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.
55 Chapter 3 Assessment of Vehicle Characteristics In order to identify vehicle body styles and structural characteristics which were influential during crashes with roadside systems, a review of full scale crash tests was conducted. This review provided a clear indication of the roadside systems that performed best under a series of test conditions with the chosen NCHRP test vehicles (i.e. 820kg and 2000kg vehicles). The review provided the research team with an understanding of the characteristic behavior of vehicles during these crash events. Since only these two vehicle classes have been observed during tests, little was learned about the vehicle attributes that influence crash performance during roadside impacts. Influential characteristics would be recognized if two tests of identical roadside systems were conducted using different impacting vehicles. Under these conditions, a direct comparison of geometric and dynamic vehicle properties indicates possible sources of incompatibility. Alternative methods to study the effect of vehicle attributes on compatibility with roadside hardware using analytical modeling of vehicles and barrier systems is included in Section of this report. Information regarding characteristics of passenger vehicles that are influential during vehicle crashes with roadside structures was gathered through individual crash case reviews shown in Chapter 2 of this report and through information compiled during the literature review for this project. These sources provided the basis for the following list of vehicle attributes that potentially are influential during roadside hardware crash events. 1. Vehicle mass 2. Height of vehicle front structure and profile 3. Stiffness and geometry of vehicle front and side structure 4. Frontal overhang ahead of front wheels 5. Front and rear suspension characteristics 6. Vehicle door rocker geometry 7. Vehicle door latch/structural geometry 8. Vehicle wheelbase
56 9. Vehicle Static Stability Factor In addition, the literature review provided insight into the most appropriate characteristics which should be considered to assess vehicle performance during roadside crashes. A comprehensive FHWA project conducted by the Texas Transportation Institute (TTI) was reviewed. The objective of this project was to develop protocols that could be used to identify compatibility issues caused by changes in the future motor vehicle fleet. The final report of this project included many relevant findings and recommendations regarding vehicle compatibility with roadside hardware. Some highlights from this project are shown below [2]. 1. Vehicle platforms will be face lifted every 3 - 4 years with new platforms every 5 - 7 1/2 years. A protocol needs to be in place to categorize the vehicle fleet to assess the level of performance. 2. Light truck population will continue to increase from its current exceedance of 50% of total vehicle markets. The greater vehicle height which is unregulated will make vehicle stability a continuing concern. 3. Curb weight and size of the 820 kg class vehicle will continue to increase requiring a selection of heavier vehicles for the lower weight class. 4. Driver and passenger side airbags will approach 100 percent in the next decade. It may be appropriate to consider this and the increased safety restraint usage (i.e. over 70% seat belt usage) when evaluating roadside hardware. 5. Recently introduced crumple zones in light truck subclasses have shown a significant reduction in occupant compartment deformation. 6. Vehicle manufacturers are producing less full-size passenger cars. 7. Market share of the two midsize car platforms continue to increase above the two small car platforms. 8. Large pickups (1/2 ton and 3/4 ton) continue to dominate the sub-class in terms of market share among light trucks. 9. Some of the more significant characteristics identified are: Total mass, front overhang, height of vehicle center of gravity, suspension height, bumper height, geometric profile, and frontal crush stiffness.
57 10. Because wheelbase, weight, overall length, overall width, and front track width were highly correlated, by retaining one of them, all of the statistical information contained in original data was preserved. Many of the vehicle characteristics highlighted by the TTI study were further reviewed to understand their correlation with real world crash outcomes and results of full scale tests. Further, it was determined that a thorough survey of the current vehicle fleet to understand the variability and range of characteristics that exist today was necessary. The following section outline the methodology used to gather those relevant characteristics. 3.1 Geometric Characteristics During the literature survey portion of this project, trade magazines and engineering resources were compiled to document a series of vehicle characteristics for US model automobiles. Some of those resources include: The Mitchell Automotive Repair Series by Mitchell Automotive and the "Consumer Review 2001 Car Prices" By Harris Publications. The Mitchell Series documents dimensions of all vehicle sub structures for body shop repair professionals. The "Consumer Review" Magazine documents consumer information such as vehicle weight, height, wheelbase and engine type. Following review of these resources, a large amount of data was compiled however, a series of critical vehicle attributes were still unknown. Since this necessary data was not available directly from the manufacturer, the research team performed measurements by hand on a large number of new and used vehicle structures. Those attributes and procedures for these measurements were conducted as follows. 1. Frame rail spread- The frame rail spread is the distance between the left and right frame rails. When viewing the car from the front, this measurement is taken from the inside of the left frame rail to the inside of the right frame rail at the point closest to the front of the car possible. This vehicle attribute is important during oblique and frontal impact events. During oblique impacts, including interaction with longitudinal barriers, the proximity of this stiff body structure to the impacting device often dictates the acceleration and crush profile exhibited by the body structure. A soft outer body structure surrounding frame rails positioned well inboard (close to the vehicle longitudinal centerline) often leads to high body deformation and a high likelihood of snagging with barrier systems. Conversely, if the stiff vehicle structure is positioned more outboard, the stiff vehicle structure will engage with the rigid or flexible barrier without absorbing large amounts of impact energy. Higher levels of lateral acceleration result in this case.
