Recommended Test Vehicle Update for Manual for Assessing Safety Hardware

Abstract

Passenger vehicles in Manual for Assessing Safety Hardware (MASH) crash testing are required to be representative of the modern vehicle fleet and practical worst-case impact scenarios. The objective of this paper was to identify potential updates for standard test vehicle selection criteria for MASH as well as develop a preferred strategy for performing future standard test vehicle reviews. Representative vehicles were documented using sales data, and registration and crash data were observed to validate the primary use of sales data. Curb weights were plotted against cumulative market share of new vehicle sales to identify the 5th and 95th percentile “practical worst-case” weights of 2,800 lb and 5,850 lb, respectively, consistent with MASH philosophy. Suitable test vehicle options were found at the 5th percentile weight; however, a pickup near the 92.5 percentile weight (5,400 lb) was recommended to ensure vehicles are both representative of the fleet and obtainable for crash sites. MASH test vehicle specifications were recommended based on a review of geometrical and inertial properties of candidate vehicles near these target weights. Potential mid-size test vehicles were also explored, and four vehicle classes (two mid-size car and two crossover utility vehicle [CUV] classes) were identified as test vehicle candidates. Research on vehicle impact behavior of mid-size cars and CUVs are desired to determine impact behavior of each vehicle with different roadside hardware. Future revisions to MASH test vehicle selection criteria were outlined and should use analysis and attributes of new vehicle sales.

The American Association of State Highway and Transportation Officials’ (AASHTO’s) Manual for Assessing Safety Hardware (MASH) provides standardized full-scale testing evaluation parameters and criteria for determining crashworthiness of roadside features, such as longitudinal barriers and crash cushions in the United States ( 1 ). Full-scale crash testing procedures are intended to represent real world, “practical worst-case” impact scenarios, which are separated into test matrices specific to the roadside feature being evaluated. Test matrices include vehicle selection guidelines, impact speed, angle, and location reflective of “practical worst-case” scenarios for each crash-tested feature. Criteria such as vehicle deformation and intrusions, impact accelerations and velocities, and vehicle containment and redirection must be satisfied to verify roadside structure crashworthiness.

Impact conditions and crash test vehicles should be representative of current highway conditions. Vehicle geometries and dimensions are constantly evolving and may affect vehicle stability and barrier impact loading. The composition of traffic and personal vehicle ownership has changed drastically the last 50 years. In the 1970s, cars comprised nearly 80% of the vehicle fleet, and no light truck vehicle was crash tested because of their scarcity. For example, full-scale crash testing procedures established in TRC 191 ( 2 ) and NCHRP Report No. 230 ( 3 ) used only small and large sedans to represent passenger vehicles. Pickup trucks were introduced to full-scale crash testing in 1989 under AASHTO’s Guide Specifications for Bridge Railings ( 4 ). AASHTO’s document included seven passenger vehicle sizes (four automobiles and three vans/pickups) in addition to commercial vehicles such as the single-unit truck and tractor-trailer. Three barrier performance levels (PL) were established, PL-1, PL-2, and PL-3, to evaluate different impact scenarios ranging from small car to tractor-trailer. The implementation of NCHRP Report No. 350 ( 5 ) established two small cars and a pickup as passenger test vehicles. Test conditions and evaluation criteria included a gradation of performance levels ranging from TL-1 (31 mph impact speed pickup and small car) to TL-6 (62 mph impact for passenger vehicles and 50 mph impact for a tank-trailer vehicle).

Modern full-scale crash testing guidelines described in AASHTO’s MASH strive to capture “worst practical conditions” for vehicle-to-hardware impact scenarios ( 1 ). Three passenger vehicle sizes are included in MASH crash testing matrices:

2,420-lb Small Car (1100C)

3,300-lb Mid-Size Car (1500A)

5,000-lb Pickup Truck (2270P)

MASH intent is to capture practical worst-case impact scenarios by using representative small and large test vehicles at 5th and 95th percentile weights, respectively. It is believed that if systems acceptably contain and redirect vehicles of representatively small and large weights, all vehicles in between those weights should also perform acceptably with similar impact conditions. The distributions of weights of vehicles in use and which are involved in crashes are of pivotal importance. However, given the difficulty of determining weights of vehicles in use, and accommodating changes in vehicle fleet over time, sales data are commonly used as a surrogate or estimate of the crashed vehicle weight distribution. Comprehensive U.S. vehicle fleet analysis was last completed in 2009 when MASH used 2002 weight distribution data to select test vehicles. Test vehicles were selected near the 2nd and 90th percentile weights as vehicle weights had increased dramatically in the 10 years before MASH implementation, and to be consistent with common, inexpensive vehicle models. MASH suggests test vehicles minimally sell 50,000 units per year, with 100,000 unit sales being desired ( 1 ). In addition, a review of vehicle properties should occur approximately every 5 years to ensure that vehicle evaluation criteria continue to be representative of the vehicle fleet, but revisions to vehicle selection criteria are recommended only when significant changes occur.

Since NCHRP Report No. 350, lightweight cars such as the 820C test vehicle have been essentially phased out of U.S. automobile production. Compact cars have highest fatality rates of all vehicle classes when considering all crashes (including rollover); however, mid-size and large sport utility vehicles (SUVs) have the highest fatality rates during guardrail and cable barrier impacts ( 6 ). This finding suggests “practical worst-case scenario” impacts for each system tested are dependent on vehicle type. In addition, instability of mid-size and large SUVs during longitudinal barrier impacts indicated their crash testing use may be more suitable than pickups to capture instability concerns. Although these findings support SUV use in crash testing, car performance during guardrail impacts is also critical, as vehicle impact location influences wheel snag probability which may lead to undesirable impact performance ( 7 ). Cars also represent the smallest vehicle profile and may be associated with increased risk of guardrail underride and penetration.

Currently, the MASH mid-size test vehicle is a mid-size sedan, and its weight percentile was not a selection criterion. Different vehicle types may be considered for mid-size test vehicle selection to encompass practical worst-case impact scenarios and be representative of the modern U.S. fleet. For example, vehicle stability is an evaluation factor for vehicle crash performance. It was found that in collisions with concrete and metal barriers, the most important predictors of rollover are vehicle type and whether the vehicle was tracking before barrier impact. Cars had the lowest chance of rollover, and (SUVs) and crossover utility vehicles (CUVs; equivalently “compact utility vehicles”) were observed nearly twice as likely as pickups to rollover ( 8 ). CUVs are the newest class of passenger vehicle with geometries similar to SUVs, but are typically produced using the same unibody or “body-in-white” techniques as modern as passenger cars.

U.S. crash testing laboratories have demonstrated concerns about insufficient availability of modern mass-production vehicles satisfying MASH small car criteria. Furthermore, little to no research has evaluated changes in modern light trucks and whether current MASH mid-size test vehicle replicates suitable “worst practical conditions” as described in MASH ( 1 ). Passenger car sales decline, vehicle dimension changes, and emergence of CUVs have prompted need to determine whether modern crash test vehicles remain reflective of the attributes of real-world vehicles during crashes with roadside safety hardware. Thus, evaluation of current passenger vehicle trends such as sales volumes, inertial properties, and physical dimensions is desired to recommend selection criteria of MASH passenger test vehicles.

