Citation
Pendulum Impact Testing of Metallic, Non-Metallic, and Hybrid Sign Posts

Material Information

Title:
Pendulum Impact Testing of Metallic, Non-Metallic, and Hybrid Sign Posts
Creator:
Groetaers, Matias
Publisher:
University of Florida
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( M.E.)
Degree Grantor:
University of Florida
Degree Disciplines:
Civil Engineering
Civil and Coastal Engineering
Committee Chair:
CONSOLAZIO,GARY R
Committee Co-Chair:
HAMILTON,HOMER ROBERT,III
Committee Members:
RICE,JENNIFER ANNE
FERRARO,CHRISTOPHER CHARLES
Graduation Date:
5/3/2014

Subjects

Subjects / Keywords:
Acceleration ( jstor )
Aluminum ( jstor )
Boundary layer separation ( jstor )
Deformation ( jstor )
High speed cameras ( jstor )
Honeycombs ( jstor )
Modeling ( jstor )
Posts ( jstor )
Simulations ( jstor )
Statics ( jstor )
hybrid
Genre:
Unknown ( sobekcm )

Notes

General Note:
The primary objective of this study was to develop a new breakaway connection-for use in large roadside sign structures-with improved performance characteristics relative to existing breakaway connection designs. By using a hybrid combination of steel and fiber reinforced polymer (FRP) structural components, a new system was designed to resist statically equivalent hurricane wind loading, but also breakaway under direct vehicle impact loading. By replacing the heavier steel sign posts that have been used in previous designs with lighter FRP posts, and simultaneously introducing the use of innovative steel cutting surfaces that are engaged only during direct vehicle impact loading, occupant risks associated with collisions were reduced in severity, thereby producing improving safety. Since the performance standard for the design of roadside breakaway connections is currently the AASHTO Manual for Assessing Safety Hardware (MASH), the new connection system developed in this study was designed to meet the MASH requirements. However, since MASH is a relatively new standard, existing (i.e., previously-developed) designs of surrogate vehicles-used in impact testing of breakaway hardware-that meet the current MASH requirements were not available in the literature. Therefore, to facilitate physical testing of newly developed prototype connections, a MASH-compliant surrogate vehicle first had to be designed, fabricated, and validated. Hence, a secondary objective of this study was to develop a new 'small car' (1100 kg) surrogate vehicle for use in conjunction with the FDOT pendulum impact testing facility. In the course of pursuing this objective, several innovative and improved methods for designing surrogate impact vehicles were developed in this study.. A key improvement was using tapered (trapezoidal) aluminum honeycomb cartridges, rather than the traditionally employed rectangular (constant cross-section) cartridges, to the reproduce the force-deformation behavior of the target production (market) vehicle (a 2006 Kia Rio in this study).. Additionally, the behavior of aluminum honeycomb at high levels of compressive stress (and deformation) was investigated using experimental testing and the findings were integrated into the surrogate vehicle design process. Validation crash-testing of the final surrogate vehicle (i.e., a combination of an impact block and a crushable nose constructed from aluminum honeycomb cartridges) demonstrated an acceptable match to the behavior of the target vehicle. Subsequently, the newly developed MASH-compliant surrogate vehicle was used to experimentally evaluate, using pendulum impact testing, multiple prototypes of a new hybrid (FRP and steel) breakaway connection design. The final design was demonstrated to meet the impact risk requirements stipulated in MASH as well as the hurricane wind resistance requirements specified by AASHTO and FDOT.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Groetaers, Matias. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/31/2016

Downloads

This item has the following downloads:


Full Text

PAGE 1

PENDULUM IMPACT TESTING OF METALLIC, NON METALLIC, AND HYBRID SIGN POSTS By MATIAS GROETAERS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2014

PAGE 2

2014 Matas Groetaers

PAGE 3

To my father and mothe r

PAGE 4

4 ACKNOWLEDGMENTS I would like to express sincere gratitude to my advisor D r. Gary Consolazio for ex tending the opportunity to work under him and develop professional ly in the field of s tructural e ngineering. Furthermore, his constant guidance throughout this research, most importantly his instruct ion with regards to decision making and judgment allowed me to successfully complete this thesis In addition, I would like to thank my graduate research assistant mentor Dr. Daniel Getter for his valuable training in technical and non technical areas of research Gratitude should also be expressed towards my c ommittee members, Dr. Trey Hamilton III Dr. Jennifer Ri c e and Dr. Chris topher Ferraro for the ir feed back on this research and s ervice o n my committee. A special thanks to my family, specifically mother and father, for their unconditional support and enorm ous efforts assisting me in the pursuit of each goal that has been presented in my life. Finally, many thanks to all of my colleagues and friends who in one way or another

PAGE 5

5 TABLE OF CON TENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 8 LIST OF FIGURES ................................ ................................ ................................ ......................... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 14 CHAPTE R 1 BACKGROUND ................................ ................................ ................................ .................... 16 2 DESIGN BASIS ................................ ................................ ................................ ..................... 21 2.1 Introduction ................................ ................................ ................................ ....................... 21 2.2 Selection Of Sign Structure ................................ ................................ .............................. 21 2.3 Wind Loading ................................ ................................ ................................ ................... 23 2.4 Impact Loading ................................ ................................ ................................ ................. 24 3 VEHICLE IMPACT TESTING ................................ ................................ ............................. 28 3.1 Introduction ................................ ................................ ................................ ....................... 28 3.2 Market Vehicle Impact Testing Of Break away Devices ................................ .................. 28 3.3 Surrogate Vehicle Impact Testing Of Breakaway Devices ................................ .............. 30 4 CRUSHABLE NOSE: MATERIALS AND DESIGN CONCEPTS ................................ ..... 34 4.1 Introduction ................................ ................................ ................................ ....................... 34 4.2 Elastic Plastic Aluminum Honeycomb Behavior Used In Surrogate Vehicles ................ 34 4.3 Existing Surrogate Vehicles ................................ ................................ ............................. 39 4.4 Development Of 1100 kg UF/FDOT Surrogate Vehicle ................................ .................. 40 4.4 .1 Tapered Sections ................................ ................................ ................................ .... 40 4.4.2 Aluminum Honeycomb Behavior Beyond The Fully Buckled Stage .................... 41 4.4.3 Force deformation Behavior Of T apered Cartridges Arranged In Series .............. 48 5 CRUSHABLE NOSE: DESIGN BY WAY OF NUMERICAL SIMULATION .................. 50 5.1 Introduction ................................ ................................ ................................ ....................... 50 5.2 High Resolution Modeling Of Front Block And Tubes ................................ ................... 51 5.3 High Resolution FEA Modeling And Analysis Of Aluminum Honeycomb Cartridges ................................ ................................ ................................ ............................ 54 5.4 High Resolution Nonlinear Spring Crushable Nose Model ................................ ............. 57 5.5 Simplified 8 DOF Crushable Nose Model ................................ ................................ ....... 60

PAGE 6

6 5.6 Final Crushable Nose Design ................................ ................................ ........................... 62 5.7 Impact F D Curves: Numerical Simulation Vs. MASH Compliant Car (Kia) ................. 64 6 CRUSHABLE NOSE: EXPERIMENTAL VALIDATION ................................ .................. 66 6.1 Introduction to Pendulum Impact Testing ................................ ................................ ........ 66 6.2 Testi ................................ ................................ ..... 67 6.3 Instrumentation ................................ ................................ ................................ ................. 68 6.3.1 High Speed Cameras ................................ ................................ .............................. 68 6.3.2 Break Beams ................................ ................................ ................................ ........... 70 6.3.3 Tape Switch ................................ ................................ ................................ ............ 71 6.3.4 Accelerometers ................................ ................................ ................................ ....... 71 6.4 Impact Test Results ................................ ................................ ................................ ... 72 6.5 Comparison Of Numerical Predictions And Experimental Results ................................ 75 7 H YBRID BREAKAWAY CONNECTION DESIGN ................................ ........................... 77 7.1 Introduction ................................ ................................ ................................ ....................... 77 7.1.1 FRP Pultruded Section ................................ ................................ ........................... 77 7.1.2 Section Design Requirements ................................ ................................ ................. 78 7.1.3 Section Properties ................................ ................................ ................................ ... 78 7.2 Static Flexural Testing ................................ ................................ ................................ ...... 79 7.2.1 Test Setup ................................ ................................ ................................ ............... 79 7.2.2 Results ................................ ................................ ................................ .................... 80 7.3 Section Fundamental Frequency ................................ ................................ ....................... 83 7.4 Hybrid Breakaway Connection System Concept ................................ ............................. 83 7.5 Impact Shear Test ................................ ................................ ................................ ............. 87 7.5.1 Test Setup ................................ ................................ ................................ ............... 87 7.5.2 Results ................................ ................................ ................................ .................... 88 8 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS ................................ ......... 96 8.1 Introduction ................................ ................................ ................................ ....................... 96 8.2 UF/FDOT Crushable Nose ................................ ................................ ............................... 96 8.3 Hybrid Breakaway Connection ................................ ................................ ......................... 97 APPENDIX A FABRICATION DRAWINGS FOR IMPACT BLOCK ................................ ....................... 98 B FABRICATION DRAWINGS FOR CRUSHABLE NOSE ................................ ................ 109 C CALCULATION OF POST STRENGTH REQUIRED FOR WIND LOADING .............. 115 D CALCULATION OF POST FLEXURAL STRENGTH FROM TEST DATA .................. 122 E FABRICATION DRAWINGS FOR HYBRID BREAKAWAY CONNECTION .............. 125

PAGE 7

7 F FRP POST STATIC SHEAR TESTING ................................ ................................ .............. 131 G MANUFACTURER SP ECIFIED PROPERTIES OF FRP POST ................................ ....... 139 H GEOMETRIC DIMENSIONS OF FRP POST ................................ ................................ .... 141 I CALCULATION OF FRP POST FUNDAMENTAL FREQUENCY ................................ 143 LIST OF REFERENCES ................................ ................................ ................................ ............. 145 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 146

PAGE 8

8 LIST OF TABLES Table page 2 1 project (Pinelli and Subramanian 1999). ................................ ................................ ........... 22 5 1 S ummary of crushable nose components ................................ ................................ ........... 63 7 1 Sign post strength design requirements. ................................ ................................ ............ 78 7 2 FRP octagonal section properties as specifie d by Creative Pultrusions. ........................... 79 7 3 FRP octagonal section properties by experimental testing. ................................ ............... 83 7 4 FRP octagonal section experim ental impact shear test results. ................................ ......... 95

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Loading conditions for large sign post slip base breakaway connection. .......................... 17 1 2 Breakaway behavior of sign supports. ................................ ................................ ............... 18 1 3 Failure of sign structure with slip base connection under hurricane wind loading. .......... 19 1 4 Shear controlled moment collar breakaway connection developed previously in study BDK75 977 40. ................................ ................................ ................................ ........ 19 2 1 Panel dimensions and clearance height of selected sign. ................................ ................... 22 3 1 Low speed (per NCHRP 350) impact. A) Test of 1979 VW Rabbit B) C orrespondi ng force deformation behavior. ................................ ................................ 29 3 2 Low speed (per MASH 2009) impact. A) Test of 2006 Kia Rio. B) Corresponding force deformation behavior. ................................ ................................ ............................... 30 3 3 Portion of 2006 Kia Rio impact test force deformation curve needed to develop a low speed, MASH compliant surrogate vehicle. ................................ ............................... 31 3 4 Surrogate vehicles characteristics. A) Photo of 820 kg MwRSF crushable nose. B) St epwise linear approximation. ................................ ................................ ..................... 32 4 1 Aluminum honeycomb cartridges. A) Rectangular. B) Rectangular with punch out. C) Hexagonal honeycomb cell structure. ................................ ................................ ........... 35 4 2 Aluminum honeycomb cartridge installed in Tinius Olsen test machine to conduct a static compression test. ................................ ................................ ................................ ...... 36 4 3 Hypothetical force deformation curve for compression of a rectangular aluminum honeycomb cartridge. ................................ ................................ ................................ ......... 36 4 4 Stages of aluminum honeycomb cartridge behavior. A) Original. B) Upset. C) Progression of buckling. D) Fully buckled. ................................ ................................ .. 37 4 5 Measured force deformation curve for compression of a rectangular aluminum honeycomb cartridge ................................ ................................ ................................ .......... 37 4 6 Schematic diagram of impact block and attached crushable nose ................................ ..... 38 4 7 Existing 820 kg surrogate vehicle. A) Stepwise theoretical design curve and vehicle curve. B) Crushable nose attached to impact block. ................................ .......................... 39 4 8 Tapered (trapezoidal) aluminum honeycomb cartridge with linearly varying cross sectional area. ................................ ................................ ................................ ..................... 40