58 During frontal impacts with narrow objects, the position of these frame rails is important when considering optimal engagement of the pole/post with rigid structure (engine) or deformable structures (rails). 2. Bumper structure (lower and upper)- The bumper structure is defined as the hard portion of the bumper that will not deform in a minor accident. Usually the bumper structure is made of steel or a hardened plastic. Foam and light plastic have less significant effect on the impact and are not included in the bumper structure measurements. In some cases, when the vehicle could not be disassembled or direct measurements of the front bumper structure could not be performed, the actual bumper structure height was estimated by the measurement of the outer fascia. The bumper structure location as well its overall height could have significant effect on the outcome of a crash. The bottom and top aspects of the bumper structure are important to determine the approximate region of first engagement with guardrail devices. These beams or U- shaped channels are responsible for transferring a large percentage of loading during frontal impacts to the vehicle structure before crushing occurs. The size (height) of the structure is important during pole impacts to understand the likelihood of pole bending, fracture or collapse as well as the likelihood of release of breakaway devices during those impact conditions. 3. Bumper fascia (lower and upper)- The bumper fascia is defined as the continuous metal or plastic cover surrounding the bumper structure. The measurements of the fascia are always taken at the center of a vehicle from the ground to the upper most and the lowest point on the front fascia. These measurements do not include structures such as chin spoilers unless these spoilers are directly cast into the fascia (i.e. it does not include bolt on spoilers). If the grill is continuously integrated into the bumper fascia, the measurements are taken to the top of the grill. However if there is a gap between the bumper and the grill, the measurements do not include the grill area. The geometry of this fascia is important to determine the likelihood of post snagging with the vehicle structure. Also, this "flexible" structure that is often plastic gives the impression that impact forces will be distributed over a larger area than the bumper structure explained above. 4. Rail height (lower and upper)- The rail height is the height of the frame rail measured at the most forward point possible. The frame rails are two longitudinal members that carry most of the
59 frontal impact force during impact. These rails are often tubular, box or c-channels welded to the vehicle structure in the case of unibody constructed vehicles. The dimensions of these members are important to understand the probable center of force that results during frontal impacts with a wide variety of devices. The lowest and upper-most points on the frame rail will indicate the likelihood of favorable interaction with guardrails, end terminals and semi-rigid longitudinal barriers during high-energy impacts. Often during these types of impacts, the outer body and bumper structure collapse and all remaining engagement with the barriers occurs with the engine or frame structures. 5. Free Space- Free space is measured from the aft most point of the radiator to most forward hard point of the engine. Hard points are defined as engine components and frame components (Plastic fans, belts and pulleys are not considered hard points in this measurement). If the engine protrudes underneath the radiator, the free space is defined to be 0. This dimension is important during frontal impact with narrow objects and partner vehicles. Often vehicle crash sensors deploy airbags based on sudden deceleration of the vehicle structure. Usual deceleration levels experienced by the vehicle during deformation of the bumper structure and the radiator often fail to trigger airbag sensors. The larger the free space is, the later the airbag deployment will occur. If the sensors do not trigger airbags before the pole structure begins interacting with the engine block, a sudden peak in deceleration forces will take place leading to airbag deployment. In some cases, the occupant has moved forward or out of position relative to the deploying airbag causing an unfavorable late deployment crash scenario. During interaction with partner vehicles, a large amount of free space creates a more favorable situation for impacted vehicles as this region is more compliant than the engine block itself. 6. Frontal Overhang- The frontal overhang is the distance from the lowermost potion of the front fender to the most forward position of the vehicle. This gives an indication of the exposure of the wheel, suspension and power train to objects struck during frontal impact conditions. The ride height combined with the front overhang dictate the level of interaction seen between impacted and rotating tires/suspension structures. In the case of Pickup Trucks and SUVs, a short frontal overhang and higher ride height often lead to higher potential for snagging with guardrail posts and rail members themselves. This condition is prevalent during guardrail impacts with
