National Academies Press: OpenBook
« Previous: II. State of the Practice
Page 12
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 12
Page 13
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 13
Page 14
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 14
Page 15
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 15
Page 16
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 16
Page 17
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 17
Page 18
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 18
Page 19
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 19
Page 20
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 20
Page 21
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 21
Page 22
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 22
Page 23
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 23
Page 24
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 24
Page 25
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 25
Page 26
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 26
Page 27
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 27
Page 28
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 28
Page 29
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 29
Page 30
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 30
Page 31
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 31
Page 32
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 32
Page 33
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 33
Page 34
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 34
Page 35
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 35
Page 36
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 36
Page 37
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 37
Page 38
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 38
Page 39
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 39
Page 40
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 40
Page 41
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 41
Page 42
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 42
Page 43
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 43
Page 44
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 44
Page 45
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 45
Page 46
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 46
Page 47
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 47
Page 48
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 48
Page 49
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 49
Page 50
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 50
Page 51
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 51
Page 52
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 52
Page 53
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 53
Page 54
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 54
Page 55
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 55
Page 56
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 56
Page 57
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 57
Page 58
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 58
Page 59
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 59
Page 60
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 60
Page 61
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 61
Page 62
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 62
Page 63
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 63
Page 64
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 64
Page 65
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 65
Page 66
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 66
Page 67
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 67
Page 68
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 68
Page 69
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 69
Page 70
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 70
Page 71
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 71
Page 72
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 72
Page 73
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 73
Page 74
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 74
Page 75
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 75
Page 76
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 76
Page 77
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 77
Page 78
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 78
Page 79
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 79
Page 80
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 80
Page 81
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 81
Page 82
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 82
Page 83
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 83
Page 84
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 84
Page 85
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 85
Page 86
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 86
Page 87
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 87
Page 88
Suggested Citation:"III. Assessment of Roadside Safety Hardware ." National Academies of Sciences, Engineering, and Medicine. 2010. Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report. Washington, DC: The National Academies Press. doi: 10.17226/22938.
×
Page 88

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

12 III. ASSESSMENT OF ROADSIDE SAFETY HARDWARE A limited number of full-scale crash tests were performed under NCHRP Project 22-14(02) to help understand and evaluate the consequences of adopting the recommended changes on current hardware. A summary of these tests is presented in Table 3. It should be noted that several of the tests listed in Table 3 involve a 5000-lb, 3/4-ton, standard cab pickup. This vehicle was initially selected as the new design vehicle for MASH. The heavy design test vehicle was later changed to a 5000-lb, 1/2-ton, 4-door pickup to be more representative of large sport-utility vehicles (SUVs) in terms of center-of-gravity (C.G.) height and body torsional stiffness. Several barrier systems that had previously been tested with the 3/4-ton, standard cab pickup were retested with the 1/2-ton, 4-door pickup. In the subsequent sections of this chapter, the results of these and other tests performed to date in accordance with MASH are used in combination with engineering analysis and engineering judgment to provide an initial assessment of the ability of other non-proprietary roadside safety hardware to comply with MASH. This initial evaluation is intended to help prioritize future research and testing needs to demonstrate compliance of these devices with MASH and to provide information that would assist understanding of the implications of adopting MASH as it progressed through the AASHTO review and publication process. For ease of reference, the review is divided by category or application of roadside safety hardware (e.g., guardrail, median barrier, transitions, etc.). GENERAL PERFORMANCE CONSIDERATIONS The criteria used to assess the impact performance of roadside safety hardware in regard to MASH are those recommended for evaluation of full-scale crash tests under both NCHRP Report 350 and MASH. The assessment of a given device may include various qualitative and quantitative factors depending on the nature of the device and the availability of data. Experience testing under NCHRP Report 350 has identified three primary concerns or modes of failure: structural adequacy, vehicle stability, and occupant risk. The evaluation criteria for occupant impact velocity and occupant ridedown accelerations remain consistent with NCHRP Report 350 and will not be addressed herein. However, occupant risk in the form of occupant compartment deformation has changed and will be addressed. Discussion of these three evaluation criteria will be helpful prior to assessing individual roadside safety devices.

13 Table 3. Summary of Crash Tests Conducted under NCHRP Project 22-14(02). Ref. Test No.* Agency Test No. Test Designation Test Article Vehicle Make and Model Vehicle Mass (lb) Impact Speed (mi/h) Impact Angle (deg) Pass/Fail 1 2214WB-1 3-11 Modified G4(1S) Guardrail 2002 GMC 2500 3/4-ton Pickup 5000 61.1 25.6 Pass 2 2214WB-2 3-11 Modified G4(1S) Guardrail 2002 Dodge Ram 1500 Quad Cab Pickup 5000 62.4 26.0 Pass 3 2214MG-1 3-11 Midwest Guardrail System (MGS) 2002 GMC 2500 3/4-ton Pickup 5000 62.6 25.2 Pass 4 2214MG-2 3-11 MGS 2002 Dodge Ram 1500 Quad Cab Pickup 5000 62.8 25.5 Pass 5 2214MG-3 3-10 MGS (Max. Height) 2002 Kia Rio 2588 60.8 25.4 Pass 6 2214TB-1 3-11 Free-Standing Temporary F-Shape Barrier 2002 GMC 2500 3/4-ton Pickup 5000 61.8 25.7 Pass 7 2214TB-2 3-11 Free-Standing Temporary F-Shape Barrier 2002 Dodge Ram 1500 Quad Cab Pickup 5000 61.9 25.4 Pass 8 2214NJ-1 3-10 32-inch Permanent New Jersey Safety Shape Barrier 2002 Kia Rio 2579 60.8 26.1 Pass 9 2214T-1 3-21 Guardrail to Concrete Barrier Transition 2002 Chevrolet C1500HD Crew Cab Pickup 5083 60.3 24.8 Pass 10 2214TT-1 3-34 Sequential Kinking Terminal (SKT)- MGS (Tangent) 2002 Kia Rio 2597 64.4 14.5 Pass 11 2214NJ-2 4-12 32-inch Permanent New Jersey Safety Shape Barrier 1989 Ford F-800 22,045 56.5 16.2 Fail1 * For reference purposes within this report 1 Truck rolled over rail

14 Structural Adequacy In regard to longitudinal barrier impacts, structural adequacy is evaluated with respect to a barrier’s ability to contain the impacting vehicle and either redirect it or capture it and bring it to a controlled stop. The vehicle is not permitted to penetrate, underride, or override the barrier although controlled lateral deflection is acceptable. Structural adequacy of a barrier is often equated to its ultimate strength or capacity to resist lateral impact forces. Engineering analyses based on yield line theory or plastic design procedures can be used to compute the load capacity of rigid or semi-rigid barriers (e.g. bridge rails and concrete median barriers). Figure 1 illustrates such a yield line failure analysis procedure for a vertical concrete parapet. Structural adequacy can then be assessed by comparing the capacity of a barrier to a design force corresponding to a desired test or performance level. Figure 1. Yield Line Failure Analysis for Concrete Parapet(9).

15 Data from two instrumented wall studies(32, 33) were used to derive barrier design loads for various test or performance levels included in the AASHTO LRFD Bridge Design Specifications: Section 13 – Railings. The test levels correspond to those contained in NCHRP Report 350. In these research studies, instrumented concrete walls were designed to measure the magnitude and location of vehicle impact forces. In this first study(32), eight full-scale crash tests were conducted using various sizes of passenger cars and buses. The wall consisted of four 10-ft long panels laterally supported by four load cells. Each of the 42-inch tall x 24-inch thick panels was also instrumented with an accelerometer to account for inertia effects. Surfaces in contact with the supporting foundation and adjacent panels were Teflon coated to minimize friction. In the second such study(33), a new wall with a height of 90 inches was constructed using similar design details; crash tests with a variety of trucks (up to and including an 80,000-lb tractor with tank-type trailer) were conducted. Speeds in these tests ranged from 50 mi/h to 60 mi/h, and the impact angles ranged from 15 degrees to 25 degrees. The design load calculated for both TL-3 and TL-4 is 54 kips. Note that this design force is derived from an impact with a nearly rigid instrumented wall barrier and, therefore, is considered to represent the upper bound of forces that would be expected on actual barriers. The design loads established for TL-5 and TL-6, which include consideration of 80,000-lb tractor trailers, are 124 kips and 175 kips, respectively. During the course of the instrumented wall work, the researchers derived relationships that use a measured lateral impact force resulting from a vehicle-barrier collision to estimate the impact force associated with a collision involving a different vehicle and/or impact conditions. The relationship is given as:                                   = 1 2 1 2 2 1 1 2 2 1 2 12 sin sin W W K K L L V VFF θ θ Where: F = impact force, V = impact velocity, θ = impact angle L = longitudinal distance from front of vehicle to C.G. K = barrier contact area or stiffness W = vehicle weight Using 54 kips as the design impact force for NCHRP Report 350 test 3-11, the impact force corresponding to MASH test 3-11 with the 1/2-ton, 4-door pickup truck can be estimated. The impact speed and angle used in MASH test 3-11 are the same as those prescribed under NCHRP Report 350 and, therefore, will not influence the impact force. Assuming the contact area associated with impacts by both pickup trucks is essentially the same for a given longitudinal barrier system, the change in impact force becomes a function of vehicle weight and

16 vehicle length. Using measured vehicle lengths of test vehicles (from the front bumper to the C.G.) and the nominal vehicle weights specified for the respective pickup trucks, the impact force associated with MASH test 3-11 can be estimated as follows:               = 1 2 2 1 12 W W L LFF kips lb lb in inF 52 4409 5000 100 90542 =           = The estimated impact force of 52 kips for MASH test 3-11 represents a 4 percent decrease from the 54 kip design load used for NCHRP Report 350 test 3-11. This result is somewhat unexpected considering the 13 percent increase in vehicle weight and impact severity associated with this test. It leads to the conclusion that the structural adequacy of TL-3 barriers that comply with NCHRP Report 350 guidelines should be sufficient to comply with the same test level under MASH. A similar analysis can be conducted for Test Level 4. MASH recommends increasing the weight of the TL-4 single-unit truck (SUT) from 17,640 lb to 22,050 lb and increasing impact speed from 50 mi/h to 56 mi/h. The impact angle will remain unchanged and, therefore, will not influence the impact force. Since the dimensions of the SUT have not changed, the vehicle length and the contact area associated with an impact into a given longitudinal barrier system will not be factors. Using 54 kips as the design impact force for NCHRP Report 350 test 4-12, and nominal vehicle weights and impact speeds specified for the respective TL-4 tests, the impact force associated with MASH test 4-12 can be estimated as follows:               = 1 2 2 1 2 12 W W V VFF kips lb lb mph mphF 76 640,17 050,22 50 5654 2 2 =            = The estimated impact force of 76 kips for MASH test 4-12 represents a 41 percent increase from the 54 kip design load used for NCHRP Report 350 test 4-12. Consequently, some barriers that meet the NCHRP Report 350 guidelines as a TL-4 barrier may not have adequate strength to comply with the same test level under MASH. Another aspect of the structural adequacy criteria is that the test vehicle should not override the barrier. Adequate barrier height is required to prevent heavy trucks with high centers of gravity from rolling over a barrier. Full-scale crash testing has shown that 32-inch tall barriers are capable of meeting TL-4 impact conditions under NCHRP Report 350. However,

17 when MASH Test 4-12 was conducted on a 32-inch tall New Jersey safety-shape concrete barrier (see Test 11 in Table 3), the SUT rolled over the top of the barrier. After the unsatisfactory outcome of this test, it was proposed to reduce the C.G. height of the ballast of the SUT from 67 inches to 63 inches. This effectively decreases the overturning moment by decreasing the moment arm between the C.G. of the truck and the reactive force applied by the barrier. A test conducted under this project with the reduced ballast height failed due to roll of the truck over the barrier. Additional testing is required to determine what barrier height is required to contain the SUT under MASH test conditions. Vehicle Stability For all tests involving passenger vehicles, a key requirement for the safety of vehicle occupants is for the impacting vehicle to remain upright during and after the collision. Criterion F of NCHRP Report 350 states that moderate roll, pitching, and yawing are acceptable. The commentary in Section A5.2 further explains that “Violent roll or rollover, pitching, or spinout of the vehicle reveal unstable and unpredictable dynamic interaction, behavior that is unacceptable.” However, the term “moderate” used in Criterion F is not defined, thereby leaving evaluation of this criterion somewhat subjective. MASH retains language that the impacting vehicle should remain upright during and after an impact. However, to provide a further indication of vehicle stability, and to make evaluation of Criterion F more quantitative, the maximum roll and pitch angles are not to exceed a threshold of 75 degrees. Since the adoption of a 3/4-ton pickup truck as the design test vehicle for structural adequacy tests, vehicle instability and rollover has been a common failure mode associated with longitudinal barrier impacts including guardrails, bridge rails, and transitions. Compared to passenger cars, pickup trucks have a higher C.G., a shorter front overhang, and greater bumper height (see Table 4). All of these factors combine to make the pickup truck a more critical vehicle than a passenger car in regard to impact performance with roadside safety features. The propensity for wheel snagging, occupant compartment deformation, and vehicle instability (i.e., rollover) are greater for the pickup truck than most passenger cars. National Highway Traffic Safety Administration (NHTSA) officials believe that the static stability factor (SSF) is one of the most reliable indicators of rollover risk in single-vehicle crashes. The formula for calculating SSF is: SSF = T/2h, where T = track width and h = C.G. height A statistical study using data from six states showed that there is a strong correlation between a vehicle’s SSF and its likelihood of being involved in a rollover. A higher SSF indicates a more stable vehicle with less propensity for rollover. As expected, the pickup truck design vehicles have a lower SSF than the passenger sedan previously used under NCHRP Report 230 (see Table 4). More interesting is that although the new 2270P has a slightly greater

18 C.G. height than the 2000P, its SSF is actually greater than the 2000P. This is an indicator that the 2270P may be more stable in barrier impacts than the 2000P. Further, the longer front overhang of the 2270P makes it less critical than the 2000P in terms of snagging severity and snagging-induced instability. TTI researchers also believe the improved stability of the 2270P can be attributed to increased torsional rigidity provided by its different frame design and longer crew cab body. Table 4. Comparison of Critical Test Vehicle Dimensions. Vehicle Property Vehicle Type 4500S1 2000P2 2270P3 C.G. Height (inches) 22 27 28 Front Overhang (inches) 43 32 39 Bumper Height4 (inches) 12-21 16-25 14-27 Wheelbase (inches) 120 132 140 Track Width (inches) 62 64 68 Static Stability Factor5 1.41 1.19 1.21 1 4500-lb passenger sedan; NCHRP Report 230 design vehicle 2 4409-lb, 3/4-ton, standard cab pickup truck; NCHRP Report 350 design vehicle 3 5000-lb, 1/2-ton, 4-door, quad-cab pickup truck; MASH design vehicle 4 Range: bottom edge – upper edge 5 SSF = T/2h, where T = track width and h = C.G. height Although the data are very limited at this point, these observations regarding the relative stability of the two pickup truck design vehicles are supported by crash test data. Test 6 and Test 7 in Table 3 are nominally identical tests of a precast, F-shape, pin-and-loop, concrete median barrier. The only difference is the type of pickup. Test 6 was conducted with a 5000-lb, 3/4-ton, standard cab, GMC 2500 pickup; Test 7 involved a 5000-lb, 1/2-ton, 4-door, Dodge Ram 1500 quad-cab pickup. While both vehicles were contained and redirected, the 3/4-ton, standard cab pickup exhibited much greater roll and was noticeably less stable than the 1/2-ton, quad-cab pickup. Thus, devices that have stably contained and redirected the 2000P pickup under NCHRP Report 350 guidelines would not be expected to have stability concerns with the new 2270P pickup in MASH. In fact, it is possible that some devices that failed to comply with NCHRP Report 350 due to instability and rollover of the pickup truck might satisfy MASH. Occupant Compartment Deformation Another common mode of failure for bridge rails and guardrail-to-bridge rail transitions tested in accordance with the guidelines of NCHRP Report 350 is excessive occupant compartment deformation. This type of failure is most often associated with severe snagging of

19 the front, impact-side wheel at a joint, splice, or transition that results in the wheel being pushed into the fire wall and toe pan area of the occupant compartment. While such behavior was rarely observed when testing with large passenger sedans under NCHRP Report 230, the short front overhang of the pickup truck exposed the wheel and made snagging contact between the wheel and structural components of barriers a common occurrence. Evaluation Criterion D of NCHRP Report 350 states that “Deformations of, or intrusions into, the occupant compartment that could cause serious injuries should not be permitted.” Because the extent of deformation that can cause serious injury was not defined, this criterion was subjective in nature. Testing houses routinely had internal and external discussions regarding the magnitude and location of deformation that should constitute a pass or fail. To reduce the level of subjectivity associated with evaluating this criterion, the FHWA established a 6-inch threshold for occupant compartment deformation or intrusion. This threshold subsequently became the standard by which testing houses evaluated occupant compartment deformation. While MASH adopts a similar quantitative approach, it significantly relaxes the failure threshold established by FHWA. The revised criteria are founded largely on work performed by the Insurance Institute for Highway Safety and the National Highway Traffic Safety Administration (NHTSA). NCHRP study 22-14(02) documents the establishment of these criteria. The limiting extent of deformation varies by area of the vehicle damaged as follows: • roof < 3.9 inches, • windshield < 3.0 inches, • side windows – no shattering resulting from direct contact with structural member of test article, • wheel/foot well/toe pan < 8.9 inches, • side front panel (forward of A-pillar) < 11.8 inches, • front side door area (above seat) < 8.9 inches, • front side door (below seat) < 11.8 inches, • floor pan and transmission tunnel area < 11.8 inches. In addition to establishing maximum acceptable deformation thresholds to establish pass/fail criteria, a damage rating scale was introduced for further indication of vehicle damage and barrier performance. The damage scale has the following ratings and associated ranges of intrusion/deformation:

20 Rating Extent of Intrusion Good <5.9 inches Acceptable 5.9 inches – 8.9 inches Marginal 8.9 inches – 11.8 inches Poor >11.8 inches MASH also makes a clear distinction between: “(a) penetration, in which a component of the test article actually penetrates into the occupant compartment; and (b) intrusion or deformation, in which the occupant compartment is deformed and reduced in size, but no actual penetration is observed.” Penetration by any element of the test article into the occupant compartment of the vehicle is not allowed. The change in deformation thresholds notwithstanding, design characteristics of the 2270P will decrease its propensity for severe snagging and excessive occupant compartment deformation. Improved vehicle design and vehicle crashworthiness (e.g., introduction of energy managed crumple zones and other energy management strategies) will reduce occupant compartment deformation in a variety of crash scenarios. Furthermore, the longer front overhang of the 2270P makes it less critical than the 2000P in terms of snagging severity and snagging-induced occupant compartment deformation. Consequently, researchers believe that, as a result of the relaxed deformation thresholds, improved vehicle design, and the longer front overhang of the 2270P pickup, occupant compartment deformation will cease to be a critical factor in the evaluation of roadside safety devices. Devices that have contained and redirected the 2000P pickup under NCHRP Report 350 guidelines without excessive occupant compartment deformation (i.e., < 6 inches) would not be expected to have occupant compartment intrusion or deformation concerns with the new 2270P pickup proposed under MASH. In fact, it is possible that some devices that failed to comply with NCHRP Report 350 due to excessive occupant compartment deformation inside the pickup truck might satisfy MASH. GUARDRAILS In the mid 1990s, TTI researchers performed full-scale crash tests of all commonly used guardrail systems in accordance with NCHRP Report 350 Test 3-11 under a pooled fund study administered by FHWA(34). It was under this testing program that performance issues associated with light trucks impacting commonly used guardrail systems such as the standard strong steel-post W-beam guardrail system, G4(1S), the weak-post W-beam guardrail system, (G2), and the thrie-beam guardrail system (G9) were first identified.