PAGE 10

10 4 9 Comparison of constant cross section and tapered cross section aluminum honeycomb cartridges. A) Constant rectangular cross section. B) Tapered trapezoidal cross section. C) Rectangular cartridge approximation. D) Trapezoidal cartridge approximation. ................................ ................................ ................................ ................... 41 4 10 Stag es of aluminum honeycomb deformation. A) Mechanical upset. B) Buckling of cell walls. C) Fully buckled stage. D) Group buckling stage. E) Fully compressed stage. ................................ ................................ ................................ ................................ .. 43 4 11 High stress compressio n schematics. A) Rectangular geometry. B) Trapezoidal geometry. ................................ ................................ ................................ ........................... 44 4 12 Deformation stages and geometric dimensions used in computing force simplified force deformation curves. A) Initial buck ling. B) Fully buckled stage. C) Group buckling. D) Fully compacted stage. ................................ ................................ ................. 47 4 13 Comparison of computed and measured force deformation curves. A) Rectangular cartridge. B) Trapezoidal cartri dge. ................................ ................................ ................... 47 4 14 Simultaneous compression of tapered cartridges in series. A) Cartridge 1 behavior. B) Cartridge 2 behavior. C) Behavior of cartridges in series compression. ...................... 49 5 1 Kia Rio force deformation curves. ................................ ................................ .................... 50 5 2 Preliminary finite element model used to assess structural demands in front block and telescoping guid e tubes. ................................ ................................ .............................. 52 5 3 Finite element model of UF/FDOT surrogate vehicle front block. A) Mesh. B C) Maximum Von Mises stresses during impact simulation ................................ ......... 53 5 4 Finite element model of UF/FDOT surrogate vehicle guide tubes. A) Mesh; B C) Worst case Von Mises stresses during impact. D E) Next to worst case Von Mises stresses during impact ................................ ................................ ............................. 54 5 5 Finite element model and experimental test of trapezoidal cartridge. A B) Un deformed. C D) Intermediate stages of compression. E) Fully buckled state. ... 55 5 6 Sharp tipped pe ntagonal aluminum honeycomb cartridge. A) Un deformed. B) Experimental test specimen. C) Initial condition of compression simulation. D) Partially crushed condition during compression simulation. ................................ ........ 56 5 7 UF/FDOT surrogate vehicle finite element model with grids of nonlinear springs used to model each aluminum honeycomb cartridge. A) Overall view. B) Elevation (side) view. C) Plan (top) view. ................................ ................................ ......................... 57 5 8 Behavior of sample nonlinear springs grids. A) Side view of honeycomb modeling simplification. B) Illustration of material curves. ................................ .............................. 58

PAGE 11

11 5 10 FEA model of UF/FDOT crushable nose using in. thick Garolite G 11 spacers. ......... 59 5 11 UF/FDOT surrogate vehicle high resolution FEA model results. ................................ ..... 60 5 12 UF/FDOT su rrogate vehicle simplified FEA model results. ................................ ............. 61 5 13 UF/FDOT surrogate vehicle simplified FEA model results. ................................ ............. 61 5 14 UF/FDOT cr ushable nose and impact block. A) Side view. B) Top view. C) Isometric view. ................................ ................................ ................................ .............. 62 5 15 UF/FDOT surrogate vehicle high resolution and simplified FEA model results. ............. 65 5 16 UF/FDOT surrogate vehicle FEA model results and Kia Rio test data curves. ................ 65 6 1 UF/FDOT crushable nose and impact block. A) Isometric view. B) Crushab le nose side view. ................................ ................................ ................................ ........................... 66 6 2 Pendulum structure. A) Schematic of pendulum swing path. B) Pendulum release photo. ................................ ................................ ................................ ................................ 66 6 3 Impactor bl ock and crushable nose structures. A) Isometric schematic of surrogate vehicle. B) Crushable nose post compression. ................................ ................................ .. 67 6 4 Crushable nose validation test. ................................ ................................ .......................... 67 6 5 Camera 1 and 2 wide test view. ................................ ................................ ....................... 68 6 6 Camera 2 and 3 detailed view of crushable nose area from both sides. .......................... 68 6 7 Top view of test set up for cameras layout. ................................ ................................ ....... 69 6 8 High speed camera frames of tracking points procedure using ProAnalyst software. ...... 70 6 9 Position of infrared optical break beam sensor. ................................ ................................ 70 6 10 Position of tape switch installed on rigid pole. ................................ ................................ .. 71 6 11 Locations of accelerometers installed on the surrogate vehicle. ................................ ........ 72 6 12 High speed camera video frames of validation test. ................................ .......................... 72 6 13 Filtered and averaged acceleration data from back block accelerometers. ........................ 73 6 14 Filtered acceleration data from front block accelerometer. ................................ ............... 73 6 15 Force contribution to the total system by the front and back block structures. ................. 74 6 16 Back block displacement data from high speed cameras. ................................ .................. 75

PAGE 12

12 6 17 UF/FDOT surrogate vehicle impact test results. ................................ ................................ 75 6 18 Comparison between test results and Kia Rio force deformation curves. ......................... 76 6 19 Comparison between experimental test and models results force deformation curves. .... 76 7 1 Specimen tested. A) Two 3ft long an d three 8ft long sections donat ed. B) Cross sectional detail. ................................ ................................ ................................ .................. 78 7 2 Static three point flexural test schematic. ................................ ................................ .......... 80 7 3 Static t hree point flexural test setup. ................................ ................................ ................. 80 7 4 Section deformation during static flexural test. ................................ ................................ 82 7 5 Enlarged view of FRP section at f ailure. ................................ ................................ ........... 82 7 6 Static flexural test results. ................................ ................................ ................................ .. 82 7 7 Hybrid breakaway system exploded view. ................................ ................................ ........ 84 7 8 Hybrid breakaway system concept ................................ ................................ .................... 85 7 9 Hybrid breakaway system in field. ................................ ................................ .................... 86 7 10 Impact shear test setup schematic. ................................ ................................ ..................... 87 7 11 Impact shear test setup. ................................ ................................ ................................ ...... 88 7 12 High speed camera video frames of breakaway impact shear test. ................................ .... 88 7 13 Specimen post impact shear test. ................................ ................................ ....................... 89 7 14 Back block smooth displacement time plot from high speed cameras. ............................. 90 7 15 Back block smooth acceleration time plot. ................................ ................................ ........ 90 7 16 Front block smooth displacement time plot from high speed cameras. ............................. 91 7 17 Front block smooth acceleration time plot. ................................ ................................ ........ 91 7 18 Force contribution to the total system by the front and back block structures. ................. 91 7 19 Crushable nose deformation time plot from high speed cameras. ................................ ..... 92 7 20 Forces generated by the surrogate vehicle on the breakaway impact shear test. ............... 92 7 21 Comparison between crushable nose impact shear test and Kia Rio dynamic behavior. ................................ ................................ ................................ ............................. 92

PAGE 13

13 7 22 Back block c.g. acceleration time history using 10 (ms) moving average technique. ....... 94 7 23 Back block c.g. relative velocity time history from trapezoidal integration of acceleration time data. ................................ ................................ ................................ ....... 94 7 24 Back block c.g. relative displacement time history from double trapezoidal integration of acceleration time data. ................................ ................................ ................ 94 F. 1 Schemati c of static double shear test setup ................................ ................................ ...... 132 F. 2 Static double shear test setup ................................ ................................ ........................... 133 F. 3 Section deformation during static double shea r test ................................ ........................ 133 F. 4 S tatic double shear test results ................................ ................................ ......................... 134

PAGE 14

14 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partia l Fulfillment of the Requirements for the Degree of Master of Engineering PENDULUM IMPACT TESTING OF METALLIC, NON METALLIC, AND HYBRID SIGN POSTS By Matas Groetaers May 2014 Chair: Gary R. Consolazio Cochair: Homer R. Hamilton III Major: Civil Engi neering The primary objective of this study was to develop a n ew breakaway connection for use in large roadside sign structures with improved performance characteristics relative to existing breakaway connection designs. By using a hybrid combination of s teel and fiber reinforced polymer (FRP) structural components, a new system was designed to resist statically equivalent hurricane wind loading, but also breakaway under direct vehicle impact loading. By replacing the heavier steel sign posts that have bee n used in previous designs with lighter FRP posts, and simultaneously introducing the use of innovative steel cutting surfaces that are engaged only during direct vehicle impact loading, occupant risks associated with collisions were reduced in severity, t hereby producing improving safety. Since the performance standard for the design of roadside breakaway connections is currently the AASHTO Manual for Assessing Safety Hardware (MASH), the new connection system developed in this study was designed to meet t he MASH requirements. However, since MASH is a relatively new standard, existing (i.e., previously developed) designs of surrogate vehicles used in impact testing of breakaway hardware that meet the current MASH requirements were not available in the liter ature. Therefore, to facilitate physical testing of newly

PAGE 15

15 developed prototype connections, a MASH compliant surrogate vehicle first had to be designed, 1100 kg) surrogate vehicle for use in conjunction with the FDOT pendulum impact testing facility. In the course of pursuing this objective, several innovative and improved methods for designing surrogate impact vehicles were developed in this study. A key improvement was using tapered (trapezoidal) aluminum honeycomb cartridges, rather than the traditionally employed rectangular (constant cross section) cartridges, to the reproduce the force deformation behavior of the target production (market) vehicle (a 2006 Kia Rio in this study). Additionally, the behavior of aluminum honeycomb at high levels of compressive stress (and deformation) was investigated using experimental testing and the findings were integrated into the surrogate vehicle design process. V alidation crash testing of the final surrogate vehicle (i.e., a combination of an impact block and a crushable nose constructed from aluminum honeycomb cartridges) demonstrated an acceptable match to the behavior of the target vehicle. Subsequently, the ne wly developed MASH compliant surrogate vehicle was used to experimentally evaluate, using pendulum impact testing, multiple prototypes of a new hybrid (FRP and steel) breakaway connection design. The final design was demonstrated to meet the impact risk re quirements stipulated in MASH as well as the hurricane wind resistance requirements specified by AASHTO and FDOT.

PAGE 16

16 CHAPTER 1 BACKGROUND While measures to prevent vehicles from running off roadways are necessary, measures to build a more forgiving roadside environment, where collision severity can be reduced, are also very important. From the perspective of roadside safety, it is desired that an errant vehicle can encroach onto the roadside without striking fixed, rigid object s. Ideally, positioning fixed objects at less vulnerable locations is typically a good strategy for minimizing collision consequences. However, as a component of traffic control systems, multi post ground signs must be located adjacent to roadways. As a re sult, such ground signs will pose potential hazards to passengers if they are not designed, fabricated, and installed so that they break away during a vehicular collision. Survivability requirements for ground signs require that they must be strong enough to resist both hurricane wind loading and gravity loading (self weight). One means of meeting this requirement is to use sign supports (posts) that are rigidly connected to a ground level foundation system. However, satisfying wind and gravity loading req uirements in this manner unfortunately makes sign structures more dangerous to vehicle occupants should a vehicle sign collision occur. A vehicle striking a fixed object usually leads to abrupt deceleration or excessive compartment deformation that can ca use fatal injuries. To prevent collision related occupant fatalities, roadside hardware such as utility poles and ground level sign posts are designed to yield or break away under vehicle impact loading. That is, posts are designed to fail in a specific manner so as to allow a vehicle to pass through without abrupt deceleration. Most utility poles, such as light poles or small sign structures, can incorporate breakaway features without significantly affecting functionality. Large ground signs, however, pr esent a challenging engineering problem in that the sign supports must possess significant strength to be able to

PAGE 17

17 speed vehicle impact loading (Figure 1 1 ). Due to the large surface area over which wind load will act, moderate to large sign structures usually require large post sizes and rigid foundation connections to transfer wind loads to the ground. Such features, howev er, tend to work against the goal of providing a system that breaks away at low impact load during a vehicle impact. Figure 1 1 Loading conditions for large sign post slip base breakaway connection (C ourtesy of Consolazio et al. 2012) To prevent fatal injuries to vehicle occupants, breakaway support design concepts have been applied for most types of objects that are located adjacent to roadways In man y states, including Florida, multi post ground signs typically utilize a breakaway slip base connection to minimize potential collision hazards. The slip base connection is designed so that under vehicle impact, the post detaches at the slip surface betwee n base plates and rotates about a hinge on the post near the bottom of the sign panel (Figure 1 2 ). Conversely, wind loading is transferred to the

PAGE 18

18 foundation through base plates, clamping mechanism, and friction force that is developed at the slipping surface. Figure 1 2 Breakaway behavior of sign supports (C ourtesy of Consolazio et al. 2012 ) Because several large ground signs inco rporating this slip base connection collapsed (Figure 1 3 ) during hurricane events in 2004 and 2005, a research project was undertaken ( BDK75 977 40 Consolazio et al. 2012) to develop a n ew breakaway connection (Figure 1 4 ) with improved performance characteristics (greater strength, less sensitivity to installation procedures, etc.). Based on analytical modeling and simulation, the newl y developed connection requirements for breakaway systems (as set forth in AASHTO MASH 2009) and reduced the performance sensitivity to installation and maintenanc e issues (e.g., installed and maintained bolt torques). However, while the shear controlled moment collar which was fabricated solely from structural steel was deemed to perform satisfactorily (based on simulation results), it was also

PAGE 19

19 determined that even further improvements in performance might be achievable if alternative materials, such as lightweight fiber reinforced polymer (FRP), were also incorporated into the connection design. Figure 1 3 Failure of sign structure with slip base connection under hurricane wind loading (C ourtesy of Consolazio et al. 2012) Figure 1 4 Shear control led moment collar breakaway connection developed previously in study BDK75 977 40 ( C ourtesy of Consolazio et al. 2012)

PAGE 20

20 Hence, a primary objective of the research presented in this report was to develop an alternative multi post ground sign breakaway conne ction that incorporated the best hybrid connection that possessed performance characteristics which were better than could be achi eved using solely steel or FRP.

