Paper to be presented at the 1999 SAE International Congress & Exposition , March 1- 4, 1999 at the Cobo Center in Detroit, Michigan
The New Car Assessment Program: Has It Led to Stiffer Light Trucks and Vans over the Years?
Brian T. Park, James R. Hackney, Richard M. Morgan, and Hansun Chan
National Highway Traffic Safety Administration
Johanna C. Lowrie and Heather E. Devlin
Since model year 1983, one hundred and seventy five light trucks, vans, and sport utility vehicles (LTVs) have been included in the New Car Assessment Program (NCAP) frontal crash tests. In this frontal test, vehicles are crashed at 35 mph such that the entire front impacts against a rigid, fixed barrier. Instrumented anthropometric dummies are placed in the driver and right front passenger seats. Accelerometers are placed on the vehicle to record the response of the structure during the crash.
A number of recent papers have examined the compatibility of LTVs and cars in vehicle-to-vehicle collisions. The studies in these papers, generally, consider three factors for vehicle-to-vehicle compatibility: (1) mass, (2) stiffness, and (3) geometry. On June 5, 1998, Transport Canada and the National Highway Traffic Safety Administration held a forum entitled "Transport-NHTSA International Dialogue on Vehicle Compatibility," in Windsor, Canada. At the forum, representatives of major vehicle manufacturers expressed the view that NCAP has led to the design of more aggressive stiffness parameters for LTV front structures. With crash data on 175 LTVs distributed over the last 15 model years, it is possible to examine and analyze the actual trends of the stiffness and structural characteristics of this class of vehicles. In this paper, the acceleration data from accelerometers in the occupant compartment and from the dummies are analyzed to determine:
Three parameters, that are associated with energy management and with the crushing of the front structure (the crash signature of crash pulse) of the vehicle, are studied relative to the total stiffness or aggressivity of LTVs. These three parameters are (1) maximum dynamic crush, (2) maximum acceleration of the vehicle structure, and (3) the time period of the acceleration pulse. Approximate linear stiffness values are calculated from the occupant compartment accelerations and the vehicle masses for the first 200 mm of vehicle crush during the rigid barrier impact. The stiffness of these first 200 mm is an approximate measure of the structural aggressivity of the striking vehicle in front to side impacts. The effect of these aggressivity and stiffness parameters on the dummy responses is examined. It is shown that, on the average,
(1) the maximum dynamic crush and the time period of the acceleration pulse have increased over time. The maximum acceleration of the vehicle structure has decreased over time.
(2) the approximate linear stiffness in the first 200 mm of crush has decreased over time, and
(3) there is a correlation between lower stiffness and aggressivity parameters and better NCAP scores.
These findings indicate that NCAP may have influenced manufacturers to design less aggressive, and, therefore, more compatible LTV front structures for both front to front and front to side impacts.
In the USA, NCAP has appraised the safety performance of passenger cars in frontal impact tests since model year 1979.  NCAP scores have improved since the beginning of the program. As shown in Figure 1, significant reductions in average head injury criterion (HIC) from approximately 1300 in 1980 to about 600 in 1998 have occurred in passenger cars. These lower HIC's suggest a reduction in the probability of serious head injury from approximately 40 percent for a HIC of 1300 to 5 percent for a HIC of 600. [2, 3]
A similar reduction in fatality rates (fatalities per 100 million vehicle miles traveled) was observed in the USA from 1979 through 1996.  The fatality rates for passenger cars are shown in Figure 2.
The Congressionally-mandated National Academy of Science study, Shopping for Safety -- Providing Consumer Automotive Safety Information,  states that "NCAP scores have improved steadily since the inception of the program, with the greatest improvements in the early years ... Improvements in test performance have been matched by real-world reductions in fatality likelihood for drivers in head-on crashes similar to those simulated by the NCAP test.
The NCAP is not the sole stimulus for this improvement in safety; the 1984 regulations leading to automotive passenger restraint systems and air bags were another important factor. But the program can claim part of the credit." In addition, a General Accounting Office (GAO) study  found "Nonetheless, it seems reasonable to conclude that manufacturers' successful efforts to improve their products' performance in NHTSA crash tests, particularly in NCAP, have contributed to improved occupant protection in real-world crashes."