21 Strong-Post W-Beam Guardrail [modified G4(1S) and G4(2W)] The strong steel post W-beam guardrail system, G4(1S), failed due to snagging of the pickup truck’s wheel on the steel support posts. The snagging was aggravated by the collapse of the W6x9 steel offset blocks, which precipitated rollover of the truck as it exited the barrier. Subsequent testing demonstrated that a modified G4(1S) system with 8-inch deep wood or structural plastic offset blocks between the W-beam rail element and W6x9 steel posts in lieu of the original W6x9 steel offset block was able to accommodate the 3/4-ton, 2-door, pickup truck design vehicle (denoted 2000P) and comply with NCHRP Report 350 guidelines(34-36). The strong wood post W-beam guardrail system, G4(2W), which utilizes 6 inch x 8 inch wood posts and offset blocks, contained and redirected the 2000P pickup(34). However, instability of the pickup truck resulted in the test being classified as marginally acceptable. Both of these strong-post W-beam guardrail systems are national standards. A cross-section of a typical W-beam guardrail is shown in Figure 2. The guardrail is constructed with 12-gauge W-beam rail mounted at a height of 21 inches to the center on 6-ft long W6x9 steel or 6 inch x 8 inch wood posts spaced at 6 ft-3 inches. The 8-inch deep offset blocks inserted between the rail and posts may be fabricated from wood or an approved alternative. Figure 2. Typical Cross-Section of Strong-Post W-Beam Guardrail.

22 These strong-post W-beam guardrail systems are at or near their performance limits under NCHRP Report 350 impact conditions. The increase in the weight of the proposed ½-ton, 4-door, pickup truck (designated 2270P) increases the impact severity of the structural adequacy test (Test 3-11) for longitudinal barriers by 13 percent. Under NCHRP Project 22-14(02), a series of crash tests were performed to assess the impact performance of strong-post W-beam guardrail when subjected to the revised impact conditions. As indicated in Test 1 of Table 3, a standard 27-inch tall, modified G4(1S) steel post W-beam guardrail failed due to rail rupture when impacted by a 5000-lb, 3/4-ton pickup truck(37). In a subsequent test of the same system with the 5000-lb, 1/2-ton, four door pickup truck that is currently proposed as the design test vehicle for MASH, the guardrail successfully contained and redirected the vehicle(38). However, the rail was torn through approximately half of its cross-section, indicating that the modified G4(1S) guardrail is at its performance limits with no factor of safety. The same sequence of tests with the two different pickup trucks was performed on a modified guardrail design known as the Midwest Guardrail System (MGS)(39,40). This modified guardrail increases the W-beam rail height from 27 inches to 31 inches, increases the depth of the offset blocks between the rail and posts from 8 inches to 12 inches, and moves the rail splice locations from the posts to mid-span between posts. In both tests, the pickup truck was successfully contained and redirected. The MGS guardrail was also successfully tested under modified Test 3-10 impact conditions with the proposed 2425-lb small car (designated 1100C) at a speed of 62 mi/h and a modified angle of 25 degrees (41). Thrie-Beam Guardrail (G9) The thrie-beam (G9) guardrail system is constructed of 6 ft-6-inch long W6x9 steel posts spaced 6 ft-3 inches on center and W6x9 offset blocks. The blockouts are 6 inches long x 18 inches deep and 4 inches wide at the flanges. A cross-section of a typical thrie-beam guardrail system (G9) is shown in Figure 3. The thrie-beam guardrail (G9) system contained and redirected the 2000P vehicle when tested in accordance with NCHRP Report 350(34). However, upon exiting the test installation at a high roll angle, the pickup truck subsequently rolled two and a quarter revolutions. During the impact event, the left front wheel severely caught the flanges of two posts and had direct contact with as many as five posts total. The post-to-wheel interaction severely twisted five posts and caused severe damage to the left front of the pickup truck. These events caused the pickup truck to subsequently rollover. This system does not meet NCHRP Report 350 and no additional work is warranted on this version of the thrie-beam guardrail system.

23 Figure 3. Typical Cross-Section of Thrie-Beam Guardrail. Thrie-Beam Guardrail (steel posts and routed wood blockouts) Following the failure of the standard G9 thrie-beam guardrail system described above, a steel post thrie-beam guardrail system with routed wood blocks was tested and evaluated. The 6 inch x 8 inch x 22 inch wood offset blocks were routed 4 inches wide x 3/8 inches deep to fit over the flange of the W6x9 steel posts. The steel post thrie-beam guardrail system with routed wood blocks successfully contained and redirected the 2000P vehicle in accordance with NCHRP Report 350 Test 3-11(42). Thrie-Beam Guardrail on Strong Wood Posts The strong wood post thrie-beam guardrail system with wood blocks is constructed of 6 ft-9-1/4-inch long x 6 inch x 8 inch wood posts spaced 6 ft-3 inches on center and using 6 inch x 8 inch x 22 inch wood offset blocks. The strong wood post thrie-beam guardrail system with wood blocks successfully contained and redirected the 2000P vehicle in accordance with NCHRP Report 350 Test 3-11(42).

24 Modified Thrie-Beam Guardrail The modified thrie-beam guardrail system was originally developed by TTI researchers in the mid-1980s to contain buses(43). Changes that increase the capacity of the modified thrie- beam include raising the rail height to 34 inches and incorporating different blockouts. The M14x18 offset blocks are 17 inches long x 14 inches deep and 6 inches wide at the flanges. The blockouts are modified by cutting a section out of the blockout web measuring 6 inches at the bottom and angling up at 40 degrees to the flange upon which the thrie-beam is attached. A cross-section of a typical modified thrie-beam guardrail system is shown in Figure 4. Figure 4. Typical Cross-Section of Modified Thrie-Beam Guardrail System. This system successfully contained and redirected a 20,000-lb bus impacting at a speed of 60 mi/h and an angle of 15 degrees(43). The modified thrie-beam guardrail was subsequently successfully crash tested in accordance with NCHRP Report 350 Test Level 4 (TL-4) impact conditions with both the 2000P pickup(34) and the 8000S single-unit truck(42). As described above, an NCHRP Report 350 Test 3-11 thrie-beam guardrail system is available using either strong wood or steel posts. The modified thrie-beam guardrail system has been successfully tested to NCHRP Report 350 Test 4-12. All tests involving the final design versions of these systems performed successfully and the test vehicles were contained and redirected in a stable manner. No further testing for Test Level 3 performance is anticipated for compliance with MASH. However, MASH requirements for Test Level 4 are much more severe than those in NCHRP Report 350. The increase in the weight of the proposed SUT (designated 10000S) and increase in impact speed from 50 mi/h to 56 mi/h increases the impact severity of the TL-4 structural adequacy test (Test 4-12) for longitudinal barriers by 58 percent. Therefore,

25 the researchers recommend performing MASH test 4-12 on the modified thrie-beam guardrail system if a TL-4 guardrail system is desired. Weak-Post W-Beam Guardrail (G2) The weak-post W-beam guardrail system (G2) failed to contain and redirect the 2000P vehicle at Test Level 3 due to the guardrail dropping ahead of the test vehicle and allowing the vehicle to override the guardrail. Subsequent testing demonstrated the same weak-post W-beam guardrail system (G2) could successfully contain and redirect the 2000P vehicle at Test Level 2 conditions(34). A cross-section of a typical weak-post W-beam guardrail system (G2) is shown in Figure 5. The guardrail is constructed with 12-gauge, W-beam rail mounted at a height of 30 inches to the top and supported on 5 ft-3 inch long S3x5.7 steel posts spaced at 12 ft-6 inches. The rail is attached to the posts with 5/16-inch diameter bolts. Additionally, a 1/2-inch diameter shelf bolt is used as a rail rest during construction and to provide some vertical support to the rail. Figure 5. Typical Cross-Section of Weak Post W-Beam Guardrail System. Pennsylvania Department of Transportation (PDOT) developed a variation of the weak post guardrail system (G2) that they refer to as their Type 2 system. This modified G2 guardrail successfully met NCHRP Report 350 test conditions 3-10 and 3-11(44, 45), thus fully qualifying it as a TL-3 rail system. The primary differences between the PDOT Type 2 guardrail system and the G2 include an increase in the W-beam rail mounting height to 32.3 inches, the use of W-beam backup plates at the posts, and the relocation of the rail splices from the posts to mid-span between posts. Additionally, the rail mounting bolts and washers and the post shelf bolt details differ from the G2 system.

26 The MASH 2270P test vehicle has demonstrated sensitivity to rail height. In addition, previous testing has shown that the impact performance of this system and other weak-post guardrail systems is sensitive to the post-to-rail attachment detail. TTI researchers believe the weak-post W-beam guardrail system (G2) may warrant consideration for re-evaluation with the MASH 2270P vehicle due to the height of the system and the opportunity for the weak-post systems to drop the rail off the posts in advance of the impacting vehicle, thus allowing the vehicle to travel over the rail element and behind the installation. MASH test 3-11 is recommended for the weak-post W-beam guardrail system (G2). Low-Tension Cable Guardrail (G1) High-tension cable roadside and median barrier systems have rapidly gained in popularity. The median application of cable barrier has gained exceptional attention as a cost-effective alternative for shielding motorists from crossover crashes. The relatively low cost makes cable barrier appealing for treating long expanses of highway. Additionally, the flexibility of these systems results in lower decelerations to an impacting vehicle, which lowers the probability of injury to occupants. However, sufficient space must be available to accommodate the greater design deflections associated with these systems and more maintenance may be required initially to keep the cables appropriately tensioned. The low-tension cable guardrail system (G1) successfully contained and smoothly redirected the 2000P vehicle at Test Level 3(34). The maximum dynamic deflection was 7.8 ft. A cross-section of the low-tension cable guardrail system (G1) is shown in Figure 6. The cable guardrail is constructed with three 3/4-inch diameter 3x7 wire ropes mounted on S3x5.7 steel posts. The mounting heights of the three cables were 23-1/2 inches, 26-1/2 inches, and 29-1/2 inches. The cables were attached to the posts with 5/16-inch diameter hook bolts. A New York cable anchor was used to terminate the system on each end. Figure 6. Typical Cross-Section of Low-Tension Cable Guardrail System.

27 Presently, there are five proprietary high-tension cable barriers in the market place. All of these systems are proprietary and, thus, will not be discussed herein. It is fully expected that the low-tension cable guardrail system (G1) will be capable of successfully containing and redirecting the new 5000-lb, 1/2-ton, 4-door pickup truck specified in MASH. The 13 percent increase in impact severity associated with MASH test 3-11 will likely increase dynamic deflection of this system and its proprietary counterparts. If desired, the modest increase in deflection can be offset through the use of reduced post spacing or other means. It should be noted that placement issues have been identified through simulation performed by the National Crash Analysis Center and other testing laboratories, in-service crash investigation, and full-scale crash testing. It is recommended that the low-tension cable guardrail system (G1) be tested to MASH with the 2270P vehicle on a 6:1 slope if it is to continue to be used on the NHS on slopes greater than 10:1. It is believed the low-tension cable guardrail system (G1) will continue to perform acceptably on 10:1 or flatter slopes if tested to MASH Test Level 3 conditions. It should be noted, that installation details for field-applied cable fittings are lacking. If use of this system is to be continued in the future, an investigation into consistent fabrication and installation instructions of field-applied cable fittings should be performed. Weak-Post Box-Beam Guardrail (G3) The weak-post box-beam guardrail system (G3) successfully contained and redirected the 2000P vehicle in compliance with NCHRP Report 350 test 3-11(34). The maximum dynamic deflection of the guardrail was 3.8 ft. The vehicle sustained moderate damage with very minimal deformation into the occupant compartment. The weak-post box-beam guardrail system (G3) is constructed of 5 ft-4 inch-long S3x5.7 steel posts spaced 6 ft on center. An L5 inch x 3-1/2 inch x 3/8 inch x 4-1/2 inch long shelf angle is attached to the post with a 1/2-inch diameter x 1-1/2-inch long hex bolt with washer and nut. A TS6 inch x 6 inch x 3/16-inch box-beam rail element is attached to the support angle with a 3/8-inch diameter x 7-1/2-inch long hex through bolt with washer and nut. The mounting height of the box-beam rail was 27 inches to the top of the box-beam rail element. A cross-section of a typical weak-post box-beam guardrail system (G3) is shown in Figure 7. The MASH 2270P test vehicle has demonstrated sensitivity to rail height. In addition, a structurally adequate rail attachment to the post used in this system and other weak post guardrail applications has proven critical in past developmental testing. TTI researchers believe the weak post box-beam guardrail system (G3) may warrant consideration for re-evaluation with the MASH 2270P vehicle due to the height of the system and the opportunity for the weak post systems to drop the rail off the posts in advance of the impacting vehicle, thus allowing the vehicle to travel over the rail element and behind the installation. MASH test 3-11 is recommended for the weak post box-beam guardrail system (G3).

28 Figure 7. Typical Cross-Section of Weak-Post Box-Beam Guardrail System. AESTHETIC BARRIERS Aesthetic barriers have special features added that do not necessarily add to the performance of the barrier. The features added are often done so to make the barrier better fit visually within a particular environment. Context sensitive design of the barrier may include adding colors or shapes that fit with the cultural values of a specific community or area or help the barrier blend with the surrounding environment. Many of the aesthetic barriers that will be presented herein were developed by the FHWA, Federal Lands Highway Divisions for use in national parks. Surface discontinuities and irregular shapes, wood and stone construction materials, and other methodologies are used to accomplish aesthetically pleasing barrier designs. The method of introducing wood and stone construction materials into a barrier can create an impact surface that may potentially cause snagging of components on the vehicle due to gouging in wood rail elements and posts or the surface discontinuities created by grout joints between stones of varying size and texture. Guidance for introducing surface discontinuities into concrete barrier faces is addressed in NCHRP Report 554, “Aesthetic Concrete Barrier Design”(46). FHWA performed Guardrail Testing Program IV, DTFH71-99-C-00035 to research the crashworthy performance of several longitudinal barriers, bridge rails, and transitions used on Federal Lands Highways when evaluated in accordance with NCHRP Report 350 evaluation criteria(47). The Rough Stone Masonry Guardwall and the Type A Steel-Backed Timber Guardrail were tested and evaluated to NCHRP Report 350 Test Level 3. The Steel-Backed Timber Round Log Rail and the Type B Steel-Backed Timber Guardrail were tested and evaluated to NCHRP Report 350 Test Level 2 and the Glacier Removable Rail and the Glacier Round Log Removable Rail were tested and evaluated to NCHRP Report 350 Test Level 1.

29 Additionally, Connecticut designed the Merritt Parkway wood guiderail and performed crash tests on it in accordance with NCHRP Report 350 under contract DTFH61-95-C-00119(48). Test Level 3 Rough Stone Masonry Guardwall The Rough Stone Masonry Guardwall is a vertical-face rigid barrier constructed of a precast or cast-in-place concrete core that is covered with a native stone and mortar veneer. The finished dimensions of the wall are 27 inches tall x 24 inches deep. A photograph of the Rough Stone Masonry Guardwall is shown in Figure 8. The Rough Stone Masonry Guardwall contained and redirected the 2000P pickup truck vehicle under TL-3 impact conditions. The occupant compartment deformation was 5 inches and the maximum vehicle roll angle was approximately 34 degrees(47). The Rough Stone Masonry Guardwall performed well and it is the opinion of the researchers that despite its 27-inch height, this system would perform acceptably under Test Level 3 conditions of MASH. Figure 8. Rough Stone Masonry Guardwall System. Type A Steel-Backed Timber Guardrail The Type A Steel-Backed Timber Guardrail is a semi-rigid rough sawn wood post, wood blockout, and wood rail element barrier with a steel plate rail bolted to the rear of the wood rail guardrail. The wood used is typically either Southern Yellow Pine or Douglas Fir. The steel plate provides the required tensile strength for the system. The wood posts are 10 inches deep x 12 inches wide x 6 ft-11 inches long and placed on 9 ft-10 inch centers. The wood rail elements are 6 inches deep x 10 inches tall x 9 ft-10 inches long. The rail is bolted to the post with a 4 inch-deep x 9 inch-tall x 12 inch-wide wood blockout mounted between the rail and the post. Each rail element is backed with a 3/8 inch x 6 inch A588 weathering steel plate attached with lag screws. The overall installed rail height is 30 inches. A photograph of the Type A Steel-Backed Timber Guardrail is shown in Figure 9.

