Impact Assessment

1
PEDESTRIAN KINEMATICS AND INJURIES IN COLLISIONS
WITH VEHICLES
Dr Ciaran Simms
2

• PEDESTRIANS AND CYCLISTS VULNERABLE NO
  PROTECTION
• IRELAND 2004 ROAD COLLISION FACTS IRELAND 2004
• 21 OF FATALIIES (70 DEATHS)
• 17 OF INJURIES

3
ROAD COLLISION FACTS IRELAND 2004
4
PEDESTRIAN VEHICLE IMPACTS

• AT COLLISION SPEEDS LESS THAN 20KM/H, PEDESTRIANS
  USUALLY SUSTAIN ONLY MINOR INJURIES.
• BUT AT SPEEDS ABOVE 45KM/H, COLLISIONS WITH
  PEDESTRIANS ARE MOSTLY FATAL WOOD, 1990 OTTE,
  1999.
• THE REASON FOR THE DOMINANCE OF SPEED IS THAT THE
  COLLISION ENERGY INCREASES WITH THE SQUARE OF THE
  IMPACT SPEED

5
PEDESTRIAN FATALITIES VERSUS IMPACT SPEED
6

AUSTRALIAN STUDY EFFECTOF REDUCING IMPACT SPEED
URBAN SPEED LIMITS REDUCING SPEED LIMITS TO
50 KM/H FROM 60KM/H DEATH RATE OF ADULT
PEDESTRIANS 30 LOWER
7
ANALYSIS OF PEDESTRIAN KINEMATICS

• YIELDS INSIGHTS
• IMPACT LOCATIONS ON VEHICLE
• IMPACT LOCATIONS ON THE BODY
• VEHICLE SPEED FOR CRASH RECONSTRUCTION

• CLASSIFICATION
• 1. WRAP PROJECTION
• 2. FENDER VAULT
• 3. FORWARD PROJECTION
• 4. ROOF VAULT

8
WRAP PROJECTION
FENDER VAULT
9
FORWARD PROJECTION
ROOF VAULT
10
PEDESTRIAN IMPACT SIMULATION WRAP PROJECTION
11
PEDESTRIAN INJURIES

• LOWER LEG AND HEAD
• KNEE
• PELVIS
• CAUSED BY
• BUMPER AND BONNET WINDSCREEN STIFFNESS
• IMPACT WITH THE GROUND

12
PEDESTRIAN IMPACT SIMULATION WRAP PROJECTION
13
PEDESTRIAN INJURIES ARE COMPLEX MANY INJURIES
CAN OCCUR IN ONE COLLISION
OTTE, IRCOBI, 2005
14
INTRODUCTION OF REGULATORY TESTS
http//www1.tpgi.com.au/users/mpaine/ped_veh.html
15
IMPROVEMENTS IN KNEE PROTECTION THROUGH VEHICLE
DESIGN
OTTE, IRCOBI, 2005
16
EFFECT OF DESIGN IMPROVEMENTSON KNEE INJURIES
OTTE, IRCOBI, 2005
17
ACCIDENT RECONSTRUCTION VEHICLE SPEED
ESTIMATION FROM THROW DISTANCE

• ACCIDENT INVESTIGATION TECHNIQUES
• CORRELATION OF INJURIES WITH IMPACT SPEED
• ABS REDUCES TYRE SKID
• LEGAL CASES

18
SEARLES PARTICLE MODEL IMECHE, 1993
SLIDE/ROLL/BOUNCE DISTANCE NOT TRIVIAL OFTEN gt
FLIGHT
19
SEARLES PARTICLE MODEL

• µ - coefficient of retardation
• M- mass of pedestrian
• S - distance to rest
• H - height drop of cg to rest
• U, v horiz vert comp launch velocity
• Launch angle generally unknown

