OFFSET CRASH TESTS
OBSERVATIONS ABOUT VEHICLE DESIGN AND STRUCTURAL PERFORMANCE
Michael Paine, Vehicle
Design and Research Pty Limited email: email@example.com
Donal McGrane Crashlab, NSW Roads and Traffic Authority
Jack Haley NRMA
Sixteenth International Technical Conference of the Enhanced Safety
of Vehicles (ESV)
Windsor, Canada, June 1998
Paper Number 98-S1-W-21
(PDF version) Published by NHTSA,
Absorption of crash energy
Integrity of passenger compartment
Offset frontal crashes can place severe demands on the structure of vehicles.
Offset crash tests conducted in the USA, Europe and Australia are revealing
that some vehicle models perform exceptionally well in these severe tests.
Between them, the authors have been involved in the assessment of more
than 45 offset crash tests conducted under the Australian New Car Assessment
Program (ANCAP). They
have also evaluated data on a similar number of offset crash tests conducted
in the USA and Europe.
This paper sets out some general observations about the structural performance
of cars, passenger vans and four-wheel-drive vehicles in offset crash tests.
The design features which appear to contribute to good structural performance
are discussed. Likely reasons for poor performance are noted.
This paper sets out some general observations about the structural performance
of cars, passenger vans and four-wheel-drives in the offset crash test
conducted by Australian NCAP (New Car Assessment Program), the US Insurance
Institute for Highway Safety (IIHS) and Euro-NCAP.
The observations are intended to be constructive and should assist vehicle
designers improve the crashworthiness of vehicles.
The Offset Crash Test
In the offset crash test the vehicle is travelling at 64km/h when it collides
with a crushable aluminium barrier. The barrier initially makes contact
with 40% of the width of the front of the vehicle, on the driver's side
(Lowne 1996). The resulting crash forces place severe demands on the structure
of the vehicle, particularly on the driver's side.
The vehicle structure affects the outcome of an offset frontal crash
in two main ways:
Absorption and dissipation of crash energy
Integrity of the passenger compartment
Figure 1. Overhead view of an offset test into a deformable barrier
at 64km/h (ANCAP).
ABSORPTION OF CRASH ENERGY
The offset crash test is intended to simulate a collision between two similar-sized
vehicles with similar crush characteristics. In these types of crashes
it is desirable that most of the crash energy is absorbed and dissipated
in the deformation of components within the front metre or so of each vehicle.
The increasing use of engine/suspension cradles has allowed designers
to better control this deformation and to by-pass very rigid components
such as engine blocks which are not effective energy absorbers.
To avoid load concentrations it is important that the crash forces are
spread across the face of the deformable barrier. In several cases it has
been observed that box-section structures at the front of the vehicle have
punched through the barrier and relatively little energy is absorbed through
deformation of the barrier. These box section structures appear to be designed
to achieve better performance during a full-frontal crash into a rigid
barrier but they can be much less effective during offset crashes into
deformable objects, including other motor vehicles. Conversely, some box
sections which crush efficiently in a full-frontal crash do not perform
as well under the asymmetric loads of an offset crash - they tend to buckle
rather than concertina (also a problem with bull-bars).
Some four-wheel-drive recreational vehicles have relatively stiff front
structures. This can result in a very high deceleration of the passenger
compartment and high loads on the occupants. A stiff front structure can
also place excessive demands on the deformable barrier, causing the barrier
to bottom out early in the crash sequence. In a collision between two vehicles
the occupants of the heavier vehicle would generally be better off, due
to the physics of the collision. In the case of four-wheel-drive vehicles
colliding with passenger cars, however, this advantage can be diminished
by a stiff front structure. Analysis of crashes in Australia have shown
that, on average, the driver of a four-wheel-drive vehicle has a greater
likelihood of being killed or seriously injured in a crash than the driver
of a large car (Newstead et al , 1997).
The front structure of a vehicle also has a strong influence on aggressivity
- the degree to which individual vehicle models cause injuries to occupants
of other vehicles. The Monash University Accident Research Centre (MUARC)
has recently conducted an analysis of aggressivity in more than 300,000
on-road crashes (Cameron et al 1998). Key results related to front structure
are set out below:
The cases where, in the offset crash test, the front structure imposed
concentrated loads on the deformable barrier could also be expected to
be hazardous to the occupants of other vehicles due to increased penetration
and intrusion. This might partly explain the adverse result for vans. Other
factors might be the very high proportion of these vehicles fitted with
"bull bars" in Australia (estimated at 50% or higher - Traffic and Transport
Surveys, 1994) and the higher laden mass, compared with cars (the analysis
was based on kerb mass).