30 Figure 9. Type A Steel-Backed Timber Guardrail System. The Type A Steel-Backed Timber Guardrail contained and redirected the 2000P vehicle under TL-3 impact conditions. Maximum dynamic deflection of the rail was 22.8 inches and maximum permanent deflection was 12.4 inches. The occupant compartment deformation was 3.5 inches and the vehicle maximum roll and pitch angles were minor. The vehicle was very stable throughout the event(47). The Type A Steel-Backed Timber Guardrail performed well and it is the opinion of the researchers that this system would perform acceptably under TL-3 conditions of MASH. Merritt Parkway Steel-Backed Timber Guardrail The Merritt Parkway Steel-Backed Timber Guardrail is a semi-rigid rough sawn steel post, wood blockout, and wood rail element barrier with a steel plate rail bolted to the rear of the wood rail guardrail. The wood used is typically either Southern Yellow Pine or Douglas Fir. The steel plate provides the required tensile strength for the system. The steel posts in the length of need section are W6x15 x 6 ft-6 inches long and placed on 10 ft centers. The wood rail elements are 6 inches deep x 12 inches tall x 9 ft-11-1/2 inches long. The rail is bolted to the post with a 4 inch-deep x 8 inch-wide x 11-inch tall wood blockout mounted between the rail and the post. Each rail element is backed with a 3/8 inch x 6 inch x 9 ft-6 inch A588 weathering steel plate attached with lag screws. The overall installed rail height is 30 inches. A photograph of the Merritt Parkway Steel-Backed Timber Guardrail is shown in Figure 10.

31 Figure 10. Merritt Parkway Steel-Backed Timber Guardrail System. NCHRP Report 350 Tests 3-10, 3-11, and 3-21 were performed on the Merritt Parkway Steel-Backed Timber. In addition, test 3-11 was performed both with a 4 inch x 6 inch wide concrete curb placed 12 inches forward of the face of the rail and without the curb. The Merritt Parkway Steel-Backed Timber Guardrail contained and redirected the 820C passenger car and the 2000P pickup truck vehicle in all tests. Test 3-11 without the curb produced the greatest damage to the occupant compartment and the maximum deflected rail distance of the three length-of-need tests performed. Maximum dynamic deflection of the rail was 45.3 inches and maximum permanent deflection was 33.1 inches. The occupant compartment deformation was 1.9 inches and the vehicle maximum roll and pitch angles were minor. The vehicle was very stable throughout the event. Test 3-21 of the transition to a concrete safety shape barrier produced a maximum dynamic deflection of the rail of 5.9 inches and maximum permanent deflection was 2.0 inches. The occupant compartment deformation was 2.2 inches and the vehicle maximum roll and pitch angles were minor. The vehicle was stable throughout the event(48). The Merritt Parkway Steel-Backed Timber Guardrail performed well and it is the opinion of the researchers that this system would perform acceptably under Test Level 3 conditions of MASH.

32 Test Level 2 Type B Steel-Backed Timber Guardrail The Type B Steel-Backed Timber Guardrail is a semi-rigid rough sawn wood post and wood rail element barrier with a steel plate rail bolted to the rear of the wood rail guardrail. The Type A and Type B Steel-Backed Timber Guardrails only differ in the inclusion or omission of the wood blockout. A photograph of the Type B Steel-Backed Timber Guardrail is shown in Figure 11. Figure 11. Type B Steel-Backed Timber Guardrail System. The Type B Steel-Backed Timber Guardrail contained and redirected the 2000P pickup truck vehicle under TL-2 impact conditions. Maximum dynamic deflection of the rail was 9.8 inches and maximum permanent deflection was 8.5 inches. There was no measurable occupant compartment deformation. The vehicle maximum roll and pitch angles were minor and the vehicle was very stable throughout the event(47). The Type B Steel-Backed Timber Guardrail performed well and it is the opinion of the researchers that this system would perform acceptably under Test Level 2 conditions of MASH. The omission of the blockout will likely produce more occupant compartment deformation than the Type A version. However, with the relaxed occupant compartment deformation evaluation criterion, the Type B Steel-Backed Timber Guardrail would be a candidate for evaluating to the higher Test Level 3 conditions. Round Steel-Backed Timber Guardrail The Round Steel-Backed Timber Guardrail is a semi-rigid rough sawn round wood post, wood blockout, and round (uniform diameter) wood rail element barrier with a steel plate rail

33 bolted to the rear of the wood rail guardrail. The wood used is typically either Southern Pine or Douglas Fir. The steel plate provides the required tensile strength for the system. The wood posts are 12-inch diameter x 6 ft-11-inches long and placed on 9 ft-10-inch centers. The wood rail elements are 10-inch diameter x nominally 9 ft-10 inches long. The rail is bolted to the post with a 5-1/2-inch deep x 8-inch tall x 8-inch wide wood blockout mounted between the rail and the post. Each rail element is backed with a 3/8-inch x 6-inch A588 weathering steel plate attached with lag screws. The overall installed rail height is 30.5 inches. A photograph of the Round Steel-Backed Timber Guardrail is shown in Figure 12. Figure 12. Round Steel-Backed Timber Guardrail System. NCHRP Report 350 Tests 2-10 and 2-11 were performed on the Round Steel-Backed Timber Guardrail. The Round Steel-Backed Timber Guardrail contained and redirected both the 820C passenger car and 2000P pickup truck vehicle. Test 2-11 represents the most severe test for evaluating rail deflection. Maximum dynamic deflection of the rail in Test 2-11 was 13.3 inches and maximum permanent deflection was 4.3 inches. Additionally, the maximum occupant compartment deformation was also observed in Test 2-11 with the pickup truck test vehicle. The occupant compartment deformation was 2.6 inches. The vehicle maximum roll and pitch angles were minor and the vehicle was very stable throughout both the 2-10 and 2-11 tests(47). The Round Steel-Backed Timber Guardrail performed well and it is the opinion of the researchers that this system would perform acceptably under Test Level 2 conditions of MASH and possibly would be a candidate for evaluating to the higher Test Level 3 conditions. Deception Pass State Park Log Rail The Deception Pass State Park Log Rail is an emulated historic rail developed under the joint sponsorship of the Washington Department of Transportation, Washington State Historic

34 Preservation Office, and Washington State Parks and Recreation. The rail was developed to emulate the appearance of the original 1935 Civilian Conservation Corps (CCC) constructed log rail supported on stone and mortar bollards. This rail was specially designed for the Deception Pass State Park area in Washington. The Deception Pass State Park Log Rail is a steel-backed, round turned wood log rail supported by concrete bollards faced with native stone veneer and mortar and intermediate steel pipe posts. The wood used is Douglas Fir. The steel plate provides the required tensile strength for the system. The finished veneer face of the stone bollards are nominally 4 ft-6inches wide x 2 ft-4 inches deep x 2 ft-10-1/2 inches tall and placed on 18-ft centers. One intermediate 8-inch diameter extra strong steel pipe support is used between the stone and mortar bollards to provide additional capacity to the rail. The wood log rail elements are nominally 12 inches in diameter x 17 ft-11-1/2 inches long. The rail is bolted to the bollards and intermediate support posts without the need for a wood blockout. Each rail element is backed with a 3/8-inch x 6-inch A588 weathering steel plate attached with lag screws. The overall installed rail height is 27 inches. A photograph of the Deception Pass State Park Log Rail is shown in Figure 13. Figure 13. Deception Pass State Park Log Rail System. The Deception Pass State Park Log Rail contained and redirected the 2000P pickup truck vehicle. Maximum dynamic deflection of the rail was unobtainable and maximum permanent deflection was 0.4 inch. The occupant compartment deformation was 2.3 inch and the vehicle maximum roll and pitch angles were minor. The vehicle was stable throughout the event. There was significant gouging of the log rail in the region of impact and minor wheel snagging on the leading edge of the stone bollard(49). The Deception Pass State Park Log Rail performed well and it is the opinion of the researchers that this system would perform acceptably under Test Level 2 conditions of MASH.

35 Test Level 1 Glacier Park Removable Timber Guardrail The Glacier Park Removable Timber Guardrail is a semi-rigid removable rough-sawn wood rail and steel post barrier with a 4-inch tall x 6-inch deep concrete curb placed in front of the face of the barrier. This barrier was designed to be removed annually during the closed winter period of the park to prevent snow accumulation. Removable guardrail posts are mounted atop foundation beams embedded in a concrete footing. The posts are fabricated from base plated W8x31 steel, are approximately 24 inches tall, and spaced 83.7 inches on center. A clamping plate and bolts are used to anchor the post to the foundation beam. The wood rail elements are 6 inches thick x 12 inches tall x 83 inches long and attached to 3/8-inch thick L-shaped bent weathering steel backup plates. The wood used is typically either Southern Pine or Douglas Fir. The L-shaped bent steel backup plates are attached with lag screws and provide the required tensile strength for the system. The overall installed rail height is 24 inches to the paved surface. A photograph of the Glacier Park Removable Timber Guardrail is shown in Figure 14. Figure 14. Glacier Park Removable Timber Guardrail System. NCHRP Report 350 Tests 1-10 and 1-11 were performed on the Glacier Park Removable Timber Guardrail. The Glacier Park Removable Timber Guardrail contained and redirected both the 820C passenger car and 2000P pickup truck vehicle. The measured dynamic rail deflections were very similar (test 1-10 was 2.5 inches and test 1-11 was 2.4 inches) in both tests due to the movement in the system, attributed primarily to the mounting methodology of the posts to their foundation beam anchor plates. Likewise, the occupant compartment deformation measurements

36 were very similar; test 1-10 was 0.28 inch and test 1-11 was 0.35 inch. The vehicle maximum roll and pitch angles were minor and the vehicle was very stable throughout both the 1-10 and 1-11 tests(47). The Glacier Park Removable Timber Guardrail performed well and it is the opinion of the researchers that this system would perform acceptably under Test Level 1 conditions of MASH. Due to the very limited use of Test Level 1 barriers, no additional testing or evaluation of this barrier is recommended. Glacier Park Round Removable Timber Guardrail The Glacier Park Round Removable Timber Guardrail is almost identical in detail to the previously presented Glacier Park Removable Timber Guardrail with the exception of the round wood rail elements and the absence of a curb. The wood rail elements are 12.5 inches in diameter x 83 inches long and attached to 3/8-inch thick L-shaped bent weathering steel backup plates. The round logs are flat on the back and bottom sides. The wood used is typically either Southern Pine or Douglas Fir. The L-shaped bent steel backup plates are attached with lag screws and provide the required tensile strength for the system. The overall installed rail height is 24 inches to the paved surface. A photograph of the Glacier Park Round Removable Timber Guardrail is shown in Figure 15. Figure 15. Glacier Park Round Removable Timber Guardrail System. NCHRP Report 350 Tests 1-10 and 1-11 were performed on the Glacier Park Round Removable Timber Guardrail. The Glacier Park Round Removable Timber Guardrail contained and redirected both the 820C passenger vehicle and 2000P pickup truck vehicle. No measurable dynamic or permanent rail deflection was noted in test 1-10. In test 1-11, dynamic rail deflection was 3.0 inches and permanent rail deflection was 1.7 inches. No occupant compartment deformation occurred in either test. The vehicle was stable throughout both the 1-10 and 1-11

37 tests(47). The Glacier Park Round Removable Timber Guardrail performed well and it is the opinion of the researchers that this system would perform acceptably under Test Level 1 conditions of MASH. Due to the very limited use of Test Level 1 barriers, no additional testing or evaluation of this barrier is recommended. MEDIAN BARRIERS Many of the roadside barriers previously discussed are also acceptable for use in median applications and therefore will not be repeated. The following roadside barriers are accepted for use as NCHRP Report 350 accepted median barriers by FHWA as outlined in Horne’s memorandum B64(5): Test Level 3 (TL-3) • Weak Steel Post Cable (3-strand) Guardrail and Median Barrier (G1;SGRO1a-b tested with New York Terminal by Washington State) • Weak-Post Box-Beam Median Guardrail and Barrier (SGRO3 and SGMO3) • Strong-Post (Wood) W-Beam Guardrail and Median Barrier with wood or approved plastic blockout (SGRO4b,SGMO4b, and SGMO6b) • Strong-Post (Steel) W-Beam Guardrail and Median Barrier with routed wood or approved plastic blockout (SGRO4a and SGMO4a with non-steel blocks, and SGMO6a with steel, wood or plastic blocks) • Strong-Post (Wood) Thrie-Beam Guardrail and Median Barrier with wood or approved plastic block (SGRO9c and SGMO9c) • Strong-Post (Steel) Thrie-Beam Guardrail and Median Barrier with routed wood or approved routed plastic block (SGRO9a and SGMO9a with non-steel blocks) • Merritt Parkway (CT) Steel-backed Timber Guiderail (Acceptance Letter B-45) Test Level 4 (TL-4) • Strong-Post Modified Thrie-Beam Guardrail and Median Barrier (SGRO9b and SGMO9b). Note: the correct length of the modified spacer block is 17 inches and not the 22 inches shown on PWBO3 in the AASHTO Roadside Design Guide. • 32-inch tall Safety Shape (New Jersey) Median Barrier (SGM11a) • 32-inch tall F-shape Median Barrier (SGM10a) • 32-inch tall Vertical Concrete Barrier* • 32-inch tall Constant Slope Barrier (TX and CA designs – see Acceptance Letters B17 and B-45)

38 Test Level 5 (TL-5) • 42-inch tall Safety Shape (New Jersey) Median Barrier (SGM11b) • 42-inch tall F-shape Median Barrier (SGM10b) • 42-inch tall Vertical Concrete Barrier* • 42-inch tall Constant Slope Barrier (TX and CA designs)** • 42-inch tall Ontario Tall Wall Median Barrier (SGM12 and Acceptance Letter B-19) * These two designs were tested as bridge railings. They may be used as roadside or median barriers if reinforcing and foundation details are equivalent to the crash tested installations. ** The Constant Slope Barriers were not tested to the TL-5 level, but may be considered TL- 5 barriers when cast in place or slip formed if the dimensions, reinforcing, and foundation details are equivalent to designs that have been successfully tested. Cable Median Barrier As previously presented, high-tension cable median barrier systems have rapidly gained popularity as a cost-effective alternative for shielding motorists from crossover crashes. The relatively low cost makes cable median barrier appealing for treating long expanses of highway. Additionally, the flexibility of these systems results in lower decelerations to an impacting vehicle, which lowers the probability of injury to occupants. However, sufficient space must be available to accommodate the greater design deflections associated with these systems. A cross-section of a typical low-tension cable median barrier is shown in Figure 16. Presently, there are five high-tension cable barriers in the market place. All of these systems are proprietary and, thus, will not be discussed herein. However, it is fully expected that these and the G1 system will be capable of successfully containing and redirecting the new 5000-lb, 1/2-ton, 4-door pickup truck specified in MASH. The 13 percent increase in impact severity associated with MASH test 3-11 will likely increase dynamic deflections of these systems. If desired, the modest increase in deflection can be offset through the use of reduced post spacing or other means.

39 Figure 16. Typical Cross-Section of a Low-Tension Cable Median Barrier System. Concrete Median Barrier Concrete barriers are frequently used in narrow medians along high-speed, high-volume roadways due to their negligible deflection, low life-cycle cost, and maintenance-free characteristics. The rigid nature of these concrete barriers results in essentially no dynamic deflection. Thus, vehicle deceleration rates and probability of injury are greater for concrete barriers than for more flexible systems. Although the installation cost is relatively high, concrete barriers require little maintenance or repair after an impact. This reduces the risk of maintenance personnel on high-volume, high-speed roadways. Concrete median barriers that meet NCHRP Report 350 include the New Jersey, F-shape, constant- or single-slope barrier, vertical wall, and the Ontario tall wall. Each of these barriers meet NCHRP Report 350 Test Level 4 when constructed 32 inches tall. Note that the Ontario tall wall was designed and used only as a 42-inch tall barrier, but should meet TL-4 requirements if constructed 32 inches tall. When these same barriers are constructed 42 inches tall, they all meet NCHRP Report 350 Test Level 5. The constant (single) slope barrier has not been tested to TL-5, but is considered by FHWA a TL-5 accepted barrier when it is cast-in-place or slip formed and the

40 dimensions, reinforcing, and foundation details are equivalent to the other barrier designs that have been successfully tested(5). The New Jersey profile has a long history of widespread use. However, it has been falling out of favor in recent years based on the realization that it can impart significant climb and instability to impacting vehicles. A vertical wall barrier eliminates issues of vehicle instability, but will impart slightly higher decelerations and cause more damage than the other barrier types. The F-shape and single-slope barriers have comparable impact performance and fall between the New Jersey safety shape and vertical wall parapet in terms of vehicle climb and decelerations. Basic dimensions of the New Jersey and F-shape concrete safety barrier are presented in Figures 17 and 18, respectively. The barriers are both 32 inches tall and the top width may vary to accommodate lighting and signage when necessary. A cross-section of the single-slope concrete barrier is shown in Figure 19. The barrier is 42 inches tall and has a top width and bottom width of 8 inches and 24 inches, respectively. The taller height and constant slope profile permit this barrier to accommodate multiple pavement overlays without affecting its impact performance with passenger vehicles. Figure 17. Typical Cross-Section of New Jersey Shape Concrete Safety Barrier.

