The document summarizes the doctoral dissertation defense of Iván R. Canino regarding aerodynamic load testing of fiber reinforced polymer connections and metal fasteners for hurricane damage mitigation. It provides background on structural deficiencies in roof-to-wall connections, describes development and testing of a new FRP connection, and outlines wind tunnel testing of a full-scale building specimen at various angles of attack to evaluate the FRP connection's performance. Instrumentation and testing protocols are also summarized.
1. Aerodynamic Load Characteristics Evaluation and Tri-Axial Performance Testing on Fiber Reinforced Polymer Connections and Metal Fasteners to Promote Hurricane Damage Mitigation Doctoral Dissertation Defense By Iván R. Canino Major Advisor: Dr. Arindam Gan Chowdhury November 13, 2009 Civil & Environmental Engineering Florida International University Miami, Florida
2. Outline Background Structural Timber Roof-to-Wall Connection Deficiency Current Connection and Testing Methods Previous Development of Fiber Reinforced Polymer (FRP) Connection Development of FRP Connection using Component and Full-Scale Specimens Wall of Wind Testing Test Specimen Instrumentation Test Protocol Full-Scale Aerodynamic Load Characteristics for Roof-to-Wall Connections Full-Scale Test Results Discussion on Aerodynamic Characteristics Titan Structures and Construction Laboratory tri-Axial Testing New Tri-Axial Test Protocol Test Setup Results Conclusions Future Work & Recommendations FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
4. Background Comparison of Hurricane Losses (Dantin, 2006) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
5.
6. Background Wind Dynamics Around A Building FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
7. Background Toe-Nailed Connection Hurricane Clip Past and Current Roof-to-Wall Connection Systems FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
8. Background Miami Dade County Notice of Approval (NOA) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
9. Background Up-Lift Tests Up-Lift Tests Close-Up Current Roof-to-Wall Connection Component Testing Methods (Simpson) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
10. Background Parallel to the Wall (Top-Plate) L1 Current Roof-to-Wall Connection Component Testing Methods (Simpson) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
11. Background Perpendicular to the Wall (Top-Plate) L2 Unidirectional Component Tests Up-lift Lateral to wall L1 Perpendicular to wall Note: Current testing methods do not take into consideration the effects of simultaneous loading. Current Roof-to-Wall Connection Component Testing Methods (Simpson) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
12. Background NIST Full Scale Up-Lift & Lateral Tests Current Roof-to-Wall Connection Full-Scale Testing Methods (NIST) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
13. Background NAHB Lateral Tests NAHB Lateral Tests Current Roof-to-Wall Connection Full-Scale Testing Methods (NHAB) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
14. Previous Development of FRP Connection Up-Lift Component Test Set-Up Sample Component Specimen FRP Roof-to-wall Connection Development (Canbek, 2009) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
15. Previous Development of FRP Connection Carbon FRP (CFRP) Glass FRP (GFRP) FRP Test Specimens used in the Roof-to-wall Connection Development (Canbek, 2009) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
16. Previous Development of FRP Connection Full-Scale Specimen Fink Trusses GFRP Tie Connection Full-Scale Uplift Specimen (Canbek, 2009) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
17. Previous Development of FRP Connection FRP Results (Canbek, 2009) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
18. Previous Development of FRP Connection Results of FRP Tie Connection Development (Canbek, 2009) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
19. Previous Development of FRP Connection Cost Analysis for Configuration A with GFRP and CFRP (Canbek, 2009) NOTE: THE COST OF A CFRP TIE CONNECTION IS APPROXIMATELY 5 TIMES HIGHER THAN THAT OF A GFRP TIE CONNECTION, SO GFRP WAS SELECTED FOR FURTHER DEVELOPMENT (CANBEK, 2009). FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
20. Wall of Wind Testing -- Test Specimen WoW and Test Specimen, Ready for Testing at 90 Degrees FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
21. Floor Membrane Wall with Door & Window Wall of Wind Testing -- Test Specimen WoW Test Specimen Construction FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