PAGE 21

21 CHAPTER 2 DESIGN BASIS 2 1 Introduction Design of breakaway support devices is governed in part by the AASHTO Standard Specifications for Structural Su pports for Highway Signs, Luminaires and Traffic Signals (AASHTO 2001) and the AASHTO Manual for Assessing Safety Hardware (MASH 2009) The AASHTO 2001 specifications require that breakaway supports be designed to meet both structural and dynamic performan ce requirements. For sign structures located in Florida, breakaway supports must be structurally capable of carrying dead load, wind load, and combinations of dead and wind loads Load tests are required if the structural capacity of the support structure is potentially diminished by the introduction of breakaway features. 2 2 Selection O f Sign S tructure The primary objective of this research project was to develop a hybrid breakaway base con nection for multi post ground signs. Multi post systems are usually used for large roadside signs. As sign panel area and clearance height increase, it becomes more difficult to satisfy both of the conflicting requirements of being wind resistant and bei ng able to break away during impact Consequently, the sign panel size used in developing the new breakaway connection was selected so that it was representative of large signs used in Florida. In a previous study conducted by Pinelli and Subramanian (1999 ) seven typical sign configurations used along Florida highways were identified The typical signs had panel depths varying from 5 ft to 12 ft; panel widths varying from 10 ft to 20 ft; and clearance heights varying from 9 ft to 15 ft. Sign panel areas an d first moments of area with respect to ground surface are presented in Table 2 1

PAGE 22

22 Table 2 1 Study of Break project (Pinelli and Subramanian 1999) Sign No Panel Width ( ft ) Panel Height ( ft ) Clearance Height ( ft ) Sign Area ( ft 2 ) First Moment of Area ( ft 3 ) 1 12 8 12 96 1536 2 20 12 11 240 4080 3 12 8 15 9 6 1824 4 19.5 6 13 117 1872 5 15.5 5 9 77.5 891 6 14 5.5 10 77 98 2 7 10 10 12 100 1700 For use in the present study, a sign system with panel dim ensions of 12 ft x 20 ft (depth x wid th) was selected. An overview of the sign structure configuration fo r breakaway connection development is presented in (Figure 2 1 ). It should be noted that the selected sign size has a panel area of 240 ft 2 whic h is much greater than the 54 ft 2 that is required to b e classified as a large road sign according to AASHTO. Figure 2 1 Panel dimensions and clearance height of selected sign (C ourtesy of Consolazio et al. 2012)

PAGE 23

23 2 3 Wind L oading A sign support structure must be designed to resist dead load, wind load and combinations of dead and wind that act on the structure during its service life. Desi gns of multi post ground signs are required to conform to the AASHTO Standard Specification for Structural Supports for Highway Signs, Luminaires and Traffic Signals (AASHTO 2001). Multi post ground signs in Florida must also be designed to meet the requir ements specified in the FDOT Modifications to Standard Specifications for Structural Supports for Highway Signs, Luminaires and Traffic Signals, 4 th Edition (FDOT Modifications to AASHTO 2001), Florida Department of Transportation Structures Manual, Vol. 9 (FDOT 2009). Note that ice loading is not applicable to the design of signs in the state of Florida, thus, the loads acting on the sign structure are wind loading and self weight. The first step in determining wind loading is to select the basic wind spee d. AASHTO (2001) provides a design wind speed for each county in the state of Florida. The design wind pressure at any point of the sign structure is then calculated as follows: ( 2 1 ) where P is design wind pressure (psf), K z is a height and exposure factor, V is the basic wind speed (mph), and G is a gust effect factor used to correct the effective velocity pressure for the dynamic interaction of the structure wi th the gust characteristics of the wind. AASHTO permits the gust effect factor to be taken as a minimum of 1.14. C d is the wind drag coefficient used to account for the effects of geometry of the element and the Reynolds number of the flow. The importance I r is a factor included to convert wind pressures associated with a 10 year mean recurrence interval to wind pressures associated with o ther mean recurrence intervals.

PAGE 24

24 According to AASHTO (2001), for a height above ground z, or 16.4 ft, (whichever is great er) K z is computed by the following equation: ( 2 2 ) where is the height above the ground at which the wind pressure is computed, and are constants that vary with exposure condition. AASHTO adopted exposure C for use in designing sign structural supports since it is considered to provide an accurate and conservative approach for such s tructures. Exposure C represents open terrain with scattered obstructions having heights less than 30 ft. For exposure C, and are taken as 9.5 and 900 ft correspondingly. Wind loading analysis is conduct ed by multiplying wind pressures by corresponding projected areas and then applying the resulting loads statically and horizontally to the sign structure of interest. From sign size, material, configuration and preliminary post section selection, self weig ht of the sign structure can be determined and combined with wind load for structural design. The combined effects of axial, shear, and bending moment due to wind loading and self weight are then analyzed. However, for most large multi post ground sign str uctures, structural design is controlled primarily by wind loading. 2 4 Impact L oading The typical design process for most structures begins with the determination of loads and load combina tions that will act on the structure. Loads acting on a sign structure in Florida can be dead load, wind load, and impact load. Load determination procedures for each type of load can be found in appropriate design provisions, guidelines, and design manual s. Static or dynamic analysis can then be performed using the determined loads to quantify member level structural

PAGE 25

25 design forces (e.g., axial, shear, moment). As presented previously, sign structure wind loads in Florida can be determined from the AASHTO p rovisions and the FDOT Structures Manual. Quantifying vehicle impact loads on a sign support structure, however, is not described prescriptively by the relevant codes. In fact, sign structures are not designed to resist vehicle impact loading, but rather t o breakaway or yield under such loading conditions. As such, quantifying impact load is not a design requirement. Instead, primary focus is the safety of vehicle occupants should a collision occur. Hence, dynamic performance criteria are employed to addres s the vehicle impact loading condition. AASHTO requires sign support structures to be designed to meet the dynamic performance criteria specified in MASH ( 2009 ). The evaluation criteria consist of three primary aspects: structural adequacy, occupant risk, and vehicle trajectory after collision. MASH provides guidelines for crash testing of highway safety features and performance criteria to evaluate test results. It therefore provides roadside safety hardware developers and user agencies with a basis to com pare the impact performance of proposed safety features. With the goal of providing uniform guidelines, MASH covers standardized test parameters, such as test facility, test article, test vehicles and surrogate occupants. In addition to test parameters, te st conditions for different roadside safety devices are assigned with suitable test levels. Test levels are further divided into different features and test number in which vehicle type, impact speed, impact angle, acceptable kinetic energy range, impact p oint and evaluation criteria are specified. Recommended data acquisition systems and parameters that need to be determined during different phases of the testing process are also included. The breakaway connection developed in this study conforms to the cr iteria set forth in MASH

PAGE 26

26 Evaluation criteria for dynamic performance of breakaway supports specified in MASH include: structural adequacy; occupant risk; and post impact vehicular response. Structural adequacy under impact conditions requires that the br eakaway support shall readily fail in a predictable manner by breaking away, fracturing, or yielding when struck head on (or at an angle of 25 degrees) by a standard vehicle with a mass of 1,100 kg (2420 lbf.), or its equivalent, at a nominal speed of 30 k m/h (19 mph) for low speed impact (AASHTO MASH Test 60). The mass of 1100 kg specified by MASH is equivalent to that of a small car. It is often more critical to evaluate breakaway performance using a small car since smaller mass leads to higher occupant i mpact velocities which can increase the risks posed to vehicle occupants. For most sign supports, and for the purpose of evaluating the breakaway mechanism and occupant risk measures, the low speed test is more critical than the high speed test since less kinetic energy is available to break the post. However, to evaluate post impact vehicle instability, test article trajectories and intrusion of structural components into the vehicle windshield, the high speed test can be more critical. AASHTO MASH Test 61 and 62 assess these behaviors by testing designated impact speeds of 50 km/h (31 mph), 70 km/h (44 mph) and 100 km/h (62 mph), with vehicle masses specified at 1,100 kg (2,420 lbf) and 2,270 kg (5000 lbf) respectively. In regard to occupant risk factors a nd evaluation, MASH limits the longitudinal component of occupant impact velocity (with respect to the interior surface of the passenger compartment) to no greater than 16 ft/s (4.9 m/s) with values less than 10 ft/s (3.0 m/s) being preferred. Maximum allo wable longitudinal and lateral components of occupant deceleration are limited to 20.49 g and preferably should not exceed 15 g. In addition to limits on velocity and deceleration at impact, detached elements (debris) from the breakaway support are not per mitted to penetrate, or show potential for penetrating, the vehicle occupant compartment or present an

PAGE 27

27 undue hazard to other traffic, pedestrians or personnel in a nearby work zone. Potential for serious injury to vehicle occupants, due to deformation of t he occupant compartment, is also not acceptable. Satisfactory performance of a support requires that, after striking the breakaway support, the vehicle remain upright (although the criteria allows moderate roll, pitch, and yaw rotations of the vehicle to occur). The post impact vehicular response evaluation criteria also require that after impact, the trajectory of the vehicle should not excessively intrude into an adjacent traffic lane. In addition to the MASH evaluation criteria cited above, AASHTO (2001 ) also provides additional requirements to ensure predictable and safe performance of breakaway supports. If full scale crash testing is not performed, the combined mass of the post and fixtures attached to breakaway supports is limited to a maximum of 450 kg (992 lbf). To prevent a vehicle from snagging after breaking a support away from its base, AASHTO (2001) limits the maximum stub height to 4 in. This specified limit also helps to prevent instability of the vehicle should a wheel of the vehicle strike the stub. For multi post breakaway sign supports, the hinge must be located at least 2.1 m (84 in) above ground level to prevent penetration of the sign into the windshield of the impacting vehicle. For testing of a sign support system, MASH recommends tha t the test be conducted with the panel that has the largest area among sign panels to be used on the support system, and that the aspect ratio of the sign should be typical of the largest panel. The sign panel material should also be that normally used in the support system.

PAGE 28

28 CHAPTER 3 VEHICLE IMPACT TESTING 3 1 Introduction Due to the significant costs associated with full scale vehicle impact testing, sur rogate vehicles that reproduce a specific car behavior are used to minimize test expenses without compromising the quality of results. Surrogate vehicles consist of a crushable nose which reproduces a vehicle impact behavior, and an impact block which prov ides the total mass. Due to continuing evolution of the national car fleet, vehicles characteristics in current standards (AASHTO MASH 2009) have been revised relative to previous standards (NCHRP 350) These changes created a need to design a new surrogat e vehicle from the modified specifications, and improving on the capabilities of previously developed crushable impact nose s 3 2 Market Vehicle Impact Testing O f Breakaway D evices The upda te from NCHRP 350 to MASH changed several test vehicle characteristics. The vehicle modification that most significantly influences testing of sign support structures was the change in total weight (mass) of the vehicle. The vehicle mass ( weight ) required by NCHRP 350 code was 820 kg (1 800 lbs) and the market vehicle used for testing was typically the 1979 Volkswagen Rabbit. Also, the speed of impact affects the inertial forces that are developed, thus changing the behavior of the car at different impact v elocities. High speed film frames of a Volkswagen Rabbit rigid pole impact test are presented in Figure 3 1 where the impact speed of the vehicle matched the NCHRP 350 low speed impact te st for sign support structures of 35 km/h (22 mph) The force deformation behavior of this vehicle during the impact test is presented in Figure 3 1

PAGE 29

29 A) B) Figure 3 1 Low speed (per NCHRP 350) impact. A) Test of 1979 VW Rabbit (Courtesy of NHTSA) B) Corresponding force deformation behavior (Courtesy of Engineering Systems Division 1990 ) The vehicle mass ( weight ) required by the MASH code is 1100 kg ( 2,420 lbf ) and one possible market vehicle used for testing is the 2006 Kia Rio High speed video frames of a 2006 Kia Rio rigid pole impact test are presented in Figure 3 2 where the impact speed of the vehicle was 50 km/h ( 31 mph) The force deformation behavior of this vehicle during the impact test is presented in Figure 3 2 MASH specifies a low speed impact velocity of 3 0 km/h ( 19 mph) which differs from the velocity at which this test was conducted This test was carried out at a higher velocity of 50 km/h (31 mph) with the on the basis that the force deformation respon se produced by the same vehicle at 3 0 km/h would not differ significantly.

PAGE 30

30 A) B) Figure 3 2 L ow speed (per MASH 2009) impact. A) Test of 2006 Kia Rio B) C orresponding force def ormation behavior (C ourtesy of Marzougui et al. 2011) 3 3 Su rrogate Vehicle Impact Testing O f Breakaway D evices MASH allows the use of surrogate vehicles for low speed impact testing of bre akaway sign support structures. These surrogate vehicles can be either pendulums or bogies. Conditions that must be met to permit the use of these surrogate test vehicles are : 1) the surrogate vehicle must possess the same essential properties as the produ ction vehicle it is intended to replicate; and 2) the production vehicle model that is replicated must meet the vehicle recommendations In other words, if the behavior of the surrogate vehicle matches the force deformation curve shown in the previous sect ion for the rigid pole test of the 2006 Kia Rio, then the use of this surrogate vehicle would be permitted. It must be noted that because the specifications of MASH Test 60 requires an impact speed of 3 0 km/h whereas the speed used in the 2006 Kia Rio tes t was 50 km/h the total deformation of the surrogate vehicle at the lower impact speed will be less

PAGE 31

31 than that shown in Figure 3 2 The total deformation related to the impact of the surr ogate vehicle will be a function of the kinetic energy of the surrogate vehicle before impact and the energy absorbed by the deformation of this vehicle (eq ual to the area under the force deformation curve). Therefore, the portion of the 2006 Kia Rio force deformation curve that will be used to develop a low speed impact surrogate vehicle is shown in Figure 3 3 where the maximum deformation is 17.75 in. Figure 3 3 Portion of 2006 Kia Rio impact test force deformation curve needed to develop a low speed, MASH compliant surrogate vehicle Surrogate vehicles developed in the past have most often been designed to reproduce the b ehavior of the 1979 VW Rabbit, which complied with the earlier NCHRP 350 specifications. Shown in Figure 3 4 are high speed video frames from a Midwest Roadside Safety Facility (MwRSF) pendulum impac t test conducted with a surrogate vehicle that consisted of a back block and a crushable nose. In Figure 3 4 the force deformation curve obtained experimentally from impact testing of the surrogate vehicle (impact block and crushable nose) is compared to a simplified stepwise linear approximation curve that can be used as a conceptual basis for designing such systems.