60 pickups and may occur during impacts between barriers and similarly configured sport utility vehicles. 7. (Window) Sill Length- The Sill length is measured from the front most position of the lower portion of the driver's side window to the rear most position of the driver's window. If the rear view mirror is incorporated in the main frame of the window, the measurement begins at the beginning of the rear view mirror housing. During crash events with narrow objects (posts or poles) or end terminals in a side impact configuration, the length of the door or window sill will indicate some potential for occupant compartment intrusion. A door structure securely fixed at door hinge points and the door latch point which are closer together are likely to resist intrusion well. Conversely, a structure where these points are further apart often has a more compliant door allowing for greater levels of intrusion. Also, as the ratio of windowsill length to total vehicle body length increases, the likelihood of contact between the deforming door and nearside occupants also increases. 8. (Window) Sill Longitudinal Location- The longitudinal location is the distance from the gap between the hood and the front fascia/fender and ending at the lower portion of the driver's side window. This measurement indicates two characteristics. First, this distance provides a metric for location of the front door versus the front of the car. Second, the distance from the front most impact point to the base of the windshield can be estimated as well. During frontal impacts with small sign support structures, the likelihood of contact between the sign blank and the windshield are a direct function of this distance. Other factors that indicate this are vehicle bumper height, ride height and vehicle mass. In some cases the sign blank strikes may strike the hood, the roof or the windshield. Windshield contact is least desirable. 9. (Window) Sill Height- The sill height is the height from the ground to the lower part of the driver's side window. This measurement is taken at the rear most portion of the driver's window. Plastic sheathings are not included in the measurement of sill height. This metric provides an estimate of occupant head position in the event of a side impact. A life threatening situation exists if the occupants head strikes and breaks the driver side window during a near side collision. In this situation, there is potential for contact of the head with the stiff
61 impacted device. This information is critical to properly determine barrier heights including longitudinal and end terminals in use. 10. Rocker height (lower and upper)- The measurement for the lower rocker height is taken from the ground to the beginning of the rocker panel. This height does not include the jack mount points or the rail channel below the vehicle. The upper rocker height is measured from the ground to the upper most portion of the rocker panel. The measurement of the upper rocker panel only measures the metal portion of the rocker panel. Vinyl and plastic coatings are not included. During side impact events, a critical factor determining crash severity is the degree of structural interaction of the vehicle rocker and pillars with the impact partner. If the center of force generated by the impacted device is above or below the rocker panel, poor engagement and high levels of compartment intrusion are likely. Trends in new vehicle design indicate increased overall height of rocker panels in order to maximize potential interacting space. The Volvo Side Impact Protection System (SIPS) is an example of this design enhancement without compromising the ease of vehicle entry and exit. 11. Striker Height- The distance from the ground to the lowest portion of the striker perpendicular to the doorframe (i.e. from the ground to the lowest portion of the striker that engages the door). The striker or latch point is a structurally rigid point where a positive connection is made between the door structure and the B-Pillar. Often manufacturers will attach side impact door beams at this rigid point and the hinge attachment points at the vehicle A-Pillar. Knowledge of the striker height, provides an indication of the potential for interaction between the door's side impact beam and the impacted structure. 12. Static Stability Factor- The Rollover Resistance Ratings assigned by NHTSA are based on the Static Stability Factor (SSF). The SSF is essentially a measure of how top heavy a vehicle is. This factor is the ratio of one half the track width to the center of gravity (c.g.) height. The Rollover Resistance Ratings of vehicles were compared to 220,000 actual single vehicle crashes, and the ratings were found to relate very closely to the real-world rollover experience of vehicles. Based on these studies, NHTSA found that taller, narrower vehicles, such as sport utility vehicles (SUVs), are more likely than lower, wider vehicles, such as passenger cars, to trip and roll over once they leave the roadway. Accordingly, NHTSA awards more stars to wider and/or lower
