of the stucco returning to the window is difficult to achieve. The stucco return in these areas cannot be sealed like the rest of the window perimeter because an unimpeded drainage channel must be maintained. The result is an inconsistency in the exterior moisture barrier of the window -wall assembly. Liquid water is ab le to infiltrate the stucco rendering at these points and if the rate of wetting exceeds the rate of drying for the stucco, accumulation w ill occur and possibly lead to water intrusion.
130 Water barrier installation methods simply employ a continuous exteri or seal around the window perimeter. These installation methods preformed well on the CMU wall specimens because the continuous seal preserves the continuity of both the water shedding ba rrier and the exterior moisture barrier a t the surface of the wall a ssembly. Although the redundancy of the interior and exterior seal is lost the critical barriers are maintained. For windows installed into the wood frame walls both water barrier and drainage installation methods provided sufficient water penetration res istance. This is due to the fact that both methods achieve adequate consistency of the critical barriers. The water shedding barrier of the wood frame walls is created by the surface of the exterior cladding. In drainage method installations because the drainage channels are located on the backside of the mounting flange of the window coincident with the concealed WRB of the wall the consistency of the water shedding barrier is preserved. The continuity of the exterior moisture barrier is achieved thr ough the proper shingle lap flashing of the sill pan and the WRB of the wall. In water barrier installations for wood frame walls the critical barriers are simply maintained by the use of the exterior perimeter seal to integrate the window into the concea led WRB of the wall. Interior Moisture/Air Barrier The effective water penetration resistance of drainage method installations relies heavily on the ability of the interior moisture/air barrier to redirect water to the drainage plane of the wall. In the drainage installation procedures of the FMA/AAMA standards the interior seal may either be created using a backer rod and gunnable sealant or a low expansion foam. Both options were tested in the original four rounds of pressure and water testing. The var ying water penetration performance of these interior seal options through the original four pressure loading sequences prompted additional testing. During static and cyclic pressure tests on the wall specimens utilizing drainage installations, it was repe atedly observed
131 that a water column developed inside the jamb of window frame. At increased pressures this liquid water began to bypass the interior seal at the interface between the sealant and the sill flashing membrane Leakage of this type was attrib uted to a lack of adhesion between the sealant and the low surface energy HDPE laminate of the sill flashing membrane In order to test this supposition a selection of seven gunnable sealants and five low expansion foams were tested to determine their hyd raulic leakage pressures according to the procedures described in Chapter 4. Furthermore, the gunnable sealants were tested to determine the strength of adhesionin peel according to ASTM C 794. The peel strengths of the gunnable sealants were then compa red with their hydraulic leakage pressures to determine if a significant correlation existed. Sealant Characteristics for U se in the Installation of Fenestration While the results disproved a definitive correlation between the leakage results and the adhes ion alone, the adhesion of a sealant is thought to be one of the key factors determining its performance. Three characteristics specific to the sealants applied to the creped flashing material of this study were reasoned to contribute to the water penetra tion resistance of the interior moisture/air seal. These characteristics include the viscosity of the sealant, the shrinkage incurred during curing, and the adhesion of the sealant to the surrounding substrates. Viscosity is a measure of the resistance of a fluid to flow. A greater force is required to initiate flow in highly viscous fluids than less viscous fluids. The viscosity of a gunnable sealant is an indication of its workability, the less viscous a sealant the more likely that it will fill all of the grooves in the flexible sill flashing membrane material when tooled by the installer. High viscosity sealants are less likely to fill these grooves when tooled with the same force. The shrinkage incurred during curing is also critical to the water pe netration performance of the sealants used in this study. While the viscosity of the sealant suggests how well it will fill the striations of the creped flashing when tooled, the shrinkage properties of the sealant
132 determine how well these striations remain filled after curing. If a low viscosity sealant with high shrinkage is used for the interior seal the grooves may be completely filled after joint is tooled, but the sealant will gradually pull out of these groves as it shrink during the curing process Finally, the adhesion of a sealant is important to ensure the durability of the sealant. Without appropriate adhesion to the joint materials the previous two sealant characteristics are negated, never getting the chance to occur. In order to ensure the durability of a sealant during its service life proper adhesion is required. Considering these three factors the ideal sealant for this study would be one that displayed low viscosity, low shrinkage, and high adhesion. This would ensure that the grooves in the sill flashing membrane were completely filled during application, remained filled during the curing process, and that the sealant maintained these properties while being subjected to increased pressure loadings and unloading through its service life. Use of Low Expansion Foams in Drainage Method Installations An important inconsistency was noticed with the use of low expansion foams in drainage method installations. Drainage method installations techniques are developed under the expectations th at water may enter the window -wall interface at some point during the service life of an installed window. Provisions are made in these methods to manage this water by redirecting it to the building exterior through two drainage channels provided in the e xterior seal at the sill of the window. A crucial component of this drainage technique is a staunch interior moisture/air barrier that ensures water accumulated in the window -wall interface is prevented from infiltrating to the building interior. Accordi ng to the manufacturers application instructions for low expansion foams these materials are intended to not only seal but fill the gap between the rough opening and the window. To properly install these low expansion foams the
133 rough opening gap is to b e filled from at its deepest portion with an amount of foam equivalent to 1/3 to 1/2 the volume of the gap, based on the expansion properties of the particular product used. The foam then expands out of the gap as it cures filling all the voids within the cavity of the window frame. Applied in this manner, the expansion of the foam acts to impede the functionality of the drainage channels as well as prevents any incidental water from entering the window opening. The apparent contradiction between the properties of low expansion foam sealants inherent to their application and the moisture management theory behind drainage method installations leads to the conclusion that the use of low expansion foams for interior seals may be ill advised for drainage inst allations. Factors Affecting the Water Penetration Resistance of Fenestration Installations The water penetration resistance of fenestration products is a crucial issue in mitigating the costly damages incurred from the wind-driven rain conditions associat ed with landfalling hurricanes. In order to ensure the proper integration of the fenestration unit into the building envelope, several key factors must be considered. Maintaining continuity between the four critical barriers of the window assembly and the wall systems. The ability of the building envelope to properly function relies on the presence and continuity of four critical barriers including the water shedding barrier, the exterior moisture barrier, the vapor barrier, and the air barrier. All of these barriers must be thoroughly considered when integrating a window into the surrounding wall system in order to preserve their continuity. Inconsistencies developed between the window and the wall assemblies resulting from poor installation details wi ll ultimately result in water intrusion. Selection of the proper sealant used to integrate the window into the surrounding wall assembly. The selection of an appropriate sealant for the installation of fenestration products is paramount to their water penetration performance. A high quality sealant with good workability, low shrinkage, and superior adhesion to the materi als used in the installation was found to work the best Information of the durability and movement capabilities of a particular sealant should be reviewed before making a selection. Selection of compatible flashing components. The flashing components including the WRB, sill flashing membrane and window flashing should all be tested for water permeability and adhesion to each other and the other building components used in the installation.