41 Figure 18. Typical Cross-Section of F-Shape Concrete Safety Barrier. Figure 19. Typical Cross-Section of Single Slope Concrete Barrier.

42 Given the estimated impact force associated with MASH test 3-11 with the new pickup truck is comparable to the design impact force used for NCHRP Report 350 test 3-11, the New Jersey, F-shape, constant- or single-slope barrier, vertical wall, and the Ontario tall wall barrier should all easily meet the structural adequacy requirements for the MASH Test Level 3 (TL-3) impact conditions. The only possible problem that might exist with the “safety-shape” concrete median barriers is in regard to the stability of the new 5000-lb, 1/2-ton, 4-door, quad-cab pickup. However, as discussed earlier in this chapter, the new pickup truck design vehicle has a greater static stability factor (SSF) than the 3/4-ton, 2-door, standard cab pickup and limited full-scale crash testing conducted under NCHRP Project 22-14(02) with both vehicle types indicates that the 1/2-ton, 4-door quad-cab pickup is more stable than the 3/4-ton, 2-door standard cab pickup. Crash test data further supports the argument that instability of the pickup truck should not be an issue with any of the concrete barriers discussed. Under NCHRP Project 22-14(02), two tests (Test 6 and Test 7 in Table 3) were performed on a precast, F-shape, pin-and-loop, concrete median barrier. Test 6 was conducted with a 5000-lb, 3/4-ton, standard cab, GMC 2500 pickup(50) and Test 7 involved a 5000-lb, 1/2-ton, 4-door, Dodge Ram 1500 quad-cab pickup(51). In both tests, the vehicles were successfully contained and redirected. In another test, a 5000-lb, 3/4-ton, standard cab pickup was successfully contained and redirected after impacting a precast, Texas F-shape concrete maintenance barrier with X-bolt connection and 10-ft long segments(52). Testing has shown that a precast barrier system will impart more motion and instability to an impacting vehicle than a rigid, permanent barrier with the same profile. This is due to the increased deflection of the precast barrier system, which increases the effective impact angle between the pickup and the precast barrier segments downstream from the initial point of contact. Therefore, given that two different versions of precast, F-shape barriers successfully contained and redirected the more critical 5000-lb, 3/4-ton, standard cab, pickup, it can be concluded that the permanent F-shape concrete safety barrier will successfully contain and redirect a 5000-lb, 1/2-ton, 4-door, quad-cab pickup in an upright and even more stable manner. Further, the single slope concrete barrier, which previous testing has shown to have comparable dynamic vehicle behavior to the F-shape profile, should also demonstrate satisfactory impact performance for MASH test 3-11. Although the focus of the discussion has been the pickup truck redirection test (test 3-11), consideration must also be given to the small car redirection test (test 3-10). As previously discussed, MASH test 3-10 has been revised to include a heavier 2425-lb passenger car (denoted 1100C) and a higher 25 degree impact angle. This is compared to NCHRP Report 350 test 3-10, which involves an 1800-lb vehicle impacting the barrier at an angle of 20 degrees. Considering both the increase in weight and impact angle, the impact severity of the revised small car redirection test (MASH Test 3-10) has increased by 206 percent. Since the impact severity of the pickup truck redirection test is still twice that of the small car redirection test, the revised small car redirection test will not pose a problem in terms of structural adequacy. However, the effect of the increase in angle and impact severity on vehicle stability and occupant risk was a concern, particularly for shaped rigid barriers such as the New Jersey profile.

43 MASH test 3-10 was conducted on a permanent New Jersey profile barrier under NCHRP Project 22-14(02) to investigate this impact performance concern (see Test 8 of Table 3). In this test, a 2002 Kia Rio was successfully contained and redirected in an upright and stable manner and occupant risk measures were within acceptable limits(53). The New Jersey profile is known to impart more vehicle climb than the more stable F-shape, single-slope, and vertical profiles. Therefore, the success of this test can be used to conclude that the impact performance of the F-shape concrete safety barrier, the single slope barrier, and the vertical barrier will be satisfactory for MASH test 3-10. The performance of concrete barriers under MASH Test Level 4 (TL-4) impact conditions merits discussion. The change in the weight and vertical C.G. height of the single-unit truck (SUT), and impact speed associated with MASH TL-4 can adversely affect the performance of 32-inch tall concrete barriers that currently comply with TL-4 under NCHRP Report 350. As previously discussed, MASH recommends increasing the weight of the TL-4 single-unit truck from 17,640 lb to 22,050 lb and increasing the impact speed from 50 mi/h to 56 mi/h. The estimated impact force of 76 kips for MASH test 4-12 represents a 41 percent increase from the 54 kip design load used for NCHRP Report 350 test 4-12. Consequently, some barriers that meet the NCHRP Report 350 guidelines as a TL-4 barrier may not have adequate strength to comply with the same test level under MASH. Further, the increase in height of the SUT vertical C.G. may contribute to the SUT rolling over the top of the barrier. Historically, full-scale crash testing has shown that 32-inch tall barriers are capable of meeting TL-4 impact conditions under NCHRP Report 350. However, when MASH Test 4-12 was performed on a 32-inch tall New Jersey safety-shape concrete barrier (see Test 11 in Table 3), the 22,045-lb SUT, traveling 56 mi/h and impacting the barrier at a nominal 15 degrees rolled over the top of the barrier(54). After the unsatisfactory outcome of this test, it was proposed to reduce the C.G. height of the ballast of the SUT from 67 inches to 63 inches. This effectively decreases the overturning moment by decreasing the moment arm between the C.G. of the truck and the reactive force applied by the barrier. In a retest of the New Jersey safety shape performed under this project, the SUT with reduced ballast height still rolled over the top of the 32-inch tall barrier. Additional testing is required to determine what minimum barrier height is required to contain the SUT under the impact conditions specified for MASH test 4-12. In summary, concrete median barriers should readily comply with MASH Test Level 3 conditions. Further testing and evaluation does not appear necessary at this time to satisfy the MASH Test Level 3 conditions and, consequently, is given a low priority. However, testing has shown that the MASH Test Level 4 conditions are problematic for 32-inch tall barriers with regard to containment of the SUT vehicle. Additional testing is necessary to determine the minimum barrier height required to satisfy MASH test 4-12 impact conditions. No change in performance is expected with regard to the MASH Test Level 5 conditions for 42-inch tall barriers, as the test conditions remain unchanged.

44 Portable and Precast Concrete Median Barrier Portable and precast concrete median barriers are often used in work zones to shield motorists from hazards in the work area (e.g., pavement edge drops, excavations, equipment, etc.), provide positive protection for workers, and separate two-way traffic. Due to the temporary and frequently changing nature of work zones, these barriers are designed to be easily transported, placed, and relocated. Unlike permanent concrete barriers, these free-standing temporary barriers can undergo large displacements when subjected to a vehicular impact. Thus, vehicle deceleration rates will typically be less for portable and precast concrete median barriers than for rigid, permanent concrete barriers. On the other hand, the deflection of the free-standing barrier systems imparts more motion and instability to an impacting vehicle than a rigid, permanent barrier with the same profile due to an increase in the effective impact angle between the vehicle and precast barrier segments downstream from the initial point of contact. Low-Profile Barrier – Test Level 2 The low-profile barrier system is a 20-inch high precast concrete barrier system that incorporates a negative slope on the impact face. The low-profile barrier was originally developed for use in low-speed work zones where the use of a traditional 32-inch high concrete barrier system would significantly limit visibility. This is particularly important in urban areas where it is often necessary to have frequent openings in the barrier system that allow cross-traffic vehicles to enter the main traffic stream and vehicles in the main traffic stream to exit. Unlike the other barriers presented throughout this document, the low-profile barrier is a proprietary barrier. However, any agency or contractor may obtain a license to produce the barrier. Despite its proprietary nature, the researchers believe the popularity and practicality of the low-profile barrier warrant its presentation in this study. The low-profile barrier system consists of two different types of barrier segments: the primary low-profile segment and the end-treatment segment. The primary low-profile barrier segment is produced in 20-ft lengths. Figure 20 illustrates the low-profile barrier segment cross-section. The low-profile end-treatment is a 20 ft-long segment that tapers from a height of 20 inches at the high end to a height of 4 inches at the low end. Complete fabrication details for the low-profile barrier segment are presented in TxDOT standard detail sheet LPCB(1)-92 and complete fabrication details for the low-profile end-treatment are presented in TxDOT standard detail sheet LPCB(2)-92. The low-profile barrier system has been successfully tested and accepted for NCHRP Report 350 Test Level 2(55).

45 Figure 20. Typical Cross-Section of Low-Profile Barrier Segment. In addition to the low-profile barrier system meeting the qualifications for NCHRP Report 350 TL-2 impact conditions, further testing conducted at TTI has shown that the low-profile barrier segment can also successfully redirect a 4500-lb full-size passenger vehicle impacting with a speed of 60 mi/h and an angle of 25 degrees. These impact conditions correspond to the full-service impact criteria presented in NCHRP Report 230. The impact severity (IS) associated with this more severe impact can be determined to be 96,647 ft-lb. This impact severity is considerably higher (71 percent) than the impact severity associated with the revised TL-2 criteria recommended in MASH, which is calculated to be 56,508 ft-lb. Therefore, it is believed that the low-profile barrier system can easily meet the structural requirements for the MASH TL-2 testing criteria. The only possible problem that may exist with the low-profile barrier in regard to the new TL-2 testing criteria involves the stability of the new 5000-lb, 1/2-ton, 4-door, quad-cab pickup. Based on developmental research performed at TTI, it has been observed that when impacted at speeds greater than or equal to 50 mi/h, the low-profile barrier has a tendency to cause the NCHRP Report 350 3/4-ton pickup to gently roll onto its side, slide down the roadway and come to a stop. Because there is a 13 percent increase in the IS associated with the MASH pickup impact, there is a minimal chance that the pickup will become unstable under the new impact criteria. However, the impact severity associated with this 50 mi/h impact is 47 percent greater than the impact severity associated with the MASH TL-2 impact conditions. Further, even though the new pickup truck design vehicle proposed in MASH has a vertical C.G. approximately 1 inch greater than the 2000P, 3/4-ton, standard cab pickup of NCHRP Report 350, it has a greater static stability factor (SSF) than the 3/4-ton, 2-door pickup. Limited full-scale crash testing conducted under NCHRP Project 22-14(02) with both vehicle types indicates that the 1/2-ton, 4-door pickup is inherently more stable than the 3/4-ton, 2-door pickup.

46 For these reasons, it is the opinion of the researchers that the low-profile barrier system should be able to successfully redirect the new pickup under TL-2 impact conditions. However, this assertion may ultimately have to be demonstrated through full-scale crash testing. Based on the above discussion, the researchers have assigned a low priority to the retesting of the low-profile barrier system based on safety considerations alone. However, in light of the increasing popularity of this barrier system and its growing use, the testing priority of the low-profile barrier should perhaps be given consideration. Portable and Precast Median Barrier Connections Portable and precast barriers, such as the safety shape, F-shape, and single-slope barrier are connected to one another in work zones and other temporary application environments using one of a wide variety of end connections. Tables 5 and 6 show the non-proprietary portable concrete barrier connections that have been crash tested in accordance with NCHRP Report 350 and have received FHWA acceptance. Only NCHRP Report 350 non-proprietary connections are shown. Unlike permanent concrete barriers, these free-standing temporary barriers can undergo large displacements when subjected to a vehicular impact. Thus, vehicle deceleration rates will typically be less for portable and precast concrete median barriers than for rigid, permanent concrete barriers. However, the deflection of the free-standing barrier system imparts more motion and instability to an impacting vehicle than a rigid, permanent barrier with the same profile. This is due to the increase in the effective impact angle that arises between the vehicle and downstream barrier segments as the barrier segments displace during impact. In addition, in the case of pin-and-loop connections, the barriers not only displace laterally relative to one another but also rotate about the longitudinal axis relative to one another. The rotation of the barrier allows the impacting vehicle to more readily mount and climb the face of the barrier, thus sometimes resulting in very high vehicular pitch and roll angles. It is common for portable and precast barrier connections to be strong in shear but be weak in moment and/or torsional strength.

47 Table 5. FHWA Accepted Temporary Concrete Median Barriers with Pin Connection FHWA Approval Letter Agency/ Manufacturer Barrier Profile Segment Length (ft) Pin Connector Max. Barrier Deflection (ft) Comments Dia.(in) Restrained/ Unrestrained B-41 Univ. of Nebraska F-Shape 12.5 1.25 Restrained 3.74 marginal test; 49 deg vehicular roll angle B-54 Virginia DOT F-Shape 20.0 1.0 Restrained 6.00 B-61 CalTrans New Jersey 20.0 1.25 Unrestrained 0.85 barrier segments staked to ground with four 1 inch dia. x 24 inch long steel stakes B-67 Georgia DOT New Jersey 10.0 1.25 Restrained 6.33 large deflection due to joint failure (rupture of rebar loop); 38 deg vehicular roll angle B70 Idaho DOT New Jersey 20.0 1.25 Restrained 3.28 25 inch long bolt with heavy hex nut Unrestrained 3.61 26 inch long pin B-84 Indiana DOT F-Shape 10 1.19 Restrained 5.25 spacer tubes placed in gap between barrier to help limit free rotation B-86 Oregon DOT F-Shape 12.5 1.0 Unrestrained 2.49 standard barrier with two sets of three steel bar loops B-86A Oregon DOT F-Shape 10.0 1.0 Restrained 2.66 42 inch tall barrier; pin passed through C- channel connectors B-90 CalTrans Single- Slope 13.1 1.25 Unrestrained 2.46 dual pin connection through horizontal steel plates B-93 Ohio DOT New Jersey 10.0 1.25 Restrained 5.48 B-98 North Carolina DOT New Jersey 10.0 1.25 Unrestrained 5.05 two sets of three steel bar loops Montana New Jersey 10.0 1.25 Unrestrained 4.17 three sets of two steel bar loops Washington New Jersey 12.5 1.0 Unrestrained 4.53 marginal test; 52 deg vehicular roll angle Washington New Jersey 12.5 1.25 Unrestrained 4.10 marginal test; 59 deg vehicular roll angle

48 Table 6. FHWA Accepted Temporary Concrete Median Barriers with Miscellaneous Connections FHWA Approval Letter Agency/ Manufacturer Barrier Profile Segment Length (ft) Connection Type Max. Barrier Deflection (ft) Comments B-79 Pennsylvania DOT modified F- shape 12.5 Plate connector 8.38 grooves/slots cast into bottom of barrier ends fit over steel plates; requires free barrier ends to be restrained with dowels anchored into pavement; marginal performance - one barrier joint opened during test B-94 New York DOT New Jersey 20.0 I-beam connector 4.17 flanges of fabricated I-beam connector drop into sleeves cast into each end of adjacent barrier segments Montana DOT New Jersey 10.0 Bolted vertical plates 3.61 Two sets of 1 inch thick x 4 inch wide lapped steel plates oriented in vertical direction and connected using 1 inch diameter high-strength bolts Texas DOT F-Shape 30.0 Type X-bolt 1.51 Two 0.88 inch diameter high-strength cross bolts through pipes cast into barrier ends Texas DOT F-Shape 10.0 Type X-bolt 2.25 Two 0.88 inch diameter high-strength cross bolts through pipes cast into barrier ends

49 F-Shape Type X Connection As is noted in Table 6, the most recently developed portable concrete barrier connection and also the lowest deflection connection type is the TxDOT F-shape Type X or “cross-bolted” connection. The Type X connection utilizes two threaded rods/bolts to form the connection. The bolts are placed in different horizontal planes in the barrier at a prescribed angle with respect to the longitudinal axis of the barrier. The bolts pass through guide pipes cast into the ends of the barrier segments. The bolts exit one barrier segment and enter the adjacent barrier segment at the vertical center line of the barrier section. In plan view, the two connection rods/bolts form an “X” across the joint between adjacent barrier segments. Triangular wedges are cast into the barrier to permit the exposed ends of the cross bolts to be recessed and, thus, prevent vehicle snagging. The guide pipes through which the cross bolts pass are oversized to provide connection tolerance for barrier fabrication, installation, and placement of the barrier on horizontal and vertical curves. The tight moment connection provided by the cross-bolted design minimizes barrier deflections while maintaining constructability. Standard 32-inch height F-shape profile precast segments constructed with the Type X connection have been crash tested to NCHRP Report 350 in both 10-ft and 30-ft length segments. The F-shape segments were 24 inches wide at the base and 9-1/2 inches wide at the top(52). A photograph of the F-shape concrete safety barrier with Type X connections is shown in Figure 21. For complete fabrication details of this precast barrier system, the reader is referred to TxDOT standard detail sheets CSB(1)-04, CSB(2)-04, and CSB(8)-04. As previously mentioned, 10-ft and 30-ft segments of the F-shape concrete median barrier with Type X connection were successfully tested to NCHRP Report 350(51). Occupant risk measures were below desirable levels, and the maximum roll angle was 30 degrees for the 10-ft long segments and 23.3 degrees for the 30-ft long segments. Maximum dynamic deflection of the barrier was only 27 inches for the 10-ft long segments and 19 inches for the 30-ft long segments, which is the lowest deflection of any free-standing, unanchored concrete barrier system accepted under NCHRP Report 350 guidelines. It should also be noted, in anticipation of the MASH test conditions, the 10-ft long barrier segment test installation was impacted with a 5000-lb single cab pickup truck with a 27-inch C.G. Recall from previous discussions, this vehicle is more critical than the 5000-lb, 1/2-ton, 4-door, quad-cab pickup truck currently proposed under MASH in terms of both structural adequacy and stability. Thus, the F-shape barrier with Type X connection and 10-ft long segments is considered to have met the requirements of MASH. Further, since the F-shape barrier with Type X connection and 30-ft long segments offers improved vehicle stability compared to the version of the barrier, the successful MASH test of the F-shape barrier with 10-ft long segments can be used to infer compliance of the F-shape barrier with 30-ft long segments with MASH.

