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.

133   8.1 Overview In addition to the collection and recommendation of ZOI envelope data for use in the update to the RDG, researchers collected recommendations that would be useful for expediting and increasing the accuracy of future revisions involving ZOI determination. The following repre- sents recommendations for best practice for roadside safety laboratories and computer simulation experts to improve the fidelity of future results. 8.2 ZOI Measurements and Working Width Throughout this research effort, researchers identified boundaries corresponding to the estimated ZOIs based on the movement of stiff vehicle components during an impact event. These measure- ments were rarely extracted directly from technical summary reports or journal papers because the published values were most consistent with working width, dynamic deflection, or vehicle position and did not include the heights of components generating the working widths. It is recommended that standardized working widths reported in summary reports be revised to include a figure show- ing the combined ZOI/working width combination, such as the one shown in Figure 124. Note that the results of this study excluded all tests with barrier deflections greater than 10 in. As a result, the ZOI envelopes shown in this study are intended for use in combination with rigid barriers. Reporting the working width similar to the method shown for reporting ZOI in this study will result in standardized outputs that transcend rigid barriers and apply to semi-rigid and flexible barriers. Adding the additional detail may increase costs and effort associated with reporting but will greatly improve the accuracy and usability of ZOI data. 8.3 Camera Positioning for ZOI Measurement Optimizing camera positioning is critical for improving accurate ZOI measurement from future full-scale crash tests. Overhead views are typically focused on the impact point; however, maximum lateral extent sometimes occurs far enough downstream from the impact point such that the overhead view is obscured. This can result in inaccurate estimates for ZOI and work- ing width for MASH test designations no. 4-12 and 5-12 due to the size of the test vehicles. The overhead camera should be positioned high enough above ground to capture vehicle-barrier interaction downstream from impact without significantly impeding the accuracy of the view. Enough space behind the nontraffic side of the barrier should be in view to ensure the lateral extent does not occur laterally out of view. In some cases, view distortions occurred in which dimensions of the vehicle appeared dif- ferent at different locations in the camera view. For example, for full-scale tests involving the C H A P T E R 8 Test Documentation and Modeling Improvements Recommendations

134 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments 10000S TL-4 vehicle, the cargo box corners could appear wider near the top than at the bottom even though cargo boxes have a constant width, as shown in Figure 138. During impact, it appeared as if the top box corner is protruding over the barrier front face (Figure 138a), even though the downstream view indicates that the top-front box corner was far from the barrier’s front face (Figure 138b). It is recommended that the location of the overhead camera be closely aligned with the primary contact face of the barrier, such as the leading top edge of the barrier system. Overhead camera views angled such that the front barrier face is visible will exaggerate lateral intrusions, while overhead camera views angled such that the barrier’s back face is vis- ible will suggest lower lateral intrusions. While this can be compensated for using high-speed digital video analysis, lens corrections, and frame realignment techniques, the difficulty and time required to perform these steps were beyond the scope of this project. Therefore, to expedite future analyses involving ZOI or working width measurements, the researchers recommend that the camera be centered vertically with a reference location on the barrier corresponding to the primary impact surface, the top traffic-side edge of the barrier. Upstream and downstream camera positioning is extremely important for ZOI measure- ments as these are the only frames where vertical measurements can be collected. At a mini- mum, the upstream camera view should be positioned to record accurate ZOI measurements. These camera views should be colinear with a theoretical axis positioned at the center of the downstream barrier face that spans longitudinally with the barrier length. While some deviation from the nominal alignment with the barrier may be desired for observing different events that occur during impact, these deviations decrease the accuracy of ZOI measurements. If the camera view axis is angled relative to the test article’s horizontal plane located at half the barrier height, vertical intrusion will be distorted. Additionally, if the camera view axis is angled about the test article’s vertical plane, lateral intrusion will be distorted. Downstream or upstream cameras positioned to allow accurate lateral intrusion measurements are important so that these mea- surements can be compared with overhead data. It is recommended that at least two views be precisely aligned to ensure accurate ZOI and working width measurements that can be recorded graphically and used in future iterations of this study. Figure 139 shows examples of potential camera alignment for upstream and downstream (139a) and overhead views (139b). (a) (b) Figure 138. Overhead camera distortion, test. no. 110MASH4S19-02 (Her et al. 2017).