134 Proper maintenance and upkeep of the building exterior. A major source of incidental water in the window -wall interface is superficial cracks in the exterior cladding of the wall assembly. Particularl y in stucco finishing systems small cracks in the faade serve as entry points for infiltrating water that ultimately migrate to the window -wall interface. Stucco is not a water impermeable building product; however, it can be a successful component in a water resistant wall system if properly maintained. The performance of stucco rendering relies largely on proper application of high quality paint. If maintained accordingly the paint application will help to seal any small cracks in the stucco endured during its service. Coordination of trades and proper sequencing of the construction process. Drainage method installations require a significant amount of coordination between trades. The contractor must make sure that the window installer leaves the ap propriate sized gaps at the proper locations within the exterior seal for the drainage channels. Care must then be taken to guarantee that t he drainage channels remain free from obstruction during the application of the exterior cladding and paint. This becomes a difficult task as the drainage channels are provided behind the mounting flange of the window and cannot be seen. Craftsmanship and quality of construction. Craftsmanship and quality of construction are more influential than any other factor s af fecting the water penetration resistance of window installations. The flashing sequences and techniques used in many installation procedures are highly involved and reliant on proper implementation. Regardless of the robustness of an installation method incorrect execution will always lead to water intrusion. Recommendations Improved water penetration performance of installed windows may be achieved through continual development of more robust installation practices. In order to accomplish this, builders and design professionals must be first educated on the issues associated with the water penetration resistance of fenestration products. Hence, recommendations to ensure that all of the factors influencing fenestration installations are addressed are listed herein. Selection of the proper sealant used to integrate the window into the surrounding wall assembly. More definitive guidance must be offered in the selection of sealants used in the installation of fenestration products. Currently ASTM E 2112 leaves the selection of the sealant to the discretion of the builder or designer based on the vague guidance provided in Tables A4.1 and A4.2 of the standard. Other installation standards simply require that sealants be chosen based on compliance with AS TM C 920 or AAMA 808.3 for elastomeric joint sealants and AAMA 812 for low expansion foam sealants. Clear and distinct guidelines should be given for the selection of sealants used for the installation of fenestration products. Further testing requiremen ts may also be required that evaluate the sealant under in situ conditions such as the testing p er formed in this study.
135 Selection of compatible flashing components. Further evaluation and advancement is required for flashing products. The adhesion of the se materials to other building components used in the installation of the windows is critical to their water penetration resistance. These products should be developed with the adhesion of polyurethane and siliconized acrylic latex sealants in mind. Coordination of trades and proper sequencing of the construction process. A high degree of coordination is required for drainage installations to ensure that the drainage channels provided at the sill and remain unrestricted throughout the building construction process. To accomplish this, efforts must be made among all of the trades involved in the installation of fenestration products. Window manufacture r s can incorporate designations in the backside of mounting flanges to denote the appropriate size and loc ations of gaps required in the exterior seal for drainage channels. The party responsible for the installation of the window can apply a small piece of tape over the location of the drainage channel to inform subsequent trades to their location. The build er responsible for the coordination of the trades associated with the installation can create project stickers that clearly indicate the installation method to be placed on the windows. The contractor responsible for the exterior cladding must use extreme care to prevent the exterior finishing system from constricting the drainage of the channels. The education of the industry and the public must serve as the foundation for whatever steps are taken to develop more robust window installation standards. It is important that issues influencing the water penetration resistance of fenestration products be brought to the attention of building officials, designers, contractors, manufacturers, homeowners, and the various industry associations in order to take steps toward mitigating the effects of water intrusion through the building envelope during tropical storm events. Once all aspects of the industry are informed and knowledge on the matters of water ingress through buildings a collaborative effort may be unde rtaken toward the advancement of fenestration installation techniques.