PAGE 32

32 A) B) Figure 3 4 Sur rogate vehicles characteristics A ) Photo o f 820 kg MwRSF crushab le nose. B) S tepwise linear approximation ( C ourtesy of Bielenberg et al. 2009) To reproduce a specific force deformation curve previous surrogate vehicles have use d materia ls that can deform with a predictable constant force. The basic design concept of using these materials is that by alternating the stiffness of t he material to stepwise linearly match the behavior needed, the surrogate vehicle behavior ( Figure 3 4 ) can be made to closely match the actual market vehicle behavior ( Figure 3 1 ) In order to develop and test a new hybrid breakaway sign su pport structure that was MASH compliant, a new surrogate vehicle that exhibited the behavior shown in Figure 3 3 first needed to be designed, constructed, and validated (experimentally) Earlier 820 kg surrogate desi gns (e.g. NCAC, TTI, MwRSF) were used as a starting point for develop ment of the new UF/FDOT 1100 kg surrogate vehicle. Several modifications (enhancements) were made to the

PAGE 33

33 design concepts used in developing earlier crushabl e noses so that an accurate 2006 Kia Rio force deformation curve could be reproduced by the newly developed crushable nose and overall surrogate vehicle.

PAGE 34

34 CHAPTER 4 CRUSHABLE NOSE: MATERIALS AND DESIGN CONCEPTS 4 1 Introduction In order to successfully design a new surrogate vehicle crushable nose, the behavior of the materials to be used in the design and construction process (e.g. aluminum honeycomb) had to be understood well enough that calculation procedures could be developed to predict the anticipated material response under impact loading. Once such calculation procedures were developed, individual components of the new crushable nose (e.g., crushabl e cartridges) could be efficiently designed using analytical methods, and then evaluated using physical testing. To achieve the required level of understanding of the material behavior, both a literature review and physical material testing were carried ou t. 4 2 Elastic Plastic Aluminum Honeycomb Behavior Used I n Surrogate V ehicles In order to reliably reproduce the force deformation response of a Kia Rio to within an acceptable level of err or, a material capable of deforming in a predictable manner was needed. Furthermore, the material had to also be capable of dissipating impact energy in a manner that could approximate that of the vehicle. For this purpose, previously developed surrogate v ehicles have typically made use of aluminum honeycomb cartridge s of constant rectangular cross sectional shape (Figure 4 1 a and Figure 4 1 b). Such cartridges have the convenient characteristic of compressing at a constant crush load (force) due to their constant cross sectional shape. Stated alternatively, they exhibit a perfectly plastic force deformation response. The magnitude of the crush lo ad can be modified either by changing the cross sectional area of the overall cartridge or by changing the crush strength of the aluminum honeycomb material itself. In the literature, aluminum honeycomb crush strength is often described as the axial load r equired to locally buckle the walls of the hexagonal cells that run through the th ickness of the

PAGE 35

35 materia l (Figure 4 1 d). Since this type of buckling is related both to the characteristic width of the hexagonal cells, as well as to the cell wall thickness, either (or both) of these parameters can be altered to modify the honeycomb crush strength. Figure 4 1 Aluminum honeycomb cartridge s A ) R ectangular. B ) R ectangular with punch out. C ) Hexagonal h oneycomb cell structure ( C ourtesy of Matas Groetaers ) To experimentally confirm the elastic perfectly plastic behavior of the aluminum honeycomb material that is widely d escribed in the relevant technical literature, a series of static compression test s were performed at the Civil and Coastal Engineering Structures Laboratory at the University of Florida. Each test was conducted using a 400 kip Tinius Olsen (T.O.) universa l test machine (Figure 4 2 ). The tests consisted on the compression of rectangula r aluminum honeycomb cartridges Cartridges are manufactured using corrugated aluminum sheets that produce cell structures with no imperfections. In order to initiate the compressive buckling process of the cells, a force larger than that required to continue this process is needed (Figure 4 3 ). However, to avoid an initial spike i (Figure 4 4 ). The upsetting process involved tapping the surface with a hammer to initiate minimal buckling of the cell walls. After this upse tting process was completed, the cartridge was ready to be tes ted. A ) B ) C )

PAGE 36

36 Figure 4 2 Aluminum honeycomb cartridge install ed in Tinius Olsen test machine to conduct a static compression test (Co urtesy of Matas Groetaers ) Figure 4 3 Hypothetical force deformation curve for compression of a rectangular aluminum honeycomb cartridge Using the T.O. machine, the upset cartridge was compressed at a constant deformation rate. Force levels measured during the test at different stages of deformation exhibited a nearly perfectly plastic response (Figure 4 5 ) as the collapse d (buckled) zone progressed downward through the sample thickness (Figure 4 4 c). The final stage of the compression test was the fully buckled stage (Figure 4 4 d). This stage begins when there is no more unbuckled aluminum cell height left in the cartridge thickness. Compression of the cartridge beyond this point will not be at the plastic crushing load, but instead involves compress ion of the already crushed aluminum

PAGE 37

37 cell walls. Consequently, force levels beyond this point do not remain constant, but instead increase dramatically (this deformation phase will be discussed in more detail later in this chapter). A ) B ) C ) D ) F igure 4 4 Stages of alumin um honeycomb cartridge behavior. A) Original. B) Upset. C ) Progression of buckling. D ) F ully buckled (Courtesy of Matas Groetaers ) Figure 4 5 Measured force deformation curve for compression of a rectangular aluminum compressive load)

PAGE 38

38 To create a surrogate vehicle with this material and to be able to reproduce the behavior of a specific car, it is necessary to place different aluminum honeycomb cartridges in series and with different cross sectional areas in order to produce the desired force deformation curve. This series of differently configured cartridges, together with additional structural components (spacer combined sequence of aluminum honeycombs cartridges to p roduce the target curve, a second major component needs to be included in the surrogate vehicle: the mass block (Figure 4 6 ). This part represents most of the mass of the vehicle (i.e., there is very little mass in the crushable nose structure). Surrogate vehicles consisting of a mass block and crushable nose are most typically produce the desired im pact speed at the bottom of the swing motion where the crushable nose makes contact with the article being tested (e.g., a breakaway connection). Figure 4 6 Schematic diagram of impact b lock and attached crushable nose (Note: cartridges have differing strengths, achieved either through selection of different cell size and 4 1 b)

PAGE 39

39 4 3 Existing Surrogate V ehicles Researchers have previously produced surrogate vehicles that behaved in a manner similar to the VW Rabbit that satisfied the NCHRP Report 350 small car test requirement. These surrogate vehicles were created as described above in that a stepwise linear function was used to approximate the target vehicle curve (Figure 4 7 ). In such a syst em, each aluminum honeycomb cartridge produces a force contribution that will be an average for the specific deformation level they will be subjected to, creating an overall curve similar to the VW Rabbit. In previous crushable noses, the method used to al ter the force level of the honeycomb cartridge was by changing the cross sectional area, the crush strength of the cartridges, and/or finally creating punch outs in cartridges (Figure 4 1 b). Punched out are as are typically created by pre compressing a specific portion of the cartridge cross sectional area so that the pre crushed cells do not contribute to the buckling resistance of the cartridge. Punch outs are done at the center of cartridges to keep the structure symmetric and therefore stable. Figure 4 7 b shows an existing crushable nose built by The Midwest Roadside Safety Facility (MwRSF) in Nebraska which replaces the VW Rabbit on impact t ests. A ) B ) Figure 4 7 Existing 820 kg surrogate vehicle A ) Stepwise theoretical design curve and vehicle curve B ) Crushable nose attached to impact block (C ourtesy of Bielenberg et al. 2009 )

PAGE 40

40 4 4 Development O f 1100 kg UF/FDOT Surrogate V ehicle Following a review of previously developed crushable nose designs, and after conducting a series of experimental cartridge com pression tests, several areas were identified where design innovation and improvement became feasible. 4 4 1 Tapered S ections Instead of relying on the step ped (piecewise constant) force deformation approximation that results from using constant cross section rectangular cartridges (of varying strengths and sizes), cartridges with linearly varying cross sectional areas were instead investigated (Figure 4 8 ) By using tapering cartridges, a vehicle force deformation can be approximated in a piecewise linear manner, instead of a step wise constant manner ( Figure 4 9 ) Smooth transitions of impact force from crushing of one cartridge to the next results in a more realistic behavior of the surrogate vehicle. Figure 4 8 Tapered (t rapezoidal) a luminum honeycomb cartridge with linearly varying cross sectional area (Courtesy of Matas Groetaers )

PAGE 41

41 Figure 4 9 Comparison of constant cross section and tapered cross sectio n aluminum honeycomb cartridges A ) Con stant rectangular cross section B ) Ta pered trapezoidal cross section C ) Rect angular cartridge approximation D ) Trapezoidal cartridge approximation 4 4 2 Aluminum Honeycomb Behavior Beyond T he Fully Buckled S tage When developing the UF/FDOT crushable nose, a first configuration design of aluminum cartridges was dynamically tested. The force levels for specific deformations were lower than predicted, less overall cartridges were crushed, and the cartridges that engaged in compression deformed more than expected. These results showed that the aluminum honeycomb cartridges were absorbing considerately rectangular and trapezoidal shapes. A ) B ) C ) D )

PAGE 42

42 The compression of the aluminum cartridges exhibited the sa me behavior as explained earlier in this chapter, but higher forces were applied to the sample. Therefore, a higher level of compressive deformation was produced and new behaviors past the fully buckled stage were observed. Figure 4 10 illustrates the different stages of cartridge compression that describe the full range of behavior of the aluminum cartridge structure. (Note that descriptions of deformation stages beyond the fully bucked state w ere not found in the literature review that was conducted in this study). The first three stages shown are no different from the stages described earlier. However, with further compression of the aluminum cartridge, two additional deformation stages become evident. After reaching the fully buckled stage, the stiffness of the cartridge increased considerably as the buckled aluminum cells compressed, creating a solid and thicker second cell structure. With continued loading, these thicker columns also buckle d (but at a much higher force level than was required to buckle the original cell walls). This stage will be referred to as as deformation continues to incre ase. This stage continued until the cell groups buckled and effect of the group cells was also seen through experimental observations. At the end of this stage of deformation, the aluminum honeycomb reached a condition that will be referred to as force deformation response. As this last stage of deformation continued, and the remaining void area in the material is filled in, the observed behavior approaches that of a solid block of aluminum.

PAGE 43

43 Schematic diagram Experimental test photo Honeycomb cellular structure A ) B ) C ) D ) E ) Fi gure 4 10 Stages of aluminum honeycomb def ormation. A) Mechanical upset. B ) Buckling of cell walls. C) Fully buckled stage. D) Group buckling stage. E ) Fully compressed stage (Courtesy of M atas Groetaers ) During an initial dynamic impact test of the UF/FDOT crushable nose, the second force plateau that occurs during the group buckling stage, produced a lower force level than was expected. This was due to the fact that the increased force le vel of the second plateau produced a greater energy absorption per cartridge (particularly in the weakest cartridges used) than was predicted by the elastic perfectly plastic behavior assumption that is commonly described in the literature. Figure 4 11 shows the extra capability for each cartridge to absorb more energy.

PAGE 44

44 Figure 4 11 a shows the behavior of a rectangular cartridge which can be re lated to the images presented in Figure 4 10 In Figure 4 11 b, the behavior shown corresponds to the compression of a tapered trapezoid al cartridge. In a tapered cartridge, the difference between the initial plateau force level and the second plateau force level is greater than it is in a similarly sized rectangular cartridge. The reason for this is that the initial plateau force level is associated with the cross sectional area of the smaller (e.g., top) surface of the cartridge (this top cross sectional area increases through a tapered cartridge compression), whereas the second plateau force level even though is still associated to the i nitial smallest surface of the cartridge (seen by the residual material compressed in the cartridge) the difference between the original smaller surface and the current smaller surface cross sectional area is contributing to the st rength by cell buckling a ction. A ) B ) Figure 4 11 Hig h stress compression schematics. A ) R e ctangular geometry. B ) T rapezoidal geometry After characterizing the full range of deformation behaviors of the alumi num honeycomb cartridges, and introducing the use of tapered cartridges, it became evident that the process of designing an overall crushable nose would benefit from the use of nonlinear finite element analysis (FEA). For each trial set of cartridge config urations, FEA was used to numerically assess whether the anticipated response of the crushable nose would reproduce the target vehicle curve closely enough.