Test Documentation and Modeling Improvements Recommendations 135   8.4 View Calibration and Documentation Improvements ZOI measurement using full-scale crash testing video analysis requires calibration to an object in the camera view with a known dimension. During the video analysis of upstream and downstream video footage conducted in this study, measurements were calibrated to the dimen- sion on the downstream and upstream faces of the test article. However, measuring distances far from the measurement calibration location causes distortion as the vehicle appears larger as it approaches the camera. Although this was an acceptable calibration method, a better option is to calibrate video measurements to dimensions on the vehicle during the ZOI time. This could not be done to calibrate upstream and downstream views in this study due to the absence of vehicle measurements, particularly for crash tests with heavy vehicles. Currently, the box length, width, and height above ground are documented, but the height of the box of 10000S vehicles is not typically included in vehicle dimension summaries. It is recommended that box heights be recorded and included on vehicle dimension summary sheets, both for use in calibrations and for evaluating differences in vehicle dimensions and their effects on ZOI envelopes, vehicle fleet evolution, and vehicle roll behaviors. One procedure that test facilities use to gather accurate measurements is to place targets on the vehicle with known distances. These targets are useful when dimensions are available since video measurements can be calibrated on the vehicle and have less distortion than calibrat- ing to an object closer to the camera view. Targets have been observed on the roof of vehicles, which aids in calibrating overhead view measurements. Longitudinal target placement has been observed but cannot be used in ZOI analysis since the vehicle is angled relative to the camera and yaws substantially during impact. Additional targets placed vertically on the vehicle would allow better calibration of the downstream and upstream views relative to a component on the vehicle. Generally, testing laboratories have not recorded views of impacts from within suspension compartments due to logistical difficulty, the uncertainty of need, lack of synchronization with Figure 139. Recommended camera alignment for ZOI measurement. (a) (b)

136 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments external views, the potential for damage, and little correlation with occupant survivability. MwRSF researchers have used high-speed and suspension-mounted video to estimate component failure times during full-scale crash testing for use in tuning simulation models. Post-test damage photos are frequently used to document the suspension damage, but the time of damage and contributing circumstances may not be known. When feasible, it is recommended that suspension documen- tation be collected through camera views mounted inside the wheel wells to gain a better under- standing of the timing and sequence of events contributing to suspension failure. An example of a suspension documentation photograph is shown in Figure 140. Likewise, it is recommended to monitor damage to and potential failure of the tractor-trailer fifth-wheel attachment during tractor-trailer testing. This is especially important since this connection is not clearly visible in any typical camera view. 8.5 Large-Truck Documentation Improvements It is recommended to measure the height from the ground to the front and back corners on both sides of the boxes of 10000S vehicles and the trailers of 36000V and 36000T tractor-trailer vehicles. The box or trailer height above the ground is critical since this dimension affects the dynamics of the impact event, particularly during impact with barriers 42 in. tall or shorter. Due to air ride suspensions and varying frame geometries, vehicles that are consistent with MASH (AASHTO 2016) dimension and height criteria may still have different heights at front and back corners and may even vary from left to right. On impact, a lateral load is applied to the vehicle, creating a moment about the c.g. caused by the distance between the impact load application on the vehicle and the vehicle c.g. height. When the bottom box or trailer corner is below the barrier-top surface, the box or trailer side slides across the front barrier face, and the box or trailer can pivot near the front of the barrier. If the box or trailer edge extends over the top surface of the barrier, the effective pivot point may be on top of or even behind the barrier, which has a significant effect on vehicle roll and ZOI encroachment. Further recommendations for documentation improvements are noted below. 8.5.1 10000S Documentation For 10000s vehicles, documentation of accelerometer locations and attachment of the cargo box would improve the consistency of results. Accelerometer locations are currently documented; however, additional documentation on their mounting attachments and locations should be Figure 140. GoPro suspension camera view (Stolle et al. 2014).