136 CHAPTER 8 SUGGESTIONS FOR F UTURE R ESEARCH The results for the initial four pressure loading sequences as well as the interior moisture/air barrier testing were analyzed. Conclusions were drawn from this data regarding the affects of the multiple variants included in the test matrix on the water penetration resistance of the wall specimens. Significant finding s were achieved by simply compar ing the initial leakage pressure of the specimens. However, certain limitations were inherent to the testing. Some conclusions remained unexplained because the origin of the leakage path could not be determined for fear of damaging the wall specimen for subsequent tests. Without knowledge of the origin of a leakage path it is difficult to determine the building component responsible for the water intrusion. Furthermore, because the wall specimens of this testing were constructed by and under the supervi sion of industry professionals in a laboratory environment the efficacy of their results to typical construction pr actices are questioned Hence, suggestions are made to resolve these issues in future research. Identification of Leakage Paths One of the main limitations encountered during the testing of the wall specimens was the inability to definitively locate the origin of the leakage paths. Without knowing the starting points of the leaks it is difficult to pinpoint the weak areas of the wall assembly. This is of particular concern when comparing the drainage and barrier method installations. The key feature that separates drainage method installations from water barrier installations is the provisions made to control any water that may leak into the window opening through the adjacent wall assembly. While various leaks were observed in the testing, because the source was unknown it was difficult to differentiate the leakage through the wall assembly from the
137 leakage through the window itself. In o rder to locate the origin of these leaks further testing protocols are suggested. The source of a particular leak may be exposed by subjecting the wall specimens to additional rounds of pressure testing utilizing a controlled spray. When a leak was observ ed during the four initial rounds of pressure loading the location and the pressure at which the leak occurred was recorded. To analyze a particular leakage path, the static pressure corresponding to that leak may be applied to the wall specimen while the wetted surface of the wall is restricted. By systematically wetting isolated sections of the wall the possible sources of the leak may be confined to a particular area. Careful scrutiny of this area may help to reveal the origin of the leak and conseque ntly the path it followed to the interior surface of the wall assembly. Another option that may be used to locate the leakage points in the wall assembly is the deconstruction of the specimen s directly following a test. A concentration of tracer dye was a dded to the reservoir of the wetting system used for the static and cyclic tests to aid in the detection of leaks. As the water evaporates on the wall specimens after a test, the pigment of the tracer dye remains and a stain is left. If a particular wall is re tested under either the static or the pulsating pressure tests, the tracer dye laden water will wet the known leakage paths highlighting the leaks. After the completion of the test these leakage paths should be apparent as the wall specimen is carefully disassembled. While this procedure involves little speculation in the derivation of the leakage paths, the walls specimens are sacrificed in the process. After the deconstruction process the test specimens will no longer be useful for any subsequ ent evaluations. For this reason, it is recommended that this leakage analysis technique be applied to a select few walls displaying representative leaks. Conclusions from these walls may then be extrapolated to the remaining test specimens.
138 Environmenta l Aging The thirty -four wall specimens in the testing matrix were designed to serve as a representative sample of residential and light commercial construction. From the commencement of design to the completion of painting, every facet of thes e wall spe cimens was supervised and performed by professionals in the industry. Consequently, the issue has been raised that these test specimens may not be indicative of what would currently be observed on a building in a hurricane prone region. The wall specimen s were tested with a code -compliant stucco application sealed with a fresh coat of paint. Throughout the 134 tests that were preformed on the 18 wall specimens neither cracking in the stucco nor peeling of the paint was observed. It is a well documented fact that overtime stucco may crack and paint may peel without proper maintenance, thus resulting in possible leakage points during extreme winddriven rain conditions. The cause of these factors may be traced back to either quality of construction, envi ronmental exposure, or some combination thereof. While it would be unreasonable to test poor craftsmanship, it is possible to control the environment to which the walls are exposed. The wall specimens as of now have been tested in their ideal conditions The walls were constructed inside of a well ventilated laboratory facility shielded from the sun, and the majority of the testing was performed in this same facility immediately following their completion. The intent is to now relocate these walls to a n environment that more closely replicates the weather conditions found in hurricane prone areas. After a specified period the wall specimens will then be re -tested and the results will be compared to those of the previous tests to evaluate the aging effe cts on water intrusion. The natural weathering of building materials and its components is dependent on a multitude of factors that influence aging, each varying in degree based on geographical location.
139 These factors include but are not limited to solar radiation, cyclical temperature oscillations, changes in pressure, changes in relative humidity, microbiological degradation, and caustic chemicals present in the air. The success of artificial aging procedures is measured on their ability to accurately r eplicate these factors in the appropriate combinations. The two main approaches to simulate aging involve accelerated aging and natural aging. Accelerated aging seeks to replicate the aging effects of the indigenous environment over a short period of time This is usually accomplished by placing the specimen inside some form of incubation chamber where the temperature is cycled from extreme highs to extreme lows for a specified number of cycles. In some cases these chambers may be equipped with spray racks to simulate rain wetting, ultra -violet light sources to imitate the effects of the sun, and in some cases, equipment to moderate the internal pressure. While accelerated aging techniques greatly reduce the time required to produce the desired effects, they can be very costly to construct and maintain. Also, several factors crucial to the actual aging of specimens cannot be recreated using these techniques such as microbial degradation, chemicals present in the air, and eroding particles. For this reas on, a more natural aging approach will be taken to weather the components of the thirty-four wall specimens of this project In natural (or environmental) aging the specimen is exposed to its indigenous environment for a period of time sufficient to model the desired weathering effects. While this approach may take significantly longer than accelerated aging regimens, the factors influencing aging are more closely replicated because the specimen is exposed to its working environment. For the walls of this project a modified natural aging approach was developed. In this approach, the wall specimens will be housed in an environmental aging facility customized to enhance the aging effects driven by temperature and humidity differentials.