There is the expected trend of increased aggressivity with increased vehicle
kerb mass but there is a large amount of scatter, with some high-mass vehicles
showing low (good) aggressivity.
Four-wheel-drives generally show high aggressivity but there is a large
amount of scatter and some commendable exceptions. Some of the lighter
four-wheel-drives are no more aggressive than cars of the same mass.
Passenger vans and commercial vans tend to have high aggressivity for their
One criticism of consumer offset crash tests is that they are claimed
to push manufacturers towards building stiffer vehicles, in order to protect
their own occupants, and that these stiffer vehicles are more aggressive
to other vehicles. The MUARC study, and results of recent offset crash
tests suggest that this criticism is unfounded. Some vehicles performed
exceptionally well at protecting their occupants in the offset test while
apparently having low aggressivity towards the occupants of other vehicles.
Evidently this was achieved by efficiently absorbing crash energy in the
front structure while retaining the integrity of the passenger compartment.
Occupants of other vehicles should be at less risk in collisions with
these low-aggressivity vehicles. Vehicle design is therefore a crucial
factor in achieving good crashworthiness and low aggressivity.
If the all-vehicle average aggressivity observed in the MUARC study
had been reduced to the average for small cars then it is estimated there
would have been a 30% reduction in fatal/serious injuries to the occupants
of "other" vehicles. Since aggressivity is not a quality easily affected
by consumer pressure this may be an area which requires legislation.
INTEGRITY OF THE PASSENGER COMPARTMENT
The passenger compartment should keep its shape in the crash test. The
steering column, dash, roof, roof pillars, pedals and floor panels should
not be pushed excessively inwards, where they are more likely to injure
the occupants. Doors should remain closed during the crash and should be
able to be opened after the crash to assist quick rescue.
There is a temptation, when observing a frontal crash test, to concentrate
on what is happening at the front of the vehicle, since this is where most
of the deformation is occurring. This might give the impression that the
front of the vehicle is being forced back into the passenger compartment.
While this is an important source of passenger compartment intrusion it
is only part of the story.
At the height of a frontal crash test the front of the vehicle has come
to a halt but the remainder of the vehicle is still undergoing a high deceleration
- typically around 40g (up to 60g with some four-wheel-drive vehicles).
Substantial compression forces are generated between the front and rear
of the vehicle at this time.
Consider a transverse vertical plane in line with the dash. The resulting
cross-section might include the a-pillars, side doors, door sills and floor.
About 50% of the vehicle's weight will usually be rearward of this plane.
The compression forces arising in these components due to a 40g deceleration
are therefore equivalent to about 20 times the weight of the vehicle. This
places a severe demand on the structure. Furthermore, there is usually
very little structural redundancy in the design and if any one of these
components has a major failure then catastrophic collapse can occur (and
has been observed).
For commercial vehicles and four-wheel-drives the proportion of mass
in the rear is usually greater than with passenger cars, particularly when
laden. This, combined with a stiffer front structure, can place a very
severe demand on the structure of the passenger compartment .
Figure 2. Unladen light commercial vehicle during an offset crash test
at 64km/h (ANCAP).
During the 56km/h full-frontal test of several four-wheel-drive vehicles
there was a noticeable sleeving effect, where the sides of the passenger
compartment tended to slide around the engine compartment, resulting in
substantial firewall and dash intrusion at the height of the crash (figure
3). The deformation usually appears to be elastic (noticeable in the Figure
1 animation) and therefore the post-crash residual movement of the firewall
and dash might not indicate a problem.
Figure 3. Overhead view of a full-frontal crash test into a solid barrier
at 56km/h (ANCAP).
Front suspension components such as lower control arm pivots are commonly
located on, or just ahead of, the footwell toepan area and above the floor
level. The suspension components which are attached to these points are
usually rigid in a longitudinal direction and therefore, during the crush
of the front of the vehicle, they tend push the mounting points rearwards
into the toepan. In addition to the risk of lower leg injury due to intrusion,
dynamic movement of the toe-pan can cause the legs to lift suddenly, sometimes
resulting in a violent head to knee strike.
