<%BANNER%>

Development of Time-Varying Wind Uplift Test Protocols for Residential Wood Roof Sheathing Panels

Permanent Link: http://ufdc.ufl.edu/UFE0041332/00001

Material Information

Title: Development of Time-Varying Wind Uplift Test Protocols for Residential Wood Roof Sheathing Panels
Physical Description: 1 online resource (199 p.)
Language: english
Creator: Hill, Kenneth
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: pressure, wind, wood
Civil and Coastal Engineering -- Dissertations, Academic -- UF
Genre: Civil Engineering thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: DEVELOPMENT OF TIME-VARYING WIND UPLIFT TEST PROTOCOLS FOR RESIDENTIAL WOOD ROOF SHEATHING PANELS Kenneth M. Hill (352) 392-9537 Ext. 1502 Civil and Coastal Engineering Dr. David O. Prevatt Master of Engineering December, 2009 This thesis investigates the effects of hurricane loading on residential wood roof sheathing systems. Historically the majority of damage to residential structures in hurricanes has been a direct result of roof system failures. Specifically this investigation develops improved methods for testing the resistance of wood roofs to hurricane loading and verifies results against previous studies. The use of methods developed will enable further more standardized testing of wood roof sheathing systems. Results of standardized testing can then be used to better design wood roofs to resist extreme winds helping to mitigate the billions of dollars spent annually in recovery. The benefit to society lies in the potential savings from residential structures not losing roof sheathing, which in the state of Florida alone represents a population of over $1.5 trillion dollars of existing homes in the line of fire of hurricanes.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kenneth Hill.
Thesis: Thesis (M.E.)--University of Florida, 2009.
Local: Adviser: Prevatt, David O.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041332:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041332/00001

Material Information

Title: Development of Time-Varying Wind Uplift Test Protocols for Residential Wood Roof Sheathing Panels
Physical Description: 1 online resource (199 p.)
Language: english
Creator: Hill, Kenneth
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: pressure, wind, wood
Civil and Coastal Engineering -- Dissertations, Academic -- UF
Genre: Civil Engineering thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: DEVELOPMENT OF TIME-VARYING WIND UPLIFT TEST PROTOCOLS FOR RESIDENTIAL WOOD ROOF SHEATHING PANELS Kenneth M. Hill (352) 392-9537 Ext. 1502 Civil and Coastal Engineering Dr. David O. Prevatt Master of Engineering December, 2009 This thesis investigates the effects of hurricane loading on residential wood roof sheathing systems. Historically the majority of damage to residential structures in hurricanes has been a direct result of roof system failures. Specifically this investigation develops improved methods for testing the resistance of wood roofs to hurricane loading and verifies results against previous studies. The use of methods developed will enable further more standardized testing of wood roof sheathing systems. Results of standardized testing can then be used to better design wood roofs to resist extreme winds helping to mitigate the billions of dollars spent annually in recovery. The benefit to society lies in the potential savings from residential structures not losing roof sheathing, which in the state of Florida alone represents a population of over $1.5 trillion dollars of existing homes in the line of fire of hurricanes.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kenneth Hill.
Thesis: Thesis (M.E.)--University of Florida, 2009.
Local: Adviser: Prevatt, David O.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041332:00001


This item has the following downloads:


Full Text

PAGE 1

DEVELOPMENT OF TIME-VARYING WIND UPLIFT TEST PROTOCOLS FOR RESIDENTIAL WOOD ROOF SHEATHING PANELS By KENNETH M. HILL A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Kenneth M. Hill 2

PAGE 3

To Patricia, Jeff, Christin a, Fitz, Mimi and Laddie 3

PAGE 4

ACKNOWLEDGMENTS The completion of this thesis is due to the hard work of several individuals. I would like to thank Dr. David O. Prevatt for all the support an d guidance. Further I would like to thank my committee Dr. Kurtis Gurley and Dr. Forrest Mast ers for their considerab le contributions through experimental results and valuable insights. I would like to thank the Department of Civil and Coastal Engineering for tuition support and the Florida Departme nt of Community affairs for their support of the research. I would like to extend th anks to a number of students who have aided in the effort needed to complete this work: Peter Datin, Bill Dugary, Zack Farrell, Carl Harrigan, Laun Chau, Jared Easterlin and Johann Weeks who ha ve all in some way contributed to the testing of the many panels. I would also like to thank James Je steadt and Chuck Broward for their help in conducting tests. Finally I would like to thank my family and fr iends for their support a nd confidence. I am especially thankful for Christina whos e support has made all the difference. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .......................................................................................................................11 ABSTRACT ...................................................................................................................................16 CHAPTER 1 INTRODUCTION ................................................................................................................ ..17 Motivation ...............................................................................................................................17 Objectives ...............................................................................................................................19 Organization of Thesis ............................................................................................................19 2 LITERATURE REVIEW .......................................................................................................22 Wind Loads on Low-Rise Buildings ......................................................................................22 Response to Earlier Historical Hurricane Damage .................................................................24 Recent Hurricane Damage to the United States .....................................................................25 Construction of Residential Houses ........................................................................................27 Wind Uplift Behavior of Reside ntial Wood Roof Sheathing Panels ......................................29 Variability in Uplift Capacity Testing .............................................................................29 Retrofit Measures ............................................................................................................31 In Field Testing ................................................................................................................31 Building Codes .......................................................................................................................32 Wind Uplift Resistance of Dowel Type Fastener Connections in Wood ...............................32 ASTM D-1761 Protocol ..................................................................................................32 Effect of Rate of Loadin g on Withdrawal Resistance .....................................................33 Pull-Through Testing .......................................................................................................34 Age and Weathering Effects ............................................................................................34 Nail Withdrawal Tests on Existing Residential Buildings ..............................................35 Dynamic Loading ...................................................................................................................36 Related Research at the University of Florida ........................................................................37 Summary .................................................................................................................................37 3 DEVELOPMENT OF STANDARD WIND UPLIFT METHOD WITH STATIC AND DYNAMIC LOADING ..........................................................................................................46 UF Wood Roof Sheathing Uplif t Test (UF-WRSUT) Protocol .............................................47 The UF-WRSUT, Method A Static Pressure Test Protocol ............................................48 UF WRSUT, Method B Dynami c Pressure Test Protocol ..............................................49 Failure of Wood Roof Sheathing Panels ................................................................................51 5

PAGE 6

Determination of Panel Failure Pressure .........................................................................52 4 EXPERIMENTAL SETUP ....................................................................................................57 Panel Test Series and Construction ........................................................................................57 Panel Test Series ..............................................................................................................57 Laboratory Fabricated Panels (New) ...............................................................................58 Harvested Panels (Aged) .................................................................................................59 Specific Gravity and Moisture Content Measurements ..........................................................61 Uplift Testing Equipment .......................................................................................................63 Panel Installation .............................................................................................................64 Manual Control ................................................................................................................64 Pressure Load Actuator ...................................................................................................65 5 RESULTS AND ANALYSIS.................................................................................................77 Laboratory Static Panel Uplift Tests .......................................................................................77 Evaluation of Fastener Size and Spacing on Uplift Capacity ..........................................77 Comparison of Results with Previous Studies .................................................................79 Evaluation of Calculated Fastener Resistance .................................................................81 Evaluation of ccSPF Retrofit ...........................................................................................82 Failure Mechanisms Observed ........................................................................................83 Harvested Static Panel Uplift Test ..........................................................................................84 Evaluation of Existing Panel Uplift Resistance ...............................................................84 Evaluation of Retrofit of Existing Panels ........................................................................86 Comparison of Laboratory vs. Harv ested Specific Gravity Samples .....................................87 Static vs. Dynamic Panel Uplift Test ......................................................................................88 Comparison of Static vs. Dynamic La boratory Fabricated Panel Uplift Test .................89 Comparison of Static vs. Dynamic Harvested Panel Uplift Test ....................................90 Peak Pressure vs. Failure Pressure ..................................................................................91 Analysis of Variance of Results .............................................................................................91 Procedure .........................................................................................................................92 Laboratory Fabricated Pa nels Tested Statically ..............................................................94 Harvested Panels Tested Statically ..................................................................................95 Laboratory Fabricated Panels Te sted Statically vs. Dynamically ...................................95 6 DISCUSSION .................................................................................................................. .....118 Analysis of Design Wind Speeds .........................................................................................118 Effect of Aging or Weathering on Wind Uplift Resistance ..................................................121 Effect of Dynamic Nature of Wind Uplift Loading ..............................................................122 Wind Uplift Behavior of Resi dential Wood Roof Sheathing ...............................................122 7 CONCLUSIONS AND RECOMMENDATIONS ...............................................................127 Conclusions ...........................................................................................................................127 Comparison of Results with Previous Studies ...............................................................127 Effect of In-Service and Environmen tal Effects on Roof Panel Strength .....................127 6

PAGE 7

Dynamic Load Effects on Wood Panel Strength ...........................................................128 Recommendations .................................................................................................................128 APPENDIX A FULL PANEL UPLIFT RESULTS ......................................................................................130 B FULL FAILURE MODE AND LOCATION INFORMATION ..........................................148 C PANEL CONSTRUCTION OF STATIC VS. DYNAMIC TESTING OF HARVESTED PANELS ......................................................................................................160 D FULL SPECIFIC GRAVITY MEASUREMENTS ..............................................................161 E STATIC VS. DYNAMIC PANEL TESTING, TARGET AND ACTUAL PRESSURE TIME-HISTORIES ...............................................................................................................1 64 LIST OF REFERENCES .............................................................................................................195 BIOGRAPHICAL SKETCH .......................................................................................................199 7

PAGE 8

LIST OF TABLES Table page 2-1 Wind Uplift Failure Pressure Research Conducted on Wood Roof Sheathing Panels (1993 through 2004) ..........................................................................................................42 2-2 Summary of selected uplift pressure tes ting results from Jone s (1998) investigation into adhesive retrofit methods for residential wood roof sheathing ..................................43 2-3 Summary of existing residential roof sh eathing uplift testing c onducted by Judge and Reinhold (2002) .................................................................................................................43 2-4 Summary of building code requirements for roof sheathing design in Florida (1988 to current) ...........................................................................................................................44 2-5 Percentage of original strength summary of Chow et al. in vestigation of ageing effects on fastener withdrawal and pull-through resistance, (Chow et al. 1990) ...............44 4-1 Laboratory fabricated (New) panel series tested, constructed with in. OSB and 2 in. by 4 in. southern yellow pine # 2 or better ...................................................................69 4-2 Harvested panel series tested .............................................................................................70 4-2 Summary of harvested pa nels tested statically an d statically vs. dynamically ..................73 5-1 Results of static UF-WRSUT of laboratory fabricated panels fastened with 2-3/8 in. long 6d smooth shank, 8d smooth shank and 8d ring shank nails .....................................97 5-2 Comparison of mean and 5% exclusion valu e failure pressures for panels fabricated in the lab tested statically UF-WRSUT vs. previous studies .............................................98 5-3 Mean of calculated maximum fast ener loads based on tributary area ...............................98 5-4 Results of static UF-WRSUT of laborato ry fabricated panels fastened with 8d ring shank nails at 6 in. / 12 in. retrofitted with ccSPF adhesive ..............................................99 5-5 Measured failure pressure and calculated fa stener failure load of statically tested panels harvested from existing structures in Central Florida ...........................................102 5-6 Mean failure pressures of static UF-W RSUT of panels harvested from existing LFWS located in Central Florida and retrofitted existing panels ....................................105 5-7 Summary of specific gravities calculated to investigate effect of specific gravity on panel wind uplift resistance .............................................................................................107 5-8 Mean failure pressure of laboratory fabricated panels attach ed with 2 in. long 6d common nails tested statically and dynamically per UF-WRSUT ..................................108 8

PAGE 9

5-9 Mean failure pressure of panels harvested from existing construction attached with 1.5 in. long staples at 4 in. / 4 in. tested statically and dynamically per UF-WRUT .......109 5-10 Summary of comparison of peak vs. failure pressure for all static vs. dynamic panel results (34 panels) ............................................................................................................110 5-11 Summary of reduction from st atically tested to dynamically tested panel resistance for both peak and failure pressures ..................................................................................111 5-12 Summary of series used in analysis of variance ..............................................................111 5-13 ANOVA table for all laboratory fabricat ed static phase re sults (alpha = 0.05, therefore 95% confidence level) ......................................................................................112 5-14 Bonferroni test full results for laboratory fa bricated statically tested panels (alpha = 0.05) .................................................................................................................................113 5-15 Bonferroni test summary fo r laboratory fabricated static tested results (alpha = 0.05) ...115 5-16 ANOVA table for all panels fastened with 6d smooth shank nails spaced at 6 in. / 12 in. (alpha = 0.05, therefore 95% confidence level) ..........................................................116 5-17 Bonferroni test full results for all panels fastened with 6d smoo th shank nails spaced at 6 in / 12 in. (alpha = 0.5) ..............................................................................................116 5-18 T-test result for panels retrofitted with 8d ring shank nails at 6 in. / 12 in. vs. panels with only 8d ring shank nails at 6 in. / 12 in. ...................................................................116 5-19 ANOVA table for all panels tested in static vs. dynami c phase using peak pressure (alpha = 0.05, therefor e 95% confidence level) ...............................................................117 5-20 Bonferroni test full results for panels te sted in static vs. dynamic phase using peak pressure (alpha = 0.05) .....................................................................................................117 5-21 ANOVA table for all panels tested in static vs. dynamic phase using failure pressure (alpha = 0.05, therefor e 95% confidence level) ...............................................................117 5-22 Bonferroni test full results for panels test ed in static vs. dynamic phase using failure pressure (alpha = 0.05) .....................................................................................................117 6-1 Comparison of design wind speeds calc ulated per ASCE 7-05 Method 2 for UFWRSUT Method A results based on A) a factor of safety of 2.0 applied to the mean and B) 5% exclusion of the data (enc losed gable roof building in exposure B assumed, with a mean roof height of 15 ft) .....................................................................125 9

PAGE 10

6-2 Comparison of design wind speeds calc ulated per ASCE 7-05 Method 2 for UFWRSUT Method B results A) statically test ed panels and B) dynamically tested panels (enclosed gable roof building in e xposure B assumed, with a mean roof height of 15 ft) ............................................................................................................................126 A-1 Summary of pane l uplift test series ..................................................................................130 A-2 Full uplift test resu lts of panels fastened with 1.5 in. staples ..........................................132 A-3 Full uplift test resu lts of panels fastened with 2.5 in. staples ..........................................133 A-4 Full uplift test results of panels fastened with 6d smooth shank nails .............................134 A-5 Full uplift test results of panels fastened with 8d smooth shank nails .............................138 A-6 Full uplift test resu lts of panels fastened with 8d ring shank nails ..................................142 D-1 All specific gravity measurements take n from laboratory fabricated static vs. dynamic uplift testing ......................................................................................................161 D-2 All specific gravity measurements take n from harvested static vs. dynamic uplift testing ...............................................................................................................................162 E-1 Summary of dynamic pressure trace for 6d SS at 6 in. / 12 in. panels (blue actual and red target) ..................................................................................................................167 E-2 Summary of dynamic pressure trace for 6d SS at 6 in. / 6 in. panels (blue actual and red target) ..................................................................................................................175 E-3 Summary of dynamic pressure trace for 6d SS at 6 in. / 12 in. retrofitted panels (blue actual and red target) .....................................................................................................182 E-4 Summary of dynamic pressure trace for 1.5 in. Staples at 4/4 panels (blue actual and red target) ..................................................................................................................186 E-5 Summary of dynamic pressure trace for 1.5 in. Staples at 4/4 with Ret. A-2 panels (blue actual and red target) ...........................................................................................192 10

PAGE 11

LIST OF FIGURES Figure page 1-1 Total damages due to hurricane events by decade normalized to 2005 US currency by population vs. housing units presented in a paper by Pielke et al. (2008) ....................21 1-2 Structural damage due to loss of roof sheathing below design wind speeds, design wind speed 140 mph actual wind speed 120 mph Ono Island, AL (FEMA et al. 2005b) ................................................................................................................................21 2-1 Resulting pressure loads from wi nd loading of residential structures ...............................39 2-2 Structural damage due to loss of roof sheathing below design wind speeds, design wind speed 140 mph actual wind speed 120 mph Ono Island, AL (FEMA et al. 2005b) ................................................................................................................................39 2-3 Typical residential structure construction with wood roof sheathing attached to metal plate trusses or rafter, which are attached to walls with either metal straps or toe-nails ...40 2-4 Fastener schedule and construction of roof panel designed by pre-1994 building code ...40 2-5 Loss of roof sheathing at overhang locations, Hurricane Katrina (130 mph) long beach Mississippi (FEMA et al. 2006) ..............................................................................41 2-6 Tributary areas for individual fastener installed in a 6 in. / 12 in. spaced roof sheathing panel ...................................................................................................................41 2-7 Nail Extraction Device Developed at Clemson University by Sutt (2000) .......................45 3-1 UF-WRSUT static pressure trace. (5 psf initial static pressure is included in the trace) ..................................................................................................................................53 3-2 Gable roof model A) installed in Wind Tunnel and B) close-up of model ........................53 3-3 Pressure tap location A) 1: 50 model scale (62 ft by 35 ft full scale) and B) full scale representation of panel .......................................................................................................54 3-4 Comparison of static vs. dynamic pressure traces .............................................................54 3-5 Diagram of instantaneous failure pressure for panel tested statically ................................55 3-6 Sample of target and actual pressure tim e-history for panels tested with the PLA system A) full time-history and B) close up of instantaneous peak and deviation from target pressure ....................................................................................................................56 4-1 Summary of panel series tested in the static phase of this study .......................................66 4-2 Summary of panel series tested in th e static vs. dynamic phase of this study ...................66 11

PAGE 12

4-3 Laboratory fabricated panel constructi on, 6 in. / 12 in fastener schedule shown ..............67 4-4 Laboratory panel fastening schedule (6 in. / 12 in., 6 in. / 8 in. and 6 in. / 6 in.) ..............67 4-5 Static ccSPF retrofit constr uction of A) fillet sample set and B) full 3 in. sample set ......68 4-6 Panel construction for static vs. dynamic testing with 2 mils. thick plastic sheet placed between sheathing and wood framing member during fabrication ........................68 4-7 Locations of residential struct ures where panels were harvested ......................................71 4-8 Pictures of harvested panel removal, A) Expose framing members, B) Cut framing members and C) Roof after panels removed ......................................................................71 4-9 Comparison of retrofit measures A-1 and A-2 ..................................................................72 4-10 ccSPF retrofit of an existing residential structure in Port Orange, FL ...............................72 4-11 Laboratory fabricated panel sp ecific gravity sample locations ..........................................73 4-12 Harvested panel specific gravity and moisture content sample locations ..........................74 4-13 Panel installed in pressure ch amber with negative pressure setup ....................................74 4-14 Panel installed in pressure ch amber with positive pressure setup .....................................75 4-15 Manual control of pressure chamber (A ga te valve control and (B vacuum pump in series ..................................................................................................................................75 4-16 Pressure Load Actuator (PLA) ...........................................................................................76 4-17 Comparison of dynamic target vs. actual chamber pressure used with the PLA ...............76 5-1 Mean failure pressures of laborator y fabricated panels tested statically ...........................97 5-2 Summary of mean panel failure (psf) vs. calculated fastener failure (lbs) for A) 8d ring shank nails and B) 8d smooth shank nail s attached at 6 in. / 12 in., 6 in. / 8 in. and 6 in. / 6 in. schedules ...................................................................................................99 5-3 Sample of failure mode and location calcu lations from Statically tested 8d smooth shank nails at 6 in. /12 in. ................................................................................................100 5-4 Distribution of Laborator y Fabricated Statically Te sted Panels Dominated by Withdrawal or Pull-through Failure modes A) 6 in. / 12 in. spacing, B) ccSPF retrofit of 6 in. / 12 spacing, C) 6 in. / 8 i n. spacing and D) 6 in. / 6 in. spacing .........................101 5-5 Comparison of statically tested harves ted and new panels full results of failure pressure ............................................................................................................................103 12

PAGE 13

5-6 Comparison of mean uplift capacities with mean calculated fastener failure loads for harvested panels attached w ith a) staples and b) nails .....................................................104 5-7 Comparison of statically tested harveste d panels in existing vs. retrofit conditions separated by individual source houses .............................................................................106 5-8 Probability distribution of Specific Gravity of Panels Tested in Phase 2 for (A New Panels and (B Harvested Panels 15 years old ..................................................................107 5-9 Comparison of statically vs. dynamically loaded laboratory fabricated panels attached with 2 in. long 6d smooth shank nail s at 6 in. / 12 in. (existing and retrofit) and 6 in. / 6 in. .................................................................................................................108 5-10 Comparison of statically vs. dynamically loaded panels harvested from existing construction attached with 1.5 in. long staple s at 4 in. / 4 in. (e xisting and retrofit) .......109 6-1 Comparison of lab static mean and 5% exclusion failure pressures to studies by Kallem (1997) w/ plywood, IHRC (2004) w/ plywood and Murphy et al. (1996) w/ OSB (IHRC 8d SS @ 6/12 was not a normal distribution so 5% exclusion value is not provided) ....................................................................................................................124 B-1 Failure mode / location for panels fastened with 6d smooth shank nail spaced at 6 in. / 12 in. tested statically .....................................................................................................148 B-2 Failure mode / location for panels fastened with 6d smooth shank nail spaced at 6 in. / 12 in., 6 in. / 12 in. retrofitted with ret. A-2, and 6 in. / 6 in. tested statically ...............149 B-3 Failure mode / location for panels fastened with 6d smooth shank nail spaced at 6 in. / 12 in., 6 in. / 12 in. retrofitted with ret. A-2, and 6 in. / 6 in. tested dynamically .........150 B-4 Failure mode / location for panels fastened with 8d smooth shank nail spaced at 6 in. / 12 in. tested statically .....................................................................................................151 B-5 Failure mode / location for panels fastened with 8d smooth shank nails spaced at 6 in. / 8 in. tested statically .................................................................................................152 B-6 Failure mode / location for panels fasten ed with 8d smooth shank spaced at 6 in. / 6 in. tested statically ............................................................................................................153 B-7 Failure mode / location for panels fastened with 8d ring shank nails spaced at 6 in. / 12 in. tested statically .......................................................................................................154 B-8 Failure mode / location for panels fastened with 8d ring shank nails spaced at 6 in. / 12 in. retrofitted with 1-1/2 in. fillet of ccSPF tested statically .......................................155 B-9 Failure mode / location for panels fastened with 8d ring shank nails spaced at 6 in. / 12 in. retrofitted with full 3 in. thick layer of ccSPF tested statically .............................156 13

PAGE 14

B-10 Failure mode / location of panels fastened with 8d ring shank nails spaced at 6 in. / 6 in. tested statically ............................................................................................................157 B-11 Failure mode / location for panels fastened with 8d ring shank nails spaced at 6 in. / 8 in. tested statically .........................................................................................................158 B-12 Failure mode / location for panels fastened with 8d ring shank nails spaced at 6 in. / 12 in. tested statically .......................................................................................................159 C-1 Summary of missing fasteners from stat ic vs. dynamic testing, Debary #1 series ..........160 E-1 Summary of pressure time-history for static 6d SS at 6/12 panel (6d SS-38) .................164 E-2 Summary of pressure time-history for static 6d SS at 6/12 panel (6d SS-40) .................164 E-3 Summary of pressure time-history for static 6d SS at 6/12 panel (6d SS-41) .................165 E-4 Summary of pressure time-history for static 6d SS at 6/12 panel (6d SS-42) .................165 E-5 Summary of pressure time-history for static 6d SS at 6/12 panel (6d SS-39) .................166 E-6 Summary of pressure time-histories fo r dynamic 6d SS at 6/12 panel (6d SS-43) A) full time-history and B) close up of failure ......................................................................168 E-7 Summary of pressure time-histories fo r dynamic 6d SS at 6/12 panel (6d SS-44) A) full time-history and B) close up of failure ......................................................................169 E-8 Summary of pressure time-histories fo r dynamic 6d SS at 6/12 panel (6D SS-45) A) full time-history and B) close up of failure ......................................................................170 E-9 Summary of pressure time-histories fo r dynamic 6d SS at 6/12 panel (6d SS-46) full time-history and B) close up of failure ............................................................................171 E-10 Summary of pressure time-histories fo r dynamic 6d SS at 6/12 panel (6d SS-47) A) full time-history and B) close up of failure ......................................................................172 E-11 Summary of pressure time-history for static 6d SS at 6/6 panel (6d SS-53) ...................173 E-12 Summary of pressure time-history for static 6d SS at 6/6 panel (6d SS-52) ...................173 E-13 Summary of pressure time-history for static 6d SS at 6/6 panel (6d SS-56) ...................174 E-14 Summary of pressure time-history for static 6d SS at 6/6 panel (6d SS-55) ...................174 E-15 Summary of pressure time-history for static 6d SS at 6/6 panel (6d SS-54) ...................175 E-16 Summary of pressure time-histories fo r dynamic 6d SS at 6/6 panel (6d SS-57) A) full time-history and B) close up of failure ......................................................................176 14

PAGE 15

E-17 Summary of pressure time-histories fo r dynamic 6d SS at 6/6 panel (6d SS-58) A) full time-history and B) close up of failure ......................................................................177 E-18 Summary of pressure time-histories fo r dynamic 6d SS at 6/6 panel (6d SS-59) A) full time-history and B) close up of failure ......................................................................178 E-19 Summary of pressure time-histories fo r dynamic 6d SS at 6/6 panel (6d SS-60) A) full time-history and B) close up of failure ......................................................................179 E-20 Summary of pressure time-histories fo r dynamic 6d SS at 6/6 panel (6d SS-61) A) full time-history and B) close up of failure ......................................................................180 E-21 Summary of pressure time-history for static 6d SS at 6/12 ret. A-2 panel (6d SS-48) ....181 E-22 Summary of pressure time-history for static 6d SS at 6/12 Ret. A-2 panel (6d SS-49) ..181 E-23 Summary of pressure time-history for dynamic 6d SS at 6/12 ret. A-2 panel (6d SS50) A) full time-history and B) close up of failure ..........................................................183 E-24 Summary of pressure time-history for dynamic 6d SS at 6/12 ret. A-2 panel (6d SS51) A) full time-history and B) close up of failure ..........................................................184 E-25 Summary of pressure time-history for st atic 1.5 in. Staple at 4/4 (1.5 in. Staple-7) .......185 E-26 Summary of pressure time-history for st atic 1.5 in. Staple at 4/4 (1.5 in. Staple-8) .......185 E-27 Summary of pressure time-history for dyna mic 1.5 in. Staple at 4/4 (1.5 in Staple-9) A) full time-history and B) close up of failure ................................................................187 E-28 Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 (1.5 in. Staple10) A) full time-history and B) close up of failure ..........................................................188 E-29 Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 (1.5 in. Staple11) A) full time-history and B) close up of failure ..........................................................189 E-30 Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 (1.5 in. Staple12) A) full time-history and B) close up of failure ..........................................................190 E-31 Summary of pressure time-history for sta tic 1.5 in. Staple at 4/ 4 with Ret. A-2 (1.5 in. Staple-13) ....................................................................................................................191 E-32 Summary of pressure time-history for sta tic 1.5 in. Staple at 4/ 4 with Ret. A-2 (1.5 in. Staple-14) ....................................................................................................................191 E-33 Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 with Ret. A-2 (1.5 in. Staple-15) A) full time-history and B) close up of failure ..................................193 E-34 Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 with Ret. A-2 (1.5 in. Staple-16) A) full time-history and B) close up of failure ..................................194 15

PAGE 16

16 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering DEVELOPMENT OF TIME-VARYING WIND UPLIFT TEST PROTOCOLS FOR RESIDENTIAL WOOD ROOF SHEATHING PANELS By Kenneth M. Hill December 2009 Chair: David O. Prevatt Major: Civil Engineering Following the damage caused by Hurricane Andr ew (1992) several re search institutions and universities made an effort to improve the performance of reside ntial wood roof sheathing panels in extreme wind events. The outcome of this research was predicted panel uplift resistance based on observed performance under vary ing static test methods. Changes in panel construction have resulted in improved wind uplift resistance, however failures are still observed below design level wind speeds. It is the hypothe sis of this investigatio n that some residential wood roof sheathing panel failure s are due to current test methods overestimating panel resistance. Previous wood roof sheathing panel uplift tests have us ed uniform static pressure on newly constructed panels. However aged w ood roof sheathing panels are subjected to temporally varying pressure loading in the field. As part of this res earch standardized test protocols are developed for static and temporally varying (dynamic) pressure loading. Panels fabricated in the lab and harves ted from existing construction are tested with static and dynamic pressure. Over 170 panels were tested static ally, 38 of which were harvested from existing construction, and 34 panels were tested static ally vs. dynamically. It is found that dynamic loading reduces uplift ca pacity of wood roof sheathing panels approximately 20%, and that age or weathering do not appear to eff ect the uplift resistance of panels.