140 The environmental aging facility will be a four sided structure with a low slope roof. The front wall will be constructed with 2.4 m x 2.4 m ( 8 ft x 8 ft ) removable sections spaced a minimum of 609.6 mm ( 2 ft ) apart to accept the wall specimens for the duration of the test To ensure that the test specimens will receive maximum exposure to the sun, the building will be situated so that the front wall has a southern exposure and a reflective surface will be placed at the base of the walls. Temperature and humidity inside t he environmental aging facility will be controlled. The facility will be kept at ambient room temperature, 24oC (75oF ), with a feasibly low relative humidity. By creating a dramatic difference in temperature and humidity between the interior and exterior faces of the wall moisture will be driven through the wall assembly decreasing the aging period. The wall specimens will be aged for a period of six months and then re -subjected to the four rounds of pressure loading. The leakage results from this seri es of testing may then be compared to the original leakage results to evaluate the effects of aging and material degradation on the leakage paths.
141 APPENDIX WALL SPECIMEN LEAKAGE RESULTS Wall: Wood Exterior Finish: Stucco Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: ASTM E 2112 Flashing: 228.6 mm mechanical Interior Seal: N/A N/A N/A N/A N/A N/A N/A N/A N/A 4:25 459.6-1024.6 sill corner window 4:30 459.6-1024.6 right jamb installation 4:30 459.6-1024.6 head corner window N/A N/A N/A N/A 017 STATIC N/A CYCLIC N/A through window joinery, insufficient sealant approximately 457.2 mm up, through sealant Time (min:sec) Pressure (Pa) Location Description Infiltrated Component leak through window joinery, insufficient sealant DYNAMIC STATIC 2 N/A Wall: Wood Exterior Finish: Stucco Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 100 Flashing: 101.6 mm self-adhered Interior Seal: Backer Rod/Sealant 17:45 842.7 sill installation 16:30 287.3-861.8 sill installation 17:30 287.3-861.8 sill installation 19:39 335.2-1005.5 sill corner installation 20:40 335.2-1005.5 sill installation N/A N/A N/A N/A 20:00 957.6 sill installation 017B Time (min:sec) Pressure (Pa) Location Description Infiltrated Component STATIC STATIC 2 CYCLIC 254.0 mm 304.8 mm from right jamb, between sealant and flashing 254.0 mm from right, between sealant and flash 50.8 mm from left, between sealant and flash 406.4 mm from right, between sealant and flash 19.1 mm from left corner, between sealant and flash bottom right side, between sealant and flash N/A DYNAMIC
142 Wall: Wood Exterior Finish: Stucco Sill: Flush Sill Window: Single Hung DP Rating: 2633 Pa Window Material: Aluminum Installation: FMA/AAMA 100 Flashing: 101.6 mm self-adhered Interior Seal: Low Expansion Foam 9:30 493.2 sill corner window 12:00 670.3 meeting rail window 15:00 948.0 sill window 18:50 1225.7 jamb window 9:24 191.5-574.6 sill corner window 12:45 239.4-718.2 meeting rail window 13:50 239.4-718.2 sill corner window 1:47 172.4-316.0 meeting rail window 1:58 373.5-593.7 operable sash window 2:10 344.7-632.0 operable sash window 3:35 517.1-948.0 meeting rail window 4:00 459.6-1024.6 meeting rail window 9:11 464.4 sill corner window 12:19 713.4 sill window 15:40 976.8 sill window 17:45 1149.1 meeting rail window 017C STATIC sill overflowing at corners drops traveling upward on right jamb at corner of meeting rail entire sill overflowing left jamb leak from weather-stripping approximately 114.3 mm from sill. CYCLIC Time (min:sec) Pressure (Pa) Location Description Infiltrated Component bottom right corner, leak coming over sill pan past weather-stripping water is being driven up between meeting rail at left corner leak has spread to various spots along the meeting rail STATIC 2 right jamb meets sill, seems to climb up the weather stripping approximately 254.0 mm from left, overflow from sill DYNAMIC droplets rising over sill during high pressure pulse droplets traveling upward between meeting rail; through fixed joint droplets rising over sill during high pressure pulse water is being driven up between meeting rail at right corner bottom left corner, leak coming over sill pan past weather-stripping entire sill overflowing droplets traveling upward on right jamb at corner of meeting rail Wall: Wood Exterior Finish: Stucco Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 100 Flashing: 228.6 mm mechanical Interior Seal: Backer Rod/Sealant 19:30 943.2 sill corner installation 16:00 287.3-861.8 sill installation N/A N/A N/A N/A 17:27 818.8 sill corner installation STATIC 2 DYNAMIC CYCLIC STATIC 017D Time (min:sec) Pressure (Pa) Location Description Infiltrated Component bottom right corner of sill, between sealant and flashing 25.4 mm from corner of sill, between sealant and flashing N/A bottom right corner of sill, between sealant and flashing (grooves)