22. Bottom Structure Assembly Bottom Structure Inside Wall Assembly Wall of Wind Testing -- Test Specimen WoW Test Specimen Construction FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
23. Roof Trusses Assembly of Roof Trusses Wall of Wind Testing -- Test Specimen WoW Test Specimen Construction FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
24. Mitered Top-Plate & GFRP Connection Typical 2 x 1.5 inch GFRP Wall of Wind Testing -- Test Specimen WoW Test Specimen Construction FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
25. Bottom Structure and Truss Roof Assembly Test Specimen Assembly Wall of Wind Testing -- Test Specimen WoW Test Specimen Construction FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
26. Eave Flashing Roof Flashing Wall of Wind Testing -- Test Specimen WoW Test Specimen Construction FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
27. Soffit Connection Notched Soffit Wall of Wind Testing -- Test Specimen WoW Test Specimen Construction FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
28. Wall of Wind Testing -- Instrumentation The test specimen was instrumented with the following sensors: 6 Load Cellsunder the trusses; sandwiched between double top plate of walls and single top plate of roof . The recorded forces are FX, FY, and FZ corresponding to the in-plane shear (parallel to the side walls), out-of-plane shear (perpendicular to the side walls), and uplift, respectively. 6 Linear Voltage Differential Transformers (LVDT)to measure horizontal displacements of GFRP truss connection (parallel to the side walls); placed on roof single top-plate. 12 String Potentiometers (String Pots) to measure vertical deflection and horizontal deflection (perpendicular to the side walls) of the connections; placed on roof single top-plate. 8 Strain Gaugesto measure strain in the vertical portion of each GFRP connection. 2 Compact-Rios were used for all data acquisition, installed on the inside of the walls of the test specimen and controlled using a common laptop through an Ethernet connection. FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
29. Wall of Wind Testing -- Instrumentation WoW Test Specimen Instrumentation and Connection Numbers FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
30. Wall of Wind Testing -- Instrumentation Typical Connection Instruments FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
31. Resurfaced Aluminum Plates Resurfaced Aluminum Plates ±0.01” Wall of Wind Testing -- Instrumentation WoW Test Specimen Instrumentation FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
32. Load Cells Alignment Mounted Load Cells & Aluminum Plates Wall of Wind Testing -- Instrumentation WoW Test Specimen Instrumentation FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
33. Load Cells & Aluminum Plates Mounted Load Cells & Aluminum Plates Wall of Wind Testing -- Instrumentation WoW Test Specimen Instrumentation FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
34. Typical Connection Instrumentation Wall of Wind Testing -- Instrumentation String Pots & Stain Gauge WoW Test Specimen Instrumentation FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
35. Pressure Transducers Between the Trusses Wall of Wind Testing -- Instrumentation Pressure Transducer on the Center of Test Specimen WoW Test Specimen Instrumentation FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
36. Wall of Wind Testing -- Instrumentation Pressure Transducers #4 Between Middle & Rear Trusses Pressure Transducer Manifold WoW Test Specimen Instrumentation FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
37. Wall of Wind Testing -- Instrumentation Compact Rios (Data Acquisition) WoW Test Specimen Instrumentation FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
38. Wall of Wind Testing -- Protocol 30 tests were performed 5 angles of attack (AOA) Enclosed and partially-enclosed building conditions Wind without rain condition, and wind-driven rain (WDR) condition Each test was performed using a 1 minute flat waveform (at maximum rpm of the WoW engines and generating high frequency turbulence only) 3 minutes quasi-periodic waveform (generating low frequency turbulence in addition to high frequency turbulence). WoW Testing Protocol for GFRP Roof-to-Wall Connections FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
39. 0º AOA, Enclosed Wall of Wind Testing -- Protocol 0º AOA, Enclosed & With Wind Driven Rain FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
40. 90º AOA, Partially Enclosed, One Window Removed 45º AOA, Partially Enclosed, Two Windows Removed Wall of Wind Testing -- Protocol FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
41.
42.
43.
44.
45.