The research objective of NCHRP Project No. 20-07 Task 372, Evaluation and Update of MASH Test Vehicles, was to investigate the composition of the current passenger vehicle fleet and determine what revisions, if any, are needed for MASH test vehicle selection criteria to be representative of the contemporary vehicle fleet. In addition, a method for rapidly conducting a similar study in the future was documented. This research paper summarizes the analysis of sales data and vehicle selection recommendations of that study.

Methodology

MASH previously combined vehicle model sales data and curb weights to approximate distribution of passenger vehicle weights for a given sales year. Vehicle weight distribution was used to identify the 5th and 95th percentile weights, which are guidelines for small and large test vehicle recommendations. The MASH mid-size vehicle was selected to evaluate staging of energy-absorbing terminals, crash cushions, and truck-mounted attenuators ( 1 ), and little supplemental information aids in mid-size vehicle selection.

Analysis for this study included vehicle sales data from 1980 to 2018. Sales from 2018 were not available until late February 2019; therefore, this study primarily focuses on 2017 sales. Validation of sales data use as surrogate for determining vehicle fleet composition was done through comparison of sales data to vehicle registration and crash data.

Unfortunately, available U.S. sales data are rarely differentiated by trim level, and sometimes not differentiated by pickup payload capacity (e.g., Ford F-150, F-250, and F-350). Methods for allocating vehicle model sales among trim levels were explored to obtain discrete weight distribution approximations. Boundary curves of the lightest- and heaviest-weight distributions were created to observe the total possible variation in vehicle weights based on sales, and a median-weight sales distribution model was used to create a discrete weight distribution to obtain 5th and 95th percentile weights. Vehicle classes near the 5th and 95th percentile weights were documented to identify whether adequate vehicle availability existed for crash testing in accordance with MASH philosophy. Dimensional properties of vehicle classes near 5th and 95th percentile weights were tabulated using a high and low method so updated vehicle properties could be recommended.

Few mid-size vehicle selection criteria exist, so multiple vehicle classes were identified as potential mid-size test vehicles, including two mid-size sedan and two CUV classes. Potential mid-size vehicle classes were determined based on previous mid-size vehicle selection and the emergence of CUVs as common mid-size vehicles. Further research is needed to ensure the MASH mid-size test vehicle is representative and encompassing of practical worst-case impact scenarios. As well, careful consideration should be provided to the evolution of new vehicle safety features that could affect vehicle interaction with roadside barrier systems.

Differentiating between classes of vehicles by body style is challenging because of varying attributes and the independence of vehicles within those attributes. For consistency, researchers utilized the Wards’ Vehicle Segmentation Criteria for all vehicle makes and models ( 9 ). A description of additional vehicle classification systems and the justification for selecting the Wards system is provided in detail in the NCHRP Project 20-07 Task 372 final report. Eight passenger vehicle types were specified. Car sizes (small, mid-size, and large) were segmented by overall length, and luxury cars were defined with a base price greater than $34,000. Vans and pickups are generally classified by their body style, whereas CUVs and SUVs were differentiated by their frames. Wards identifies CUVs as wagon body style with unibody construction, and SUVs typically use body-on-frame construction with minimum 7.5-in. ground clearance. For purposes of this study, “cars” were generally used to refer to sedans, coupés, convertibles, luxury cars, and hatchbacks, and “light trucks” were used to refer to CUVs, SUVs, pickups, and vans; this was consistent with Wards segmentation criteria.

Results

Sales Analysis

Wards Intelligence sales data were observed, and nearly 80% of passenger vehicles sold in 1980 were cars and the remaining 20% were light trucks. Since 1980, a steady shift from passenger cars to light trucks occurred, although a leveling of passenger car and light truck demand occurred in the early 2000s because of fuel efficiency concerns, the 2001 domestic terror attack on the World Trade Center and federal properties, and two economic recessions. Light truck sales began to increase again in 2012, and by 2018, nearly 69% of passenger vehicles purchased were light trucks, and the remaining 31% were cars. Moreover, in 2018, pickup trucks accounted for more sales than small and mid-size cars combined. The sales trend from 1980 to 2018 is shown in Figure 1.

Figure 1.

U.S. passenger vehicle sales.

The four highest sales volume vehicles types since 2005 in descending order are: CUVs, small cars, mid-size cars, and pickups. Since 2009, these four vehicle types have comprised at least 75% of all new passenger vehicle sales. Specifically, CUV shares of total new vehicle sales rose from 12.0% in 2005 to 38.7% in 2018, whereas pickups and small cars experienced approximately 2.0% reduction in market share and mid-size cars had a 6.1% decrease in market share. In 2018, large cars, luxury cars, SUVs, and vans combined for approximately 21.1% share of new vehicle sales, a cumulative sales share decrease of 16.5% since 2005. Passenger vehicle sales shares by vehicle type are shown in Figure 2.

Figure 2.

Passenger vehicle sales shares by type.

Vehicle manufacturers such as Ford, Chevy, and Buick have been removing car models from production and increasing utility vehicle production. In 2005, there were about 41 CUV models and 164 car models available for purchase in the United States, compared with 2018 when there were about 97 CUV models and 160 car models available. Sales trends and vehicle availability suggest light truck sales will continue to outpace cars sales in the near future unless other factors infringe on light truck popularity and availability. The increase in light truck sales has been catalyzed by CUVs, which comprise the largest vehicle sales share since 2009. Though not previously used in crash testing, CUVs warrant consideration for inclusion as a crash test vehicle because of their popularity.

Vehicle Registration and Crash Data Analysis

Registration and crash data analyses were conducted to validate sales data use for crash test passenger vehicle selection. Bulk registration data from the Bureau of Transportation Statistics ( 10 ) and Federal Highway Administration (FHWA) ( 11 ) provided registered shares of passenger vehicles and approximate ages of vehicles on roadways. Vehicle registration data were available from 1994 to 2016. Proportional registrations and passenger vehicle sales are shown in Figure 3. In addition, year-by-year (1995–2014) analysis showed average light trucks and car ages were similar ( 10 ). Using the sales data as a predictive model of the registration data by applying x-axis time shifts to the new sales data, researchers calculated differences between the predicted registrations and actual registrations. The highest correlation coefficient was associated with a 12-year time offset between new sales and vehicle registrations (0.97), and is shown in Figure 3. Findings suggest likelihood that registration of U.S. light trucks and cars in 2029 may be proportionally similar to 2017 light truck and car sales.

Figure 3.

Registered vehicles relationship to vehicle sales: (a) Relationship between registration and new vehicle sales data, and (b) modeling registration data with sales data by implementing a 12-year delay.