PAGE 17

CHAPTER 1 INTRODUCTION Motivation The damaging effects of hurricanes making la ndfall along the eastern seaboard of the United States have been well documented in the past 20 years, beginning with Hurricane Hugo (1989). Hurricanes annually cause billions of do llars in property loss in coastal regions of the south-east United States (van de Lindt et al. 2007). Examples of hurricane damage were observed in Hurricane Hugo in 1989, Hurricane Andrew in 1992 and most recently during the eight land falling hurricanes of 2004 and 2005 se asons. Approximately 160 billion dollars (normalized to 2005) in damages occurred in the 2004-05 hurricanes from Hurricanes Jeanne, Frances, Ivan, Charley, Rita, Wilm a and Katrina (Pielke et al. 2008) This significant damage highlights a need to better unders tand the effects of Hurricane ev ents on society. Figure 1-1 summarizes the total damages due to wind events by decade from 1900-2005 normalized to 2005 US dollar by Pielke et al. (2008), which suggests that this level of impact is not uncommon. Wind damage to light-frame wood structural systems is responsible for a significant proportion of the observed damage from hurricane ev ents. Sparks (1991b) observed that extreme winds in Hurricane Hugo were responsible for 60% of the insured loss es. Further, a large proportion of the damage caused to light-framed w ood structures is concentrated in damage to the roof systems. Baskaran and Dutt (1997) es timated that 95% of all monetary losses from Hurricanes Iniki and Andrew in 1992 resulted fro m the failure of roof system materials. The construction of wood roofs typically consis ts of wood sheathing nailed to wood trusses or rafters that are attached to the walls of th e building. Wood roof sheathing panels provide a structural diaphragm which resists lateral load s and encloses the structure when used in conjunction with a roof covering system. Post disaster investigations have found that roof 17

PAGE 18

sheathing wind uplift failures lead to structural damage, due to loss of lateral stiffness (FEMA 1993), and extensive water damage (FEMA et al. 2005a). The damage caused by wood roof sheathing failure presents considerable monetary risk to the state of Florida considering that approximately $1.5 trillion in buildin g stock are exposed to hurricane events (Pinelli et al. 2004). Typically, the wind uplift tests used for commer cial roofs and roofing systems (i.e. ASTM E15-92, UL 1897 and FM 4470) are based on physical tests using uniform static pressure applied to a sample until the specimen fails. However no such test exists for residential roof structures. The current prescriptive standa rds for wind uplift resistance of wood roof construction in the International Building Code (ICC 2006) were developed from experimental research conducted by Cunningham (1993), Schiff et al. (1994) and others. These tests which are basis for current design against wind uplift, which were developed five to seventeen years ago, lack the dynamic characteristics necessary to replicate actual wind forces generated during a hurricane. It has been observed from post hurricane inve stigations that roof systems fail at wind speeds below design level. For example FEMA (2006) reported roof damage to homes in Alabama and Mississippi during Hurricane Katr ina occurred in 120 mph winds although the design wind speeds for the areas we re 140 mph (Figure 1-2). Reasons suggested for this include poor construction quality, environmental effect s which accelerate loss of strength, natural variability in wood properties and under-estimation of the actual wind forces that occur during hurricanes. Another reason for this disconnect between in-field and predicted performance may be related to limitations of curre nt wind uplift test protocols th at predict uplift performance. Baskaran and Dutt (1997) demonstrated that dynamic loads along individual dowel-type fasteners in wood reduced their w ithdrawal capacity by 10 to 30%. This effect is not currently accounted for in panel uplift testing and it may re sult in overly conservati ve prediction of wind 18

PAGE 19

uplift capacity of wood roof sheathing. Similarl y the in service conditions of residential wood roof sheathing may reduce the uplif t resistance. Studies have shown that moisture and high humidity can cause 38% to 75% reduction in the strength of nails and staples installed in sheathing to wood framing connection (Chow et al. 1990; Feldborg 1989; Pye 1995). Judge and Reinhold (2002) conducted one of the first studies to inve stigate the wind uplift capacity of roof sheathing installed in residential structures. While their limited study of twelve harvested wood roof sheathing panels from nine homes in Horry County, SC is not able suggest any conclusions it provides preced ence for uplift testing of harvested panels conducted as part of this investigation. It is the hypothesis of this investigation that some premature failures of roof sheathing panels occur because current wind uplift test methods overestimate the resistance of wood roof sheathing panels to real wind lo ading. A second hypothesis is that wood roof sheathing panel uplift resistance is reduced due to in servic e environmental effects which may degrade wood material properties over the typical life (0-50 years) of a residential roof. Objectives In order to test these hypotheses, wind uplift pressure testing will be conducted to Evaluate the uplift resistance of new vs. aged (harvested) wood roof sheathing panels in order to assess the effect of in-service conditions such as moisture cycles and, Compare the wind uplift resistance of wood roof sheathing panels tested statically vs. dynamically in order to assess the effects of dynamic loading. Organization of Thesis Chapter 2 presents a literatu re review of the current knowledge of wind uplift pressure testing for residential roofs, construction and vul nerability of wood residential buildings. The structural load path associat ed with wind events in typical residential wood roof sheathing system construction and failure mechanisms of wood roof sheathing systems are established. 19

PAGE 20

Standardized commercial wood roof sheathing test protocols are reviewed to provide basis for development of the wind uplift test pr otocol to be used in testing. A new standardized wind uplift test protocol is developed in Chapter 3 to evaluate the wind uplift resistance of wood roof sheathing panels. The protocol include a static pressure test method that is used to compare re sults with previously reported studies in the literature, and a dynamic pressure test method that provides more realistic simulation of actual wind forces. Through wind uplift testing this investigation will a) evalua te the uplift pe rformance of new panels and aged wood roof sheathing panels harvested from existing re sidential structures, b) compare the failure mechanisms and failure pressures of roof panels tested using new static vs. dynamic pressure test protocol s, c) evaluate the effect of retrofit on existing roof wind uplift capacity and d) evaluate the effect of specific gr avity on failure pressure of wood roof sheathing. Descriptions of sample construction/harvesting a nd test equipment are presented in Chapter 4. Chapter 5 presents the results a nd analysis of wind uplift test s performed. A discussion of findings relating to the objectives is then pres ented in Chapter 6. Chapter 7 makes conclusions based on results and provides recommendations fo r future use in a sta ndardized test method. 20

PAGE 21

21 0 50,000 100,000 150,000 200,000 250,0001900-1905 19 0 6-1 9 15 1 9 1 6-1 9 25 1 9 2 6-1 93 5 1 9 36 194 5 1946 195 5 1956-1965 1966-1 9 75 19 7 6-1 9 85 1 9 8 6-1 9 95 1 9 96 -2 00 5Year Ran g eMillions of US Dollars (2005) Normalized by Population Normalized by Housing Units Figure 1-1. Total damages due to hurricane events by decade normalized to 2005 US currency by population vs. housing units presented in a paper by Pielke et al. (2008) Figure 1-2. Structural damage due to loss of roof sheathing below design wind speeds, design wind speed 140 mph actual wind speed 120 mph Ono Island, AL (FEMA et al. 2005b)

PAGE 22

CHAPTER 2 LITERATURE REVIEW Wind Loads on Low-Rise Buildings Natural winds exhibit considerable and rapid fl uctuation in its veloci ty and direction, and this fluctuation (or turbulence) of the wind produces pressures on bluff bodies in its path that vary both spatially and in time. Wind tunnel stud ies have determined that the wind pressure vary over the building surface, with ge nerally positive pressures being created on the windward walls and negative pressures on the leeward and side wa lls of the structure. In low-rise buildings which have sloped roof structur es, fairly high negative pressu res can be created in flow separation zones (at roof corners, ridges or edges) (Holmes 2001), see Figure 2-1. These negative pressures or suctions can be several times greater th an the positive pressures on the windward walls. These suction forces are responsi ble for the wind damage to roof structures of low-rise structures observed afte r recent hurricanes. To estimate the pressure in the regions of flow separation Bernoullis e quation in the form (eq. 2-1), 2 01 U U Cp (2-1) is used where Cp is a non-dimensional pressure coeffici ent, U is the wind speed at the location being measured and U0 is the reference wind speed. Stric tly speaking Bernoullis equation is not applicable in separated flow regions but it has been found that reasonab le predictions can be found. The relationship of wind ve locity to the resulting pressure force variations is described by the aerodynamic admittance function. Wind speeds are highly variable and the result ing pressure forces depend greatly on the geometry of the structure making it impossible to solve for loads deterministically. Therefore stochastic methods are employed to envelope results to account for all possible combinations of 22

PAGE 23

variables in determining design pressure loads. This enveloping process is conducted by analysis of both wind tunnel testing of structural models of different roof shapes and expected wind speeds for a particular location. Wind t unnel tests result in non-dimensional Cp pressure coefficients for a particular roof shape to be scaled by the design wind speed. Design wind speeds are developed by computer modeling to de termine the probability of a wind speed being experienced at any location in the US. These wi nd speeds are defined at 10 meters (33 ft) above ground level in open exposure ,and must be adjusted to both the mean roof height and exposure condition to be applied to wind tunnel determined pressure coefficients. Th is stochastic analysis is summarized ASCE 7 (2006), which is the cu rrent reference manual used to determine wind uplift loading on low-rise residential structures. The turbulent nature of wind loading results in gusts which can be significantly higher than mean wind speeds. These gusts are incorporated into design loads by categorizing the surface area a connection is design ed to resist pressure loading over or effective wind area. The smaller the effective wind area subjected to a gust the larger the effect that gust has on the connection, because the gust load is not able to be distributed. This is accounted for in ASCE 7 by two categories main wind force resisting systems (MWFRS) and components and cladding (C&C), which are further broken down by actua l effective wind areas. Design wind speeds are developed by computer modeling to determine th e probability of a wind speed being experienced at any location in the US. These wind speeds are defined at 10 meters (33 ft) above ground level and must be adjusted to the mean roof height to be applied to wind tunnel determined pressure coefficients. Pressure loads are summarized in figures by cross-referencing effective wind areas by roof zones. The zones on a roof are determined from the analysis of wind tunnel studies and include field, edge, corner and overhang zone s, where corner and overhang zones have the 23

PAGE 24

highest load. Described above ar e the external pressure coeffi cients which are a result of building geometry, as pressure loading is a result of difference in pressure it is also necessary to know the internal pressure to determine a load. The internal pressure during wind loading is dependant on the openings in a structure through which wind can affect internal pressures. ASCE 7 accounts for this by using three categories open, partially enclosed and enclosed. ASCE 7 design pressures are determined from roof ge ometry, location, exposure, mean roof height, effective wind area and openings. Response to Earlier Historical Hurricane Damage Following Hurricane Hugo (in 1989) which cau sed $7 billion dollars in losses and Hurricane Andrew (in 1992) which caused $26 billi on dollars in losses, several organizations initiated wind uplift testing on wood roof structur al components. The results from those tests have informed many of the wind load design requir ements for residential roof construction. The research was initiated after num erous observations of premature roof failures indicated the vulnerability of wood roofs on singlefamily residential structures. Sparks (1991a) estimated that approximately 60% of the total damage from Hurricane Hugo occurred to residential buildings, the majority of which occurred to the roof structure. An analysis of insurance claims from Hugo found that for nearly all case-studies the vast majority (95%) of losses were a direct result of failure of roof materials (Amirkhanian et al. 1994). Keith and Rose (1992) observed that 24% of wood-fr amed residential homes in their study area (Miami, FL) lost at least one roof sheathing pa nel during Hurricane Andrew. The consequences of roof sheathing failure is significant for two reasons: 1) as sheat hing fails the roof structure can lose its structural integrity re sulting in structural failure a nd collapse and 2) sheathing loss creates openings in the roof that allow water in trusion that causes severe water damage to interior partitions and bu ilding contents (Cook Jr 1991) 24

PAGE 25

The connection of wood sheathing to wood fr aming members was identified as a key factor in wood roof system failure. Expe rimental studies were conducted by Cunningham (1993), Schiff et al. (1994) and others to investig ate and improve this connection. These studies used roof panels installed on a rigid pressure chamber where the chamber pressure is increased until failure occurs. Failure defined as the first failure of a nailed connection or fracture of the wood panel or wood framing member Test results are reported as the mean failure pressure (psf) and the design (or allowable) wind uplift capacity is determined by dividing the mean failure pressure by a factor of safety (i.e. 2.0) or using the 5% exclusion value of the data. A full treatment of experimental studies into wind uplift loading of re sidential wood roof sheathing is provided in a later section. Recent Hurricane Damage to the United States Since 1900 the monetary loss associated w ith hurricanes in the United States is approximately $10 billion normalized to 2005 USD (Pielke et al. 2008), which suggests consistent potential for loss. Most recently the 2004 and 2005 storms have totaled over $150 billion dollars in damages (Pie lke et al. 2008). Hurricane Katr ina was the largest contributor causing over $125 billion of damage followed by Charle y ($15 billion) and Ivan ($14 billion). Wind damage during hurricane events has hist orically shown to be responsible for the majority of losses. For example Baskaran and Dutt (1997) estimated that 95% of losses from Hurricanes Iniki and Andrew in 1992 resulted fro m the failure of roof system materials under wind loading. However significant losses can occur from flood damage, which is caused by an increase in water levels which is referred to as a storm surge. Hurricane Katrina is an example where flood damage has far exceeded wind damage The damaging effects of wind loading can reach further inland than storm surge limits and ther efore present a risk to larger areas than flood damage. 25

PAGE 26

Wood roof sheathing panels fail in hurricane wi nd loading by removal or separation of the sheathing from the wood framing members. Failure occurs when individual fastener connections fail or to a lesser extent when wood components fracture. The co mplete removal of a panel is not necessary to consider a panel to have failed, as water intrusion will in crease even with partial panel removal as observed in Hurricane Hugo (Mur den 1991) and elsewhere. Nail withdrawal from the framing member is the pr imary mode of failure observed in roof sheathing panels due to wind loading, with nail pull-th rough and fracture of the sheathi ng and/or wood framing observed to a lesser extent. Nail pull-through is a punchi ng failure of the sheathing where the nail shank remains embedded in the wood framing wh ile the sheathing is pulled away. Hurricane Charley a category 4 hurricane on the saffir-simpson scale struck Punta Gorda, FL on August 13th 2004 (FEMA et al. 2005a). The stor m tracked north-west across Florida exiting in the Daytona Beach area approxima tely 10 hours later causing total flood and wind damages of approximately $16 billion. From a po st disaster investigation it was observed that the majority of damages to residential structures occurred in homes cons tructed prior to code changes in 1994. This suggests the reduced fasten er spacing and increased size have resulted in improved performance as compared to homes built after 1994. Hurricane Ivan a category 3 hurricane on the saffir-simpson scale struck Alabama and Florida coastline on September 19th, 2004 (FEMA et al. 2005a). The storm tracked north west over the next week eventually exiting from the Delaware, Maryland, Virginia peninsula causing approximately $15 billion in losses. Again it was observed that the majority of damage occurred to construction built after 1994. It was observe d that wood sheathing failures occurred in wind speeds below design level winds. For example r oof sheathing failure was observed in 120 mph winds (estimated from FEMA HAZUS maps) at Ono Island, Alabama where current design wind 26

PAGE 27

speed is 140 mph (Figure 2-2). Damage to roof sheathing was also observed in newer construction below the design wind speeds. Hurricane Katrina struck sout hern Florida as a category 1 hurricane on August 25, 2005 then after strengthening in the Gu lf of Mexico struck the Louisi ana and Mississippi coasts as a category 3 hurricane (FEMA et al. 2006). As mentioned before the majority of the total estimated $125 billion in losses was caused by flooding. The majority of wind damage was observed in pre-1994 construction. Wind damage was also observed in new construction below design level wind speeds (van de Lindt et al. 2007). The extensive monetary losses experienced from recent hurricanes have put the affected communities under considerable strain to cope with recovery. Whats more the homes destroyed by either the resulting structural or water intrusion damage cause disruptions to thousands of residents. It is estimated that Hurricane Ka trina destroyed over 300,000 single family homes in Louisiana and Mississippi displacing 450,000 t housand residents (FEMA et al. 2006). Construction of Residential Houses Low-rise residential wood roof structures are the focus of this investigation and the load path associated with resisting wind uplift pressure loads defined above is discussed in this section. The typical residential structure is composed of a roof system which transfers wind load to the wall system and then on to the foundation (Figure 2-3). The stru ctural roof sheathing system is composed of wood sheathing attached to wood framing members. Roof sheathing is the part of the building envelope which provides weather resisting barrier to the structures and in many cases its lateral stiffness. Wind loading on roof sheathing panels ar e resisted by individual fasteners (nails or staples), wh ich are installed into the wood fr aming. Roof structures can consist of individual rafters or more commonly metal-plate connected trusses. 27

PAGE 28

These components are fastened to the walls us ing toe-nailed connecti on in older homes or using metal strap connections in contemporary c onstruction. Typical sp acing of wood framing roof members is 24 in. on center. Oriented strand board (OSB) sheathing have been used as roof sheathing the 1980s (Feng et al. 2008). OSB is a composite material consisting of small wood chips glued together and oriented in random directions. Plywood is also a composite material composed of layers of thin wood sheets adhered together in alternati ng orthogonal directions. Current International Building Code specifies the use of in. thick plywood sheathing (ICC 2006). The code also specifies the minimu m fastener sizes and spacing to meet design wind uplift capacity for residen tial wood roofs. Fastening sche dule written as x in. / y in., defines the spacing of fasteners installed along the exterior panel edge the nail spacings on interior panel members. The minimum code requirements for installing wood roof sheathing systems for building construction before 1994 were 2 in. long nails with 0.113 in. diameter (6d Common) at 6 in. / 12 in, see Figure 2-4. Appr oximately 86% of the current building stock are constructed before building code modifications were introduced in 1994 to improve the wind resistance of structures (US Census Bureau 2003). Wood roof sheathing systems built after 1994 in high wind zones in Florida are required to use a 2-1/2 in. long 0.113 in. diameter annularly threaded nails (8d ring shank) at 6 in. / 6 in or 4 in. / 4 in. at gable ends (ICC 2004). Many residential roofs overhang the exterior walls along the eaves and at gable ends. These overhangs are typically 12 to 18 in. long. Light soffit materials are installed below the cantilever wood members and the soffits also have ve nts that allow air flow into the attics. This overhang allows pressure to act on both faces of the sheathing extending into the overhang which has been observed to be the location of increased failures (Figure 2-5). 28

PAGE 29

Wind Uplift Behavior of Resident ial Wood Roof Sheathing Panels Design of roof sheathing systems currently re lies on full-scale system testing to measure roof wind uplift capacities. Residential wood roof systems have no industry-recognized test protocols for determining their wind uplift capacities. Currently accepted loading methods use static pressures applied uniformly to the full-scale test specimen. It appears that residential wood roof sheathing design is based on a limited numb er test studies conducted in the 1990s in response to extensive damage associated with roof materi als observed in Hurricanes Hugo (1989), Andrew (1992), & Iniki (1992). The majority of the published work from 1994 through 2008 on residential wood wind uplift testing was carried at re search institutions; Clemson Univ ersity (Jones 1998; Mizzell 1994; Murphy et al. 1996; Sutt 2000), FIU (IHRC 2004), while other testing was performed at the APA (Cunningham 1993), NAHB (NAHB 2003) and Stanle y-Bostich (Reinhold et al. 2003). The wind uplift capacity or stru ctural resistance of wood roof panels can only be determined by wind uplift testing. Currently there are no analytica lly based methods for predicting the wind uplift capacity of wood roof sheathing panels. The current design method is based on static uplift pressure tests to determine failure capacities of specimens. Fastener load is calculated by the product of the failur e pressure (psf) and the tributary area (ft2), which results in units of pounds. A sample of tributary areas for a panel fastened with a 6 in. / 12 in. spacing can be seen in Figure 2-6. Variability in Uplift Capacity Testing The lack of a standard resident ial roof sheathing uplift test pr otocol is a possible cause for the variability in pressure test results. The natural va riation of wood streng th, moisture content and the wide variety of fasters wh ich may also contribute to observed variability in results, which range from 26 psf for a 6d smooth shank at 6 in. / 12 in. to 397 psf for a 8d ring shank nail at 6 29

PAGE 30

in. / 6in. Tests were performed using pneumatic ai r pressure or air bags and the failure pressure was defined as the pressure at which the first sign of structural distress or failure occurred. The results were reported as the ultimate failure pres sure from which nail type and fastener schedule design values were determined by dividing the ultim ate pressure by a factor of safety (typically 2.0) or 5% exclusion values (Kallem 1997). Table 2-1 summarizes results of studies into wind uplift failure capacities conducted from 1993 to 2004. It can be seen that the mean failu re pressure increases w ith nail size and reduced fastener spacing, and there appe ars to be similar performance of wood roof panels fabricated using OSB or plywood sheathing. Comparison of these results are somewhat misleading because all of the differences in individual test methods used to obtain failure pressures such th at reported test s lack a common basis to make comparisons. NAHB (2003) test s were conducted using air bags, while the remaining test used actual air pressure. The use of air-bags may result in different transfer of load to the sheathing that pressure loading does. Cunningham (1993) tested only one specimen per roof configuration. A further difference was observed even within institutions, such as Mizzell (1994) used a 40 psf/min rate increase vers us the near-instantaneous loading that Kallem (1997) and Sutt (2000) relied upon. Mean failure pressures vary significantly be tween studies. Panel a ttached with 6d smooth shank fasteners at a 6 in. / 12 in. schedule ha ve mean failure pressures of 55 psf (15/32 plywood, Cunningham 1993), 65 psf (5/8 plywood, Cunningham 1993)) and 33 psf (plywood, Kallem 1997). Panels attach with 8d smooth sha nk fasteners attached at 6 in. / 12 in. had mean failure capacities of 79 psf (15/32 plywood, Sutt 2000), 67 ps f (7/16 OSB, Sutt 2000) and 110 psf (1/2 plywood, IHRC 2004). 30

PAGE 31

All told 22 wind uplift failure pressure studies f ound involved a total of 220 test specimens and the author was not able to find any other studies available in literature. Despite possible changes to materials that have been made over past 16 years the results from the 1990s and early 2000s are the basis of recent prob abilistic studies and reliabilit y models (i.e. Li and Ellingwood 2009 etc.). The current minimu m fastener schedules in the Florida Building Code (ICC 2007) appear to still be based on these early test resu lts. In successive versions of the Florida Building Code fastener schedules strengthened and nail di mensions increased in order to increase the wind uplift capacity of roofs. However, given the wide variation in methodologies used and spread of failure capacities observed it is difficult to have confidence in the test values used to develop these prescriptive guidelines. Retrofit Measures Retrofit of wood roof sheathing has been inve stigated at Clemson University for spray applied adhesives (Jones 1998). Samples were co nstructed with 4 ft by 8 ft by 15/32 in. plywood or 19/32 in. OSB attached to 2 in. by 4 in. framin g members with 6d smooth shank nails at 6 in. / 12 in. Spray applied adhesive was applied in seve ral combinations of full-bead and partial-bead between sheathing and framing members. Table 2-2 pr esents selected results from this study. It was found that significant improvement can be achieved through this approach. In Field Testing Judge and Reinhold (2002) reports on the dest ructive testing of several homes in Horray County, SC. Ten wood roof panels were test ed and are presented in Table 2-3. Panel construction tested was connected with staples, 8d sinker nails, and 16d nails. Sheathing was 0.465 in. plywood, 7/16 in. OSB and 8 in. wide w ood planking. Failure pressures ranged from 110 psf for staples spaced 3 in. to 5 in. apart to an extremely high 450 psf failure pressure for the planking. 31

PAGE 32

Building Codes Building codes specify the minimum requirement s for construction of buildings within a jurisdiction. Their purpose is to protect li fe safety and provide a minimum standard for performance to protect public/p rivate property. The design and construction of low-rise residential structures in Florida is governed by the Florida Building Code (ICC 2007) which has adopted ASCE 7 for reference as the minimum design loads of buildings. Currently the FBC specifies the construction of re sidential wood roof sheathing in high wind zones to be in. plywood attached with 8d ring shank nails spaced at 6 in. Prior to 1994 equivalent provisions specified the connection of wood roof sheathing to be 6d smooth shank nails spaced at 6 in. / 12 in. Table 2-4 summarizes residential wood roof sheathing attachment in Florida from 1988 to present. Wind Uplift Resistance of Dowel Type Fastener Connections in Wood In previous sections the failure modes of roof sheathing panels are defined as the individual fastener connections failing in withdrawal or pull-through. This section reviews the current state of knowledge on withdrawal strength of roof fastener connections. The National Design Specifications for Wood Construction (NDS) (AF&PA 2005) requires that withdrawal capacities of all nail types & sizes which shall be used to resist uplift in roof sheathing be experimentally determined per ASTM D-1761 Sta ndard Test Methods for Mechanical Fasteners in Wood (ASTM 2006a). In addition the na il pull-through capacity of wood based panel products is considered in the design of w ood roof sheathing, and ASTM D-1037 Evaluating Properties of Wood-Based Fiber and Particle Panel Materials (ASTM 2006c) is specified. ASTM D-1761 Protocol ASTM D-1761 test protocol S tandard Test Methods for Mechanical Fasteners in Wood determines the withdrawal strength of nails under a constant withdrawal rate of 0.1 in. / minute, 32

PAGE 33

which is applied to failure. The test specimens are wood prisms with th e nail driven at right angles to the face. Load is applied directly to the nail head wood prism is clamped to the testing machine. A minimum of 10 samples are tested wi th expected range of coefficient of variation (COV) between 15 to 30%. Larger number of samp les is needed to obtain COVs in the 5 to 10% range due to the inherent variability in wood. Effect of Rate of Loading on Withdrawal Resistance Kallem (1997) conducted an investigation to dete rmine the effect of nail withdrawal rate on the nail ultimate failure capacity. The test procedure used by Kallem was based on ASTM 1761 protocol and he compared stre ngth versus six nail withdrawal rate; namely 0.1, 0.5, 1.0, 5.0, 10.0 in. per minute and near instantaneous~1/8 to a 1/10 of a second. Results were then analyzed for each 10 sample series. Samples we re constructed with 2-1/2 in. 8d smooth shank nails (0.131 in. diameter) installed in spruce pine fir. Mean wit hdrawal capacities ranged from 153 to 165 lbs and COVs ranged from 21 to 34%. It found that for the given data there was no apparent relationship between rate of loading an d withdrawal strength; however they note that the study is not conclusive and further testing should include conditioning of wood. Further investigation was conducted at by Sh erman (2000) which used four stages of loading (0.05, 0.1, 1.0, and 10.0 in./min.) to de termine the effect of loading rates on fastener withdrawal strength in wood. Sample were constructed with tw o wood species, southern yellow pine and spruce pine fir, and three fastener, ha d driven 8d smooth shank nails, power driven 8d smooth shank nails and hand driven 8d ring shank nails, for 20 repeats each. Again there was no increase in strength as rate of loading increased, in fact withdrawal loads were slightly higher for the slowest loading rate. Results do not suggest a relationship between loading rate and fastener withdrawal strength. 33

PAGE 34

Pull-Through Testing The pull-through strength of OSB and plywood panels and effects of aging have been investigated in order to determine the appropria te failure mode for wood roof sheathing. Chui and Craft (2002) investigated pull-through streng ths of roof sheathing system with plywood or OSB per ASTM D-1037 (ASTM 2000). It was fo und that unless sheathing, both OSB and plywood, is sufficiently thick (i .e. greater than 1/2 in.) the pull-through strength of 2-3/8 in. common and 3 in. common roof sheathing fasteners will be the cont rolling failure mechanism. Roof sheathing fastener strength is a function of both pull-through & with drawal failure modes, where the lower capacity controls. Age and Weathering Effects Age and weathering effects on roof sheathing fastener withdrawal strength has been investigated in laboratory se ttings in several studies disc ussed below. Feldborg (1989) conducted long term withdrawal test ing of annularly threaded and smooth shank nails to evaluate the effects of long term loading and humidity cy cling. Nails were driv en to three different depths and loaded with consta nt withdrawal load for two y ears while using five different humidity cycling treatments. After two years th e withdrawal strengths of samples which did not fail were obtained by short term tests. It is found that long term loadi ng has little effect on withdrawal capacity but that alternating humidity does reduce withdrawal capacity. Chow et al. (1990) tested 6d smooth shank nails and 2 in. long 16 gauge staples for withdrawal and pull-through streng ths after being exposed to (1) long term (5 years) weathering outside and (2) accelerated agi ng in the laboratory per ASTM D-1037 (1978) by soaking in water and drying in cycles. Results from plywood and OSB samples are summarized in Table 2-5. Mean withdrawal strength from 16 samples ranged from 31% to 125% of the corresponding original strength, and mean pull-through resistance ranged from 75% to 99% of the 34

PAGE 35

corresponding original pull thr ough strength. A possible reas on for increased withdrawal strength is suggested to be due to corrosion. Pye (1995) addressed the effects of nail drivin g/coating, shank type, framing species, and heat cycles. In testing he used 2 in. by 5 in by 15/32 in. plywood sheathing attached to a 2 in. by 4 in. by 4-1/2 in. wood framing members of southern yellow pine and spruce pine fir. Pye found by testing different fasteners that coated or power driven nails had a hi gher withdrawal capacity than uncoated, hammer-driven nails. Ring shank nails had a 51% higher capacity than similar diameter smooth shank nails. Fasteners installe d in southern yellow pi ne wood had the highest capacity except in the case of heat testing. Heat cycles ra nging from 1 to 48 hours in duration were found to have the most significant eff ect, producing approximately 56% reduction, on fastener withdrawal capacity as compared with samples not exposed to heat cycling. Nail Withdrawal Tests on Existing Residential Buildings Laboratory investigations (Feldborg 1989, Chow et al. 1990 and Pye 1995) have found that humidity and heat reduces ultimate withdrawal strength of nails and st aples installed in wood framing members under laboratory c onditions. However the effect of humidity and heat cycles on the in-service performance of mechanical fasteners has not been well defined. Sutt et al. (2000) developed a portable nail extraction device to measure nail withdrawal strength of nails installed in existing roof stru ctures. The device uses a 2000 lb load cell connected to metal jaws that engage the nail h eads (after sheathing has been removed), and lever arms used to pull this assembly vertically upw ards. The load cell was connected to a digital readout that provides peak withdrawal load from the nail (Figur e 2-7). Sutt et al. collected nail withdrawal loads from 200 samples taken from a single residential struct ure in Anderson, SC. The 2-3/8 in. long 0.113 in. diameter nails were installed in SYP and SPF. The mean nail withdrawal load and coefficient of varia tion for SYP results were 68 lbs/in. and 56% 35

PAGE 36

respectively. The mean nail wit hdrawal load and coefficient of variation for SPF results were 42 lbs/in. and 59% respectively. The rather high variability in withdrawal loads and low mean capacities obtained by Sutts work was attributed to aging effects associated with in-service conditions. Sutt also suggested that the withdrawal capacity of the nails were lower than expected from NDS (AF&PA 2005) design values al so due to the effects of age and weathering of the wood. Dynamic Loading Investigations into the eff ects of uniform dynamic wind loading on flexible thermoplastic roofing systems attached with mechanical fast eners have been conducted by (Baskaran and Dutt 1997; Baskaran et al. 1999b). In this study component testing of individual nails installed in a in. of plywood and OSB sheathing was conducted by applying sinusoidal load to the nails. When compared to static loading it was found th at the mean withdrawal loads for the dynamic loading were reduced by 13 to 30 %. The effects of dynamic loading on axially loaded dowel type fastener in wood is relatively unknown. The only such research has been c onducted by Baskaran a nd Dutt (1997) described above. The varying load may cause damage to th e frictional interface of the nail surface and the wood material. This type of damage may be sim ilar to fatiguing effects in other materials where the number and amplitude of cycles determines the amount of damage. The cumulative damage model used to describe the interaction of multiple durations at different amplitude was proposed by Miner (B enham and Warnock 1976) can be described by, 1...2 2 1 1 N n N n (2-2) Where n1 is the number of cycles at one stress level which would require N1 cycles to cause failure, and n2 is the number of cycles at a s econd stress level which would require N2 36

PAGE 37

cycles to cause failure. The idea is that the sum of the ratio of numbers of cycles at each stress level should sum to 1.0 or total damage. In order to use this fatigue model for dowel t ype fasteners installed in wood it is necessary to determine the failure or stress level number of cycles (S-N) curve, which would define how many cycles at a stress level it w ould take to fail a connection. To extend this fatigue model to wood roof sheathing uplift capacity it is first necessary to un derstand how pressure load is distributed to individual fasteners. As both the withdrawal strength of individual nails and the distribution of pressure load through a sheathing panel are extr emely variable processes the problem would need to be solved through stochastic methods requiring large numbers of repeats. Related Research at th e University of Florida This thesis seeks to advance the understanding of the structur al performance of wood roof sheathing systems through component based test ing of new and existing wood roof panel specimens. The study follows a research dire ction initiated through the Florida Coastal Monitoring Program (FCMP). FCMP was initiate d at Clemson University in 1998 with three main goals; (a) the characterization of wind field in real loading conditions, (b) quantifying the resistance of existing residential constructi on structural components and (c) laboratory simulation of hurricane loading characteristics in component testing (J esteadt 2006). Jesteadt developed and summarized methods to harvest ex isting residential wood roof sheathing panels. These procedures were used for removal and testi ng in this investigation to evaluate the effects of aging and weathering. Summary This chapter presented testing procedures, review of damage levels observed to wood structures in hurricanes, wind lo ad on low-rise structures, stru ctural behavior of wood roof 37

PAGE 38

panels at ultimate wind uplift loads, and source of the previous wood r oof sheathing panel uplift tests used as comparison in this study. 38

PAGE 39

Ground Level Flow Separation Zones Wind Flow Figure 2-1. Resulting pressure loads from wind loading of residential structures Figure 2-2. Structural damage due to loss of roof sheathing below design wind speeds, design wind speed 140 mph actual wind speed 120 mph Ono Island, AL (FEMA et al. 2005b) 39

PAGE 40

Figure 2-3. Typical residential structure construction with wood r oof sheathing attached to metal plate trusses or rafter, which are attached to walls with either metal straps or toe-nails 6 in. / 12 in. 6 in. 12 in. 24 in. Wood framing member nominal 2 in. wide by 4 in. Figure 2-4. Fastener schedul e and construction of roof panel designed by pre-1994 building code 40

PAGE 41

41 Figure 2-5. Loss of roof sheathing at overh ang locations, Hurricane Katrina (130 mph) long beach Mississippi (FEMA et al. 2006) Figure 2-6. Tributary areas fo r individual fastener installed in a 6 in. / 12 in. spaced roof sheathing panel