HOW TO RELATE LAUNCH VELOCITY TO VEHICLE SPEED?
20
UNCERTAINTY IS MAJOR FACTOR

• WHY NOT SAY HERES THE MASS, HERES THE FRICTION
  ETC,- HENCE THE A SPEED - BUT THAT DOESNT WORK
  IN PRACTICE
• MONTE CARLO SIMULATION
• REQUIRE PARAMETERS DISTRIBUTION
• NORMAL DISTRIBUTIONS FOR MC, MP, µ

21
FORWARD PROJECTION
22
DISTANCE CARRIED IN INITIAL IMPACT
PROJECTION VELOCITY
COLLISION VELOCITY
MOMENTUM AND RESTITUTION
23
FALLING OVER DISTANCE
EQUATIONS OF MOTION FOR PLANAR SYSTEM
3 DOF
CONSTRAINTFEET IN CONTACT WITH THE GROUND
NUMERICAL INTEGRATION OF THIS SYSTEM TO YIELD
DISTANCE TRAVELLED IN FALLING OVER
24
DISTANCE TRAVELLED IN SLIDE/ROLL/BOUNCE TO REST
SPEED LOSS DUE TO IMPACT WHEN CG HITS THE GROUND
EQUATION OF UNIFORM DECELERATION
25
FORWARD PROJECTION MODEL RESULTS
WOOD, SIMMS AND WALSH, IMECHE 2004
26
WRAP PROJECTION SIMILAR BASIS FOLLOWED

• TOTAL THROW DISTANCE IS COMBINATION OF INITIAL
  CONTACT, FLIGHT AND SLIDE TO REST PHASES
• MOVEMENT MORE COMPLEX DUE TO ROTATION ONTO THE
  BONNET AND SECONDARY IMPACT TO THE HEAD
• ONLY PRESENT RESULTS HERE

27
WRAP PROJECTION MODEL RESULTS
WOOD, SIMMS AND WALSH, IMECHE 2004
28
MEAN AND VARIABILITY
29
CONFIDENCE LIMIT CRITERIA
The 50ile or probable range of value in
injury and in civil law The 95ile range -
application in general civil law and in depth
research The 99.8ile range - the Overall
confidence limits, corresponds to beyond
reasonable doubt required for criminal
law cases.
30
TABLES FOR RECONSTRUCTION
SIMMS, WOOD AND WALSH, IJCRASH 2004
31
PEDESTRIAN IMPACT THE EFFECT OF PEDESTRIAN
MOTION ON HEAD CONTACT FORCES WITH VEHICLE AND
GROUND
SIMMS AND WOOD PRESENTED SEPTEMBER 2005 AT
INTERNATIONAL RESEARCH COUNCIL ON BIOMECHANICS
OF IMPACT CONFERENCE PRAGUE SIMMS AND WOOD,
IJCRASH 2006
32
BACKGROUND

• 1980S REAL WORLD PEDESTRIAN ACCIDENTS
• HIGH IMPACT SPEEDS, HEAD INJURIES FROM VEHICLE
  RATHER THAN ROAD IMPACT.
• AT LOW SPEED THE GROUND IMPACT INCREASES IN
  IMPORTANCE
• (LESTRELIN ET AL, 1985).
• BELOW 7 M/S IMPACT SPEED, INJURY FROM THE GROUND
  IMPACT WAS HIGHER THAN FROM VEHICLE IMPACT, BUT
  THIS REVERSED AT HIGHER SPEEDS
• (ASHTON, 1982 ASTON AND MACKAY, 1983).
• SINCE THEN, CONSIDERABLE DEVELOPMENT OF CAR
  FRONTS.
• HAS THE ROLE OF THE GROUND CONTACT CHANGED IN
  IMPORTANCE?