Test Documentation and Modeling Improvements Recommendations 137   provided as methods of attaching accelerometers to vehicles and the stiffnesses of those attach- ments may vary between different laboratories. Some facilities mount accelerometers to the cargo box wooden floor, while others mount them to a lateral brace located between the frame rails. These differences may affect the accelerometer traces used in computer simulation verifica- tion and validation procedures. Most testing facilities do not provide documentation on the attachment of the cargo box to the chassis for SUTs. The box connection type, number of connections, and location of box- to-chassis connections affect the amount of box roll relative to the truck. Figure 141 illustrates the separation of the box bed rails, colored pink, from the truck frame rails, colored brown, for an SUT simulation. Details of these connection methods should be documented and reported by testing facilities. Although standard locations for attachments are recommended by vehicle guides, there is still a need to report and confirm the locations, sizes, and number of attachments used. If it is shown that variations in sizes or standards exist between different vehicle makes and models, standardizing attachment locations, numbers, sizes, and methods may be warranted to improve consistency. 8.5.2 Tractor-Trailer Documentation Tractor-trailers are typically fitted with either a stationary or sliding fifth-wheel mechanism, as shown in Figure 142. The stationary fifth wheel is bolted directly to the tractor frame, while the sliding fifth wheel consists of a rail bolted to the tractor frame and a mount that latches a lock mechanism. The type of fifth wheel a test vehicle is fitted with could affect the amount of trailer roll due to the compliance of these mechanisms and their attachments to the vehicle. Addition- ally, certain fifth wheels could be more likely to fail during an impact event. This information is currently unknown due to a lack of documentation. The review of tractor-trailer crash tests indicated significant trailer roll occurs for some tests, specifically at 42-in.-barrier heights, but it is unknown if this is due to the fifth-wheel connection itself or the twisting of components near this interface. Test facilities should document the type of fifth-wheel connection fitted on test vehicles and include post-test damage photos in their documentation. This will improve researchers’ ability to accurately model these connections in crash test simulations. 8.6 Considerations for Future ZOI Studies This research effort extensively utilized available video and working width data when estimat- ing ZOI envelopes. The database of full-scale crash test and simulation data should continue to be maintained and updated following this study. ZOI envelope recommendations may be adjusted as more information is obtained from crash testing configurations with different heights. Figure 141. F800 model box rails separation from frame rails.

138 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments Most SUTs and tractors use fiberglass hood and fender panels, which typically have a low shear loading capacity and tear during full-scale crash testing. Because of this, the intrusion of these components past the front barrier face may not pose significant snag concerns for the cab compartment. Future ZOI studies that consider MASH TL-4 impacts should study the interac- tion between these components and rigid elements to identify snag risk. Crash testing under MASH criteria typically used RAMs for pickup truck tests, and therefore most pickup truck crash test data in this study consisted of RAM trucks. Although these trucks meet MASH requirements, it is not certain if they provide the most critical evaluation of snag on structures located in the ZOI. Chevrolet and Ford trucks typically have more planar hoods com- pared to the RAM, especially older RAM models on which the hood does not extend to the front corners. This may cause hoods on Chevrolet or Ford trucks to extend further laterally past the front barrier face than the curved, shallow RAM hoods. In Chapter 2, two crash tests conducted with Chevrolets generated a slightly larger ZOI. In circumstances where ZOI engagement could be critical, it may be important to consider the effects of a vehicle model on engagement with structures in the ZOI. 8.7 RAM Model Improvements 8.7.1 Steering Mechanics During pickup truck model validation, researchers became aware of two discrepancies related to the modeled steering system, which are not consistent with truck vehicles in production. Figure 142. Stationary tractor-trailer fifth wheel (top) (Truck Parts Inventory n.d.) and sliding tractor- trailer fifth wheel (bottom) (My Little Salesman n.d.).

Test Documentation and Modeling Improvements Recommendations 139   First, through sequential comparisons, it was discovered the simulated vehicle’s wheels steered away from the barrier shortly after impact. This behavior was not observed in full-scale crash testing. Test no. H34BR-2, an example of a test performed according to MASH test designation no. 3-11, is shown in Figure 143 (Bielenberg 2019). Recently tested RAM trucks were equipped with electric and hydraulic steering assist systems, which only function properly when power is supplied to the vehicle systems. Electrically actuated steering assistance systems have pro- liferated in the development of Advanced Driver Assistance Systems, including lane-keeping assistance and crash avoidance system technologies, and it is unclear what effect power steering has on wheel behavior during impact. The lateral wheel behavior in simulation models has not replicated real-world wheel reactions for power-off vehicles. In addition, deformations of the steering arms in the pickup truck model are unrealistic, as shown in Figure 144. Steering rods were modeled with 2-mm shell elements although the steer- ing arm has a solid, circular cross section. Recent computer simulation efforts performed at MwRSF have implemented alternative modeling structures for the steering arms to better reflect their axial stiffness. It is uncertain if the suspension inconsistencies affected the ZOI envelope estimates presented in this study. Revised models of the pickup truck with updated suspension Figure 143. Test no. H34BR-2 (Bielenberg 2019) and simulation wheel steer comparison. Figure 144. Pickup truck model steering arm deflection.