PAGE 42

Table 2-1. Wind Uplift Failure Pressure Research C onducted on Wood Roof Sheathing Panels (1993 through 2004) Reference Sheathing Thickness, in. Sheathing Type Wood Member Size Wood Member Species Nail Size Nail Spacing, in./in. Number of Samples Average Uplift, psf COV Distribution Loading Regime (Cunningham 1993) 15/32 5-ply plywood 2x4 Douglas-fir or larch 6d Common 6/12 1 55 N/A N/A Monotonic 30-40 psf/min 15/32 5-ply plywood 2x4 Douglas-fir or larch 8d Common 6/12 1 130 N/A N/A 7/16 OSB 2x4 Douglas-fir or larch 6d Common 6/12 1 65 N/A N/A 5/8 4-ply plywood 2x4 Douglas-fir or larch 8d Common 6/12 1 105 N/A N/A 15/32 5-ply plywood 2x4 Douglas-fir or larch 6d Common 6/6 1 120 N/A N/A 5/8 4-ply plywood 2x4 Douglas-fir or larch 8d Common 6/6 1 218 N/A N/A 5/8 4-ply plywood 2x4 Douglas-fir or larch 8d Ring Shank 6/6 1 397 N/A N/A (Mizzell 1994) 15/32 Plywood 2x4 SPF #2 or better 6d Common 6/12 4 26 0.09 Normal Step: Increase 1 psf increments and hold for 1.5 sec 15/32 Plywood 2x4 SPF#2 or better 8d Common 6/12 10 61 0.11 Normal 15/32 Plywood 2x4 SPF #2 or better 8d Common 6/6 10 107 0.16 Lognormal 19/32 Plywood 2x4 SPF#2 or better 8d Common 6/6 10 115 0.28 Lognormal 19/32 OSB 2x4 SPF #2 or better 8d Common 6/6 10 77 0.27 Normal (Murphy et al. 1996) 15/32 OSB N/A SYP #2 or better 8d Common 6/6 30 131 0.14 Normal Step: Increase 1 psf for 1.5 sec (Kallem 1997) NA 4-ply Plywood 2x4 SYP 6d Common 6/12 14 33 0.22 Normal Rapid Monotonic (failure in less than 16 seconds) (Jones 1998) 19/32 OSB 2x4 SYP/SPF #2 or better 8d Common 6/12 10 87 0.28 Lognormal Monotonic 15/32 CDX Plywood 2x4 SYP/SPF #2 or better 8d Common 6/12 9 80 0.17 Normal (Sutt 2000) 15/32 Plywood 2x4 SYP 8d Common 6/12 7 79 0.09 N/A Rapid Monotonic Failure occurs within 10-45 sec 7/16 OSB 2x4 SYP 8d Common 6/12 7 67 0.15 N/A (NAHB 2003) 7/16 OSB 2x6 SPF #1/#2 8d Common 6/12 3 228 0.07 N/A Monotonic: 20 psf/min (IHRC 2004) 1/2 CDX Plywood 2x4 SYP 8d Common 6/12 49 110 0.17 Lognormal Monotonic 1/2 CDX Plywood 2x4 SYP 8d Ring Shank 6/12 50 140 0.17 Normal 42

PAGE 43

Table 2-2. Summary of selected uplift pressure tes ting results from Jone s (1998) investigation into adhesive retrofit methods for residential wood roof sheathing Control Full-bead both sides % Increase in Mean Failure Pressure # of Panels Mean Failure Pressure (psf) COV # of Panels Mean Failure Pressure (psf) COV OSB 10 87 28% 4 185 16% 113% Plywood 9 72 10 213 196% Table 2-3. Summary of existing residential r oof sheathing uplift testing conducted by Judge and Reinhold (2002) House Sheathing and Fr aming Fastener Spacing # of Panels Failure Pressure (psf) 1 0.465 Ply, 2" x 4" @ 24" o.c. 2.5" 0.113" dia. 6" / 8-12" 1 127 2 0.465 Ply, 2" x 6" @ 16" o.c. 2.5" 0.113" dia. 6" / 8-10" 2 232 3 7/16" OSB, rafter @ 24" o.c. Staples 3-5" 1 105 4 7/16" OSB, rafter @ 24" o.c. Staples 3-5" 2 110 5 rafter @ 24" o.c. 2.5" 0.113" dia. 7-9"/5.5-7.5" 1 196 6 rafter @ 24" o.c. 2.5" 0.113" dia. 7-9"/5.5-7.5" 2 119 7 8" Plank, rafter @ 16" o.c. two 8d and one 16d nail per rafter 1 450 43

PAGE 44

Table 2-4. Summary of buildi ng code requirements for roof sh eathing design in Florida (1988 to current) Year Building Code Nail Min. Sheathing Thickness Fastening Schedule 1988 South Florida Building Code (Dade County 1988) 6d common 8d common in. or less greater than in. 6 in. / 12 in. 1994 South Florida Building Code(Dade County 1994) 8d common up to 19/32 in. 6 in. / 12 in., and 4 at gable ends 1997 Standard Building Code (SBCCI 1997) 6d common 8d common in. or less 19/32 in. or greater 6 in. / 12 in. 2000 International Building Code (ICC 2000) 8d common in. or less 6 in. / 12 in. 2004 Florida Building Code (ICC 2004) 8d common in. or less 6 in. / 6 in., and 4 at roof corners 2004 Florida Building Code High-Velocity Hurricane Zone (ICC 2004) 8d ring shank Minimum 19/32 in. 6 in. / 6 in., and 4 at gable ends 2006 International Building Code (ICC 2006) 8d common in. or less 4 in. / 8 in. 2007 Florida Building Code High-Velocity Hurricane Zone (ICC 2007) 8d ring shank Minimum 19/32 in. 6 in. / 6 in., and 4 at gable ends Table 2-5. Percentage of orig inal strength summary of Chow et al. investigation of ageing effects on fastener withdrawal and pullthrough resistance, (Chow et al. 1990) Sheathing Fastener Withdrawal Pull-Through Outdoor Lab Outdoor Lab Plywood C-D grade Nail 38% 38% 85% 77% Staple 61% 30% 87% 99% OSB (pine) Nail 31% 125% 75% 82% Staple 56% 18% 83% 97% Notes: Nail-6d smooth shank; Staple-2 in. long 16 gauge staples; Lab aging per ASTM D-1037 (78) 44

PAGE 45

45 Figure 2-7. Nail Extraction Device Devel oped at Clemson University by Sutt (2000)

PAGE 46

CHAPTER 3 DEVELOPMENT OF STANDARD WIND UP LIFT METHOD WITH STATIC AND DYNAMIC LOADING The intent of any standard wind uplift test protoc ol should be to replicat e the real effects of wind loading on building components, under controll ed laboratory conditi ons, that the building component would experience in an extreme wind event. The test prot ocol should also be repeatable, consistently reproducing the same test conditions in every test so that reasonable comparisons can be made for performance of any roof or wall system. For this reason simple uniform static test protocols have been used for the majority of commerci al and residential roof sheathing tests. These tests assume that extrem e wind loading can be mode led as a pseudo-static pressure load on a rigid building. Most wind uplift tests ignore the dynamic characteristics of the wind, assuming that structural component are ri gid and therefore will not have a significant response to fluctuating loads (Holmes 2001). However it may not be an accurate assumption because it relies on the rigidity of the structure not the components. Residential wood structures are rigid systems but the stiffness of individu al building cladding components (i.e. wood roof sheathing panels) may not be as rigid. Wind flow produces pressures and structural loads that vary temporally and spatially but these features are not represented in st atic test methods. In order to evaluate wind uplift performance of wood roof sheathing panels a test protocol must be selected that reasonably recreates wind load conditions. As discussed in the previous chapter, no such protocol currently exists and still no widely accepted test methods are available for wood roof sheathing. Thus a first stage of this research was to develop a suitable wind uplift test protocol. It was decided to develop both a st atic test protocol having clear-cut guidelines, as well as a dynamic test protocol, to test the hypo thesis that fluctuating wind load reduces the ultimate resistance of wood roof panels. 46

PAGE 47

This chapter presents the development of the test protocol and two loading functions that will be used in the experimental studies following this chapter. The test protocol is called the University of Florida W ood Roof Sheathing Uplift Test (UF-WRSUT) Protocol with Method A Static Pressure and Method B Dynamic Pre ssure. The main features of the UF-WRSUT Protocol are as follows: In order to provide a basis for comparison th e parameters effecting panel uplift resistance addressed are, The pressure traces are fully described to enable accurate reproduction, The number of repeats in each test series is established (a minimum of 10 repeats is recommended) Failure modes are observed and recorded, The material properties (specific gravities and moisture contents) of the wood sheathing and components are recorded, Details of geometric dimensions and characte ristics of the fasteners are recorded, During this investigation the above requireme nts were added incrementally as knowledge of critical parameters was gain ed through (1) a literature review of previous findings and (2) experience gained during testing. Therefore not a ll testing presented in this investigation adheres to these requirements. Additionally some requi rements are not able to be followed due to limitations such as a lack of sufficient harvested samples, this is noted in the discussion. UF Wood Roof Sheathing Uplift Test (UF-WRSUT) Protocol Several experimental studies have been pr eviously used to determine the wind uplift resistance of wood panels (Cunningham 1993; IHRC 2004; Kallem 1997; Mizzell 1994). For the purposes of this research a sta tic pressure test is defined wh en the chamber pressure does not cycle or fluctuate. Static pressure tests may include periods of constant pressures held for specific lengths of time or peri ods where pressure is monotonica lly increasing or decreasing. 47

PAGE 48

The UF-WRUST test protocol was adapted from the existing standard test protocol, ASTM E330-02 for determining the structural performan ce of exterior curtain walls and windows and doors (ASTM 2004b). ASTM E330 is intended onl y for evaluating the structural performance associated with the specified test specimen and not the structural performance of adjacent construction. This test method was the selected starting point instead of using wind uplift test protocols for commercial roofing tests because of the typical duration of each test was comparable to durations of previous studies. ASTM E330 acknowledges the time-dependency of strength and deflection characteristics in some materials and so it recommends testing assemblies for the actual time duration to which it would be exposed. The test is conducted by seali ng a test specimen against one face of a closed test chamber to which air is supplied or exhaus ted at a sufficient rate to mainta in the pressure difference across the specimen. The ultimate failure pressure of the panel is the maximum pressure that it sustains without failure. Failure of the panel can be by nail withdrawal or pull through or by fracture of one or more wood framing members or of the sheathing. During a hur ricane event a roof structure would typically expe rience several hours of elevat ed wind speeds and increasing gustiness, however typically wind speeds appro ach or exceed the design wind speeds of the house only during a few time periods. It is assu med that the extreme wind speeds are the sole cause of roof damage, and therefore the eff ect of lower intensity pressure duration and fluctuations are neglected. The 10 second period wind pressure is then representing the period in which peak pressure acts on the structure to cau se damage. The static test will reproduce the most severe peak pressure but not the sustained buffeting that the roof may experience. The UF-WRSUT, Method A Static Pressure Test Protocol The UF-WRSUT, Method A Static test prot ocol uses the step-and-hold procedure identified by ASTM E-330 procedure B with a 10-second pressure plateau. The pressure control 48

PAGE 49

can be applied manually or under computer-contr ol. For the computer-c ontrol approach it is found that two 20 second long initial stabilizing pressure steps are needed; at 5 psf and again at 15 psf. The chamber pressure is increased in 15 psf increments, held for 10 seconds and increased again until failure. A typical pressure trace is shown in Figure 3-1. The following adjustments to the ASTM E330 test protocol were made: Include initial pressure stabilization time steps of 5 psf and 15 psf for 20 sec. each. Apply pressure in one direction only; i.e. either suction (reduced chamber pressure) or pressure (increased chamber pressure), but not both. Eliminate deflection readings to obs erve permanent panel deformations. Eliminate the 60 second recovery peri od for stabilizati on during testing. Determine and report the moisture content and specific gravit y of wood members. Record and report the nail prop erties (length, shank diameter, head diameter, coating material and deformed shank pattern). Panel failure is defined as any permanen t separation between sheathing and framing member, sheathing fracture, split in framing member or failure of nails (in withdrawal or pull through). The chamber pressure is monitored and continuously recorded during the test and the peak instantaneous pressure at failure is recorded. The determination of the peak instantaneous failure pressure is discussed in a subsequent sect ion. Each panel is insp ected after the tests and the locations and failure mechanisms of the fasteners are noted. UF WRSUT, Method B Dynamic Pressure Test Protocol The dynamic wind pressure trace was developed to better simulate the wind pressure fluctuations observed in actual wind loading. There are a few existing dynamic pressure test protocols used for roof sheathing, for example SIDGERS (Baskaran et al. 1999a). These test traces apply uniformly distribute d pressures in regular cycles, am plitude and constant frequency, to the specimen. The more recent SIDGERS wa s developed using rain-flow analysis of wind 49

PAGE 50

tunnel pressure data. The UF-WRSUT, Method B was also developed from a wind tunnel study but the characteristics of the fluctuations were retained. Development of the UF-WRSUT, Method B dynamic trace is based on time histories of pressure fluctuations collected from wind t unnel data conducted at Clemson Universities boundary layer wind tunnel facility. The study (D atin and Prevatt 2007) used a 1:50 scale model of a 30 ft by 60 ft gable roof residential structure. The building had a mean roof height of 13 ft 6 in. and a 4 in 12 (18.4) slope, Figure 3-2, and 387 taps were installed. Upwind terrain was modeled as suburban terrain exposure, (zo = 0.22 m) with a turbulence intensity of 24% at mean roof he ight in the tunnel (with building model removed). Only selected highlights of this experiment are reported here, and further details can be obtained in Datin and Prevatt (2009). For this study a simulated pr essure trace was developed from the measured pressure coefficients at pressure tap #002, located at (16.75 in., 7.75 in.) from the ridge corner at full-scale, see Figure 3-3. The dynamic pressure trace was developed usin g the reduced frequency relationship to provide an equivalent full-scale time step. Th is is done because wind tunnel measurements are taken at a faster rate than the pressure control system can process, therefore requiring that the data be compressed while retaining the frequenc y content of the wind tunnel measurements. A 10 second period pressure fluctuation was chosen so that it contained at leas t three peak pressure excursions of near equal and highest magnitude during the period. The wi nd pressure coefficient trace was converted to full-scale pressure trace an d the highest peak pressure was matched to a pressure plateau (i.e. 15 psf, 30 psf, 45 psf, etc.) as in Method A static pressure trace, see Figure 3-4. 50

PAGE 51

Failure of Wood Roof Sheathing Panels The predominant failure mode for wood roof panels is nail withdrawal, followed by nail pull through failures and to a lesse r extent fracture of the wood framing member or sheathing. Typically connections fail in fastener withdrawal or fastener head pullthrough failure modes. The failure of an entire panel is a result of multiple connecti ons failing in which individual connections progressively fail afte r the initial fastener. However the removal of an entire roof sheathing panel from the framing members is not n ecessary to consider the panel as failed since even partial removal provides a path for water fl ow into the structure. Additionally the nailed (dowel-type) fastener connections to wood, loaded axially in withdrawal is a very brittle one in which nearly all strength is lost after minima l displacement (Forest Products Laboratory 1999). Currently it is not possible to relate the load di stribution behavior of fa steners in wood sheathing panels to failure pressures because there have been no studies conducted that directly measured the nail withdrawal load. Further it is not alwa ys possible to relate fi eld performance of roof panels to failure pressures given the large num ber of parameters (con struction qua lity, actual wind speed, wood specific gravity and condition, nail type and length, and spacings etc.) for which no data is available. Further post-disaste r studies do not usually detail the actual failure mechanisms observed in the panels. It is proposed that to establish the relationship of fastener type and sp acing, sheathing, and framing condition on roof panel performance the lo cation and modes of failu re for all fastener failures must be observed. Qualitative consider ations such as the location of the connection failures within the panel and the type of failure (r apid versus slow) may also be important. It is noted however, that such visual observations of failure modes after the panel has failed cannot determine the location of the initial fastener to fail. 51

PAGE 52

Determination of Panel Failure Pressure The panel failure pressure recorded is the inst antaneous peak pressure recorded at failure where panel failure is determined by examination of pressure trace. Pressure for panels tested in the static phase of the investigation were pr edominantly recorded with a peak measurement pressure gauge, therefore failure pressure for these samples is simply the peak instantaneous pressure from the test. The peak instantaneous pr essure is selected as the panel failure pressure because it is the largest pressure the panel could withstand before permanent damage was inflicted. A pressure transducer was incorporated to recorded time-historie s for the ccSPF retrofit test series from which the peak instantaneous pr essure was found (Figure 3-5). The benefit of the pressure time-history is that if for some reason the pressure spiked after panel failure this false peak pressure could be identified and discarded. Panels tested in the static vs. dynamic pha se of the investigat ion used a computer controlled system which measured the pressure time-history. The program shuts off if it determined there was significant leakage in the system resulting in a deviation from the target pressure. Pressure time-histories were evaluated to determine the peak instantaneous pressure which preceded failure. Figure 3-6 presents a sa mple of the pressure time-history for a panel fastened at 6 in. / 12 in. All time-histories of dynamically tested panels are presented in Appendix E. 52

PAGE 53

I n c r e a s e t o F a i l u r e Figure 3-1. UF-WRSUT static pr essure trace. (5 psf initial sta tic pressure is included in the trace) A B Figure 3-2. Gable roof model A) installed in Wind Tunnel and B) close-up of model 53

PAGE 54

4.1 2.9 18.4 7.2 14.9 4.2" 4.2" 14. 9 4.2 4.2 A 4 ft 8 ft Ridge Line X Y 027 001 023 026025024 005 004003002 B Figure 3-3. Pressure tap locati on A) 1:50 model scale ( 62 ft by 35 ft full scale) and B) full scale representation of panel Figure 3-4. Comparison of sta tic vs. dynamic pressure traces 54

PAGE 55

Failure Pressure (60 psf) Figure 3-5. Diagram of instantaneous failure pressure for panel tested statically 55

PAGE 56

56 A Instantaneous Peak Initial deviation from target pressure (failure) Target Actual B Figure 3-6. Sample of target and actual pressure time-history for panels tested with the PLA system A) full time-history and B) close up of instantaneous peak and deviation from target pressure

PAGE 57

CHAPTER 4 EXPERIMENTAL SETUP Wind uplift pressure tests were conducted using 4 ft by 8 ft r oof sheathing panels tested on a steel pressure chamber. Tests were conducted on both laboratory-fabri cated panels and roof panels harvested from existing structures. The description of sample panel construction, harvesting, preparation, instrumentation, test equipment and procedures used in testing are presented in this chapter. Panel Test Series and Construction Laboratory fabricated panels were constructed in two groups with slightly different arrangements and different retrofit methods. Th e first group evaluated th e effects of fastener type and spacing on uplift resistance of panels test ed statically. The second compared the uplift resistance of panel tested stat ically vs. dynamically. Aged roof panels were harvested from 12 existing residential houses thr ough the FCMP program over a year period from 2006 to 2009, and tested in two groups with di fferent retrofit methods. Harvested panels have relatively small data sets due to the difficulties in procuring them, which is detailed below. Panel Test Series The first phase of testing is conducted on both laboratory fabricated panels and harvested panels, which is summarized in Figure 4-1. Testing was conducted to evaluate the developed test method (UF-WRSUT) and the effect of agi ng on panel uplift capacity. The second phase of testing is also conducted on both laboratory fabricated panels and harvested panels, which is summarized in Figure 4-2. Testing was conducted to evaluate the effect of dynamic loading on panel uplift capacity. 57

PAGE 58

Laboratory Fabricated Panels (New) Test panel specimens were fabricated using a 4 ft by 8 ft by in. thick oriented strand board (OSB) sheathing fastened to 2 in. by 4 in. no minal dimensions southern yellow pine (SYP) framing members spaced 24 in. apart. The framing members were at least 4 ft-6 in. long in order to span the short dimension of the pressure chambe r, see Figure 4-3. Nails were installed using a air compressor driven nail-gun, Stanley Bostitc h Model FL21P. The air pressure was set at 70 psi to 80 psi and adjusted periodically to avoid over-driving the nails. In the static phase of testing panels were c onstructed with three nail sizes; a) smooth shank 0.131 in. diameter 2-1/2 in. long nail (8d smooth shank nail) b) annularly threaded 0.113 in. diameter 2-1/2 in. long nail (8d ring shank nail) and c) smooth sh ank 0.113 in. diameter 2-3/8 in. long nail (2-3/8 in. 6d smooth shank nail). The 8d smooth shank and 8d ring shank nails were installed in three fastener schedules 6 in. / 12 in., 6 in. / 8 in., and 6 in. / 6 in. with approximately 15 replications in each set. The 2-3/8 in. 6d smoo th shank nails were installed at 6 in. / 12 in. only with 15 replications, see Fi gure 4-4. To evaluate the stru ctural benefit of retrofitting existing panels two arrangements of closed cell spray applied polyurethane foam (ccSPF) adhesive were tested. The ccSPF was installed on panels originally constructed with 8d ring shank nails spaced at 6 in. / 12 in. The panels we re installed with (1) a fillet of ccSPF along each side of framing member to sheathing interface (see Figure 4-5 a) and (2) a full 3 in. thick layer of ccSPF filling each space between framing members (see Figure 4-5 b). A series of control panels were also tested with the same materials to provide a direct comparison, see Table 4-1 for a summary of panel series. In the static vs. dynamic phase of testing labo ratory fabricated panels were constructed with one nail type attached w ith two fastening schedules and one retrofit method. Hot dipped galvanized 2 in. 6d (0.113 in. diameter) smooth sh ank nails were used to match dimension of the 58

PAGE 59

nails found in harvested panels. Two fastening sc hedules were used 6 in./12 in. and 6 in. /6 in. with 10 replications each, where half were tested statically and the other half were tested dynamically (Table 4-1). A 2 mils. thick plasti c sheet was inserted between the wood stud and OSB sheet during fabrication to allow samples to be tested with positive chamber pressure (Figure 4-6), discussed below in installation of pa nel section. To evaluate the effect of dynamic loading on retrofitted panels four additional panels were constructed with the 2 in. 6d common nails at 6 in. /12 in., then installed with 8d ri ng shank nails between the 6d smooth shank nails. Harvested Panels (Aged) To evaluate the effect of in-service conditi ons on wind uplift resist ance of residential structures panels were harvested from existing homes. Panels were removed from thirteen existing residential wood roof st ructures located in four comm unities throughout Central Florida, see Figure 4-7 for locations of structures. A to tal of 48 panels were harvested from 2006 thru 2009. Testing is broken into two groups the first conducted statically to ev aluate the effect of age and two retrofit measures. The second test gr oup is tested statically vs. dynamically in order to evaluate the effect of dynamic loading on roof sheathing wind uplift capacity. Panel attachment varies significant, in fastener type or spacing, resulting in limited sample sizes and relatively few direct comparisons to laboratory fabricated samples. In order to obtain panels researchers must wait for homes purchased by local or state governments to be made available before demolition. This is a slow process and re sults in researches getting what is available, which is rarely the ideal sample. Additionally the process of removing panels prohibits large numbers of panels to be removed from a single roof. To access panels shingles are removed from th e necessary section of roof. Panels are removed by cutting the sheathing around the sample panel with a circular-s aw to expose framing members (Figure 4-8 a). This prevents any adjacent panel to be used as a sample panel therefore 59

PAGE 60

keeping the number of panels removed from a singl e roof small, see Figure 4-8 c. The sample panel is then removed from the roof structure by cutting the exposed fr aming (Figure 4-8 b). Care is taken to brace the pa nel only on the sheathing during removal to prevent damage to fastener connections. The panel is then lowere d to ground level to remove any additional truss members connected with metal plates, while only holding the sheathing. Metal truss plates are cut with a grinder while bracing the framing member to protect the fastener connections. Panels are braced only by the sheathing during transpor tation to prevent racking of the framing. All harvested panels tested in the first group (38) to evalua te the effect of in-service conditions on wood roof sheathing wind uplift resistance consisted of 4 ft by 8 ft by in. thick plywood sheathing. Sheathing is attached to 2 in. by 4 in. wood truss or rafter members spaced 24 in. apart with 1.5 in. staples, 2.5 in. staples, 2.5 in. long 0.131 in. diameter (8d) smooth shank nails and 2 in. long 0.113 in. diamet er (6d) smooth shank nails at various spacing ranging from 4 in. / 4 in. to 6 in. / 12 in. detailed in Table 4-2. Panels were harvested from twelve homes located in three cities in Central Florida (Port Orange, Crystal City and Bartow) and ranged in age from 29 to 33 years. Twenty five harvested of the pa nels were tested in th eir existing configuration and thirteen panels were retrof itted. Retrofitting of panels was performed to simulate the economical strengthening of existing roof sheathing. Panels were retrof itted using one of two methods; a) twelve were retrofi tted using 8d-ring shank nails inst alled at a 6 in. / 12 in. spacing which can be performed when a roof covering is replaced (retrofit A-1, see Figure 4-9), and b) one panel was retrofitted using a 3 in. thick laye r of closed-cell spray-ap plied polyurethane foam (ccSPF) which can be installed through the attic at any time (Figure 4-10). Ten panels were harvested from the same home in order to evaluate the effect of dynamic loading on wind uplift resistance of wood roof sheathing. Construction of panels are 4 ft by 8 ft 60

PAGE 61

by 7/16 in. thick OSB attached to 2 in. by 4 in. framing members spaced 24 in. apart. Sheathing was fastened with 1-1/2 in. long 14 gauge staples sp aced at 4 in. interior and exterior. Six panels were tested in existing conditions, two statica lly and four dynamically. The remaining four panels, two static and two dynamic, were retrofitted with 8d ring shank nails spaced between a 6 in. / 12 in. fastening schedule (retrofit A-2, see Figure 4-9) to correspond to static vs. dynamic testing of laboratory fabricated panels. Table 4-2 summarizes the test series. Specific Gravity and Moisture Content Measurements Specific gravity and moisture content of w ood members are identified in Chapter 2 as parameters effecting the wind up lift resistance of wood roof sheathing. Therefore specific gravity measurements were taken for all static vs. dynamic testing to ev aluate their effect on uplift resistance. Moisture content measuremen ts were taken only for sample harvested from existing structures. Each reported framing member specific gravity or moisture content is the average of three samples, but methods for coll ecting samples were diffe rent between laboratory fabricated panels and harvested panels. Oven dry specific gravity measurements were determined for framing members following procedures outlined in ASTM D 2395 (2006b), Me thod A. The equation for specific gravity presented in Method A (#2) is Lwt M KW SG 100 1 ( which 4.1) calculates the green specific gravity or in other words the sp ecific gravity relative to the moisture content at the time of testing. It wa s that the definition used in the wood design manual NDS (AF&PA 2005) would be selected, which specifi es the oven dry specific gravity. The term 61

PAGE 62

100 1 M W ( is defi 4.2) ned as the calculated oven-dry wei ght of specimen based on moisture content t of testing measurements. Due to the selection of oven dry sp ecific gravity as the primary resul the oven-dry weight was directly meas ured. The resulting equation is Lwt KW SG (4.3) where K = 27.68 t = ss (in.). The oven used for drying is a 3488M Model Lab-line Instruments convection thod B was selected due to its ted ared and allowed to dry for 2-3 weeks. Each 10 ft board was cut to ma t unit conversion, W = dry weight (lbs ), L = length (in.), w = width (in.), and thickne oven. Vernier calipers were used to meas ure the specimen volume and an Omega WSB 8150 scale with a maximum capacity of 15 Kg was used to measure mass. Moisture content measurements were conducte d on harvested static vs. dynamic tested samples per ASTM D4442 Method B secondary ove n dry method. Me similarity to the specifi c gravity method used, therefore sa mples were simultaneously tes for specific gravity and moisture content. The precision of measurements is assumed to be the nearest whole percent because precision and bias calculations (section 5) were not made. Additionally endpoint measurements were not us ed because the drying time specified in the specific gravity was followed. Framing members of panels fabricated in th e laboratory are constructed with ten foot boards purchased together, prep ke 2 framing members. Three samples were collected from each ten foot board, one on either end and one in the middle (Figure 4-11), a nd used to determine the specific gravity of both framing members cut from the ten foot board. AS TM D2395 specifies that specimens be cut a 62

PAGE 63

least 18 in. from the ends of boards to avoid effects of end drying, however as oven dry specific gravity was being calculated this was not followed. Framing members of panels harvested from ex isting structures are cut longer than ne for uplift testing in order to get specific gravity/moisture con eded tent samples. Three samples were remov thesis are to evalua te the effects of dynamic loading and in-service conditions on wood istance. In or der to do this a test metho ith a e ed from the ends of framing members, s ee Figure 4-12. End drying effects do not apply due to the fact that framing members were cut from the middle of the roof structure, therefore moisture content measurements were taken. Uplift Testing Equipment The objectives of testing as part of this roof sheathing wind uplift res d was developed with static and dynamic load traces, which is detailed in Chapter 3. The equipment with which testing is conducted is developed through-out the testing conducted. All panel samples were tested in the same pressure chamber. The pressure chamber is a steel chamber measuring 4 ft. 6 in. wi de by 8 ft. 6 in. long and 6 in deep. Laboratory fabricated and harvested panels tested statically are test ed with manual control of chamber pressure. Pressure is measured with a peak measurement ga uge for all statically tested harvested panels and most statically tested laborat ory fabricated panels. The pre ssure gauge was replaced w pressure transducer for some statically tested la boratory fabricated panels. Panel tested with manual pressure control the pres sure is supplied with two vac uum pumps in series, detailed below. Static vs. dynamic testing required higher accuracy of pressure control therefore a different system was used (Pressure Load Actuato r), which is described below. As part of th feed-back control system pressure time-histories were recorded with a pressure transducer. 63

PAGE 64

Panel Installation Pressure is applied in one of two different methods, either negative or positive applied pressure depending on whether the specimen was la b built or harvested from an existing house. In both arrangements an uplift lo ad is applied to the panel. The negative pressure arrangement was used for all panels tested statically and harvested panels tested statically vs. dynamically. Panels are placed on top of the chamber then covered with 2 mil plastic leaving extra folds at fram ing to allow continuous contact of plastic and sheathing. Duct tape is used to seal plastic to chamber, see Figur e 4-12. Then a shop-vac is used to remove the excess air in system, which can affect pressure control. After loading is completed the panel is removed from chamber and inspec ted for fastener failure type and location. Laboratory fabricated panels te sted statically vs. dynamically are tested in the positive pressure arrangement. Panels tested with positive pressure are constructe d with plastic between the sheathing and framing. Panels are placed in side the chamber sheathing up to restrain the panel. Shims and wood blocking are then used to restrain the panel from movement during loading. Finally a steel frame is placed over the plastic and clamped tight, see Figure 4-13. After loading, the panel is removed from the cham ber and inspected for fastener failure type and location. Manual Control Laboratory fabricated and harves ted panels tested statically were controlled manually. Pressure was supplied with two 15 CFM US Vacuum pumps (model CP15) connected in series with PVC tubing to the pressure chamber supply pressure, see Figure 4-14. Pressure is controlled manually with a gate valve (Figure 414 a) which is closed to reduce the chamber pressure. Pressure measurements were initi ally recorded with an Omega DPG8000 general 64

