Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments (2022)

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.

147   9.1 Summary The lateral extension of vehicle components behind, above, or below the front face of each traffic barrier is referred to as zone of intrusion, or ZOI. Project NCHRP 22-34 was executed to collect background data related to previous full-scale testing of roadside barriers, augment full- scale crash test data with calibrated computer simulation data, and generate recommended ZOI envelopes for updating the AASHTO RDG (AASHTO 2011). 9.1.1 Background Data Collection A literature review was conducted on previous ZOI research, full-scale crash testing, and simulation studies for various rigid traffic barriers. This review included 95 successful crash tests conducted according to MASH (AASHTO 2016) TL-2 through -5 and NCHRP Report 350 TL-5 (Ross et al. 1993). For the purposes of this study, a barrier was considered rigid if the dynamic deflection was less than or equal to 10 in. during impact. Barrier profiles considered included New Jersey, F-shape, single-slope, vertical, and low-profile concrete barriers as well as steel bridge rails and steel-concrete combination rails. During evaluation, similar barrier shapes were grouped together for ZOI envelope development. Eighty of the 95 tested barriers deflected less than 10 in. and were therefore considered rigid; of these, 47 had enough publicly available video and data to obtain ZOI measurements. Twenty-four studies involving FEA of vehicle-rigid barrier impacts were reviewed, most of which were consistent with MASH TL-3 impact conditions. Rigid barrier shapes used for com- parison included the RESTORE, Jersey and F-shape temporary and permanent, single-slope, and steel bridge rail barriers. Most simulated systems used for model validation ranged in height from 32 in. to 42 in. Lateral vehicle extent over the front barrier face was not the primary concern for authors of these studies and thus was not frequently compared between simulations and crash tests. These simulation efforts were documented to identify factors potentially affecting ZOI, including barrier shape and height, vehicle-barrier friction, vehicle suspension failure, and tire deflation. Past ISPEs and anecdotal cases were reviewed to identify anecdotal severity and frequency of crashes involving fixed objects in the ZOI. Most state DOTs had not conducted ISPEs on attach- ments to rigid barriers or rigid barriers themselves. Existing ISPEs focused on general barrier performance and did not specifically relate to ZOI. A detailed ISPE through examination of individual crashes was not conducted as it would be difficult to determine (1) if a fixed object adjacent to a barrier was impacted, (2) barrier height and shape, (3) location of the fixed object laterally and longitudinally in relation to the crash, and (4) if the fixed object contributed to injuries. It was likely that crash frequency involving fixed objects behind a barrier was relatively C H A P T E R 9 Summary, Conclusions, and Recommendations

148 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments low, although no specific data confirmed this hypothesis. Fatal and severe-injury crashes were disproportionately likely to become news headlines, and therefore the anecdotal analysis con- ducted using search engine queries was inherently biased. Additionally, fewer crashes with fixed objects adjacent to barriers may occur as many transportation agencies adopt design standards and implement safety improvements consistent with the RDG (AASHTO 2011). Full-scale crash testing and simulation data for rigid barriers were reviewed and, along with the agency survey, analyzed to categorize each barrier and identify minimum and maximum barrier heights used at each test level. The recommended heights are shown in Table 6 in Sec- tion 2.8. Full-scale crash test data were also used to determine whether a functional relationship exists between barrier geometry and ZOI envelope. Generally, the available data favored some barrier shapes at specific heights with little data for similar barrier shapes at other heights; for example, few barriers that are consistent with F-shape or New Jersey shape profiles have been tested with heights above 42 in. or below 32 in. Based on the available data, no clear trends were observed between barrier geometry and ZOI envelope estimated from test data. 9.1.2 Agency Survey Thirty-one state DOTs completed an agency survey to gather information on barrier shapes and heights used at each MASH test level. Many state DOTs used ZOI criteria if necessary to place a fixed object adjacent to a barrier but avoided this placement when possible. Single-slope barriers were most common across all test levels; various barrier heights were used at each test level, many exceeding the minimum requirement. Thus, it was recommended to develop ZOI envelopes to accommodate the wide variety of barrier heights and shapes used by state DOTs. Some state DOTs noted objects such as luminaires, overhead sign structures, noise walls, and MSE walls were often mounted on or near barriers in the ZOI and therefore were contacted by vehicles. These crashes were generally not severe, but some fatalities and serious injuries have occurred. 9.1.3 Preliminary ZOI Envelopes The ZOI was evaluated from 47 rigid barrier crash tests to establish intrusion limits for each barrier class. The primary focus of the literature review was pickup truck tests, as they were observed to have higher ZOI intrusions than small car vehicles. As well, MASH test designation nos. 4-11 and 5-11, consisting of a 2270P pickup truck impacting at 62.1 mph and 25 degrees, were identical to MASH test designation no. 3-11 and were included in recommendations for MASH TL-3 ZOI envelopes. Of these tests, two were at MASH TL-2 conditions (test designa- tion no. 2-11), 31 were at MASH TL-3 (test designation nos. 3-11, 4-11, and 5-11), eight were at MASH TL-4 (test designation no. 4-12 only), and six were at MASH TL-5 (test designation no. 5-12 only). Data from two validated simulations of full-scale crash tests were also considered. 9.1.4 FEA Computer Simulation LS-DYNA simulations using existing vehicle models were used to fill gaps in existing crash test data to develop final MASH ZOI envelopes. To improve model stability, minor modifica- tions were made to the existing pickup truck, SUT, and tractor-trailer models. Although major updates were beyond the scope of this project, the implemented modifications improved model calibration by correlating more closely with some extremes of test data. These models were validated against full-scale crash tests using the V&V procedure when data were available (Ray, Plaxico, and Anghileri 2011). Once models were validated, other barrier configurations were simulated to supplement crash data. Barrier configurations commonly used by state DOTs,