140 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments component properties tuned to match the compressive, tensile, and bending response of the realistic components are recommended. 8.7.2 Front-End Connections Small bolts and plastic clips are commonly used to connect plastic components such as head- lights, grilles, and other trim pieces in modern cars. Similarly, metal fenders are connected to the vehicle using small bolts. In the pickup truck model, these connections were modeled using constrained nodal rigid bodies, as shown in Figure 145, which did not fail and allow component separation during impact. These connections may have affected the predicted ZOI envelope intru- sion. In the simulation shown in Figure 146, the fender achieved the maximum lateral extent, but its intrusion could be restricted by its connections to the headlight, grille, and plastic trim, each of which typically detaches in full-scale crash testing. These parts should be constrained to the vehicles using methods that allow detachment. 8.8 SUT Model Improvements 8.8.1 Vehicle Model Comparison to Test Vehicles SUTs used in crash testing were compared to the F800 model used in simulation, as sum- marized in Table 22. The model’s overall height was comparable to older SUTs, but newer test vehicles were found to be up to 11 in. taller for some models. Cab height, roof-to-hood distance, and roof width varied between test vehicles and the vehicle model. Most test vehicles in this data set were manufactured by International, but the simulation model was a Ford F800, as shown in Figure 147. Based on the results of this study, further research to investigate SUT properties and develop a new SUT vehicle model to better match the current vehicle fleet is recommended. As discussed previously, most SUTs and tractor-trailer test vehicles have fiberglass hood and fender panels, but the F800 model’s hood and fender panels were steel. Furthermore, since the SUT model reflects the properties of a 1996 vehicle, its suspension prop- erties may be outdated and should also be reviewed. Figure 145. Pickup truck model front-end rigid body connections.

Test Documentation and Modeling Improvements Recommendations 141   Figure 146. Pickup truck front-end interaction with a barrier. Test No. Test Vehicle Vehicle Dimensions (in.) Height Cab Height Roof-Hood Distance Roof Width Wheelbase FEA 1996 Ford F800 134.0 88.0 20.0 58.8 208.2 420020-9b 1991 International 4700 133.0 96.0 26.5 73.0 187.5 467491-1-1 2003 International 4200 134.0 98.5 30.0 71.0 206.0 110MASH4S16-03 2004 Freightliner M2 150.3 97.8 33.5 70.9 234.9 490027-2-1 2004 International 4300 133.0 98.5 30.0 71.0 201.0 469467-3-1 2005 International 4300 145.0 98.5 30.0 71.0 207.5 4CBR-1 2005 International 4300 146.1 98.9 30.1 72.8 229.5 STBR-4 2007 Freightliner M2 106 153.0 100.0 34.0 71.9 216.3 469689-3-3 2009 International 4300 143.5 98.5 30.0 71.0 204.8 469680-2-3 2009 International 4300 151.5 98.5 30.0 71.0 204.8 608331-1-1A 2011 International 4300 143.0 98.5 30.0 71.0 204.8 MNCBR-1 2013 International DuraStar 4300 146.0 98.9 31.0 70.5 236.8 Table 22. SUT crash test vehicle dimensions (Bligh, Menges, and Kuhn 2018; Sheikh, Bligh, and Menges 2011; Pena et al. 2020; Rosenbaugh et al. 2021; Her et al. 2017; Williams, Sheikh, et al. 2018; Sheikh et al. 2020; Williams, Menges, and Griffith 2019; Hinojosa et al. 2021).

142 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments 8.8.2 Mesh Density Many SUT vehicle components have a coarse mesh density, including critical components, such as the gas tank, frame rails, and cab compartment. Components with coarse meshes may not accurately replicate the behavior of physical components during full-scale crash testing. Coarse mesh densities are also prone to producing numerical instabilities, especially when contacting other components. Components of the F800 vehicle model with coarse mesh densities are shown in Figure 148. Note that researchers utilized an updated model of the F800 that was modified at MwRSF, and many coarse meshes of front cab components were remeshed with a finer mesh. These components are not shown in Figure 148. Figure 148. SUT model components with coarse mesh densities. Figure 147. 2016 International 4300 (Ryder n.d.) (left) and Ford F800 model (right).