PAGE 65

purpose digital peak load measurem ent pressure gauge, later pressure measured with an Omega, Model # PX243A-2.5BG5V pressure transducer. Pressure Load Actuator Static vs. dynamic testing requires a more accurate system to control chamber pressure. The accurate development and repeatability of the dynamic pressure trace was only possible because of the unique capabilities of the test apparatus used. The Pressure Loading Actuator (PLA) system utilizes a 12 hp regenerative blow er and computer-controlled feedback loop to actuate a 3 port valve with 5 arrangements, see Figure 4-15. The active valve control system and blower enable highly responsive co ntrol of chamber pressure result ing in a chamber pressure that closely follows the large amplitude pressure fluc tuations and 10 Hz response of the simulated dynamic wind pressure. The PLA system was developed by Cambridge Consultants for the University of Western Ontario to investigate wind loading of full scale structures. Development of the PLA system can be found in Kemp (2008). The system was designed to provide pressure loading to boxes (chambers) ranging from 1 to 64 ft2, which are placed on a structure allowing spatially varying dynamic pressure loading of an en tire structure. This ongoing study is detailed in Bartlett et al. (2007). The PLA system enab led accurate repeatable dynamic pressure loading of wood roof sheathing pa nels, see Figure 4-16. 65

PAGE 66

Static Testing (171) Lab Built (133) Harvested (38) Existing (103) 6d SS @ 6/12 (15) 8d SS @ 6/12 (15) 8d SS @ 6/8 (15) 8d SS @ 6/6 (15) 8d RS @ 6/12(15) 8d RS @ 6/8 (15) 8d RS @ 6/6 (13) Retrofit (30) 8d RS @ 6/12 (10) ccSPF fillet (10) ccSPF full 3 (10) Existing (25) 1.5S @ 3/6 (4) 2.5S @ 3/6 (2) 2.5S @ 6/12 (2) 8d SS @ 6/12 (2) 6d SS @ 6/12 (10) 6d SS @ 8/8 (2) 6d SS @ 3/6 (2) 6d SS @ 4/4 (1 ) Retrofit (13) 8d RS @ 6/12 (12) ccSPF full 3 (1) Figure 4-1. Summary of pane l series tested in the static phase of this study Static vs. Dynamic Testing (34) Lab Built (24) Harvested (10) Existing (20) 6d SS @ 6/12 6d SS @ 6/6 Retrofit (4) 8d RS @ 6/12 Existing (6) 1.5Staple @ 4/4 Retrofit (4) 8d RS @ 6/12 Static(5) (5) Dynamic(5) (5) Static(2) Dynamic(2) Static(2) Dynamic(4) Static(2) Dynamic(2) Figure 4-2. Summary of panel se ries tested in the static vs dynamic phase of this study 66

PAGE 67

Figure 4-3. Laboratory fabricat ed panel construction, 6 in. / 12 in fastener schedule shown 6 in. / 12 in. 6 in. / 8 in. 6 in. / 6 in. 6 in. 12 in. 6 in. 6 in. 8 in. 6 in. Figure 4-4. Laboratory panel fast ening schedule (6 in. / 12 in., 6 in. / 8 in. and 6 in. / 6 in.) 67

PAGE 68

A B Figure 4-5. Static ccSPF retrofit construction of A) fillet sample set and B) full 3 in. sample set Figure 4-6. Panel construction for static vs. dynamic testing with 2 mils. thick plastic sheet placed between sheathing and wood fr aming member during fabrication 68

PAGE 69

Table 4-1. Laboratory fabricated (New) panel series tested, cons tructed with in. OSB and 2 in. by 4 in. southern yellow pine # 2 or better Fastener Spacing # of Panels Static Panel Series 2-3/8 in. 6d smooth shank nail 6 in. / 12 in. 15 2-1/2 in. 8d smooth shank nail 6 in. / 12 in. 15 2-1/2 in. 8d smooth shank nail 6 in. / 8 in. 15 2-1/2 in. 8d smooth shank nail 6 in. / 8 in. 15 2-1/2 in. 8d ring shank nail 6 in. / 12 in. 15 2-1/2 in. 8d ring shank nail 6 in. / 8 in. 15 2-1/2 in. 8d ring shank nail 6 in. / 6 in. 13 2-1/2 in. 8d ring shank nail 6 in. / 12 in. 10 2-1/2 in. 8d ring shank nail, with Ret. B-1 6 in. / 12 in. 10 2-1/2 in. 8d ring shank nail, with Ret. B-2 6 in. / 12 in. 10 Static vs. Dynamic Panel Series 2 in. 6d smooth shank nail 6 in. / 12 in. 10 2 in. 6d smooth shank nail 6 in. / 6 in. 10 2 in. 6d smooth shank nail, with Ret. A-2 6 in. / 12 in. 10 69

PAGE 70

Table 4-2. Harvested panel series tested Source Age (years) Existing Retrofit # of Panels # of Panels Retrofit Method Static Tested Panel Series 1.5 in. staple @ 3 in. / 6 in. spacing Port Orange #1 4 2 A-1 2.5 in. staple @ 3 in. / 6 in. spacing Bartow #1 2 Bartow #2 2 2 A-1 2.5 in. smooth shank nail @ 6 in. / 12 in. spacing Crystal River #1 36 2 2 A-1 2 in. smooth shank nail @ 6 in. / 12 in. spacing Port Orange #2 29 3 1 B-2 Bartow #3 33 2 2 A-1 Port Orange #3 2 1 A-1 Port Orange #4 2 2 in. smooth shank nail @ 8 in. / 8 in. spacing Port Orange #5 2 2 in. smooth shank nail @ 3 in. / 6 in. spacing Bartow #4 1 2 A-1 Port Orange #6 2 2 in. smooth shank nail @ 4 in. / 4 in. spacing Bartow #5 33 1 1 A-1 Static vs. Dynamic Tested Panel Series 1.5 in. staple @ 4 in. / 4 in. spacing Debary #1 15 6 4 A-2 Notes: Retrofit Method A-1. 8d ring shank nail @ 6 in. / 12 in. spacing Retrofit Method A-2. 8d ring shank nail between 6 in. / 12 in. spacing Retrofit Method B-2. 3 in. thick layer of spray applied polyurethane adhesive 70

PAGE 71

Figure 4-7. Locations of residential st ructures where panels were harvested A B C Figure 4-8. Pictures of harvested panel remova l, A) Expose framing members, B) Cut framing members and C) Roof after panels removed 71

PAGE 72

6 in. 6 in. 12 in. 12 in. Retrofit A-1 Retrofit A-2 Existing Retrofit Figure 4-9. Comparison of re trofit measures A-1 and A-2 Figure 4-10. ccSPF retrofit of an existing residential stru cture in Port Orange, FL 72

PAGE 73

Table 4-2. Summary of harves ted panels tested statically and statically vs. dynamically Sheathing Attachment Method House Address Existing Retrofit # of Panels # of Panels Retrofit Method Statically vs. Dynamically Tested Pa nels, constructed with 7/16 in. OSB 1.5 in. staple @ 4 in. / 4 in. Debary #1 6 4 A-2 Statically Tested Panels, constructed with in. Plywood 1.5 in. staple @ 3 in. / 6 in. Port Orange #1 4 2 A-1 2.5 in. staple @ 3 in. / 6 in. Bartow #1 2 NA NA 2.5 in. staple @ 6 in. / 12 in. Bartow #2 2 2 A-1 2.5 in. smooth shank nail @ 6 in. / 12 in. Crystal River #1 2 2 A-1 2 in. smooth shank nail @ 6 in. / 12 in. Port Orange #2 3 1 B-2 2 in. smooth shank nail @ 6 in. / 12 in. Bartow #3 2 2 A-1 2 in. smooth shank nail @ 6 in. / 12 in. Port Orange #3 2 1 A-1 2 in. smooth shank nail @ 6 in. / 12 in. Bartow #4 1 2 A-1 2 in. smooth shank nail @ 4 in. / 4 in. Bartow #5 1 1 A-1 2 in. smooth shank nail @ 3 in. / 6 in. Port Orange #6 2 NA NA 2 in. smooth shank nail @ 8 in. / 8 in. Port Orange #5 2 NA NA 2 in. smooth shank nail @ 6 in. / 12 in. Port Orange #4 2 NA NA Notes: Retrofit A-1 is 8d ring shank nails installed at 6 in. / 12 in. spacing, Retrofit A-2 is 8d ring shank nails installed between a 6 in. / 12 in. spacing, and Retrofit B-2 is ccSPF installed in a full 3 in. thick layer Figure 4-11. Laboratory fa bricated panel specific gravity sample locations 73

PAGE 74

Framing SG Samples Sheathing Figure 4-12. Harvested panel specific grav ity and moisture content sample locations Sheathing Member Framing Plastic Negative Pressure Tape Figure 4-13. Panel installed in pressure chamber with negative pressure setup 74

PAGE 75

C-Clamp Steel Frame & Gasket Sheathing Wood Blocking Framing Member Shim Plastic Positive Pressure Figure 4-14. Panel installed in pressure chamber with positiv e pressure setup A B Figure 4-15. Manual control of pr essure chamber (A gate valve c ontrol and (B vacuum pump in series 75

PAGE 76

76 Figure 4-16. Pressure Load Actuator (PLA) Figure 4-17. Comparison of dynamic target vs. actual chamber pressure used with the PLA

PAGE 77

CHAPTER 5 RESULTS AND ANALYSIS This chapter presents results from wood roof sheathing uplift pressure testing of panels fabricated in the laboratory and harvested from existing LFWS in Florida and analysis corresponding analysis. The effects of fastener type, fastener spacing, in-service conditions, retrofit measures and dynamic loading on panel up lift resistance are evaluated. Testing is broken into three groups (1) panels fabricated in the la boratory tested statica lly (2) panels harvested tested statically and (3) panels fabricated in the laboratory and harvested tested to directly compare static vs. dynamic loading. Laboratory Static Panel Uplift Tests Static testing of laboratory fa bricated panels is conducted using the UF-WRSUT developed in Chapter 3. The maximum chamber pressure during testing is measured with the Omega peak measurement pressure gauge and is reported as fa ilure capacity. The effects of fastener type, fastener spacing and ccSPF retrofit on newly c onstructed roof sheathing panel uplift resistance are investigated. Evaluation of Fastener Size and Spacing on Uplift Capacity Table 5-1 presents mean uplift capacities of panels constructed with 8d ring shank, 8d smooth shank and 2-3/8 in. long 6d smooth shank nails. Panels fast ened with 8d ring shank and 8d smooth shank nails are constructed with 6 in. / 12 in., 6 in. / 8 in. and 6 in. / 6 in. fastener schedules. Panels constructed wi th 2-3/8 in. 6d smooth shank nails are constructed with only a 6 in. / 12 in. fastener schedule. A control series of 8d ring shank nails which is used for evaluation of ccSPF retrofit discussed belo w is also presented. Figure 5-1 shows graphically the mean uplift failure capacities of laboratory fabricated panels tested with 8d ring shank, 8d smooth sha nk and 6d smooth shank nails at the three nailing 77

PAGE 78

schedules used. It is seen that as interior fastener spacing decreases the mean uplift capacity increases. Panels fastened with 8d ring shank nails increase 45% to 57% in mean uplift strength (174/161 to 252 psf) from 6 in. / 12 in. to 6 in. / 6 in. fastening schedules. Panels fastened with 8d smooth shank nails increased 59% in mean uplif t strength (129 psf to 205 psf) from 6 in. / 12 in. to 6 in. / 6 in. fastening schedules. Additio nally it is found that this relationship between uplift capacity and fastener spacing appears to be linear for panels attached with 8d ring shank or smooth shank nails. The slopes for 8d ring shank and 8d smooth shank are 13.6 psf/in. and 12.5 psf/in. respectively suggesting that the relations hip between fastener sp acing and uplift capacity is independent of fastener type. Ring shank nails are found to have higher uplif t strength than smooth shank nails. Results show a 27% mean increase in st rength from 8d smooth shank to 8d ring shank. Additionally it is found that 8d smooth shank nails have a higher uplift capacity than 2-3/8 in. 6d smooth shank nails. Only the 6 in. / 12 in. fastener schedule was tested for 2-3/8 in. long 6d common nails so no conclusive trends can be f ound; however it was found that uplift strength is increased 74% from 6d smooth shank nails to 8d smooth shank nails. Failure pressure results of la boratory fabricated panel tested statically provide a control series which can be used for comparative purposes. In the following section results are compared to Previous studies for assessmen t of developed test method, Design fastener withdrawal strengths fo r evaluation of design assumptions, ccSPF attached panel for evaluation of strength benefits of ccSPF, Harvested panels for evaluation of in-service effects, Retrofitted harvested panels to evaluate strength benefits, Dynamically tested panels to determ ine the effects of dynamic loading. 78

PAGE 79

Comparison of Results with Previous Studies Results from static UF-WRSUT are compared w ith results identified in Chapter 2 in order to assess the relative performance of the devel oped test method with previous methods. Due to difficulties in making direct comparisons previously discussed in Chapter 2 results presented in Table 5-1 are compared to ranges of similar constr uction presented in litera ture review. Previous studies mean and 5% exclusion failure pressu res with sufficient numbers of samples are compared to results. Panel results from previous studies constructe d with 6d smooth shank fasteners attached at 6 in. / 12 in. yielded mean failure pressures rangi ng from 26 psf to 55 psf, which is lower than static UF-WRSUT results having a mean failure pre ssure of 74 psf. Panel results from previous studies constructed with 8d smooth shank fasteners attached at 6 in / 12 in. and 6 in. / 6 in. had ranges of mean failure pressures of 61 psf to 228 psf and 77 psf to 218 psf respectively. Static UF-WRSUT results of panels fastened with 8d smooth shank nails at 6 in. / 12 in. yielded a mean failure pressure of 129 psf and panels attached at 6 in. / 6 in. had a mean failure pressure of 205 psf; both of which fall within the corres ponding ranges found from literature. Panels constructed with 8d ring shank fasteners had onl y one data set for each spacing of 6 in. / 12 in. and 6 in. / 6 in. Results from previous testing of panels constructed with 8d ring shank nails at 6 in. / 12 in. had a mean failure pressure of 140 ps f, and the one panel fastened with 8d ring shank nails at 6 in. / 6 in. had a failure pressure of 397 psf. Panels constructed with 8d ring shank nails tested per UF-WRSUT for 6 in / 12 in. and 6 in / 6 in. fastener schedules resulted in mean failure pressures of 174 psf and 252 psf, which ar e lower than results from previous testing. Table 5-2 presents a comparison of the mean and 5% exclusion fa ilure pressures of laboratory fabricated static UF-WRSUT results and results of previous studies which had a sufficiently large sample size. Sample sizes from statically tested laboratory fabricated panels as 79

PAGE 80

part of this study were 13 to 15 panels and for comparative purposes prev ious studies selected had similar or larger sample sizes ; details of selected studies can be found in Chapter 2. The 5% exclusion failure pressure is presented to give a comparison of the variability in results. A statistical analysis of laboratory fabricated stat ic UF-WRSUT found that a ll data sets fit a normal distribution at the 0.05 confidence level using an Anderson-Da rin test. Using the normal cumulative distribution function the 5% value was calculated for results fr om static UF-WRSUT and all previous studies sele cted had normal distributions. Static UF-WRSUT results from panels instal led with 6d smooth shank fasteners at 6 in. / 12 in. are significantly higher than results from Kallem (1997), which had 14 samples, for both mean and 5% exclusion failure pressures. Panels attached with 8d smooth shank fasteners at 6 in. / 12 in. UF-WRSUT mean failure pressure re sult is within 18% of a 49 panel data set by IHRC (IHRC 2004) of similar construction. Stati cally tested UF-WRSUT panels attached with 8d smooth shank fastener spaced at 6 in. / 6 in (205 psf mean and 170 psf 5% exclusion) are significantly higher than a 30 pane l data set (131 psf mean and 101 psf 5% exclusion) of similar construction by Murphy et al. (1996) Statically tested UF-WRSUT panels attached with 8d ring shank fastener spaced at 6 in. / 12 in. (174 psf mean and 137 psf 5% exclusion) are significantly higher than a 50 panel data set (140 psf mean and 101 psf 5% excl usion) of similar construction by IHRC (IHRC 2004). Results presented in Tables 5-1 and 5-2 show th at laboratory fabricated panels tested with static UF-WRSUT yield failure pressures whic h in the same range as previous results considering that direct comparis ons cannot be made. First, UF -WRSUT results are both above and below results from previous studies. Second results from previous studies vary by the same magnitude between each other as they do to UF-W RSUT results. Third, the coefficients of 80

PAGE 81

variation for static UF-WRSUT results (7% to 22%) are similar to previous testing (7% to 28%). Finally, while it is found that while difference in mean failure pressures can vary significantly the closest match with previous studies are from data sets as large or larger than those tested with UF-WRSUT. Given that result are comparable to previous studies, which wood roof sheathing panel design is based on, UF-WRSUT result s can be compared with design values. Evaluation of Calculated Fastener Resistance Table 5-3 presents mean fastener failure loads calculated based on the assumption of tributary area for laboratory fabric ated panels tested statically, where the maximum tributary area of fasteners installed in the panel is used. Panel failure pressure are dived by a factor of safety of 2.0 for design pressure then multiplied by the maximum tributary area to obtain calculated withdrawal strengths. Resulting failure loads are compared to design withdrawal strength based on NDS strengths (Table 11.2C) (AF&PA 2005). Withdrawal strengt hs from the NDS are multiplied by the built in factor of safety (5) and depth of penetration for each fastener, which is 2 in. for 8d smooth shank or ring shank nails and 1.875 in. for 2-3/8 in. 6d smooth shank nails. For example, 8d smooth shank fasteners have a design withdrawal value of 41 (lbs/in.) and a depth of penetration of 2 in. when installed through in. OSB, is lbs SFin in lbs 410)(5.)(2 41 (5-1) Figure 5-2 summarizes the effect of fastener spacing on mean panel failure pressure vs. mean fastener failure load. Calculated fastener failure load results show that as nail spacing is reduced mean withdrawal loads decrease, for bot h the 8d ring shank and sm ooth shank nails. It is observed that failure loads for the 8d ring sh ank nail decreased from 175 lbs to 126 lbs as the interior fastener spacing reduced from 12 in. to 6 in. o.c. Similarly nail failure loads decreased from 130 lbs to 103 lbs for 8d common nail as fasten ing spacing decreased from 12 in. to 6 in. 81

PAGE 82

Calculated fastener failure resu lts are significantly lower than design withdrawal strengths found from NDS strengths, ranging from 67% to 77% less for 8d ring shank and 2-3/8 in. 6d smooth shank respectively. Results suggest that 8d ri ng shank nails are roughly 27% greater than 8d smooth shank nails, however design withdrawal strength for 8d ring shank nails are 6% less than 8d smooth shank nails. Results suggest that fastener withdrawal strength increases as spacing increases. However a study by Dao and van de Lindt (2008), in which the effect of fastener spac ing is investigated by quantifying the applied moment of a nail due to the changes in eccen tricity/spacing, suggests that fastener withdrawal loads would decrease with in creased fastener spacing. These findings, while preliminary are important as they presents an al ternative interpretation of test results from the commonly held notion that nail failure load is not changeable by the structural system itself. Results suggest that current approaches are not m odeling the behavior of fa stener strength within a panel. The limitation of the results is that no measurement of fastener load at the time failure were made to confirm the estimates based on failure pressure. It is likely that a large sample set is needed in order to investigate this behavior therefore a standardized test method is necessary to provide consistent results. Additionally it is likely that dynamic characteristics of wind loading effect this behavior and are theref ore necessary to captur e the true behavior. Evaluation of ccSPF Retrofit The need to address deficiencies in uplift resistance of the current building stock is established in Chapter 2. Two retrofit measures are investigat ed within this study first 8d ring shank nails installed at a fastener schedule of 6 in. / 12 in. and second ccSPF adhesive. Laboratory fabricated panels attached with ccSPF adhesive are tested statically to establish a baseline of their performance, Table 5-4. 82

PAGE 83

The mean uplift capacity of panels connected with 8d ring shank nail at 6 in. / 12 in. was 161 psf with a coefficient of va riation of 18%, which compares well with results presented in Table 5-1. This mean uplift capacity is increased to 202 psf when a fillet of ccSPF (ret. B-1) is applied resulting in a 25 % increase capacity. When ccSPF was applied in a full 3 in. thick layer (ret. B-2) the increased effect wa s negligible with the mean uplift capacity increasing to 209 psf. Results suggest that from the standpoint of improving structur al strength there is little added benefit from the fillet to full coverage of ccSPF. Results provide a baseline for comparison with, Harvested panels retrofitted ccSPF to de termine the effects of adhesion with aged panels, New and harvested panels retrofitted w ith 8d ring shank nails to compare the strength benefits. Failure Mechanisms Observed As detailed in Chapter 2 the uplift desi gn of wood roof sheathing is based on the withdrawal strength of the indivi dual fasteners. In order to assess the validity of this assumption that fastener connecting wood roof sheathing to framing fail in wit hdrawal panels were examined after testing for failure mode and location of all failures. Figure 5-3 presents a sample of the failure mode and location information of 8d smoo th shank and ring shank nails at 6 in. / 6 in., which is collected for all laboratory fabricated panel tested statically. The framing members attached to the 4 ft. by 8 ft. sheathing is represented by the lines, on-top of which a shape corresponding to a particular failure mode is plac ed on each fastener that failed. Observations were made after panel was removed from the ch amber. Full results are found in Appendix B. The dominant failure mode of laboratory fabricated panels tested with the static UFWRSUT is presented in Figure 5-4. It is found that failure of OSB panels fastened with smooth shank nail tend to be dominated by the withdrawal failure mode. Altern atively it is found that 83

PAGE 84

panels attached with ring shank nails tend to be dominated by the pull-through failure mode. Additionally it is seen that as fastener spacing decreases pull-through failure mode becomes more dominant, however the effect of spacing is less significant than fastener type. The addition of ccSPF as a retrofit measure is seen to cause a more even distribution of panel failure modes. Results will provide a basis for comparison of behavior of fasteners during failure. Comparisons of failure behavior will be made for new and harvested panels attached with different fasteners tested statically vs. dynamically. Harvested Static Panel Uplift Test Static testing of panels harvested fro m 12 homes throughout Central Florida was conducted using the UF-WRSUT developed in Ch apter 3. The maximum chamber pressure during testing recorded with the Omega peak meas urement pressure gauge is reported as failure capacity. The effects of age/weathering, fastener type, fastener spac ing and retrofitting on existing panel constructions uplif t resistance are investigated. Evaluation of Existing Panel Uplift Resistance Table 5-5 presents full results for static UF -WRSUT panels harvested from LFWS located in Central Florida in their existing condition, a nd the fastener failure loads calculated by the assumption of tributary area. Fastener failure loads are calculated by multiplying the failure pressure by the maximum tributary area in the panel; for example the 1.5 in. staples at 4 in. / 4 in. have a maximum tribut ary area of 2/3 ft2 which is multiplied by the failure pressure of 93 psf to get a fastener failure load of 62 lbs. It is seen that panels attached w ith 2 in. smooth shank nails at 4 in. / 4 in. had the highest uplift capacity, which corresponds to the smallest tributary area seen. The fastener with the highest withdrawal capacity is the 2.5 in. (8d) smooth shank nail, which corresponds to the larges t tributary area and fastener size. The most common panel attachment harvested was 6d smooth shank nails spaced at 6 in. / 12 in. having a range of mean 84

PAGE 85

uplift capacities of 39 psf to 102 psf, which contai ns the laboratory fabricat ed result of 74 psf. Harvested panels attached with 8d common smooth shank nails at 6 in. / 12 in. had a mean uplift capacity of 107 psf which considering the small sa mple size (2) is sim ilar to the laboratory fabricated result of 129 psf Figure 5-5 presents a comparison of the sca tter of static uplift capacity results from harvested vs. new panels. Harvested panels att ached with 1.5 in. staples are attached at two fastener schedules having tributar y areas of two thirds and one ft2, which yield similar results. Harvested panels attached with 2.5 in. staples ar e attached at a 6 in. / 12 in. schedule and yield results lower than the 1.5 in. staples, which is reasonable considering that due to tributary area the 2.5 in. staples experience load levels as much as three times greater than the 1.5 in. staples. Harvested panels attached with 2 in. 6d smooth sh ank fasteners are attached with four fastener schedules (2, 1-1/3, 1 and 2/3 ft2). As discussed in Chapter 4 panels constructed with 6d smooth shank should nominally be attached with a 6 in / 12 in. schedule however spacing is somewhat variable in the field. Comparing harvested and new 6d smooth shank fastened panels it is seen that harvested panels have a sli ghtly higher mean but contain laborat ory fabricated panel results. This increased capacity of harvested panels can be attributed to the smaller loads on individual fasteners due to the closer spacing of harveste d panels. Harvested panels attached with 8d smooth shank fasteners are attached at a 6 in. / 12 in. schedule and have 17% lower mean uplift capacity than new panels, however only two samp les exist for harvested results and both fall within range of new panel results. Figure 5-6 presents a comparison of mean uplift capacity vs. calculated fastener withdrawal load for panels attach ed with staples and nails at different fastening schedules. The change in fastening is best thought of as change in tributary area becaus e fastener withdrawal 85

PAGE 86

loads are calculated by multiplying the tributary area by panel uplift capacity. Panels attached with 1.5 in. staples experienced a decrease in mean uplift capacity and calculated fastener load as tributary area decreases. Panels attached with 2 in. smooth shank nails had similar mean uplift capacities for fastening schedules 6 in. / 12 in. to 3 in. / 6 in., but calculate d fastener withdrawal strength steadily decreased. Fr om fastening schedules 3 in. / 6 in. to 4 in. / 4 in. panel uplift capacity increased and fastener withdrawal stre ngth increased slightly, however only a single sample was tested at 4 in. / 4 in. spacing so this result may not be representa tive. This trend of decreasing fastener load as tri butary area decreases is confirmed in laboratory fabricated static test results. Results from panels harvested from existing structures and laborator y fabricated panels presented above provide a data for the evalua tion of age or weathering effects on wood roof sheathing panel uplift capacity. Th ese results show that statica lly tested uplift capacities of existing wood roof sheathing panels have similar uplift capacities as newl y fabricated panels. This suggests that laboratory testing of newly fabricated roof sheathing panels can sufficiently approximate the in-service strength of wood roof sheathing panels. Evaluation of Retrofit of Existing Panels Table 5-6 presents mean uplift capacity re sults from a comparative study of panels harvested from existing LFWS located in Central Fl orida in existing condition vs. retrofitted. All panels are tested statically per UF-WRSUT. Pa nels are retrofitted with 8d ring shank nails spaced at 6 in. / 12 in. installed as if a panel had no existing fasteners (Retrofit A-1). When existing fastener locations coincided with the 8d ring shank retrofit fastener locations nails were installed directly next to the existing fastener. One panel was retrofitted with a full 3 in. thick layer of ccSPF adhesive (Retrofit B-2). 86

PAGE 87

Figure 5-7 presents a comparison of statically tested harvested panels in existing and retrofit conditions. Results are averaged by indi vidual source houses to account for any variation between source houses of similar construction in the comparison. Harvested panel which were retrofitted were found to have sign ificantly increased uplif t capacities. Panels retrofitted with 8d ring shank nails spaced at 6 in. / 12 in. had an average increase in capacity of over 2.5 times. Variations in retrofitted uplift capacity are seen to be independent of the existing fastener. Panels retrofitted with ccSPF retrofitted increased 2.4 time howev er only 1 sample was tested. Results presented in this section provide da ta for the evaluation of the initial strength benefits of retrofitting existing construction. It is seen that retrofitting with one of the two methods used can increase the up lift capacity of existing panels immediately after retrofits are installed significantly. Comparison of Laboratory vs. Harvested Specific Gravity Samples Specific gravity measurements (167 laborator y samples and 178 harvested samples totaling 345 samples) were taken for all panels tested in static vs. dynamic testing. The mean specific gravity of laboratory samples is 0.52 and 0.57 for harvested panels, which is similar to the NDS specified specific gravity for southern yellow pine of 0.55. However, the mean values of two sets of specific gravity are not statisti cally equal (p-value on the order of 10-8). The specific gravities of the framing members used in the lab-built panels were best fit by a lognormal distribution (Figure 5-8a), and for the harvested panels (15 years old), a normal distribution was a better fit (Figure 5-8b). Ra mmer et al. (2001) found that norm al and lognormal distributions fit the specific gravity of southern yellow pine. Table 5-7 presents specific gravities calculat ed in an attempt to determine if framing member specific gravity affects the wind uplift cap acity of wood roof sheathing. Three specific gravities are calculated; the first is a mean of all framing member s, the second is a mean of all 87

PAGE 88

framing members which failed and the third is the mean of the framing members which were observed or determined to fail first. In the th ird case if it was not possible to observe which framing member failed first, then the framing members with relatively larger numbers of fastener failures were used. The normalized panel failure pressure was calculated by dividing each panel failure pressure by the maximum failure pressure for the given data set. It is found that the specific gravity of fr aming members collected from in-service wood roof systems are higher than wood framing member s which were structural #2 grade or higher. This suggests that there is no degradation of wood properties once installed in the field. An analysis of the effects of specific gravity to uplift capacity suggests no correlation, however limitations in testing make it difficult to pinpo int the initial failure and corresponding specific gravity. It is found that if specific gravity effects the uplift capacity of wood roof sheathing it is not a dominant factor. Static vs. Dynamic Panel Uplift Test The effects of dynamic loading characteristics on existing and retrof itted panels both fabricated in the laboratory and harvested from existing construction are investigated. Laboratory fabricated panels fast ened with 6d smooth shank fasteners at 6 in. / 12 in. and 6 in. / 6 in. schedules from the same sample population are tested both statically and dynamically. Then laboratory fabricated existing panels attached with the 6d smooth shank fasteners at 6 in. / 12 in. retrofitted with 8d ring sh ank nails are tested both statically and dynamically. Panels harvested from existing construction in Central Florid a fastened with 1.5 in. staples at 4 in. / 4 in. are tested both statically and dynamically. Harvested panels ar e retrofitted with 8d ring shank nails and tested statically and dynamically. All static vs. dynamic samples are tested with the PLA system detailed in Chapter 4. Results reco rded are pressure time-histories framing member 88

PAGE 89

specific gravity, failure mode and location, mo isture content for only harvested panels and defects in harvested panel constr uctions (i.e. missing fasteners). Comparison of Static vs. Dynamic Laboratory Fabricated Panel Uplift Test Table 5-8 presents results from static vs dynamic UF-WRSUT of laboratory fabricated panels. Failure pressures of statically tested laboratory fabricated panels attached with 2 in. long 6d smooth shank nails spaced at 6 in. / 12 in. (62 psf) are similar to results obtained in static UFWRSUT results (74 psf). Similarly statically te sted panels retrofitted with 8d ring shank nails yield a mean failure pressure (183 psf) close to that of static UF-WRS UT results of similar construction (174 psf and 161 psf). Figure 5-9 presents a comparison of statically vs. dynamically tested panels fabricated in the lab. As expected from sta tic testing of panels fabricated lab, panels having 6 in. / 12 in. fastening schedules failed at lower pressures than the panels having 6 in. / 6 in. fastening schedules for both static test s (62 psf vs. 108 psf) and dynami c tests (52 psf vs. 90 psf). Similarly as expected from static testing of panels retrofitted w ith 8d ring shank nails, panels retrofitted with 8d ring shank nails had a higher failure pressure than in existing condition for both static (62 psf vs. 183 psf) and dynamic (52 psf 167 psf) loading. A reduction of 16% was found in uplift capacity of pane ls in existing condition tested dynamically. The capacity of retrofitted panels was reduced 9% from static to dynamic loading. Results from this section provide a comparison of statically vs. dynamically tested panels, and the effect of dynamic loading on new 8d ring shank retrofitted panels It is seen that dynamic loading reduces the uplif t capacity of new panels co nstructed with smooth shank fasteners 16% and new panels constructed with ring shank fasteners 9%. This suggests that dynamic loading reduces wood roof sheathing uplift strength and that the e ffects are lessened for annularly threaded nails. 89

