Mass Change, Complexity and Fleet
Impact Response
Guy Nusholtz
Chrysler Group LLC
February 25, 2011
Overview
• Review of historical mass/fatalities studies
• Elaborating on complexity of mass reduction
• Description of fleet model
• Results of fleet model
2
Mass Effect History
• Evans: R1/R2 = (M2/M1)3.58
• Kahane 2003: mass was the dominant
parameter, stiffness plays significantly less of
a role and geometry has no statistically
significant contribution.
• Padmanaban: Mass is the most significant
vehicle parameter determining relative fatality
outcome.
3
Effect of Mass Ratio on Relative Fatality Risk
Occupants in lighter
vehicles are at
greater risk
R : driver fatalities in lighter car / driver fatalities in heavier car
μ : mass of the heavier car / mass of the lighter car
L. Evans, February 2000
4
Effect of Mass on Velocity Change
m1
v1
m2
( v1 + v2 )
∆v1 =
m1 + m2
v2
m2
m1
∆v2 =
( v1 + v2 )
m1 + m2
For m1 = 2500 lb, m2 = 3500 lb, and v1 + v2 = 50 mph,
∆v1 = 29.2 mph,
∆v2 = 20.8 mph.
5
Effect of Velocity Change on Fatality
Percentage of fatalities
20
18
16
14
r2: heavier car
Delta V2=20.8
12
10
8
6
r1: lighter car
Delta V1 =29.2
4
2
0
15
20
Source: Evans 2000
25
30
35
40
45
50
55
Delta V (mph)
Ratio of fatality risk ~ 4%/1.5%=2.7
Occupants in the lighter vehicle are at
2.7 times greater risk
6
Relative Contribution of Variables
to Odds of Fatality Car-to-Car, Frontal Crashes
Percent of Total Contribution
50
Mass is the dominant
vehicle factor in
car-to-car crashes
40
30
20
10
0
Log (mass
Struck
ratio)
Driver Age
Source: Padmanaban 2003
Striking
Vehicle
FAW
Struck
Struck
Airbag Vehicle Age
Deployed
Struck
Driver
Drinking
Striking Log
Stiff. (Ke2)
x Bumper
Ht. 7
Relative
Contribution
of Variables
Relative
contribution
of varies
to odds of
Relative
Contribution
of variables
to
to Odds
of
Fatality
Car-to-Truck,
Frontal
Crashes
fatality (front-front, car-truck or truck-car)
odds of fatality (front-to- front, car-car)
25.00
Percent of total
Percent of total
contribution
35
20.00
30
contribution
15.00
Mass is the dominant
vehicle factor in
car-to-truck crashes
25
10.00
20
Struck
driv er age
Vehicle
ty pe
Height ratio
Gender
Wheelbase
Belt use
Front overhang
Mass ratio
ruck driver airbag
0
Mass ratio
5
struck driver age
100.00
Belt use
155.00
Striking
driv er
Gender
Fleet Model
• Analyze and understand the general effects of
a set of different vehicle and behavioral
attributes on overall crash outcome.
• Attributes considered are: mass, stiffness,
crush, intrusion, pulse shape, seatbelt usage,
velocity distribution, etc.
9
Original 2003 Model
• Data-based fleet model
– Original version:
• Investigate the effect of mass and size
• Fleet samples created based on 22 NCAP
tests.
• Pulse: Two-step, Linear and Linear/Plastic
• Average acceleration is used to estimate the
fatality risk.
Source: Nusholtz SAE 2003-01-0905
10
Updated Current Model
• Data-based fleet model
• Introduced a significant number of new variables
(intrusion, belt use, airbag, driver behavior, etc.)
• Fleet samples created based on over 300 NCAP tests
(1993~2010).
• Non-NCAP response included: Car to Car, offset, lower
velocity impacts, etc.
• Pulse: Linear, Piece-wise Linear, Non-linear and NCAP
• Average acceleration is used to estimate the fatality risk
(no change).
11
Examples of Data Input to Model
Data source: NCAP (35 MPH)
12
Data Input: Initial Velocity Spectrum
α = 5.37
β = 2.38
µ = 12.76
σ = 5.51
13
Data Input: Intrusion
8.3
Data source: Nusholtz SAE 2006-01-1134
14
Model Validation: Intrusion
15
Model Validation: Fatality Risk
16
Model Assumptions (1 of 3)
• 70% belted.
• No behavioral changes.
• Front Impacts: car to car, truck to truck and
truck to car.
• Risk is monotonically increasing with
velocity change (all other conditions fixed).
• Risk is a function of velocity change and
the average rate of velocity change (all
other conditions fixed).
17
Model Assumptions (3 of 3)
• Fleet turn over at a constant rate with
complete turn over in 20 years.
• The national and state accident data bases
are an accurate representation of the real
world.
• Scaling laws apply during down massing
and stiffening and adding crush space.
19
Effect of Mean Mass and Mean Crush Change
Mass change
dominates fatality
risk compared to
associated vehicle
crush change
20
Complexity Effect of Belted vs. Unbelted
Belt usage is
consequential
21
Fixed 200 lb. Mass Reduction
-3 ( no other changes)
9.2
x 10
Fleet mean risk
9
8.8
8.6
8.4
Fatality risk
increases as reduced
mass vehicles enter
the fleet
8.2
8
7.8
0
5
10
Year
15
20
22
Scaled Stiffness, Mass, and Crush
(intrusion held constant)
-3
9.2
x 10
Fleet mean risk
9
Fixed mass reduction,
no other changes
8.8
Fatality risk
reduced but not
eliminated
stiffness, crush and mass scaling
8.6
8.4
8.2
8
7.8
0
5
10
Year
15
20
23
Conclusions
• The following conclusions are based on the
assumptions used to construct the fleet
model.
24
Conclusion for
Fixed 200 lb. Mass Reduction
• A constant 200lb. mass reduction across the fleet
with no other changes (average stiffness, crush,
vehicle size, functional aspects and impact force
deflection characteristics stay the same) results in
increased risk of fatality.
25
Conclusion for
Scaled Mass and Stiffness Reductions
• A 3/2 power law scaled mass reduction (heavier vehicles
have a greater amount of mass removed across the fleet)
• Scaled reductions based on known impact response
• An average stiffness reduction, proportional to the mass
with a comparative force deflection modification
• Crush increase obtained from downsizing components as
a result of mass reduction while holding vehicle size and
intrusion constant
• Risk of fatality slowed compared to constant mass
reduction, but fuel economy improvement is not as
significant and fatality risk still increases as reduced
26
mass vehicles enter the fleet.
Thank you!
27