A MASH Compliant W-Beam Median Guardrail Systemdocs.trb.org/prp/14-5558.pdf · 2 The W-beam median guardrail system is a common guardrail system used by many states as a ... The simulation

1

A MASH Compliant W-Beam Median Guardrail System 1 2

By A. Y. Abu-Odeh, R. P. Bligh, W. Odell, A. Meza, and W. L. Menges 3 Submitted: July 30, 2013 4 Word Count: 2,624 + (14 figures + 2 tables=4,000) =6,624 words 5

6 Authors: 7 Akram Y. Abu-Odeh (Corresponding Author) 8 Research Scientist 9 Texas A&M Transportation Institute 10 College Station, Texas 77843 11 O: (979) 862-3379 12 F: (979) 845-6107 13 Email: [email protected] 14 15 Roger P. Bligh 16 Research Engineer 17 Texas A&M Transportation Institute 18 College Station, Texas 77843 19 O: (979) 845-4377 20 F: (979) 845-6107 21 Email: [email protected] 22 23 Wade Odell, P.E. 24 Research Engineer 25 Safety & Operations 26 Structures & Hydraulics 27 Texas Department of Transportation 28 Research & Technology Implementation Office 29 P.O. Box 5080 30 Austin, Texas 78763-5080 31 O: (512) 416-4737 32 F: (512) 416-4734 33 Email: [email protected] 34 35 Rory Meza 36 Director, Roadway Design Section 37 Texas Department of Transportation 38 Director, Roadway Design Section 39 P.O. Box 5080 40 Austin, Texas 78763-5080 41 Email: [email protected] 42 O: (512) 416-2678 43 F: (512) 416-2686 44 45 Wanda L. Menges 46 Research Specialist 47 TTI Proving Ground 48 Texas A&M Transportation Institute 49 College Station, Texas 77843 50 O: (979) 845-6157 51 F: (979) 845-6107 52

TRB 2014 Annual Meeting Original paper submittal - not revised by author

Abu-Odeh, et al . 2

Abstract 1 The W-beam median guardrail system is a common guardrail system used by many states as a 2 median barrier in ditches and other median configurations. Unfortunately, the 27-inch design did 3 not pass MASH evaluation criteria in the crash testing program of NCHRP project 22-14(03). In 4 this paper, the researchers modeled and simulated MASH test 3-11 of the 27-inch W-beam 5 median barrier and used the model in the evaluation of a new design of the W-beam median 6 barrier. The simulation indicated that the new 31-inch W-beam median barrier would pass 7 MASH evaluation criteria. Subsequently, MASH tests 3-10 and 3-11 were performed to evaluate 8 the new design. Both tests passed MASH evaluation criteria. Hence, a new MASH complaint W-9 beam median barrier system is available for TxDOT and other state DOT’s to utilize as a median 10 guardrail system. 11

12 Keywords: MASH, Guardrails, Median Barrier, Simulation. 13


Abu-Odeh, et al . 3

INTRODUCTION 1 In NCHRP project 22-14(03) (1), Texas A&M Transportation Institute (TTI) researchers 2 performed crash tests for several non-proprietary hardware. One of the systems tested under that 3 project was the G4(1S) median barrier. 4

This system presents a unique performance challenge due to the additional constraint of the 5 posts imposed by the double-sided G4(1S) W-beam median barrier. The added post constraint 6 delays the release of the post from the rail, which results in vehicle climb and vaulting due to a 7 localized drop in rail height. Hence, there was a need to re-design the W-beam median guardrail 8 system to pass MASH (Manual for Assessing Safety Hardware (2)), evaluation criteria. 9

BACKGROUND 10 The G4(1S) W-beam median barrier (American Association of State Highway and 11 Transportation Officials (AASHTO) designation SGM04a with non-steel blocks) is a 27-inch 12 tall, strong steel post, W-beam median barrier. In NCHRP project 22-14(03), the median barrier 13 was constructed using 12-gauge W-beam guardrail elements attached to 6 ft long W6×8.5 steel 14 posts spaced 6 ft-3 inch on center. The W-beam guardrail elements were offset from the posts 15 using non-steel blockouts nominally 6 inch × 8 inch × 14 inch long. 16