PAGE 90

Comparison of Static vs. Dynami c Harvested Panel Uplift Test Table 5-9 presents results from static vs. dynamic UF-WRSUT of panels harvested from a single existing LFWS in Florida. Failure pressures of statically tested harvested panels attached with 1.5 in. long staples (93 psf) are less than results obtained in static UF-WRSUT results (120 psf). Similarly statically tested panels retrofitt ed with 8d ring shank nails yield a mean failure pressure (168 psf) less than that of static UF -WRSUT results of similar construction (270 psf). However retrofitted panel uplift capacity fall with in the range of statically tested laboratory fabricated panels attached w ith 8d ring shank nails at 6 in. / 12 in. (174 psf and 161 psf). Figure 5-10 presents a comparis on of statically vs. dynamically tested panels harvested existing construction. As expected from static testing of panels retrof itted with 8d ring shank nails, panels retrofitted with 8d ring shank nails had a higher failu re pressure than in existing condition for both static (93 ps f vs. 168 psf) and dynamic (72 psf vs. 126 psf) loading. A reduction of 23% was found in uplift capacity of pane ls in existing conditi on tested dynamically. The capacity of retrofitted panels was reduced 26% from static to dynamic loading. Results from this section provide data for th e evaluation of the effect of dynamic loading on harvested panel in existing and retrofitted conditi ons. It is seen that dynamic loading reduces harvested panels in existing conditions by 23% and panels in retrofitted conditions by 26%. This increase in effect for panels retrofitted with 8d ring shank nails is contradictory to laboratory fabricated panel results, however as detailed in Chapter 4 the panels used for retrofitting were the panel with most missing fasteners. Thus harv ested retrofitted panels effectively had larger tributary areas than those of laboratory fabricated retrofitted panels making them more susceptible to dynamic loading. However limited sa mple sizes prevent conclusive findings. It is clear that dynamic loading effects the up lift capacity of wood roof sheathing panels. 90

PAGE 91

Peak Pressure vs. Failure Pressure Table 5-10 presents a summary of the peak pres sures and the failure pressure of all panels tested statically vs. dynamically. This is investigated to determine if panel failure occurs at a peak pressure or later at some reduced pressure. Dynamically tested results are the focus of this comparison but statically tested results are provid ed for comparison. If a panel fails at a peak pressure or pressure spike then the failure is assumed to be due to the large loading rate. However if a panel fails after a pr essure spike at a lower pressure then failure is assumed to be due to some accumulation of damage in the nail to wood connection. It is found that half (9) of the dynamically tested panels failed at a pressure spike and are therefore due to the loading rate. It is found that the half of the dynamically tested panels failed after a pressure spike at a reduced pressure, which were on average is 87 % of the peak pressure recorded. For dynamically tested panels the failure pres sure compared to peak pressure it can be seen that the failure pressure is lower than the pe ak pressure by 3-9%. St atically tested panels show little difference from peak to failure pressure. Therefore the reduction from dynamically to statically tested panel resistance is increased wh en using the failure pressure instead of peak pressure, which is what has typically been used in previous static tests. Table 5-11 presents the reductions from static to dynamic re sistance for both the peak and fa ilure pressures. As expected the magnitude of the reductions from static pane l resistance are increased when calculated with the failure pressures Analysis of Variance of Results Results are analyzed to determine if the contro l variables fastener type, fastener spacing, age, date tested, retrofit and type of loading have a statistically signi ficant effect on failure pressure in this section. This is done by analyzin g the variance within data series to test the null hypothesis that all the samples co me from the same sample population. Analysis of Variance 91

PAGE 92

(ANOVA) test is a measure of the error betw een some number of data series by judging statistically if the differences in the sample means are different. With the limited number of samples there is a risk that the theoretical behavior of a combination of variables is not being captured in results therefore the analysis is cond ucted for datasets of four samples or greater. Similarly the methods employed assume that the distribution of the results is normal which has only been verified for laboratory fabricated samp les in the static phase of testing constructed with 8d ring shank, 8d smooth shank and 6d smooth shank. All series used in analysis of variation are presented in Table 5-12. Procedure To analyze the variance in the different data sets the comparison is broken down into two parts (1) the error due to comparison or treatmen t and (2) the error due to the differences in individual samples within a given dataset. The general procedure to analyze the variance (ANOVA) of several datasets is to first calculated the sum of th e squares error for the treatment (SST) by, i iyynSST2 ... (5-2) Where iy is the mean of an individual series, ..y is the mean of all the series combined, n is the number of samples in the series and i is the number of the current treatment. Next the sum of the squares error for those variations which are not due to the treatment (SSE) by, ij iijyySSE2 (5-3) Where is the jth term of the ith series or treatment and ijy .iy is the mean of the ith series or treatment. Then unbiased estimates of the varian ce of the error between se ries are calculated for the error due to the treatment (M ST) and the error due to differ ences in sample means (MSE), 92

PAGE 93

1 t SST MST (5-4) tN SSE MSE (5-5) Where t is the number of treatments and N is the total number of samples in all series. Therefore the ratio of the two unbiased mean square errors MST to MSE (F-value) should approach 1.0 if the series were from similar sa mple populations expressed as, MSE MST valueF (5-6) Simply stated this means that if the variations of the mean values of each dataset about the overall mean are the same as the variations of the individual sample s to their corresponding mean, then the samples can be said to come fr om the same population. To assess the probability that the F-value is close to 1.0 a p-value analysis is used where the probability that the F-value will be close to the expected value (p-value) is determined and compared to a level of confidence. To assess the probabili ty that the F-value will be cl ose to 1.0 the p-value must be less than the confidence limit For example if a p-value is gr eater than 0.05 then one can say the data is from the same population with 95% confidence ( %1001%95 ). A critical Fcritical-value is calculated corresponds to the confidence limit meaning that if the F-value is greater than the Fcritical-value then the datasets come from different sample populations. To further analyze the result the Bonferroni Meth od (Ott and Longnecker 2004) is applied to determine if individual datase ts are similar to each other. The Bonferroni Method compares the variance between datasets in the same way described above except that it does this for each individual dataset. This is done by adjus ting the confidence interval for the individual comparisons by dividing the by C, where C is, 93

PAGE 94

2 1 II C (5-7) and I is the number of treatments or comparisons. By dividing the by C the problem becomes more conservative, therefore with large numbers of comparisons th ere is a risk of rejecting the null hypothesis when it is true. The problem is evaluated by adjusting the comparison of the series means by adding and subtracting the critical t-value (eq. 5-8). ji CNjiJJ MSE tXX 112/,1.. (5-8) Where is calculated from the Students t-table, )2/(,1CNt iX is the mean of the series i, is the number of observation in series i, iJ jX is the mean of the series j and is the number of observation in series j. If zero lies between the limits of equation 6 then the series means are found from the same population because the differe nce is zero within th e confidence interval. jJLaboratory Fabricated Panels Tested Statically A comparison of the variances between all laborato ry fabricated panel datasets in the static phase of testing is conducted using the procedure described ab ove with to determine if the different datasets came from the same sample population. Table 5-13 presents the ANOVA table which finds that the datasets come from different sample populations. Table 5-14 presents a full comparison of the individual datasets to each other using the Bonferroni Method, which is summarized in Table 5-15. It was found that the two 8d ring shank series spaced at 6 in. / 12 in. are from the same sample population. It wa s found that the two ccSPF retrofit arrangements (fillet and full 3 in.) were from the same samp le population. Additi onally it was found that 8d ring shank nails at 6 in. / 12 in. are from the same population as 8d smooth shank nails at 6 in. / 8 94

PAGE 95

in. Similarly 8d ring shank nails at 6 in. / 8 in are found to be from the same population as 8d smooth shank nails at 6 in. / 6 in. All rema ining populations were found to be different. Harvested Panels Tested Statically A comparison of variances for all panels constructed with 6d smooth shank panel spaced at 6 in. / 12 in. was conducted to determine if harves ted panel results came from the same sample population as newly constructed results. This comparison included ha rvested and laboratory fabricated panels from the static phase of testing, as well as laboratory fabr icated panels from the static vs. dynamic phase of testing. Table 5-16 presents the ANOVA table for the four datasets. It is found that all datasets constructed with 6d smooth shank nails at 6 in. / 12 in. come from the same sample population. Table 5-17 presents the full results for comparison using the Bonferroni Method, which suggest again that all datasets come from the same sample population. A comparison of all statically loaded panels retrofitted with 8d ring shank nails spaced at 6 in. / 12 in. versus existing panel with 8d ring sh ank nails spaced at 6 in. / 12 in. was conducted using the unequal variance t-test. The t-test is identical to the analysis of variance procedure described above except that only two samples are compared. The full T-table is presented in Table 5-18 and it is found that the two results are not from the same sample population for a confidence level of 95%. This suggests that th e original fasteners in stalled at the time of construction make a difference Laboratory Fabricated Panels Tested Statically vs. Dynamically A comparison of variances for panels fastened with 6d smooth shank nails spaced at both 6 in. / 12 in. and 6 in. / 6 in. from the static vs. dynamic phase of testing are compared. The peak pressure recorded during a test and the pressure which the panel failed at were both used. Table 5-19 and Table 5-21 present ANOVA tables for th e peak and failure pressures respectively, 95

PAGE 96

which both suggests that all stat ically and dynamically tested pa nels regardless of fastener spacing come from the same sample population. Table 5-20 and Tabl e 5-22 present the Bonferroni analysis for the peak and failure pr essures respectively, and again both suggest that statically and dynamically tested panels are from the same population for both 6 in. / 12 in. and 6 in. / 6 in. spacings. 96

PAGE 97

Table 5-1. Results of static UF-WRSUT of labora tory fabricated panels fastened with 2-3/8 in. long 6d smooth shank, 8d smooth shank and 8d ring shank nails Fastener Schedule 8d Ring Shank Nails 8d Smooth Shank Nails 6d Smooth Shank Nails # of Panels Mean Failure (psf) COV # of Panels Mean Failure (psf) COV # of Panels Mean Failure (psf) COV 6 in. / 12 in. 10 161.0 18% 15 129.4 11% 15 74.3 22% 15 174.4 13% 6 in. / 8 in. 15 216.1 17% 15 175.1 11% NA NA NA 6 in. / 6 in. 13 252.4 7% 15 205.3 10% NA NA NA Notes: Notes: Ring Shank Nails 2.5 in. long 0.113 in. dia., 8d Smooth Shank Nails 2.5 in. long 0.131 in. dia., and 6d Smooth shank 2.375 in. long 0.113 in. dia. Figure 5-1. Mean failure pre ssures of laboratory fabricated panels tested statically 97

PAGE 98

Table 5-2. Comparison of mean and 5% exclusion value failure pr essures for panels fabricated in the lab tested statically UF-WRSUT vs. previous studies Panel Attachment Current Study Previous Results Mean Capacity 5% Exclusion Value Mean Capacity 5% Exclusion Value 6d Smooth Shank @ 6"/12" 74 47 (A) 33 21 8d Smooth Shank @ 6"/12" 129 105 (B) 110 N/A 8d Smooth Shank @ 6"/6" 205 170 (C) 131 101 8d Ring Shank @ 6"/12" 174 137 (B) 140 101 NOTES: A Study by Kallem (1997) w/ plywood, B Study by IHRC (1004) w/ plywood and C Study by Murphy et al. (1996) w/ OSB Table 5-3. Mean of calculated maximu m fastener loads based on tributary area Fastening Schedule 6 in. / 12 in. (lbs) 6 in. / 8 in. (lbs) 6 in. / 6 in. (lbs) Design Withdrawal Load (lbs) 8d Ring Shank 175 144 126 387 8d Smooth Shank 130 117 103 410 6d Smooth Shank 75 NA NA 328 Notes: (1) Specific gravity of 0.55 assumed for design withdrawal load (2) Ring shank (0.113 in. diameter) design withdrawal strength determined by linear ex trapolation (38.67 lbs/in.) (AF&PA 2005) 98

PAGE 99

99 0 50 100 150 200 250 300 350 400 6 in. / 12 in.6 in. / 8 in.6 in. / 6 in Fastener ScheduleMean Failure Pressure (psf). Mean Failure Pressure (psf) Mean Fastener Load (lbs)Mean Failure Load (lbs)0 50 100 150 200 250 300 350 400 A 0 50 100 150 200 250 300 350 400 6 in. / 12 in.6 in. / 8 in.6 in. / 6 in. Fastener ScheduleMean Failure Pressure (psf) Mean Failure Pressure (psf) Mean Fastener Load (lbs)Mean Failure Load (lbs)0 50 100 150 200 250 300 350 400 B Figure 5-2. Summary of mean pa nel failure (psf) vs. calculated fa stener failure (lbs) for A) 8d ring shank nails and B) 8d smooth shank nail s attached at 6 in. / 12 in., 6 in. / 8 in. and 6 in. / 6 in. schedules Table 5-4. Results of static UF-WRSUT of labo ratory fabricated panels fastened with 8d ring shank nails at 6 in. / 12 in. re trofitted with ccSPF adhesive # of Panels Mean Failure Pressure (psf) COV Control Series 10 161 18% Fillet Retrofit (B-1) 10 202 14% Full 3 in. Retrofit (B-2) 10 209 19%

PAGE 100

Board Split Partial Withdraw Pull Through Full Withdraw 195.2 psf Moderate#39 193.8 psf Fast failure#38 222.1 psf Fast failure#37 195.9 psf Moderate#36 232.7 psf Fast failure#35 239.1 psf Fast failure#33 176.1 psf Fast failure#34 213.6 psf Moderate#44 193.1 psf Fast failure#45 188.8 psf Fast failure#31 163.4 psf Moderate#32 207.2 psf Fast failure#43 226.3 psf Fast failure#41 223.5 psf Fast failure#42 Legend 208.6 psf Fast failure#40 8 in 4 ft 8 ft 2 ft15 samples Mean : 205.3 psf Max : 239.1 psf Min : 163.4 psf STD : 21.35 psf COV : 0.104 Test dates : 10/22/2007 and 10/25/2007 Spacing : 6 inches O.C. interior ( 9 total) and at 6 inches O.C. edge (9 total) Fastener : 8d smooth shank 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 8d SS 8d RS12 0 3 Total Number of Panels Withdrawal Even PullThrough Fastener% of Panel Failures PTPT WWWWWWWW PT WWWW Panel Failure Mode 34 2013 6 11 6879 10 78 13 9 # of Withdrawal Failures 6575352142 11 2352 # of Pull-Through Failures 4544 4342 4140393837 3635 3433 3231 Panel Figure 5-3. Sample of failure mode and location calculations from Statically tested 8d smooth shank nails at 6 in. /12 in. 100

PAGE 101

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 6d SS8d SS Fastener% of Failure Mode 8d RS A 6 in. / 12 in. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 8d RS fillet8d RS full Fastener% of Failure Mode B ccSPF Retrofit 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 8d SS8d RS Fastener% of Failure Mode C 6 in. / 8 in. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 8d SS8d RS Fastener% of Failure Mode D 6 in. / 6 in. Legend Pull-Through Dominated Even Failures Withdrawal Dominated Figure 5-4. Distributi on of Laboratory Fabricated Stati cally Tested Panels Dominated by Withdrawal or Pull-through Failure modes A) 6 in. / 12 in. spacing, B) ccSPF retrofit of 6 in. / 12 spacing, C) 6 in. / 8 in spacing and D) 6 in. / 6 in. spacing 101

PAGE 102

Table 5-5. Measured failure pressure and calcula ted fastener failure load of statically tested panels harvested from existing structures in Central Florida House Failure Pressure (psf) Tributary Area (ft2) Fastener Withdrawal Load (lbs) Staples: 1.5 in. staple @ 4 in. / 4 in. spacing, in. OSB 1.5 in. Staple-7 93 0.67 62 1.5 in. Staple-8 93 0.67 62 Mean 93 62 Staples: 1.5 in. staple @ 3 in. / 6 in. spacing, in. Plywood 1.5 in. Staple-1 137 1 137 1.5 in. Staple-2 132 1 132 1.5 in. Staple-3 117 1 117 1.5 in. Staple-4 93 1 93 Mean 120 120 Staples: 2.5 in. staple @ 3 in. / 6 in. spacing, in. Plywood 2.5 in. Staple-1 66 1 66 2.5 in. Staple-2 97 1 97 2.5 in. Staple-3 30 1 30 2.5 in. Staple-4 39 1 39 Mean 58 58 Nails: 2.5 in. smooth shank nail @ 6 in. / 12 in. spacing, in. Plywood 8d Smooth Shank-46 112 2 224 8d Smooth Shank-47 101 2 202 Mean 107 213 Nails: 2 in. smooth shank nail @ 6 in. / 12 in. spacing, in. Plywood 6d Smooth Shank-16 134 2 268 6d Smooth Shank-17 98 2 196 6d Smooth Shank-18 74 2 148 6d Smooth Shank-20 53 2 106 6d Smooth Shank-21 132 2 264 6d Smooth Shank-29 46 2 92 6d Smooth Shank-30 31 2 62 6d Smooth Shank-36 59 2 118 6d Smooth Shank-37 90 2 180 Mean 80 159 Nails: 2 in. smooth shank nail @ 8 in. / 8 in. spacing, in. Plywood 6d Smooth Shank-34 83 1.33 111 6d Smooth Shank-35 62 1.33 83 Mean 73 97 102

PAGE 103

Table 5-5. Continued House Failure Pressure (psf) Tributary Area (ft2) Fastener Withdrawal Load (lbs) Nails: 2 in. smooth shank nail @ 3 in. / 6 in. spacing, in. Plywood 6d Smooth Shank-24 117 1 117 6d Smooth Shank-32 62 1 62 6d Smooth Shank-33 64 1 64 Mean 81 81 Nails: 2 in. smooth shank nail @ 4 in. / 4 in. spacing, in. Plywood 6d Smooth Shank-27 149 0.67 99 0 20 40 60 80 100 120 140 160 1800.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9. 5 Panel Attachment Failure Pressure (psf) 1.5 in. Staple (Harvested) 2.5 in. Staple (Harvested) 2 in. 6d Smooth Shank (Harvested) 8d Smooth Shank (Harvested) 2.375 in. 6d Smooth Shank (New) 8d Smooth Shank (New)LEGEND 3" / 6"3" / 6" 6" / 12" 8" / 8"3" / 6"4" / 4"6" / 12" 6" / 12" 6" / 12" Harvested New New Harvested Figure 5-5. Comparison of static ally tested harvested and new panels full results of failure pressure 103

PAGE 104

0 25 50 75 100 125 150 175 200 225 2.5" @ 3"/6"1.5" @ 3"/6"1.5" @ 4"/4 Staple Size and SpacingMean Failure Pressure (psf) Failure Pressure (psf) Fastener Withdrawal Load (lbs)0 25 50 75 100 125 150 175 200 225Mean Failure Load (lbs) 2.5 in. STAPLE 1.5 in. STAPLE A 0 25 50 75 100 125 150 175 200 225 2.5" @ 6"/12"2" @ 6"/12"2" @ 8"/8"2" @ 3"/6"2" @ 4"/4" Nail Size and SpacingMean Failure Pressure (psf) Failure Pressure (psf) Fastener Withdrawal Load (lbs)Mean Failure Load (lbs)0 25 50 75 100 125 150 175 200 225 2.5 in. NAIL 2 in. NAIL B Figure 5-6. Comparison of mean uplift capacities with mean calculated fastener failure loads for harvested panels attached w ith a) staples and b) nails 104

PAGE 105

Table 5-6. Mean failure pre ssures of static UF-WRSUT of panels harvested from existing LFWS located in Central Florida and retrofitted existing panels Source Existing Retrofit % increase in Mean Failure Pressure # of Panels Mean Failure Pressure (psf) # of Panels Retrofit Method Mean Failure Pressure (psf) Staples: 1.5 in. staple @ 3 in. / 6 in. spacing, in. plywood Port Orange #1 4 120 2 A-1 270 125% Staples: 2.5 in. staple @ 3 in. / 6 in. spacing, in. plywood Bartow #2 2 35 2 A-1 167 377% Nails: 2.5 in. smooth shank nail @ 6 in. / 12 in. spacing, in. plywood Crystal River #1 2 107 2 A-1 200 87% Nails: 2 in. smooth shank nail @ 6 in. / 12 in. spacing, in. plywood Port Orange #2 3 102 1 B-2 250 145% Bartow #3 2 93 2 A-1 218 134% Port Orange #3 2 39 1 A-1 140 259% Nails: 2 in. smooth shank nail @ 3 in. / 6 in. spacing, in. plywood Bartow #4 1 117 2 A-1 169 44% Nails: 2 in. smooth shank nail @ 4 in. / 4 in. spacing, in. plywood Bartow #5 1 149 1 A-1 197 32% Notes: Retrofit Method A-1. 8d ring shank nail @ 6 in. / 12 in. spacing Retrofit Method B-2. 3 in. thick layer of spray applied polyurethane adhesive 105

PAGE 106

0 50 100 150 200 250 3001.5" Staple @ 3"/6" (ret. A1) 2.5" Staple @ 3"/6" (ret. A1) 2.5" Nail @ 6"/12" (ret. A-1) 2" Nail @ 6"/12" (ret. B) 2" Nail @ 6"/12" (ret. A-1) 2" Nail @ 6"/12" (ret. A-1) 2" Nail @ 3"/6" (ret. A1) 2" Nail @ 4"/4" (ret. A1)Panel Attachment and RetrofitMean Failure Pressure (psf) Existing Retrofit 1.5 in. STAPLE 2.5 in. STAPLE 2.5 in. NAIL 2 in. NAIL Figure 5-7. Comparison of static ally tested harvested panels in existing vs. retrofit conditions separated by individual source houses 106

PAGE 107

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 Specific GravityProbability Density n = 167 Specific Gravity Data Normal Distribution Lognormal Distribution A 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 Specific GravityProbability Density n = 178 Specific Gravity Data Normal Distribution Lognormal Distribution B Figure 5-8. Probability distribution of Specific Gravity of Panels Tested in Phase 2 for (A New Panels and (B Harvested Panels 15 years old Table 5-7. Summary of specific gravities calculated to inve stigate effect of specific gravity on panel wind uplift resistance ID Statically Tested Dynamically Tested Mean Failure Pressure (psf) SG 1 SG 2 SG 3 Mean Failure Pressure (psf) SG 1 SG 2 SG 3 6d Com. at 6"/12" 62 0.47 0.55 0.57 52 0.53 0.52 0.52 6d Com. At 6"/6" 108 0.52 0.55 0.54 90 0.54 0.55 0.54 6d Com. At 6"/12" Ret. A 183 0.47 0.46 0.48 167 0.55 0.58 0.63 1.5" 14 gauge Staple at 4"/4" 93 0.60 0.62 0.60 72 0.56 0.55 0.50 Notes: SG 1 calculated as mean of all framing members, SG 2 calc. as mean of all framing members which failed, and SG 3 calc. as mean of initial framing member(s) that failed 107

PAGE 108

Table 5-8. Mean failure pressu re of laboratory fabricated pane ls attached with 2 in. long 6d common nails tested statically and dynamically per UF-WRSUT Fastener Schedule Mean Specific Gravity of Framing Static Dynamic Ratio of Dynamic to Static Mean Failure Pressures # of Panels Mean Failure Pressure (psf) # of Panels Mean Failure Pressure (psf) 6 in. / 12 in. 0.50 5 62 5 52 0.84 6 in. / 6 in. 0.53 5 108 5 90 0.83 6 in. / 12 in. ret. A-2 0.51 2 183 2 167 0.91 Notes: Retrofit Method A-2. 8d ring shank nails installed between an existing fastener spacing of 6 in. / 12 in. 0 20 40 60 80 100 120 140 160 180 200 6 in. / 12 in. 6 in. / 6 in.6 in. / 12 in. ret. A-2 Fastener SpacingMean Uplift Capacity (psf) Static Dynamic Existing Retrofit Figure 5-9. Comparison of sta tically vs. dynamically loaded laboratory fabricated panels attached with 2 in. long 6d smooth shank nail s at 6 in. / 12 in. (existing and retrofit) and 6 in. / 6 in. 108

PAGE 109

Table 5-9. Mean failure pressure of panels ha rvested from existing construction attached with 1.5 in. long staples at 4 in. / 4 in. test ed statically and dyna mically per UF-WRUT Fastener Schedule Mean Specific Gravity of Framing Mean Moisture Content of Framing Static Dynamic Ratio of Dynamic to Static Mean Failure Pressures # of Panels Mean Failure Pressure (psf) # of Panels Mean Failure Pressure (psf) 4 in. / 4 in. 0.56 9% 2 93 4 72 0.77 4 in. / 4 in. ret. A-2 0.58 9% 2 168 2 126 0.75 Notes: Retrofit Method A-2. 8d ring shank nails installed between an existing fastener spacing of 6 in. / 12 in. 0 20 40 60 80 100 120 140 160 180 4 in. / 4 in. 4 in. / 4 in. ret. A-2 Fastener ScheduleMean Uplift Capacity (psf) Static Dynamic Existing Retrofit Figure 5-10. Comparison of statically vs. dynami cally loaded panels harvested from existing construction attached with 1.5 in. long staples at 4 in. / 4 in. (e xisting and retrofit) 109

PAGE 110

Table 5-10. Summary of comparis on of peak vs. failure pressure for all static vs. dynamic panel results (34 panels) Dynamically Tested Panels Statically Tested Panels Panel Peak Failure Ratio of Failure to Peak Panel Peak Failure Ratio of Failure to Peak (psf) (sec.) (psf) (sec.) (psf) (sec.) (psf) (sec.) 6d SS at 6 in. / 12 in. (Lab) 6d SS-43 34.5 52 33 60 0.96 6d SS-38 47 52 47 52 6d SS-44 51 68 48 72 0.94 6d SS-40 58 61 58 61 6d SS-45 50 62 46 67.5 0.92 6d SS-41 64.8 61 63.3 70 98% 6d SS-46 58 72 58 72 6d SS-42 65 62 65 62 6d SS-47 66 82 66 82 6d SS-39 73 70 73 70 6d SS at 6 in. / 6 in. (Lab) 6d SS-57 44 62 44 62 6d SS-53 106 90 106 90 6d SS-58 103 111 103 111 6d SS-52 64 61 64 61 6d SS-59 105 111.5 105 111.5 6d SS-56 134 111 134 111 6d SS-60 72 82 53 85 0.74 6d SS-55 91 81 91 81 6d SS-61 124 121 104 124 0.84 6d SS-54 135 110 135 110 6d SS at 6 in. / 12 in. Retrofitted with A-2 (Lab) 6d SS-50 164 171 164 171 6d SS-48 199 150 195 154 98% 6d SS-51 171 173.5 162 173.8 0.95 6d SS-49 167 130 167 130 1.5 in. Staple at 4 in. / 4 in. (Harvested) 1.5 in. Staple-9 93 102 93 102 1.5 in. Staple-7 93 80 90 82 97% 1.5 in. Staple-10 61 72 61 72 1.5 in. Staple-8 93 80 90 84 97% 1.5 in. Staple-11 52 62 37 66 0.71 1.5 in. Staple-12 82 82 71 89 0.87 1.5 in. Staple at 4 in. / 4 in. w ith Retrofitted with A-2 (Harvested) 1.5 in. Staple-15 150 161 150 161 1.5 in. Staple-13 168 130 165 13 2 98% 1.5 in. Staple-16 102 102 90 105 0.88 1.5 in. Staple-14 169 130 165 13 6 98% 110

PAGE 111

Table 5-11. Summary of reduction from statically tested to dynamically tested panel resistance for both peak and failure pressures Peak Pressure Failure Pressure 6d SS at 6 in. / 12 in. (Lab) 16% 20% 6d SS at 6 in. / 6 in. (Lab) 15% 23% 6d SS at 6 in. / 12 in. Retrofitted with A-2 (Lab) 8% 10% 1.5 in. Staple at 4 in. / 4 in. (Harvested) 23% 27% 1.5 in. Staple at 4 in. / 4 in. with Re trofitted with A-2 (Harvested) 25% 27% Table 5-12. Summary of series used in analysis of variance Series # of Panels Mean (psf) St. Dev. Laboratory Fabricated Static 8d RS 6/12 w/ ccSPF fillet 10 202 28 8d RS 6/12 w/ full ccSPF 10 209 39 8d RS 6/12 15 174 23 8d RS 6/12 ccSPF Control 10 161 28 8d RS 6/8 15 216 36 8d RS 6/6 13 252 18 8d SS 6/12 15 129 15 8d SS 6/8 15 175 20 8d SS 6/6 15 205 21 6d SS 6/12 15 74 16 Harvested Static 6d SS 6/12 harvest 9 80 37 6d SS 6/12 lab 15 74 16 Laboratory Fabricated Static vs. Dynamic 6d SS 6/12 S 5 62 10 6d SS 6/12 D 5 52 12 6d SS 6/6 S 5 108 29 6d SS 6/6 D 5 90 31 111

PAGE 112

Table 5-13. ANOVA table for all laboratory fabr icated static phase re sults (alpha = 0.05, therefore 95% confidence level) Source of Variation SS df MS F P-value F crit Between Groups 301684 7 43097.72 81.87163 1.27E-39 2.098005 Within Groups 55272.63 105 526.406 Total 356956.7 112 112

PAGE 113

Table 5-14. Bonferroni test full results for laboratory fabricated st atically tested panels (alpha = 0.05) Series A Series B Lower Bound Upper Bound Null Hypothesis 8d RS 6/12 w/ ccSPF fillet 8d RS 6/12 w/ full ccSPF -43.98 155544 30.181 55544 Accept 8d RS 6/12 -6.284 007303 61.417 34064 Accept 8d RS 6/12 ccSPF control 3.91844456 78.08155544 Reject 8d RS 6/8 -47.91 067397 19.790 67397 Accept 8d RS 6/6 -85.26899174 -15.51562365 Reject 8d SS 6/12 38.7699927 106.4713406 Reject 8d SS 6/8 -6.944 007303 60.757 34064 Accept 8d SS 6/6 -37.14 40073 30.557 34064 Accept 6d SS 6/12 93.88265936 161.5840073 Reject 8d RS 6/12 w/ full ccSPF 8d RS 6/12 0.615992697 68.31734064 Reject 8d RS 6/12 ccSPF control 10.81844456 84. 98155544 Reject 8d RS 6/8 -41.01 067397 26.690 67397 Accept 8d RS 6/6 -78.36899174 -8.615623647 Reject 8d SS 6/12 45.6699927 113.3713406 Reject 8d SS 6/8 -0.044 007303 67.657 34064 Accept 8d SS 6/6 -30.24 40073 37.457 34064 Accept 6d SS 6/12 100.7826594 168.4840073 Reject 8d RS 6/12 8d RS 6/12 ccSPF co ntrol -20.41734064 47.2840073 Accept 8d RS 6/8 -71.9036299 -11.34970343 Reject 8d RS 6/6 -109.378864 -46.53908473 Reject 8d SS 6/12 14.77703677 75.33096323 Reject 8d SS 6/8 -30.93 696323 29.616 96323 Accept 8d SS 6/6 -61.13696323 -0.583036768 Reject 6d SS 6/12 69.88970343 130.4436299 Reject 8d RS 6/12 ccSPF control 8d RS 6/8 -88.91067397 -2 1.20932603 Reject 8d RS 6/6 -126.2689917 -56.51562365 Reject 8d SS 6/12 -2.230 007303 65.471 34064 Accept 8d SS 6/8 -47.94 40073 19.757 34064 Accept 8d SS 6/6 -78.1440073 -10.44265936 Reject 6d SS 6/12 52.88265936 120.5840073 Reject 8d RS 6/8 8d RS 6/6 -67.75219732 -4.912418063 Reject 8d SS 6/12 56.40370343 116.9576299 Reject 8d SS 6/8 10.68970343 71.2436299 Reject 8d SS 6/6 -19.51 029657 41.043 6299 Accept 6d SS 6/12 111.5163701 172.0702966 Reject 113