50 Figure 21. F-Shape Concrete Safety Barrier with X-Bolt Connection. Although MASH test 3-10 has not been conducted on a portable, F-shape concrete median barrier, it is not believed to pose an impact performance problem for the Texas F-shape barriers with Type X connections. As mentioned previously, MASH test 3-10 was performed on a permanent New Jersey profile barrier under NCHRP Project 22-14(02) (Test 8 of Table 3). In this test, a 2002 Kia Rio was successfully contained and redirected in an upright and stable manner and occupant risk measures were within acceptable limits. The New Jersey profile is known to impart more vehicle climb than the more stable F-shape profile, and the deflections of

51 the F-shape barriers with Type X connections for impacts with the small car under MASH test 3-10 impact conditions will be small. Therefore, both F-shape barriers with Type X connections should meet the impact performance requirements for MASH test 3-10. Portable and Precast Median Barrier Connections Summary Although portable and precast median barriers inherently can produce more vehicle instability than their permanently mounted counterparts, the MASH pickup truck vehicle has demonstrated improved stability over the 2000P pickup truck vehicle and is not believed to impose any additional load on the barrier. As presented in Tables 5 and 6, the current selection of NCHRP Report 350 tested portable and precast median barrier connections should comply with the revised impact performance guidelines proposed under the MASH conditions. Further testing and evaluation does not appear necessary at this time and, consequently, is given a low priority. TRANSITIONS Bridge rails are longitudinal barriers designed to keep vehicles from encroaching off bridge structures and encountering underlying hazards. Bridge rails are typically rigid in nature due to the lack of space on bridge structures to accommodate barrier deflection. Common types of bridge rails include continuous concrete barriers, metal rails mounted on concrete parapets, and both concrete and metal beam and post systems. Transition sections are commonly used to connect a flexible approach guardrail to a more rigid bridge rail. The purpose of the transition is to gradually change the stiffness of the rail section so a vehicle impacting the flexible approach rail does not pocket or snag severely on the end of the stiffer bridge rail. The change in stiffness is generally accomplished through a combination of increased post strength, reduced post spacing, and/or increased rail strength. Many of the guardrail-to-bridge rail transition designs tested and accepted under NCHRP Report 230 were unable to accommodate the 2000P pickup truck adopted as the design test vehicle for structural adequacy tests in NCHRP Report 350. The most common failure modes observed in full-scale crash tests of transitions with the pickup truck were excessive occupant compartment deformation and vehicle instability (i.e., rollover). It was found that the transition systems needed to be further stiffened to limit vehicle snagging to tolerable levels and avoid vehicle overturn. Snagging on one or more posts or the end of the bridge parapet frequently contributed to the front wheel being displaced into the floor and toe pan area thus causing excessive occupant compartment deformation and, in some instances, loss of the front wheel assembly which also contributed to vehicle instability. It was further determined that the clear opening beneath the transition rail element had to be reduced through the addition of a rub rail or curb to prevent the wheel of the pickup from intruding underneath the transition rail and snagging on the stiff transition posts or end of the bridge rail parapet. The use of 10-gauge or nested 12-gauge thrie-beam guardrail also sometimes accomplished the objective of minimizing the open space under the rail without the use of a rub

52 rail or curb. However, this was not always a definitive solution. As an example, a full-scale crash test was conducted to determine if the curb detail could be eliminated from TxDOT’s Test Level 3 (TL-3) nested 12-gauge thrie-beam transition system without adversely affecting impact performance. Test Designation 3-21 was performed in accordance with NCHRP Report 350. This test consisted of a 2000P pickup truck impacting the transition at a speed of 62.2 mi/h and an angle of 25 degrees. The test vehicle rolled over while exiting the test installation and, as a result, the nested thrie-beam transition system without curb failed to meet the impact performance criteria of NCHRP Report 350. Therefore, the curb had to be retained as part of the overall transition system(56). A few states use two different transition designs: a Test Level 3 (TL-3) system which is used on high-speed roadways (i.e., speeds > 50 mi/h), and a TL-2 system which is used on roadways with speeds of 45 mi/h or less. At the time of the writing of this report, 21 acceptance letters related to transitions had been posted to FHWA’s Roadway Departure Safety web site (keyword: bridge rail transitions) (http://safety.fhwa.dot.gov/roadway_dept/policy_guide/road_hardware/barriers). Table 7 outlines some general characteristics of the transitions referred to in the FHWA acceptance letters. Additionally, other transition designs may have been successfully crash tested but do not appear because a formal FHWA acceptance letter was not requested by the user agency. Table 7 also contains a few miscellaneous transitions the researchers are aware of that have been successfully crash tested but do not have a formal FHWA acceptance letter. A detailed discussion will not be presented for each of the transitions listed in Table 7. However, for purposes of general transition discussion, the transition system tested as part of NCHRP Project 22-14(02) is presented. A Test Level 2 transition is also presented. TL-3 Transition A schematic of TxDOT’s TL-3 transition is shown in Figure 22. This guardrail-to- concrete bridge rail transition consists of a nested thrie-beam rail supported on 7-ft long steel or wood posts spaced at 18-3/4 inches. A 4-inch tall curb runs along the length of the nested thrie-beam section. The front face of the curb is aligned with the traffic face of the wood blockout that offsets the thrie-beam from the support posts. A thrie-beam terminal connector is used to attach the downstream end of the transition to the concrete bridge rail parapet. On the upstream end, a 6 ft-3 inch, 10-gauge, thrie-beam-to-W-beam transition element is used to transition the thrie-beam to the W-beam rail element of the approach guardrail. Additional details of the TL-3 transition are presented in TxDOT standard detail sheet MBGF (TR)-02. This transition system was originally designed and tested at the Midwest Roadside Safety Facility (MwRSF) at the University of Nebraska under sponsorship of the Midwest State’s Regional Pooled Fund Program(57). Both steel post and wood post versions of the transition were successfully tested with a 3/4-ton pickup truck following NCHRP Report 350 test 3-21 impact conditions.

53 Table 7. Summary of NCHRP Report 350 Transition Tests Description Rail Rubrail Top Height of Rail (inch) Posts Post Spacing (inch) Max. Defl. (inch) Bridge Rail BOX BEAM TL-3 WYDOT Box- Beam FHWA B-143 & B37A (TTI report 473610 & MWRSF reports) 6 inch Box- beam TS6 x 2 x .188 28 min W6x9x64 Refer to specific system drawing F-shape, NJ-shape, vertical & single slope concrete barriers, & WY steel post & beam rail TL-4 NYSDOT Box- Beam FHWA B-127 (TTI Report 401021-7) Two - 6 inch Box-beams TS6 x 2 x .188 varies 16 posts; 4 - W6x9x72; 12 –S3x5.7x72 Refer to specific system drawing 4.72 NYDOT Four rail steel bridge rail THRIE-BEAM TL-3 & 4 WYDOT FHWA B-151 Symmetric transition - 12-ga W- beam to nested 12-ga thrie-beam None 31.0 9 posts; 7-W6x8.5x78 1-W6x8.5x78 1-W6x8.5x72 P1 47.5 from B1; P1-P6 18.75; P6-P9 37.5 N/A Wyoming Two-Tube Bridge Rail TL-4 CalTrans Thrie-Beam FHWA B-106 Symmetric transition – 10-ga W- beam to nested 10-& 12-ga thrie- beam with add’l field side 12-ga thrie-beam None 31.88 6 wood posts; 5 - 8x8x96 1 – 8x8x72 37.5 Concrete barrier or steel railing

54 Description Rail Rubrail Top Height of Rail (inch) Posts Post Spacing (inch) Max. Defl. (inch) Bridge Rail TL-3 NDOT Thrie- Beam FHWA B-105 TTI Report 404211-7 Symmetric transition – 12-ga W- beam to nested 12-ga thrie-beam None 31.65 6 posts; 2 – W6x25x102 4 - W6x25x84 48 37.5 37.5 (last space is 75) 3.15 Vertical face concrete barrier TL-4 ODOT Thrie- Beam FHWA B-99 TTI Report 401021-2a TTI Report 401021-5 Symmetric transition - W-beam to 10-ga thrie- beam 4 inch curb 31.60 6 posts; 2 - W8x24x96 1 - W6x25x96 3 - W6x25x72 49.6 37.5 37.5 75 21.65 (TL3) 7.08 (TL-4) Vertical face concrete barrier TL-4 MWRSF Report No. TRP-03-71-01- MWRSF Test TRBR-3 Symmetric transition – 12-ga W- beam to 10- ga thrie- beam Tapered 6.75 inch wood 31.65 10 posts; 4-8x8x78 6-8x8x72 48 to P1 P1-P7 @18.75; P7- P10@37. 5 6.42 Wood rail on wood posts on transverse glue laminated timber decks TL-4 MWRSF Report No. TRP-03-71-01- MWRSF-STTR-3 Symmetric transition – 12-ga W- beam to 10- ga thrie- beam None (TS8x3x3/16 transition cap rail at bridge rail used) 31.65 7 posts; 5-W6x15x84 2-W6x9x78 48 to P1 P1-P7 @37.5 5.63 Thrie beam on steel posts on transverse glue laminated timber decks TL-3 ITRANS MWRSF Report No. TRP-03-69-98 FHWA B-47 & B47A MWRSF-ITNJ-2 Symmetric transition – 12-ga W- beam to nested 12-ga thrie-beam - 4 inch triangular concrete curb 31.45 8 posts; 6-W6x9x80 2-W6x9x72 11.5 to P1; P1- P6@ 18.75; P6- P9@37.5 5.24 Safety shape w/ toe cut & Vertical wall parapet transition to New Jersey shape concrete

55 Description Rail Rubrail Top Height of Rail (inch) Posts Post Spacing (inch) Max. Defl. (inch) Bridge Rail TL-3 ITRANS MWRSF Report No. TRP-03-69-98 FHWA B-47, B47A & B47B MWRSF-ITNJ-4 Symmetric transition - 12-ga W- beam to nested 12-ga thrie-beam - 4 inch triangular concrete curb 31.45 8 posts; 6-6x8x84 2-6x8x72 or 7 inch diameter post may be used in place of 6x8 wood posts 11.5 to P1; P1- P6@ 18.75; P6- P9@37.5 3.90 Safety shape w/ toe cut & Vertical wall parapet transition to New Jersey shape concrete TL-3 MoDOT MWRSF Report TRP- 03-47-95 MWRSF-MTSS-2 Symmetric transition – 12-ga W- beam to 10- ga thrie- beam (both sides) None 31.0 9 posts; W6x9x72 11.5 to P1; P1-P6 @18.75; P6- P9@37.5 7.52 42 inch single slope median barrier TL-3 MoDOT FHWA letter 06/04/1999 to Ron Faller, MWRSF Symmetric transition – 12-ga W- beam to 10- ga thrie- beam or nested 12-ga thrie-beam None (C8x11.5 transition cap rail at bridge rail used) 31.65 8 posts; 5-W6x15x84 3-W6x9x78 48 to P1 P1-P8 @37.5 5.63 Thrie beam on steel posts TL-3 Nebraska FHWA B105 TTI Report 404211-7 Symmetric transition – 12-ga W- beam to nested 12-ga thrie-beam - TS4x4x5/16 behind first span None 31.65 5 posts; 2- W6x25x102 4- W6x15x84 49.25 to P1;P1- P5@ 37.5 3.15 Vertical concrete wall with bevel

56 Description Rail Rubrail Top Height of Rail (inch) Posts Post Spacing (inch) Max. Defl. (inch) Bridge Rail TL-4 AKDOT FHWA B-55A TTI Report 404311-5 Symmetric transition - 12-ga W- beam to nested 12-ga thrie-beam None 31.0 9 posts; 7-W6x8.5x78 2-W6x8.5x78 P1 45 from B1; P1-P6 18.75; P6-P9 37.5 5.15 Alaska Two-Rail Bridge Rail W-BEAM TL-3 ODOT GR3.4 FHWA B-127 Nested 12-ga W-beam None 27.75 6x8x72 wood 4 posts @ 18.75 and 4 posts @ 37.5 TL-3 MNDOT FHWA B-83 TTI Report 473390-3 Nested 12-ga W-beam with 4 inch curb Curb 27.0 11 posts; 2-10x10x96 (or W8x21x96) 4-6x8x84 3-6x8x72 2-6x8x72 (may substitute W6x8.5 for all 6x8 wood posts) 11.40 18.75 18.75 37.5 75 4.09 F-shape concrete TL-3 MNDOT FHWA B-83 TTI Report 473390-3 Nested 12-ga W-beam with 6 inch curb C6x8.2 Tapers from 32 to 27 11 posts; 2-10x10x96 (or W8x21x96) 3-6x8x84 6-6x8x72 (may substitute W6x8.5 for all 6x8 wood posts) 7.68 18.75 18.75 37.5 6.93 New Jersey shape concrete

57 Description Rail Rubrail Top Height of Rail (inch) Posts Post Spacing (inch) Max. Defl. (inch) Bridge Rail TL-4 PDOT FHWA B-81 & 81a TTI Report 404211-3 & 401301-1 Nested 12-ga W-beam C6x8.2 31.1 11 posts; 4-W6x9x84 7-W6x9x72 10.35 P1-P7 18.75 P8-P11 37.5 6.97 Flared F-shape concrete TL-4 PDOT FHWA B-81& 81a TTI Report 404211-3 & 401301-1 Nested 12-ga W-beam with 8 inch curb & drainage inlet Curb 31.1 9 posts; 2-W8x21x96 3-W6x9x84 4-W6x9x72 P1 & P2 16.9 & 49.0 from parapet P3-P5 @ 18.75 P6-P9@ 37.5 Flared F-shape concrete TL-3 AKDOT FHWA B-78 TTI Report 404311- 7&8 Nested 12-ga W-beam Varies from 29 to 27.8 7 posts; 3-W8x13x82 4-W6x8.5x78 P1 45 from B1; P1-P6 18.75; P7 37.5 4.57 Alaska Two-Rail Bridge Rail TL-3 CDOT FHWA B-77 TTI Report 404211-9 Nested 12-ga W-beam with 4 inch curb C6x8 and curb 27.8 9 posts; 2- W8x13x90 7- W6x8.5x72 12 P1- P5@18.7 5; P5- P9@37.5 3.03 New Jersey shape concrete TL-3 FHWA B-65 TTI Report 404211-12 12-ga W- beam with 12-ga W- beam rubrail 12 ga. W-beam 27.7 8 posts; 2-W8x13x90 6-W6x9x72 P1- P4@18.7 P4- P8@37.5 2.76 Vertical wall parapet transition to New Jersey shape concrete TL-3 TXDOT TTI Report 1804-9,10&11 Tubular W- beam with pipe inserts None 27 7" round wood 37.5 5.52 & 13.8 New Jersey safety shape w/ cut toe

58 Description Rail Rubrail Top Height of Rail (inch) Posts Post Spacing (inch) Max. Defl. (inch) Bridge Rail AESTHETIC TL-3 FHWA B-64D2 TTI Report 405181-22 6 x 10 Steel- backed timber guardrail with TS4x4x.25 backup None 27 6 posts; W6x20x78 B1- P1&P1- P4@29.5 ;P4- P6@59 5.5 Tubular Steel-Backed Timber Bridge Rail TL-3 FHWA B-64D2 TTI Report 405181- 18&5a (TL-2) 6 x 10 Steel- backed timber guardrail 6x6 timber 27 7 posts; 4-10x12x96 3-10x12x84 P1- P4@30; P4- P7@60 1.18 Straight Stone Masonry Guardwall with Tapered End TL-2 FHWA B-64D2 6 x 10 Steel- backed timber guardrail 6x6 timber 27 7 posts; 4-10x12x96 3-10x12x84 P1- P4@30; P4- P7@60 Curved Stone Masonry Guardwall TL-3 FHWA B-64D2 TTI Report 405501-4 6 x 12 Steel- backed timber guardrail Tapered curb 30 7 posts; W6x15x68 P1- P4@30; P4- P7@60 5.9 New Jersey shaped concrete

59 Figure 22. Elevation of Texas TL-3 Guardrail-to-Concrete Bridge Rail Transition. Under NCHRP Project 22-14(02), MASH test 3-21 was conducted on the steel post version of this guardrail-to-concrete bridge rail transition to evaluate its impact performance with the 5000-lb, 1/2-ton, 4-door pickup and assess its compliance with MASH. In this test (Test 9 in Table 3), a 2002 Chevrolet C1500HD crew cab pickup weighing 5084 lb impacted the transition at its critical impact point at a speed of 60.3 mi/h and an angle of 24.8 degrees. The pickup was successfully contained and redirected in an upright manner(58). Consequently, the TxDOT TL-3 transition system complies with MASH, and no further testing is necessary. TL-2 Transition Most transition systems have been crash tested under Test Level 3 (TL-3) of NCHRP Report 350, which is the basic test level required to receive approval of the system for use on high-speed roadways. Since there are no national transition designs that have been developed for lower speed conditions, most states typically apply the same transition standard to all roadways regardless of speed and traffic volume. However, the new transition designs developed to comply with NCHRP Report 350 represented a significant increase in installation cost and complexity over designs previously acceptable under NCHRP Report 230. Thus, it can be cost prohibitive to require use of the high-speed, TL-3 guardrail-to-concrete bridge rail transition systems on low-speed roadways.