33
METHODS MADYMO MULTIBODY PEDESTRIAN MODEL
SIMPLIFIED VEHICLE GEOMETRY
SIMPLIFIED CONTACT FUNCTIONS
COLEY ET AL, 2001
34
MODEL CONFIGURATIONS
1
2
3

• 1. FACING SIDE
• 2. 45 DEGREES
• 3. FACING VEHICLE
• 4. SIDE LEFT LEG BACK
• 5. SIDE RIGHT LEG BACK

5
4
VEHICLE IMPACT SPEED 5, 10 20m/s
35
FACING SIDEWAYS HIGHER EFFECTIVE RADIUS OF
ROTATION ABOUT THE BONNET LEADING EDGE YIELDS
SLOWER ROTATION THAN FRONT BACK CASE
36
HEAD VEHICLE IMPACT FORCE
FACING VEHICLE YIELDS MORE SEVERE HEAD IMPACT
DUE TO BODY GEMOETRY
37
HEAD GROUND IMPACT FORCE

• ALMOST RANDOM VARIATIONS IN TIMING AND MAGNITUDE

38
VEHICLE GROUND CONTACT COMPARISONFORCE AND
VELOCITY CHANGE (dV)
39
EFFECT OF PEDESTRIAN TRANSVERSE VELOCITY
AT LOW VEHICLE SPEEDS, PEDESTRIAN TRANSVERSE
VELOCITY CAN PREVENT HEAD CONTACT WITH VEHICLE
ALTOGETHER
40
CONCLUSIONS

• INITIAL PEDESTRIAN STANCE AND SPEED HAVE A
  SIGNIFICANT EFFECT ON THE VEHICLE/HEAD CONTACT
  FORCE.
• FOR PEDESTRIAN/GROUND CONTACT, VERY LARGE
  RANDOM VARIATIONS IN CONTACT FORCE OCCUR AS A
  RESULT OF DIFFERENT BODY PARTS ABSORBING THE
  GROUND IMPACT.
• HEAD CONTACT WITH THE GROUND RESULTS IN HIGHER
  FORCES ACTING OVER A SHORTER DURATION THAN THE
  VEHICLE HEAD CONTACT FORCE.
• CONTAINMENT OF THE PEDESTRIAN ON THE VEHICLE TO
  PREVENT GROUND IMPACTS CAN YIELD OPTIMUM RESULTS
  AT LOW SPEEDS. ENCOURAGING, AS LOW SPEED
  CONTAINMENT IS A MORE REALISTIC PROSPECT THAN AT
  HIGH SPEED

41
THE INCREASED INJURY RISK TO PEDESTRIANS FROM
SUVS COMPARED TO CARS
SIMMS AND WOOD PRESENTED AT WORLD CONGRESS OF
BIOMECHANICS, MUNICH 2006
42
Context of research

• In Europe, SUVs represent 15 of new vehicle
  registrations
• PriceWaterhouseCoopers, 2004
• SUVs have different mass and shape than passenger
  cars.
• What is the effect of differences between cars
  and SUVs on injury patterns of struck
  pedestrians?
• IMPROVED DESIGN OF VEHICLE FRONTS

43
Review of empirical evidence of SUV risks

• Ballesteros et al real accident data 1995-1999
  AAP, 2004
• lt50km/h odds ratio for pedestrian risk from SUVs
  compared to cars
• 2.0 for traumatic brain injury
• 2.0 for thoracic injury
• 2.5 for abdominal injuries.
• However, only 4.5 of cases actually involved an
  SUV, compared to 66 of cases involving cars.

44
Review of empirical evidence of SUV risks

• Lefler Gabler AAP, 2004 real world data from
  the US
• 11.5 of pedestrians struck by large SUVs killed
• 4.5 for pedestrians struck by cars killed
• Roudsari et al IP, 2004
• Light truck type vehicles (LTVs) threefold
  higher risk of severe injuries to pedestrians
  than cars.
• Effect most pronounced at lower speeds

45
Roudsari et al TIP, 2005 PCDS 3146 injuries
in 386 pedestrians

• - No difference in impact speed between LTVs and
  cars.
• - 159 adults with head injuries, of which 46
  struck by LTVs