A study has established statistically significant correlation between NCAP performance and the fatality risk of belted drivers in actual head-on collisions. A restrained driver of a car which performs well in NCAP is, on average, about 15 to 30 percent less likely to be killed than the driver of a car which performs poorly in NCAP. 
Beginning with the model year 1994 vehicles, the agency adopted a simplified nonnumeric format, the five star rating, for the frontal NCAP results. NHTSA wanted to give the US consumer easily grasped vehicle safety performance information. This nonnumeric format is based on the use of injury risk functions, that relate the Hybrid III dummy measurements to injury probabilities. The head and chest injury data is combined into a single rating, reflected by the number of stars.
|= 10% or less chance of any serious injury to the head or chest|
|= 11% to 20% chance of serious injury|
|= 21% to 35% chance of serious injury|
|= 36% to 45% chance of serious injury|
|= 46% or greater chance of serious injury|
From a historical perspective, the passenger cars, in frontal NCAP, received many one star ratings in 1979. Over the years, the star ratings in the passenger cars moved to five and four stars as shown in Figure 3.
In model year 1997, 85 percent of the passenger cars had four or five stars in frontal NCAP testing compared to only 33 percent in 1979. The LTVs, in frontal NCAP, had a parallel showing from 1984 to 1997, as shown in Figure 4. In the NCAP testing, crash protection is provided to the driver and right front seat passenger using all occupant protection equipment, including safety belts.
With LTVs becoming a much greater percentage of the vehicle fleet, there is a increasing concern about their compatibility in crashes with other types of vehicles. A number of recent papers [9, 10, 11, 12, and 13] have examined the compatibility of LTVs and cars in vehicle-to-vehicle collisions. The studies in these papers, generally, consider three factors for vehicle-to-vehicle compatibility: (1) mass, (2) stiffness, and (3) geometry. In addition, at the Transport-NHTSA International Dialogue on Vehicle Compatibility, held in June, 1998, representatives of major vehicle manufacturers expressed the view that NCAP has led to the design of more aggressive stiffness parameters for LTV front structures. This study examined 175 LTVs to answer two questions. The two are (1) have LTVs become stiffer over the fourteen years that NCAP has been testing these vehicles and (2) do stiffer LTVs do better in NCAP?
STIFFNESS IN AN AUTOMOTIVE CRASH TEST
In engineering terms, stiffness is the ability of a structure to resist a deformation within its elastic (structure returns to its original form) range of behavior. A high-speed automotive crash is highly non elastic with permanent deformation of the LTV. Figure 5 shows the total force-crush (non-linear stiffness) of three vans as derived from the acceleration data from the vehicles' occupant compartments. ( The electronic data from accelerometers positioned at different locations on the vehicle structure were examined. It was decided that the accelerometers located closer to the rear of the vehicle gave a better indication of the overall motion of the vehicle.) The 1987 and 1992 Ford Aerostar vans are seen to rise to a higher force level and to have less dynamic crush. Ford's replacement van, the Windstar has a lower force level and higher dynamic crush in 1995 and 1998. The later Ford Windstar is less "stiff" and should be less aggressive in vehicle to vehicle crashes than the earlier Ford Aerostar.
Similarly, Figure 6 shows the force-crush of the Ford F150 light truck. Ford's 1984 F150 rises to a higher force level and a lower dynamic crush. In contrast, the 1994 and 1998 F150s rise to a lower force level and higher dynamic crush. Ford's later F150 is less "stiff" than the earlier F150.
Finally, Figure 7 exhibits the force-crush for the Chevrolet C/K Pickup. From 1988 through 1996, the structure of the Chevrolet C/K Pickup appears to have remained approximately constant. In reviewing Figures 5 through 7, it can also be noted that the force-crush curve is approximately linear during the first 200 mm of crush.
EFFECT OF STRUCTURAL DESIGN ON SAFETY PERFORMANCE
Hackney and Kahane (8) examined energy management capabilities of frontal structure in NCAP. The crash pulse of the NCAP vehicles, as measured by accelerometers in the occupant compartment, was used to establish how energy management of the front structure reduced dummy loading. They found that three parameters--associated with crushing of the front structure during the impact--were correlated with lower risk and higher risk cars. The three parameters are (1) peak deceleration as measured in the occupant compartment, (2) maximum vehicle dynamic crush, and (3) the time duration of the crash pulse, from instant of impact until the vehicle reaches maximum rebound velocity.