PAGE 45

45 To efficiently analyze crushable nose finite element models, the force deformation response curve for each cartridge needed to be computed prior to FEA model development. Accuracy on determining this individual curves was critical to produce an effective and practical model of the system. Figure 4 12 illustrates the calculation procedure used to compute simplified curves representing how the aluminum cartridges behave in compression. The procedure illustrated was developed based on physical observations from high stress compression tests performed o n several cartridges. From the experimental data collected, the deformation level was identified at each stage described earlier (this was a unique property of each aluminum honeycomb that depended on aluminum foil thickness, cell size, etc). Next, appropr iate force levels were associated with each range of deformation. The force deformation behavior curve for each cartridge was simplified down to five (6) key points. Point 1 represented the initial crush force required to buckle the honeycomb cells from th e original smaller cross sectional area compressed, assuming the cartridge is mechanically upset (Figure 4 12 a). To calculate the deformation level several test results were seen for a specific type of honeycomb. To calculate the force the initial cartridge surface area was multiplied by the crush strength of the honeycomb material. Point 2 represented the force level corresponding to the fully buckled condition (Figure 4 12 b) The smaller surface area compressed at the time where the fully buckled condition was reached, got multiplied by the crush strength of the honeycomb material. The deformation was taken from experimental test results. Point 3 r epresented the condition at which the second force plateau associated with group buckling occurred (Figure 4 12 c). The deformation was measured at the point where the group cells buckled alleviating th e force. The force was calculated by multiplying the cross sectional area related to the material engaged in the fully buckled state (from geometry)

PAGE 46

46 multiplied by the crush strength of the material and a coefficient determined from testing that relates to this stage, plus the contribution of the cross sectional area undergoing buckling at tha t specific time (from geometry) Point 4 represented the same condition as point 3 (Figure 4 12 c) with the differ ence that the deformation was measured at the point where the group cells buckling is over, increasing the material stiffness dramatically from experimental tests. Finally Point 5 and 6 represented the condition (Figure 4 12 d) at which the fully compacted stage occurred. These points were selected arbitrarily, together with points 3 and 4, although to a lesser degree. In order to keep the energy absorption of the cartridge into a realistic level, points where positioned on the described condition with special care to adjust small variations in data and follow test data closely. Point 5 was selected in order to maximize the curve fitting from test results and to produce a realistic final stiffness for the cartridge at its fully compacted stage represented by point 6. The force related to point 5 and 6 were calculated by multiplying the cross sectional area of the material engaged into the fully buckled stage at that specific point with the crush strength o f the material and a coefficient calculated for the specific point selected. The deformation was arbitrary as mentioned earlier. (The cartridge area that was still undergoing re be neglected). From experimental test results, it was observed that the force levels of both the group buckling and fully compacted stages followed a predictable pattern. Even though different aluminum honeycomb material strengths exhibited different be haviors, it was determined that the force and deformation levels at each stage (fully buckled; fully compacted) were related to the cartridge crush force (crush strength and specific cross sectional area) by a specific coefficient that could be empirically determined. Hence, as long as the properties of the aluminum honeycomb deformation level at each stage; crush strength of cartridge; and coefficients can

PAGE 47

47 be quantified all six points on the simplified force deformation curve can be readily computed. When force deformation curves constructed in this manner were compared to results from experimental cartridge compression test data, favorable agreement was observed ( Figure 4 13 ) Schematic diagram Geometric dimensions used in calculation procedure A ) B ) C ) D ) Figure 4 12 Deformation stages and geometric dimensions used in computing force simp lified force de formation curves A) Initial buckling B ) Fully buckled stage C) Group buckling D ) Fully compacted stage A ) B ) Figure 4 13 Comparison of computed and measured force deformation curv es A) Rectangular cartridge B ) Trapezoidal cartridge

PAGE 48

48 4 4 3 Force deformation Behavior O f Tapered C artridges Arranged I n S eries When designing a crushable nose with rectangular cartridges, as has been the typical practice in previous research studies, each cartridge in the nose generally has a unique strength (or stiffness) (different from than all other cartridges). Consequently, when the cartridges are pl aced in series and compressed, they crush sequentially, starting from weakest and progressing to the strongest. This process produces a step wise force deformation curve, as noted earlier (Figure 4 7 ). Since each cartridge strength is unique, in general, only one cartridge undergoes compressive crushing at a given point in time. In contrast, when tapered (trapezoidal) cartridges of varying material strength and varying top and bottom surface sizes (c ross sectional sizes) are placed in series and compressed, multiple cartridges may compress simultaneously rather than sequentially (one following another). For example, as a tapered cartridge is compressed, it becomes stiffer with increasing deformation ( as the effective cross sectional area of the crush zone increases). If the stiffness (and associated crush force) grows large enough, it is possible that crushing on the small surface of the next cartridge in line may be initiated, thereby producing a cond ition where two cartridges crush simultaneously. Such a condition increases the complexity of the crushable nose design process because account the possibilit y of simultaneous crushing. However, simultaneous crushing is also advantageous crushable nose force deformation curve that is produced. For example, through careful sizing of m ultiple tapered cartridges placed in series, nearly any desired piecewise linear force deformation curve can be constructed, including control over whether the curve is concave up or down at various deformation levels. A simplified example of this concept is illustrated in

PAGE 49

49 Figure 4 14 for two tapered cartridges in series. Simultaneous interaction (crushing) of the two cartridges produces a series combined force deformation curve that is mor e complex in shape than the force deformation curves of either of the cartridges individually. Given the design advantages of combining tapered cartridges in series, this approach was used in the design of the UF/FDOT crushable nose that is discussed in mo re detail in the following chapter. A ) B ) C ) Figure 4 14 Simultaneous compression of tapered cartridges in series. A ) Cartridge 1 behavior B ) Cartridge 2 behavior C ) Behavior of c artridges in series compression

PAGE 50

50 CHAPTER 5 CRUSHABLE NOSE : DESIGN BY WAY OF NUMERICAL SIMULATION 5 1 Introduction As discussed in Chapter 4 the UF/FDOT crushable nose, targeting a Kia Rio 2006 at 19 mph ( 3 0 km/h ), was designed using trapezoidal cartridges. Due to the nonlinear curves describing the honeycomb material, FEA analysis was re quired to achieve an accurate design. After modeling and analyzing the FEA numerical simulation, the results were expected to be very close to the force deformation curve of the target vehicle. Figure 5 1 illustrates the behavioral curve of the Kia Rio vehicle and also an average curve that was produced from the were removed using smoothing, but the main character istics were retained for purposes of designing the crushable nose. The averaged curve also has an energy dissipation level that is equivalent to that of the Kia Rio at the point of maximum expected deformation. Therefore, the curves are not only similar in behavior but also in energy dissipation characteristics. Hence, the FEA models that will be discussed in this chapter were designed to reproduce the averaged Kia Rio curve. Figure 5 1 Kia Rio force deformation curves

PAGE 51

51 Although the choice of cartridge configuration was extremely important in producing an acceptable crushable nose design, most of the main structural components of the surrogate vehicle (excluding the cartridges) also ne eded to be reusable and capable surviving hundreds of impact tests. FEA modeling and analysis was thus used to analyze complex design geometries for different structural members and to evaluate their capacities, taking into consideration different material properties, such as the aluminum front block, fiberglass spacer plates, and the aluminum guide tubes. 5 2 High Reso lution Modeling Of Front Block A nd T ubes Structural demands (e.g., stress es and strains) in the front block and telescoping guide tubes of the crushable nose were assessed, for design purposes, using a preliminary finite element model (Figure 5 2 ) that employe d rectangular, rather than tapered, aluminum honeycomb cartridges. In this preliminary model, all honeycomb cartridges were modeled using LS DYNA (LSTC 2007) material t ype 26 (* MAT_HONEYCOMB ). As will be discussed later in this chapter, modeling tapered (a s opposed to rectangular) cartridges with material t ype 26 required the use of very high resolution meshes, which then lead to numerical inefficiency in the simulation process. Therefore, solely for purposes of designing the solid structural aluminum fron t block and tubes, geometrically simpler rectangular cartridges which produced force levels similar to the final tapered cartridges were used in the model instead. (Note that a more numerically efficient method for modeling tapered cartridges was later dev eloped and was used to evaluate the dynamic impact performance of the overall crushable nose. This modeling method, utilizing nonlinear spring elements, is also discussed later in this chapter). Major structural components of the impact back block were m odeled with material type 1 (* MAT_ELASTIC ) and were assigned the material properties (modulus and mass density) of either concrete or steel, as appropriate. High resolution meshes were used to model two

PAGE 52

52 embedded steel tube s, which were then nodally merged to the rest of the impact back block components Fiberglass spacer plates, 0.5 in. in thickness, were also modeled using material type 1 (* MAT_ELASTIC ). Material properties of the fiberglass were obtained from the literature. For purposes of conducting rig id pole impact simulations, a nearly rigid steel pole was modeled using material model type 1 (* MAT_ELASTIC ) with steel material properties and a wall thickness of 1 in. Shell elements were used to model the pole to decrease the computational time (relativ e to using a higher resolution method of solid elements). Figure 5 2 Preliminary finite element model used to assess structural demands in front block and telescoping guide tubes The fron t block was designed on the basis of limiting Von Mises (effective) stresses to levels that were deemed acceptable for 6061 T6 aluminum. A high resolution front block finite element model consisting of solid (brick) element was developed (Figure 5 3 ) to quantify stresses that arise during surrogate vehicle impact against the rigid pole. The structure was modeled using material type 24 (* MAT_ PIECEWISE_ LINEAR_ PLASTICITY) in order to represent the nonlinear, inelastic properties of aluminum 6061 T6. In Figure 5 3 b and Figure 5 3 c, maximum (worst case) Von Mises stress es arising in the front block during rigid pole impact are presented. Using this type of analysis, a set of structurally adequate front block dimensions were determined that kept the effective stresses, and inelastic strains, to sufficiently low levels.

PAGE 53

53 De tailed structural (fabrication) drawings for the final dimensioned front block structure are included in Appendix B A ) B ) C ) Figure 5 3 Finite element model of UF/FDO T surrogate vehicle front block. A) Mesh. B C ) Maximum Von Mises stresses during impact simulation (Note: blue = 0 ksi Von Mises stress; red = 40 ksi Von Mises stress) The telescoping guide tubes (Figure 5 4 ) were also modeled using material type 24 (* MAT_ PIECEWISE_ LINEAR_ PLASTICITY) with aluminum 6061 T6 properties In order to simulate sliding of the guide tubes inside the steel sleeve tub es that are embedded in the concrete back block, a surface to surface contact detection technique was used in the finite element model. To design (i.e., size) the guide tubes such they had adequate structural capacity, a worst on this analysis, the finite element model of the surrogate vehicle was configured to impact the rigid degrees (deemed to be a reasonable upper limit). Maximu m Von Mises stresses in the guide tubes (Figure 5 4 ) were computed from analyses of this type to determine a tube diameter and wall thickness that kept effective stresses, and inelastic strai ns, to sufficiently low levels. Ultimately 3 in. diameter, schedule 80 aluminum 6061 T6 guide tubes were found to be structurally adequate (see Appendix B for final fabrication drawings).

PAGE 54

54 A ) B ) C ) D ) E ) Figure 5 4 Finite element model of UF/FDOT sur rogate vehicle guide tubes. A) Mesh; B C ) Worst case V on Mises stresses during impact. D E ) Next to worst case V on Mis es stresses during impact (Note: blue = 0 ksi Von Mises stress; red = 40 ksi Von Mises stress) 5 3 High Resolution FEA Modeling And Analysis O f Aluminum H oneycomb C artridges When analyzing the preliminary finite element model noted previously in Figure 5 2 which made use of rectangular cartridges the LS DYNA material type 26 (* MAT_HONEYCOMB ) was successfully used to model t he material behavior of the honeycomb. However, as discussed in Chapter 4 tapered aluminum honeycomb cartridges were used in the final crushable nose design because they are able to repr oduce a target vehicle curve more accurately than rectangular cartridges. To determine whether material type 26 would also be applicable to the analysis of tapered cartridges, two test models were developed; used to conduct compressive crushing simulations and compared to corresponding experimental test data. The first test model involved trapezoidal geometry. In Figure 5 5 the finite element model and the corresponding experimental test specimen are presented at different stages of compression. In general, the progression from top of cartridge to bottom of cartridge of the

PAGE 55

55 buckled zone predicted by finite element analysis agreed favorably with the experimental test photos. Importantly, in the final state presented in Figure 5 5 e, the finite element model accurately predicted the distribution of fully buckled material (spanning most of the cartridge width) versus partially b uckled m aterial (near the outer edges). Finite element model Experimental test A ) B ) C ) D ) E) Figure 5 5 Finite element model and experimenta l test of trapezo idal cartridge A B) Un deformed. C D ) Int ermediate stages of compression. E ) Fully buckled state (Courtesy of Matas Groetaers ) The second test model involved a sharp tipped pentagonal geometry. The initial undeformed finite element mesh of the cartridge is shown in Figure 5 6 a, and a corresponding physical specimen which was subjected to experimental compression testing is shown in

PAGE 56

56 Figure 5 6 b. In contrast to the compression simulation of the tapered cartridge discussed above, the process of simulating compression of the pentagonal cartridge was found to be exceedingly computationally demanding and slow. Due to the sharp geom etry and refinement of the mesh near the cartridge tip, high level localized deformations occur. As a result, the time step size required to maintain analysis stability decreases dramatically and the overall simulation time therefore grows rapidly. For exa mple, simulating cartridge compression from the initial condition shown in Figure 5 6 c to the condition shown in Figure 5 6 d took more than a week on a multi core computer. Such analysis times precluded efficient impact simulation of an overall crushable nose consisting of approximately a half dozen cartridges. A) B) C) D) Figure 5 6 Sharp tipped pentagonal aluminum honeycomb cartridge. A) Un deformed. B) Experimental test specimen. (Courtesy of Matas Groetaers ) C) Initial condition of compression simulation. D) Partially crushed condition during compr ession simulation. Two potential methods considered for reducing the analysis time were: the use of mass scaling, and the use of a coarser mesh. Both of these techniques, however, must be used with great caution if accurate results are to be obtained. Inst ead, the approach taken in the present study was to not use material type 26 (* MAT_HONEYCOMB ) to model the aluminum honeycomb