46. Wall of Wind -- Testing AOA 0º, 4000 RPM, Enclosed & With Water FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
47. Wall of Wind -- Testing AOA 45º, Quasi-Periodic RPM, Enclosed & With Water FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
48. Full-Scale Aerodynamic Load Characteristics for Connections WoW Test Results for GFRP Roof-to-Wall Connections WoW test results are represented as: Graphs of the 3-second time averaged histories of individual load cells Bar graphs with mean results of all the conditions per load cell Scatter plots with mean force results of individual load cells Graphs of the 3-second time averaged histories of strain in connections Graphs of the 3-second time averaged histories for LVDTs (Displacements parallel to side walls in connections) Graphs of the 3-second time averaged histories for String Pots (Displacements perpendicular to side walls in connections) Graphs of the 3-second time averaged histories for String Pots (Vertical displacements in connections) Graphs of the 3-second time averaged histories of Internal Pressures FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
49. Full-Scale Aerodynamic Load Characteristics for Connections WoW Test Results for GFRP Roof-to-Wall Connections The nomenclature used to determine the type of test runs in the graphs is as follows: E_FT: enclosed condition and wind speeds at full-throttle E(W)_FT: enclosed condition with wind driven rain and at full-throttle PE_FT: partially enclosed condition where 1 (for AOA 90º test) or 2 (for AOA 45º test) windows have been removed and 1 window and the door (for AOA 0º test) and at full-throttle PE’_FT: partially enclosed condition, where the windows and the test specimen door have been removed at full-throttle (for AOA 45º & 60º tests) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
50. Full-Scale Aerodynamic Load Characteristics for Connections WoW Test Results for GFRP Roof-to-Wall Connections The nomenclature used to determine the type of test runs in the graphs is as follows: E_QP : enclosed condition and a quasi-periodic ramp function E (W)_QP : enclosed condition with wind driven rain and quasi-periodic ramp function PE_QP: corresponds to a partially enclosed condition where 1 (for AOA 0º & 90º tests) or 2 (for AOA 45º test) windows have been removed and a quasi-periodic ramp function PE’_QP: corresponds to a partially enclosed condition where 2 (for AOA 45º & 60º tests) windows and the test specimen door have been removed and a quasi-periodic ramp function FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
51. WoW Full-Scale Test Results Time Histories of Fz -- LC #1 -- AOA 0º FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
52. WoW Full-Scale Test Results Time Histories of Fz -- LC #1 0º 30º 30º 90º 60º 45º
53. WoW Full-Scale Test Results Time Histories of Fz -- LC #2 0º 30º 90º 60º 45º
54. WoW Full-Scale Test Results Time Histories of Fz -- LC #3 0º 30º 90º 60º 45º
55. WoW Full-Scale Test Results Time Histories of Fz -- LC #4 0º 30º 90º 45º 60º
56. WoW Full-Scale Test Results Time Histories of Fz -- LC #5 0º 30º 90º 45º 60º
57. WoW Full-Scale Test Results Time Histories of Fz -- LC #6 0º 30º 90º 60º 45º
58. WoW Full-Scale Test Results LC #1 -- Bar Graph of Mean Fx at 4400 RPM for all Conditions & AOAs
59. BAR GRAPHS ALL CONDITIONS & AOAs: Mean FX (4000 RPM) LC 1 LC 4 LC 2 LC 5 LC 3 LC 6
60. BAR GRAPHS ALL CONDITIONS & AOAs: Mean Fy (4000 RPM) LC 1 LC 4 LC 2 LC 5 LC 3 LC 6
61. BAR GRAPHS ALL CONDITIONS & AOAs: Mean Fz (4000 RPM) LC 1 LC 4 LC 2 LC 5 LC 3 LC 6
62. WoW Full-Scale Test Results LC #5 – Scatter Plot of Mean Fx, Fy & Fz at 4000 RPM (Enclosed and all AOAs)
67. WoW Full-Scale Test Results LC #5 - Time Histories of Strain -- AOA 60º, All Conditions FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
68. WoW Full-Scale Test Results Time Histories of Strain at All AOAs -- LC #5 0º 30º 90º 60º 45º
69. WoW Full-Scale Test Results LC #5 - Time Histories of Displacements Parallel to Side Walls (LVDT measurements) – AOA 0º, All Conditions FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
70. WoW Full-Scale Test Results Time Histories of Displacements Parallel to Side Walls (LVDT measurements) -- LC #5 0º 30º 90º 45º 60º
71. WoW Full-Scale Test Results LC #1 - Time Histories of Displacements Perpendicular to Side Walls (String Pots measurements) – AOA 0º, All Conditions FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
72. WoW Full-Scale Test Results Time Histories of Displacements Perpendicular to Side Walls (String Pots measurements) -- LC #1 0º 30º 90º 60º 45º
73. WoW Full-Scale Test Results LC #1 - Time History of Vertical Displacements (String Pots measurements) – AOA 0º, All Conditions FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
74. WoW Full-Scale Test Results Time History of Vertical Displacements (String Pots measurements) -- LC #1 0º 30º 90º 60º 45º
75. Full-Scale Aerodynamic Load Characteristics for Connections Effect of Internal Pressure due to Breach of Envelope Time History of Highest Recorded Loads in Connection #5; Partially Enclosed, 0º AOA Partially Enclosed, 0º AOA FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
