III. AIR BAG AGGRESSIVENESS TESTING

Introduction

Limited testing has been performed to determine the effect of depowering air bags on the types of fatalities that occur as a result of aggressive air bags. These fatalities include infants in rear-facing child seats, improperly-belted or unbelted children who were out-of-position (OOP) when the air bag deployed due to pre-impact maneuvers, and small statured drivers who sit close to the steering wheel in order to drive comfortably. Moving the seat forward puts the occupant in closer proximity to the air bag and increases the risk of air bag induced injuries. The punch out forces and other dynamic effects resulting from the inflation rate of the air bag can cause large forces to act on the occupant's head, neck and chest area, if they are very close to the air bag, resulting in the potential for severe to fatal injuries.

To evaluate the options available to the agency to reduce the potential for air bag deployment injuries, a test program was developed to assess air bag aggressiveness. Specifically, the objectives of the test program were:

1. Conduct a fleet characterization study of air bag aggressiveness on the driver and passenger-side for OOP occupants and examine air bag system characteristics that contribute to aggressiveness,

2. Compare current 96 MY air bag systems with earlier (94 MY) air bag systems where a change was made to the air bag system to reduce aggressiveness to OOP occupants,

3. Compare baseline and depowered air bag systems in a static test configuration with OOP driver and passenger occupants,

4. Compare baseline and depowered air bag systems in an FMVSS 208 sled test to determine whether, and to what extent, safety for the correctly positioned occupant is jeopardized by using a depowered system,

5. Examine the trade-offs in going to depowered inflators.

Out-of-Position Testing

Dummy Selection

Passenger-Side

For infants in rear facing child restraints, the 9-month old dummy, which is instrumented to record HIC and chest g's was used in testing.

Most child fatal injuries (excluding rear facing infants) have occurred by some mechanism involving applied air bag loads on the child's neck. It was imperative that the dummies selected for testing have a means to measure those types of loads. The current child 3 and 6 year-old dummies used for compliance testing of FMVSS No. 213 do not have a mount for a neck load cell, nor is the neck biofidelic. The most advanced child dummies are the H-III 3 and 6 year-old dummies produced by First Technology Safety Systems (FTSS). These dummies have H-III necks scaled down from the H-III 50th male, with neck load cells and the neck biofidelity is much better than the current dummies used in compliance testing.

Driver Side

The fatalities found in the real-world have been to predominantly small stature adults. The H-III 5th percentile female is a surrogate for small stature adults and was chosen for testing OOP drivers. Measurements of the dummy upper and lower neck loads, head and spine accelerations, and chest deflection were recorded in the tests.

Test Buck Configuration

Passenger-Side

Infants

Different rear facing child restraints were put in front of a mid-mount air bag and a top-mount air bag. The vehicle seat position was moved from full forward, to mid-seat, to rear positions. Very high dummy readings were recorded with the seat forward and in one case with the vehicle seat in the mid position. None of the dummy readings exceeded 1000 HIC or 60 chest g's with the vehicle seat in the rear position. However, the agency has not conducted any testing in these similar situations with a depowered air bag to determine whether any benefits could be achieved for rear-facing infant restraints leaning on the air bag cover. Table III-1 shows the high dummy readings found in some of the tests.

Table III-1 Dummy readings in those tests that exceeded FMVSS 213 criteria
	HIC	Chest G
Rear-Facing 1, 9-mo. Dummy - Seat Fwd - Test 1, Mid-Mount	3156	50
Rear-Facing 1, 9-mo. Dummy - Seat Fwd - Test 2, Mid-Mount	2857	54
Rear-Facing 2, 9-mo. Dummy - Seat Fwd - Test 1, Mid-Mount	2046	86
Rear-Facing 2, 9-mo. Dummy - Seat Fwd - Test 2, Mid-Mount	958	83
Rear-Facing 2, 9-mo. Dummy - Seat Mid, Top-Mount	1190	49
Rear-Facing 2, 9-mo. Dummy - Seat Fwd, Top Mount	3015	61

Child Passengers

Testing with the OOP child dummies was conducted on a static test fixture. Typically, the vehicle seat, instrument panel (IP), and tubular frame supporting the IP were mounted on the test fixture to re-create the vehicle interior dimensions. A plexiglass windshield was mounted so that height and angle could be adjusted to that of the actual vehicle over the passenger compartment. The dummy was placed in a pre-described position in front of the air bag and the bag deployed. The position in front of the bag represents an OOP situation attainable during pre-impact maneuvers (see Figure 1).