PAGE 57

57 cartridges, but instead to developed a simplified and much more numerically efficient approach involving the use of nonlinear spri ngs. 5 4 High Resolution Nonlinear Spring Crushable Nose M odel In an attempt to use a simplified method to model aluminum honeycomb cartridges that could produce reliable results and be ef ficient in terms of numerical simulation time, material type 121 (* MAT_ GENERAL_ NONLINEAR_ 1DOF_ DISCRETE_BEAM ) was implemented. Figure 5 7 illustrates the FEA final model used to design the alu minum honeycomb cartridge configuration using several one degree of freedom spring elements per cartridge. A ) B ) C ) Figure 5 7 UF/FDOT surrogate vehicle finite element model with gr ids of nonlinear springs used to model ea ch aluminum honeycomb cartridge. A ) Overall view. B) Elevation (side) view. C ) P lan (top) view A high resolution grid of nonlinear spring elements was created for every cartridge (Figure 5 7 ). The stiffness of individual springs corresponded to the average resistance of the tributary area of each element (Figure 5 8 a) and the properties of the specific honeycomb. In

PAGE 58

58 Figure 5 8 both parameters are equal for all springs (same tributary area and honeycomb properties), therefore, the behavior in terms of stiffness is the sam e for all. Springs located at the slope of the cartridge do not extend to the full height of the honeycomb, simulating the physical geometry of the tapered trapezoidal cartridges. Due to the shorter length of these elements compared to the springs located at the center, the deformation behavior was proportional to the difference in length of the spring, having the same force level as explained earlier (Figure 5 8 b). Also, these elements di d not engage in compression until the distance between both spacer plates reached the element height, making the side springs contribute to the stiffness of the cartridge in a delayed manner just like in a real scenario. A) B) Figure 5 8 Behavior of sample nonlinear springs grids. A) Side view of honeycomb modeling simplification. B) Illustration of material curves.

PAGE 59

59 Full size springs simulate the material behavior accurately (only limit ed to the material curve introduced to the model), whereas springs located at the cartridge slope simulate an average strength of the real material. This suggests that the grid resolution of elements located under slopes can increase the accuracy of result s. A discretization of 10 elements was used in sloped areas of every cartridge and a minimum of 5 elements for full size springs as they do not compromise results. Numerical simulations determined that the flexibility of spacer plates separating aluminum h oneycomb cartridges had a significant effect in the results of simulations. In contrast to a stiffer spacer plate, a flexible spacer would deform under the compression loads applied by two honeycomb cartridges. This compression produces higher global defor mations of the overall system at equivalent force levels (Figure 5 9 b), compared to the existing surrogate vehicles having a rectangular cartridge configuration with punch outs (Figure 5 2 ) Hence, stiffer spacer plates were required by a tapered trapezoidal cartridge configuration due to the larger cantilever distributed force acting on spacer plates from one cartridge to the next. Garolite G 11 was a stiff and light material considered for the implementation of spacer plates. This material was tested in the Civil Engineering lab at the University of Florida A modulus of elasticity of 3109 ksi was recorded. Further finite ele ment analysis determined that a thickness of in. would keep the deformation levels of the spacer plates at an acceptable level (Figure 5 10 ). Figure 5 10 FEA model of UF/FDOT crushable nose using in. thick Garolite G 11 spacers

PAGE 60

60 Solid elements with material type 1 (* MAT_ELASTIC ) were used to model the spacer plates as failure stresses were far from occurring. The use of node to surface contact detection technique was applied to the FEA model to separate spring elements and spacer plates under compression. This contact definition was also used for the interphase between the last spacer plate and the back block. A modificat ion to the properties of the rigid impactor was implemented from the preliminary FEA model previously described. A material type 20 (* MAT_ RIGID ) was used in addition to a much coarser mesh that merged both concrete and steel materials. A density which prod uces the total weight of the impact block was given to the rigid material Due to the high compression strength of both materials, this assumption was valid, allowing the model to reduce simulation efficiency by 1/6 th of the original simulation time. Resul ts do not differ from a comparable model with the prior properties, thus validating the changes (Figure 5 11 ). Figure 5 11 UF/FDOT surrogate vehicle high resolution FEA model results 5 5 Simplified 8 DOF Crushable Nose M odel A second UF/ FDOT surrogate vehicle FEA model was developed. Figure 5 12 illustrates a diagram describing the 8DOF simplified model. The advantage of using this model is the computational time savings. The simulation time of the simplified model is about 10 seconds compared to the 4 h ours of the high resolution model. This was accomplished by using material

PAGE 61

61 type 121 (* MAT_ GENERAL_NONLINEAR_1DOF_DISCRETE_BEAM ) (LSTC 2007) and introducing the force deformation curves of the aluminum honeycomb cartridges developed in Chapter 4 It also required the use of nodal masses and different nodal constrains where the input of masses, velocities and curves were critical for proper results. Figure 5 12 UF/FDOT surrogate vehicle simplified FEA model results Figure 5 13 shows the results of the simplified FEA crushable nose model where the data collection a nd analysis matches the procedure to obtain experimental test results from Chapter 6 Figure 5 13 UF/FDOT surrogate veh icle simplified FEA model results

PAGE 62

62 5 6 Final Crushable N ose D esign Numerical models previously described have determined the UF/FDOT surrogate vehicle design. Figure 5 14 displays different views of the surrogate vehicle final design and cartridge configuration. Technical drawings of the impact block and the crushable nose frame structure can be found on Appendix A and Appendix B respectively. Lastly, Table 5 1 presents a detailed list of the mate rial properties, geometries, and order of the aluminum cartridge and spacer configuration. A ) B ) C ) Figure 5 14 UF/FDOT crush able nose and impact block. A) Side view. B) Top view. C ) I sometric view

PAGE 63

63 Table 5 1 Summary of crushable nose components Part Number Schematic Dimensions Material Weight (lbs) 1 st Cartridge Aluminum Honeycomb S trength: 242psi 0.1855 2 nd Cartridge Aluminum Honeycomb Strength: 242psi 0.1080 2 nd Spacer Garolite G 11 7.8280 2 nd Spacer Aluminum 2 nd Spacer Garolite G 11 3 rd Cartridge Aluminum Honeycomb Strength: 242psi 0.2025 3 rd Spacer Garolite G 11 4.6715 4 th Cartridge Aluminum Hone ycomb Strength: 242psi 0.2520 4 th Spacer Garolite G 11 9.3605 5 th Cartridge Aluminum Honeycomb Strength: 242psi 0.4700 5 th Spacer Garolite G 11 10.1300

PAGE 64

64 Table 5 1 C ontinued Part Number Schematic Dimensions Material Weight (lbs) 6 th Cartridge Aluminum Honeycomb Strength: 513psi 1.2780 6 th Spacer H: 16.5 Garolite G 11 11.0570 7 th Cartridge Aluminum Honeycomb Strength: 513psi 1.5900 7 th Spacer (Optional) Garolite G 11 11.4460 8 th Cartridge (Optional) W: 10.375 Aluminum Honeycomb Strength: 513psi 1.9665 8 th Spacer Structural Fiberglass (FRP) 6.8735 5 7 Impact F D C urves: N umerical Simulation V s. MASH Complian t C ar (Kia) Both FEA models described earlier produced force deformation curves shown in Figure 5 15 The high resolution and simplified crushable nose models agree very closely. It is

PAGE 65

65 important to emphasi ze the fact that the calculations for the honeycomb cartridges curves in each model used different approaches. The simplified model calculated curves using the trapezoidal cartridge calculation procedure and the high resolution model used only the rectangu lar cartridge calculation procedure due to the high resolution beam elements grid produced. Figure 5 15 UF/FDOT surrogate vehicle high resolution and simplified FEA model results The nume rical simulations results not only matched but also closely agreed with the Kia Rio test curve, and most importantly the Averaged Kia Rio curve which the models were intended to reproduce. Figure 5 16 illustrates the plot comparing both FEA models with the Kia Rio curves. Some oscillations differ from the Kia Rio curve but the general characteristics of the behavior are met very closely. Figure 5 16 UF/FDOT surrogate vehicle FEA model results and Kia Rio test data curves

PAGE 66

66 CHAPTER 6 CRUSHABLE NOSE: EXPERIMENTAL VALIDATION 6 1 Int roduction to P endulum Impact T esting Once the design of the crushable nose and the FEA simulations were finished, an experimental test to validate the MASH standard surrogate vehicle was ready to be executed. The test was performed on May 9 th of 2013 at th e FDOT structures lab in Tallahassee, Florida. Photos of the UF/FDOT surrogate vehicle can be seen in Figure 6 1 A ) B ) Figure 6 1 UF/FDOT crushable nose and impact block A) Isometric view. B ) C rushable nose side view (Courtesy of G ary Consolazio ) The use of three 50 ft. tall pendulum t owers were required to swing the surrogate vehicle and reach a specific impact veloci ty. Figure 6 3 and Figure 6 3 show photos and schematics of the test befor e and after it was carried out. A ) B ) Figure 6 2 Pendulum structure. A ) S chematic of pendulum swing path. B) Pendulum release photo. (Courtesy of G ary Consolazio )

PAGE 67

67 A ) B ) Figure 6 3 I mpactor bloc k and crushable nose structures. A ) Isometric schematic of surrogate vehicle B ) C rushable nose post compression (Courtesy of G ary Consolazio ) 6 2 eight, E tc.) The validation test was conducted by the conditions specified in MASH for test number 60 an 1100 kg vehicle impacting a test article at a speed of 19 mph. The impact velocity corresponded to a drop height of 12 ft. using a surrogate vehicle in a pendulum facility. The test article in this test was a 10 in. diameter rigid pole, following the NCAC test 11005 which assesses the impact behavior of a 2006 Kia Rio vehicle. Test 11005 was conducted at a speed of 30 mph, but with the intent to be used in the desi gn of an 1100 kg surrogate vehicles traveling at 19 mph such as the UF/FDOT crushable nose. See Figure 6 4 for a test set up illustration. Figure 6 4 Crushable nose validation test

PAGE 68

68 6 3 Instrumentation Instrumentation needed for the rigid pole validation test using the FDOT crushable impact nose consisted of the following devices : 6 3 1 High Speed C ameras Two (2) high speed cameras were needed to record the behavior of the rigid pole validation test. The se cameras (high speed cameras 1 and 2) recorded detailed views including the crushable nose to rigid pole contact area and compression of the honeycomb cartridges (Figures 6 5 and 6 6 ). Even though they recorded the same view, each camera was set looking against each other in order to record both lateral sides of the impact (Figure 6 7 ). All high speed cameras captured images at a frame rate of 2000 frames per second. A third camera (normal speed) was used to capture a wider view in order to show in presentations (Figure 6 7 ). Figu re 6 5 Camera 1 and 2 wide test view Figure 6 6 Camera 2 and 3 detailed view of crushable nose area from both sides

PAGE 69

69 Figure 6 7 Top v iew of test set up for cameras layout A software program called ProAnalyst was used to track points from the footage of both high speed cameras. Initial calibration and image filtering were critical to successfully produce displacement time data points. Tracking points were produced by creating white squares surrounded by black color which produced contrast that was enhanced by different filtering techniques available in the software (Figure 6 6 ). Tracking points were placed on both sides of the back block, spacer plates, front block, and rigid pole. The recorded data points were used to calculate deformations at spe cific time intervals. They also provided a very accurate redundant measurement of acceleration for both front and back block by integrating the displacement time curve twice. Figure 6 8 shows the tracking point process carried out by ProAnalyst to get the displacements of all points of interest as data points.