143 Wall: Wood Exterior Finish: Stucco Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 100 Flashing: 228.6 mm mechanical Interior Seal: Low Expansion Foam 6:20 215.5 sill installation 7:15 272.9 sill installation 9:20 368.7 sill installation 17:30 833.1 jamb installation 19:20 924.1 jamb installation 1:46 67.0-205.9 sill installation 3:15 95.8-287.3 sill installation 18:00 335.2-1005.5 jamb installation 20:15 335.2-1005.5 jamb installation 2:53 306.4-679.9 sill installation 4:00 459.6-1024.6 sill installation 3:18 138.9 sill installation 8:06 306.4 jamb installation 20:00 957.6 jamb installation right jamb 31.8 mm above sill, between foam and flashing 017E STATIC 2 DYNAMIC CYCLIC STATIC 152.4 mm 203.2 mm from right corner, between foam and flashing 228.6 mm from left jamb, between foam and flashing Time (min:sec) Pressure (Pa) Location Description Infiltrated Component right jamb 76.2 mm above sill, between foam and flashing 127.0 mm from left jamb, between foam and flashing leaking along entire foam and flashing interface 127.0 mm from right jamb, multiple leaks between foam and flash ensue right jamb 12.7 mm above sill, between foam and flashing right jamb 31.8 mm above sill, between foam and flashing center of sill, between foam and flashing right jamb 12.7 mm up from sill, between foam and flashing right jamb 50.8 mm up from sill, between foam and flashing 482.6 mm from left jamb, between foam and flashing leaking along entire sill between foam and flashing interface Wall: Wood Exterior Finish: FCB Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: ASTM E 2112 Flashing: 101.6 mm self-adhered Interior Seal: N/A 18:17 866.6 sill corner installation N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 018 STATIC CYCLIC DYNAMIC STATIC 2 bottom right corner of window N/A N/A N/A Time (min:sec) Pressure (Pa) Location Description Infiltrated Component Wall: Wood Exterior Finish: FCB Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 100 Flashing: 101.6 mm self-adhered Interior Seal: Backer Rod/Sealant N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 018B STATIC CYCLIC STATIC 2 N/A N/A N/A N/A DYNAMIC Time (min:sec) Pressure (Pa) Location Description Infiltrated Component
144 Wall: Wood Exterior Finish: FCB Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 100 Flashing: 101.6 mm self-adhered Interior Seal: Low Expansion Foam N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A Time (min:sec) Pressure (Pa) Location Description Infiltrated Component 018C N/A STATIC CYCLIC DYNAMIC STATIC 2 N/A N/A N/A Wall: Wood Exterior Finish: FCB Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 100 Flashing: 228.6 mm mechanical Interior Seal: Backer Rod/Sealant N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 018D N/A Time (min:sec) Pressure (Pa) Location Description Infiltrated Component STATIC CYCLIC DYNAMIC STATIC 2 N/A N/A N/A Wall: Wood Exterior Finish: FCB Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 100 Flashing: 228.6 mm mechanical Interior Seal: Low Expansion Foam N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 018E STATIC N/A Time (min:sec) Pressure (Pa) Location Description Infiltrated Component CYCLIC DYNAMIC STATIC 2 N/A N/A N/A
145 Wall: CMU Exterior Finish: DCC Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: ASTM E 2112 Flashing: N/A Interior Seal: N/A 15:00 957.6 jamb installation 11:25 191.5-574.6 jamb installation pre-wetting 0.0 jamb installation 3:15 565.0-890.6 jamb installation 8:55 354.3 jamb bucking installation 16:00 742.1 jamb installation left jamb leaking through bucking, leakage shows at sill corner left jamb 254.0 mm above sill, between seal and bucking, shows at sill Time (min:sec) Pressure (Pa) Location Description 019 STATIC CYCLIC DYNAMIC STATIC 2 left jamb 254.0 mm above sill, between seal and bucking, shows at sill left jamb 254.0 mm above sill, between seal and bucking, shows at sill moisture noticed at sill, suspected to come from jamb leak left jamb 254.0 mm above sill, between seal and bucking, shows at sill Infiltrated Component Wall: CMU Exterior Finish: DCC Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 200 Barrier Flashing: LAF & Wire Lath Interior Seal: Backer Rod/Sealant N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A Description Infiltrated Component N/A N/A 019B Time (min:sec) Pressure (Pa) Location STATIC 2 DYNAMIC CYCLIC STATIC N/A N/A Wall: CMU Exterior Finish: DCC Sill: Face Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 200 Drainage Flashing: LAF & Bonding Agent Interior Seal: Backer Rod/Sealant 7:20 272.9 glazing window 10:30 430.9 glazing window 17:20 814.0 sill installation 6:20 143.6-430.9 glazing window 12:00 239.4-718.2 sill installation 0:40 181.9-296.9 glazing window 2:00 373.5-593.7 sill installation 6:00 191.5 glazing window 019C STATIC CYCLIC DYNAMIC head jamb 25.4 mm 177.8 mm from left, between glazing and glazing stop Time (min:sec) Pressure (Pa) Location Description Infiltrated Component STATIC 2 head jamb 76.2 mm 203.2 mm from left, between glazing and glazing stop head jamb, leak behind glazing completely filled sill 25.4 mm from left jamb, between sealant and sill head jamb 76.2 mm 101.6 mm from left, between glazing and glazing stop 25.4 mm from left jamb, between sealant and sill head jamb 25.4 mm 152.4 mm from left, between glazing and glazing stop 25.4 mm from left jamb, between sealant and sill