60 For these reasons, TxDOT developed a cost-effective TL-2 transition for use on low-speed roadways. The TL-2 transition, shown in Figure 23, is entirely comprised of standard hardware components and is significantly less expensive and complex to install than the high-speed, TL-3 transition system. This transition consists of 12 ft-6 inches of nested W-beam rail supported on 6-ft long steel or wood posts spaced at 37-1/2 inches. The 27-inch mounting height greatly simplifies the ability to connect the transition to some existing bridge rails. A W-beam terminal connector is used to attach the downstream end of the transition to the concrete bridge rail parapet. Additional details of the TL-2 transition are presented in TxDOT standard detail sheet MBGF (TL2)-05. Figure 23. Texas TL-2 Guardrail-to-Concrete Bridge Rail Transition. Test Designation 2-21 was performed in accordance with the guidelines and procedures set forth in NCHRP Report 350(56). This test consisted of a 4409-lb, 3/4-ton pickup truck impacting the critical impact point of the transition at a speed of 43.5 mi/h and an angle of 25 degrees. The test vehicle was successfully contained and redirected in a stable manner and the TL-2 transition system met all applicable NCHRP Report 350 evaluation criteria. The maximum dynamic deflection of the transition rail was only 2.6 inches. The maximum roll angle of the pickup truck was 13.4 degrees, and the maximum occupant compartment deformation was only 0.4 inch. Based on the performance of this transition system, the researchers believe it will also perform acceptably to the MASH test conditions. Additional discussion of the general performance of transitions to the MASH conditions follows. General Transition Discussion The researchers believe the propensity for wheel snagging, excessive occupant compartment deformation, and vehicle instability (i.e., rollover) are greater for the 3/4-ton

61 pickup truck of NCHRP Report 350 than the 1/2-ton, 4-door pickup truck designated in MASH. Although the 13 percent increase in vehicle weight and impact severity may slightly increase dynamic deflections, increases, if any, in vehicle roll angles and occupant compartment deformations resulting from NCHRP Report 350 test 3-21 should be minimal. Most transition systems should be capable of safely accommodating the increase in impact severity without imparting excessive occupant compartment deformation (OCD) or vehicle instability. Additionally, even if the OCD were to modestly increase, it would unquestionably be below the 9- to 12-inch threshold established in MASH. With these factors in mind, it is the opinion of the researchers that most NCHRP Report 350 transition designs will comply with MASH test 3-21. Further testing and evaluation does not appear necessary at this time and, consequently, is given a low priority. END TERMINALS AND CRASH CUSHIONS Crashworthy end terminals and crash cushions are installed to shield some discrete hazard or the terminus of something rigid installed within the clear zone, such as the end of a flexible W-beam guardrail, the end of a rigid concrete barrier, a bridge pier, overhead sign structure, or a gore area. End terminals reduce the impact severity of an errant vehicle striking the terminus of a longitudinal barrier. If the terminal is struck along its side, the terminal may act to contain and redirect the striking vehicle or permit the vehicle to pass behind or through in a controlled manner if struck near its end at an angle. Crash cushions when impacted head-on reduce the impact severity by attenuating the energy of the errant vehicle by various means, such as, momentum transfer, material deformation, and friction. When a crash cushion is struck along its side, it may also contain/capture and/or redirect the striking vehicle or permit the vehicle to pass behind or through in a controlled manner. When a crash cushion is impacted along its nose and contains/captures or redirects the vehicle, it is referred to as a “non-gating” crash cushion. If the crash cushion allows a vehicle impacting at or near the nose of the crash cushion to pass through and travel behind, it is referred to as a “gating” crash cushion. Crashworthy end terminals are required to safely terminate guardrail ends. Currently, all but four W-beam guardrail end terminals that satisfy the safety evaluation criteria of NCHRP Report 350 are proprietary. Of the four W-beam guardrail end terminals that are non- proprietary, only two satisfy NCHRP Report 350 Test Level 3; the buried-in-backslope and the eccentric loader terminal. The crash cushions and guardrail end treatments that are proprietary in nature will not be discussed herein. The manufacturers of these devices will be required to assess the impact performance of their devices and ultimately demonstrate compliance of their devices with the new test and evaluation guidelines as determined necessary by the user agencies and FHWA. However, the researchers do note that the dramatic increase in impact severity of the pickup truck redirection tests and other changes in the test matrices for terminals and crash cushions will likely necessitate the modification of some of these systems.

62 End Terminals Test Level 3 Guardrail Terminals Buried-in-Backslope Guardrail Terminal (G4 System Guardrail). The W-beam guardrail buried-in-backslope terminal is used where the natural terrain backslope is in close proximity to the point where the barrier is introduced. This terminal was tested to NCHRP Report 350 3-35 on a 1:10 foreslope (with and without a flat-bottomed ditch), and on a 1:6 foreslope and a 1:4 foreslope forming a v-ditch with a 1:4 backslope. This terminal eliminates the possibility of an end-on impact with the end of the rail and reduces the likelihood of the vehicle traversing behind the rail. The W-beam guardrail buried-in-backslope terminal is used to terminate the G4(2W) or modified G4(1S) guardrail systems. The guardrail is flared across the ditch with its end anchored to a concrete anchor block buried in the backslope. The post height varies the last 50 ft of the installation as the rail tapers into the backslope. A W-beam rubrail is used where the bottom height of the guardrail beam exceeds 18 inches. Depending on the guardrail system it is attached to, the terminal posts are either 6 inch x 8 inch wood or W6x8.5 steel posts. A photograph of the Buried-in-Backslope Guardrail Terminal is shown in Figure 24. Figure 24. Buried-in-Backslope Guardrail Terminal.

63 As noted in FHWA acceptance letter CC53A(17): Key elements common to all buried-in-backslope include: 1) using a flare rate that is appropriate for the design speed until the flow line is reached; 2) keeping the W-beam rail height constant relative to the roadway grade until the barrier crosses the ditch flow line (and beyond where practical); 3) adding a rubrail whenever the clearance from the bottom of the W-beam to the ground line exceeds approximately 450 mm; 4) providing at least 22 m of barrier upstream from the beginning of the area of concern to the point where the barrier crosses the ditch flow line (to allow some recovery area for an impacting vehicle that may ride up a relatively flat backslope and get behind the barrier); and 5) using an anchor (concrete block or steel posts) that is capable of developing the full tensile strength of the W-beam rail. The buried-in-backslope G4 guardrail terminal contained and redirected the 2000P pickup truck vehicle in each of the three tests. Maximum dynamic deflections of the rail were 29.5 inches, 31.4 inches, and 26.4 inches and maximum permanent deflections were 9.4 inches, 19.7 inches, and 16.1 inches. The occupant compartment deformations were 2.6 inches, 1.3 inches, and 8.1 inches, and the vehicles were very stable throughout each impact event(59, 42). The buried-in-backslope G4 guardrail terminal performed well and is it is the opinion of the researchers that this system would perform acceptably under Test Level 3 conditions of MASH. Eccentric Loader Terminal (ELT). The Minnesota Department of Transportation sponsored crash testing the Eccentric Loader Terminal (ELT) to NCHRP Report 350 test conditions in 1998(60). In consultation with FHWA, only NCHRP Report 350 tests 3-31 (pickup head-on) and 3-35 (pickup redirect) tests were performed. The other tests were waived or were considered to be adequately addressed in prior tests performed when the ELT was evaluated in accordance with NCHRP Report 230. The ELT is a 37 ft-6-inch long flared W-beam guardrail terminal. The terminal head has a lateral offset of 4 ft. Attached to the end of the W-beam, at the nose of the terminal, is a vertically oriented corrugated pipe section with a fabricated structural steel loader mounted inside of the pipe section. The ELT is fabricated from three 12 ft-6-inch sections of 12-gauge W-beam mounted to two 6 inch x 8 inch Breakaway Cable Terminal (BCT) posts mounted in foundation tubes, interconnected with a ground channel strut, and 6 inch x 8 inch wood Controlled Releasing Terminal (CRT) posts in post locations 3 through 7. Posts 3 through 6 are spaced 50 inches on center. All other posts are spaced the standard 6 ft-3 inches. All posts, except the first post, are blocked out with 6 inch x 8 inch wood blocks. A photograph of the ELT is shown in Figure 25.

64 Figure 25. Eccentric Loader Terminal. The ELT marginally met the safety performance evaluation criteria for NCHRP Report 350 tests 3-31 and 3-35. The ELT contained and redirected the 2000P pickup truck vehicle in test 3-35. However, the rail partially tore at the splice located at post 6. Maximum dynamic deflection of the rail was 39.0 inches and the maximum permanent deflection was 25.2 inches. The occupant compartment deformation was 2.4 inches. The vehicle was very stable throughout the event. In test 3-31, the ELT brought the 2000P pickup truck vehicle to a stop. However, the pickup truck mounted and rode on top of the guardrail a distance of 149.2 ft from the point of impact. There was no occupant compartment deformation. The vehicle exhibited moderate roll and pitch as it mounted and rode the rail. As noted in FHWA acceptance letter CC56(19), the NCHRP Report 350 ELT differs from the NCHRP Report 230 version in that: 1) a steel post equivalent version of the ELT is not permitted; 2) post 7 is a CRT post (before it was a standard line post); and 3) post 2 offset distance changes from 25 inches to 26 inches. In addition, the pickup truck rode the rail for 147.6 ft. Each barrier installation terminated with an ELT should have a length-of-need sufficiently long to prevent an impacting vehicle from reaching a shielded fixed-object hazard that is directly behind the guardrail. In summary, the ELT guardrail terminal performed marginally in the tests performed to evaluate its compliance with NCHRP Report 350. The guardrail partially ruptured at a splice during the redirect test and rode on top of the guardrail installation a distance of 149.2 ft before coming to a stop in the end-on impact test. In consideration of other guardrail tests performed using the MASH pickup truck vehicle, it is the opinion of the researchers that this system has a high probability of rail rupture in the terminal redirection test with the MASH pickup truck. The end-on test resulted in a very long stopping distance and will likely not improve with the heavier and higher C.G. 5000-lb MASH pickup truck. The researchers do not recommend testing the

65 ELT to the MASH conditions. The probability of the ELT performing successfully to the MASH conditions is low and the terminal has seen only limited use. New York Cable Guardrail Terminal. The New York three-cable guardrail terminal is the only non-proprietary low-tension three-cable end anchor system crash tested and accepted for use on the NHS. The terminal was tested under a research program conducted by the New York State Department of Transportation in accordance with NCHRP Report 230. The research report documenting the results of this study is the March 1990 NYSDOT Research Report 148, “Cable Guiderail Breakaway Terminal Ends”(61). A photograph of the New York Cable Guardrail Terminal is shown in Figure 26. Figure 26. New York Cable Guardrail Terminal. FHWA compared the 12 tests performed by NYSDOT under the NCHRP Report 230 evaluation criteria to the seven tests required for the evaluation of a gating terminal under NCHRP Report 350. The seven NCHRP Report 350 tests required for a gating terminal are 3-30, 3-31, 3-32, 3-33, 3-34, 3-35, and 3-39. FHWA determined that only one additional test, NCHRP Report 350 test 3-34, would be required for the New York cable guardrail terminal to satisfy the safety performance requirements of NCHRP Report 350. Test 3-34 was successfully performed at TTI and FHWA subsequently accepted the terminal system for use on the NHS as a TL-3 terminal. Additional discussion of the FHWA analyses for this terminal is presented in FHWA acceptance letter CC-63(22). The researchers believe the New York cable guardrail terminal will continue to perform acceptably and there is no evidence to support any additional analysis is required to satisfy the MASH conditions.

66 Test Level 2 Guardrail Terminals Vermont Low-Speed Strong Post W-Beam Guardrail Terminal – (G-1d). The Vermont Department of Transportation sponsored crash testing the Vermont G1-d terminal to NCHRP Report 350 TL-2 test conditions. The Vermont G1-d terminal had not previously been crash tested and initially NCHRP Report 350 test 1-30 was performed. Upon successful completion of this test, Vermont DOT decided to evaluate the terminal to TL-2. Three additional tests were performed at TTI, tests 2-30, 2-34, and 2-35(62). Upon successful completion of these tests, FHWA accepted the Vermont G1-d terminal for use on the NHS where the speeds are 44 mi/h or less(21). The Vermont G1-d terminal is a 12 ft-6 inch-long flared strong post W-beam guardrail terminal. The terminal end has a lateral offset of 5 ft. The first W-beam rail section is shop bent to a 16-ft radius. Post 2 is positioned at the midpoint of the curved guardrail section. The terminal is anchored using a steel rod attached to the guardrail at post 3 and anchored to a concrete block in the ground between posts 2 and 3. The W-beam terminal is supported on W6x8.5 x 72-inch steel posts. All posts are spaced the standard 6 ft-3 inches apart and blocked out with W6x8.5 x 14-inch standard G4(1S) guardrail blockouts. A photograph of the Vermont Low-Speed Guardrail Terminal is shown in Figure 27. Figure 27. Vermont Low-Speed Guardrail Terminal. The Vermont G1-d terminal successfully met the safety performance evaluation criteria for NCHRP Report 350 tests 1-30, 2-30, 2-34, and 2-35. In test 1-30 (end-on impact), the Vermont G1-d terminal stopped the 820C vehicle with the vehicle still in contact with the terminal end. Maximum dynamic deflection of the rail was 46.1 inches and the maximum permanent deflection was 12.6 inches. There was no occupant compartment deformation. The vehicle was relatively stable throughout the event.

67 Upon successful completion of the 1-30, Vermont DOT decided to evaluate the terminal to NCHRP Report 350 TL-2. In test 2-30 (end-on impact), the Vermont G1-d terminal yielded to the 820C vehicle and permitted the vehicle to gate through the guardrail. The vehicle came to rest behind the rail near post 12. There was no occupant compartment deformation. The vehicle was very stable throughout the event. In test 2-34 (CIP), the Vermont G1-d terminal contained the 820C vehicle on the traffic side of the guardrail. The vehicle impacted the traffic side of the rail at post 2. The rail deformed, and the vehicle impacted the anchor rod near post 3 causing the vehicle to pitch and yaw out toward the traffic side of the rail. The vehicle came to rest in front of the rail between posts 2 and 3. There was 1.7 inches of occupant compartment deformation. The vehicle experienced moderate pitch and yaw during the event. The Vermont G1-d terminal performed marginally in test 2-34 in regard to the occupant risk values. The longitudinal occupant impact velocity was 37.1 ft/s (39.4 ft/s maximum allowed) and the occupant ridedown acceleration was 19.5 Gs (maximum allowed 20 Gs). In test 2-35 (LON), the Vermont G1-d terminal contained and redirected the 2000P pickup truck vehicle on the traffic side of the guardrail. The beginning of length of need for this terminal is post 3. The vehicle impacted the traffic side of the rail at post 3. The vehicle was smoothly redirected and came to rest 46 ft from the end of the terminal. There was no occupant compartment deformation. The vehicle was very stable throughout the event. The Vermont G1-d terminal performed marginally in test 2-34 in regard to the occupant risk values. The researchers are uncertain how this terminal would perform if test 2-34 were performed using the MASH 1100C vehicle. However, since this system is not known to be used in other states, the researchers give this terminal low priority for testing to the MASH conditions. Modified Eccentric Loader Terminal. The Modified Eccentric Loader Terminal (MELT) failed to satisfy the safety performance criteria in NCHRP Report 350 for performance Test Level 3. Due to the large quantities of this terminal already installed along roadway facilities operating at lower travel speed, the New England Transportation Consortium chose to sponsor crash testing the MELT to NCHRP Report 350 TL-2 test conditions(63). NCHRP Report 350 tests 2-30 and 2-31 were conducted at TTI. NCHRP Report 350 test 2-35 was conducted at Southwest Research Institute. Additionally, due to the fact the anchor detail and post spacing are essentially identical to the design used for the ELT TL-3 tests, additional impacts on the side of the MELT were not required by FHWA. Tests 2-32 and 2-33 were also waived by FHWA due to experience showing the angle impacts on the nose of gating terminals similar to the MELT are generally less severe than the head-on tests that were performed. Upon successful completion of tests 2-11, 2-30, and 2-31, FHWA accepted the MELT terminal for use on the NHS where the speeds are 44 mi/h or less(42). The MELT terminal is a 37 ft-6-inch long flared strong-post W-beam guardrail terminal. The terminal head is offset 4 ft. The first W-beam rail section is shop bent to a 38-ft radius over the first 6 ft-3 inches of its length and is bent to a 90-ft radius over the second 6 ft-3 inches of its

68 length. The second 12 ft-6 inch section of W-beam is shop bent to a radius of 90 ft over its entire length. Posts 1 and 2 are breakaway wood BCT posts set in steel ground tubes with soil plates and anchored together using a C6x8.2 ground strut. Posts 3 through 9 of the MELT terminal are 6 inch x 8 inch x 72 inch wood CRT posts. Posts 3 through 8 are spaced 50 inches apart. Post 9 is a standard 6 inch x 8 inch wood line post also spaced 50 inches from post 8. All posts, except post 1, use a 6 inch x 8 inch wood blockout. The rail was bolted to posts 1 and 9 only. W-beam backup plates were used at posts 4, 5, 7, and 8. The MELT was attached to a modified G4(1S) guardrail system. A photograph of the MELT is shown in Figure 28. Figure 28. Modified Eccentric Loader Terminal. In test 2-30 (end-on impact), the MELT yielded to the 820C vehicle, allowing the vehicle to pass behind the guardrail installation and come to rest 25.4 ft from the point of impact. Maximum dynamic deflection and maximum permanent deflection of the rail was 10.7 ft. The occupant compartment was deformed 1.5 inches. The vehicle was relatively stable throughout the event. In test 2-31 (end-on impact), the MELT yielded to the 2000P vehicle, allowing the vehicle to pass behind the guardrail installation. The truck turned back into the rear side of the guardrail, impacting post 14, climbing on top of the guardrail, and subsequently coming to rest 112.5 ft from the point of impact at post 21, straddling the guardrail. Maximum dynamic deflection and maximum permanent deflection of the rail was 2.3 ft. There was no occupant compartment deformation. The vehicle was moderately stable throughout the event. Although the impact angle for the beginning length-of-need test in MASH has increased to 25 degrees, the researchers believe there is adequate reserve capacity in the guardrail system at the energy level demanded by Test Level 2 conditions to successfully pass the MASH test. The

69 MELT should perform satisfactorily to the MASH conditions for Test Level 2. However, the researchers give this terminal a low priority for testing to the MASH conditions given the limited number of states that currently use the MELT. Crash Cushions Connecticut Impact Attenuator System The Connecticut Impact Attenuator System (CIAS) was developed in the early 1980’s and tested in accordance with Transportation Research Circular 191 and NCHRP Report 230. Additionally, an in-service performance evaluation of the CIAS was performed up through 1987. The CIAS is the only NCHRP Report 350 non-proprietary crash cushion used on the NHS. The CIAS consists of 14 steel cylinders arranged in a seven row matrix of cylinders affixed to a rigid backup structure that crush upon impact, thus attenuating the energy of an errant vehicle. Twelve of the steel cylinders are 4.0 ft in diameter and the two cylinders in the second row of the crash cushion are 3.0 ft in diameter. All the cylinders are 4.0 ft tall. The CIAS is constructed so that the wall thickness of the three cylinders anchored to the backup structure is 1/4 inch. The next two cylinders have a wall thickness of 3/8 inch followed by six cylinders (three rows of two cylinders each) with a 3/16-inch wall. The second row cylinders are 8-gauge steel and the single nose cylinder is 3/16 inch. Stiffening members constructed of 0.125 x 5.0 steel straps are placed in the rear most seven cylinders. Additionally, 1.5-inch diameter schedule 40 pipe is placed transversely in the four cylinders in front of the last row of three cylinders. All the cylinder contact points are interconnected with 7/8-inch diameter x 2-inch long A307 bolts, washers, and nuts. A photograph of the CIAS is shown in Figure 29. The CDOT standard drawings should be referred to for additional construction and anchoring details. Figure 29. Connecticut Impact Attenuator System.