The height of the G4(1S) W-beam median barrier test installation was 27 inches. The 17 length-of-need for the installation was 100 ft. The front (impacted) rail was constructed with 37 18 ft-6 inch long terminals on each end and the rear rail was constructed with 50 ft long terminals 19 on each end. The total overall test installation length was 200 ft. 20

A cross section of the G4(1S) W-beam median barrier is shown in Figure 1. 21 22

23 24 25

FIGURE 1 Cross-section of W-beam median barrier. 26 27 Two tests were performed under NCHRP Project 22-14(03) (1),, MASH test 3-10 and 28

MASH test 3-11. MASH test 3-10 involves a 1100C vehicle with test inertial mass of 2420 lb 29 impacting the median barrier at an impact speed of 62.2 mi/h and at an angle of 25 degrees. 30 MASH test 3-11 involves a 2270P vehicle weighing 5000 lb and impacting the median barrier at 31 an impact speed of 62.2 mi/h and at an angle of 25 degrees. The W-beam median barrier was 32 able to contain and redirect the 1100C vehicle in MASH test 3-10. However, it did not contain or 33 redirect the 2270P vehicle in MASH test 3-11. During TTI test 476460-1-9(1),, the left front tire 34


Abu-Odeh, et al . 4

and wheel of the 2270P vehicle rode up on post 14. Then, the front of the vehicle became 1 airborne above the median barrier, and the vehicle lost contact with the barrier as it was airborne 2 over the median barrier. This vaulting behavior is shown in Figure 2. 3

4

5 FIGURE 2 The 2270P vehicle vaulting the 27-inch W-beam median barrier during 6

MASH 3-11 test. 7

RESEARCH APPROACH 8 9

To improve the performance of the 27-in high median W-beam, TTI researchers analyzed the 10 failed test and incorporated design changes that have the potential of rectifying the performance 11 of the median W-beam barrier. First, the research team developed a detailed finite element model 12 of the W-beam median rail to calibrate the model under the MASH test already conducted. The 13 new Silverado vehicle model developed by the National Crash Analysis Center (NCAC, (3)) was 14 used to simulate the MASH 2270P test vehicle. 15

In the model, the post comprised of different thicknesses to accurately represent the shape 16 of a W6×9 steel post. A total of 18,240 shell elements were used for modeling the posts. 17 Additionally, the W-beam model contains a more refined element mesh than the previously used 18 W-beam models, so it can capture deformation more realistically. A total of 182,304 shell 19 elements were used for modeling the W-beam segments (4). Both the post and the W-beam 20 models are shown in Figure 3. The end terminals and the remaining portion of the length-of-21 need rail were represented by spring elements connected to each end of the modeled W-beam. 22 These springs elements have a combined stiffness representative of typical end terminals. 23

24

25


Abu-Odeh, et al . 5

FIGURE 3 Meshing scheme of the 8-ft Post Model (left) and the 12 gauge W-beam rail 1 (right). 2

The vehicle model used for simulation was the Chevrolet Silverado model, which was 3 developed by NCAC. This vehicle model represents the MASH 2270P test vehicle. The finite 4 element model for the MASH 1100C test vehicle was not available at the time when this research 5 was performed. The vehicle and 27-inch median W-beam barrier models are shown in Figure 4. 6

7 8

FIGURE 4 Finite Element model of the 27-inch median barrier. 9 10 The research team started by simulating the failed test using LS-DYNA (5) finite element 11

code. The vaulting phenomena of the vehicle captured in the simulation as shown in Figure 5 12 below. Hence, the model is considered corroborated with the failed MASH test 3-11 and can be 13 used as a tool to investigate the system performance once modified. 14


Abu-Odeh, et al . 6

1 FIGURE 5 Simulation of the MASH 3-11 where the 2270P vehicle vaults over the 2

27-inch median barrier. 3 Design modifications included increasing the rail height from 27 inches to 31 inches and 4

moving the splice location from at post to mid-span. The cross-section views of the new system 5 design and the model are shown in Figure 6. 6