Driver-Side

For driver's-side testing, the 5th female H-III was placed on a generic seat that can be moved forwards and backwards. A non-deformable steering-wheel column was used to mount each vehicle's steering wheel and the height and angle adjusted to the vehicle's dimensions. A typical set-up for this position is shown in Figure 2.

Positioning Dummy for Static Test

Figure 2 Set-up for OOP driver-side Test

Passenger-Side -- 3 and 6 year old dummy testing

The position to set the child dummies was developed using the ISO DTR 10982 as a guideline.

The ISO document lists 4 positions for setting a 3-yr old dummy for OOP air bag deployment tests. For this test series, 3 of the 4 positions were adopted as guidelines for setting dummies. The fourth position in the ISO document was developed to test the aggressiveness of low-mounted air bag systems on OOP occupants. Since no manufacturer currently uses low-mounted air bags, that position was not included. While the ISO positions were used as guidelines, NHTSA developed specific procedures and justifications for those procedures to represent three scenarios that were deemed as reasonable conditions that occupants may move into during pre-impact maneuvers. It is typically unknown what the exact position of the occupant is when the air bag deploys. The following lists each of the three positions and gives a brief explanation of their development.

OOP Child Position 1

figure 3a

figure 3b

figure 4

Position 1 is designed primarily to evaluate contact forces of the deploying air bag on the chest. However, head accelerations and neck loading are significant factors in this test position. The positioning is intended to represent a standardized worst case condition in which the child has been thrown against the frontal structures of the vehicle's interior due to pre-impact braking and/or vehicle impact (Figure 3 and 4). Position 1 could also resemble a 3 year old child standing in front of the air bag.

OOP Child Position 2

figure 5

Position 2 is designed to primarily address the contact forces and loading forces of the deploying air bag on the head and loading forces on the neck. The Child Position Number 2 is intended to represent a standardized worse case scenario in which the child slides forward or is sitting forward on the seat while the upper torso jack-knifes downward into the dashboard. The final positioning may not necessarily place the head into direct contact with the air bag's cover but does reflect a reasonable positioning based on estimated body kinematics resulting from pre-impact braking (Figure 5).

OOP Child Position 3

figure 6

Position 3 addresses the loading and contact forces on the head and loading forces on the neck of the dummy. The positioning is intended to simulate a worse case scenario in which the child is thrown to the floor of the vehicle and is in a kneeling posture. The final positioning may not necessarily place the head into direct contact with the air bag's cover, especially for a top mounted air bag, but does reflect a reasonable positioning for a child that has fallen from the seat into the foot well of a vehicle.

Driver's-Side

The DTR ISO 10982 also gives recommended practices for testing OOP adults in the driver's seating position. There are two prescribed positions to place an adult dummy to test air bag aggressiveness in either a static deployment scenario, or a HYGE sled. The first Position places the head in close proximity to the air bag module (OOP Driver Position 1). The second position places the dummy's chest in close proximity to the air bag module. Position 1 represents a realistic position that may be attainable during pre-impact maneuvers or during the crash. Position 2 tests the chest injury causing potential of the air bag. This position may not be realistically attainable during a crash, but it tests the worst case scenario by putting the occupant's chest on the air bag.

Static Test Results

Three manufacturers supplied air bags to be tested in time for the analysis in this FRE. Other manufacturers have agreed to supply more depowered air bags which have not been tested at this date. Table III-2 shows the vehicles types and the air bags tested (94 and 96 are the model years of the systems tested).

 

Table III-2 Vehicles with Baseline and Depowered Air Bags
Vehicles	Air bag Inflation
B-94	Baseline
	30 % Depowered
	60 % Depowered
D-96	Baseline
	18 % Depowered
I-96	Baseline
	23 % Depowered
	38 % Depowered

Depowering is defined by the percent reduction in the peak pressure of the air bag inflator. In most cases, the percent reduction in peak pressure correlates to approximately the same percentage reduction in the maximum inflator pressure rise rate. In none of the depowered air bags did the air bag itself undergo any re-design. Results of the depowering will be discussed on a vehicle-by-vehicle basis.