146 Wall: CMU Exterior Finish: DCC Sill: Flush Sill Window: Awning DP Rating: 2523 Pa Window Material: Vinyl Installation: FMA/AAMA 200 Drainage Flashing: LAF on Interior Interior Seal: Low Expansion Foam N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A Time (min:sec) Pressure (Pa) Location Description Infiltrated Component 035 STATIC CYCLIC DYNAMIC STATIC 2 N/A N/A N/A Wall: CMU Exterior Finish: Stucco Sill: Face Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: ASTM E 2112 Flashing: N/A Interior Seal: N/A N/A N/A N/A N/A 18:00 335.2-1005.5 sill wall N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 020 STATIC CYCLIC DYNAMIC STATIC 2 Time (min:sec) Pressure (Pa) Location N/A moisture observed from interior at bottom left corner of sill Description Infiltrated Component Wall: CMU Exterior Finish: Stucco Sill: Flush Sill Window: Awning DP Rating: 2523 Pa Window Material: Aluminum Installation: FMA/AAMA 200 Barrier Flashing: LAF on Interior Interior Seal: Backer Rod/Sealant N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 016 STATIC CYCLIC DYNAMIC STATIC 2 Time (min:sec) Pressure (Pa) Location Description Infiltrated Component
147 Wall: CMU Exterior Finish: Stucco Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 200 Drainage Flashing: LAF & Wire Lath Interior Seal: Backer Rod/Sealant 6:38 239.4 sill corner installation 2:30 67.0-205.9 sill corner installation 10:05 191.5-574.6 jamb window 2:47 306.4-679.9 jamb window 4:15 138.9 sill corner installation STATIC CYCLIC DYNAMIC STATIC 2 020C Time (min:sec) Pressure (Pa) Location Description Infiltrated Component bottom right corner, leaked through skip in sealant bottom right corner, between sealant and liquid applied flashing bottom screw in left jamb, flows into sill bottom screw in left jamb, flows into sill bottom right corner, leaked through skip in sealant Wall: CMU Exterior Finish: Stucco Sill: Flush Sill Window: Fixed DP Rating: 1915 Pa Window Material: Aluminum Installation: FMA/AAMA 200 Drainage Flashing: LAF & Wire Lath Interior Seal: Low Expansion Foam 10:45 502.7 jamb installation 16:15 746.9 jamb window 6:30 143.6-430.9 jamb installation 11:00 191.5-574.6 jamb installation 12:30 239.4-718.2 jamb window 12:40 239.4-718.2 jamb window 1:45 373.5-593.7 jamb installation 3:00 565.0-890.6 jamb installation 3:54 459.6-1024.6 jamb window 3:55 459.6-1024.6 jamb window 11:09 521.9 jamb installation 16:15 756.5 jamb window 17:40 833.1 jamb window 020D STATIC CYCLIC Infiltrated Component leaking from bottom screw in right jamb leaking from bottom screw in left jamb right jamb 25.4 mm above sill, between jamb and foam leaking from bottom screw in left jamb leaking from bottom screw in right jamb Time (min:sec) Pressure (Pa) Location Description DYNAMIC STATIC 2 right jamb 44.5 mm above sill, between window frame and foam leaking from bottom screw in left jamb right jamb 38.1 mm 50.8 mm above sill, between jamb and foam interface right jamb 76.2 mm 88.9 mm above sill, between foam and bucking leaking from bottom screw in left jamb leaking from bottom screw in right jamb right jamb 25.4 mm above sill, between jamb and foam right jamb 63.5 mm above sill, foam and bucking
148 LIST OF REFERENCES Anis, W., Quirouette, R., and Rousseau, J. (2006). Air and Vapor Barriers for Southern Buidlings. Building Envelope Forum Online Newsletter (April 04, 2009). American Architectural Manufacturers Association (AAMA). (1992). Voluntary Specification and Test Methods for Sealants: Exterior Perimeter Sealing Compounds. AAMA 808.31992, American Architectural Manufac turers Association, Schaumburg, Illinois. AAMA. (2004). Voluntary Specification, Performance Requirements and Test Voluntary Practice for Assessment of Single Component Aerosol Expanding Polyurethane Foams for Sealing Rough Openings of Fenestration Inst allations. AAMA 812 -04, American Architectural Manufacturers Association, Schaumburg, Illinois. AAMA. (2005). Standard/Specification for Windows, Doors, and Unit Skylights. AAMA/WDMA/CSA 101/I.S.2/A440 -2005, American Architectural Manufacturers Asso ciation, Schaumburg, Illinois. AAMA. (Draft). Voluntary Specification for Rating the Severe Wind Driven Rain Resistance of Windows Doors and Unit Skylights. AAMA XXX-XX American Architectural Manufacturers Association, Schaumburg, Illinois. American Concrete Institute (ACI). (2005). Specification for Masonry Structures. ACI 530.1-05/ASCE 6 -05/TMS 602-05, American Concrete Institute, Farmington Hills, Michigan. American Society of Civil Engineers (ASCE). (2006). Minimum Design Loads for Buildings and Other Structures ASCE 7 -05, American Society of Civil Engineers, Reston, Virginia. American Society for Testing and Materials (ASTM). (1995). Standard Test Method for Water Penetration of Exterior Windows, Skyligh ts, and Doors by Rapid Pulsed Air Pressure Difference. ASTM E 2268-04, West Conshohocken, Pennsylvania. ASTM. (2000a ). "Standard Test Method for Water Penetration of Exterior Windows, Skylights, Doors, and Curtain Walls by Uniform Static Air Pressure D ifference. ASTM E 331-00, American Society for Testing and Materials, West Conshohocken, Pennsylvania. ASTM. (2000b ). Standard Test Method for Field Determination of Water Penetration of Installed Exterior Windows, Skylights, Doors, and Curtain Walls by Uniform or Cyclic Static Air Pressure Difference. ASTM E 1105-00, American Society for Testing and Materials, West Conshohocken, Pennsylvania. ASTM. (2001a). Standard Specification for Load Bearing Concrete Masonry Units. ASTM C 90-01a American S ociety for Testing and Materials West Conshohocken, Pennsylvania. ASTM. (2001b). Standard Specification for Mortar for Unit Masonry. ASTM C 270-01a, American Society for Testing and Materials West Conshohocken, Pennsylvania.