7 FIGURE 6 Model of the 31-inch median W-beam guardrail with the 2270P vehicle 8

model. 9

SIMULATION RESULTS 10 Two simulations were conducted using LS-DYNA (5) finite element code. One was conducted 11 with vehicular impact at post and the other with vehicular impat at mid-span. 12

1. TL 3-11 Mid-Span Impact 13 14

In the first simulation case, the analysis represents vehicular impact at mid-span of the guardrail. 15 The modified barrier system was impacted by 2270P vehicle model at 62.2 mi/h and an angle of 16 25 degrees. The occupant risk assessments for this model are provided in 17 TABLE 1, and vehicle behavior is shown in Figure 7. 18


Abu-Odeh, et al . 7

TABLE 1 TRAP Output Summary for Mid-Span Impact Case 1

Occupant Risk Factors Impact Velocity (m/s) at 0.1695 sec on left side of interior x-direction: 6.7

Rec: <9 m/s

y-direction: -4.7 Max: <12 m/s THIV (km/hr): 26.9 at 0.1633 sec on left side of interior

THIV (m/s): 7.5

Ridedown Acceleration (G's) x-direction: -10.6 (0.2471 - 0.2571 sec) Rec: <15 G's y-direction: 9.9 (0.2959 - 0.3059 sec) Max: <20 G's PHD (G's): 11.8 (0.2471 - 0.2571 sec)

ASI: 0.78 (0.1477 - 0.1977 sec) Maximum 50 msec Moving Average Acceleration (G's) x-direction: -7.2 (0.1432 - 0.1932 sec) y-direction: 4.9 (0.1883 - 0.2383 sec) z-direction: 2.3 (0.4747 - 0.5247 sec)

2 FIGURE 7 Views of vehicle behavior in the at mid-span impact simulation case. 3

2. TL 3-11 At-Post Impact 4 The modified system wassimulated under impact by the 2270P test at a post location instead 5 of the mid-span using the same MASH TL 3-11 initial conditions of 62.2 mi/h and 25 6 degrees. The occupant risk assessment for this model is provided in TABLE 2 TRAP 7 Output Summaryand vehicle behavior is shown in Figure 8. 8

TABLE 2 TRAP Output Summary for At-Post Case 9

Occupant Risk Factors Impact Velocity (m/s) at 0.1738 sec on left side of interior x-direction: 6.1

Rec: <9 m/s

y-direction: -4.7 Max: <12 m/s THIV (km/hr): 27.1 at 0.1676 sec on left side of interior

THIV (m/s): 7.5

Ride down Acceleration (G's) x-direction: -9.5 (0.2046 - 0.2146 sec) Rec: <15 G's y-direction: 9.7 (0.2896 - 0.2996 sec) Max: <20 G's


Abu-Odeh, et al . 8

PHD (G's): 11.8 (0.2046 - 0.2146 sec)

ASI: 0.68 (0.1134 - 0.1634 sec) Maximum 50 msec Moving Average Acceleration (G's) x-direction: -6.4 (0.1128 - 0.1628 sec) y-direction: 5.2 (0.2499 - 0.2999 sec) z-direction: 3.1 (0.3532 - 0.4032 sec)

1

2 FIGURE 8 Views of vehicle behavior in the at post impact simulation case. 3

4 Both simulation cases indicated that the 31-inch W-beam median barrier is able to contain 5

and redirect the test vehicle, and able to pass MASH evaluation criteria as presented in 6 TABLE 1 and TABLE 2. Hence, the research team used the new design for the full-scale 7

crash testing phase of the project. 8

FULL SCALE CRASH TESTS 9

Test Article Design and Construction 10 11 The constructed system is a 31˗inch tall, strong steel post, W-beam median barrier. This 12 (TxDOT) median barrier was constructed using 12˗gauge W-beam guardrails attached to 6 ft 13 long W6×8.5 steel posts spaced 6 ft-3 inch on center. The W˗beam guardrails are offset from the 14 posts using wood blockouts nominally 6 inch × 8 inch × 14 inch long. For this installation, the 15 W-beam rail element joints were moved off the posts and centered mid-span between posts. 16