46
Vehicle factors for pedestrian riskMass,
Geometry Stiffness

• Lefler Gabler, AAP 2004 Pedestrians mass ltlt
  vehicle mass
• suggest frontal geometry is controlling factor
  but no elaboration
• Ballesteros et al AAP, 2004 Higher bumper
  bonnet heights in SUVs dictate initial contact
  points.. but no comment on momentum transfer
• Roudsari et al TIP, 2005 trajectory governed
  by pedestrian cg and bonnet leading edge height
  but no comment on effect of direct impact against
  the pelvic/abdomen region.
• Stiffness important but no data

47
Summary of empirical evidence of SUV risks

• Empirical studies show substantially increased
  risk for pedestrians when struck by high fronted
  vehicles compared to a passenger car.
• However, conflicting evidence on relative risk of
  head injuries from these different vehicle types,
  and no agreement on the source of the increased
  risk of LTVs for pedestrians.

Current work aims to answer these questions
48
Methods Madymo pedestrian and vehicle models to
simulate dummy impacts of OKAMOTO ET AL, 2001

• VALIDATION POLAR DUMMY IMPACT AT 40KM/H WITH CAR
  SUV
• UPPER LEG REACTION TORQUES
• HIGH SPEED VIDEO

SUV
car
MIZUNO KAJZER, 2000 LIU ET AL 2002
49
Validation car impactOKAMOTO ET AL, 2001
0ms 20ms 40ms
60ms 80ms
50
Validation Suv ImpactOKAMOTO ET AL, 2001
0ms 20ms 40ms
60ms 80ms
51
Simulation matrix

• pedestrian facing car
• pedestrian facing SUV
• pedestrian sideways to car (walking stance,
  struck leg back)
• pedestrian sideways to SUV (walking stance,
  struck leg back
• 5, 10 and 15m/s impacts, braking

52
10m/s snapshots
53
Results side struck pedestrian head
resultant acceleration
54
HEAD IMPACT ON WINDSCREEN
STIFFNESS VARIES DEPENDING ON IMPACT LOCATION
55
Results Pelvis resultant acceleration
56
Head injury predictions HIC
57
Head injury predictions HIC
Head Injury Predictions using HIC criterion
Injury reference level 1000
58
Pelvis injury predictions acceleration
Pelvis Injury Predictions using peak acceleration
criterion (m/s2) Injury reference level 716m/s2
59
THE EFFECT OF SPEED
MAJORITY OF PEDESTRIAN ACCIDENTS BELOW SPEEDS OF
50KM/H ASHTON AND MACKAY, 1983
VELOCITY CHANGE
ACCELERATION
CRUSH
INCREASED RISK OF SUVS MOST IMPORTANT AT MEDIUM
/ LOW SPEED,
60
Effects of vehicle mass
Primary impact with vehicle Wood IMechE, 1988
k pedestrian radius gyration h vertical
offset between bonnet leading edge height and
pedestrian cg
61
Effects of vehicle mass
head
pelvis
62
Conclusions on the effect of vehicle front shape
on pedestrian injuries

• Head injuries similar or slightly lower from
  contact with SUVs compared to cars
• Injuries to mid body regions are substantially
  higher.
• Primary reason for increased hazard to
  pedestrians from SUVs is the high front shape of
  the bumper and bonnet.
• Location of primary impact means mid body region
  is directly struck in a SUV/pedestrian collision,
  allowing less rotation of the body.

63
Conclusions on the effect of vehicle front shape
on pedestrian injuries

• For pedestrians struck by SUVs there is the
  combination of a harder primary impact which
  occurs directly with the critical mid body
  region.
• The mass difference between cars and SUVs not
  significant for pedestrians
• Lowering the bumper and bonnet and reducing
  bonnet stiffness for SUVs would help to reduce
  injuries to these mid body regions.

Simms and ONeill, British Medical Journal,
2005 Simms Wood, IMechE 2006
64
SIZE DOES MATTER
65
Thank you