PAGE 114

114 Table 5-14. Continued Series A Series B Lower Bound Upper Bound Null Hypothesis 8d RS 6/6 8d SS 6/12 91.59308473 154.432864 Reject 8d SS 6/8 45.87908473 108.718864 Reject 8d SS 6/6 15.67908473 78.51886399 Reject 6d SS 6/12 146.7057514 209.5455307 Reject 8d SS 6/12 8d SS 6/8 -75.99096323 -15.43703677 Reject 8d SS 6/6 -106.1909632 -45.63703677 Reject 6d SS 6/12 24.83570343 85.3896299 Reject 8d SS 6/8 8d SS 6/6 -6 0.47696323 0.0 76963232 Accept 6d SS 6/12 70.54970343 131.1036299 Reject 8d SS 6/6 6d SS 6/12 100.7497034 161.3036299 Reject

PAGE 115

Table 5-15. Bonferroni test summ ary for laboratory fabricated static tested results (alpha = 0.05) 8d RS at 6/12, ccSPF fillet 8d RS at 6/12, ccSPF full 8d RS at 6/12 8d RS at 6/12 ccSPF control 8d RS at 6/8 8d RS at 6/6 8d SS at 6/12 8d SS at 6/8 8d SS at 6/6 6d SS at 6/12 8d RS at 6/12, ccSPF fillet 1 8d RS at 6/12, ccSPF full 1 1 8d RS at 6/12 1 0 1 8d RS at 6/12 ccSPF control 0 0 1 1 8d RS at 6/8 1 1 0 0 1 8d RS at 6/6 0 0 0 0 0 1 8d SS at 6/12 0 0 0 1 0 0 1 8d SS at 6/8 1 1 1 1 0 0 0 1 8d SS at 6/6 1 1 0 0 1 0 0 1 1 6d SS at 6/12 0 0 0 0 0 0 0 0 0 1 Notes: 1 = same population (accept null hypothesis), and 0 = different populatio n (reject null hypothesis) 115

PAGE 116

Table 5-16. ANOVA table for all pa nels fastened with 6d smooth shank nails spaced at 6 in. / 12 in. (alpha = 0.05, therefore 95% confidence level) Source of Variation SS df MS F P-value F crit Between Groups 3076.918 3 1025.639 1.986536 0.137212 2.922277 Within Groups 15488.86 30 516.2954 Total 18565.78 33 Table 5-17. Bonferroni test full results for all panels fastened with 6d smooth shank nails spaced at 6 in / 12 in. (alpha = 0.5) Series A Series B Lower Bound Upper Bound Null Hypothesis 6d SS 6/12 harvest 6d SS 6/12 lab -21. 66197671 32.461 97671 Accept 6d SS 6/12 S -17. 63496351 53.964 29685 Accept 6d SS 6/12 D -8. 176963514 63.422 29685 Accept 6d SS 6/12 lab 6d SS 6/12 S -20.37 935052 45.908 68385 Accept 6d SS 6/12 D -10. 92135052 55.366 68385 Accept 6d SS 6/12 S 6d SS 6/12 D -31.13496 506 50.05096506 Accept Table 5-18. T-test result for panels retrofitted with 8d ring shank na ils at 6 in. / 12 in. vs. panels with only 8d ring shank nails at 6 in. / 12 in. 8d RS 6/12 ret. 8d RS 6/12 exist Mean 192.789375 169.06 Variance 1886.39914 648.3875 Observations 16 25 Hypothesized Mean Difference 0 df 22 t Stat 1.978578735 P(T<=t) one-tail 0.030258876 t Critical one-tail 1.717144335 P(T<=t) two-tail 0.060517752 t Critical two-tail 2.073873058 116

PAGE 117

117 Table 5-19. ANOVA table for all pa nels tested in static vs. dyna mic phase using peak pressure (alpha = 0.05, therefor e 95% confidence level) Source of Variation SS df MS F P-value F crit Between Groups 9952.879 3 3317.626 6.436434 0.004583 3.238872 Within Groups 8247.117 16 515.4448 Total 18200 19 Table 5-20. Bonferroni test full results for panels tested in static vs. dynamic phase using peak pressure (alpha = 0.05) Series A Series B Lower Bound Upper Bound Null Hypothesis 6d SS 6/12 S 6d SS 6/12 D -33.73835 588 52.65435588 Accept 6d SS 6/6 S -89.85435588 -3.461644119 Reject 6d SS 6/6 D -71. 32035588 15.072 35588 Accept 6d SS 6/12 D 6d SS 6/6 S -99.31235588 -12.91964412 Reject 6d SS 6/6 D -80.77835588 5.6143 55881 Accept 6d SS 6/6 S 6d SS 6/6 D -24.66235588 61. 73035588 Accept Table 5-21. ANOVA table for all pa nels tested in static vs. dyna mic phase using failure pressure (alpha = 0.05, therefor e 95% confidence level) Source of Variation SS df MS F P-value F crit Between Groups 9054.6535 3 3018.2178 5.782337 0.007095 3.238872 Within Groups 8351.552 16 521.972 Total 17406.2055 19 Table 5-22. Bonferroni test full results for panels tested in static vs. dynamic phase using failure pressure (alpha = 0.05) Series A Series B Lower Bound Upper Bound Null Hypothesis 6d SS 6/12 S 6d SS 6/12 D -32.40899 8 54.52899801 Accept 6d SS 6/6 S -88.208998 -1.27100199 Reject 6d SS 6/6 D -64. 008998 22.928 99801 Accept 6d SS 6/12 D 6d SS 6/6 S -99.268998 -12.331002 Reject 6d SS 6/6 D -75.068998 11.868 99801 Accept 6d SS 6/6 S 6d SS 6/6 D -19.268998 67. 66899801 Accept

PAGE 118

CHAPTER 6 DISCUSSION Wood roof sheathing panel syst em wind uplift design is based on prescriptive codes which specify acceptable panel construction materials and attachment. The determination of panel construction is based on pseudo-sta tic pressure testing of full si ze newly constructed panels to failure. Wind uplift loading of wood roof sheat hing systems in existing conditions is both dynamic and subjected to aging or weathering. It is the hypothesis of th is investigation that pseudo-static pressure testing of newly constructed panels overestimates the wind uplift resistance, which accounts for a portion of the observed failures below design wind speeds. In order to do this a standard test protoc ol was developed (UF-WRSUT) as none currently exist. The protocol has two procedures, Method A (static) and Method B (dynamic), in order to (1) provide a direct comparison of static and dynamic loading and (2 ) create a test protocol using simple equipment which can be utilized in most la boratories. The standard protocol is compared with previous studies which currently form the basis for wood roof sheathing uplift design. The static method is then used to evaluate the upl ift resistance of panels harvested from existing structures in Florida. Finally laboratory constructed and harves ted panels are tested with both methods in order to compare the uplift resistan ce of wood roof sheathing under dynamic loading. This chapter discusses findings from the experi mental study to evaluate the effect of dynamic uplift loading and age or w eathering on wood roof sheathi ng panel uplift capacity. Analysis of Design Wind Speeds As detailed in Chapter 2 the current wind up lift design of wood roof sheathing is based on previous wind uplift pressure testing. To compar e results from this experimental study to design wind speeds ASCE 7-05 (2006) Method 2 Analyti cal Procedure is worked in reverse. The author was unable to find a standard method fo r determining design uplif t pressure, therefore a 118

PAGE 119

factor of safety is applied to the mean failure pressure to obtain the de sign pressure. The 2.0 factor of safety is selected based on a paper by Sutt et al. (2003) whic h summarizes applicable research to panel uplift resistan ce. The paper proposes a desi gn equation for residential wood roof sheathing based on individual fastener capacity, which applies a factor of safety of 2.0 for panel design pressures in addition to the factor of safety of 5.0 used in nail withdrawal design loads. Additionally 5% exclusion values are used for design pressure where available for comparison. The structure which the panels are assumed to be installed in is a partially enclosed gable roof building with a mean roof height of 15 ft in exposure B conditions. Equation 6-22 (eq. 6-2) is rearranged to solve for the velocity pressure (qz) (eq. 6-3), where the design pressure (p) is divided by the sum of the internal (GCpi = +0.55) and external (GCp = -2.6) product of gust effect factor and pressure coefficien ts for an effective wind area of 2 ft2. pi pzGCGCqp (6-2) pi p zGCGC p q (6-3) Then equation 6-15 (eq. 6-4) is rearranged to solve for the design velocity (Vdesign) (eq. 6-5) = (6-4) where qz is divided by the product of a unit conv ersion (0.00256), elevation coefficient (Kz = 0.7), topographic factor (Kzt = 1.0), directionality factory (Kd = 0.85) and importance factor (I 1.0). Then the roof of th at is the design velocity. IVKKK qdesign dztz z200256.0 IKKK q Vdztz z design 00256.0 (6-5) 119

PAGE 120

Table 6-1 summarizes the calculated design wind speeds for laboratory fabricated UFWRSUT Method A results with both 2.0 factor of safety and 5% exclusion value design pressures. From the analysis it is seen that the 2.0 factor of safety design pressure is more conservative than the 5% exclusion design pre ssure, where wind speeds are 11-25% less for the factor of safety design pressure. The current connection of residentia l wood roof sheathing in coastal areas is in Florida is 8d ring shank at 6 in. / 6 in. (ICC 2004) and 8d smooth shank at 4 in. / 8 in. (ICC 2006). ASCE 7-05 (ASCE 2006) design wind speeds for coastal areas in Florida range from 130-150 mph. Results corresponding to current construction practices, 8d ring shank at 6 in. / 6 in. and 8d smooth shank at 6 in. / in. panels, are compared to design wind speeds in coastal areas. Wind speeds calcula ted with the 5% exclusion design pressure are sufficient for both connections. It is seen that current connections of 8d ring shank nails spaced at 6 in. / 6 in. are sufficient to resist design wind speed in coas tal areas. It is found that 6d smooth shank nails provide an insufficient resist ance design wind speeds above 100 m ph. This suggests that the current building stock constructe d before 1994 is in need of retrofit measures to prevent significant losses due to hurricane loading. Design pressure calculated with th e 2.0 factor of safety are cons idered herein due to lack of 5% exclusion values for all re sults. Table 6-2 summarizes ca lculated design wind speeds for panels tested per UF-WRSUT Method B. It is s een that no panel configur ation used is sufficient to be used in coastal areas. Panels retrofitte d with 8d ring shank nails at 6 in. / 12 in. have increased calculated design wind speeds but do not meet ASCE 7-05 design wind speeds. Dynamic testing was not conducted on current construction so conclusions cannot be made regarding the effect of dynamic loading on curr ent construction. However it is shown that 120

PAGE 121

dynamic loading reduces similar wood roof sheathing uplift resistance, theref ore it is reasonable to assume that current construction up lift resistance would be reduced. Effect of Aging or Weathering on Wind Uplift Resistance Laboratory studies of wood roof sheathing conne ctions have suggested that withdrawal strength can be effected by conditions which may be present in the field, detailed in Chapter 2. As part of this study the effect s of aging or weathering on wind uplift resistance is investigated by testing uplift capacities of panel harvested from existing structures. Results do not suggest any significant reduction in wind uplift resistance of residential wood roof sheathing due to aging effects. Results show that harvested panels are within the range of similar laboratory fabricated results. Comparisons are difficult to make due to the limited sample sizes and variation in construction of harvested panel. As discussed in Chapter 4 it is diffi cult to obtain harvested panels and the researcher must take what is availa ble. The effect of age or weathering is shown not to have an effect on the specific gravity of the wood framing of harvested samples. The specific gravity samples collected from existing st ructures is actually sh own to have a higher specific gravity (0.57) than those of la boratory constructed samples (0.52). Panels retrofitted with all of the four methods used experience increased wind uplift resistance. Retrofitting panels with 8d ring shank nails can be done when the roof covering is being replace with little effort. Considering the finding that age or weathering do not effect panel uplift capacity when fastened with nails, results suggests that the signifi cant increase in capacity found in testing will be su stained over the life of the roof. Panels retrofitted with ccSPF have similar structural benefits and provide an option to retrofit the roof at any time, not just when being re-roofed. However the effects of age or weather on ccSPF structural capacity are not known. 121

PAGE 122

Effect of Dynamic Nature of Wind Uplift Loading The comparison of wind uplift performance of st atically tested panels to similar panels tested dynamically is conducted in order to evaluate the effect of the dynamic nature of wind uplift loading on reside ntial wood roof sheathi ng. Results showed approximately a 20% reduction in uplift capacity when dynamically tested panels are compared w ith statically tested ones. The amplitude and frequency content of the dynamic pressure trace derived from wind tunnel tests was representative of actual wind pressure fluctuations to represent as accurately as possible real wind flow. Current pseudo-static wind uplif t testing is based on the assumption that the residential wood structure is a rigid which results in a high natural frequency, therefore the amplifying effects of dynamic loading can be neglected. While the natural frequency of residential wood structures is known the natural frequency of wood roof sheathing is currently unknown and may not be a rigid system. Howeve r displacement measurements were not taken from sheathing during testing making conclusions a bout the rigidity of sh eathing impossible, and small sample sizes limit the statis tical merit of the conclusions. Results show that 50% of panels tested dyna mically fail below the highest instantaneous pressure during testing suggesting a cumulative damage effect. It is thought that if a panel is able to take load after it receives a pressure sp ike then the panel failure is not due to the large loading rate associated with the pressure spike. When failure then occurs at a pressure lower than the preceding spike it is suggested that this is evidence of damage to the nail/wood interface accumulating. These results are preliminary and further testing is necessary to confirm the findings. Wind Uplift Behavior of Resi dential Wood Roof Sheathing Results showed an interesting fact in that current tributary area model used to estimate the nail withdrawal load of sheathing behavior. For all data sets the calculated fastener failure load 122

PAGE 123

increased with decreasing fastener spacings. It wa s expected that the with drawal load would be the same regardless of nail spacing since the same wood is used. A new model developed by Dao and van de Lindt (2008) suggests that withdraw al capacity of fasteners installed in sheathing is non-linearly dependant on fastener spacing, wher e as spacing increases the moment applied to the fastener increases. It is found that under uniform loading panel di splacement is increased significantly and therefore field fastener strengt hs are reduced. This model is particularly interesting when considering the spatial variations present in wind uplift lo ading, where pressure may vary significantly over a 4 ft by 8 ft su rface. Considering that moment reduces the withdrawal capacity of fastener s installed in wood then fiel d nails would also experience moment. Wood roof sheathing panels behavi or is based on the assumption that fastener withdrawal strength with be the limiting factor in panel failure capacity. As discussed in Chapter 2 pullthrough strengths of common wood roof sheathing are similar to fast ener withdrawal strengths. It is therefore expected that as fastener withdrawal capacity incr eases, based on fastener type or spacing, the panel failure mode will change from withdrawal to pull-through. This trend is observed in test results for fasten er type, where panels fastened w ith ring shank nails have more pull-through failures. It is obser ved that as fastener spacing in creases the number of pull-through failures decrease, which suggests an inverse relationship of fast ener spacing and withdrawal capacity. This finding supports Dao and van de Lindt findings that spacing effects fastener withdrawal capacity. The conf irmation of fastener withdraw al strength reducing as spacing increases and the discrepancy of this finding with the tributary area calculated fastener load suggest that tributary is not a ppropriate for wood roof sheathing. 123

PAGE 124

0 25 50 75 100 125 150 175 200 225Lab Static 6d SS @ 6"/12" Kallem Static 6d SS @ 6"/12" Lab Static 8d SS @ 6"/12" IHRC 8d SS @ 6"/12" Lab Static 8d SS @ 6"/6" Murhpy et al. 8d SS @ 6"/6" Lab Static 8d RS @ 6"/12" IHRC 8d RS @ 6"/12"Series IDPressure (psf) Mean Failure Pressure 5% Exclusion Failure Pressure 6d SS at 6 in. / 12 in.8d SS at 6 in. / 12 in.8d SS at 6 in. / 6 in.8d RS at 6 in. / 12 in. Figure 6-1. Comparison of lab static mean a nd 5% exclusion failure pressures to studies by Kallem (1997) w/ plywood, IHRC (2004) w/ plywood and Murphy et al. (1996) w/ OSB (IHRC 8d SS @ 6/12 was not a normal distribution so 5% exclusion value is not provided) 124

PAGE 125

Table 6-1. Comparison of design wind speeds calculated per ASCE 7-05 Method 2 for UFWRSUT Method A results based on A) a factor of safety of 2.0 applied to the mean and B) 5% exclusion of the data (enc losed gable roof building in exposure B assumed, with a mean roof height of 15 ft) Series Mean Failure Pressure (psf) Field Panels (mph) Overhang Panels (mph) Zone 1 Zone 2 Zone 3 Zone 2 Zone 3 6d SS @ 6"/12" 74.3 130 104 88 105 81 8d SS 6"/12" 129.4 171 137 116 139 107 8d SS 6"/8" 175.1 199 160 135 162 125 8d SS 6"/6" 205.3 >200 173 146 175 135 8d RS @ 6"/12" 174.4 191 153 130 155 120 8d RS @ 6"/12" 161.0 199 160 135 161 124 8d RS @ 6"/8" 216.1 >200 178 150 180 138 8d RS @ 6"/6" 252.4 >200 192 162 194 150 Fillet of ccSPF (B-1) 202.0 >200 172 145 174 134 Full 3" of ccSPF (B-2) 209.0 >200 175 148 177 136 A Factor of safety = 2.0 Series 5% Exclusion Failure Pressure (psf) Field Panels (mph) Overhang Panels (mph) Zone 1 Zone 2 Zone 3 Zone 2 Zone 3 6d SS @ 6"/12" 74.3 146 117 99 119 92 8d SS @ 6"/12" 129.4 >200 175 148 177 136 8d SS @ 6"/8" 175.1 >200 >200 172 >200 159 8d SS @ 6"/6" 205.3 >200 >200 188 >200 174 8d RS @ 6"/12" 174.4 >200 >200 169 >200 156 8d RS @ 6"/8" 216.1 >200 >200 181 >200 167 8d RS @ 6"/6" 252.4 >200 >200 >200 >200 198 B 5% exclusion 125

PAGE 126

126 Table 6-2. Comparison of design wind speeds calculated per ASCE 7-05 Method 2 for UFWRSUT Method B results A) statically tested panels and B) dynamically tested panels (enclosed gable roof building in expos ure B assumed, with a mean roof height of 15 ft) Mean Failure Pressure (psf) Field Panel Design Wind Speed (MPH) Overhang Panel Design Wind Speed (MPH) Zone 1 Zone 2 Zone 3 Zone 2 Zone 3 Laboratory Fabricated Panels 2" 6d SS @ 6"/12" 62 118 95 80 96 74 2" SS @ 6"/6" 108 156 126 106 127 98 2" SS @ 6"/12" Ret. A-2 183 204 163 138 165 127 Harvested Panels 1.5" S @ 4"/4" 93 145 116 98 118 91 1.5" S @ 4"/4" Ret. A-2 168 195 157 132 158 122 A Static Mean Failure Pressure (psf) Field Panel Design Wind Speed (MPH) Overhang Panel Design Wind Speed (MPH) Zone 1 Zone 2 Zone 3 Zone 2 Zone 3 Laboratory Fabricated Panels 2" 6d SS @ 6"/12" 52 108 87 74 88 68 2" SS @ 6"/6" 90 143 115 97 116 89 2" SS @ 6"/12" Ret. A-2 167 194 156 132 158 122 Harvested Panels 1.5" S @ 4"/4" 72 128 102 87 104 80 1.5" S @ 4"/4" Ret. A-2 126 169 136 115 137 106 B Dynamic

PAGE 127

CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS Conclusions The purpose of this research was to establ ish a relationship between laboratory test performance of wood roof sheathing and the fa ilure mechanisms observed in the field after hurricanes. The reviewed literature identified no recognized test method for evaluating the wind uplift performance of these systems, and so a te st method was proposed that standardizes current approaches used for wind uplift testing of resi dential wood roof sheathing. A dynamic pressure test protocol was also developed using pre ssure time histories co llected in wind tunnel experiments. These two test methods (a Static and a Dynamic one) were used to compare the ultimate wind uplift resistance of new and exis ting wood roof sheathing panels. The design wind speeds associated with the failure pressures were determined in accordance with ASCE 7 minimum load provisions, and the following conclusions were made: Comparison of Results with Previous Studies Panels tested with standard wind uplift test protocol developed (UF-WRSUT) have failure pressures in the same order of magnitude as prev iously reported test re sults. For example 8d smooth shank nails at 6 in. / 12 in. yielded a mean failure pressure of 129 psf which is similar to a study by IHRC (IHRC 2004) which had a mean failure pressure of 110 psf. Additionally coefficients of variation from panel uplift results from UF-WRS UT (7% to 22%) are in similar range as observed in previous studies (7% to 28%). Effect of In-Service and Environmen tal Effects on Roof Panel Strength Test results have not shown si gnificant loss in wind uplift stre ngth of roof panels due to environmental conditions. Comparing panels fast ened with 6d smooth shank nails at 6 in. / 12 in. spacing, harvested panels have a higher mean uplift capacity (80 psf) than laboratory 127

PAGE 128

fabricated panels (74 psf). In the case of panels fastened with 8d smooth shank nails at 6 in. / 12 in. spacing harvested panels have a lower m ean uplift capacity (107 psf) than laboratory fabricated panels (129 psf), however only two harvested panels were tested. A larger data set of existing panels would need to be tested to establish a statistical basis for this, however experience with wood materials su ggests that once its moisture content is maintained below about 19% (as was the case in all the houses tested) the strength of the wood should not deteriorate. Further low moistu re content in the wood also prev ents the growth of fungi that permanently degrade wood strength. Dynamic Load Effects on Wood Panel Strength Dynamic loading is shown to reduce failure pressure results from 9% to 23%, which suggests a potential reduction due to dynamic load ing. The reduction from static to dynamic testing is observed in both new panel constructio n and panels harvested fro m existing structures. An analysis of variance finds that all static vs. dynamic results come from the same sample population and a limited sample size is tested so no conclusive findings can be made. Further testing with larger numbers of samples are needed to extrapolate results to current wind uplift design of wood roof sheathing panels. Recommendations Previous residential wood roof sheathing uplif t testing have used various methods which have contributed to a varying body results. The us e of a standard test method is suggested as a for uplift testing of residential wood roof sheathing in order to reduce the error in comparisons due the test method. It is recommended that further testing of both dynamic and age or weathering effects be continued as sample sizes were relatively small due to logistical and time constraints. It is recommended that an investiga tion be conducted to determine if cumulative damage occurs in 128

PAGE 129

129 the connection between individual dowel type fa steners and the wood it is embedded in. As wood is found in to be an elastic material at low levels of energy cumulative damage effects may only occur above some threshold. Quantifying cu mulative damage would require lab studies of nail samples subjected to cyclical loadi ng of varying load levels and amplitude. It is recommended that the distribution of pr essure loads to individual fasteners in wood roof sheathing panels be invest igated to relate the previous recommendation to panel system behavior.

PAGE 130

APPENDIX A FULL PANEL UPLIFT RESULTS Table A-1. Summary of panel uplift test series Panel Series Fastener Schedule Dates of Testing Retrofit Type Description 1.5 in. Staple Port Orange #1 3 in. / 6 in. A-1 Six harvested panels tested statically in EC (4) and RC (2) Debary #1 4 in. / 4 in. 15 yrs A-1 Ten harvested panels tested statically in EC (2) and RC (2) vs. dynamically in EC (4) and RC (2) 2.5 in. Staple Bartow #1 3 in. / 6 in. Two harvested panels tested statically in EC Bartow #2 6 in. / 12 in. A-1 Four harvested panels tested statically in EC (2) and RC (2) 6d (0.113 in. diameter) Smooth Shank Nails, 2 in. and 2.375 in. long Laboratory #1-A 6 in. / 12 in. New Five laboratory fabricated panel tested statically Laboratory #1-B 6 in. / 12 in. New Ten lab-built panels tested statically to have similar sample size as other lab-built panels Port Orange #2 6 in. / 12 in. 28 yrs B-2 Four harvested panels tested statically in EC (3) and RC (1) Bartow #3 6 in. / 12 in. 32 yrs A-1 Four harvested panels tested statically in EC (2) and RC (2) Bartow #4 3 in. / 6 in. 32 yrs A-1 Three harvested panels tested statically in EC (1) and RC (2) Bartow #5 4 in. / 4 in. 32 yrs A-1 Two harvested panels tested statically in EC (1) and RC (1) Port Orange #3 6 in. / 12 in. A-1 Three harvested panels tested statically in EC (2) and RC (1) Port Orange #4 3 in. / 6 in. Two harvested panels tested statically in EC (2) Port Orange #5 8 in. / 8 in. Two harvested panels tested statically in EC (2) Port Orange #6 6 in. / 12 in. Two harvested panels tested statically in EC (2) Laboratory #2-A 6 in. / 12 in. New Ten laboratory fabricat ed panels tested statically (5) and dynamically (5) in EC Laboratory #2-B 6 in. / 12 in. New A-2 Four laboratory fabri cated panels tested statically (2) and dynamically (2) in RC Laboratory #2-C 6 in. / 6 in. New Ten laboratory fabricated panels tested statically (5) and dynamically (5) in EC Notes: EC = existing condition, RC = retrofit condition; Retrofit Methods A-1. 8d ring shank nail @ 6 in. / 12 in. spacing, A-2. 8d ring shank nails installed between an existing fastener spacing of 6 in. / 12 in., B-1. fillet of spray applied polyurethane adhesive, B-2. 3 in. thick la yer of spray applied polyurethane adhesive 130

PAGE 131

131 Table A-1. Continued Panel Series Fastener Schedule Dates of Testing Retrofit Type Description 8d (0.131 in. diameter) Smooth Shank Nails, 2.5 in. long Laboratory #3-A 6 in. / 12 in. New Fifteen labora tory fabricated panels tested statically in EC Laboratory #3-B 6 in. / 8 in. New Fifteen laborato ry fabricated panels tested statically in EC Laboratory #3-C 6 in. / 6 in. New Fifteen laborato ry fabricated panels tested statically in EC Crystal River #1 6 in. / 12 in. 35 yrs A-1 Four harvested panel tested statically in EC (2) and RC (2) 8d (0.113 in. diameter) Ring Shank Nails, 2.5 in. long Laboratory #4-A 6 in. / 12 in. New Ten la boratory fabricated panels tested stati cally in EC, for comparison with B and C Laboratory #4-B 6 in. / 12 in. New B-1 Ten labora tory fabricated panels tested statically in RC Laboratory #4-C 6 in. / 12 in. New B-2 Ten labora tory fabricated panels tested statically in RC Laboratory #5-A 6 in. / 12 in. New Fifteen labora tory fabricated panels tested statically in EC Laboratory #5-B 6 in. / 8 in. New Fifteen laborato ry fabricated panels tested statically in EC Laboratory #5-C 6 in. / 6 in. New Fifteen laborato ry fabricated panels tested statically in EC Notes: EC = existing condition, RC = retrofit condition; Retrofit Methods A-1. 8d ring shank nail @ 6 in. / 12 in. spacing, A-2. 8d ring shank nails installed between an existing fastener spacing of 6 in. / 12 in., B-1. fillet of spray applied polyurethane adhesive, B-2. 3 in. thick la yer of spray applied polyurethane adhesive

PAGE 132

Table A-2. Full uplift test results of panels fastened with 1.5 in. staples Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Port Orange #1 1.5 in. Staple-1 1.5 in. 3 in. / 6 in. 1/2 in. Ply X 137 1.5 in. Staple-2 X 132 1.5 in. Staple-3 X 117 1.5 in. Staple-4 X 93 1.5 in. Staple-5 8d RS 6 in. / 12 in. X 221 1.5 in. Staple-6 X 319 Debary #1 1.5 in. Staple-7 15 yrs 0.56 8% 1.5 in. (14 gauge) 4 in. / 4 in. 7/16 in. OSB X W 93 76 1.5 in. Staple-8 0.64 9% X W 93 76 1.5 in. Staple-9 15 yrs 0.56 10% 4 in. / 4 in. X W 93 57 1.5 in. Staple-10 0.55 9% X PT 61 33 1.5 in. Staple-11 0.54 9% X W 52 25 1.5 in. Staple-12 0.58 9% X W 82 42 1.5 in. Staple-13 15 yrs 0.61 10% 4 in. / 4 in. 8d RS 6 in. / 12 in. X PT 168 152 1.5 in. Staple-14 0.56 9% X W 169 152 1.5 in. Staple-15 15 yrs 0.62 9% 4 in. / 4 in. 8d RS 6 in. / 12 in. X PT 150 90 1.5 in. Staple-16 0.52 9% X PT 102 57 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even and RS = ring shank nail 132

PAGE 133

Table A-3. Full uplift test results of panels fastened with 2.5 in. staples Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Bartow # 1 2.5 in. Staple-1 2.5 in. 3 in. / 6 in. 1/2 in. Ply X 66 2.5 in. Staple-2 X 97 Bartow # 2 2.5 in. Staple-3 2.5 in. 6 in. / 12 in. 1/2 in. Ply X 30 2.5 in. Staple-4 X 39 2.5 in. Staple-5 8d RS 6 in. / 12 in. X 165 2.5 in. Staple-6 X 169 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even and RS = ring shank nail 133

PAGE 134

Table A-4. Full uplift test results of pa nels fastened with 6d smooth shank nails Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #1-A 6d SS-1 New 2.375 in. (0.113 in.) 6 in. / 12 in. 1/2 in. OSB X 75 6d SS-2 X 105 6d SS-3 X 71 6d SS-4 X 76 6d SS-5 X 47 Laboratory #1-B 6d SS-6 New X W 62 6d SS-7 X W 59 6d SS-8 X W 96 6d SS-9 X W 69 6d SS-10 X W 79 6d SS-11 X W 87 6d SS-12 X W 95 6d SS-13 X W 57 6d SS -14 X W 59 6d SS-15 X W 76 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even ; RS = ring shank nail and SS = smooth shank nail 134

PAGE 135

Table A-4. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Port Orange #2 6d SS-16 28 yrs 2 in. (0.113 in.) 6 in. / 12 in. 1/2 in. Ply X 134 6d SS-17 X 98 6d SS-18 X 74 6d SS-19 ccSPF Full (3 in.) X 250 Bartow #3 6d SS-20 32 yrs 2 in. (0.113 in.) 6 in. / 12 in. 1/2 in. Ply X 53 6d SS-21 X 132 6d SS-22 8d RS 6 in. / 12 in. X 229 6d SS-23 X 206 Bartow #4 6d SS-24 32 yrs 2 in. (0.113 in. ) 3 in. / 6 in. 1/2 in. Ply X 117 6d SS-25 8d RS 6 in. / 12 in. X 136 6d SS-26 X 202 Bartow #5 6d SS-27 32 yrs 2 in. (0.11 3 in. ) 4 in. / 4 in. 1/2 in. Ply X 149 6d SS-28 8d RS 6 in. / 12 in. X 197 Port Orange #3 6d SS-29 2 in. (0.113 in. ) 6 in. / 12 in. 1/2 in. Ply X 46 6d SS-30 X 31 6d SS-31 8d RS 6 in. / 12 in. X 140 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS = smooth s hank nail 135

PAGE 136

Table A-4. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Port Orange #4 6d SS-32 2 in. (0.113 in.) 3 in. / 6 in. 1/2 in. Ply X 62 6d SS-33 X 64 Port Orange #5 6d SS-34 8 in. / 8 in. X 83 6d SS-35 X 62 Port Orange #6 6d SS-36 6 in. / 12 in. X 59 6d SS-37 X 90 Laboratory #2-A 6d SS-38 New 0.48 2 in. (0.113 in.) 6 in. / 12 in. 1/2 in. OSB X W 47 32 6d SS-39 0.49 X W 73 61 6d SS-40 0.44 X W 58 46 6d SS-41 0.44 X W 65 60 6d SS-42 0.51 X W 65 46 6d SS -43 New 0.50 2 in. (0.113 in.) 6 in. / 12 in. 1/2 in. OSB X W 35 23 6d SS-44 0.56 X W 51 32 6d SS-45 0.52 X W 51 29 6d SS-46 0.51 X W 58 32 6d SS-47 0.54 X W 66 40 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS = smooth s hank nail 136