149 ASTM. (2001c). Standa rd Specification for Grout for Masonry. ASTM C 476-01, American Society for Testing and Materials West Conshohocken, Pennsylvania. ASTM. (2001d ). Standard Test Method for Adhesion -in -Peel of Elastomeric Joint Sealants. ASTM C 794-01, American Soci ety for Testing and Materials, West Conshohocken, Pennsylvania. ASTM. (2005). Standard Specification for Elastomeric Joint Sealants. ASTM C 920-05, American Society for Testing and Materials West Conshohocken, Pennsylvania. ASTM. (2006a). Standard Specification for Application of Portland Cement Based Plaster. ASTM C 926-06, American Society for Testing and Materials West Conshohocken, Pennsylvania. ASTM. (2006b). Standard Specification for Installation of Lathing and Furring to Receive Interior and Exterior Portland. ASTM C 1063-06, American Society for Testing and Materials West Conshohocken, Pennsylvania. ASTM. (2007). Standard Practice fo r Installation of Exterior Windows, Doors, and Skylights. ASTM E 2112-07, American Society for Testing and Materials, West Conshohocken, Pennsylvania. Australian/New Zealand Standard (AS/NZS). (1995). Testing of Building Facades. AS/NZS Standard 428 4 -1995, Standards Australia, Homebush, NSW, Australia, and Standards New Zealand, Wellington, New Zealand. Blake, E. S., Rappaport, E. N., and Landsea, C. W. (2007). The Deadliest, Costliest, and Most Intense United States Tropical Cyclones from 1851 to 2004 (and Other Frequently Requested Hurricane Facts) NOAA Technical Memorandum, NWS TPC 5, National Oceanic and Atmospheric Administration, Tropical Prediction Center, National Hurricane Center, Miami, Florida, 1 24. Building Code Compliance Office (BCCO ). (2006). Post Hurricane Wilma Progress Assessment Miami -Dade County Building Code Compliance Office, Miami, Florida. Crossett, K. M., Culliton, T. J., Wiley, P. C., and Goodspeed, T. R. (2004). Population Trends Along the Coastal United States: 1980 -2008. Coastal trends report series, U.S. Dept. of Commerce, National Oceanic and Atmospheric Administration, National Ocean Service, Management and Budget Office, Special Projects, Silver Spring, Maryland. Crowder Moore, B. J., Weston, T. A., and Katsar os, J. D. (2006). Performance Testing of Flashing Installation Methods for Brick Mold and Nonflanged Windows. Journal of ASTM Internaional American Society for Testing and Materials, West Conshohocken, Pennsylvania, 3(1), 133152. Federal Emergency Ma nagement Agency (FEMA). (1992). Building Performance: Hurricane Andrew in Florida Observations, Recommendations, and Technical Guidance FIA 22,
150 Federal Emergency Management Agency, Federal Insurance Administration, Washington, DC. FEMA. (1999). Building Envelope Performance. Building Performance Assessment: Hurricane Georges in Puerto Rico, FEMA 339, Federal Emergency Management Agency, Washington, DC. FEMA. (2005a ). Building Envelope Performance. Mitigation Assessment Team Report: Hurrica ne Charley in Florida, FEMA 488, Federal Emergency Management Agency, Washington, DC. FEMA. (2005b ). Building Envelope Performance. Mitigation Assessment Team Report: Hurricane Ivan in Alabama and Florida, FEMA 489, Federal Emergency Management Agency, Washington, DC. FEMA. (2005c ). Summary Report on Building Performance: 2004 Hurricane Season, FEMA 490, Federal Emergency Management Agency, Washington, DC. FEMA. (2006a ). Hurricane Katrina in the Gulf Coast Building Performance Observations, Recommendations, and Technical Guidance: Mitigation Assessment Team Report FEMA 548, Federal Emergency Management Agency, Washington, DC. FEMA. (2006b ). Summary Report on Building Performance: Hurricane Katrina 2005, FEMA 549, Federal Emergency Managem ent Agency, Washington, DC. Fenestration Manufacturers Association (FMA) (2007). Standard Practice for the Installation of Windows with Flanges or Mounting Fins in Wood Frame Construction. FMA/AAMA 100 -07, Fenestration Manufacturers Association, Tallahassee, Florida. FMA. (200X Draft). Standard Practice for the Installation of Windows with Frontal Flanges for Surface Barrier Masonry Construction. FMA/AAMA Standard 200-0X Fenestration Manufacturers Association, Tallahassee, Florida. Grier, J. ( 1996). Florida Residents are Still Weathering the Storm in the Aftermath of Hurricane Andrew: The Cost of Starting Over is Prohibitive for Many. Partners, Federal Reserve Bank of Atlanta, Atlanta, Georgia, 6(2). Gutro, R., Patzert, B. (2008). 2007 H urricane Season Starts Early, Ends Late. NASA's Goddard Space Flight Center/Jet Propulsion Lab, (Feb. 20, 2009). Institute for Business and Home Safety (IBHS). (2004). Preliminary Damage Observations: Hurricanes Charley, Frances & Ivan 2004, Institute for Business and Home Safety, Tampa, Florida. Insurance Information Institute (III). (2008a). Florida Property Insurance Facts. Research and Analysis < http:// www.iii.org/media/research/floridafacts08/> (Feb. 18, 2009). III. (2008b). Florida Hurricane Insurance Fact File. Research and Analysis < http://www.iii.org/media/research/FLfactfile/> (Feb. 18, 2009).