Summary, Conclusions, and Recommendations 149   especially barrier heights near the maximum at each test level, were prioritized during simula- tion. Vertical, 10.8-degree single-slope and F-shape barriers were considered. MASH TL-2 simulations were conducted on barriers ranging from 24 to 54 in. tall. Although vertical barriers produced the maximum lateral extent for all barrier heights, the difference was not significant by barrier shape. Lateral extent decreased as barrier height increased, from a maximum of 15.8 in. for the 24-in.-tall vertical barrier to 0.0 in. for the 54-in.-tall F-shape and single-slope barriers. When data were available, F-shape barriers produced the maximum vertical extent. At 24-, 27-, and 29-in. barrier heights where F-shape simulation was not con- ducted, vertical barriers produced the maximum vertical extent. Maximum height above ground decreased as barrier height increased, from a maximum of 65.33 in. for the 24- and 27-in.-tall single-slope simulation to 0.0 in. for the 54-in.-tall F-shape and single-slope barriers. MASH TL-3 simulations were conducted on barriers ranging from 29 to 54 in. tall. Although vertical barriers typically produced the maximum lateral extent across barrier heights, the dif- ference was not significant by barrier shape. Lateral extent decreased as barrier height increased, from a maximum of 16.9 in. for the 29-in.-tall vertical barrier to a minimum of 0.1 in. for the 54-in.-tall vertical barrier. Although single-slope barriers typically produced the maximum ver- tical extent, the difference was not significant by barrier shape. Maximum height above ground decreased as barrier height increased, from a maximum of 76.9 in. for the 32-in.-tall single-slope barrier to a minimum of 55.4 in. for the 54-in.-tall vertical barrier. MASH TL-4 simulations were conducted on barriers ranging from 36 to 90 in. tall. MASH TL-4 ZOIs were separated into two zones based on the likelihood of occupant injury: a “cab zone” wherein the greatest likelihood of injury occurred for occupants in the vehicle cab; and a “box zone” that consisted of the extension of the SUT box below, behind, and above the barrier. The maximum lateral extent of the cab observed in simulation was 26.1 in. for 36-in.-tall F-shape barriers. Cab lateral encroachment decreased significantly with increasing barrier height, and no cab ZOI encroachment was observed at barrier heights of 48 in. or taller. Cab vertical extent was comparable among all three barrier types, reaching a maximum height above ground of 104.7 in. for 36-in.-tall F-shape barriers. Maximum lateral extent of the box was 76.4 in. for 36-in.-tall F-shape barriers. When averaged across the three barrier types, maximum lateral extent of the box for 42-in.-tall barriers was 54.1% lower than for 48-in.-tall barriers, with comparable results at 48-in. or taller barriers. Similar trends were observed for vertical extent, with a significant decrease from 42- to 48-in.-tall barriers. Maximum box height above ground was 164.9 in. for 36-in.-tall F-shape barriers. MASH TL-5 simulations were conducted on barriers ranging from 42 to 90 in. tall. The max- imum lateral extent of the cab was 13.3 in. for 42-in.-tall single-slope barriers. Single-slope barriers that were 42 in. tall also displayed the maximum cab protrusion above ground, 113.4 in. The maximum lateral extent of the box was 71.4 in. for 42-in.-tall single-slope barriers. Lateral encroach- ment decreased significantly from 42- to 48-in. barriers; when averaged across the three barrier types, maximum lateral extent of the box for 48-in.-tall barriers was 64.3% lower than for 42-in.-tall barriers. Maximum height above ground was 184.7 in. for 42-in.-tall single-slope barriers. 9.1.5 Development of ZOI Envelope Recommendations Data collected from rigid barrier crash tests and simulations were compiled to form ZOI enve- lopes and compared to the crash test data used to develop ZOI envelopes in the RDG (AASHTO 2011). Based on trends observed, the relationship between barrier height, and ZOI maximum width, a series of ZOI envelopes were created that best captured maximum practical results from simulations and full-scale testing. The ZOI envelope recommendations are shown in Chapter 6 and are recommended for updates to the RDG (AASHTO 2011).