76. WoW Full-Scale Test Results LC #5 – Load Difference Between an Enclosed and Partially-Enclosed Conditions – AOA 0º Connection #5 Up-Lift Time Histories for Enclosed & Partially Enclosed Conditions Connection #5 Mean Up-Lift Forces for Enclosed & Partially Enclosed Conditions The maximum difference, related to the uplift loading between enclosed and partially-enclosed conditions was recorded to be 528 lbs FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
77. WoW Full-Scale Test Results Time Histories and Mean Internal Pressure (Transducers #3, 4 & 5) -- Near Connection 5 Transducer#3 Transducers# 3, 4 & 5 (E & PE at FT) Transducer#4 Mean Internal Pressures: Transducers #3, 4, & 5 (E&PE@FT) Transducer#5
78. Discussion on Aerodynamic Characteristics Effect of Internal Pressure Increase on Uplift Loading Connection #5, 117 lbs and 645 lbs for E and PE conditions, respectively -- the difference being 528 lbs Mean internal pressures difference was 0.08 psi The tributary area 35 sq. ft. - load of about 400 lbs The measured uplift increment on the connection is higher than the estimated value (both showed significant increase) The difference could be due to approximation of tributary area and spatial correlation of internal pressure The experiments indicate how severe can be the effects of breach of building envelope (5.5 times of uplift load increase) FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
79. Discussion on Aerodynamic Characteristics Effect of Turbulence (Low Frequency Vs High Frequency) The proportionalities between the mean uplift, in-plane, and out-of-plane forces are very similar for flat and quasi-periodic waveforms Higher turbulence (TI 24% vs 5%) generated by the low frequency fluctuations of the wind does not affect the proportionalities between the mean uplift and lateral forces induced on the connections Thus rigorous generation of turbulence intensity and integral length scale (as always attempted in wind tunnels) may not be necessary for tri-axial loading evaluation for roof-to-wall connections For further testing of the GFRP connections to failure under tri-axial loading in SCL, only the data obtained for the flat waveform tests are used FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
80. Discussion on Aerodynamic Characteristics Effect of Wind Driven Rain Hurricane winds are accompanied by wind-driven rain Generally wind tunnels cannot be used for comprehensive research into this phenomenon WoW was used to determine if there is any significant difference between aerodynamic and aero-hydrodynamic loading induced on the GFRP connections No significant increase in load was observed during the wind-driven rain tests as compared to wind with no rain The data used for failure testing in SCL were obtained from the wind tests FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
81. Discussion on Aerodynamic Characteristics Database for Tri-Axial Loading on Connections Load cells measured uplift, in-plane (parallel to the side walls), and out-of-plane (perpendicular to the side walls) loads experienced by each GFRP connection The results were used for tri-axial loading of the GFRP connection in SCL till failure at the component level For each test the three force components were converted to a resultant mean load in order to test the GFRP connections more realistically using aerodynamic loading obtained from WoW tests A total of 36 resultant forces were obtained from the loads recorded at the WoW and were used to test the newly developed GFRP connections and metal hurricane clips FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
82.
83. Under real storms a fastener will experience simultaneous uplift, in-plane, and out-of-plane loading which will have specific ratios based on several factors (e.g., location of the connection, type of the roof, etc)
84. The common practice of Uni-Axial testing can lead to incorrect specifications of the allowable capacity of a mechanical fastener
85. To circumvent the above limitations the new testing approach is based on simultaneous aerodynamic tri-axes loads with proportionalities obtained from realistic full-scale WoW testingFLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
86.
87. New tri-axial test protocol was established and connections tested to failure at the SCL using ratios of uplift to in-plane lateral and out-of-plane lateral loads
88. Each test in SCL represented a particular tri-axial aerodynamic loading obtained at the WoW for specific parameters: connection location, angle of attack, and internal pressure condition (enclosed or partially enclosed condition) -- results were compared with those from testing using individual loading
89. Series of resultant mean forces were used to test the GFRP component connections in the SCL up to failure -- 23 of the 36 resultant forces were used due to limitations in the SCL system
90. Hurricane clips were tested to provide a comparison of performance between GFRP and metal connections subjected to simultaneous tri-axial loading FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
91.