These three structural parameters were examined in this paper. The dependent variable was the probability of a severe head or chest injury. The independent variable was one of the three structural parameters. In Figure 8, for air bag and belt restrained occupants (139 data points) in LTVs, the dummy responses are plotted versus the maximum vehicle dynamic crush, Xpeak. A linear regression analysis was used to find the correlation of Xpeak with the probability of AIS 4 head or chest injury. A R-squared value of 0.21 was generated based on the LTV data (belt and air bag restrained dummies). These data indicate a general trend of lower injury probability with increasing vehicle crush.
Figure 9 is a bar graph for all the 175 LTVs tested in NCAP. The abscissa divides the period from 1983 through 1998 so that roughly the same number of vehicles was crashed in each of the three spans. The average mass for the three time spans are 1841 kg, 1994 kg, and 2090 kg respectively. Plus and minus one Standard Deviation is denoted in Figure 9 and subsequent bar charts. The ordinate is the average of the maximum dynamic crush, Xpeak. Over the past fourteen years of NCAP testing, on the average, the total crush of the LTVs has increased. This trend is consistent with LTVs incorporating softer structures over the fourteen years that NCAP has been testing these vehicles.
Similarly, it was established that, for belt and air bag restrained occupants in LTVs, peak vehicle deceleration is correlated with the probability of a severe head or chest injury. Also, the time duration of the vehicle crash pulse is inversely correlated with the probability of a severe head or chest injury. Over the past fourteen years, the peak deceleration has decreased on the average and the time duration of the crash pulse has increased on the average. These two trends also are consistent with LTVs incorporating softer structures. The details of the peak deceleration and time pulse analyses are presented in Appendix A.
EFFECT OF STRUCTURAL RESPONSE ON SAFETY RATING
A structure is only one part of the safety system that controls the forces going into the driver and passenger dummies. The safety system includes the seat belt and the air bag. For the frontal NCAP, NHTSA reports crash test results in a range of one to five stars, with five stars showing the best crash protection for LTVs. Dummy head and chest data, which show the chance of a life-threatening injury, are combined into a single rating, reflected by the number of stars. These represent a vehicle's relative level of crash protection in a high speed frontal crash.
In the 175 LTV tests, 93 driver and passenger occupants were (1) seat belt and air bag restrained and (2) received a high or a low safety rating. A high safety rating means four or five stars. A low safety rating means one or two stars. Figure 10 is a bar graph for the 93 occupants.
The abscissa divides the occupants into a low and high safety rating. The ordinate is the average of the maximum dynamic crush. Higher rated occupants are associated with, on the average, higher crush of the LTVs. This trend is consistent with softer LTVs doing better in NCAP.
Other indicators of structural behavior suggest that softer LTVs are associated with higher safety ratings. Figure 11 suggests that, on the average, the peak value of vehicle acceleration is lower for higher rated vehicles. Similarly, Figure 12 suggests that, on the average, vehicles that take a longer time to crush have a higher safety rating. For air bag-equipped LTVs with a high safety rating, the peak value of vehicle acceleration is lower and the time that the vehicle takes to crush is higher.
AVERAGE STIFFNESS IN THE FIRST 200 MM OF CRUSH
A number of recent papers (9, 10, 11, 12, and 13) have examined the compatibility of LTVs and cars in vehicle-to-vehicle collisions. In general, these studies consider three factors for vehicle-to-vehicle compatibility: (1) mass, (2) stiffness, and (3) geometry. In addition, at the Transport-NHTSA International Dialogue on Vehicle Compatibility held in June, 1998, representatives of major vehicle manufacturers expressed the view that NCAP has led to the design of more aggressive stiffness parameters for LTV front structures. While Figure 9 and Appendix A indicate that the average total stiffness of LTVs has decreased over the fourteen years of NCAP testing, no insight is offered about the stiffness at the very front of the LTV. Initial frontal stiffness may be considered as an approximate measure of the structural aggressivity of the striking vehicle in front to side impacts. This section will examine the approximate linear stiffness in the first 200 mm of vehicle crush.