Table III-3 shows the peak values for an evaluation of B-94 with the 30 and 60 percent depowered air bags. The most challenging position to reduce injury measures is Position 1 for the 6YO. In this position, the bag concentrates under the chin and on the neck before the bag begins to expand with gas. Once the bag begins to fill, the dummy is actually lifted by the inflating bag, causing its head to be thrown back, resulting in high levels of neck tension, extension, and shear. Each level of depowering reduces injury measures significantly. Table III-3 continued gives the chance of fatality assuming head and neck injury curves and an overall combined chance of fatality given the severe test conditions.

Table III-3

Air Bag B-94 H-III 6-yr old
	Position 1				Position 2				Position 3
	Base	30%	60%		Base	30%	60%		Base	30%	60%
HIC	900	161	20		1531	298	66		854	34	2
Shear	4709	1889	544		3387	969	417		2540	689	48
Tension	6184	2383	1279		6661	2069	836		2152	1257	146
Compression	2	1	2		31	20	26		2152	1257	146
Flexion	0	1	0		32	21	10		10	3	2
Extension	175	83	31		68	51	20		40	39	5
N (i.j.)	6.85	3.80	1.11		4.38	2.62	1.21		2.69	2.00	0.41
H-III 3-yr old
					Position 2
					Base	30%	60%
HIC					3511	861	161
Shear					1142	855	269
Tension					2999	1951	902
Compression					103	105	34
Flexion					2	1	1
Extension					111	64	33
N (I. j.)					4.77	3.28	1.75

H-III 3 and 6-YO results of baseline and 30% and 60% depowered air bags for B-94.

Table III-3 cont. Percent Chance of Fatality (%) Air Bag B-94
H-III 6-yr Old
	Position 1			Position 2			Position 3
Head	Base	30%	60%	Base	30 %	60%	Base	30%	60%
Prasad/Mertz	.0625	.0004	0	2.36	.0013	0	.0476	0	0
Lognormal	4.72	0	0	23.6	.02	0	3.82	0	0
Neck
Overall	100	100	0.95	100	95	1.1	95	5	0.5
Prasad/Mertz	100	100	0.95	100	95	1.1	95	5	0.5
Lognormal	100	100	.95	100	95	1.1	95.22	5	.5
H-III 3 yr old
				Position 2
Head				Base	30%	60%
Prasad/Mertz				99.95	.0496	.0004
Lognormal				78.15	3.96	0
Neck
Overall
Prasad Mertz
Lognormal

Table III-4 Baseline and 18% depowered static test results for D-96
Air Bag D-96
H-III 6-yr old
	Position 1			Position 2
	Base	18%		Base	18%
HIC	865	43		300	109
Shear	777	414		651	575
Tension	1985	730		242	803
Compression	98	2		2735	1149
Flexion	27	4		19	38
Extension	29	22		9	16
N (I. j.)	1.73	1.44		1.97	1.29
H-III 3-yr old
	Position 1			Position 2
	Base	18%		Base	18%
HIC	299	135		649	237
Shear	724	475		601	372
Tension	1782	914		773	1239
Compression	0	0		1023	346
Flexion	30	15		9	6
Extension	30	15		43	27
N (I. j.)	2.60	1.14		1.75	1.88

Table III-4 continued Percent Chance of Fatality % Air Bag D-96
Air Bag D-96
H-III 6-yr old
	Position 1			Position 2
Head	Base	18%		Base	18%
Prasad Mertz	.0508	0		.0014	.0001
Lognormal	4.03	0		.02	0
Neck
Overall	1.75	1.4		4.5	1.2
Prasad Mertz	1.8	1.4		4.5	1.2
Lognormal	5.71	1.4		4.52	1.2
H-III 3-yr old
	Position 1			Position 2
Head	Base	18%		Base	18%
Prasad Mertz	.0013	.0002		.0139	.0008
Lognormal	.02	0		1.17	0
Neck
Overall
Prasad Mertz
Lognormal

Baseline and 18% depowered static test results for D-96

D-96 did not have a significantly aggressive air bag compared to B-94. There are design features that have been incorporated that seem to benefit the D-96 air bag over the B-94. Particularly, the placement of the air bag module has made it difficult to be in close proximity to the air bag deployment.