Nationally from 2014 to 2015, 48.3% of all registered vehicles were light trucks and about 43.3% were cars ( 11 ). Although national registration numbers are useful, vehicle fleet composition differs between states. For example, Wyoming’s registered vehicle shares significantly deviate from the national average. From 2014 to 2015, about 66.7% of all registered vehicles in Wyoming were light trucks and about 27.6% were cars. Some states were more consistent with national registration numbers. In Ohio, nearly 47.5% of all registered vehicles were light trucks and about 47.1% were cars, and in Utah, 51.4% of registered vehicles were light trucks and 43.5% were cars.

Crash records from individual states indicated that in Ohio, Utah, and Wyoming, cars were 2%–9% more likely to be involved in accidents than light trucks. Conclusions from the crash data analysis indicated that sales data predicted future crash trends by approximately 8–15 years after the purchase, such that the median age of the crashed vehicle was between 9 and 12 years.

Acquisition of uniform, national crash records was not possible, and limited crash data from state Departments of Transportation (DOTs) could be analyzed. Several difficulties arose during analysis: each state documents crashes using their own criteria; datasets were large and time-consuming to process; and crash data from individual states may not be representative of aggregate national statistics. Additional information on crashed vehicles and their relationship to registered vehicles can be found in the NCHRP final summary report for project 20-07. Although crash data are an excellent evaluation of hardware in-service and an accurate reflection of the distribution of vehicles which engaged roadside features and their impact conditions, the difficulties encountered using crash data analysis, and the limitations of the conclusions, indicated that sales data may be an economical, simpler, and more efficient technique for identifying attributes of MASH test vehicle selection criteria.

Vehicle Model Sales Allocation

MASH was last updated in 2009, and sales data from 2002 were analyzed to create a weight distribution and select passenger vehicles for crash testing. A consistent method for vehicle model sales allocation was needed to obtain a discrete mass distribution. Sales data were available by vehicle model, but typically, model sales were not annotated by trim level. For example, 117,596 new Kia Forte cars were sold in 2017 and distributed among six trim level weights (two coupés, two hatchbacks, and two sedans), but sales were not distributed among model trims. Pickups experienced similar lack of differentiation among trim levels and payload capacities. The Ford F-series pickup sold 834,445 units in 2017; however, the number of sales among half-ton (F-150), three-quarter-ton (F-250), and one-ton (F-350) pickups was not specified.

A dataset obtained from Dominion Cross-Sell in Lexington, Kentucky was used to determine the approximate share of pickups sold by payload capacity. The dataset included the number of new pickups sold by car dealerships in 23 states and included both private and business sales. Distributions of payload capacity by pickup model were identified. For example, Ford F-series pickups sold 483,605 units in the Dominion Cross-Sell market area. Of these, 332,165 (68.7%) were Ford F-150s, 102,828 (21.3%) were F-250s, and 48,612 (10.1%) were F-350s. Percent shares of model/payload sold were multiplied by national model units sold to obtain estimated national sales by payload capacity. Additional pickup sales estimates are shown in Table 1.

Table 1.

National Sales Estimates of Pickups by Payload Capacity

Make	Model/payload	Market area units sold	% share of models sold	Total national units sold	Estimated national units sold
Chevrolet	Silverado 1500	364,430	75.8	585,864	444,085
	Silverado 2500	84,740	17.6		103,112
	Silverado 3500	31,796	6.6		38,667
Ford	F-150	332,165	68.7	834,445	573,264
	F-250	102,828	21.3		177,737
	F-350	48,612	10.0		83,445
GMC	Sierra 1500	94,402	73.7	217,943	160,624
	Sierra 2500	25,788	20.1		43,807
	Sierra 3500	7,867	6.1		13,295
Ram	1500	204,677	72.0	483,520	348,134
	2500	52,845	18.6		89,935
	3500	26,864	9.4		45,451

Weight Distributions

Weight distribution creation requires fair allotment of vehicle model sales among model trim levels. Vehicle curb weights and other dimensions were obtained from the Canadian Vehicle Specifications database (also used by NHTSA to obtain vehicle specifications) ( 12 ). Weight distributions were produced by arranging passenger vehicles from lightest to heaviest weights. Individual vehicle model trim sales were divided by total number of passenger vehicles sold to acquire proportional sales shares for each vehicle model trim. The cumulative proportional share was plotted against curb weight to obtain mass distribution curves for different sales estimate techniques.

Passenger vehicle model sales from 2017 were allotted to trim levels within each model series using several weight-based sales distribution methods, to approximate passenger vehicle weights at the 5th and 95th percentiles. Researchers proposed multiple methods of identifying the weight distributions:

1. Bracketing

1.1. All sales allocated to the lowest-weight trim level (“lightest” distribution)

1.2. All sales allocated to the highest-weight trim level (“heaviest” distribution)

2. Distributed Sales

2.1. Quasi-normal distribution around median trim weights

2.2. Equal trim representation (mean trim weight)

2.3. Inventory counting (e.g., compiled vehicle seller data, online sales clearinghouse)

3. Simplifications

3.1. Allocate all sales to median-weight trim level

3.2. Reference to “default” values in online resources or repositories

3.3. Mean, median, or representative curb weight value from New Car Assessment Program (NCAP) testing for all trim levels of a model

Several of these techniques were explored, but results did not vary significantly. Methods and results are discussed in the NCHRP 20-07 final report. The median curb weight sales estimate was believed to be the most effective in capturing vehicle model weight ranges, was simple to review and plot, and was recommended when performing subsequent MASH vehicle reviews. Thus, only the median weight distribution results are included in this paper.

The 2017 passenger vehicle weight distribution and the 2002 weight distribution are shown in comparison with current MASH vehicle weights in Figure 4. It should be noted that 2002 sales data corresponding to vehicle weights between 2,750 and 4,000 lb were not available, but vehicle sales data were known and plotted for vehicles less than 2,750 lb and greater than 4,000 lb. The dashed line was a linear interpolation estimate in the unknown sales and weight region.

Figure 4.

Weight distributions of all passenger vehicles.

The intent of MASH is to select small and large passenger test vehicles at the 5th and 95th percentile weights, and a median-weight sales distribution technique identified discrete weights at each percentile. The 5th percentile mass (2,789 lb) suggests about 5% of all passenger vehicles sold in 2017 were less than 2,789 lb. This represents a 208 lb weight increase since 2002 when the 5th percentile weight was 2,581 lb. The 95th percentile mass (5,847 lb) suggests nearly 5% of all passenger vehicles sold in 2017 were greater than 5,847 lb. This represents a 283 lb weight increase since 2002 when the 95th percentile weight was 5,564 lb. For simplicity, the 5th and 95th percentile masses will be rounded to 2,800 lb and 5,850 lb, respectively, for the remainder of this paper.