62 vehicles. The Rollover Resistance Rating, however, does not address the causes of the driver losing control and the vehicle leaving the roadway in the first place. One criticism for the static stability factor is the fact that it is an oversimplification of the true structure of the vehicle. It does not include the effects of suspension deflections, tire traction and electronic stability control (ESC). The above vehicle characteristics are shown graphically in Figure 3.1. Figure 3.1: Vehicle Characteristics As Measured Tables 3.1, 3.2, and 3.3 below contain average vehicle specifications for each class reviewed. All available resources were used to obtain this data. It is believed that if a vehicle with attributes closest to the class average is chosen for future crash testing, the entire class should be well represented. However,
63 current practices utilize the "worst case vehicle" approach where the attributes of the test vehicle lie at the boundary of the population. To aid the selection of an average vehicle, Appendix B lists over 342 vehicle makes and models and their corresponding design attributes. Average of Moments of Inertia Vehicle Type Class Pitch Roll Yaw Avg. SSF Car Compact 1584 374 1685 1.342 Midsize 2438 495 2544 1.354 Large 2946 560 3081 1.346 Car Total 2208 460 2320 1.347 SUV Compact 2059 515 2143 1.064 Midsize 3353 692 3399 1.083 Large 5165 1019 5206 1.076 SUV Total 3172 674 3233 1.074 Truck Compact 2627 474 2669 1.205 Large 4644 846 4693 1.172 Truck Total 3782 676 3824 1.171 Van Large Van 5953 1198 5912 1.110 Minivan 3481 822 3536 1.154 Van Total 3991 884 3996 1.145 Grand Total 3152 640 3212 1.187 Table 3.1: Average Inertial Properties per Vehicle Type and Class
64 Length Width Ht Whlbase Curb Wgt. Front Ovrhng Rear Ovrhng Ft. Rock Height CAR compact 168.19 65.21 52.88 96.42 2380.01 34.75 36.93 7.56 mid 186.68 70.11 53.43 104.41 3159.74 38.86 43.44 7.87 large 206.27 74.46 55.40 114.21 3831.85 41.43 50.56 8.45 CAR Tot 184.19 69.23 53.72 103.68 3012.77 37.91 42.75 7.88 SUV compact 157.92 66.33 66.61 94.89 2849.49 28.17 34.56 10.99 mid 177.68 69.59 68.83 104.54 4022.32 31.12 41.67 15.07 large 195.89 78.19 72.56 116.08 4907.71 33.62 46.02 15.59 SUV Tot 178.06 71.56 69.48 105.63 3977.77 31.08 41.00 13.44 TRU compact 186.55 66.94 63.58 112.79 3038.79 30.97 43.09 11.89 large 212.66 77.32 71.36 132.18 4269.49 34.47 46.03 13.18 TRU Tot 196.46 70.88 66.49 120.15 3505.77 32.33 44.24 12.15 VAN mid 186.51 72.34 66.92 112.25 3547.82 35.77 38.63 9.89 large 200.33 77.56 77.75 121.18 4426.65 33.35 45.53 VAN Tot 191.71 74.30 70.91 115.61 3878.47 34.90 41.11 9.89 Grand Tot 184.78 69.84 56.81 105.46 3183.29 36.78 42.56 8.35 Table 3.2: Structural Properties per Vehicle Type and Class (Averages) Rr. Rocker Height Ft. Bumper Height Rr. Bumper Height Door to Ground Front Track Ft. Wght Percent Rr. Wght Percent CAR compact 7.35 11.23 11.68 10.95 56.98 60.7% 39.3% mid 7.62 11.18 11.83 11.30 59.17 59.8% 40.3% large 8.37 11.55 12.51 11.19 61.46 59.2% 40.8% CAR Tot 7.69 11.29 11.94 11.10 59.12 60.0% 40.0% SUV compact 11.21 12.83 13.42 15.75 57.18 54.7% 45.3% mid 15.23 16.64 17.04 18.41 58.45 53.1% 46.9% large 16.57 15.89 18.50 19.49 64.73 52.9% 47.1% SUV Tot 13.72 15.15 15.85 17.94 59.92 53.6% 46.4% TRU compact 13.34 15.03 13.92 14.70 57.21 61.0% 39.0% large 14.74 18.08 16.91 64.50 0.0% 0.0% TRU Tot 13.65 15.59 14.95 14.70 60.50 61.0% 39.0% VAN mid 10.41 10.10 12.13 12.80 61.61 57.7% 42.3% large 65.55 55.8% 44.2% VAN Tot 10.41 10.10 12.13 12.80 62.80 57.4% 42.6% Grand Tot 8.26 11.55 12.19 11.44 59.67 59.5% 40.5% Table 3.3: Average Structural Properties per Vehicle Type and Class- (Population Weighted Averages)
65 3.2 Barrier Force Data Vehicle to vehicle crash incompatibility has been attributed to three factors: (1) mass incompatibility, (2) stiffness incompatibility, and (3) geometric incompatibility [14]. These factors may be effectively applied when considering compatibility between vehicles and roadside hardware objects as well. The measurement of vehicle mass is relatively straightforward. However, measurement of stiffness and geometric compatibility needs further definition. Without exhaustive investigation of individual vehicle attributes as shown in the following section, a method has been developed to understand vehicle metrics critical to the interface between striking vehicles and objects struck. This method is repeatable and objective making it ideal for side by side comparison of a variety of structures. It has been suggested that the height of the forward-most load-bearing member of the vehicle structure as a metric for geometric incompatibility. Since this element has no precise definition, the rocker panel height was used as the geometric metric. For the stiffness metric, the vehicle crush at the maximum barrier force during a 35-mph rigid barrier crash was utilized. [14] NHTSA's