It was observed earlier that, for the initial 200 mm of crush, the force is approximately a linear function of crush. For each of these 175 LTVs, approximate linear stiffness values are calculated from the occupant compartment accelerations and the vehicle masses for the first 200 mm of vehicle crush during the rigid barrier impact. Figure13 shows the average stiffness in the first 200 mm of crush for 175 LTVs. The abscissa divides the period from 1983 through 1998 so that roughly the same number of vehicles was crashed in each of the three spans. Over the past fourteen years of NCAP testing, on the average, the initial stiffness of the first 200 mm of the front structure of the LTVs has decreased. This trend is consistent with LTVs incorporating softer structures at the front over the fourteen years that NCAP has been testing these vehicles. These data indicate that, relative to vehicle stiffness, the aggressivity of LTVs in side impacts has not increased.
Over the past fourteen years of NCAP frontal testing, 175 LTVs were crashed. The total force-versus-crush of LTVs in high speed barrier impacts is non-linear. Examination of the force-versus-crush curves suggests three parameters to assess stiffness. These three structural parameters are (1) maximum vehicle dynamic crush, (2) peak deceleration as measured in the occupant compartment, and (3) the time duration of the crash pulse, from instant of impact until the vehicle reaches maximum rebound velocity.
Over the past fourteen years of NCAP testing, on the average, the total crush of the LTVs has increased, the peak deceleration in the occupant compartment has decreased, and the time duration of the crash pulse has increased. The trend of each of these three parameters is consistent with a reduction in the total stiffness of frontal structures of LTVs.
For LTVs with air bags, on the average, the vehicles with greater total crush, lower peak occupant compartment acceleration, and longer time duration of the crash pulse have a higher safety rating. This is consistent with higher NCAP rated LTVs having a softer structure.
An approximate linear structural stiffness can be computed for the first 200 mm of vehicle crush. This initial frontal stiffness is considered an approximate measure of the structural aggressivity of the striking vehicle in front to side impacts. Over the past fourteen years of NCAP testing, on the average, the initial stiffness of the LTVs has decreased. This is consistent with LTVs incorporating softer structures at the very front.
As stated earlier, this study examined 175 LTVs to answer two questions (1) have LTVs become stiffer over the fourteen years that NCAP has been testing these vehicles and (2) do stiffer LTVs do better in NCAP? From these data and the assumptions of these analyses, it is concluded that, on average, LTVs have become less stiff and, therefore, potentially less aggressive in vehicle-to-vehicle crashes, and the less stiff LTVs have higher NCAP rating.
1. Hackney, J. R., "The Effects of FMVSS No. 208 and NCAP on Safety as Determined from Crash Test Results," Proceedings of the Thirteenth International Conference on Experimental Safety Vehicles, Paris, France, November 1991.
2. Prasad, P., and Mertz, H., "The Position of the United States Delegation to the ISO Working Group 6 on the Use of HIC in the Automotive Environment," SAE Paper 851246, presented at the SAE Government/Industry Meeting and Exposition, Washington, DC, May 1985.
3. Viano, D. C., and Arepally, S., "Assessing the Safety Performance of Occupant Restraint Systems," Proceedings of the 34th Stapp Car Crash Conference, SAE Paper 902328, Warrendale, PA, November 1990.
4. Traffic Safety Facts 1996, National Highway Traffic Safety Administration, U.S. Department of Transportation, Report No. DOT HS 808 649, pg. 17, December 1997.
5. Shopping for Safety: Providing Consumer Automotive Safety Information, Transportation Research Board, National Research Council, Special Report 248, National Academy Press, Washington, DC, 1996.
6. GAO, Highway Safety: Reliability and Validity of DOT Crash Tests, GAO/PEMD-95-5, May 1995.
7. Kahane, C. J. , Hackney, J. R., and Berkowitz, A. M.., Correlation of NCAP Performance with Fatality Risk in Actual Head-On Collisions, Report No. DOT HS 808 061, National Highway Traffic Safety Administration, Washington, DC, 1994.
8. Hackney, J. R., and Kahane, C. J., "The New Car Assessment Program: Five Star Rating System and Vehicle Safety Performance Characteristics," SAE Paper 950888, presented at the SAE International Congress and Exposition, Detroit, Michigan, February 1995.
9. Hollowell, W. T., and Gabler, H. C., "NHTSA's Vehicle Agressivity and Compatibility Research Program," Proceedings of the Fifteenth International Enhanced Safety Vehicle Conference, Melbourne, Australia 1996.