Small and Large Test Vehicle Candidates

Traditionally, small cars have been used as small passenger test vehicles to bracket occupant risk and vehicle capture and snag for small, lightweight vehicles with narrow, sharp profiles, low bumper heights, and small occupant compartments. During barrier impacts, small car test failure typically results from excessive occupant accelerations and velocities, inadequate vehicle capture, or snag. The current MASH small car has an approximate 2.2% weight tolerance from 2,365 lb to 2,475 lb. Application of 2.2% weight tolerance to the 2017 5th percentile weight (2,800 lb) yields a range from 2,735 lb to 2,865 lb. Vehicles in this range were tabulated, and criteria were developed to isolate test vehicle candidates. Four-door sedans are the most common small car body style ( 13 ), thus hatchbacks, coupés, and convertibles were eliminated as candidates. Hybrid, electric, and luxury cars were also removed because of lack of availability and high associated costs. Note that the growth of the alternative power source (APS) class of vehicles, including hybrid, fuel cell, and electric vehicles, warrants careful consideration in future revisions for MASH standardized test vehicle criteria. The remaining small car test vehicle candidates included the following base trim level, automatic transmission, front-wheel drive, compact four-door sedans which, when combined, summed 1,332,645 total model sales in 2017:

Honda Civic DX/LX/EX 4DR Sedan (2,743 lb)

Chevrolet Cruze 4DR Sedan (2,756 lb) (discontinued)

Toyota Corolla 4DR Sedan (2,789 lb)

Kia Forte LX 4DR Sedan (2,804 lb)

Volkswagen Jetta 4DR Sedan 2.0L (2,804 lb)

Hyundai Elantra 4DR Sedan (2,811 lb)

Chevrolet Sonic 4DR Sedan (2,848 lb) (discontinued)

Dimensional properties of seven small car test vehicle candidates were tabulated, and a high- and low-value method was used to obtain proposed MASH property values. For example, the largest (106 in.) and smallest (99 in.) wheelbases of the recommended small cars were identified. The whole-number midpoint between these two values was selected as the proposed MASH wheelbase (103 in.), and the tolerance currently allowed in MASH was used (±5 in. for wheelbase). High and low values and tolerances were obtained for other dimensional properties, and other proposed MASH small car properties were obtained similarly. The most notable recommended changes to the MASH small car are an increase in weight, wheelbase, overall length, hood height, and width. It is unclear exactly how these changes will affect impact behavior, but vehicle weight increase may result in less severe occupant impact velocities and accelerations. Researchers used the Canadian Vehicle Specifications (also referenced by Expert Autostats) definitions of hood geometries, which measured the leading edge of the hood at the centerline of the vehicle. MASH has traditionally used the top of the radiator support.

Pickups have historically been the MASH large test vehicle because of availability and replicability of high mass, high center of gravity (c.g.) vehicles such as SUVs and vans. The 95th percentile weight was identified at 5,850 lb, and vehicles near this weight range included heavy-duty pickups, large SUVs, and commercial vans. SUVs and vans combined for fewer sales than pickups and were excluded as potential large test vehicle candidates ( 13 ). In addition, one-ton pickups did not yield enough sales for consideration, and specialty/luxury trims are too expensive for crash test use.

Three-quarter ton, regular cab, two-wheel drive, long-box pickups were desired as large test vehicles because their weights ranged from 5,631 lb to 5,966 lb. Analysis of available pickups from model years 2013 to 2019 using Edmunds and Cars.com showed the desired heavy-duty pickup class was not adequately available for crash test consideration. Crew and quad cab configurations and four-wheel drive trucks were found to be most common pickup styles. Although every truck manufacturer has different differentiation between body styles, generally “Crew” or “Quad” cab pickup trucks have a rear seat with either a half-door or full door. Some tabulated differences between Quad Cab and Crew Cab designations for the Dodge Ram 1500, which has been the most common pickup truck model used in MASH crash testing, are shown in Table 2. The crew cab trims had approximately 9 in. longer total length and 290 lb increase in curb weight compared with quad cab trim levels. Multiple trim levels were compared with similar results, but for consistency, only the “Big Horn” style with 4-wheel drive, standard box (6.3 ft), and 2019 model year specifications for both quad and crew cab configurations are shown.

Table 2.

Comparison of 2019 “Big Horn” Ram 1500 Quad Cab and Crew Cab Dimensions ( 14 , 15 )

	4 door 4WD Quad cab	4 door 4WD Crew cab	Difference
Length	229.0 in.	237.9 in.	+8.9 in.
Maximum towing capacity	10,150 lb	10,140 lb	−10 lb
Curb weight	5,100 lb	5,390 lb	+290 lb
Maximum payload	1,700 lb	1,510 lb	−190 lb
Height	78.8 in.	77.4 in.	−1.4 in.
Wheelbase	140.5 in.	149.5 in.	+9.0 in.
Width	79.4 in.	79.4 in.	+0.0 in.

Note: 4WD = four wheel drive. Bold values indicate potentially significant differences in vehicle properties.

Three-quarter ton, four-wheel drive, crew cab pickups typically weighed over 6,300 lb whereas Dodge, Chevrolet, GMC, and Toyota half-ton counterparts weighed 5,300 lb to 5,400 lb. Note that Ford half-ton pickup trucks were considerably lighter than Chevrolet or Dodge counterparts, and many trim levels had curb weights less than 5,000 lb. Because an eligible vehicle class could not be identified at the 95th percentile weight, the large passenger test vehicle is recommended at the 92.5 percentile weight of 5,400 lb. One of the most common pickups on roadways is the half-ton, crew cab, four-wheel drive, medium-box pickup, and Chevy, GMC, Ram, Toyota, and Nissan each produce pickups near to the 92.5 percentile weight that were eligible candidates. Pickups for sale nationally on cars.com were used to create a body style distribution for half-ton pickups, which is shown in Figure 5. Note, Ram does not produce an extended cab pickup, and Ram crew/quad cab sales were combined. Although there are differences in the total vehicle length and cab size between Crew Cab and Quad Cab options, the quad cab options were better aligned with target curb weights and preferred for consistency. It was found four-wheel drive pickups outnumber rear-wheel drive pickups by a magnitude of about 8:1. Chevy, GMC, and Ram half-ton pickups accrued about 952,843 total model unit sales in 2017, and recommended candidates for crash testing are as follows:

Chevrolet Silverado 1500 Crew Cab M/Box 4 × 4 (5,359 lb)

GMC Sierra 1500 Crew Cab M/Box 4 × 4 (5,359 lb)

Ram 1500 Crew Cab 6.4-ft Box 4 × 4 (5,386 lb)

Figure 5.

Cab and drivetrain distribution of test pickup candidates: (a) Distribution of cab type by pickup truck model, and (b) distribution of transmission type by pickup truck model.