70 To evaluate the CIAS in accordance with NCHRP Report 350, two crash tests were initially performed, 3-32 and 3-38(64). In test 3-32, an 1808-lb passenger car traveling 62.1 mi/h impacted the nose of the CIAS at 15.8 degrees and deformed the cushion 12.1 ft. In test 3-38, a 4409-lb pickup truck traveling 62.6 mi/h impacted near the midpoint of the side of the cushion at 19.9 degrees. The critical impact point was selected so that the centerline of the impacting vehicle was aligned with the center-rear of the cushion. The change in longitudinal occupant impact velocity was 37.0 ft/s and the occupant compartment was deformed 7.3 inches. The test was considered unsuccessful due to excessive deformation of the occupant compartment. CDOT modified the CIAS by placing the steel cylinder array on top of two steel skid rails anchored to the pavement surface and offsetting the rear-most outside cylinders 33 inches from the edge of the backup structure using steel L-brackets. The L-brackets allowed the cylinder to be extended out past the backup structure edge an additional 6 inches. Other changes were in the classification of the CIAS. The CIAS was originally tested as a “non-gating” crash cushion, in consultation with FHWA the classification was changed to a “redirective/gating” crash cushion. Thereafter, NCHRP Report 350 tests 3-33 (15 degree impact on the nose) and 3-35 (impact at the beginning of length-of-need) were performed(65). In test 3-35, a 4409-lb pickup truck traveling 61.8 mi/h impacted near the midpoint of the side of the cushion at 20.5 degrees. The change in longitudinal occupant impact velocity was 35.7 ft/s, the 10-millisecond longitudinal ridedown acceleration was -18.8 Gs, and the occupant compartment was deformed 4.3 inches. The test performance was considered acceptable. In test 3-33, a 4409-lb pickup truck traveling 62.1 mi/h impacted the nose of the cushion at 14.7 degrees. The cushion deformed 13.7 ft and the pickup passed through the cushion while yawing minimally. The change in longitudinal occupant impact velocity was 24.7 ft/s, the 10-millisecond longitudinal ridedown acceleration was -6.1 Gs, and the occupant compartment was deformed 0.3 inches. The test performance was considered acceptable. One additional test, test 3-34, was performed to demonstrate satisfactory performance of the 1808-lb passenger car when an impact occurred between the nose of the cushion and the beginning length-of–need along the side of the cushion(66). In test 3-34, an 1808-lb passenger car traveling 61.3 mi/h impacted the side of the cushion at the third row cylinder at 15.4 degrees. The cushion deformed laterally 3.1 ft and the car came to rest alongside the cushion. The change in longitudinal occupant impact velocity was 35.1 ft/s, the 10-millisecond longitudinal ridedown acceleration was -20.5 Gs, and the occupant compartment was deformed 1.3 inches. The test performance was considered marginal due the 10-millisecond longitudinal ridedown acceleration being -20.5 Gs. MASH is changing the impact angle for test designation 3-35 and 3-38 from 20 degrees to 25 degrees. The combination of the increase in the impact angle from 20 to 25 degrees and the increase in the weight of the pickup truck from 4409 to 5000 lb warrants the conduct of test 3-35 to evaluate compliance of the CIAS. The combination of the increase in test vehicle weight and impact angle increases the lateral impact severity approximately 73 percent. The researchers

71 believe there is a low probability the CIAS will satisfy the MASH evaluation criteria as currently designed. Narrow Connecticut Impact Attenuator System The Narrow Connecticut Impact Attenuator System (NCIAS) consists of eight steel cylinders arranged in a single row matrix affixed to a rigid backup structure that crush upon impact, thus attenuating the energy of an errant vehicle. All of the steel cylinders are 3.0 ft in diameter and are 4.0 ft tall. The wall thickness varies between 1/8 inch and 3/8 inch. The cylinder wall thicknesses are staged in a very specific order. Unlike the CIAS, the NCIAS has two 1-inch diameter wire ropes placed along each side of the cushion to control lateral deflection during side impacts. The cables are held vertically in place with U-bolts on the side of cylinders two through seven. Eye bolts anchor the cables to cylinder one. Two tension rods, one near the top and one near the bottom, are provided on the transverse diameter of the first cylinder. Additionally, two spacers fabricated from TS4 x 4 x 3/16 inch are placed near the top of the first and second cylinders. A stiffening member (compression struts) constructed of 1.5-inch diameter schedule 40 pipe is placed transversely in cylinders five through seven. Additionally, two compression struts, one near the top and one near the bottom, are used in the last cylinder (cylinder eight). All the cylinder contact points are interconnected with two 7/8-inch diameter x 2-inch long A307 bolts, washers, and nuts and cylinder eight is attached to the backup structure with four 7/8-inch diameter x 2-inch long A307 bolts, washers, and nuts. A photograph of the NCIAS is shown in Figure 30. The CDOT standard drawings should be referred to for additional construction and anchoring details. Figure 30. Narrow Connecticut Impact Attenuator System. To evaluate the NCIAS in accordance with NCHRP Report 350, five crash tests were performed. The tests performed were 3-32, 3-33, 3-37, 3-38, and 3-39. Test 3-39 failed due to excessive occupant compartment deformation and high longitudinal ridedown acceleration(67). Therefore the NCIAS cannot be used where a reverse direction impact may occur.

72 In test 3-32, an 1808-lb passenger car traveling 61.5 mi/h impacted the nose of the NCIAS at 14.4 degrees and deformed the cushion 12.9 ft. The vehicle yawed 109 degrees and came to rest 14.8 ft laterally from the cushion. All the occupant risk criteria were satisfied. In test 3-33, a 4409-lb pickup truck traveling 61.7 mi/h impacted the nose of the cushion at 14.7 degrees and deformed the cushion 7.4 ft. The vehicle yawed 72 degrees and came to rest 15.0 ft laterally from the cushion. All the occupant risk criteria were satisfied. In test 3-37, a 4409-lb pickup truck traveling 60.4 mi/h impacted the beginning length-of- need (BLON) along the side of the cushion at 20.2 degrees and deformed the cushion 2.1 ft. The BLON was approximately the mating interface of cylinders one and two. The vehicle was contained and redirected. All the occupant risk criteria were satisfied. The maximum occupant compartment deformation was 1.5 inches. The lateral ridedown accelerations were was -19.5 Gs, which approached the maximum value of 20 Gs. In test 3-38, a 4409-lb pickup truck traveling 62.2 mi/h impacted the critical impact point along the side of the cushion at 19.6 degrees and deformed the cushion 0.8 ft. The critical impact point was selected so that the centerline of the impacting vehicle was aligned with the center-rear of the cushion. The change in longitudinal occupant impact velocity was 24.9 ft/s and the occupant compartment was deformed 6.6 inches. The vehicle was contained and redirected. Although the occupant compartment deformation was marginal, all occupant risk criteria nonetheless were satisfied. MASH is changing the impact angle for test designation 3-38 from 20 degrees to 25 degrees. The combination of the increase in the impact angle from 20 to 25 degrees and the increase in the weight of the pickup truck from 4409 to 5000 lb warrants the conduct of test 3-38 to evaluate impact performance in accordance with MASH. The combination of the increase in test vehicle weight and impact angle increases the lateral impact severity approximately 73 percent. The researchers believe there is a low probability the NCIAS will satisfy the MASH evaluation criteria. Thrie-Beam Bullnose Guardrail System (Bullnose Attenuator) The Thrie-Beam Bullnose Guardrail System (Bullnose Attenuator) is the last type of median hazard protection that will be discussed herein. The Bullnose is a hybrid of a guardrail terminal and a crash attenuator. The Bullnose is an enclosed guardrail envelope that wraps the hazard being protected with a thrie-beam guardrail system. A length-of-need of thrie-beam guardrail is placed parallel to each of two roadways along a divided roadway median and is terminated by joining the separate guardrail runs together with a pseudo-elliptical section of thrie-beam. Thrie-beam guardrail has long been acceptable for use on the NHS. However, the terminus (the bullnose) of connecting two thrie-beam guardrails together did not meet NCHRP Report 350. The Midwest State’s Regional Pooled Fund Program sponsored a very in-depth

73 research program to develop a bullnose terminal that would meet the safety performance evaluation criteria of NCHRP Report 350(68-70). LS-DYNA and full-scale crash testing were used to develop the bullnose attenuator ultimately found acceptable for use on the NHS by FHWA in their letter CC68(23). The bullnose attenuator acceptable for use on the NHS consists of 12-gauge thrie-beam rail supported by 28 wood posts, 14 posts on each side of the system. Posts 1 and 2 on each side are wood BCT-type posts set in foundation tubes with holes at ground level. Posts 3 through 8 are standard wood CRT posts. Posts 9 through 12 are standard line posts. The last two posts on each side are used to tension and anchor the terminal with anchor cables and grounds struts. Horizontal slots are cut in the valleys of five thrie-beam sections to aid in the capture of the vehicle and to reduce the buckling and bending capacities of the rail sections. Two 5/8-inch diameter 6 x 25 cables were placed behind the top and middle corrugations on the curved nose section thrie-beam rail element to replace the beam strength lost by slotting the rail, and to contain a vehicle impacting on the nose. The bullnose attenuator is a non-gating terminal. A photograph of the Bullnose Attenuator is shown in Figure 31. Figure 31. Bullnose Attenuator. NCHRP Report 350 tests 3-32 and 3-38 were performed on the final design of the bullnose. However, tests 3-30, 3-31, and 3-33 were performed on earlier design versions of the bullnose. Redirection tests (3-36, 3-37, and 3-39) were not performed because the thrie guardrail itself had previously demonstrated the ability to contain and redirect vehicles. In test 3-30, an 1808-lb passenger car traveling 64.2 mi/h impacted the nose of the bullnose attenuator at an angle of -3.4 degrees with a 1/4-point offset. The bullnose attenuator brought the vehicle to a controlled stop in approximately 21.5 ft. The vehicle yawed counter-clockwise and came to rest against the guardrail. The highest 10-msec longitudinal occupant ridedown acceleration was 11.37 Gs and the change in the longitudinal occupant impact velocity was 31.5 ft/s. In test 3-31, a 4409-lb pickup truck traveling 64.3 mi/h impacted the nose of the attenuator at 0.6 degrees and deformed the terminal 53.6 ft. The vehicle came to rest in the

74 terminal system. The highest 10-msec longitudinal occupant ridedown acceleration was 9.2 Gs and the change in the longitudinal occupant impact velocity was 17.7 ft/s. In test 3-32, an 1808-lb passenger car traveling 65.3 mi/h impacted the nose of the bullnose at 15.7 degrees, deformed the attenuator 21.3 ft, the vehicle yawed counter-clockwise, and the vehicle came to rest in the attenuator. The highest 10-msec longitudinal occupant ridedown acceleration was 13.9 Gs and the change in the longitudinal occupant impact velocity was 32.6 ft/s. In test 3-33, a 4409-lb pickup truck traveling 64.0 mi/h impacted the nose of the attenuator at 13.4 degrees and deformed the terminal 36.9 ft. The vehicle came to rest in the terminal system. The highest 10-msec longitudinal occupant ridedown acceleration was 10.5 Gs and the change in the longitudinal occupant impact velocity was 20.4 ft/s. In test 3-38, a 4409-lb pickup truck traveling 62.0 mi/h impacted midway between posts one and two, along the side of the attenuator, at 21.5 degrees. The vehicle came to rest in the terminal system. The highest 10-msec longitudinal occupant ridedown acceleration was 10.9 Gs and the change in the longitudinal occupant impact velocity was 29.2 ft/s. As has previously been discussed, MASH is changing the impact angle for test designation 3-38 from 20 degrees to 25 degrees. The combination of the increase in the impact angle from 20 to 25 degrees and the increase in the weight of the pickup truck from 4409 to 5000 lb warrants the conduct of test 3-38 to evaluate the impact performance of the bullnose attenuator in accordance with MASH. In addition, test 3-31 should also be conducted with the MASH pickup truck vehicle. The combination of the increase in test vehicle weight and impact angle increases the impact severity approximately 73 percent for test 3-38 and 13 percent for test 3-31. The researchers believe there is a possibility the bullnose attenuator may fail to satisfy the MASH evaluation criteria for these tests. BREAKAWAY HARDWARE - SIGN SUPPORTS AND LUMINAIRES NCHRP Report 350 and MASH each address support structures, work zone traffic control devices, and breakaway utility poles in a chapter together. Additionally, MASH has added longitudinal channelizers to the group of hardware. Longitudinal channelizers, to date, are generally comprised of proprietary water-filled plastic barrier systems, and as such, will not be discussed. Likewise, numerous types of proprietary and non-proprietary work zone traffic control devices ranging from two-piece cones, channelizing drums, vertical panels, delineators, barricades, and temporary sign supports exist. The list is far too extensive to cite and discuss each device individually. Other than temporary sign supports, it should suffice to say that there are no new or additional impact performance concerns for any of these devices with the MASH pickup truck or small car. The discussions that follow will pertain to small and large sign supports and luminaire supports only. The NCHRP Report 350 test matrix for support structures, which includes sign supports and luminaires, specifies two crash tests with an 1808-lb passenger car: a low-speed test and a

75 high speed test. For TL-3, the relevant test designations are 3-60 and 3-61, which have design impact speeds of 21.8 mi/h and 62.2 mi/h, respectively. NCHRP Report 350 allowed for the use of a lighter weight (1543-lb) passenger car. However, the lighter vehicle was never really used in practice. In addition, the 4408-lb pickup truck, although not designated by a specific test number, could be used if the primary concern regarding the impact behavior of the support system is penetration of the test installation into the occupant compartment rather than excessive occupant impact velocity or ridedown acceleration and/or vehicle instability. NCHRP Report 350 test numbers 3-60 and 3-61 specified a critical impact angle (CIA) be determined and selected from a 0 to 20 degree impact envelope, as measured from the normal traffic travel direction. If the support structure is installed at or near an intersection, then the CIA could be tested at some angle other than 0 to 20 degrees (i.e. 90 degrees for example). The NCHRP Report 350 performance evaluation criteria for support structures consist of several evaluation factors. The evaluation factors as described in NCHRP Report 350 are: B. The test article should readily activate in a predictable manner by breaking away, fracturing, or yielding. D. Detached elements, fragments or other debris from the test article should not penetrate or show potential for penetrating the occupant compartment, or present an undue hazard to other traffic, pedestrians, or personnel in a work zone. Deformations of, or intrusions into, the occupant compartment that could cause serious injuries should not be permitted. F. The vehicle should remain upright during and after collision although moderate roll, pitching, and yawing are acceptable. H. Longitudinal occupant impact velocity is preferred to be less than or equal to 9.8 ft/s and should not exceed a maximum of 16.4 ft/s. I. Longitudinal and lateral occupant ridedown accelerations are preferred to be less than or equal to 15 Gs and the maximum is 20 Gs. K. After the collision it is preferable that the vehicle’s trajectory not intrude into adjacent traffic lanes. N. Vehicle trajectory behind the test article is acceptable. Of primary concern regarding the impact behavior of a support structure is preserving the integrity of the occupant compartment. To minimize the potential for injury during impact, penetration of the test article or parts of the test article into the occupant compartment is not permitted. However, NCHRP Report 350 does not specify a quantitative threshold for permissible deformation to the occupant compartment. FHWA drafted guidance for occupant compartment deformation and windshield damage in 1999 to supplement the evaluation criteria presented in NCHRP Report 350. These documents were titled: “Draft Guidance for Analysis of Passenger Compartment Intrusion”(71) and “Windshield Damage for Category II Work Zone Traffic Control Devices: Draft Guidance for Pass/Fail”(72). However, the windshield damage criteria were only for Category II work zone devices and never used for evaluating sign supports and luminaires. From the “Draft Guidance for Analysis of Passenger Compartment Intrusion”

76 document came additional performance evaluation criteria for small sign supports and luminaires. In general, deformation to the occupant compartment in the area of the roof could not exceed 6 inches. In MASH, the 1800-lb passenger car is replaced by a heavier 2425-lb passenger car (denoted 1100C). The impact speed for the low-speed test (test 3-60) has been decreased from 22 mi/h to 18.6 mi/h. The purpose of the speed reduction in MASH test 3-60 was to maintain the same nominal kinetic energy as NCHRP Report 350 test 3-60. The kinetic energy in MASH for test 3-60 is nominally 1.5 percent lower. This low-speed test evaluates the kinetic energy required to activate the breakaway, fracture, or yield the support mechanism. In addition, the effect the support has on the occupant impact velocity is normally most profound in test 3-60. The impact speed for the high-speed test (test 3-61) remains unchanged. A third test, test 3-62, was added to the MASH matrix for evaluating breakaway support structures. Test 3-62 involves a 5000-lb pickup truck vehicle impacting the test article at 62.2 mi/h at the CIA. In NCHRP Report 350, a specified CIA was to be determined and selected from a 0 to 20 degree impact envelope, as measured from the normal traffic travel direction. In MASH, the CIA envelope range was increased to 0 to 25 degrees. As before, if the support structure is installed at or near an intersection, then the CIA could be tested at some angle other than 0 to 25 degrees (i.e. 90 degrees for example). MASH adopts evaluation criteria language similar to NCHRP Report 350 but additionally establishes deformation thresholds to make assessment of Criterion D more quantitative and objective. The language adopted is largely a result of the 1999 FHWA guidance memorandums(71, 72). Key to the evaluation of support structures is a roof deformation limit of 3.9 inches and a windshield deformation limit of 3 inches. Further, no tearing of the interior plastic liner of the laminated windshield glass is permitted. Additional verbiage regarding side windows was added: “no shattering of a side window resulting from the direct contact with a structural member of the test article” is permitted. If the side windows are laminate glass then the windshield evaluation criteria applies. Table 8 lists the non-proprietary small and large sign supports and the breakaway luminaires accepted by FHWA to date. A discussion of their performance and the affect MASH may have on their future use follows the table. Sign supports may be generally divided into three general categories in the manner in which they yield to an impacting vehicle: flexible/bending, fracturing, and controlled release. Examples of these types of sign supports and luminaire bases are illustrated in Figure 32.