crash test program produces additional measurements, which can contribute to assessing stiffness and geometric characteristics of vehicle frontal structures. For most of the 35-mph crash tests conducted under the NCAP program, the time history of the distribution of force applied by the vehicle to the barrier was measured. These measurements indicate the geometric location of "hard spots" and the amount of force the vehicle imparts to a rigid barrier. This data permits the calculation of local stiffness and of load paths at various heights. Different aggressiveness metrics may be applicable to different crash modes. The efficacy of any proposed metric would need to be verified using on-the-road crash and injury data. However, a number of metrics can be proposed and developed from the available NCAP test data. For a front to side impact, the front of the striking vehicle may crush less than 125 millimeters. The force developed in this intermediate crush range and the height of the force measured on the barrier face may be the critical parameters. For a frontal-offset crash, the force and geometry of only the left or right portion of the vehicle front may be applicable. For interaction with reasonably compliant roadside devices such as roadside hardware crush levels rarely exceed 125 millimeters unless localized intrusion by barrier sections occurs. The use of barrier force data permits a finer discrimination of vehicle stiffness and geometry that can be further investigated as appropriate aggressivity metrics. From this approach, metrics may be derived from barrier test data that may be used to assess vehicle geometric and stiffness aggressiveness in frontal type crashes. Barrier Information
66 The barrier used in the New Car Assessment Program (NCAP) is a rigid, fixed barrier with 36 force measuring load cells on its surface. The load cells array consists of 4 rows of 9 cells, as shown in Figure 3.2. The rows are designated by letters A through D, with A at the bottom. The columns are numbered 1 through 9, starting at the left, facing the barrier. The array is subdivided in 6 groupings, 1 through 6, numbered left to right, and beginning with lower left grouping (see Figure ). Figure 3.2: Configuration of Load Cells on Barrier The array of load cells provides the opportunity to assess the distribution of forces that the vehicle imposes on the barrier during the crash. In this study, the relationship between barrier forces and their geometric location are of particular interest. In offset crashes, the left or right side of the structure principally deforms and absorbs energy. In centerline impacts with narrow objects, the center response is primary. In head-on crashes with large overlap, the entire width of the force array may be required. The vertical force distribution of the vehicle structures in contact during the crash is important in assessing the geometric compatibility. To address these various requirements, the barrier measurements have been used to graphically present the forces measured by all 36-load cells. The force distributions are examined at three points during the crash. The stiffness is calculated by dividing the force measured by the load cells at a particular time by the calculated vehicle crush at that time. The vehicle crush is determined by double integration of the longitudinal acceleration measured on a structural member close to the vehicle's center of gravity. To quantify the height of the structural loading, a center of impact force was calculated for three columns of cells. The left column contained the 1 and 4 groupings, the center column the 2 and 5 groupings, and the right the 3 and 6 groupings. In addition, the height of the center of force for the total loading was calculated. For each grouping, the force on each row of cells was assumed to be uniformly
67 distributed. The height of the center of the force was calculated, applying static equilibrium relationships as shown in Figure 3.3. The center of force was calculated for vehicle crush of five inches, 10 inches and 15 inches. In the tables and figures given here all data are reported in metric units. The three crush levels are reported as the approximate metric equivalent - 125mm, 250 mm and 375 mm. In Figure 3.3, static equilibrium is first applied. The force (F) that is required to resist the sum of the load cell forces from rows A, B, C, and D is determined. The height of force F is then found by applying moment equilibrium to the barrier forces and moment arms. The height H is defined as the Center of Force. The center of force calculation is made for the entire rows of load cells as well as for the left third, the center third, and the right third of the rows. Figure 3.3: Definition of Center of Force, H The linear stiffness is sensitive to the accuracy of the zero time step selected for the barrier force data. The force level is less sensitive than the stiffness to the zero time step selection. Consequently, force rather than stiffness is a preferred metric at the selected crush values.