92. Tri-Axial Component Testing --Test Set-Up FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
93. Tri-Axial Component Testing --Test Set-Up Locations FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
94. Tri-Axial Component Testing --Results AOA 0º FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
95. Tri-Axial Component Testing –Results AOA 30º FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
96. Tri-Axial Component Testing --Results AOA 45º FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
97. Tri-Axial Component Testing --Results AOA 60º FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
98. Tri-Axial Component Testing --Results AOA 90º FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
99. Tri-Axial Component Testing – GFRP Tests FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
100. Tri-Axial Component Testing – Clip Tests FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
102. Tri-Axial Component Specimens FailureModes Case 6 FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
103. Tri-Axial Component Specimens Failure Modes Case 11 FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
104. Tri-Axial Component Specimens Failure Modes Case 15 FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
105. Tri-Axial Component Specimens Failure Modes Case 23 FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
106. Conclusions Feasibility of GFRP Connections as Substitute to Metal Connections The failure load capacity of the GFRP connection performed similar to and in most cases better than the metal fasteners test results (under tri-directional simultaneous loads obtained from aerodynamic tests at the WoW) In some cases the ultimate failure resultant load for the GFRP connection was observed to be double of that for the metal fastener The GFRP connection test results seem to demonstrate that it can be applicable to new construction as well as retrofitting of old residential buildings that require strengthening against extreme wind loads with minimally intrusive techniques. FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
107. Conclusions Failure Modes of GFRP Connections Vs Metal Connections The results show that the failure modes of connection joints are highly dependent on the type of the connection (GFRP versus metal) It was noted that as GFRP is non-intrusive it doesn’t weaken the wood members and crushing of wood is avoided (Ahmed et al., 2009). The failure mode observed was mostly detachment of GFRP from the wood surface and wood surface peeling In the case of the hurricane clip the failure mode was observed as both nail withdraw or pull-out, clip rupture FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
108. Conclusions Differences in Test Results -- Uni-Axial Vs Tri-Axial Testing The failure loads for both connectors (GFRP and metal) decreased during the tri-axial test When the coefficients FX/FZ and FY/FZ for the tri-axial testing were low the uplift capacity matched the uni-axial testing uplift capacity closely However when the coefficients were high, reduced uplift capacity was observed compared to the uni-axial testing uplift capacity This indicated that the lateral load components, if applied simultaneously with the uplift load component as experienced during real storms, the uplift load capacity of the connection is reduced – so that the uni-axial uplift test results are overestimated FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
109. Conclusions In one of the most extreme the uplift failure recorded loads were 295 (884/3FS) pounds and 250 (750/3FS) pounds as compared to the 720 pounds and 437 pounds , obtained from uni-axial testing, for the GFRP and metal clip, respectively The lateral (parallel to the walls) failure loads were 102 (307/3FS) pounds and 87 (260/3FS) pounds as compared to the 552 pounds and 165 pounds obtained from uni-axial testing for the GFRP and metal clip, respectively Results indicate the inappropriateness of the existing uni-axial testing protocol used to test connectors Design based on these erroneous allowable load capacities can cause inter-component connection failures during high wind events Improving upon current practice by taking into account the results (WoW database) reported herein and the suggested tri-axial testing will improve the performance of timber construction in high winds FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
110. Conclusions Project Contributions Based on WoW testing a database (may be used by other researchers and industry professionals) has been developed on aerodynamic and aero-hydrodynamic loading on roof-to-wall connections tested under several parameters: angles of attack, wind-turbulence content, enclosed and partially-enclosed building conditions, with and without effects of rain The research’s findings demonstrated that a GFRP connection system is a viable option for use in a timber roof-to-wall connection system A component level testing protocol and setup have been developed in SCL to test connections to failure under the influence of simultaneous tri-axial loading; such protocol eliminates the erroneous load capacity predictions from existing uni-axial test protocols A database has been developed on the uni-axial and tri-axial load capacity of the GFRP connections and of a particular type of metal fastener; simultaneous application of tri-axial loading to roof-to-wall connections in SCL is one of the first attempts to mimic realistic aerodynamic loads obtained from full-scale wind testing FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