Again, the worst condition was Position 1. The depowered air bag had lower numbers in all categories for this position, with significant reductions in the tension, shear, and HIC measures.

One result which bears explanation occurred when comparing the baseline and depowered tension and compression results in position 2. In the baseline cases, the bag tended to go over the head catching the top of the 3 year-old and 6 year-old's head to different degrees. The depowered bags tended to get underneath the chin area more significantly, changing the loading from a predominantly compression load, to a tension load on the neck. As an example, with the 3 year-old, compression was 1023 Newtons in the baseline and 346 Newtons in the 18% depowered air bag. The tension changed from 773 to 1239 for the baseline and depowered system, respectively. Similar results are seen in the 6 year old in Position 2.

Results from I-96 with depowered air bags is shown in Table III-5. While results were directionally correct for the 6 year-old in all the higher measures, the 3 year-old did not show a similar benefit from depowering when the recorded peak values are examined. Comparison of the Ni,j's for the 3 year old show mixed results. While the peak neck measures values of repeat tests were very close to the original test, the timing of the peaks were different, allowing for differences in the calculated Ni,j.

Table III-5

Air Bag I-96
H-III 6-yr old
	Position 1				Position 2				Position 3
	Base	23%	38%		Base	23%			Base	23%
HIC	467	70	63		1909	602			1476	292
Shear	2481	1146	892		938	555			878	443
Tension	3476	1673	1534		3050	1515			2817	652
Compression	6	1	28		349	199			2	90
Flexion	1	0	1		6	3			6	2
Extension	86	59	48		51	29			50	49
N (I. j.)	4.10	2.18	1.89		2.42	1.97			2.63	1.66
H-III 3-yr old
	Position 1				Position 2
	Base	23%	38%		Base	23%	38%
HIC	254	204(235)	107		56(66)	201	9
Shear	1309	1497(1242	875		379(383)	925	63
Tension	2095	2130(2019	1578		22(1)	230	20
Compression	2	0(6)	0		1655(1639	649	539
Flexion	0	0(1)	0		2(2)	9	2
Extension	53	56(52)	36		19(20)	46	5
N (I. j.)	3.11	3.78(2.77)	1.88		1.33(1.91)	1.90	0.36
Baseline with 23% and 38% Depowered Static Test Results for I-96 (Numbers in parens are repeat tests)

Figures 7, 8, and 9 show selected graphs of data from the three vehicle tests. The maximum is the highest value from positions 1, 2 and 3. The average is a simple average of the results from the three positions.

Table III-5 continued Percent Chance of Fatality (%) Air Bag I-96
H-III 6-yr Old
	Position 1				Position 2				Position 3
Head	Base	23%	38%		Base	23 %			Base	23%
Prasad/Mertz	.0044	0	0		17.38	.0104			1.736	.0013
Lognormal	.0023	0	0		37.39	.86			21.26	.02
Neck
Overall	100	25	3.5		70	4.5			92.9	1.6
Prasad/Mertz	100	25	3.5		85.21	4.51			93.0	1.6
Lognormal	100	25	3.5		81.22	5.32			94.4	1.62
H-III 3 yr old
	Position 1				Position 2
Head	Base	23%	38%		Base	23%	38%
Prasad/Mertz	.0009	.0006	.0001		0	.0006	0
Lognormal	0	0	0		0	0	0
Neck
Overall	100	97	3.5		3.5	3.5	0.4
Prasad/Mertz	100	97	3.5		3.5	3.5	0.4
Lognormal	100	97	3.5		3.5	3.5	0.4

FIGURE 7

FIGURE 8

FIGURE 9

Driver-Side

Limited crash data suggest that the newer model year vehicles may be improving their OOP performance. The agency has examined 21 driver fatalities. Several vehicles have new air bag systems for MY 96 specifically designed to reduce the aggressiveness of air bag deployment to drivers. Three vehicles were selected with known modifications. These three vehicles are A, E, and F. Other vehicles were selected to help assess the aggressiveness of the baseline fleet.