68 Figure 3.4: Total Barrier Force vs. Vehicle Crush At a crush of 200 mm, the Jeep Grand Cherokee exerts almost twice as much force as the Dodge Neon. This difference in stiffness will result in a higher extent of crush for the Dodge Neon in a frontal crash involving the two vehicles. This difference illustrates the stiffness differences between the two vehicles. These differences are shown in Figure 3.4 above. Figure 3.5: Force Deformation Relationships in Vehicle to Vehicle Frontal / Side Crashes
69 An idealized relationship between the crash forces of cars with different frontal stiffnesses is shown in Figure 3.4. In a frontal-to-frontal collision, the soft car crushes more than the stiff car at the same interface force. In the example, the interface force level is 400 kN. The crush of the soft car is 500mm and the crush of the stiff car is 250 mm. The area under the force-deformation curve is proportional to the energy absorbed. Consequently, the soft car has absorbed about twice as much crash energy as the stiff car. This difference illustrates the stiffness incompatibility of the two vehicles. As shown in Figure 3.5, the force vs. crush relationship may not be linear, as assumed in the figure. It should be noted that the difference in the geometric location of the forces generated by the vehicle structures could influence the idealized interaction presented in Figure 3.5. This difference will be addressed under the discussion of geometric compatibility. The maximum force produced during the crash and the linear stiffness based on the crush at maximum force have been suggested as metrics for stiffness incompatibility. In view of the force vs. crush non-linearities, and geometric influences during the crash, some more robust metrics may be needed. In this study, we propose to investigate the force levels at 125, 250, and 375 mm. The forces developed by the vehicle left, center, or right segments of the vehicle front may be applicable in offset collisions. Tabular Summaries of Load Cell Barrier Data This report presents summary data from 50 vehicles. The 50 vehicles are listed in Appendix B of this report. Another 14 vehicles have been analyzed, but the data was found to be of unsuitable quality. In 17 of the cases, data was not reported for three of the four rows of load cells. The data on the 50 vehicles included in this report should be considered preliminary. Several adjustments in the data will be necessary. For example, some vehicles may not have impacted the center of the barrier. Shifting of the load cell columns to the right or left will be needed in these cases. In other cases, a single load cell in the array may produce unrealistically high readings. Finally, adjustments to gain a precise zero time step may be necessary in a few cases. The vehicle characteristic table shown in Appendix B provides selected results of the barrier data analysis. The nine columns of load cells are divided into three groups as described earlier. The groups are: left, center and right. The sums of the forces left, center, right, and total are designated by FCRT, FCCT, FCLT, and FCT, respectively. The percent of the barrier force on the A, B, C, and D rows are designated in the last four columns of the tables. The values listed in the table are for a vehicle crush of 375 mm. Data Processing Procedures The acceleration data points were the average of two accelerometer readings. The two accelerometers selected were the left and right rear floor pan or the left and right rear seat accelerometers. In the event inaccurate velocity changes of the vehicle were predicted, the best available accelerometers were selected.
70 The raw data from all 36 load cells was processed. The raw acceleration and barrier load cell data points were filtered according to SAE J211 Standard, with a corner frequency of 18, using a filter supplied by NHTSA. It was assumed that the zero time steps provided in the data were accurate, and were identical for the force and acceleration data. Beginning with the zero time step, acceleration data and barrier force data were sampled every 2 ms for 120 ms. The resulting acceleration data and load cell data were the input for subsequent analysis. In examining the resulting data, several inconsistencies were observed. The most frequent was an initial force on load cells at time zero. In the event the total force at time zero was greater than 10% of the maximum barrier force, the data was rejected. A second problem was the presence load on cells outside the contact region, or unrealistically high loads on cells inside the contact region. These cases were not rejected in the event the consequence was negligible. Finally, in some cases, the acceleration readings produced a higher or lower delta-V than expected. In the event that the delta-V prediction from the accelerometers up to the time of maximum crush was reasonable, the data was not rejected. Discussion The results of the barrier data provide useful insights into the geometry and height of the stiffest portions of the vehicle structure in a barrier crash. By developing metrics for these properties, it may be possible to quantify more precisely vehicle compatibility with a variety of impacted structures. Other structures may include any aspects of opposing vehicles or roadside safety systems. The proposed metrics need to be further evaluated. The evaluation should include the assessment of a large number of vehicles and an assignment of proposed compatibility metrics based on barrier crash test data and physical measurements. The resulting metrics should be evaluated by determining the extent to which they explain the aggressiveness characteristics observed in the on-the-road crash data. The application of load cell barrier data provides valuable measurements for assessing the loading of vehicles in a crash. The metrics developed from barrier data needs to be evaluated against NASS/CDS and FARS data to assess the viability of the metrics, and their applicability to understand compatibility issues between the current vehicle fleet and existing roadside safety structures. 3.3 Application of Vehicle Characteristics For this task, the relationship between vehicle characteristics, roadside hardware design characteristics and impact scenario are studied. Metrics such as vehicle mass, geometry (bumper height, sill height, and hood profile) and structural factors such as body type and stiffness can be used in combination to assess effectiveness of roadside hardware devices during impact. Ideally, design and performance corridors for