PAGE 70

70 Figure 6 8 High speed camera frames of tracking points procedure using P roAnalyst software 6 3 2 Break B eams Two (2) pairs of optical break beam sensors (Figure 6 9 ) were used to determine the block speed just prior to impact. Each pair had a transmitter and a receiver. The first pair of sensors was located 2 inches away from the rigid pole and the second pair 12 in. away from the first pair. Figure 6 9 Position of infrared optical break beam sensor

PAGE 71

71 6 3 3 Tape S witch One (1) 1 8 in. long (ribbon) tape switch was needed in order to determine the moment in time at which the crushable nose first contacted the rigid pole. Contact was also used to set the high speed cameras to be synchronized with the accelerometers so a force deform ation curve could be accurately produced. The sampling rate for this device was 10 kHz. The tape switch was placed vertically along the impact face of the rigid pole as shown in Figure 6 10 Figure 6 10 Position of tape switch installed on rigid pole 6 3 4 Accelerometers One (1) 50 g and one (1) 30 g accelerometer with a bandwidth frequency of 400 Hz, plus one (1) 250 g with bandwidth of 10 kHz were needed. The sampling rate for these three devices (and the previously mentioned tape switch) was 10 kHz. The 50 g and 30 g ac celerometers were mounted to the top surface of the concrete block and oriented in the direction of impact. Both instruments were centered at the mid length of the block, one next to the other. The 250 g accelerometer was mounted to a front block stiffener and oriented parallel to the direction of impact. Acceleration data points were used to measure the force contribution of both the back

PAGE 72

72 block and the front block which were superimposed to represent the test contact force. See Figure 6 11 for accelerometers positioning. Figure 6 11 Locations of accelerometers installed on the surrogate vehicle 6 4 Impact Test R esults After the experimental test was carried out and the data analysis was processed, results of the UF/FDOT crushable nose behavior were obtained. Figure 6 12 shows some high speed camera frames describing the compression of the crushable nose when striking the 10 in. diameter rigid steel pole. Figure 6 12 High speed camera video frame s of validation test (Courtesy of Chris Weigly ) The accelerometers placed on the impact block were averaged and then filtered at the same bandwidth related to the frequency of the accelerometers (400Hz) using ksmooth technique. Both 50g and 30g accelerome ters closely agree with the acceleration values

PAGE 73

73 throughout the impact time. Figure 6 13 illustrates the processed acceleration time curve of the impact block. The accelerometer placed on the fron t block was also filtered, but with a bandwidth related to twice the frequency of the back block accelerometer (800Hz) using ksmooth technique. Figure 6 14 shows the acceleration data of the fro nt block structure during the same interval of time. This data was compared with the high speed camera front block acceleration data points obtained from the double integration of the displacement data taken by ProAnalyst software. Both of these outcomes m atched closely, validating the results gott en from the 250g accelerometer. Figure 6 13 Filtered and averaged acceleration data from back block accelerometers Figure 6 14 Filtered acceleration data from front block accelerometer

PAGE 74

74 The filtered acceleration data was used to calculate the force contribution of each crushable nose part by multiplying them to their corresponding mass measured before the test was carried out The resultant force time histories were super imposed to produce the total force time history of the entire system. Figure 6 15 describes t he front and back blocks force contributions and the total force of the impact. It can be seen that even though the front block was subjected to very high accelerations, the contribution of force to the overall impact was minimum relative to the total forc e produced by the system. This can be given to the relative low mass of the front block which was minimized in order to have the least effect on the o verall crushable nose behavior. Figure 6 15 Force contribution to the total system by the front and back block structures The impact test deformation was measured by the high speed cameras by measuring the difference in distance between the back block and the rigid post starting at the time o f contact. Figure 6 16 illustrates this data with respect to the impact time. The deformation of the crushable nose and the total system force throughout the entire impact were combined to show the final behavior of the UF/FDOT surrogate vehicle representing a Kia Rio car under impact at a speed of 19 mph Final force deformation plot of the crushable nose validation test is shown in Figure 6 17

PAGE 75

75 Figure 6 16 Back block displacement data from high speed cameras Figure 6 17 UF/FDOT surrogate vehicle impact test results 6 5 Compa rison O f Numerical Predictions A nd Experimental R esults When comparing the force deformation data obtained from the crushable nose validation test with both the Kia Rio cur ve from the NCAC test 11005 and the averaged Kia Rio curve (target curve) on Figure 6 18 a very close match can be seen. The test results follow the averaged Kia Rio curve force level throughout the compression of the crushable nose. It is clear from Figure 6 18 that the test results show local force oscillations that represent the normal characteristics of impact tests. The validation test results and the Kia Rio curve can be related in this matter, which makes the crushable nose behavior not only be similar to the actual car curve by the level of force throughout the compression of the surrogate vehicle, but also by having random small oscillations of force throughout impact.

PAGE 76

76 Figure 6 18 Comparison between test results and Kia Rio force deformation curves In Chapter 5 a detailed description of the crushable nose high resolution and simplified FEA models was discussed. These models resulted in close final curves that were expected to represent the behavior of the UF/FDOT surrogate vehicle behavior unde r impact. Figure 6 19 illustrates the comparison plot of the experimental test results with both FEA model results. It can be seen that both models did represent the crushable nose impact characte ristics very accurately considering that the curves introduced in both models were a simplification of the experimental curves gotten from static honeycomb compression tests. Figure 6 19 C omparison between experimental test and models results force deformation curves

PAGE 77

77 CHAPTER 7 HYBRID BREAKAWAY CONNECTION DESIGN 7 1 Introduction Previous a ttempts to produce a breakaway connection made out of steel have result in slip base connections that have a considerate amount of weight which translate to undesirable inertial resistance produce by the connection when activated under impact. Even though these connections work, and the system developed does not depend in clamping bolts, the aim of this research project is to develop a lighter breakaway connection that can preferably activate at a lower force level decelerating vehicles at a lower rate, inc reasing their chances of survival and decreasing their chances of injury. New composite materials have been used in the aerospace industry for decades expanding to other fields due to its strong and light properties. Companies have developed flag poles, li ght poles and power poles made out of Fiber Reinforced Polymers (FRP) stating several advantages over similar products made out of steel and timber. Such advantages include reduced weight, high strength to weight ratio, low maintenance, dimensional stabili ty, high dielectric strength, non toxic handling & higher service life, resistance to rot, corrosion, chemicals, pest damage and non toxic disposal. Properties of these materials can be modified depending on the fibers and resins selected as well as manufa cturing process and fabrication technique. 7 1 1 FRP Pultruded S ection Five FRP octagonal sections (CP076) with a moment capacity in t he proximity of the estimated required value was donated by Creative Pultrusions. Their Powertrusion product line provides very strong FRP poles that have a high potential to work for the development of a hybrid breakaway system. The smallest section from the Powertrusion product line CP076 was

PAGE 78

78 the most fitted section for this project. Figure 7 1 shows the three 8 ft. long and two 3 ft. long sections sent by Creative Pultrusions and a zoomed in view of the cross section of the s. A ) B ) Figure 7 1 Specimen tested. A ) Two 3ft long and three 8ft long sections donated. B ) Cross sectional detail (Courtesy of Matas Groetaers ) 7 1 2 Section Design R equirements Following section 2.2 of this report the calculations for the sign post design requirements were performed. An ear ly estimate was originally produced due to the unknown shape and size of the possible sign posts for the wind load calculation. After the decision to try Creative Pultrusions section CP076, accurate calculations were developed which can be found in Appendi x C a summary of the calculation results can be seen in Table 7 1 Table 7 1 Sign post strength design requirements Properties Value Min. required flexural capacity (wind) 657 kip in Min. required shear capacity (wind) 3.41 kip Min. required axial capacity (gravity) 0.48 kip 7 1 3 Section P roperties Creative Pultrusion specified properties summary can be seen in Table 7 2 The se parameters comply with the design requirements of the worst case scenario multi post sign required to be produced. Original section engineering specifications can be found in Appendix G

PAGE 79

79 of this report. Appendix H includes a detailed description of the cross sectional geometry of the section, with selected properties provided by Creative Pultrusions. Several additional supporting documents are also included as appendices. Table 7 2 FRP octagonal section properties as specified by Creative Pultrusions Properties Value Flexural strength 40 ksi Modulus of elasticity 3800 ksi Moment capacity 720 kip in Moment of inertia 60.87 in 4 Cross sectional area 7.72 in 2 Weight 6.33 lbs/ft 7 2 Static Flexural T esting The main objective of the o ctagonal FRP static flexural test was to measure experimental properties from the section to corroborate the values given by the manufacturer. Factors of safety are expected to be introduced in the moment capacity of the section as private companies have t o specify a safe capacity for their materials to operate. Parameters such as modulus of elasticity, moment of inertia and weight are critical to calculate the rigidity of the sign post and therefore the ability to estimate the proper resistance to wind loa ds. 7 2 1 Test S etup A static three point bending test was carried out at the FDOT structures lab in Tallahassee Florida on August 30, 2013. The size of the specimen tested was 8ft long. Calculations derived from the test data can be found in Appendix D The following instruments were used to record data during testing ; (1) En erpac hydraulic actuator with integrated data collection with a l oad rate of 100 lbs/s and a data c ollection frequency of 10 Hz ; and (2) five laser displacement sensors at section quarter points with data c ollection frequency of 10 Hz

PAGE 80

80 Figure 7 2 illustrates the basic parameters that were followed for the experimental test. A span length of 76 in. was specified with grout plugs that prevent the section to locally buckle as specified. Figure 7 3 shows the octagonal FRP section ready to be tested before the experiment. Figure 7 2 Static three point flexural test sc hematic Figure 7 3 Static three point flexural test setup (Courtesy of David Wagner ) 7 2 2 Results The three point bending test on the FRP octagonal section was carried out successfully. The overall section behavior performed as expected. Figure 7 4 shows the deflection and failure mode of the specimen seen during the test, and a zoomed in photo of the FRP failure mode can

PAGE 81

81 be seen in Figure 7 5 Instrumentation data was produced and recorded properly. Figure 7 6 illustrates the mid span force deflection curve of the specimen tested recorded by the laser sensors. In order to account for the bearing pads deformation under compress ion load, the averaged displacement at the supports was subtracted from the mid span displacement. This corrected deflection data was used to calculate the modulus of elasticity of the section recorded in Table 7 3 The stiffness of the section was assumed to be the slope of the best fitted line taken from the force deformation curve data shown. Only the portion of the curve before the FRP fibers started to fail was used, taken as a conser vative cut off data, up to 40 kips of force applied. From the simply supported mid span deflection formula, knowing the stiffness of the section, the moment of inertia given by the manufacturer and the span length, the modulus of elasticity was calculated. The FRP section properties calculations worksheet can be seen in Appendix D From visual observation and data shown in Figure 7 6 the FRP section resisted the incremental force applied by having a directly proportional deflection. The force deformation curve produced was fairly linear until the FRP fibers started to fail. This produced a lost in the section stiffness and shortly af ter the section failed. The failure was controlled by the outer most fibers having a 45 and 135 degree orientation with respect to the longitudinal direction of the pole. The outer most fiber produces confinement for the internal longitudinal fibers that r esist most of the moment capacity. Once these 45 and 135 degree fibers failed, the confinement was lost together with the integrity of the section. The internal longitudinal fibers of the FRP section opened up without failing and the grout restrain for loc al buckling inserted in the section displaced downwards resulting in a large deflection. After the load was released the section went back to its original deflection showing a completely elastic response due to the internal longitudinal fibers still intact

PAGE 82

82 Figure 7 4 Section deformation during static flexural test (Courtesy of Matas Groetaers ) Figure 7 5 Enlarge d view of FRP section at failure (Courtesy of Matas Groetaers ) Figure 7 6 Static flexural test results

PAGE 83

83 After carrying out the static FRP flexural tests, it was found that the section fl exural strength, modulus of elasticity and moment capacity (Table 7 3 ) exceed the value specified by Creative Pultrusions (Figure 7 2 ). Comparing the capacities of the FRP section (Table 7 3 ) to the design requirements (Table 7 1 ), it can be se en that the FRP section has approximately 30% more flexural capacity than is needed for wind loading. Table 7 3 FRP octagonal section properties by experimental testing Properties Value Fle xural strength 56 ksi Modulus of elasticity 4443 ksi Moment capacity 852.15 kip in Moment of inertia 60.87 in 4 Cross sectional area 7.72 in 2 Weight 6.33 lbs/ft 7 3 Section Fundamental F requency In order to assume that the sign post is rigid enough to resist wind loads, the fundamental natural frequency of the structure needs to be larger than 1 Hz. With the given section value of moment of inertia, the calculated value of modulus of el asticity, and the mass and height of the section, the frequency of the sign post was calculated (Appendix I ). 7 4 Hybrid Breakaway Connection System C oncept The main objective of the first hybrid breakaway connection system developed was to determine whether the octagonal FRP section could be severed (sheared) using metallic cutting surfaces wi thout generating e xcessive vehicle deceleration. Figure 7 7 illustrates an exploded view of the connection in order to identify all the different parts that conforms the breakaway system. These parts can b e seen in further detail on Appendix E where the technical drawings of the first system can be found. The first assembled breakaway connection concept is shown zoomed in Figure 7 8

PAGE 84

84 Figure 7 7 Hybrid breakaway system exploded view

PAGE 85

85 Figure 7 8 Hybrid breakaway system concept

PAGE 86

86 T1 is the octagonal FRP pole specimen tested. Parts T8 and T9 are tubes that have blades machined on them to act as scissors, cutting T1 under a shear impact load. The bottom outer portion of the system (T2 and botto m T8 tubes) keep the FRP pole constrained not allowing it to move under impact, assuming the future foundation is rigid. The top outer portion of the system (T3 and top T8 tubes) will be impacted by the surrogate vehicle transferring half of the shear load directly to the induced failure plane of T1 by the blade of the top T8 tube and the other half to T1 from the contact between the top of T3 and T1. T5 and T7 constrain T1 from local buckling. The direct shear produced by the top T8 tube on the induce fail ure plane should deform T1 displacing it towards the bottom T9 tube which stays inside T1 providing a scissor action that cuts the FRP frontal side (Figure X). The top T9 tube after the frontal part of T1 is cut should be pushed forward getting in contact with the back part of T1, displacing it towards the bottom T8 tube which is constrained, providing a scissor action that cuts the back side of T1. The sides of T1 with respect to the impact direction will not have blades to cut them in the same fashion as described earlier, by decreasing the shear capacity of T1, getting rid of the front and back portions of the section, the sides of T1 should be able to be cut from both top and bottom T8 tubes going all the way through, breaking the FRP cleanly from the im pact made by the Kia Rio surrogate vehicle. The physical system concept before the test can be seen on Figure 7 9 where a closer view of the blades can be seen and an overall view of the finished assembly. Figure 7 9 Hybrid breakaway system in field (Courtesy of G ary Consolazio )