Table III-6 shows the results for each vehicle in Position 1 and Position 2. Flexion and compression are very small in relation to the shear, tension and extension, and therefore are not shown. Position 2 places the chest over the air bag module and is a test of the air bag's aggressiveness on the dummy chest. Subsequently, chest measures are recorded for Position 2. In an attempt to place the chest close to the air bag module the head is placed with the chin on top of the steering wheel rim. In some situations, the neck extension and tension are a result of the chin getting caught on that portion of the steering wheel rim. This is further reason for examining only the chest measures for Position 2.

Table III-6 Comparison of 94 and 96 Model Year Vehicles
Vehicle	OOP Position 1			OOP Position 2
	Shear (N)	Tension (N)	Extension (N-m)	V*C	Chest Deflect (mm)	Chest g's
IARV	2068	2201	31	1	63	60
A-94	2153	2189	79	1.5	55	55
A-96	1253	2359	46	0.8	44	44
E-94	1283	2759	60	2.1*	66*	47*
E-96	1639	1909	67	0.8	44	54
F-94	1913	2385	70	1	43	61
F-96	784	1675	31	0.9*	36*	31*
* - Average of two tests where steering wheel broke in both tests

IARV's reported here were developed by GM, with the exception of 60 g's and 63 mm of chest deflection. The values listed refer to the peak values shown in this chart. Other factors, such as load duration are not discussed here. Generally, the newer air bag systems did show improvements, particularly in reducing chest loads in Position 2. Position 2 chest loads are created by the punch-out force of the air bag. All three systems reduced V*C to below 1, which is the accepted criteria level. Chest deflections and chest g's were also reduced to below reference values.

Loads on the neck were generally better, particularly for F-96. Overall, the loads in Position 1 were lowest for this vehicle. E-96 had an increase in neck shear and extension. Only the neck extension exceeded IARV values. Trends for A-96 showed improved shear and extension loads, but increased tension on the neck when compared to A-94. The newer MY, A-96, exceeded IARV values for Tension and Extension.

Table III-7 Comparison of MY 96 Vehicle Test Results
Vehicle	OOP Position 1			OOP Position 2
	Shear (N)	Tension (N)	Extension (N-m)	V*C	Chest Deflect (mm)	Chest g's
IARV	2068	2201	31	1.00	63	60
A-96	1253	2359	46	0.80	44	44
B-96	2198	1846	84	1.10	39	47
D-96	592	1762	31	1.80	53	37
E-96	1639	1909	67	0.80	44	54
F-96	784	1675	31	0.90*	36*	31*
H-96	1995	2209	77	2.00	56	58
I-96	424	2298	22	1.80	73	**
Avg.	1269	2008	51	1.31	49	45
Std. Dev.	699	275	25	0.53	13	10
*Average of Two Tests ** - data channel failure

Based on the results of the comparison of 94 and 96 MY vehicles with known attempts to reduce air bag aggressiveness, it was concluded that while manufacturers were improving their systems, there was still a significant level of aggressiveness that could cause serious injury or fatality.

Table III-8 shows the results of the fleet characterization study with depowered air bags conducted for driver-side systems. Two manufacturers provided depowered air bags for the driver-side OOP tests. When looking at I-96, the first level of depowering reduced the neck tension to below IARV. However, it required the 42% depowering to reach a V*C below 1 on the chest. F-96 reached all the IARV values with 35 % depowering in Position 1. Chest measurements in Position 2, however, were not reduced and appeared to be a little higher. The agency believes the higher chest measurements are the result of the way this bag was depowered. The air bag inflator had a reduced peak pressure, but not a reduced pressure rise rate. Since it is the break-out forces that contribute the most to the V*C measurement, it appears that reducing peak pressure will not reduce injuries from this type of loading without a corresponding reduction in the pressure-rise rate.