FIGURE 9 depicts the cross-section of the TxDOT 31-inch W˗Beam Median Barrier. 17 Photographs of the completed installation are shown in Figure 10. 18 19


Abu-Odeh, et al . 9

1 FIGURE 9 Cross-section of the TxDOT 31-inch W-beam median barrier. 2

3

4 FIGURE 10 TxDOT 31-inch W-beam median barrier installation. 5

6

MASH Test 3-10 7 8

MASH test 3-10 involves an 1100C vehicle weighing 2420 lb and impacting the TxDOT 31-inch 9 W-Beam Median Barrier at an impact speed of 62.2 mi/h and an angle of 25 degrees. The target 10 impact point was 21 inches upstream of post 12. The 2006 Kia Rio used in the test weighed 11 2444 lb, and the actual impact speed and angle were 62.2 mi/h and 25.0 degrees, respectively. 12 The actual impact point was 22 inches upstream of post 12. 13

The front bumper, hood, radiator and support, left front strut and tower, left front tire and 14 wheel rim, left front fender, and left front door were damaged. The hood was pushed into the 15 windshield which shattered the lower portion of the windshield. Maximum exterior crush to the 16 1100C vehicle was 13.0 inches in the side plane at the left front corner just above bumper height. 17 No occupant compartment deformation was noted. Damage to the guardrail and the vehicle is 18 shown in Figure 11. 19

20


Abu-Odeh, et al . 10

1 FIGURE 11 Vehicle and guardrail damage after MASH test 3-10. 2

3 Data from the accelerometer, located at the vehicle’s center of gravity, were digitized for 4

evaluation of occupant risk. In the longitudinal direction, the occupant impact velocity was 5 20.0 ft/s at 0.115 s, the highest 0.010-s occupant ridedown acceleration was 9.6 Gs from 0.172 to 6 0.182 s, and the maximum 0.050-s average acceleration was -7.3 Gs between 0.076 and 0.126 s. 7 In the lateral direction, the occupant impact velocity was 17.4 ft/s at 0.115 s, the highest 0.010-s 8 occupant ridedown acceleration was 8.3 Gs from 0.155 to 0.165 s, and the maximum 0.050-s 9 average was 6.9 Gs between 0.038 and 0.088s. The guardrail was able to contain and redirect 10 the vehicle. These data and other pertinent information are provided in Figure 12. 11


Abu-O

deh, et al .

11

0.000 s 0.140 s 0.280 s 0.420 s

FIGURE 12 Summary of results for MASH test 3-10 on the TxDOT 31-inch W-Beam Median Barrier.

General Information Test Agency ........................ Test Standard Test No. ....... TTI Test No. ....................... Test Date ............................Test Article Type .................................... Name .................................. Installation Length ............... Material or Key Elements .... Soil Type and Condition ......Test Vehicle Type/Designation ................ Make and Model ................. Curb .................................... Test Inertial ......................... Dummy ............................... Gross Static ........................

Texas Transportation Institute (TTI) MASH Test 3-10 490023-3 2013-06-18 Median Barrier TxDOT 31-inch W-Beam Median Barrier156 ft 12˗gauge W-beam at 31 inch height on 6 ft long W6×8.5 steel posts spaced 6 ft-3 inch on center with 6 inch × 8 inch × 14 inch routed wood blockouts with splice midspan between posts Standard soil, dry 1100C 2006 Kia Rio 2494 lb 2444 lb 180 lb 2624 lb

Impact Conditions Speed ................................. Angle................................... Location/Orientation ............ Impact Severity ...................Exit Conditions Speed ................................. Angle...................................Occupant Risk Values Impact Velocity Longitudinal ..................... Lateral .............................. Ridedown Accelerations Longitudinal ..................... Lateral .............................. THIV.................................... PHD .................................... ASI ......................................Max. 0.050-s Average Longitudinal ..................... Lateral .............................. Vertical .............................