PAGE 137

Table A-4. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #2-B 6d SS-48 New 0.49 2 in. (0.113 in.) 6 in. / 12 in. 8d RS 6 in. / 12 in. 1/2 in. OSB X W 199 186 6d SS-49 0.44 X W 167 150 6d SS-50 0.55 8d RS 6 in. / 12 in. X W 164 99 6d SS-51 0.54 X W 171 100 Laboratory #2-C 6d SS-52 New 0.50 2 in. (0.113 in.) 6 in. / 6 in. 1/2 in. OSB X PT 64 47 6d SS-53 0.58 X W 106 90 6d SS-54 0.47 X W 135 120 6d SS-55 0.57 X W 91 77 6d SS-56 0.47 X W 134 120 6d SS-57 0.56 X W 44 27 6d SS-58 0.58 X W 103 62 6d SS-59 0.56 X W 105 62 6d SS-60 0.54 X W 72 43 6d SS-61 0.48 X W 124 70 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS = smooth s hank nail 137

PAGE 138

Table A-5. Full uplift test results of pa nels fastened with 8d smooth shank nails Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #3-A 8d SS-1 New 2.5 in. (0.131 in.) 6 in. / 12 in. 1/2 in. OSB X W 127 8d SS-2 X W 162 8d SS-3 X E 131 8d SS-4 X PT 126 8d SS-5 X PT 123 8d SS-6 X E 135 8d SS-7 X W 124 8d SS-8 X W 149 8d SS-9 X E 132 8d SS-10 X E 109 8d SS-11 X W 137 8d SS-12 X W 100 8d SS-13 X W 131 8d SS-14 X E 135 8d SS-15 X W 118 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nails and SS = smooth shank nail 138

PAGE 139

Table A-5. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #3-B 8d SS-16 New 2.5 in. (0.131 in.) 6 in. / 8 in. 1/2 in. OSB X PT 151 8d SS-17 X PT 180 8d SS-18 X E 178 8d SS-19 X PT 183 8d SS-20 X W 167 8d SS-21 X W 160 8d SS-22 X W 180 8d SS-23 X PT 192 8d SS-24 X W 145 8d SS-25 X W 221 8d SS-26 X W 172 8d SS-27 X PT 151 8d SS-28 X W 168 8d SS-29 X W 178 8d SS-30 X PT 202 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS = smooth shank nail 139

PAGE 140

Table A-5. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Spacing Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Type Type Spacing Laboratory #3-C 8d SS-31 New 2.5 in. (0.131 in.) 6 in. / 6 in. 1/2 in. OSB X W 189 8d SS-32 X W 163 8d SS-33 X W 239 8d SS-34 X W 176 8d SS-35 X PT 233 8d SS-36 X W 196 8d SS-37 X W 222 8d SS-38 X W 194 8d SS-39 X W 195 8d SS-40 X W 209 8d SS-41 X W 226 8d SS-42 X W 224 8d SS-43 X W 207 8d SS-44 X PT 214 8d SS-45 X PT 193 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS = smooth s hank nail 140

PAGE 141

Table A-5. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Crystal River #1 8d SS-46 35 yrs 2.5 in. (0.131 in.) 6 in. / 12 in. 1/2 in. Ply X 112 8d SS-47 X 101 8d SS-48 8d RS 6 in. / 12 in. X 182 8d SS-49 X 217 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through and E = even ; RS = ring shank 141

PAGE 142

Table A-6. Full uplift test results of panels fastened with 8d ring shank nails Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #4-A 8d RS-1 New 2.5 in. (0.113 in.) 6 in. / 12 in. 1/2 in. OSB X W 115 8d RS-2 X W 177 8d RS-3 X E 144 8d RS-4 X PT 164 8d RS -5 X W 123 8d RS-6 X W 168 8d RS -7 X PT 153 8d RS-8 X PT 210 8d RS-9 X PT 174 8d RS-10 X PT 182 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS = smooth s hank 142

PAGE 143

Table A-6. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #4-B 8d RS-11 New 2.5 in. (0.113 in.) 6 in. / 12 in. ccSPF Fillet 1/2 in. OSB X W 204 8d RS-12 X PT 174 8d RS-13 X W 214 8d RS-14 X PT 192 8d RS15 X W 258 8d RS-16 X E 225 8d RS-17 X PT 211 8d RS-18 X W 166 8d RS-19 X PT 172 8d RS-20 X E 204 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS = smooth s hank nail 143

PAGE 144

Table A-6. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #4-C 8d RS-21 New 2.5 in. (0.113 in.) 6 in. / 12 in. ccSPF Full (3 in.) 1/2 in. OSB X E 210 8d RS-22 X PT 241 8d RS-23 X PT 273 8d RS-24 X W 202 8d RS-25 X W 151 8d RS-26 X E 175 8d RS-27 X PT 196 8d RS-28 X E 165 8d RS-29 X PT 245 8d RS-30 X W 231 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS = smooth s hank nail 144

PAGE 145

Table A-6. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #5-A 8d RS-59 New 2.5 in. (0.113 in.) 6 in. / 12 in. 1/2 in. OSB X W 186 8d RS-60 X W 151 8d RS-61 X W 115 8d RS-62 X PT 151 8d RS-63 X PT 166 8d RS-64 X PT 181 8d RS-65 X PT 188 8d RS-66 X PT 195 8d RS-67 X W 199 8d RS-68 X W 175 8d RS-69 X W 192 8d RS-70 X PT 172 8d RS-71 X PT 198 8d RS-72 X PT 186 8d RS-73 X E 161 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS =smooth sh ank nail 145

PAGE 146

Table A-6. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #5-B 8d RS-44 New 2.5 in. (0.113 in.) 6 in. / 8 in. 1/2 in. OSB X PT 185 8d RS-45 X PT 193 8d RS-46 X PT 247 8d RS-47 X W 134 8d RS-48 X PT 239 8d RS -49 X PT 269 8d RS-50 X PT 207 8d RS-51 X PT 211 8d RS-52 X PT 245 8d RS-53 X W 224 8d RS-54 X PT 163 8d RS-55 X PT 237 8d RS-56 X W 222 8d RS-57 X PT 219 8d RS-58 X PT 247 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through, E = even, RS = ring shank nail and SS = smooth s hank nail 146

PAGE 147

147 Table A-6. Continued Panel Description Age Specific Gravity Moisture Content Existing Fastener Retrofit Fastener Sheathing Static Loading Dynamic Loading Failure mode Peak Uplift (psf) 10-sec. Mean Uplift (psf) Size Spacing Type Spacing Laboratory #5-C 8d RS-31 New 2.5 in. (0.113 in. ) 6 in. / 6 in. 1/2 in. OSB X PT 241 8d RS-32 X PT 268 8d RS-33 X PT 239 8d RS-34 X PT 253 8d RS-35 X PT 284 8d RS-36 X PT 253 8d RS-37 X PT 254 8d RS-38 X PT 241 8d RS-39 X PT 255 8d RS-40 X PT 248 8d RS-41 X PT 221 8d RS-42 X PT 239 8d RS-43 X PT 286 Notes: Dominant failure mode described by W = withdrawal, PT = pull-through and E = even ; RS = ring shank

PAGE 148

APPENDIX B FULL FAILURE MODE AND LOCATION INFORMATION Figure B-1 thru B-12 present location and mode of each panel fabricated in the laboratory. Results are grouped by series. 75.7 psf Slow failure#15 62.2 psf Slow failure#6 58.7 psf Moderate#7 96.2 psf Slow failure#8 94.8 psf Slow failure#12 87.0 psf Moderate#11 78.5 psf Slow failure#10 69.3 psf Slow failure#9 Not available 105.4 psf Slow failure#2 Not available 75.0 psf Slow failure#1 59.4 psf Slow failure#14 57.3 psf Slow failure#13 Not available 46.7 psf Slow failure#5 Not available 76.4 psf Slow failure#4 Not available 71.4 psf Slow failure#3Test dates : 7/18-19/2007 and 10/23/2007 Spacing : 12 inches O.C. interior ( 5 total) and at 6 inches O.C. edge (9 total) Board Split Partial Withdraw Pull Through Full Withdraw Legend 12 in 2 ft 8 ft 4 ft15 samples Mean : 74.3 psf Max : 105.4 psf Min : 46.7 psf STD : 16.36 psf COV : 0.220 Fastener : 6d smooth shank Figure B-1. Failure mode / location for panels fa stened with 6d smooth shank nail spaced at 6 in. / 12 in. tested statically 148

PAGE 149

126.8 psf Slow failure#1028d-c-06 12 in 2 ft 8 ft 4 ft 46.5 psf 6 in. / 12 in.#38 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 73 psf6 in. / 12 in.#39 2 ft 8 ft 4 ft2 samples Mean : 182.8 psf Max : 199 psf Min : 167 psf LegendLegend Board Split Partial Withdraw Pull Through Full Withdraw Board Split Partial Withdraw Pull Through Full Withdraw Nail-head Failure 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 58 psf6 in. / 12 in.#40 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 65 psf6 in. / 12 in.#41 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 65 psf6 in. / 12 in.#42 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 167 psf6 in. / 12 in. ret.#49 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 63.9 psf6 in. / 6 in.#52 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 198.6 psf6 in. / 12 in. ret.#48 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 106 psf 6 in. / 6 in.#53 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 135 psf6 in. / 6 in.#54 Test dates : 3/13/2009 to 3/25/2009 Spacing : 12 inches O.C. interior ( 5 total) or 6 inches O.C. interior (9 total), and at 6 inches O.C. edge (9 total); one series of 6 /12retrofitted with A-2 Fastener : 2 in. 6d smooth shank, static vs. dynamic testing statically tested panels 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 134 psf 6 in. / 6 in.#56 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 102 psf6 in. / 6 in.#55 5 samples Mean : 61.5 psf Max : 73 psf Min : 47 psf STD : 10 psf COV : 0.16 5 samples Mean : 108.2 psf Max : 135 psf Min : 64 psf STD : 29 psf COV : 0.27 6/12ret. w/ A-2 6/6 6/12 Figure B-2. Failure mode / location for panels fa stened with 6d smooth shank nail spaced at 6 in. / 12 in., 6 in. / 12 in. retrofitted with ret. A-2, and 6 in. / 6 in. tested statically 149

PAGE 150

126.8 psf Slow failure#1028d-c-06 12 in 2 ft 8 ft 4 ft 34.5 psf 6 in. / 12 in.#43 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 51.2 psf6 in. / 12 in.#44 2 ft 8 ft 4 ft2 samples Mean : 167.2 psf Max : 171 psf Min : 164 psf LegendLegend Board Split Partial Withdraw Pull Through Full Withdraw Board Split Partial Withdraw Pull Through Full Withdraw Nail-head Failure 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 50.5 psf6 in. / 12 in.#45 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 58 psf6 in. / 12 in.#46 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 66 psf6 in. / 12 in.#47 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 170.7 psf6 in. / 12 in. ret.#51 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 44.4 psf6 in. / 6 in.#57 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 163.7 psf6 in. / 12 in. ret.#50 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 102.8 psf 6 in. / 6 in.#58 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 104.6 psf6 in. / 6 in.#59 Test dates : 3/13/2009 to 3/25/2009 Spacing : 12 inches O.C. interior ( 5 total) or 6 inches O.C. interior (9 total), and at 6 inches O.C. edge (9 total); one series of 6 /12retrofitted with A-2 Fastener : 2 in. 6d smooth shank, static vs. dynamic testing dynamicallytested panels 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 124 psf 6 in. / 6 in.#61 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 72.5 psf6 in. / 6 in.#60 5 samples Mean : 52 psf Max : 66 psf Min : 35 psf STD : 12 psf COV : 0.22 5 samples Mean : 89.6 psf Max : 124 psf Min : 44 psf STD : 31 psf COV : 0.35 6/12ret. w/ A-2 6/6 6/12 Figure B-3. Failure mode / location for panels fa stened with 6d smooth shank nail spaced at 6 in. / 12 in., 6 in. / 12 in. retrofitted with ret. A-2, and 6 in. / 6 in. tested dynamically 150

PAGE 151

126.8 psf Slow failure#1 162.1 psf Slow failure#2 131.1 psf Fast failure#3 126.1 psf Fast failure#4 149.4 psf Slow failure#8 124.0 psf Fast failure#7 135.3 psf Slow failure#6 123.3 psf Fast failure#5 100.1 psf Fast failure#12 137.4 psf Slow failure#11 109.2 psf Fast failure#10 131.8 psf Slow failure#9 118.4 psf Fast failure#15 134.6 psf Slow failure#14 131.1 psf Slow failure#13 Board Split Partial Withdraw Pull Through Full Withdraw Legend 12 in 2 ft 8 ft 4 ft15 samples Mean : 129.4 psf Max : 162.1 psf Min : 100.1 psf STD : 14.68 psf COV : 0.11 Test dates : 1/24/2008 to 1/31/2008 Spacing : 12 inches O.C. interior ( 5 total) and at 6 inches O.C. edge (9 total) Fastener : 8d smooth shank Figure B-4. Failure mode / location for panels fa stened with 8d smooth shank nail spaced at 6 in. / 12 in. tested statically 151

PAGE 152

Board Split Partial Withdraw Pull Through Full Withdraw 179.7 psf Fast failure#17 178.3 psf Fast failure#18 183.2 psf Fast failure#19 167.0 psf Fast failure#20 160.0 psf Slow failure#21 191.7 psf Moderate#23 179.7 psf Fast failure#22 150.8 psf Slow failure#27 171.9 psf Fast failure#26 220.6 psf Slow failure#25 144.5 psf Fast failure#24 168.4 psf Slow failure#28 201.5 psf Fast failure#30 178.3 psf Moderate#29 Legend 150.8 psf Fast failure#16 8 in 4 ft 8 ft 2 ft15 samples Mean : 175.1 psf Max : 220.6 psf Min : 144.5 psf STD : 20.07 psf COV : 0.11 Test dates : 1/24/2008 to 1/31/2008 Spacing : 8 inches O.C. interior ( 7 total) and at 6 inches O.C. edge (9 total) Fastener : 8d smooth shank Figure B-5. Failure mode / location for panels fastened with 8d smooth shank nails spaced at 6 in. / 8 in. tested statically 152

PAGE 153

Board Split Partial Withdraw Pull Through Full Withdraw 195.2 psf Moderate#39 193.8 psf Fast failure#38 222.1 psf Fast failure#37 195.9 psf Moderate#36 232.7 psf Fast failure#35 239.1 psf Fast failure#33 176.1 psf Fast failure#34 213.6 psf Moderate#44 193.1 psf Fast failure#45 188.8 psf Fast failure#31 163.4 psf Moderate#32 207.2 psf Fast failure#43 226.3 psf Fast failure#41 223.5 psf Fast failure#42 Legend 208.6 psf Fast failure#40 8 in 4 ft 8 ft 2 ft15 samples Mean : 205.3 psf Max : 239.1 psf Min : 163.4 psf STD : 21.35 psf COV : 0.104 Test dates : 10/22/2007 and 10/25/2007 Spacing : 6 inches O.C. interior ( 9 total) and at 6 inches O.C. edge (9 total) Fastener : 8d smooth shank Figure B-6. Failure mode / location for panels fastened with 8d smooth shank spaced at 6 in. / 6 in. tested statically 153

PAGE 154

126.8 psf Slow failure#1028d-c-06 12 in 2 ft 8 ft 4 ft 115 psf Slow failure#1 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 177 psfSudden failure#2 12 in 2 ft 8 ft 4 ft10 samples Mean : 161 psf Max : 210 psf Min : 115 psf STD : 28.36 psf COV : 0.176 LegendLegend Board Split Partial Withdraw Pull Through Full Withdraw Board Split Partial Withdraw Pull Through Full Withdraw Nail-head Failure 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 144 psfModerate failure#3 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 164 psfSudden Failure#4 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 123 psfSlow failure#5 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 174 psfSudden failure#9 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 182 psfSudden failure#10 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 210 psfSudden failure#8 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 153 psf Sudden failure#7 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 168 psfSudden failure#6 Test dates : 5/23/2008 to 6/4/2008 Spacing : 12 inches O.C. interior ( 5 total) and at 6 inches O.C. edge (9 total) Fastener : 8d ring shank (control for lab ccSPF retrofit testing) Figure B-7. Failure mode / location for panels fastened with 8d ring shank nails spaced at 6 in. / 12 in. tested statically 154

PAGE 155

126.8 psf Slow failure#1028d-c-06 12 in 2 ft 8 ft 4 ft 225 psf Sudden failure#16 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 211 psfSudden failure#17 12 in 2 ft 8 ft 4 ft10 samples Mean : 202 psf Max : 258 psf Min : 166 psf STD : 27.85 psf COV : 0.138 LegendLegend Board Split Partial Withdraw Pull Through Full Withdraw Board Split Partial Withdraw Pull Through Full Withdraw Nail-head Failure 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 166 psfSlow failure#18 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 204 psfSlow Failure#20 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 172 psfSlow failure#19 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 174 psfSudden failure#12 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 204 psfSudden failure#11 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 214 psfSudden failure#13 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 192 psf Sudden failure#14 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 258 psfModerate failure#15 Test dates : 2/23/2008 to 6/4/2008 Spacing : 12 inches O.C. interior ( 5 total) and at 6 inches O.C. edge (9 total) Fastener : 8d ring shank retrofitted with 1-1/2 inch fillet of ccSPF Figure B-8. Failure mode / location for panels fastened with 8d ring shank nails spaced at 6 in. / 12 in. retrofitted with 1-1/2 in. fi llet of ccSPF tested statically 155

PAGE 156

126.8 psf Slow failure#1028d-c-06 12 in 2 ft 8 ft 4 ft 273 psf Sudden failure#23 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 202 psfSlow failure#24 12 in 2 ft 8 ft 4 ft10 samples Mean : 209 psf Max : 273 psf Min : 151 psf STD : 38.84 psf COV : 0.186 LegendLegend Board Split Partial Withdraw Pull Through Full Withdraw Board Split Partial Withdraw Pull Through Full Withdraw Nail-head Failure 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 151 psfSudden failure#25 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 175 psfSudden Failure#26 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 196 psfSlow failure#27 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 210 psfSudden failure#21 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 241 psfSudden failure#22 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 245 psfSlow failure#29 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 231 psf Sudden failure#30 162.1 psf Slow failure#1058d-c-06 162.1 psf Slow failure#1058d-c-06 165 psfSudden failure#28 Test dates : 5/23/2008 to 6/4/2008 Spacing : 12 inches O.C. interior ( 5 total) and at 6 inches O.C. edge (9 total) Fastener : 8d ring shank retrofitted with full 3 in. thick ccSPF Figure B-9. Failure mode / location for panels fastened with 8d ring shank nails spaced at 6 in. / 12 in. retrofitted with full 3 in. thic k layer of ccSPF tested statically 156

PAGE 157

Board Split Partial Withdraw Pull Through Full Withdraw 239.1 psf Fast failure#42 220.7 psf Fast failure#41 248.2 psf Fast failure#40 255.3 psf Fast failure#39 253.9 psf Fast failure#37 240.5 psf Fast failure#38 252.5 psf Moderate#34 284.3 psf Fast failure#35 252.5 psf Fast failure#36 239.1 psf Fast failure#33 240.5 psf Fast failure#31 268.1 psf Fast failure#32 Legend 286.4 psf Fast failure#43 8 in 4 ft 8 ft 2 ft13 samples Mean : 252.4 psf Max : 286.4 psf Min : 220.7 psf STD : 18.48 psf COV : 0.073 Test dates : 10/23/2007 and 10/25/2007 Spacing : 6 inches O.C. interior ( 9 total) and at 6 inches O.C. edge (9 total) Fastener : 8d ring shank Figure B-10. Failure mode / location of panels fa stened with 8d ring shank nails spaced at 6 in. / 6 in. tested statically 157

PAGE 158

184.6 psf Fast failure#44 193.1 psf Fast failure#45 133.9 psf Fast failure#47 246.6 psf Fast failure#46 211.4 psf Fast failure#51 206.5 psf Fast failure#50 269.2 psf Fast failure#49 238.9 psf Fast failure#48 236.8 psf Fast failure#55 162.8 psf Fast failure#54 224.1 psf Fast failure#53 245.2 psf Fast failure#52 247.3 psf Fast failure#58 218.5 psf Fast failure#57 222.0 psf Fast failure#56Partial Withdraw Board Split Pull Through Full Withdraw Legend 2 ft 8 ft 8 in 4 ft15 samples Mean : 216.1 psf Max : 269.2 psf Min : 133.9 psf STD : 35.72 psf COV : 0.17 Test dates : 1/24/2008 to 1/31/2008 Spacing : 8 inches O.C. interior ( 7 total) and at 6 inches O.C. edge (9 total) Fastener : 8d ring shank Figure B-11. Failure mode / location for panels fa stened with 8d ring shank nails spaced at 6 in. / 8 in. tested statically 158

PAGE 159

159 186.0 psf Fast failure#59 195.2 psf Fast failure#66 192.4 psf Fast failure#69 Board Split Partial Withdraw Pull Through Full Withdraw Legend 12 in 2 ft 8 ft 4 ft 151.4 psf Slow failure#60 115.3 psf Slow failure#61 151.4 psf Slow failure#62 165.5 psf Fast failure#63 181.1 psf Fast failure#64 188.1 psf Fast failure#65 198.7 psf Fast failure#67 175.4 psf Fast failure#68 171.9 psf Fast failure#70 198.0 psf Fast failure#71 186.0 psf Fast failure#72 160.5 psf Fast failure#73 15 samples Mean : 174.5 psf Max : 198.7 psf Min : 115.3 psf STD : 22.75 psf COV : 0.13 Test dates : 2/22/2008 to 2/27/2008 Spacing : 12 inches O.C. interior ( 5 total) and at 6 inches O.C. edge (9 total) Fastener : 8d ring shank Figure B-12. Failure mode / location for panels fa stened with 8d ring shank nails spaced at 6 in. / 12 in. tested statically

PAGE 160

APPENDIX C PANEL CONSTRUCTION OF STATIC VS. DY NAMIC TESTING OF HARVESTED PANELS 2 samples Mean : 168 psf Max : 169 psf Min : 167 psf Test dates : 5/21/2009 to 5/22/2009 Spacing : 4 inches O.C. interior (13 total) and at 4 inches O.C. edge (13 total); series of retrofitted with A-2 Fastener : 1.5 in. staple tested statically & dynamically, and retrofitted with A-2 4 samples Mean : 71.9 psf Max : 93 psf Min : 52 psf STD : 19 psf COV : 0.26 4/4Dynamic ret. w/ A-2 4/4Dynamic 2 samples Mean : 93 psf Max : 93 psf Min : 93 psf 4/4Static 2 samples Mean : 126 psf Max : 150 psf Min : 102 psf 4/4Static ret. w/ A-2 12 in 2 ft 8 ft 4 ft 2 ft 8 ft 4 ft Truss 1 thru 5 Staple 1 thru 13 Figure C-1. Summary of missing fasteners from static vs. dynamic testing, Debary #1 series Figure C-1 presents results from observations of occurrences and locations of missing fasteners taken at time of harvesting. 160

PAGE 161

APPENDIX D FULL SPECIFIC GRAVITY MEASUREMENTS Table D-1. All specific gravity measurements taken from laboratory fabricated static vs. dynamic uplift testing # SG # SG # SG # SG 1 0.372 43 0.655 85 0.455 127 0.409 number 168 2 0.429 44 0.601 86 0.585 128 0.468 mean 0.52 3 0.425 45 0.598 87 0.472 129 0.465 St-Dev 0.09 4 0.632 46 0.572 88 0.479 130 0.676 COV 18% 5 0.586 47 0.603 89 0.493 131 0.429 6 0.694 48 0.506 90 0.632 132 0.421 7 0.622 49 0.498 91 0.574 133 0.432 8 0.656 50 0.548 92 0.672 134 0.539 9 0.624 51 0.602 93 0.657 135 0.651 10 0.500 52 0.532 94 0.615 136 0.587 11 0.436 53 0.529 95 0.539 137 0.757 12 0.428 54 0.509 96 0.515 138 0.787 13 0.540 55 0.495 97 0.466 139 0.667 14 0.573 56 0.480 98 0.491 140 0.539 15 0.547 57 0.582 99 0.493 141 0.575 16 0.457 58 0.593 100 0.482 142 0.582 17 0.456 59 0.479 101 0.478 143 0.473 18 0.468 60 0.644 102 0.411 144 0.508 19 0.601 61 0.495 103 0.464 145 0.446 20 0.591 62 0.609 104 0.482 146 0.457 21 0.589 63 0.689 105 0.609 147 0.477 22 0.532 64 0.523 106 0.579 148 0.377 23 0.565 65 0.517 107 0.400 149 0.368 24 0.577 66 0.486 108 0.406 150 0.378 25 0.692 67 0.573 109 0.420 151 0.458 26 0.408 68 0.574 110 0.462 152 0.440 27 0.537 69 0.536 111 0.430 153 0.393 28 0.573 70 0.390 112 0.454 154 0.589 29 0.482 71 0.419 113 0.516 155 0.449 30 0.458 72 0.412 114 0.563 156 0.523 31 0.461 73 0.483 115 0.442 157 0.552 32 0.471 74 0.613 116 0.450 158 0.537 161

PAGE 162

Table D-1. Continued # SG # SG # SG # SG 33 0.461 75 0.415 117 0.520 159 0.483 34 0.452 76 0.475 118 0.474 160 0.502 35 0.536 77 0.731 119 0.525 161 0.533 36 0.476 78 0.477 120 0.606 162 0.544 37 0.510 79 0.613 121 0.474 163 0.466 38 0.838 80 0.562 122 0.404 164 0.521 39 0.420 81 0.541 123 0.453 165 0.390 40 0.449 82 0.493 124 0.475 166 0.406 41 0.548 83 0.541 125 0.448 167 0.403 42 0.550 84 0.502 126 0.506 Table D-2. All specific gravity measurements taken from harvested static vs. dynamic uplift testing # SG # SG # SG # SG 1 0.513 46 0.519 91 0.563 135 0.592 number 178 2 0.562 47 0.590 92 0.709 136 0.570 mean 0.57 3 0.476 48 0.661 93 0.586 137 0.575 St-Dev 0.08 4 0.451 49 0.443 94 0.517 138 0.709 COV 14% 5 0.613 50 0.536 95 0.563 139 0.665 6 0.590 51 0.513 96 0.572 140 0.585 7 0.575 52 0.545 97 0.674 141 0.545 8 0.601 53 0.501 98 0.558 142 0.566 9 0.412 54 0.630 99 0.621 143 0.441 10 0.619 55 0.595 100 0.549 144 0.489 11 0.611 56 0.494 101 0.505 145 0.484 12 0.435 57 0.646 102 0.495 146 0.503 13 0.478 58 0.586 103 0.565 147 0.594 14 0.424 59 0.485 104 0.549 148 0.511 15 0.477 60 0.467 105 0.636 149 0.624 16 0.484 61 0.535 106 0.481 150 0.506 17 0.560 62 0.586 107 0.609 151 0.567 18 0.508 63 0.593 108 0.619 152 0.620 19 0.497 64 0.633 109 0.512 153 0.737 20 0.462 65 0.523 110 0.482 154 0.532 21 0.643 66 0.518 111 0.492 155 0.536 22 0.617 67 0.778 112 0.575 156 0.542 162

PAGE 163

Table D-2. Continued # SG # SG # SG # SG 23 0.619 68 0.601 113 0.523 157 0.532 24 0.470 69 0.631 114 0.652 158 0.536 25 0.646 70 0.613 115 0.599 159 0.655 26 0.693 71 0.504 116 0.441 160 0.615 27 0.634 72 0.561 117 0.675 161 0.519 28 0.529 73 0.455 118 0.462 162 0.670 29 0.705 74 0.684 119 0.599 163 0.572 30 0.452 75 0.487 120 0.644 164 0.666 31 0.425 76 0.541 121 0.713 165 0.622 32 0.526 77 0.590 122 0.611 166 0.609 33 0.285 78 0.622 123 0.590 167 0.556 34 0.562 79 0.555 124 0.437 168 0.590 35 0.596 80 0.576 125 0.450 169 0.481 36 0.567 81 0.544 126 0.573 170 0.548 37 0.564 82 0.560 127 0.698 171 0.505 38 0.632 83 0.625 128 0.668 172 0.723 39 0.667 84 0.601 129 0.571 173 0.630 40 0.439 85 0.596 130 0.571 174 0.436 41 0.529 86 0.699 131 0.670 175 0.548 42 0.672 87 0.585 132 0.546 176 0.539 43 0.657 88 0.575 133 0.655 177 0.644 44 0.605 89 0.450 134 0.549 178 0.573 45 0.457 90 0.526 163

PAGE 164

APPENDIX E STATIC VS. DYNAMIC PANEL TESTING, TARGET AND ACTUAL PRESSURE TIMEHISTORIES Laboratory Fabricated 6d Smooth Shank at 6 in. / 12 in. Static (5 panels) 0 10 20 30 40 50 60 0 10 20 30 40 50 Time (sec.)Pressure (psf)6/12Static Peak Pressure = 47 psf Failure Pressure = 47 psf Figure E-1. Summary of pressu re time-history for static 6d SS at 6/12 panel (6d SS-38) 0 10 20 30 40 50 60 70 10 0 10 20 30 40 50 60 70 Time (sec.)Pressure (psf)6d SS @ 6 in. / 12 in. Static Peak Pressure = 58 psf Failure Pressure = 58 psf Figure E-2. Summary of pressu re time-history for static 6d SS at 6/12 panel (6d SS-40) 164

PAGE 165

0 10 20 30 40 50 60 70 80 10 0 10 20 30 40 50 60 70 80 Time (sec.)Pressure (psf)6d SS @ 6 in. / 12 in. Static Peak Pressure = 64.8 psf @ 61 sec. Failure Pressure = 63.3 psf @ 70 sec. Figure E-3. Summary of pressu re time-history for static 6d SS at 6/12 panel (6d SS-41) 0 10 20 30 40 50 60 70 10 0 10 20 30 40 50 60 70 Ti m e (s e c ) Pressure (psf)6d SS@ 6 in. / 12 in. Static Peak Pressure = 65 psf Failure Pressure = 65 psf Figure E-4. Summary of pressu re time-history for static 6d SS at 6/12 panel (6d SS-42) 165

PAGE 166

0 10 20 30 40 50 60 70 80 10 0 10 20 30 40 50 60 70 80 Time (sec.)Pressure (psf)6d SS @ 6 in. / 12 in. Static Peak Pressure = 73 psf Failure Pressure = 73 psf Figure E-5. Summary of pressu re time-history for static 6d SS at 6/12 panel (6d SS-39) 166