 

Table III-8 Driver Depowered Air Bag Results
Vehicle	OOP Position 1			OOP Position 2
	Shear (N)	Tension (N)	Extension (N-m)	V*C	Chest Deflect (mm)	Chest g's
IARV	2068	2201	31	1	63	60
I-96	424	2298	22	1.8	73	**
I-96-D25	242	1216	13	1.3	57	31
I-96-D42	283	1279	9	0.5	39	41
F-96	784	1675	31	0.9*	36*	31*
F-96-D35	711	1077	28	1.3*	42*	30*
*Average of Two Tests ** - Results questionable

Sled and Vehicle Test Results with Depowered Air Bags

A sled pulse that simulated the FMVSS 208 crash compliance test was created to test each vehicle's baseline and depowered air bag module. The object was to compare the baseline air bag, which was designed to meet 208 requirements, with depowered air bags to determine how the injury measures changed. Table III-9 is the tabular data showing the results of testing vehicle I-96, B-94, and D-96 on the passenger-side. The test was conducted with an unrestrained 50th percentile H-III dummy. A sled pulse with a Delta V equal to the Barrier Equivalent Velocity in the 30 mph full-barrier crash test was created with an acceleration pulse as close to the crash as possible.

All the unrestrained results on the passenger-side show increasing injury measures as the air bags are depowered. Essentially, energy taken out of the air bag reduces the energy absorbing capacity to cushion the occupant during impact. Generally, the chest g's approach the FMVSS 208 limit of 60 g's with the least amount of depowering.

The percent chance of a fatality from HIC using two different injury curves is shown in tables following the test results. A different method will be developed in Chapter IV to estimate the impact of the more significant changes in chest g's on fatalities.

Vehicle Crash Tests with Depowered Air Bags

The agency has conducted comparison vehicle crash tests (baseline air bags and depowered air bags with passenger car model I-96) with a 25 percent depowered air bag for the driver and a 23 percent depowered air bag for the right front passenger. These tests were performed in 30 and 35 mph barrier tests with an air bag alone (unbelted) , and at 35 mph with bags and belts to determine the effects of depowering at high speed conditions. Tables III-11 and III-12 show the results of these tests on HIC, chest g's and the maximum femur loads.

Percent Chance of Fatality to the Head (%)
Vehicle I-96	Baseline 30mph	Depowered-25%	Baseline 30mph	Depowered-23%
	Driver	Driver	Passenger	Passenger
Head
Prasad/Mertz	.0013	.0024	.0021	.0039
Lognormal	.02	.08	.05	.17
Vehicle I-96	Baseline 35mph	Depowered-25%	Baseline 35mph	Depowered-23%
	Driver	Driver	Passenger	Passenger
Head
Prasad/Mertz	.0056	.025	.0042	.0157
Lognormal	.35	2.2	.20	1.35

Table III-12 Vehicle to Barrier Tests with Baseline and Depowered Air Bags (Belted)
Vehicle I-96	Baseline 35mph	Depowered-25%	Baseline 35mph	Depowered-23%
	Driver	Driver	Passenger	Passenger
HIC	814	857	788	652
Chest g's	52	59.6	52	49.6
Max. Femur	6773	5807	4643	5192

Percent Chance of Fatality to the Head (%)
Vehicle I-96	Baseline 35mph	Depowered-25%	Baseline 35mph	Depowered-23%
	Driver	Driver	Passenger	Passenger
Head
Prasad/Mertz	.0376	.0485	.0322	.0142
Lognormal	3.2	3.88	2.81	1.2

Only one manufacturer provided information comparing test scores from standard and depowered air bags (Honda in Docket No. 74-14 N108-156). Relative information was provided as shown in Table III-13 for air bags with unbelted occupants. NHTSA's interpretation of this data are that with depowering the driver side by 36.8 percent (475-300)/475, HIC increased significantly (by a factor of 2.52) and chest g's and deflection increase by 15 and 13 percent. By depowering the passenger side by only 15 percent (500-425)/500, HIC again increased significantly (by a factor of 2.78) but both chest g's and chest deflection also increased significantly (by a factor of 1.66 and 1.61). All of these results show a higher level of impact from depowering than NHTSA has seen in any of its testing or modelling work.

Table III-13 Honda Test Data
30 mph air bag without belt	Driver Air Bag, Standard	Driver Air Bag, Depowered	Passenger Side, Standard	Passenger Side, Depowered
HIC (ratio)	1	2.52	1	2.78
Chest G (ratio)	1	1.15	1	1.66
Chest Deflection (ratio)	1	1.13	1	1.61
Inflator Max. Pressure (KPA)	475	300	500	425
Inflator Onset (KPA/ms)	18.7	11.3	13	11