62.2 mi/h 25.0 degrees 27 inches upstrm of post 12 56.5 kip-ft Not obtainable Not obtainable 20.0 ft/s 17.4 ft/s 9.6 G 8.3 G 7.8 m/s 11.0 G 0.94 -7.3 G 6.9 G -3.0 G

Post-Impact Trajectory Stopping Distance ..................... Vehicle Stability Maximum Yaw Angle ................. Maximum Pitch Angle ................ Maximum Roll Angle ................. Vehicle Snagging ...................... Vehicle Pocketing ......................Test Article Deflections Dynamic .................................... Permanent ................................ Working Width ........................... Vehicle Intrusion ........................Vehicle Damage VDS .......................................... CDC .......................................... Max. Exterior Deformation ......... OCDI ......................................... Max. Occupant Compartment Deformation ........................

132.5 ft dwnstrm Aligned w/cntr rail 36 degrees 9 degrees 11 degrees No Yes/No 25.4 inches 20.25 inches 33.6 inches 21.5 inches 11KFQ5 11FLEW 4 13.0 inches LF0000000 None



MASH Test 3-11 1 MASH test 3-11 involves a 2270P vehicle weighing 5000 lb and impacting the TxDOT 31-inch 2 W-Beam Median Barrier at an impact speed of 62.2 mi/h and at an angle of 25 degrees. The 3 target impact point was 10.5 ft upstream of post 13. The 2007 Dodge Ram 1500 pickup truck 4 used in the test weighed 5017 lb, and the actual impact speed and angle were 63.0 mi/h and 25.4 5 degrees, respectively. The actual impact point was 10.8 ft upstream of post 13. 6 The left front tie rod, left front lower A˗arm, left front frame rail, and left front hub 7 assembly were deformed. Also damaged were the front bumper, grill, left front tire and wheel 8 rim, left front fender, left front door, left rear door, left rear exterior bed, left rear tire, and rear 9 bumper. Maximum exterior crush to the vehicle was 11.0 inches in the side plane at the left front 10 corner just above bumper height. No occupant compartment deformation occurred. Damage to 11 the guardrail and the vehicle is shown in Figure 13. 12 13

14 FIGURE 13 Vehicle and guardrail damage after MASH test 3-11. 15

16 Data from the accelerometer, located at the vehicle center of gravity, were digitized for 17 evaluation of occupant risk. In the longitudinal direction, the occupant impact velocity was 18 19.0 ft/s at 0.151 s, the highest 0.010-s occupant ridedown acceleration was 10.2 Gs from 0.162 19 to 0.172 s, and the maximum 0.050-s average acceleration was -6.2 Gs between 0.050 and 20 0.100 s. In the lateral direction, the occupant impact velocity was 15.1 ft/s at 0.151 s, the highest 21 0.010-s occupant ridedown acceleration was 6.9 Gs from 0.192 to 0.202 s, and the maximum 22 0.050-s average was 4.5 Gs between 0.272 and 0.322 s. These data and other pertinent 23 information from the test are provided in Figure 14. 24


Abu-O

deh, et al . 13

0.000 s 0.200 s 0.400 s 0.700 s

FIGURE 14 Summary of results for MASH test 3-11 on the TxDOT 31-inch W-Beam Median Barrier.

General Information Test Agency ........................ Test Standard Test No. ....... TTI Test No. ....................... Test Date ............................Test Article Type .................................... Name .................................. Installation Length ............... Material or Key Elements .... Soil Type and Condition ......Test Vehicle Type/Designation ................ Make and Model ................. Curb .................................... Test Inertial ......................... Dummy ............................... Gross Static ........................