111. Future Work Recommendations The GFRP connection has yet to be tested aerodynamically under Category 4 and 5 hurricane conditions; this can be simulated in the 12-fan WoW system which is presently under development A series of timber structures, within tropical cyclone prone coastal areas, could be retrofitted with the new GFRP roof-to-wall connections. Building performance under possible future storms can then provide validation of the connections -- there is no better test method than subjecting the connections to actual tropical cyclone conditions Such retrofitting can also be used to study the long term weather effects of moisture, heat and rain on the GFRP connections, which are not yet completely understood FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
112. Future Work Recommendations Studies on creep and fatigue are warranted for the GFRP connections The tri-axial testing method and system used in this research could be improved to test all 36 resultant aerodynamic forces obtained from the WoW tests; this could be done by enhancing the current testing system by enlarging the size of the system and implementing a pulley swivel system that can allow more lateral locations to be tested FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
113. END OF PRESENTATIONTHANK YOU FOR YOUR TIMEANY QUESTIONS? FLORIDA INTERNATIONAL UNIVERSITY - CIVIL & ENVIRONMENTAL ENGINEERING
Hinweis der Redaktion
Building structure integrity is compromised when the inter-component connections, such as roof-to-wall connections, wall-to-wall connections, wall-to-floor connections, or anchorage-to-foundations fail. Continuous load path and structural integrity are crucial for windstorm resiliency of residential buildings. Tropical cyclone damage has shown that wood structures tend to suffer little damage when the roof system remains intact under extreme wind loading, while major damage occurs when the roof system is partially or completely damaged (Reed et al. 1997). Post tropical cyclone inspections have noted the entire roofs detaching from buildings in some cases (Figure 2.5). This indicates a serious deficiency in the roof-to-wall connection systems, most notable in older construction. Thus the roof-to-wall connections play an important role to prevent roof failures and lessen the damages during high winds.
As wind impacts a building, it creates vortices that can overwhelm the structure on different locations (see Figure 2.1). As wind travels around sharp edges, such as wall corners, roof overhangs and roof ridgelines, a separation bubble is formed. The wind separation bubble is bounded by a free shear layer region of high velocity gradients and high turbulence (Holmes 2001). Conical vortices are formed (see Figure 2.2); as these are shed down wind, high negative pressure peaks are produced which generate suction or uplift loads on roofs up-lift loads on a roof are transferred through the roof elements (e.g., tiles, shingles) to the plywood sheathing to the roof trusses; which could lead to roof detachment if the inter-component connections are improperly designed or installed. Figure 2.3 illustrates the distribution of forces along a timber building during a high wind event. In a properly designed building the roof loads are transferred through a continuous vertical load path to the foundation.
The Florida Building Code (FBC, 2007) has special provisions for buildings in the High Velocity Hurricane Zone (HVHZ) which consists of Miami-Dade and Broward counties in the state of Florida. Due to the strict design and construction practices used in the HVHZ, all metal connectors must be approved and rated by the code compliance authorities for use in buildings. The Notice of Approval (NOA) lets structural designers know the capacity of a specific product to be used in their design (Figure 2.6). FBC-07: 2321.7.2 requires all wood to wood straps to resist a minimum uplift force of 700 pounds with 4-16d nails in each member (FBC, 2007). The nails used may change if the NOA allows it.
Compared to other configurations, Configuration A yielded more favorable results. The failure loads obtained with CFRP were 20% higher (Canbek, 2009). Nevertheless, the price of a CFRP tie connection is approximately 5 times higher than that of a GFRP tie connection. Therefore, Configuration A with GFRP was selected as the best alternative for further development (Canbek, 2009
23 GFRP and 23 metal connector specimens were testedEach specimen was built using SPF No. 2 – 2 x 6 inch lumber with two separate connection systems, GFRP or metal connectors The test system was composed of a double acting 10,000 lbs hydraulic jack that could pull on the component specimen using a cable and pulley A load cell between the specimen and pulley recorded the ultimate failure load, via a DAQ computer Each specimen was bolted to an I-beam that in turn was attached to two channels bolted to the SCL tie-downs By moving the specimen North-South and East-West the resultant loading could be simulated