71 vehicles and roadside hardware devices should be aligned to ensure optimal performance of highway systems during crashes. The following full-scale crash tests (#472580-1 and #472580-2) were performed at the Texas Transportation Institute (TTI). During this testing, two different vehicles of similar size, class and mass impacted a W-beam guardrail under the same conditions yet resulted in drastically different post impact vehicle behavior. Tables 3.4 and 3.5 present general information regarding test vehicles and test configuration. Vehicle 1: 1996 Ford Taurus: Vehicle 2: 1995 Chevrolet Lumina Mass: 1449 kg Mass: 1505 kg Speed: 99.5 km/h Speed: 98.4 km/h Impact Angle: 26.4° Impact Angle: 25° Test #: 472580-1 Test #: 472580-2 Length (m): 5.04 Length (m): 5.1 Width (m): 1.85 Width (m): 1.84 Height (m): 1.42 Height (m): 1.4 Wheelbase (m): 2.76 Wheelbase (m): 2.73 Table 3.4: Vehicle Specifications for TTI Test #472580-1 & 2 Barrier Specifications: Type: Modified G4(1S) Strong Post Installation Length: 53.4 m Barrier: W-beam (12 gauge) Rail Length: 3.82 m Post Spacing: 1.905 m (29 posts) Post Length: 1.83 m Blockouts: 140mm x 195 mm x 360 mm routed timber Rail Mount Height: 550 mm Anchorage: BCT SKT-350 Table 3.5: Barrier Specifications for TTI Test #472580-1 & 2 The guardrail system used consists of a series of 2-Space W-Beam Guardrail sections each 4130 mm long. Steel wide-flange posts are placed 1905 mm apart (2 per section) and embedded in packed soil. Timber block-outs separate the post and the guardrail by 150 mm and are mounted using a single steel bolt through the block center. The guardrail system in pre-tensioned using a BCT Cable anchor assembly in conjunction with a strut and yolk assembly.
72 During the first test (#472580-1) where the impacting vehicle was a 1996 Ford Taurus, the guardrail provided adequate protection during the 25 degree impact. The vehicle was redirected without serious deformation to large parts of the vehicle structure or excessive deceleration of the vehicle in the longitudinal or lateral direction. Conversely, the interaction of the Chevrolet Lumina and the W-beam system during test #472580-2 raises several questions regarding performance of this system. The Lumina impacted the barrier at approximately the same location as that described above (3 ft. before the thirteenth post of the complete barrier system). As the vehicle traveled longitudinally along the length of the W-beam, the first block-out released from the W-beam at its single attachment point similar to the Taurus test. Shortly following the release of the block-out, the front left corner of the vehicle reached the splice connection point between the thirteenth and fourteenth barrier sections (first and second contacted). At this time, an out pocketing of the steel W-beam is created and travels longitudinally along the rail until it reaches the splice section. This localized region of high deformation (and stress) is due to underlying structure that initiates a fracture that travels vertically from the bolt attachment point. With the failure of the W-beam, the vehicle intruded further behind the barrier and past the midline of the vehicle. Later, an off center frontal impact with the next post initiated rollover of the vehicle. It has been hypothesized that similar vehicle mass, CG height and outer body dimensions would yield similar results during crash testing. For these tests, great care was taken during guardrail installation to produce repeatable barrier behavior. One remaining factor not eliminated by identical test conditions is vehicle structural properties. These include varying stiffness of underlying structural members (frame rails, engine configuration, drive train geometry, suspension characteristics, etc.) Using vehicle characteristics sited in Task 3 described in this report, differences that may have led to divergent test behaviors has been discovered. Upon inspection of the underlying frame structure of both the Taurus and Lumina, it can be seen that geometric differences do exist. Figure 3.6 shows an overlay of schematics for the underbody structure of the two vehicles. Individual structural diagrams were obtained from the 2000 Mitchell Automotive Repair Database and images were subsequently overlaid. It can be seen that an upward distance of 12 cm exists between the lowest structural point of the forward frame structure (engine cradle) of the Chevrolet Lumina and the lowest structural point of the Taurus. In addition, the lateral location of the bumper mounts between the two vehicles indicates that the Lumina structure is 5 cm wider than the Taurus (i.e. mount points of the Lumina lie slightly outboard of the Taurus). Geometric characteristics of the Lumina show a reduced distance between the vehicle outer body and the hard point at the engine cradle mount point on the vehicle frame in the lateral direction. In other words, crush distance has been reduced in the lateral direction before direct interaction between structural members and adjacent hardware. In the
73 vertical direction, the lowest structural point of the Lumina falls at nearly the same height as the bottom edge of the W-beam section as installed. This vertical and lateral location of this hard point creates a more favorable condition for loading at splice of the W-beam section. Upon examination of crash test footage, the tear in the W-beam appears to initiate along the lower portion of the rail at the first upstream bottom bolt of the splice and subsequently travels upwards. A larger area of vehicle/beam interaction may prevent this localized rail deformation. Also, reduced levels of outer body deformation of the vehicle may have a similar positive effect. This design for the front portions of rail structures is observed in other vehicle platforms; however, it is certainly not a common feature across all passenger vehicle structures. Figure 3.6: Overlay of Chevrolet Lumina (light) and Ford Taurus (dark) lower frame structures (Permission for Reprint Given by Mitchell Automotive Repair, 2002) Geometric factors are hypothesized to have an effect on the potential for W-beam failure during these impact conditions; however other influential structural differences exist between the two vehicles as well. Upon comparison of the frontal stiffness profiles outlined earlier in this report, considerable differences may be observed. Figures 3.7 and 3.9 below show stiffness levels across the frontal structure of each vehicle at increasing levels of vehicle crush. During interaction with guardrail systems or other similar longitudinal barrier devices, crush levels rarely exceed 10 inches. Accordingly, only stiffness profiles at 2 inches, 5 inches and 10 inches will be discussed.