Application of 2.2% weight tolerance to the 95th percentile weight yielded a weight range from 5,280 lb to 5,520 lb. Resulting recommended pickup properties are similar to what is currently used in MASH. The same high- and low-value method was used to identify recommended pickup truck and small car properties, and the results are shown in Table 3. It is unknown how differences in vehicle geometries will affect pickup truck interactions with roadside hardware. Estimated c.g. height is approximately 40% of overall vehicle height ( 16 ), which indicates that the current MASH requirement that vehicle c.g. height be located no less than 28.0 in. above ground ( 1 ) is satisfied. Estimated c.g. heights of recommended pickups range from 29.6 to 31.0 in. Experimental determination of their c.g. heights is desired to ensure they satisfy MASH criteria. It is also recommended that c.g. heights of high-selling, large SUVs be experimentally determined to verify that the 28.0 in. c.g. height MASH requirement is reflective of modern SUVs.

Table 3.

Current and Proposed MASH Small and Large Test Vehicle Properties

Property	Current MASH (1100C)	Proposed MASH small car	Current MASH (2270P)	Proposed MASH pickup
	Weight, lb
Test inertial	2,420 ± 55	2,800 ± 65	5,000 ± 110	5,400 ± 120
Dummy	165	165	Optional	Optional
Max. ballast	175	175	440	440
Gross static	2,585 ± 55	2,965 ± 65	5,000 ± 110	5,400 ± 120
	Dimensions, in.
Wheelbase	98 ± 5	103 ± 5	148 ± 12	151 ± 12
Front overhang	35 ± 4	36 ± 4	39 ± 3	40 ± 3
Overall length	169 ± 8	178 ± 8	237 ± 13	235 ± 13
Overall width	65 ± 3	70 ± 4	78 ± 2	79 ± 2
Hood height^*	24 ± 4	28 ± 4	43 ± 4	46 ± 4
Track width	56 ± 2	60 ± 2	67 ± 1.5	68 ± 1.5
Front bumper height	na	19 ± 3	na	26 ± 3
	Center of mass location, in.
Behind front axle	39 ± 4	43 ± 4	63 ± 4	63 ± 4
Above ground (min.)^*	na	na	28.0	28.0
Location of engine	Front	Front	Front	Front
Location of drive axle	Front	Front	Rear	Four-wheel drive
Transmission type	Manual or auto	Manual or auto	Manual or auto	Manual or auto

Note: MASH = Manual for Assessing Safety Hardware; na = not applicable; max. = maximum; min. = minimum.

Hood heights and c.g. locations may require supplemental empirical measurement or verification.

Although the recommended MASH large vehicle weight is at the 92.5 percentile (5,400 lb), the 400 lb weight increase from current MASH pickup adheres to representative vehicle use. The 400 lb addition to the large test vehicle will increase barrier loading and could potentially result in increased pocketing and snag with existing barrier designs. Limited testing has been conducted using four-wheel drive pickups under MASH test conditions, resulting in degraded hardware performance ( 17 ), and few tests before MASH exist to determine how four-wheel drive pickups affect barrier performance. However, recent truck models have also performed excellently in safety testing; in 2016, the Ford F-150 Crew Cab with automated emergency braking was awarded an Insurance Institute for Highway Safety (IIHS) “Top Safety Pick” ( 18 ), and in 2019, the Ram 1500 pickup was the first pickup to receive an IIHS Top Safety Pick+, the highest safety rating ( 19 ). Improvements in pickup truck safety come as pickups have historically lagged passenger cars, CUVs, and SUVs in implementing safety features ( 20 ). Supplementary crash testing is recommended before implementation in MASH standards to understand the consequences of adopting a larger, different vehicle.

Mid-Size Test Vehicle Classes

MASH provides minimal guidelines for mid-size test vehicle selection. The mid-size vehicle was originally included in MASH to evaluate cable barrier penetration, staged energy-absorbing terminals, crash cushions, and truck-mounted attenuators. Several mid-size passenger test vehicle options were considered: (1) continue use of the 3,300 lb mid-size sedan; (2) increase mid-size sedan weight; (3) adopt class of high-selling, compact CUVs; (4) adopt 50th percentile weight mid-size vehicle. Individual mid-size test vehicle candidates are listed in Table 4.

Table 4.

Potential Mid-Size Test Vehicles

Potential mid-size vehicle	Make/model	Trim	Weight (lb)
3,300-lb Mid-size sedan	Kia Optima	LX/LX+/EX/EX Eco Turbo 4DR Sedan	3,225
	Toyota Camry	4DR Sedan	3,234
	Hyundai Sonata	4DR Sedan 2.4L	3,252
	Honda Accord	4DR Sedan LX/Sport/EX-L/Touring	3,298
3,500-lb Mid-size sedan	Ford Fusion^*	4DR Sedan	3,435
	Nissan Altima	3.5L 4DR Sedan	3,470
	Hyundai Sonata	4DR Sedan Sport 2.0T	3,505
	Toyota Avalon	4DR Sedan	3,549
	Honda Accord	4DR Sedan EX-L V6/Touring V6	3,560
	Nissan Maxima	4DR Sedan	3,587
	Kia Optima	SX/SXL 4DR Sedan	3,594
Compact CUV	Honda CR-V	4DR SUV	3,311
	Mazda CX-5	GS 4DR SUV	3,318
	Volkswagen Tiguan	2.0L 4DR SUV	3,393
	Nissan Rogue	4DR SUV	3,417
	Toyota RAV4	4DR SUV	3,428
	Mazda CX-5	GX 4DR SUV	3,437
	Hyundai Tucson	4DR SUV	3,439
	Hyundai Santa Fe Sport	4DR SUV	3,459
50th Percentile mass CUV	Chevrolet Equinox	4DR SUV	3,761
	Dodge Journey	4DR SUV	3,825
	GMC Terrain	4DR SUV	3,854
	Kia Sorento	EX/Limited	3,878
	Ford Edge	4DR SUV	3,911
	Chevrolet Equinox	4DR SUV 3.6L LS/LT/LTZ	3,920

Note: CUV = compact utility vehicle; SUV = sport utility vehicle; DR = number of doors.

Ford Fusion model production has been discontinued.

Continued use of the current MASH mid-size sedan would indicate mid-size vehicles have minimally evolved the last 15 years. Use of mid-size sedans near 3,300 lb is possible given their availability. Non-luxury, gas-powered, four-door sedans within current MASH mid-size weight tolerance of 3,225 to 3,375 lb were observed, and the 2017 total model sales of the vehicles in this class summed to 949,032 units. Test vehicle candidates are shown in Table 4.

Another mid-size test vehicle option is to increase mid-size sedan weight, which is reflective of vehicle weight increases from 2002 to 2017. Increased mid-size sedan weight may increase potential for cable barrier penetrations. The heaviest class of non-luxury, gas-powered, four-door mid-size sedans was found near 3,500 lb, shown in Table 4. Mid-size sedan test vehicle candidates near 3,500 lb accumulated 1,129,780 total model sales in 2017.

Current MASH mid-size evaluation criteria do not consider inherent vehicle stability, and CUV use in crash testing may better encompass stability concerns than pickups. CUVs may exhibit 10%–20% greater likelihood of rollover than mid-size sedans based on Static Stability Factor (SSF), which is correlated with vehicle stability ( 21 ). CUVs tend to be lighter and narrower than SUVs, although both are believed to have similar c.g. heights, and SUVs are historically well documented to have greater roll instability than other light truck vehicles including pickup trucks ( 8 , 22 , 23 ). Little information exists on CUV impact behavior with roadside hardware; therefore, research on CUV crashworthiness and barrier performance of existing MASH hardware is desired.

High-selling, front-wheel drive, compact CUVs are popular and widely available, making them another mid-size test vehicle option. The Honda CR-V, Nissan Rogue, and Toyota RAV4 were three of the five top-selling passenger vehicles in 2017. The compact CUV class vehicles possess curb weights within 150 lb of one another, shown in Table 4. They accumulated 1,523,354 total model sales in 2017 and are similar to the mass of the 1500A vehicle currently used in MASH.

A percentile weight is not specified for MASH mid-size vehicle, and there may be benefit associated with using a true mid-weight (50th percentile) vehicle for test evaluation. The 50th percentile passenger vehicle weight was approximately 3,850 lb, and using a 2.2% or 90-lb mass tolerance, the allowable weight range of a 50th percentile vehicle would be 3,760 lb to 3,940 lb. Researchers examined vehicles near the 50th percentile weight to determine if prospective test vehicles could be identified.

Mid-size, luxury, and sports cars in this weight range were deemed undesirable because of lower sales volumes, high cost, and large diversity of physical dimensions and attributes.

Large cars may be suitable for test vehicle selection. According to Wards’ classification system, large cars are distinguishable by a total length greater than 200 in. However, large car availability was also a concern, accounting for only 1.5% of all 2017 new vehicle sales. The 3,785-lb Chevrolet Impala (75,887 model units sold) and 3,935-lb Dodge Charger (88,351 model units sold) were potential candidate vehicles.

CUVs are also a potential 50th percentile weight vehicle. Sales and ownership of CUVs has exceeded most types of passenger cars since 2015, and because of higher c.g. heights and lower SSFs than cars, may be prone to increased instability issues. The front-wheel drive CUVs near the 50th percentile weight accumulated 707,626 total model sales in 2017.

For comparison, the 50th percentile weight CUVs are generally 7 in. longer and 450 lb heavier than the lighter compact CUVs. The contribution of wheelbase to crashworthiness is not well understood; however, larger wheelbases are generally associated with increased vehicle weights, which contribute to increase in impact severity and barrier loading and decrease in occupant impact accelerations and velocities.

Discussion

Registration and crash data were compared with sales data, and it was found national vehicle registrations were reflective of vehicle sales 12 years before registration year. Crash records were available from individual states, and it was found that in Ohio, Utah, and Wyoming, cars were 2%–9% more likely to be involved in accidents than light trucks. Sales data appeared to be an effective surrogate for approximation of vehicle fleet composition and were faster and easier to analyze, thus it is recommended that future studies primarily utilize new vehicle sales data when revising MASH standard test vehicle selection criteria.

Base trim level, front-wheel drive, compact, four-door sedans near 2,800 lb are recommended as MASH small passenger test vehicle. Half-ton, crew cab, four-wheel drive, medium-box pickups near 5,400 lb are recommended as MASH large passenger test vehicle. It is expected that the small car and pickup will increase the vehicle-to-barrier impact load. The recommended pickup is nearly identical in body style to the current MASH pickup, but the 400 lb weight increase will result in increased barrier loading and risk of pocketing and snag, potentially leading to failure performance of existing roadside structures.

Traditionally, a mid-size sedan has been used as the MASH mid-size test vehicle when necessary; however, the emergence of CUVs suggests CUVs warrant consideration for crash testing. Higher c.g. heights of CUVs, SUVs, and pickups can lead to vaulting of guardrail and concrete barriers and could contribute to vehicle rollover. In addition, if MASH were to a adopt a CUV mid-size vehicle, the compact CUV class near 3,385 lb is more available, with approximately 1.5 million annual vehicle sales. Although both vehicles exhibit similar instability, it is expected lighter weight CUVs experience increased accelerations. Currently, no MASH tests exist to explicitly evaluate the issue of mid-size vehicle stability, and it is uncertain how average CUV crash severities compare with other types of vehicles. Nonetheless, the significant volume of CUV sales and registrations compared with all other vehicle types supports the need to examine how CUVs interact with roadside features and to determine if changes to hardware design are needed to support these vehicles.

Ford produced the top-selling pickup trucks between 2017 and 2019 ( 24 ). Typical vehicle weights for the half-ton F-150 varied significantly from the three-quarter-ton F-250 and the 1-ton F-350. In 2017, the low, average, median, and high curb weights of the F-150 models were 4,050, 4,660, 4,560, and 5,700 lb, respectively, whereas the low, average, median, and high curb weights of all F-250 models were 5,680, 6,210, 6,200, and 6,700 lb, respectively. Vehicle weights went up as optional trim, suspension, box configuration, and cab configurations increased, as well with increased rated suspension capacity. However, as of 2019, the newest models Dodge Ram 1500, Chevrolet Silverado 1500, and GMC Sierra 1500 were expected to significantly reduce vehicle weights. For example, the redesigned 2019 Silverado 1500 crew cab 4×4 has a nominal curb weight of 5,090 lb, approximately 210 lb lighter than the 2018 model ( 25 ), and the Ram 2019 1500 Crew Cab 4WD (Big Horn/Lone Star) was also redesigned, such that the curb weight was reduced from 5,390 lb to 5,232 lb ( 26 ). If future models of half-ton, quad- or crew-cab pickup trucks have significantly reduced curb weights compared with 2017 distribution, the selection criteria for the light pickup truck vehicle may need to be reviewed to confirm it remains an adequate representation of heavy vehicle class for roadside hardware evaluation.

Vehicle shapes continue to evolve to adapt to aerodynamic, manufacturing, and appearance stylization of body components. Components including the front bumper, hood, and grill shapes have become increasingly difficult to characterize using a single dimension, such as “bumper height” or “hood height.” Researchers recommend further analysis on the most appropriate and robust methods of documenting vehicle shapes, geometries of critical components, and relationships including ground clearance and component heights.

To date, no clear relationship has been established which differentiates impact conditions by the vehicle class, inertial characteristics, body style, and other attributes of vehicles. Until such a relationship is established, it is generally perceived to be conservative for roadside feature design that run-off-road (ROR) impact conditions are independent of vehicle type. Thus, ROR impact conditions identified in studies such as NCHRP 17-22 and 22-14 ( 27 ) and the ongoing project NCHRP 17-43, do not affect the selection criteria for MASH crash test vehicles. The combination of vehicles selected near the anticipated performance boundaries of roadside hardware and “practical worst-case” impact conditions should provide a reasonably robust full-scale test evaluation of roadside features. Improved in-vehicle safety features may also improve survivability in all crashes. It will be necessary to review the criteria for evaluating roadside safety hardware performance as vehicle safety improves ( 20 ).

Conclusion

MASH recommends that the modern vehicle fleet be reviewed every 5 years. However, few research studies have reviewed vehicle parameters since MASH was published first in 2009, and revised in 2016. A discrete weight distribution based on 2017 passenger vehicle sales and curb weights identified 5th and 95th percentile weights of 2,800 lb and 5,850 lb, respectively. Test vehicle candidates near these weights were isolated, and dimensional properties were tabulated to make passenger vehicle recommendations for crash testing. The large test vehicle weight is recommended at the 92.5 percentile weight (5,400 lb) so that available, representative vehicles are obtainable for laboratories to perform replicable full-scale crash testing. Recommended vehicles and dimensional properties of small and large MASH passenger vehicles have been provided, and supplemental testing of recommended small and large test vehicles with existing barriers should be conducted to validate whether recommended vehicles are suitable for MASH implementation. It is anticipated that some geometrical or weight attributes of high-selling vehicles pose increased demand on roadside hardware, whereas other attributes amplify vehicle stability, occupant safety, and hardware performance. Therefore, it is critical to understand both the behavior of the roadside hardware and the impacting vehicles to determine the best practices and selection of MASH evaluation criteria and MASH crash test vehicle selection.

MASH mid-size test vehicle evaluates cable barrier penetration, staged energy-absorbing terminals, crash cushions, and truck-mounted attenuators. Mid-size sedans have traditionally been tested, but CUV sales indicate greater availability and representativeness. Viable mid-size test vehicle options include: (1) continue 3,300 lb mid-size sedan use; (2) increase mid-size sedan weight; (3) adopt class of high-selling, compact CUVs; and (4) adopt 50th percentile weight mid-size vehicle (CUV most prevalent class at 50th percentile weight). Further research on CUV impact behavior is needed to determine whether associated occupant risks require CUV implementation as a MASH test vehicle.

The next MASH revision may not occur until 2024. By that time, the vehicle fleet may have further evolved, and a similar vehicle selection study may be necessary. The following method is recommended when revising MASH standard test vehicle specifications:

1. Acquire Data by Make, Model, Trim Level (if available)

1.1. National new vehicle sales

1.2. Cab, suspension, engine configurations

1.3. Geometrical properties

1.4. Inertial properties

2. Correlate New Vehicle Sales with Curb Weight Data

2.1. When possible, use known sales with trim level, curb weight data

2.2. Use median curb weight method if sales distribution not known by trim level

3. Identify Candidate Vehicles

3.1. Identify 5th, 50th, 95th percentile weights, mid-size vehicle weight (to be determined [TBD])

3.2. Identify candidate vehicles near 5th, 50th, 95th percentile weights, mid-size vehicle weight (TBD) with a minimum of 50,000 model sales at desired trim level

3.3. Filter candidate vehicles based on engine type, cab and bed configuration (if applicable), suspension configuration, and powertrain properties

4. Revise Standardized Vehicle Selection Criteria

4.1. Identify acceptable range of variation in selection criteria (e.g., ±4 in. wheelbase)

4.2. Select nominal “target” criteria and tolerance to encompass most candidate vehicles

4.3. Reject candidate vehicles with significant deviation from nominal candidate vehicle range

Note, vehicles with APS or engine configurations had different geometrical properties and were typically a minimum of 200 lb heavier than traditional gasoline or diesel-powered vehicles of the same model. Sales of APS vehicles increased sharply around 2015; it is anticipated that public interest in electric vehicles will continue to expand. APS vehicles should be considered separately from conventional vehicles because of significant differences in structure, safety performance, vehicle risk factors (including fires or damage), and inertial properties.

Recommendations

It is recommended that exploratory testing be conducted using the new, proposed 2,800-lb small car and 5,400-lb pickup. The increased mass will increase typical IS-Values by 16% and 8%, respectively, and are likely to increase the vehicle-to-barrier impact loads. New full-scale crash test impact conditions, which are being investigated as a part of NCHRP Project No. 17-43, should be implemented into MASH at the same time as the new vehicle specifications. This research study could mimic the execution of NCHRP 22-14 during the initial adoption of MASH.

Several potential mid-size test vehicle candidates were discussed, including two mid-size sedan classes and two CUV classes. However, selection of the preferred vehicle or vehicles should be completed in conjunction with a review of in-service performance evaluation data and injury likelihood. The increase in SUV and CUV sales indicates that people enjoy CUVs and SUVs and, when presented with a choice, tend to select larger vehicles which are perceived to be safer in vehicle-to-vehicle impacts. The performance of CUVs with roadside hardware has not been established. It is not clear how long this trend will remain in the U.S. As well, it is also recommended c.g. heights of high-selling, large SUVs be experimentally determined to verify the 28.0 in. c.g. height MASH requirement is reflective of modern SUVs.

This research study was intended to review and update MASH test vehicle selection criteria, and to identify a “best-practice” for similar reviews to be conducted in the future. Recommendations from this study are similar to those which supported NCHRP Project 22-14, which experimentally investigated MASH vehicle interactions with hardware tested and approved during NCHRP Report No. 350. Future revisions for criteria for selection of MASH impact conditions at varying test levels may result from NCHRP Project 17-43; however, the selection of test vehicles is primarily based on the assumption that impact conditions are not strongly related to the type of passenger vehicle involved in a crash. Thus, revisions for test vehicle selection criteria and impact conditions for varying MASH test level evaluations should be simultaneously included in future revisions to MASH requirements.

Lastly, clarification on the most appropriate and robust methods of documenting vehicle shapes, geometries of critical components, and relationships including ground clearance and component heights is recommended. Practices for measuring critical geometrical parameters should be uniform for all laboratories conducting full-scale crash testing.

Footnotes

Acknowledgements

The authors wish to acknowledge several sources that made contributions to this project: NCHRP, 4N6XPRT Expert AutoStats, Canadian Vehicle Specifications, and Federal Highway Administration/Technical Committee on Roadside Safety. Additional acknowledgment to the following state Departments of Transportation for crash record contributions: Indiana, Nebraska, North Carolina, Ohio, South Carolina, Utah, and Wyoming. Authors also wish to acknowledge the Dwight D. Eisenhower Transportation Research Fellowship Program for supporting student research and providing the opportunity to travel to TRB.

Acknowledgment is also given to the following invididuals for contributions to the completion of this research project.

National Cooperative Highway Research Program: David Jared, Senior Program Officer, Mark Bush, Senior Program Officer. American Association of State Highway and Transportation Officials: Kelly Hardy, Safety Program Manager. Federal Highway Administration: Will Longstreet, Highway Safety Engineer. South Dakota Department of Transportation: Bernie Clocksin, P.E., Standards Engineer. Washington Department of Transportation: John Donahue, Design Analysis and Policy Manager. Wisconsin Department of Transportation: Erik Emerson, P.E., Standards Development Engineer. 4N6XPRT Expert AutoStats: Daniel Vomhof III, Accident Reconstructionist. Midwest Roadside Safety Facility: Josh McCann, Undergraduate Research Assistant, Connor Raatz, Undergraduate Research Assistant.

Data for this research effort analyzed proprietary sales databases owned by Wards Intelligence. Crash data and vehicle registration data are available through state DOT resources and the Federal Highway Administration.

Author Contributions

The authors confirm contribution to the paper as follows: study conception and design: C. Stolle, R. Bielenberg, R. Faller; data collection: K. Ronspies, C. Stolle; analysis and interpretation of results: K. Ronspies, C. Stolle, R. Bielenberg, R. Faller; draft manuscript preparation: K. Ronspies, C. Stolle. All authors reviewed the results and approved the final version of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this research effort was provided by the National Cooperative Highway Research Program, Project No. 20-07 Task 372: Evaluation of MASH Test Vehicles.

References

American Association of State Highway and Transportation Officials. Manual for Assessing Safety Hardware (MASH), 2nd ed. American Association of State Highway and Transportation Officials (AASHTO), Washington, D.C., 2016.

Transportation Research Circular 191: Recommended Procedures for Vehicle Crash Testing of Highway Appurtenances. TRB, National Research Council, Washington, D.C., Feb. 1978.

Michie

J. D.

NCHRP Report 230: Recommended Procedures for the Safety Performance Evaluation of Highway Appurtenances. TRB, National Research Council, Washington, D.C., 1981.

American Association of State Highway and Tranportation Officials. Guide Specifications for Bridge Railings. American Association of State Highway and Tranportation Officials, Washington D.C., 1989.

Ross

H. E.

Sicking

D. L.

Zimmer

R. A.

Michie

J. D.

NCHRP Report 350: Recommended Procedures for the Safety Performance Evaluation of Highway Features. Transportation Research Board of the National Academies, Washington D.C., 1993.

Eskandarian

Bahouth

Digges

Godrick

Bronstad

NCHRP Web Document 61 (Project 22-15): Improving the Compatibility of Vehicles and Roadside Hardware Safety. Transportation Research Board of the National Academies, Washington, D.C., 2004.

Buth

C. E.

Campis

W. L.

Griffin

L. I.

III Love

M. L.

Sicking

D. L.

Performance Limits of Longitudinal Barrier Systems Volume I – Summary Report. Report No. FHWA/RD-86/153. Texas Transportation Institute, College Station, TX, 1986.

Gabauer

D. J.

Gabler

H. C.

Differential Rollover Risk in Vehicle-to-Traffic Barrier Collisions. Annals of Advances in Automotive Medicine, Vol. 53, Chicago, IL, 2009.

Wards Intelligence. Wards ’19 Light-Vehicle U.S. Market Segmentation Criteria. Wards Intelligence, Southfield, MI, 2018.

10.

Bureau of Transportation Statistics. Average Age of Automobiles and Trucks in Operation in the United States. https://www.bts.gov/content/average-age-automobiles-and-trucks-operation-united-states. Accessed July 15, 2019.

11.

Policy and Government Affairs Office of Highway Policy Information. Highway Statistics Series (2000-2017). U.S. Department of Transportation, Federal Highway Administration, Washington D.C. https://www.fhwa.dot.gov/policyinformation/statistics.cfm. Accessed July 15, 2019.

12.

Transport Canada. Guide to the Canadian Vehicle Specifications Database. Transport Canada, Ottawa, 2012.

13.

Libby

. Looking Beyond Sport Utilities, Pickups and Sedans to the Other Body Types. HIS Markit, 2018. https://ihsmarkit.com/research-analysis/Looking-beyond-sport-utilities-pickups-sedans.html/.

14.

Edmunds. 2019 Ram 1500 Classic Quad Cab Features & Specs. Edmunds.com., Inc. https://www.edmunds.com/ram/1500-classic/2019/quad-cab/features-specs/.

15.

Edmunds. 2019 Ram 1500 Classic Crew Cab Features & Specs. Edmunds.com., Inc. https://www.edmunds.com/ram/1500-classic/2019/crew-cab/features-specs/.

16.

Expert AutoStats . Data Definitions, 4N6XPRT Systems, La Mesa, CA, 2018.

17.

Polivka

K. A.

Faller

R. K.

Sicking

D. L.

Rohde

J. R.

Bielenberg

B. W.

Reid

J. D.

Performance Evaluation of the Modified G4(1S) Guardrail – Update to NCHRP 350 Test No. 3-11 (2214WB-1). MwRSF Report No. TRP-03-168-06. Final Report to the National Cooperative Highway Research Program, Midwest Roadside Safety Facility, University of Nebraska-Lincoln, 2006.

18.

IIHS & HLDI. 2016 Ford F-150. Insurance Institute for Highway Safety and Highway Loss Data Institute. https://www.iihs.org/ratings/vehicle/ford/f-150-crew-cab-pickup/2016.

19.

IIHS & HLDI. 2019-20 Ram 1500 Earns Good Headlight Rating and TOP SAFETY PICK+. Insurance Institute for Highway Safety and Highway Loss Data Institute. https://www.iihs.org/news/detail/2019-20-ram-1500-earns-good-headlight-rating-and-top-safety-pick-. Accessed September 10, 2019.

20.

Atiyeh

. Three Pickup Trucks Ace IIHS Crash Tests, Yet Many Aren’t As Safe As They Should Be. Insurance Institute for Highway Safety and Highway Loss Data Institute (IIHS & HLDI). https://www.caranddriver.com/news/a26894769/full-size-pickup-trucks-crash-test-iihs/. Accessed March 21, 2019.

21.

Special Report 265: The National Highway Traffic Safety Administration’s Rating System for Rollover Resistance, An Assessment . Transportation Research Board of the National Academies, Washington, D.C., 2002.

22.

Newstead

Keal

Induced Exposure Estimates of Rollover Risk for Different Types of Passenger Vehicles. Traffic Injury Prevention, Vol. 10, No. 1, 2009, pp. 30–36.

23.

Schaal

10 Trucks and SUVs with the Highest Risk of Tipping Over. MotorBiscuit, 2019. https://www.motorbiscuit.com/trucks-suvs-highest-risk-of-tipping-over/.

24.

GoodCarBadCar Reports. 2019 U.S Pickup Truck Sales Analysis. Automotive Sales Data & Statistics. https://www.goodcarbadcar.net/2019-us-pickup-truck-sales-figures-by-model/. Accessed January 3, 2020.

25.

Edmunds. Chevrolet 1500. Edmunds.com. https://www.edmunds.com/chevrolet/silverado-1500/.

26.

Edmunds. Ram 1500. Edmunds.com. https://www.edmunds.com/ram/1500/.

27.

Mak

K. K.

Sicking

D. L.

Coon

B. A.

Albuquerque

F. D. B.

NCHRP Report 665: Identification of Vehicular Impact Conditions Associated with Serious Ran-off-Road Crashes. Transportation Research Board of the National Academies, Washington, D.C., 2010.