10. Hollowell, W. T., and Gabler, H. C., "The Aggressivity of Light Trucks and Vans in Traffic Crashes," SAE Paper No. 980908, SAE International Congress & Exposition, Detroit, Michigan, February 1998.
11. Insurance Institute for Highway Safety, "Status Report on Crash Compatibility," Arlington, VA, Vol. 33, No. 1, February 14, 1998.
12. Faerber, E., Cesari, D., Hobbs, A. C., Huibers, J., van Kampen, B., Paez, J., and Wykes, N. J., "Improvement of Crash Compatibility Between Cars," Proceedings of the Sixteenth International Enhanced Safety Vehicle Conference, Windsor, Canada, June 1998.
13. Schoenburg, R., and Pankalla, H., "Implementaion and Assessment of Measures for Compatible Crash Behavior using the Aluminum Vehicle as an Example," Proceedings of the Sixteenth International Enhanced Safety Vehicle Conference, Windsor, Canada, June 1998.
The purpose of Appendix A is to analyze the effect that peak deceleration, as measured in the occupant compartment, Apeak, has on the response of the dummies in LTVs. Also, the effect of the time duration of the crash pulse, Tpulse, is investigated in this Appendix.
As in the EFFECT OF STRUCTURAL DESIGN ON SAFETY PERFORMANCE section of this paper, the dependent variable was the probability of a severe head or chest injury. The independent variable was peak deceleration, Apeak. In Figure A1, for air bag and belt restrained occupants (139 data points) in LTVs, the dummy responses are plotted versus the maximum vehicle acceleration, Apeak. A linear regression analysis was used to learn the correlation of Apeak with the probability of AIS 4 head or chest injury. A R-squared value of 0.17 was generated based on the LTV data (belt and air bag restrained dummies). These data indicate a general tread of lower injury probability with decreasing structural deceleration.
Figure A2 is a bar graph for all the 175 LTVs tested in NCAP. The abscissa divides the period from 1983 through 1998 so that roughly the same number of vehicles was crashed in each of the three spans. The ordinate is the average of the maximum vehicle acceleration, Apeak. Over the past fourteen years of NCAP testing, on the average, the structural deceleration of the LTVs has decreased. This trend is consistent with LTVs incorporating softer structures over the fourteen years that NCAP has been testing these vehicles.
In Figure A3, for air bag and belt restrained occupants in LTVs, the dummy responses are plotted versus the time duration of the crash pulse, Tpulse. A linear regression analysis was used to ind the correlation of Tpulse with the probability of AIS 4 head or chest injury. A R-squared value of 0.16 was generated based on the LTV data (belt and air bag restrained dummies). These data indicate a general tread of lower injury probability with increasing time duration of the crash pulse.
Figure A4 is a bar graph for all the 175 LTVs tested in NCAP. The abscissa divides the period from 1983 through 1998 so that roughly the same number of vehicles was crashed in each of the three spans. The ordinate is the average of the time duration of the crash pulse, Tpulse. Over the past fourteen years of NCAP testing, on the average, the time duration of the crash pulse of the LTVs has increased. This trend is consistent with LTVs incorporating softer structures over the fourteen years that NCAP has been testing these vehicles.
|Make||Model||Body Type||Test Wt. (kg)||Apeak (G's)||Tpulse (sec)||Delta Xpeak (m)||K Value 1st 200 mm||Driver %||Pass. %||Restraint|
Body Type: MP - Multipurpose passenger vehicle (Sports utility) PU - Pickup truck VN- Vans
Test Weight Measured weight before the test
Apeak: Acceleration in G's at the peak
Tpulse: A time duration between the initial rise of the crash pulse and the pulse to reach the zero
Delta X Peak: A crush distance between the initial deformation and the maximum deformation
K Value 1st 200mm: Approximated linear stiffness value for first 200mm of crush (force divided by deformation)
Driver % Driver combined (head and chest) injury probability of serious injury (AIS 4 or greater)
Pass. % Right front passenger combined (head and chest)injury probability of serious injury (AIS 4 or greater)
Restraint: 3PT: 3-point belt only PS2: Passive two point belt
ABG: air bag with 3-point belt
ND: No data