Texas Transportation Institute (TTI) MASH Test 3-11 490023-4 2013-06-21 Median Barrier TxDOT 31-inch W-Beam Median Barrier 156 ft 12˗gauge W-beam at 31 inch height on W6×8.5 steel posts spaced 6 ft-3 inch with 6 inch × 8 inch × 14 inch routed wood blockouts with splice midspan between posts Standard soil, dry 2270P 2007 Dodge Ram 1500 Pickup 4880 lb 5017 lb No dummy 5017 lb

Impact Conditions Speed ............................... Angle ................................. Location/Orientation .......... Impact Severity .................Exit Conditions Speed ............................... Angle .................................Occupant Risk Values Impact Velocity Longitudinal.................... Lateral ............................ Ridedown Accelerations Longitudinal.................... Lateral ............................ THIV .................................. PHD .................................. ASI ....................................Max. 0.050-s Average Longitudinal.................... Lateral ............................ Vertical ...........................

63.0 mi/h 25.4 degrees 10.8 ft upstream of post 13 122.5 kip-ft 19.0 m/s 15.1 m/s 10.2 G 6.9 G 7.0 m/s 10.2 G 0.65 -6.2 G 4.5 G 3.4 G

Post-Impact Trajectory Stopping Distance ...................... Vehicle Stability Maximum Yaw Angle ................. Maximum Pitch Angle ................ Maximum Roll Angle .................. Vehicle Snagging ....................... Vehicle Pocketing.......................Test Article Deflections Dynamic ..................................... Permanent ................................. Working Width ............................ Vehicle Intrusion ........................Vehicle Damage VDS ........................................... CDC ........................................... Max. Exterior Deformation .......... OCDI .......................................... Max. Occupant Compartment Deformation .........................

119.5 ft dwnstrm 47 degrees 11 degrees 12 degrees No No 39.0 inches 29.5 inches 55.0 inches 50.3 inches 11LFQ3 11FLEW3 11.0 inches LF0000000 None



RESULTS AND CONCLUSIONS 1 2 A new MASH compliant W-beam median barrier system was developed and tested successfully 3 using MASH TL-3 conditions. Nonlinear finite element simulation was used to evaluate the 4 system performance prior to testing. The modified median barrier system passed the evaluation 5 criteria for both MASH tests 3-10 and 3-11. This modified system utilizes standard blockouts 6 and standard post spacing. The system is 31 inches tall and has splices positioned off post 7 locations. By having a MASH compliant W-beam median barrier, TxDOT and other states 8 DOT’s will have the option of specifying this new design for implementation. 9

ACKNOWLEDGEMENTS 10 11

This research project was conducted under a cooperative program between the Texas 12 Transportation Institute, the Texas Department of Transportation, and the Federal Highway 13 Administration. The authors acknowledge and appreciate TxDOT guidance and assistance. 14

REFERENCES 15 1. D. Lance Bullard, Jr., Roger P. Bligh, Wanda L. Menges, and Rebecca R. Haug, “Evaluation 16

of Existing Roadside Safety Hardware Using Updated Criteria – Technical Report,” 17 National Highway Research Cooperative Program, NCHRP Web-Only Document 157: 18 Project 22-14(03), Transportation Research Board, March 2010. 19

2. AASHTO. Manual for Assessing Safety Hardware. American Association of State Highway 20 and Transportation Official, Washington DC, 2009. 21

3. National Crash Analysis Center. NCAC Finite Element Model Archive. 22 http://www.ncac.gwu.edu/vml/models.html. Accessed: Dec 10, 2008 23

4. A. Y. Abu-Odeh, R. P. Bligh, D. L. Bullard, and W. L. Menges. Crash testing and 24 Evaluation of the modified G4(1S) W-Beam Guardrail on 2:1 Slope. TTI 405160-4-1. Texas 25 A&M Transportation Institute, College Station, TX, 2008. 26

5. J. O. Hallquist. LS-DYNA: Keyword User's Manual, Version 971, Livermore Software 27 Technology Corporation (LSTC), Livermore, California, 2007. 28 29


Documents

A MASH Compliant W-Beam Median Guardrail Systemdocs.trb.org/prp/14-5558.pdf · 2 The W-beam median guardrail system is a common guardrail system used by many states as a ... The simulation