PAGE 87

87 7 5 Impact Shear T est Th e 1100 kg (Kia Rio) crushable nose and impact block (surrogate vehicle), previously validated at the FDOT Structures Research Center, were used to pr oduce the desired impact loads. 7 5 1 Test S etup Figure 7 10 illustrates the basic parameters that were followed for the experimental test. A drop height of 12 ft. was use i n order for the surrogate vehicle to have an initial velocity of 19 mph required also in the structures validation test. The impact height in this test was determine by the fiberglass spacers used and only allowed 9 in. as a minimum clearance distance betw een the center of gravity of the surrogate vehicle and the induced failure plane. Figure 7 11 shows the surrogate vehicle ready to be tested before the experiment. Figure 7 10 Impact shear test setup schematic

PAGE 88

88 Figure 7 11 Impact shear test setup (Courtesy of Matas Groetaers ) 7 5 2 Results The pendulum impact shear test demonstrated that the metallic (steel) cutting surfaces were able to successfully cut through the FRP pipe section thereby breaking the pipe away from the stub base. Frames from high speed video taken during the test are presented in Figure 7 12 and photographs of the condition of the FRP pipe after the test are presented in Figure 7 13 Figure 7 12 High speed camera video frames of breakaway impact shear test (Cou rtesy of Chris Weigly)

PAGE 89

89 Figure 7 13 Specimen post impact shear test (Courtesy of Matas Groetaers ) Although the data acquisition system failed to measure accelerations during the t est, it was possible to use motion tracking software and high speed video to quantify time histories of impact block (surrogate vehicle) displacements, and time varying deformations of the crushable nose. The front block and back block displacements (Figur e 7 14 and 7 16 ) were then differentiated twice to obtain accelerations (Figure 7 15 and 7 17 ) which were then multiplied by the mass of the back block and front block respectively to obtain time varying surrogate vehicle force data (F igure 7 18 ). The force data from both independent structures was added together to produce the total behavior of the Kia Rio surrogate vehicle under impact at 19 mph (Figure 7 18 ). Combining the time varying force (Figure 7 18 ) and deformation data (Figure 7

PAGE 90

90 19 ) a force deformation curve corresponding to the test was produced (Figures 7 20 and 7 21 ). Also presented in Figure 7 21 are force deformation curves for the NCAC Kia Rio test, and an 7 21 that the FDOT crushable nose and impact block behaved in a manner v ery similar to the Kia Rio car and, as such, the block and nose served as a suitable small car surrogate vehicle for evaluating the performance o f the breakaway connection system. Figure 7 14 Back block smooth displacement time plot from high speed cameras Figure 7 15 Back block smooth acceleration time plot

PAGE 91

91 Figure 7 16 Front block smooth displacement time plot from high speed cameras Figure 7 17 Front block smooth acceleration time plot Figure 7 18 Force contribution to the total system by the front and back block structures

PAGE 92

92 Figure 7 19 Crushable nose deformation time plot from high speed cameras Figure 7 20 Forces generated by the surrogate vehicle on the breakaway impact shea r test Figure 7 21 Comparison between crushable nose impact shear test and Kia Rio dynamic behavior

PAGE 93

93 To evaluate the performance of the hybrid breakaway connection system, calculations of occupant impact velocity (OIV) and ridedown accelerations were carried out, as illustrated in Figures 7 22 7 23 and 7 24 The occupant ridedown acceleration limits specified by MASH are: preferred 15 g, and maximum 20.49 g. In the impact shear test (Figure 7 22 ), the maximum value was 17.1 g which is less than the MASH 20.49 g limit and therefore acceptable. For the OIV, MASH specifies a preferred limit of 10 ft/s and a maximum of 16 ft/s. In the test (Figure 7 23 ), the maximum va lue was 16.7 ft/s which is just 4% higher than allowed. In Figure 7 13 it can be noted that while the majority of the FRP pipe section was cleanly severed by the steel cutting surfaces, fibers located along the side walls of the section were not cut, but were instead broken by a combination of tension and flexure (fragments of the broken fibers can be seen extending from the shear plane). Additional section resistance associated with brea king (rather than cutting) these side fibers increased the impact forces and ridedown acceleration levels during the test. Table 7 4 displays a summary of all the values important impact sh ear test results described earlier. Importantly, the fibers in question are located near what would be the neutral axis of the pipe under flexural wind loading conditions. As such, these fibers do not contribute significantly to the flexural capacity of th e section and can be pre cut to improve breakaway performance without adversely affecting wind load capacity. Hence, design modifications to both the FRP section and the steel cutting surfaces in the breakaway connection system have being made to improve t he ridedown acceleration and OIV results and to produce a hybrid breakaway system that complies with MASH specifications.

PAGE 94

94 Figure 7 22 Back block c.g. acceleration time history using 10 (ms ) moving average technique Figure 7 23 Back block c.g. relative velocity time history from trapezoidal integration of acceleration time data Figure 7 24 Back block c.g. relative displacement time history from double trapezoidal integration of acceleration time data

PAGE 95

95 Table 7 4 FRP octagonal section experiment al impact shear test results Maximum impact force 42.042 (kip) Maximum occupant deceleration 17.133 (g) Occupant impact velocity (OIV) 16.704 (ft/s) Occupant displacement (from MASH; 2009) 2 (ft)

PAGE 96

96 CHAPTER 8 SUMMARY, CON CLUSIONS, AND RECOMMENDATIONS 8 1 Introduction In this study, the UF/FDOT crushable nose and a hybrid breakaway connection were developed. For the UF/FDOT crushable nose, a new vehicle impact behavior was replicated due to a modification in the specifications, and several improvements to prior surrogate vehicles were made. For the hybrid breakaway connection, the implementation of an FRP post with the combination of steel blades was sh own to be effective and positive results are highly promising 8 2 UF/FDOT Crushable N ose The most important finding related to the development of the UF/FDOT crushable nose was the high str esses behavior of the deformable material used to reproduce vehicle impact characteristics. By performing laboratory experimental testing on aluminum honeycomb cartridges, high energy absorption capabilities were found which were not described in literatur e. Understanding the physical compression behavior of the material lead to the development of accurate cartridge stiffness calculations. Two FEA models (a high resolution and a simplified model) allowed the acquisition of accurate simulation results that h elped with the design of the validated UF/FDOT crushable nose. The development of cartridges that change cross sectional area throughout their compression (trapezoidal and triangular cartridge geometries) was an important design improvement from previous s urrogate vehicles. This allowed a smooth transition of forces between the different cartridges aligned in series that formed the new surrogate vehicle. Furthermore, it gives the capability of accurate reproduction for any vehicle impact behavior of ascendi ng force needed to be design, avoiding piecewise constant (step function) approximations seen in prior surrogate vehicles.

PAGE 97

97 T he UF/FDOT surrogate vehicle possesses the s ame essential properties as the 2006 Kia Rio vehicle; therefore, it was permitted to be used to test the breakaway connection developed. Large savings in material costs related to impact tests was accomplished having virtually the same impact load of a market car due to the findings and developments achieved during this project. It is therefo re recommended that future developments of surrogate vehicles use the improvements achieved in this study to obtain similar or better returns. 8 3 Hybrid Breakaway C onnection The purpose of the alternative multi post ground sign breakaway connection development was to reduce the weight of the sign posts build in Florida. By achieving this goal, a reduction of inertial forces at impact would be perceived. The use of FRP material was critical i n order to achieve this goal, as well as the implementation of steel blades that enhance the failure of these posts. The combination of these two materials was demonstrated to work, avoiding the failure mechanism of typical FRP posts which do not breakaway decelerating the occupants of a vehicle dramatically. Unfortunately the results of the last test did not satisfy the OIV requirements of the design standard MASH by 4%. It is expected that a modifications to the FRP cross section would decrease the momen t capacity by less than the extra 30% of excess capacity. The impact shear resistance of the system would also be reduced, allowing the OIV requirement to fold into the accepted range by MASH and have a hybrid breakaway connection that possessed performanc e characteristics which are better than could be achieved using solely steel or FRP.

PAGE 98

98 APPENDIX A FABRICATION DRAWINGS FOR IMPACT BLOCK Presented in this appendix are fabrication drawings for the UF/FDOT 1100 kg impa ct block.

PAGE 99

99

PAGE 100

100

PAGE 101

101

PAGE 102

102

PAGE 103

103

PAGE 104

104

PAGE 105

105

PAGE 106

106

PAGE 107

107

PAGE 108

108

PAGE 109

109 APPENDIX B FABRICATION DRAWINGS FOR CRUSHABLE NOSE Presented in this appendix are fabrication drawings for the UF/FDOT crushable nose that attaches to the UF/FDOT 1100 kg impact block

PAGE 110

110

PAGE 111

111

PAGE 112

112

PAGE 113

113

PAGE 114

114

PAGE 115

115 APPENDIX C CALCULATION OF POST STRENGTH REQUIRED FOR WIND LOADING Presented in this appendix are the procedures that were used in this study to calculate the sign post strengths (flexure, shear, axial) needed to resist the combination of code specified equivalent static wind load and self weight (gravity load)

PAGE 116

116

PAGE 117

117

PAGE 118

118

PAGE 119

119

PAGE 120

120

PAGE 121

121

PAGE 122

122 APPENDIX D CALCULATION OF POST FLEXURAL STRENGTH FROM TEST DATA Presented in this appendix are the calculation procedures that were used to convert static flexural test data into a corresponding flexural strength (capacity)

PAGE 123

123

PAGE 124

124

PAGE 125

125 APPENDIX E FABRICATION DRAWINGS FOR H YBRID BREAKAWAY CONNECTION Presented in this appendix are fabrication drawings for the UF/FDOT hybrid breakaway connection system.

PAGE 126

126

PAGE 127

127

PAGE 128

128

PAGE 129

129

PAGE 130

130

PAGE 131

131 APPENDIX F FRP POST STATIC SHEAR TESTING Presented in this appe ndix are the test procedures and results obtained from static shear testing of the unmodified octagonal FRP post section. A static shear test was carried out in the University of Florida, Civil and Coastal Engineering Structures Laboratory on August 13, 2 013. The size of the specimen tested was 3ft long. Technical drawings of the test setup can be found in Appendix C of this document. The following instruments were used to record data during testing Tinius Olsen testing machine with integrated data collec tion o Position rate of 0.5 in/s o Collection data frequency of 2.7 Hz

PAGE 132

132 Figure F 1 Schematic of static double shear test setup

PAGE 133

133 Figure F 2 Static double shear test setup Figure F 3 Section deformation during static double shear test

PAGE 134

134 Figure F 4 Static double shear test results

PAGE 135

135

PAGE 136

136

PAGE 137

137

PAGE 138

138

PAGE 139

139 APPENDIX G MANUFACTURER SPECIFIED PROPERTIES OF FRP POST Presented in this appendix are design properties (material prop erties, section properties, etc.) of the FRP post section as specified by the product manufacturer (Creative Pultrusions).

PAGE 140

140

PAGE 141

141 APPENDIX H GEOMETRIC DI MENSIONS OF FRP POST Presented in this appendix are geometric d imensions of the FRP post section as specified by the product manufacturer (Creative Pultrusions).

PAGE 142

142

PAGE 143

143 APPENDIX I CALCULATION OF FRP POST FUNDAMENTAL FREQUENCY Presented in this appendix is the calculation of fund amental frequency of the proposed sign post system 2001).

PAGE 144

144

PAGE 145

145 LIST OF R EFERENCES AASHTO (American Association of State Highway and Transportation Officials) (2009), Manual for Assessin g Safety Hardware (MASH) 2009, AASHTO Washington, D.C AASHTO (American Association of State Highway and Transportation Officials) (2001), Standard Specification for Structural Supports for Highway Signs, Luminaires and Traffic Signals AASHTO, Washington D.C Bielenberg, R., Lechtenberg, K., Faller, R., and Sicking, D. (2009). Validation of the Valmont/MwRSF Pendulum with Crushable Nose Res earch Report No. TRP 03 214 09, University of Nebraska Lincoln Lincoln, NE. Consolazio, G., Bui, L., and Walters, R. (2012). Pendulum impact testing of an impact breakaway, wind resistant base connection for multi post ground signs Structures Research Report No. 2012/92174, University of Florida, Gainesville, FL. Engineering Systems Division ( 1990 ). Crush Characteris tics of the 'Breakaway' Bogie Report No. FHWA RD 89 107. The Scientex Corporation Washington, DC. FDOT (Florida Department of Transportation) (2010), FDOT Modifications to Standard Specification for Structural Supports for Highway Signs, Luminaires and Traffic Signals (LTS 5) FDOT Structures Manual, Vol. 9. LSTC (Livermore Software Technology Corporation) (200 7 ). LS Manual LSTC, Livermore CA. Marzougui, D,. Story, C., Nix, L., Kan, C., and Arispe, E. (2011). Crash Testing of Kia Ri o Sedans Into an Instrumented Rigid Pole Report No. 2011 R 005 The George Washington University, Ashburn, VA. Ross, H.E. Jr., Sicking, D.L., Zimmer, R.A., and Michie, J.D. (1993), NCHRP Report 350: Recommended Procedures for the Safety Performance Evalua tion of Highway Features, Transportation Research Board (TRB), National Research Council, Washington, D C.

PAGE 146

146 BIOGRAPHICAL SKETCH The author was born in Concepci n, Chile, in 1989 He lived in Chile until the age of 16 In August 2005 he attended the United World College of the Atlantic where he completed the International Baccalaureate Diploma Programme. In August 2007, he started his career at the University of Florida, where he received the degree of Bachelor of Sci ence in Civil Engineering in December 201 1 He then enrolled in graduate school at the University of Florida where he received a Master of Engineering in Civil Engineering in May 2014, with a n emphasis in civil structures.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EFE2UM3MV_AW2NHN INGEST_TIME 2014-10-03T23:01:14Z PACKAGE UFE0046720_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES