SLIDE 1:
INJURY CRITERIA FOR SIDE IMPACT - A RESEARCH UPDATE

Rolf Eppinger
NHTSA


SLIDE 2:
HEAD INJURY CRITERIA
[simple graphic rendering of pendulum test configuration]

- lateral head impact test data:
 McIntosh, et al (1996) IRCOBI and ESV papers
- 16 tests - skull fracture and brain injury reported
- McIntosh found Ax200 (resultant head accel. filtered to 200Hz 3dB point) was a better predictor of injury than HIC.


SLIDE 3:
REGRESSION ANALYSIS OF McINTOSH HEAD IMPACT DATA
[table]


SLIDE 4:
[graph - prob of head injury vs Ax200]
threshold@ 25%p AIS>=3: Ax200~102 G's


SLIDE 5:
[graph] Prob of head inj. vs. HIC 1000
P-value=0.06
Gamma=0.71


SLIDE 6:
THORACIC INJURY CRITERIA	
- Heidelberg type sled apparatus with left side impact
- 42 sled tests: 15 tests at OSU, 27 at MCW - of this data, 37 tests were usable.
- unembalmed fresh and frozen cadaveric human subjects: instrumented with accelerometers and chest bands


SLIDE 7:
SIDE IMPACT SLED TESTS
[simple graphic rendering of sled test configuration]

SLIDE 8:
[graph]
RESULTS USING FULL THORAX DEFLECTIONS
Age, maximum full thorax deflections, and resultant upper spine accelerations correlated best with injury outcome.


SLIDE 9:
[table]
NEW ANALYSIS INCLUDING HALF THORAX DEFLECTIONS
Model/-2logLR


SLIDE 10:
RESULTS
- full thorax maximum deflections are better predictors of thoracic injury than half thorax deflections

- the best injury criteria using half thorax deflections are:

  * linear comb. of age, upper spine ax, max. defl.
  * linear comb. of age, upper spine ax*, max. defl.

- models normalized for age=45 years


SLIDE 11:
INJURY CRITERIA

PRODUCT MODEL
AIS 3+: Logit L=-5.5267+0.15338*Acc*Dmax
AIS 4+: Logit L=-8.077+0.15338*Acc*Dmax

LINEAR MODEL
AIS 3+: L=-8.4691+7.6289*Dmax+0.0+46*Acc
AIS 4+: L=-10.9735+7.6289*Dmax+0.0+46*Acc

DEFLECTION ALONE
AIS 3+: L=-2.787+0.04146*Dmax
AIS 4+: L=-4.6697+0.04146*Dmax

 
SLIDE 12:
[scatter chart]
RESULT. SPINE ACCEL VERSUS CHEST DEFLEC.


SLIDE 13:
[graph]
PROB. OF INJURY VS. SPINE ACCEL (PRODUCT MODEL)


SLIDE 14:
[graph]
PROB. OF INJURY VS. SPINE ACCEL (lin. comb. model)


SLIDE 15:
[graph]
PROBABILITY OF AIS 3+ THORACIC INJURY VERSUS LINEAR COMBINATION OF HALF THORAX DEFLECTION AND RESULTANT SPINAL ACCELERATION


SLIDE 16:
ABDOMINAL INJURY CRITERIA
Human cadaver drop tests - Walfisch et al (France): 5th IRCOBI Conference
  - 11 cadaver drop tests onto rigid and padded armrest for a height of 1 and 2 meters
  - only 8 tests usabgle
  - ES-2 design based on these tests
  - found threshold limit = 2500N abdominal force


SLIDE 17:
[graph]
REANALYSIS OF WALFISCH DATA
probability of abdominal injury (AIS 3+)
vs abdominal force

threshold @25%p AIS>=3: abdominal forc ~ 2300Newtons


SLIDE 18:
PELVIS INJURY
Haffner et al (ESV 1988) reported pelvis accelerations of cadavers exposed to lateral impacts.
 - used to develop current FMVSS 214 pelvis acceleration tolerance limit
 - found that injury was dependent upon which structures in the bony pelvis were engaged by the vehicle interior


SLIDE 19:
PELVIS INJURY	
All current side impact dummies measure pelvis acceleration
 - presuming a biofidelic dummy acceleration response, injury threshold can be determined by relating cadaver accelerations with cadaver injury.

Most side impact dummies offer internal force measurements
 - tolerance level can be determined by performing impacts with the dummy in the same conditions as the cadaver, and relating dummy forces to cadaver injuries

NHTSA is evaluating side impact data from recent sled tests, as well as data from the literature, to develop updated injury criteria.


SLIDE 20:
CONCLUSIONS
- maximum resultant head acceleration filtered at 200Hz predicted injury better than HIC in lateral pendulum impacts

- a combination of age, maximum normalized half-chest deflection, and maximum resultant upper spine acceleration predicted thoracic injury (mostly fractures) better than age combined with chest deflection, upper spine acceleration, ASA or V*C.

- the maximum abdominal force threshold for 25% probability of abdominal injury is 2300 newtons


SLIDE 21:
INJURY CRITERIA SUMMARY

Sufficient new cadaver data available to develop both injury relationships and performance limits for all major body areas compatible with measurement cababilities of all three dummies


SLIDE 22:
THINGS TO DO
* Head Injury:  analysis of recent NHTSA sponsored lateral head drop cadaver tests

* Thoracic Injury: 
  - Expand thoracic injury criteria analysis to all NHTSA cadaver test data
  - reanalysis of other data sources

* Abdominal Injury: reanalysis of deflection-based abdominal injury studies available in the literature for application to dummies other than the ES-2

* Pelvis Injury:  review of data supporting pelvic criteria in FMVSS 214 as well as other sources (Cesari et al)