77 Table 8. Summary of FHWA-Accepted Breakaway Hardware. DEVICE VARIATIONS SIZE(S) FHWA LETTER(S) ACCEPTANCE DATE NOTES SMALL SIGN SUPPORTS Steel U-Channel Direct Bury Spliced Up to 3 lb/ft Up to 4 lb/ft SS-05, SS-36 6/15/1987, 9/3/1993 Dual, Strong Soil Up to Triple, Strong Soil Perforated Square Steel Tube Direct Bury or Sleeved Up to 2½ inch SS-05, SS-36 6/15/1987, 9/3/1993 Up to Triple Wood Post Southern Yellow Pine Western Red Cedar Douglas Fir 4 x 4 inch 4 x 6 inch 5 x 5 inch Mod 4 x 6 inch Mod 6 x 6 inch Mod 6 x 8 inch 5 inch round SS-25, SS-27, SS-32, SS-36, SS-45, SS-46A, SS-50 6/4/1991, 5/15/1992, 10/28/1992, 9/3/1993, 5/11/1994, 9/21/1995, 11/8/1994 Up to Dual Concrete (Pennsylvania) Rectangular, Uni-Directional Slip Base SS-05, SS-07, SS-36 6/15/1987, 9/1/1988, 9/3/1993 Triangular, Omni-Directional Slip Base SS-34, SS-36, SS-61 4/20/1993, 9/3/1993, 2/27/1996 Thin-Walled Aluminum Pipe 3 inch 4 inch SS-76 1/9/1998 Up to Dual ⅛” wall thickness Fiber Reinforced Plastic (FRP) Post SS-36 9/3/1993 LARGE SIGN SUPPORTS Dual, W6x12 Rectangular, Uni-Directional Triangular, Omni-Directional SS-25, SS-36 6/4/1991, 9/3/1993 Inclined or Level Single, W12x45 Rectangular, Uni-Directional Triangular, Omni-Directional SS-36 9/3/1993 Inclined or Level

78 Flexible/Bending Sign Support Fracturing Sign Support Controlled Release Sign Support Cast Aluminum Luminaire Support Figure 32. Sign Support and Luminaire Bases.

79 Flexible/bending supports may or may not ultimately release from their mounted position. If the support yields to the vehicle in crash testing by pulling completely out of the soil, careful attention should be paid when changing the founding method of anchoring the sign installation. For example, placing the same support in a concrete footing may defeat its ability to perform satisfactorily when impacted. When struck, bending supports yield by collapsing and lying down ahead of the vehicle. Fracturing supports yield by failure of the support cross- section. The failure may be controlled by the chemical/material properties of the support or by introducing a discontinuity in the support, for example in the form of a hole or saw cut in the support at a specific point. Additionally, other proprietary couplings are made frangible or have the ability to fracture by creating a shape discontinuity. Some proprietary luminaire bases are fabricated of cast aluminum and these bases rely on the material properties of the aluminum base to function properly. An aluminum alloy is selected that allows a frangible failure of the base structure when impacted by an automobile. Some large wood sign supports are fabricated with a hole in the neutral axis of the support near ground level to decrease the shear strength of the support and thus allow the support to more easily fracture. Controlled release supports intentionally construct a release point in the support(s). Examples of controlled release supports are slip base supports, support splice joints, and fuse plates. Some supports may use more than one method of breakaway activation depending on the size of the support. For example, steel U-channel supports may either yield by bending, fracturing, and/or controlled release. Additionally, it has been demonstrated in crash testing, a support type may behave in a ductile or flexible manner in low-speed impacts but behave in a brittle manner when struck at high-speed. Light gauge steel U-channel will yield by deforming and laying over in front of the vehicle. One manufacturer of steel U-channel uses high-strength steel in their posts, which make them perform and behave in a brittle manner when impacted. Likewise, as steel U-channel gets heavier per foot of cross-section, it also gets stronger and requires a joint or splice in the support to permit it to release or yield when impacted. Likewise, large structural steel supports may use a slip base at the lower attachment point where the support(s) meets with a ground anchor stub or foundation and use a fuse plate(s) to attach the lower portion of the support(s) to the upper portion of the support and sign panel. Typically, the fuse plate joins the upper and lower portions of the support together near the lower edge of the sign panel. The fuse plate(s) permit the release of the lower portion of the support from the upper portion of the support during impact. This behavior effectively: 1) permits a vehicle impacting a multi-support sign installation to pass beneath the installation when impacted; 2) reduces the effective inertial mass of the struck support; and 3) restricts the opportunity for the sign panel to impact the roof of the vehicle. Key parameters of the two small passenger cars (NCHRP Report 350 and MASH vehicle) such as bumper height, hood height, front overhang, and “wrap-around” distance are comparable. The “wrap-around” distance is the distance from the ground, up around the front of the hood, and rearward across the hood to the base of the windshield. It is a strong indicator of whether or not a sign support will contact the windshield of the impacting vehicle. This is especially important with flexible or bending sign supports. In addition, because the kinetic energy of the small passenger car test is nominally the same for MASH, the occupant impact velocity (OIV) should not increase. Due to the MASH small passenger vehicle going up in weight, the ridedown accelerations and OIV should actually be less if tested to the MASH conditions. However, the

80 level of windshield damage associated with some breakaway supports accepted under NCHRP Report 350 may not meet the more stringent criteria adopted in MASH. Further, careful consideration should be given to two new impact criteria in MASH, 1) increasing the impact angle envelope to 25 degrees, and 2) use of the pickup truck as a design test vehicle. Generally speaking, low-mounting height sign stands, such as portable work zone signs, should not pose a safety concern for the new pickup truck design vehicle. As an example, the wrap-around distance of a Dodge Ram 1500 quad-cab pickup is approximately 100 inches. Taller (i.e., high-mounting height) sign stands pose more of a concern. These systems are typically fabricated with larger support members to accommodate larger service loads (e.g., wind). If the supports do not readily fracture or release upon impact, they may deform around the front of the impacting vehicle and carry either the sign panel and/or top of supports into the windshield and/or roof of the pickup. During small car impacts with the sign support oriented 90 degrees to the travel path of the vehicle, the rigid substrate on some systems have penetrated the windshield and/or roof sheet metal. With the exception of the work performed by the Texas Department of Transportation (TxDOT)(73), the research community has very little real impact experience with regard to the pickup truck vehicle impacting sign and luminaire installations. TxDOT elected to perform MASH test 3-62 on their standard permanent small sign installations using the pickup truck vehicle. TxDOT uses two types of generic small sign support systems; a wedge anchor system and a triangular slip base system. Wedge Anchor Sign Support The wedge anchor sign support system uses a 2-7/8-inch outside diameter (O.D.) galvanized steel tubular socket cast inside a 12-inch diameter x 2 ft-6 inch deep non-reinforced concrete footing. The flattened edge of the 27-inch long socket is aligned parallel to the sign blank or perpendicular to the direction of impact. A 13 British Wire Gauge (BWG) galvanized steel tube having an outside diameter of 2-3/8-inch and a nominal wall thickness of 0.095 inches is inserted into the socket to a depth of 12 inches. An 8-1/2-inch long, 11-gauge galvanized steel wedge is driven between the socket and support post to a depth of 5-1/2 inches to secure the post in position. For the tests, a 3-ft x 3-ft x 5/8-inch thick plywood sign panel was attached to the 2-3/8-inch O.D. vertical support using two mounting clamps spaced 6 inches from the top and bottom edges of the sign panel. The mounting height from the ground to the bottom of the sign blank was 7 ft. Figures 33 and 34 show photographs and details of the wedge anchor sign support system crash tested.

81 Figure 33. Photographs of Wedge Anchor Sign Support.

82 Figure 34. Details of Wedge Anchor Sign Support.

83 Summary of Wedge Anchor Test Results The wedge anchor sign support system demonstrated satisfactory impact performance. The sign support activated by yielding to the impacting vehicle and then pulling out of its socket. The test vehicle sustained only minor damage, and there was no deformation of or intrusion into the occupant compartment. The computed occupant risk indices were below the preferred values set forth in MASH. The 2270P vehicle remained upright and stable during and after the collision event with only 1 degree of pitch and roll. The vehicle came to a controlled stop 275 ft behind the point of impact. In anticipation of minor vehicle damage, the test plan called for use of the same pickup truck for both crash tests (i.e., wedge anchor system and triangular slip base system). To reduce the probability of vehicle damage from the first test influencing the outcome of the second test, researchers planned to impact the two sign support systems at the vehicle quarter points rather than the centerline. Review of the high-speed video from the first test indicated that the trajectory of the support post was influenced by the hood geometry of the pickup. The hood of the Dodge Ram has a distinct drop in elevation at its quarter point that guided the support post toward the side of the vehicle and away from the windshield. Had the impact point been aligned with the center of the truck, the yielding support and sign panel may have contacted the windshield. Therefore, it was decided to impact a second wedge anchor system to obtain a more definitive evaluation of its impact performance. This evaluation was accomplished by impacting both a wedge anchor system and triangular slip base system in the second test. To minimize interaction between the two support systems, they were spaced 15 ft apart along the path of the vehicle with the slip base in the first position and the wedge anchor in the second position. It was theorized that the slip base would activate and rotate over the vehicle prior to the wedge anchor system contacting and yielding around the front of the vehicle. The two support systems were laterally offset 6 inches in opposite directions from the vehicle centerline to minimize the influence of vehicle damage induced in the first impact with the slip base on the outcome of the second impact with the wedge anchor. Photographs of the test installation setup are shown in Figure 35. Discussion of Test 2 of Wedge Anchor System The wedge anchor sign support system demonstrated satisfactory impact performance. The sign support activated by yielding to the impacting vehicle and then pulling out of its socket. Even with the more central impact on the bumper and hood, there was no secondary contact between the sign support structure and windshield. The height of the hood helped propel the yielding support post forward and prevented it from deflecting rearward enough to engage the windshield. The test vehicle sustained only minor damage, and there was no deformation of or intrusion into the occupant compartment resulting from the impact with the wedge anchor system. The computed occupant risk indices were below the preferred values set forth in MASH. The 2270P vehicle remained upright and stable during and after the collision event and came to a controlled stop behind the point of impact.

84 Figure 35. Vehicle/Installation Geometrics for Wedge Anchor & Triangular Slip Base Test.

85 Texas Triangular Slip Base There are two variations of the Texas triangular slip base sign support system. One version uses a 10 BWG galvanized steel tube as the vertical support and can accommodate sign panels up to 16 ft2 in area. The other version uses a schedule 80 pipe support and is acceptable for use with sign panels with areas up to 32 ft2. In absence of other factors, the heavier sign support system (i.e., the schedule 80 pipe support and larger sign panel) is typically considered to be the most critical in terms of occupant impact velocity. However, there was concern that the thin wall 10 BWG support could exhibit local buckling and collapse when impacted by the taller pickup trucks, possibly hindering the activation of the slip base mechanism. Therefore, since occupant impact velocity is not a major concern for the heavy pickup truck compared to the 1800-lb passenger cars used in previous testing of the Texas triangular slip base system, researchers decided to test the slip base with a 10 BWG support post. A 10 BWG galvanized steel tube with an outside diameter of 2-7/8-inch and a nominal wall thickness of 0.134 inches was used as the vertical support for the slip base system. A T-shaped bracket was attached to the vertical support to provide bracing for the sign panel. The T-bracket consisted of a 3-1/4-inch O.D. stub welded to a 2-3/8-inch O.D. horizontal steel tube. The stub of the T-bracket fit over the end of the 2-7/8-inch O.D. support and was secured using two 3/8-inch diameter ASTM A307 bolts. A 4 ft x 4 ft x 0.1-inch thick aluminum sign blank was attached to the 2-3/8-inch O.D. horizontal member and 2-7/8-inch O.D. vertical support using a total of three mounting clamps located 6 inches from the bottom and each edge of the sign panel. The mounting height to the bottom of the sign blank was 7 ft. The upper slip base casting consists of an integral collar and triangular base plate. The upper slip base casting slides onto the end of the steel pipe support. The lower slip base assembly consists of a 3-inch diameter x 3-ft long galvanized schedule 40 pipe stub welded to a 5/8-inch thick steel triangular base plate having the same geometry as the upper plate. The pipe stub was embedded in a 12-inch diameter x 3.5-ft deep unreinforced concrete footing such that the top face of the lower triangular slip plate was approximately 2 inches above the ground. The upper slip base unit is bolted to the lower slip base unit using three 5/8-inch x 2-1/2-inch long A325 or equivalent high-strength bolts that were tightened to a prescribed torque of 38 ft-lb. The slip base was oriented such that the direction of impact was perpendicular to one of the flat faces of the triangular plate. High-strength washers were used under both the head and nut of each bolt, and an additional washer was used to offset the two slip plates. The bolts are held in place by a keeper plate which is fabricated from 30-gauge galvanized sheet steel. Set screws in the collar of the upper slip base casting were then tightened to a prescribed torque of 60 ft-lb to secure the vertical support within the casting and keep it from rotating. Figures 36 and 37 show photographs and details of the triangular slip base sign support system.

86 Figure 36. Photographs of Texas Triangular Slip Base.

87 Figure 37. Details of the Texas Triangular Slip Base.

88 Summary of Texas Slip Base Test Results The triangular slip base sign support system demonstrated satisfactory impact performance when evaluated in accordance with MASH criteria. The slip base mechanism activated as designed. The detached supports and sign panels did not penetrate, or show potential for penetrating the occupant compartment, or to present undue hazard to others in the area. Maximum occupant compartment deformation was 3.0 inches in the roof area on the passenger side resulting from secondary contact with the sign support and sign panel. The computed occupant risk indices were below the preferred values set forth in MASH. The 2270P vehicle remained upright and stable during and after the collision event and came to a controlled stop behind the point of impact. Discussion of Texas Slip Base System Given that the triangular slip base with a 10 BWG support post was found to comply with MASH, what can be inferred regarding the impact performance of the slip base with a schedule 80 pipe support? It could be argued that the heavier mass of the schedule 80 support and its larger sign panel will produce greater occupant compartment deformation than that measured in the crash test of the lighter-weight slip base system with 10 BWG support. However, the heavier mass also increases the inertial resistance of the schedule 80 support system, which can reduce the rotational velocity imparted to the support during impact. Decreased rotational velocity will tend to shift the point of contact on the roof further rearward and decrease the deformation resulting from that contact. Furthermore, the larger sign panels typically associated with the schedule 80 support are likely to span the width of the roof and engage the door headers. The door headers are much stiffer than the central portion of the roof, and engaging them will tend to reduce the overall deformation resulting from contact with the sign panel. For these reasons, the researchers believe that the triangular slip base with a schedule 80 support post will comply with MASH. However, further testing of the system may be warranted to verify this assumption. CONCLUSIONS The researchers recommend performing engineering analysis and/or computer simulation of existing NCHRP Report 350 accepted sign and luminaire installations to evaluate their performance with regard to risk of occupant compartment deformation and intrusion when struck with the pickup truck test vehicle. Engineering analysis and/or computer simulation can be used to help predict whether or not secondary contact between a support system and an impacting vehicle will occur, and the probable location of the contact. However, the only way to reliably determine the extent of windshield damage and roof deformation from secondary contact is through full-scale testing. Sign and luminaire installations that, through engineering analysis and/or computer simulation, are found to have suspect performance would require full-scale crash testing with the pickup truck test vehicle.

Next: IV. Crash Testing Matrix and Evaluation Plan »
Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report Get This Book
×
 Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Web-Only Document 157: Volume I: Evaluation of Existing Roadside Safety Hardware Using Updated Criteria—Technical Report explores the process that was followed in developing NCHRP Research Results Digest (RRD) 349: Evaluation of Existing Roadside Safety Hardware Using Manual for Assessing Safety Hardware (MASH) Criteria.

NCHRP RRD 349 explores the safety performance of widely used non-proprietary roadside safety features by using MASH. Examples of features evaluated include longitudinal barriers (excluding bridge railings), terminals and crash cushions, transitions, and breakaway supports.

READ FREE ONLINE

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!