150 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments 9.2 Conclusions Based on the literature review and simulation analysis, lateral extent of a vehicle past the initial front barrier face was highly dependent on barrier height. Lateral extent was not strongly dependent on barrier shape for most simulations, and no relationship could be determined from full-scale crash test data. SUT impacts with F-shape barriers produced the largest lateral extent compared to other barrier shapes. Many rigid tests conducted under MASH criteria were conducted with similar test article geometries but were associated with significantly different maximum protrusions into the ZOI. For example, maximum lateral extent across five 36-in.-tall single-slope MASH TL-4 barrier tests varied from 50 in. to 81 in. In contrast, computer simulation vehicles are modeled to accu- rately reproduce the physical behavior of a single vehicle in hopes that it is representative of other vehicles with comparable makes and models. Because empirical tests have significantly larger uncertainties for similar barrier geometries and impact conditions than the differences between barrier shapes could account for, recommendations for ZOI envelopes were not barrier- shape-dependent and were only related to MASH test condition and barrier height. Full-scale crash tests involving hardware placed within the ZOI were collected, and it was observed that the interaction between these objects and vehicle components can affect the max- imum predicted ZOI. In MASH TL-3 crash tests with rails, vehicle components were sometimes contained behind the rail or interacted with rail elements, causing increased lateral encroach- ment of the stiff vehicle components. Rail elements in MASH TL-4 crash tests appeared to provide stabilizing support to vehicles in some crashes. Although the effect on ZOI was not consistent by MASH test level, results also indicated that vehicle component encroachment into the ZOI may not result in heightened occupant risk. Full-scale testing should be performed to confirm that vehicle component interaction with features behind a barrier face is acceptable. Simulations were validated using full-scale crash tests to ensure the models accurately cap- tured the ZOI for impact events with various barrier shapes and heights. ZOI data were largely dependent on vehicle model accuracy; ZOIs collected with the pickup truck model were expected to be more accurate as this model was widely used, frequently updated, and the most recently developed model. Additionally, this model was representative of vehicles used in current full-scale crash testing. The SUT and MASH TL-5 models were not as commonly used nor as frequently updated; instead, each research facility updates and modifies them based on need. Because of this, the ZOI envelope predictions according to MASH TL-4 and TL-5 impact conditions have larger uncertainties. It was noted that in some MASH TL-5 crash tests, trailer damage likely caused regions of the box to extend farther laterally than if damage did not occur. The tractor-trailer model was not designed to accommodate severe trailer damage. Results of ZOI envelopes for severely damaged trailers were not included in the final recommendations. However, because trailer damage may occur in real-world crashes, agencies are encouraged to consider if such damage may result in catastrophic outcomes to either infrastructure or the roadside environment near a crash loca- tion, such as an adjacent pedestrian walkway. 9.3 Recommendations During the course of this research study, several recommendations were developed which would improve the (1) collection, (2) accuracy, and (3) utilization of ZOI data obtained from full-scale crash testing. A summary of the recommendations is provided below. • Full-Scale Testing Measurements, Photography, and Reporting – Align cameras with front top edge of barrier (centered), orthogonal views. – Place additional targets on vehicles to promote improved tracking and calibration.