PAGE 167

Laboratory Fabricated 6d Smooth Shank at 6 in. / 12 in. Dynamic (5 panels) Table E-1. Summary of dynamic pressure trace for 6d SS at 6 in / 12 in. panels (blue actual and red target) Peak 1 Peak 2 P eak 3 Peak Failure Ratio Failure to Peak 6d SS at 6 in. / 12 in. Dynamic 6d SS-43 34.5 psf @ 52 sec. 33 psf @ 58 sec. 34 psf @ 59 sec. 34.5 psf @ 52 sec. 28 psf @ 60 sec. 0.96 37.7 psf @ 52 sec. 38 psf @ 58 sec. 51 psf @ 59 sec. 37.7 psf @ 52 sec. 33 psf @ 60 sec. 6d SS-44 48 psf @ 62 sec. 47 psf @ 67 sec. 51 psf @ 68 sec. 51 psf @ 68 sec. 48 psf @ 72 sec. 0.94 50 psf @ 62 sec. 50 psf @ 67 sec. 65 psf @ 68 sec. 65 psf @ 68 sec. 63 psf @ 72 sec. 6d SS-45 50 psf @ 62 sec. 43 psf @ 66 sec. 46 psf @ 67 sec. 50 psf @ 62 sec. 46 psf @ 67.5 sec. 0.92 50 psf @ 62 sec. 41 psf @ 66 sec. 49 psf @ 67 sec. 50 psf @ 62 sec. 66 psf @ 67.5 sec. 6d SS-46 51 psf @ 62 sec. 51 psf @ 67.5 sec. 46 psf @ 69 sec. 58 psf @ 72 sec. 58 psf @ 72 sec. 1.00 50 psf @ 62 sec. 66 psf @ 67.5 sec. 46 psf @ 69 sec. 63 psf @ 72 sec. 63 psf @ 72 sec. 6d SS-47 60 psf @ 72 sec. 62 psf @ 76.5 sec. 55 psf @ 79.5 sec. 66 psf @ 82 sec. 66 psf @ 82 sec. 1.00 63 psf @ 72 sec. 79 psf @ 76.5 sec. 62 psf @ 79.5 sec. 75 psf @ 82 sec. 75 psf @ 82 sec. 167

PAGE 168

0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 Time (sec.)Pressure (psf)6/12 Dynamic A 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 0 10 20 30 40 50 60 70 Time (sec.)Pressure ( ps f) 6/12 Dynamic Close up of Failure Peak Pressure = 34.5 psf at 52.4 sec. Failure Pressure = 33 psf at 60 sec. P1 P2 P3 B Figure E-6. Summary of pressu re time-histories for dynamic 6d SS at 6/12 panel (6d SS-43) A) full time-history and B) close up of failure 168

PAGE 169

0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 Time (sec.)Pressure (psf)6 / 12 Dynamic A 60 62 64 66 68 70 72 74 76 78 80 0 10 20 30 40 50 60 70 80 Time (sec.)Pressure (psf)6/12 Dynamic Close up Peak Pressure = 51 psf at 67.5 sec. Failure Pressure = 48 psf at 72 sec. P1 P2 P3 B Figure E-7. Summary of pressu re time-histories for dynamic 6d SS at 6/12 panel (6d SS-44) A) full time-history and B) close up of failure 169

PAGE 170

A) 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 Time (sec.)Pressure (psf)6/12 Dynamic A 50 52 54 56 58 60 62 64 66 68 70 0 10 20 30 40 50 60 70 Ti m e (s e c ) Pressure (psf)6/12 Dynamic Close up Peak Pressure = 50 psf at 62 sec. Failure Pressure = 46 psf at 67.5 sec. P1 P2 P3 B Figure E-8. Summary of pressu re time-histories for dynamic 6d SS at 6/12 panel (6D SS-45) A) full time-history and B) close up of failure 170

PAGE 171

0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 Time (sec.)Pressure (psf)6/12 Dynamic A 60 62 64 66 68 70 72 74 76 78 80 0 10 20 30 40 50 60 70 80 Time (sec.)Pressure (psf)6/12 Dynamic Peak Pressure = 58 psf at 72 sec. Failure Pressure = 58 psf at 72 sec. P1 P2 P3 B Figure E-9. Summary of pressu re time-histories for dynamic 6d SS at 6/12 panel (6d SS-46) full time-history and B) close up of failure 171

PAGE 172

0 10 20 30 40 50 60 70 80 90 10 0 10 20 30 40 50 60 70 80 Time (sec.)Pressure (psf)6d SS @ 6 in. / 12 in. Dynamic A 70 72 74 76 78 80 82 84 86 88 90 10 20 30 40 50 60 70 80 90 Time ( sec. ) Pressure (psf)6/12 Dynamic Close up Peak Pressure = 66 psf at 82 sec.Failure Pressure = 66 psf at 82 sec. P1 P2 P3 B Figure E-10. Summary of pressu re time-histories for dynamic 6d SS at 6/12 panel (6d SS-47) A) full time-history and B) close up of failure 172

PAGE 173

Laboratory Fabricated 6d Smooth Shank at 6 in. / 6 in. Static (5 panels) 0 10 20 30 40 50 60 70 80 90 100 0 20 40 60 80 100 120 Time (sec.)Pressure (psf)6/6 Static Peak Pressure = 106 psf Failure Pressure = 106 psf Figure E-11. Summary of pre ssure time-history for static 6d SS at 6/6 panel (6d SS-53) 0 10 20 30 40 50 60 70 10 0 10 20 30 40 50 60 70 Time (sec.)Pressure (psf)Static 6dC @ 6 in. / 6 in. Peak Pressure = 64 psf Failure Pressure = 64 psf Figure E-12. Summary of pre ssure time-history for static 6d SS at 6/6 panel (6d SS-52) 173

PAGE 174

0 20 40 60 80 100 120 0 20 40 60 80 100 120 140 Time (sec.)Pressure (psf) 6d SS @ 6 in. / 6 in. Static Peak Pressure = 134 psf Failure Pressure = 134 psf Figure E-13. Summary of pressu re time-history for static 6d SS at 6/6 panel (6d SS-56) 0 10 20 30 40 50 60 70 80 90 20 0 20 40 60 80 100 120 Ti m e (s e c ) Pressure (psf)6d SS @ 6 in. / 6 in. Static Peak Pressure = 91 psf Failure Pressure = 91 psf Figure E-14. Summary of pressu re time-history for static 6d SS at 6/6 panel (6d SS-55) 174

PAGE 175

0 20 40 60 80 100 120 0 20 40 60 80 100 120 140 Time (sec.)Pressure (psf)6d SS @ 6 in. / 6in. Static Peak Pressure = 135 psf Failure Pressure = 135 psf Figure E-15. Summary of pressu re time-history for static 6d SS at 6/6 panel (6d SS-54) Laboratory Fabricated 6d Smooth Shank at 6 in. / 6 in. Dynamic (5 panels) Table E-2. Summary of dynamic pr essure trace for 6d SS at 6 in. / 6 in. panels (blue actual and red target) Peak 1 Peak 2 Peak 3 Peak Failure 6d SS at 6 in. / 6 in. Dynamic 6d SS57 38 psf @ 52 sec. 35 psf @ 58 sec. 40 psf @ 59 sec. 44 psf @ 62 sec. 44 psf @ 62 sec. 38 psf @ 52 sec. 38 psf @ 58 sec. 52 psf @ 59 sec. 50 psf @ 62 sec. 50 psf @ 62 sec. 6d SS58 101 psf @ 101.5 sec. 96 psf @ 105 sec. 92 psf @ 107 sec. 103 psf @ 111 sec. 103 psf @ 111 sec. 98 psf @ 101.5 sec. 110 psf @ 105 sec. 97 psf @ 107 sec. 112 psf @ 111 sec. 112 psf @ 111 sec. 6d SS59 102 psf @ 101.5 sec. 97 psf @ 105 sec. 93 psf @ 107.5 sec. 105 psf @ 111.5 sec. 105 psf @ 111.5 sec. 98 psf @ 101.5 sec. 110 psf @ 105 sec. 97 psf @ 107.5 sec. 112 psf @ 111.5 sec. 112 psf @ 111 sec. 6d SS60 64 psf @ 72 sec. 64 psf @ 77 sec. 60 psf @ 79 sec. 72 psf @ 82 sec. 59 psf @ 82.5 sec. 63 psf @ 72 sec. 79 psf @ 77 sec. 62 psf @ 79 sec. 75 psf @ 82 sec. 60 psf @ 82.5 sec. 6d SS61 114 psf @ 111 sec. 107 psf @ 115 sec. 102 psf @ 117 sec. 124 psf @ 121 sec. 104 psf @ 124 sec. 112 psf @ 111 sec. 128 psf @ 115 sec. 114 psf @ 117 sec. 128 psf @ 121 sec. 102 psf @ 124 sec. 175

PAGE 176

0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 Time (sec.)Pressure (psf)6/6Dynamic A 50 52 54 56 58 60 62 64 66 68 70 0 10 20 30 40 50 60 70 Time (sec.)Pressure (psf)6/6 Dynamic Close up Peak Pressure = 44 psf at 62 sec. P1 P2 P3 B Figure E-16. Summary of pressu re time-histories for dynamic 6d SS at 6/6 panel (6d SS-57) A) full time-history and B) close up of failure 176

PAGE 177

0 10 20 30 40 50 60 70 80 90 100 110 120 0 20 40 60 80 100 120 Time (sec.)Pressure (psf)6/6 Dynamic A 100 102 104 106 108 110 112 114 116 118 120 0 20 40 60 80 100 120 140 Time (sec.)Pressure (psf)6/6 Dynamic Close up P1 P2 P3 Peak Pressure = 103 psf at 111 sec. Failure Pressure = 103 psf at 111 sec. B Figure E-17. Summary of pressu re time-histories for dynamic 6d SS at 6/6 panel (6d SS-58) A) full time-history and B) close up of failure 177

PAGE 178

0 20 40 60 80 100 120 0 20 40 60 80 100 120 Time (sec.)Pressure (psf)6/6 Dynamic A 100 102 104 106 108 110 112 114 116 118 120 0 20 40 60 80 100 120 140 Time (sec.)Pressure (psf)6/6 Dynamic Close up P1 P2 P3 Peak Pressure = 105 psf at 111.5 sec. Failure Pressure = 105 psf at 111.5 sec. B Figure E-18. Summary of pressu re time-histories for dynamic 6d SS at 6/6 panel (6d SS-59) A) full time-history and B) close up of failure 178

PAGE 179

0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 Time (sec.)Pressure (psf)6/6 Dynamic A 70 72 74 76 78 80 82 84 86 88 90 0 10 20 30 40 50 60 70 80 90 Time (sec.)Pressure (psf)6/6 Dynamic Close up P1 P3 Peak Pressure = 72 psf at 82 sec. Failure Pressure = 53 psf at 85 sec. P2 B Figure E-19. Summary of pressu re time-histories for dynamic 6d SS at 6/6 panel (6d SS-60) A) full time-history and B) close up of failure 179

PAGE 180

0 10 20 30 40 50 60 70 80 90 100 110 120 130 0 20 40 60 80 100 120 140 Time (sec.)Pressure (psf)6d SS @ 6 in. / 6 in. Dynamic A 110 112 114 116 118 120 122 124 126 128 130 0 20 40 60 80 100 120 140 Ti m e (s e c ) Pressure (psf)6/6 Dynamic Close up P1 P3 Peak Pressure = 124 psf at 121 sec. Failure Pressure = 104 psf at 124 sec. P2 B Figure E-20. Summary of pressu re time-histories for dynamic 6d SS at 6/6 panel (6d SS-61) A) full time-history and B) close up of failure 180

PAGE 181

Laboratory Fabricated 6d Smooth Shank at 6 in. / 12 in. Ret. A-2 Static (2 panels) 0 20 40 60 80 100 120 140 160 0 25 50 75 100 125 150 175 200 Time (sec.)Pressure (psf)Retrofitted 6dC @ 6 / 12 Static Peak Pressure = 199 psf @ 150 sec. Failure Pressure = 195 psf @ 154 sec. Figure E-21. Summary of pressu re time-history for static 6d SS at 6/12 ret. A-2 panel (6d SS48) 0 20 40 60 80 100 120 140 0 25 50 75 100 125 150 175 200 Time (sec.)Pressure (psf)Retrofitted 6d SS @ 6 in. / 12 in. Static Peak Pressure = 167 psf Failure Pressure = 167 psf Figure E-22. Summary of pressu re time-history for static 6d SS at 6/12 Ret. A-2 panel (6d SS49) 181

PAGE 182

Laboratory Fabricated 6d Smooth Shank at 6 in. / 12 in. Ret. A-2 Dynamic (2 panels) Table E-3. Summary of dynamic pr essure trace for 6d SS at 6 in. / 12 in. retrofitted panels (blue actual and red target) Peak 1 Peak 2 P eak 3 Peak Failure 6d SS at 6 in. / 12 in. Retrofit Dynamic 6d SS-50 159 psf @ 161 sec. 155 psf @ 164 sec. 162 psf @ 166 sec. 164 psf @ 171 sec. 164 psf @ 171 sec. 181 psf @ 161 sec. 175 psf @ 164 sec. 152 psf @ 166 sec. 194 psf @ 171 sec. 194 psf @ 171 sec. 6d SS-51 160 psf @ 161 sec. 157 psf @ 164 sec. 152 psf @ 166 sec. 171 psf @ 173.5 sec. 162 psf @ 173.8 sec. 181 psf @ 161 sec. 175 psf @ 164 sec. 162 psf @ 166 sec. 169 psf @ 173.5 sec. 174 psf @ 173.8 sec. 182

PAGE 183

0 20 40 60 80 100 120 140 160 180 0 25 50 75 100 125 150 175 200 Time (sec.)Pressure (psf)Retrofitted 6dSS @ 6 in. / 12 in. Dynamic A 160 162 164 166 168 170 172 174 176 178 180 0 25 50 75 100 125 150 175 200 Time (sec.)Pressure (psf)Retrofitted 6dSS @ 6 in. / 12 in. Dynamic P1 P3 Peak Pressure = 164 psf at 171 sec. Failure Pressure = 164 psf at 171 sec. P2 B Figure E-23. Summary of pressu re time-history for dynamic 6d SS at 6/12 ret. A-2 panel (6d SS-50) A) full time-history and B) close up of failure 183

PAGE 184

0 20 40 60 80 100 120 140 160 180 0 25 50 75 100 125 150 175 200 Ti m e (s e c ) Pressure (psf)Retroiftted 6dC @ 6 in. / 12 in. A 160 162 164 166 168 170 172 174 176 178 180 0 20 40 60 80 100 120 140 160 180 200 Time (sec.)Pressure (psf)Retrofitted 6/12 Dynamic Close up P1 P3 Peak Pressure = 171 psf at 173.5 sec. Failure Pressure = 162 psf at 173.8 sec. P2 B Figure E-24. Summary of pressu re time-history for dynamic 6d SS at 6/12 ret. A-2 panel (6d SS-51) A) full time-history and B) close up of failure 184

PAGE 185

Harvested 1.5 in. Staple at 4 in. / 4 in. Static (2 panels) 0 10 20 30 40 50 60 70 80 90 20 0 20 40 60 80 100 Time (sec.)Pressure (psf)4/4Static Peak Pressure = 93 psf Failure Pressure = 90 psf Figure E-25. Summary of pressure time-history for static 1.5 in. St aple at 4/4 (1.5 in. Staple-7) 0 10 20 30 40 50 60 70 80 90 20 0 20 40 60 80 100 Time (sec.)Pressure (psf)4/4 Static Peak Pressure = 93 psf Failure Pressure = 90 psf Figure E-26. Summary of pressure time-history for static 1.5 in. St aple at 4/4 (1.5 in. Staple-8) 185

PAGE 186

Harvested 1.5 in. Staple at 4 in. / 4 in. Dynamic (4 panels) Table E-4. Summary of dynamic pr essure trace for 1.5 in. Staples at 4/4 panels (blue actual and red target) Peak 1 Peak 2 P eak 3 Peak Failure 1.5 in. Staple9 92 psf @ 92 sec. 84 psf @ 96 sec. 82 psf @ 98 sec. 93 psf @ 102 sec. 93 psf @ 102 sec. 86 psf @ 92 sec. 105 psf @ 96 sec. 89 psf @ 98 sec. 98 psf @ 102 sec. 98 psf @ 102 sec. 1.5 in. Staple10 55 psf @ 62 sec. 49 psf @ 67 sec. 54 psf @ 68 sec. 61 psf @ 72 sec. 61 psf @ 72 sec. 50 psf @ 62 sec. 49 psf @ 67 sec. 65 psf @ 68 sec. 63 psf @ 72 sec. 63 psf @ 72 sec. 1.5 in. Staple11 42 psf @ 52 sec. 36 psf @ 58 sec. 42 psf @ 59 sec. 52 psf @ 62 sec. 37 psf @ 66 sec. 38 psf @ 52 sec. 38 psf @ 58 sec. 52 psf @ 59 sec. 50 psf @ 62 sec. 41 psf @ 66 sec. 1.5 in. Staple12 69 psf @ 72 sec. 63 psf @ 77 sec. 61 psf @ 79 sec. 82 psf @ 82 sec. 71 psf @ 89 sec. 63 psf @ 72 sec. 78 psf @ 77 sec. 62 psf @ 79 sec. 75 psf @ 82 sec. 75 psf @ 89 sec. 186

PAGE 187

0 10 20 30 40 50 60 70 80 90 100 110 20 0 20 40 60 80 100 120 Time (sec.)Pressure (psf)4/4 Dynamic A 90 92 94 96 98 100 102 104 106 108 110 0 20 40 60 80 100 120 Time (sec.)Pressure (psf)4/4 Dynamic Close up P1 P3 Peak Pressure = 93 psf at 102 sec. Failure Pressure = 93 psf at 102 sec. P2 B Figure E-27. Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 (1.5 in Staple9) A) full time-history and B) close up of failure 187

PAGE 188

0 10 20 30 40 50 60 70 80 20 10 0 10 20 30 40 50 60 70 Time (sec.)Pressure (psf)4/4 Dynamic A 60 62 64 66 68 70 72 74 76 78 80 0 10 20 30 40 50 60 70 80 Time (sec.)Pressure (psf)4/4Dynamic Close up P1 P3 Peak Pressure = 61 psf at 72 sec. Failure Pressure = 61 psf at 72 sec. P2 B Figure E-28. Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 (1.5 in. Staple10) A) full time-history and B) close up of failure 188

PAGE 189

0 10 20 30 40 50 60 70 20 10 0 10 20 30 40 50 60 Ti m e (s e c ) Pressure (psf)4/4Dynamic A 50 52 54 56 58 60 62 64 66 68 70 0 10 20 30 40 50 60 70 Time (sec.)Pressure (psf)4/4 Dynamic Close up P1 P3 Peak Pressure = 52 psf at 62 sec. Failure Pressure = 37 psf at 66 sec. P2 B Figure E-29. Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 (1.5 in. Staple11) A) full time-history and B) close up of failure 189

PAGE 190

0 10 20 30 40 50 60 70 80 90 20 0 20 40 60 80 100 Time (sec.)Pressure (psf)4/4Dynamic A 70 72 74 76 78 80 82 84 86 88 90 0 10 20 30 40 50 60 70 80 90 Time (sec.)Pressure (psf)4/4 Dynamic Close up P1 P3 Peak Pressure = 82 psf at 82 sec. Failure Pressure = 71 psf at 89 sec. P2 B Figure E-30. Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 (1.5 in. Staple12) A) full time-history and B) close up of failure 190

PAGE 191

Harvested 1.5 in. Staple at 4 in. / 4 in. Ret. with A-2 Static (2 panels) 0 20 40 60 80 100 120 140 20 0 20 40 60 80 100 120 140 160 180 Time (sec.)Pressure (psf)4/4 Ret. Static Peak Pressure = 168 psf Failure Pressure = 165 psf Figure E-31. Summary of pressure time-history for static 1.5 in. Staple at 4/4 with Ret. A-2 (1.5 in. Staple-13) 0 20 40 60 80 100 120 140 20 0 20 40 60 80 100 120 140 160 180 Time (sec.)Pressure (psf)4/4 Ret. Static Peak Pressure = 169 psf Failure Pressure = 165 psf Figure E-32. Summary of pressure time-history for static 1.5 in. Staple at 4/4 with Ret. A-2 (1.5 in. Staple-14) 191

PAGE 192

Harvested 1.5 in. Staple at 4 in. / 4 in. with Ret. A-2 Dynamic (2 panels) Table E-5. Summary of dynamic pr essure trace for 1.5 in. Staples at 4/4 with Ret. A-2 panels (blue actual and red target) Peak 1 Peak 2 Peak 3 Peak Pressure Failure Pressure 1.5 in. Staple15 141 psf @ 151 sec. 140 psf @ 154 sec. 144 psf @ 156 sec. 150 psf @ 161 sec. 150 psf @ 161 sec. 160 psf @ 151 sec. 164 psf @ 154 sec. 141 psf @ 156 sec. 181 psf @ 161 sec. 181 psf @ 161 sec. 1.5 in. Staple16 95 psf @ 92 sec. 85 psf @ 96 sec. 82 psf @ 98 sec. 102 psf @ 102 sec. 90 psf @ 105 sec. 86 psf @ 92 sec. 105 psf @ 96 sec. 89 psf @ 98 sec. 98 psf @ 102 sec. 94 psf @ 105 sec. 192

PAGE 193

0 20 40 60 80 100 120 140 160 180 20 0 20 40 60 80 100 120 140 160 180 200 Time (sec.)Pressure (psf)4/4 Ret. Dynamic A 150 152 154 156 158 160 162 164 166 168 170 0 20 40 60 80 100 120 140 160 180 200 Time (sec.)Pressure (psf)4/4 Ret. Dynamic P1 P3 Peak Pressure = 150 psf at 161 sec. Failure Pressure = 150 psf at 161 sec. P2 B Figure E-33. Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 with Ret. A-2 (1.5 in. Staple-15) A) full timehistory and B) close up of failure 193

PAGE 194

194 0 20 40 60 80 100 120 20 0 20 40 60 80 100 120 Ti m e (s e c ) Pressure (psf)4/4 Ret. Dynamic A 90 92 94 96 98 100 102 104 106 108 110 0 10 20 30 40 50 60 70 80 90 100 110 120 Ti m e (s e c ) Pressure (psf)4/4 Ret. Dynamic Close up P1 P3 Peak Pressure = 102 psf at 102 sec. Failure Pressure = 90 psf at 105 sec. P2 B Figure E-34. Summary of pressure time-history for dynamic 1.5 in. Staple at 4/4 with Ret. A-2 (1.5 in. Staple-16) A) full timehistory and B) close up of failure

PAGE 195

LIST OF REFERENCES AF&PA. (2005). "National Design Specification for Wood Construction." Chapter 11: DowelType Fasteners, AF & PA Ameri can Wood Council, Washington DC. Amirkhanian, S., Sparks, P., and Watford, S. (1994) "A Statistical Analysis of Wind Damage to Single-Family Dwellings Due to Hurricane Hugo." Structures Congress XII. ASCE. (2006). Minimum Design Loads for Buildings and Other Structures (ASCE/SEI Standard 7-05), American Society of Civil Engineers, Reston, VA. ASTM. (2000). "Evaluating Properties of Wood-Based Fiber and Panel Materials." Nail Head Pull-Through Test. ASTM. (2006a). "Section 4: Construction." D1761: Mechanical Fasteners in Wood. ASTM. (2006b). "Standard Test Methods for Specific Gravity of Wood and Wood-Based Materials, D 2395 -02." Annual B ook of ASTM Standards 2006. ASTM. (2006c). "Standard Test Mwethods for Ev aluating Properties of Wood-Based Fiber and Particle Panel Materials." D 1037-99. Bartlett, F. M., Galsworthy, J. K., Henderson, D., Hong, H. P., Iizumi, E., Inculet, D. R., Kopp, G. A., Morrison, M. J., Savory, E., Sabarinathan J., Sauer, A., Scott, J., St. Pierre, L. M., and Surry, D. (2007). "The Three Little Pigs Project: A New Test Facility for Full-Scale Small Buildings." 12th Intern ational Conference on Wind E ngineering, Australasian Wind Engineering Societ y, Cairns, Australia. Baskaran, A., Chen, Y., and Vilaipornsawai, U. (1999a). "A New Dynamic Wind Load Cycle to Evaluate Mechanically Attached Flexible Membrane Roofs." Journal of Testing and Evaluation, 27(4), 249-265. Baskaran, A., and Dutt, O. (1997). "Performance of roof fasteners under simulated loading conditions." Journal of Wind Engineeri ng and Industrial Aerodynamics, 72, 389-400. Baskaran, A., Lee, W., and Richardson, C. (1999b). "Dynamic Evaluation of Thermoplastic Roofing System for Wind Performance." Journal of Architectural Engineering, 16-24. Chow, P., McNatt, J. D., and Zhao, L. (1990). "E ffects of outdoor weathe ring on withdrawal and head pull-through of nails and staple s in wood-based building panels." Durability of Building Materials and Components, 259-259. Chui, Y. H., and Craft, S. (2002). "Fastener head pull-through resi stance of plywood and oriented strand board." Canadian Journal of Civil Engineering, 29(3), 384-388. Cook Jr, R. L. (1991). "Lessons learned by a roof consultant." Hurricane Hugo One Year Later. 195

PAGE 196

Cunningham, T. P. (1993). "Roof Sheathing Fastening Schedules for Wind Uplift." American Plywood Association. Dade County. (1988). South Florida Building Code, Dade County Board of County Commissioners, Miami, Dade County, FL. Dade County. (1994). South Florida Building Code, Dade County Board of County Commissioners, Miami, Dade County, FL. Dao, T. N., and Van De Lindt, J. W. (2008). "N ew nonlinear roof sheathing fastener model for use in finite-element wind load applications." Journal of Structural Engineering, 134(10), 1668-1674. Datin, P. L., and Prevatt, D. O. "Wind Uplift Reactions at Roof-to-Wa ll Connections of WoodFramed Gable Roof Assembly." International Conference on Wind Engineering, Cairns, Australia. Datin, P. L., and Prevatt, D. O. (2009). "Equi valent Roof Panel Wind Loading for Full-Scale Sheathing Testing." 11th Americas Conference on Wind Engineering, Puerto Rico. Feldborg, T. (1989). "Timber joints in tension an d nails in withdrawal under long-term loading and alternating humidity." Forest Products Journal, 39(11-12), 8-12. FEMA, Mitigation, Assessment, and Team. ( 2005a). "Hurricane Charlie in Florida." FEMA, Mitigation, Assessment, and Team. (2005b). "Hurricane Ivan in Alabama and Florida: Observations, Recommendations and Technical Guidance." FEMA, Mitigation, Assessment, and Team. (2006). "Hurricane Katrina in the Gulf Coast: Mitigation Assessment Team Report, Building Performance Observations, Recommendations, and Technical Guidance." FEMA, M. A. T. (1993). "Building Performance: Hurricane Iniki in Hawaii." Feng, Y., D'Amours, S., and Beauregard, R. (200 8). "The value of sales and operations planning in oriented strand board industry with make-to-order manufacturing system: Cross functional integrati on under deterministic demand and spot market recourse." International Journal of Production Economics, 115, 20. Forest Products Laboratory. (1999). Wood Handbook: Wood as an Engineering Material, U.S. Department of Agriculture, Forest Service, Madison, WI. Holmes, J. D. (2001). Wind Loading of Structures, Spon Press, New York. ICC. (2000). International Building Code, International Code Council, Inc., Falls Church, VA. ICC. (2004). Florida Building Code, Building, International Code Council, Inc., Falls Church, VA. 196

PAGE 197

ICC. (2006). 2006 International Building Code, International Code Council, Falls Church, VA. ICC. (2007). Florida Building Code. IHRC. (2004). "Hurricane Loss Reduction for Housi ng in Florida: Final Report." International Hurricane Research Center, Florida In ternational University, Miami, FL. Jesteadt, J. P. (2006). "Wind Data Collection, Wind Resistance of Florica Residential Structures, and Simulation of Hurricane Force WInds : 2003 to 2006 Florida Coastal Monitoring Program (FCMP)," University of Florida, Gainesville. Jones, D. T. (1998). "Retrofit Techniques Using Adhesives to Resist Wind Uplift in Wood Roof Systems," MS Thesis, Clemson University, Clemson, SC. Judge, E. K., and Reinhold, T. A. "Practicality and Effectiveness of Wind Resistant Retrofit Measures." Solutions to Coastal Di sasters 2002, Feb 24-27 2002, San Diego, CA, United States, p 983-996. Kallem, M. R. (1997). "Roof Sheathing Attach ment for High Wind Regions: Comparison of Screws and Nails," Clemson University, Clemson. Keith, E. L., and Rose, J. D. (1992). "Hurricane A ndrew Structural Perfor mance of Buidlings in Southern Florida." American Plywood Association, Report T92-21. Kemp, G. (2008). "New hurricane test facility allows repeatable full-scale building testing." Test Engineering and Management, 70(4), 10-13. Mizzell, D. P. (1994). "Wind Resistance of Sheathing for Residential Roofs," Clemson University, Clemson. Murden, J. A. (1991). "Hugo 1989. The performan ce of structures in the wind." Hurricane Hugo One Year Later. Murphy, S., Schiff, S., Rosowsky, D., and Pye, S. "System effects and uplift capacity of roof sheathing fasteners." Part 2 (of 2), Chicago, IL, 765. NAHB. (2003). "Roof Sheathing Connection Toleranc es." Prepared for the US Department of Housing and Urban Development by the NAHB Research Center, Upper Marlboro, MD. Ott, R. L., and Longnecker, M. T. (2004). A First Course in Statistical Methods. Pielke, R. A., Gratz, J., Landsea, C. W., Collin s, D., Saunders, M. A., and Musulin, R. (2008). "Normalized hurricane damage in the United States: 1900-2005." Natural Hazards Review, 9(1), 29-42. Pinelli, J.-P., Simiu, E., Gurley, K., Subramania n, C., Zhang, L., Cope, A., Filliben, J. J., and Hamid, S. (2004). "Hurricane damage predic tion model for residential structures." Journal of Structural Engineering, 130(11), 1685-1691. 197

PAGE 198

Pye, S. J. (1995). "Effect of In-Service Cond itions on the Withdrawal Capacity of Roof Sheathing Fasteners," Clemson University, Clemson. Rammer, D. R., Winistorfer, S. G., and Bender, D. A. (2001). "Withdrawal strength of threaded nails." Journal of Structural Engineering, 127(4), 442-449. Reinhold, T., Mitrani, J., Alvarez, R., and Su tt, E. "Guidlines for Design of Roof Sheathing Fastener Schedules to Resist Uplift in High Winds." 11th Internation Conference on Wind Engineering, Lubbock, TX. SBCCI. (1997). Standard Building Code, Southern Building Code Congress International, Birmingham, AL. Schiff, S. D., Mizzell, D. P., and Zaitz, M. D. "Up-Lift Resistance of Plywood and OSB Roof Sheathing." Structures Congress, Atlanta. Sherman, M. D. (2000). "Effect Of Loading Ra te On Capacities Of Nailed Connections And Fasteners In Wood," Clemson University, Clemson. Sparks, P. R. (1991a). "Damages and lessons learned from Hurricane Hugo." NIST Special Publication(n 820), p 435-450. Sparks, P. R. (1991b). "Facts about Hurricane Hugo. What it was, what it wasn't and why it caused so much damage." Hurricane Hugo One Year Later. Sutt, E. (2000). "The Effect of Combined Sh ear and Uplift Forces on Roof Sheathing Panels," Clemson University, Clemson,SC. Sutt, E., Reinhold, T., and Rosowsky, D. "The effect of in-situ conditions on nail withdrawal capacities." World Conference of Timber Engineering, Whistler, BC. US Census Bureau. (2003). "Census of populati on and housing (2000)." US Dept. of Commerce, Economics and Statistics Admini stration, Washington, D.C. van de Lindt, J. W., Graettinger, A., Gupta, R ., Skaggs, T., Pryor, S., and Fridley, K. (2007). "Performance of Woodframe Stru ctures During Hurricane Katrina." Journal of Performance of Constructed Facilities, March/April. 198

PAGE 199

199 BIOGRAPHICAL SKETCH The author was born in Davis, California, in 1983. From there he moved with his family to Nigeria then Tanzania. He relocated to the Washington DC area where he played football and skied at every chance. In 2002 he began attendi ng Clemson University where he received his Bachelors of Science in civil engineering. While at Clemson he worked at the Wind Load Test Facility conducting resear ch with the wind tunnel. In 2007 he began attending the University of Florida where he received his Master of Engineering degree in 2009.