151 III. (2008c). Catastrophes: U.S. Facts and Statistics (Feb. 18, 2009). Japanese Industrial Standard (JIS). (1984). Test Method of Water Tightness for Windows and Doors. JIS A 1517-1984, Japanese Standards Association, Tokyo, Japan. Katsaros, J. D. (2005). Adhesive Characterization and Durability of Se lf -Adhered Flashings. Pro., ASTM Symposium on Performance and Durability of the Window Wall Interface American Society for Testing and Materials, West Conshohocken, Pennsylvania. Lstiburek, J. W. (2005). Rainwater Manaement Performance of Newly Constr ucted Residential Building Enclosures During August and September 2004, Building Science Corporation, Westford, Massachusetts Lstiburek, L. W. (2008). Drainage, Holes and Moderation. Building Science Insights BSI 004, < http:// www.buildingscience.com/doctypes/building-science insights> (Jul. 30, 2008). Mullens, M., Hoekstra, B., Nahmens, I., and Martinez, F. (2006). Water Intrusion in Cent ral Florida Homes During Hurricane Jeanne in September 2004, University of Central Florida Housing Constructability Lab, Orlando, Florida. National Hurricane Center (NHC). (200 7). Hurricane History. Hurricane Preparedness (March 3, 2009). National Science Board (NSB). (2007). Hurricane Warning: The Critical Need for a National Hurricane Research Initiative. National Science Foundat ion, Arlington, Virginia. Pielke, R. A., Jr., and Landsea, C. W. (1998). Normalized Hurricane Damages in the United States: 192595. Weather and Forecasting, 13(3), 621631. Pielke, R. A., Jr., Gratz, J.,Landsea, C. W., Collins, D., Saunders, M. A., and Musulin, R. (2008). Normalized Hurricane Damage in the United States: 19002005. Natural Hazard Review 9, p. 2942. Pinelli, J. P., Simiu, E., Gurley, K., Subramanian, C., Zhang, L., Cope, A., Filliben, J. J., and Hamid, S. (2004). Hurricane Da mage Prediction Model for Residential Structures. Journal of Structural Engineering, 130(11), 1685 1691. Sparks, P. R., Schiff, S. D., and Reinhold, T. A. (1994). Wind Damage to E nvelopes of H ouses and Consequent I nsurance L osses. Journal of Wind Engineering and Industrial Aerodynamics 53(1 2), 145 155. RDH Building Engineering Limited. (2002a ). Water Penetration Resistance of Windows: Study of Manufacturing, Building Design, Installation, and Maintenance Factors RDH Building Engineering Limit ed, Vancouver, B. C., Canada. RDH Building Engineering Limited. (2002b). Water Penetration Resistance of Windows: Study of Codes, Standards, Testing, and Certification, RDH Building Engineering Limited, Vancouver, B. C., Canada.
152 Valverde, L. J., Jr., an d Andrews, M. W. (2006). Global Climate Change and Extreme Weather: An Exploration of Scientific Uncertainty and the Economics of Insurance. Insurance Information Institute Working Paper Series Insurance Information Institute, New York, New York.
153 BIOGRAPHICAL SKECTH Cory Thomas Salzano was born in the year of 1984 in Fort Lauderdale, Florida The son of a g eneral c ontractor, he was exposed to the construction industry from a young age developing a basic understanding of residential building assemblies. In the summer following his graduation from Cardinal Gibbons High School in 2002, his construction interests led him to pursue a c ivil e ngineering d egree from the Department of Civil and Coastal Engineer ing at the University of Flori da. After graduating with high honors in December 2007, he entered graduate school in the fall of 2008 at the University of Florida Under the mentorship of Dr. Forrest J. Masters, he began researching the hurricane performance of building materials with the specially designed University of Florida Hurricane Simulator. As one of the leading member s at the University of Florida Hu rricane Research Team he participated in research garnering national media exposure. In addition to his research at the University of Florida he was a team leader of the Florida Coastal Monitoring Program, a unique joint venture focusing on full -scale experimental methods to quantify hurricane wind behavior and the resultant loads on residential structures. In 2008, he depl oyed for four named storms along the Gulf Coast United States erecting wind instrumentation to quantify near -surface hurricane winds. He also participated in the documentation and assessment of post -storm damage to residential structures immediately foll owing storm landfall s Cory T. Salzano is a student member of the American Association for Wind Engineering, the American Society of Civil Engineers, American Concrete Institute, Chi Epsilon and Tau Beta Pi.