74 Figure 3.7: Ford Taurus Stiffness Profile Figure 3.8: Ford Taurus Underbody- Post Crash
75 Figure 3.9: Chevrolet Lumina Stiffness Profile Figure 3.10: Chevrolet Lumina Underbody- Post Crash At 2 inches of crush, the stiffness profile of the Ford Taurus peaks at approximately 75 N/mm and the shape of the stiffness curve spans from the 3L location to the 7R column. For the Lumina, this curve peaks at 45 N/mm and spans a narrower region across the vehicle. Upon comparison, the differences in stiffness between the two vehicles indicate that the outer-body structure of the Lumina will deform more significantly than the Taurus. This difference should be more considerable at the most outboard regions of the vehicle face. At 5 inches of crush, important differences become obvious. The stiffness profile for the Taurus, which peaks at 100 N/mm, is very broad spanning from the 2R level to the 8R level. It should be noticed that this high stiffness level evenly spans the entire front face of the vehicle. In comparison, the Lumina
76 stiffness at this level of crush also peaks at nearly 100 N/mm but spans a much smaller percentage of the vehicle frontal structure, It spans from the 3L location to the 7R location. The implication of this during an oblique guardrail impact would be high levels of deformation of the outer body structure of the Lumina at the outboard regions of the vehicle. Deformation of this structure would expose a vehicle hard- point to the opposing guardrail structure. This, in turn, creates pocketing to the metal guardrail structure, a region of increased stress concentration and higher likelihood for failure of the W-beam. In order to expose the hard-point which exists beneath the outer body of the Taurus, a larger force in an oblique direction would be required. It may be observed from the post impact photos shown in Figures 3.8 and 3.10 above, that the integrity of the front driver side structure of the Taurus remains intact throughout the test while significant deformation is observed in the frontal structure of the Lumina. This deformation exposes the underlying structural hard-point discussed previously. It should be noted that severe deformations along the centerline of the Lumina shown in the photos are the result of interaction with guardrail posts during and after beam failure. This interaction does not contribute to the failure of the system; however they indicate the severity of the resulting vehicle behavior leading to rollover. In order to investigate the nature of the rail/vehicle interaction more closely, a finite element model of the Modified G4 (1S) systems was assembled. This model accurately represents all aspects of the barrier system including accurate ground properties and post interactions, accurate geometry and material properties of posts, block-outs and rails plus accurate bolts and other attachment hardware. Further, a model of the 1995 Chevrolet Lumina, created by EASi Engineering International in 1997 exists and has been combined with the Modified G4(1S) system for the simulation cases. To understand the likelihood of rail failure during impact, stresses of each element within the W- beam have been monitored. High levels of localized stresses seen in the lower half of the W-beam section confirm excessive contact forces with the underlying engine cradle/frame hard-point. A second simulation case was created where the Lumina structure impacted the guardrail section under identical impact conditions. For this case, the vehicle structure was rigidized so that the outerbody would not deform. This stiffening of the outerbody prevented the narrow underlying hardpoint from directly interacting with the W-beam structure. During this case, it was shown that the high levels of localized stress seen in the previous case were reduced to levels where material failure is unlikely. This type of analysis provides an opportunity to vary both vehicle and roadside hardware design characteristics to confirm hypothesized mechanisms and occurrences of incompatibility.
77 Figure 3.11: Interaction of Ford Taurus and Chevrolet Lumina impacting Modified G4(1S)