Summary, Conclusions, and Recommendations 151   – Orient in-suspension dynamic camera views to document wheel movement, suspension failure progression, and timing. – Determine limits of vehicle extension into ZOI; include a diagram with the test summary page. – Document top and bottom ground heights of ballasted TL-4 box. – Document box-to-chassis attachments for TL-4 vehicles (e.g., number, location, size of box-to-chassis attachment bolts). – Measure top and bottom ground heights of ballasted TL-5 trailer. – Photograph the fifth wheel before and after the test. – Record style, make, type, and chassis attachment of large-truck fifth wheel. – Note suspension type and photograph pre-test for TL-4 and TL-5 vehicles (e.g., mechanical, pneumatic/air ride, electrical). – Document and report materials of vehicle external panels (e.g., fiberglass, aluminum). – Document ZOI envelope shapes from full-scale testing using figure examples from this report. • Computer Simulation – Update suspension models for pickup truck vehicles. – Improve connections between components of vehicles to better represent deflections into the ZOI. – Consider new model(s) of fifth-wheel attachments (fixed, sliding, pneumatically adjustable). – Consider new truck models: MASH TL-4 and TL-5. – Consider new MASH TL-5 trailer model. – Consider publicly available MASH TL-6 model with fluid slosh capability. – Use accepted and community-affirmed friction values for tire-barrier, sheet metal-barrier, and sliding friction for standardized vehicle-barrier interactions. • Further Full-Scale Crash Testing Needed – Test signs, luminaires, and breakaway features located within the ZOI. – Study alternative material property barriers: timber, stone, MSE. – Research luminaire, pole, bridge structure, infrastructure, and nonbreakaway hardware located within the ZOI. – Collect ZOI data for TL-6 vehicles with different barrier heights. – Test with alternative vehicles [e.g., newer-generation pickup truck vehicles, updated MASH test vehicles per NCHRP Project 20-07 (Task 372), electric or hybrid vehicles]. • In-Service Performance Evaluations (ISPE) – Identify examples of real-world crashes with objects located in the ZOI. – Evaluate the distribution of injuries for similar types of impacts involving and not involv- ing objects located in the ZOI. The final ZOI envelopes presented in Chapter 6 should be considered for the next revision of the RDG to guide the placement of objects near rigid barriers. These recommendations only apply to rigid systems, which are defined as systems for which dynamic deflection does not exceed 10 in. Vehicles used to test systems with larger dynamic deflections will behave in ways not considered in the ZOI development in this study. Most lateral extents collected in rigid bar- rier data analysis were measured from the front barrier face before impact; therefore, in MASH TL-3 systems with high dynamic deflections, it is possible the barrier itself will occupy a signifi- cant portion of the ZOI. Caution should be taken when these ZOI recommendations are used to place objects near transitions. If possible, the placement of objects near transitions should be avoided as vehicles respond uniquely when interacting with barrier elements of varying stiffness. Crash tests of guardrail-to-concrete transitions displayed increased lateral extent as post deflection allowed the vehicle to encroach farther past the pre-impact front barrier face. If these recommendations are used for a concrete-to-concrete transition, ZOI recommendations should be followed for the barrier that produces the largest intrusion. The more conservative “working-width” parameter

152 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments may be more appropriate for barriers with large dynamic deflections, including transitions and adjacent systems; note that working width only denotes a maximum width and extends verti- cally beyond the vehicle. If a rigid barrier has been determined to meet MASH 2016 testing requirements (AASHTO 2016), the placement of hardware outside the ZOI envelope for that system does not require further testing. When possible, placement of rigid elements within the ZOI should be avoided; if this placement is necessary, crash testing is recommended to verify the hardware placement does not pose a risk to the occupants. If applicable, simulations should be conducted on systems with hardware within the ZOI. These recommendations apply to rigid and breakaway hardware. Modifications to full-scale crash test camera setup and documentation were proposed to assist accurate measurement of vehicle intrusion. At a minimum, cameras should have a sufficient view of the space above and behind the barrier, and overhead cameras should have a sufficient view downstream from the impact point. This will ensure all vehicle movements are captured in each camera view. Vehicle model improvements discussed in Chapter 8 should be further investigated to improve the accuracy of ZOI projections. RAM model front-end connections should be assessed to verify they do not adversely affect the lateral extent. It is recommended a new SUT model be developed that better represents the current vehicle fleet, and geometric, mesh density, and suspension modeling concerns should be addressed. Further research into the TL-5 tractor-trailer connection is recommended and should determine if the current fifth-wheel model accurately emulates behavior from full-scale testing.