Test Documentation and Modeling Improvements Recommendations 143   (a) (b) Figure 149. Test no. 4CBR-1 (Rosenbaugh et al. 2021) (a) and FEA wheel damage (b). The F800 SUT model’s wheels were also coarse with an elastic material definition for the tires and rigid definition for the wheel rim. In full-scale crash testing, wheel rims typically experience plastic deformation, particularly around the wheel lip (Figure 149a). The current wheel model does not capture this deformation, as shown in Figure 149b. This could be problematic when modeling rigid barrier testing due to the impact of two relatively rigid parts. During this research study, deformable rims were investigated as a means of modifying rear suspension “tail slap,” but results did not vary significantly for deformable or rigid rims. It is believed that the coarse, simplified tires may produce more uncertainty in the model than the rigid rim. It is recommended that if a new or updated vehicle model is produced, wheel models for large trucks should also be improved. 8.8.3 U-Bolt Failure and Suspension Modeling Most full-scale crash-tested SUTs and tractor-trailer vehicles experience some level of front suspension damage when impacting rigid barriers. Research facilities use different methods of modeling and incorporating failure into these connections. The behavior of the connections is calibrated using high-speed digital video analysis tools and accelerometer data. A revised U-bolt model calibrated with physical component testing is recommended to standardize the U-bolt connection and improve modeling capabilities. As discussed, the SUT model was created to reflect a 1996 Ford F800. The modeled suspension should be evaluated to verify that its stiffness properties match current SUT vehicles. 8.9 Tractor-Trailer Model Improvements Researchers determined that for the purposes of estimating the lateral encroachment of the trailer and cab during simulations, a 2011 tractor-trailer model achieved ZOI values that more closely agreed with full-scale crash test data than an updated 2016 model. Therefore, the 2011

144 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments model was used to simulate the ZOI of all the MASH TL-5 rigid barrier impacts. It was observed that the main benefits of using the 2011 model were associated with increased trailer roll angle, as shown in Figure 150. The increased trailer roll was attributed to the tractor’s rear wheels lift- ing higher during the simulation with the older tractor-trailer model, evident by the separation between the front nonimpact-side trailer corner and the trailer’s rear wheels. Further investigation was performed to evaluate the differences between the older and newer truck models. One significant difference was related to the methods used to represent the fifth- wheel constraints, as shown in Figure 151. The 2016 model used a nodal rigid body attached to the fifth-wheel receiver and joints below the fifth-wheel receiver. In the updated 2016 model, additional spot welds were used to attach the lower fifth-wheel plate to the frame rail mounts. Because the older 2011 model used fewer connections in the fifth wheel, it resulted in decreased rigidity and increased component deformations of the fifth-wheel frame, which correlated better to large lateral intrusions of the trailer. In full-scale tractor-trailer crash testing, tractor frame twisting and fifth-wheel deformation were often observed; however, it was unknown if these behaviors were properly captured in the simu- lation. Three fifth-wheel failures are shown in Figure 152. Note in test no. MAN-1 (Rosenbaugh 2016), fifth-wheel detachment occurred due to additional impact with another concrete barrier downstream from the test installation and was not caused by the test article. A follow-up study is recommended to identify factors contributing to tractor roll and validate model fifth-wheel deformation. Currently, no fifth-wheel models based on modern fifth-wheel constructions have been made publicly available, but model accuracy and connection behavior may improve with a more detailed model. More accurate fifth-wheel models are recommended to improve confi- dence in simulated trailer dynamics. It should be noted that in a separate research effort, MwRSF began investigating fifth-wheel components and generated an example fifth-wheel model that will be made publicly available when completed. The fifth-wheel model in development is shown in Figure 153. (a) (b) (c) Figure 150. Tractor-trailer roll, test no. 510605-RYU1 (a) (Buth and Menges 2012), 2011 model (b), and 2016 model (c).

Figure 151. 2016 (top) and 2011 (bottom) fifth-wheel models. CNRB = constrained nodal right body. (a) (b) (c) (d) Figure 152. Fifth-wheel damage, test nos. MAN-1 (a and b) (Rosenbaugh et al. 2016), 510605-RYU1 (c) (Buth and Menges 2012), and TL5CMB-2 (d) (Rosenbaugh, Sicking, and Faller 2007).

146 Zone of Intrusion Envelopes Under MASH Impact Conditions for Rigid Barrier Attachments Figure 153. Example of fifth-wheel model in development at MwRSF (a) and lidar scan data (b) CAD geometry of fifth-wheel saddle plate and mounting bracket. (a) (b)