The better designs tend to locate the suspension mounting points below
floor level. In this configuration longitudinal structural members mounted
under footwell area can effectively transmit the crash forces past the
footwell area and reduce intrusion - this has been observed on several
vehicles. However, in some cases such structures have insufficient strength
to cope with forces during a 64km/h offset crash test. One possible weakness
is the location of joints or welds at a bend or change in cross-section.
These observations also apply where engine and/or suspension components
are mounted on a cradle under the front of the vehicle. These cradles are
usually attached to the body just ahead of the toepan. Their height and
method of attachment can have a significant influence on footwell intrusion.
Road wheels and tyres are relatively rigid when compared with footwell
panels and they can contribute to intrusion into the footwell area or separation
of footwell panels. Failure of the seams between the floor and door sill
or the firewall and a-pillar can greatly reduce the strength of this region.
Another factor associated with floorpan deformation is the movement of
seats. In some cases seats have tilted forward and/or to the side by substantial
amounts when the floorpan or transmission tunnel deforms. This can have
an adverse effect on occupant kinematics. This is particularly a concern
where the seat belt buckle is mounted on the seat because it can result
in extra forward movement of the occupant - thereby defeating the advantage
of mounting the buckle on the seat (better adjustment of the seat belt
to suit the occupant).
With seat-mounted seat belt buckles the loads on seat components are
much higher. Seat runners have been observed to bend, seat mounting frames
have rocked forward and floor panels have deformed. All these problems
have contributed to excessive forward movement of the occupants.
Upward movement of the steering wheel
Several of the crash-tested passenger vans and utilities experienced a
large upward movement of the steering column during the crash test. In
some cases the whole dash appeared to rotate upwards taking the steering
column with it. In other cases the vehicle structure near the bottom end
of the steering column was pushed rearwards, causing the column to rotate
and move upwards. Where an airbag is fitted this movement can adversely
affect its performance (this is evident in several of the IIHS tests of
The upward motion of the steering column often coincides with the forward
and downwards motion of the occupant's head. Where no airbag is fitted
this can cause an increase in the severity of the head strike. In one case,
this effect contributed to a head deceleration of 250g - one of the highest
In the absence of an airbag, steering wheel hub design plays an important
role in reducing head injuries. Considerable research has been undertaken
into effective, energy-absorbing hub designs but these do not appear to
have been put into production. In some cases the hub cover has flown off
just before the head impact, exposing the head to metal components.
Sideways movement of the steering wheel
In several offset crash tests the driver dummy rolled off the outboard
side of the airbag. Although some outboard motion of the dummy can be expected
due to the non-symmetrical crash forces, in most of these cases it is likely
that the steering wheel had moved inboard, relative to the driver's seat.
One possible factor is that components in the engine bay push the steering
column to one side, causing it to pivot about its mounting points. Another
possible factor can be gauged from an overhead view at the height of the
crash (figures 1 & 4). In many cases there is a substantial angular
difference between the front part of the vehicle, containing the dash and
steering column, and the rear part containing the passenger compartment
a "jack-knife" effect. The steering wheel therefore moves inboard, relative
to the passenger compartment
Figure 4. Diagram showing possible relative movement of steering wheel
This effect usually results from excessive deformation in the region
of the a-pillar on the driver's side. Rupture of the join between the dash
and the a-pillar, and collapse of the door and door sill can also contribute
to this problem.
Seat belt upper anchorages
Adjustable upper seat belt anchorages improve the fit of the seat belt.
However, some designs deform under the severe loads of the crash test and
allow additional forward movement of the occupants.
With the trend towards b-pillars which curve substantially inwards between
the door sill and the roof there is another source of seat belt "slack".
The webbing follows the curved path between the retractor unit and the
D-ring and is usually held in place by the trim. During the crash the plastic
trim can give way due to tensile forces in the webbing, which then straightens
and feeds through the D-ring.
The 64km/h offset crash test places severe demands on the structure of
the vehicle. Some vehicles perform exceptionally well during this crash
test but many exhibit excessive structural collapse and other undesirable
characteristics. The better designs appear to have the following structural
It is evident that these issues are now being taken into account during
the early stages of vehicle design. Powerful Computer Aided Engineering
(CAE) packages are now able to determine vehicle structural deformation
and occupant kinematics during a variety of crash situations, including
a simulated offset crash test (Loo and Brandini 1998 - see also "The
Crash in the Machine", Scientific American, March 1999).
a front structure which absorbs crash energy through controlled deformation
and avoids load concentrations on the impacted object,
structural components which bridge the front footwell area so that compression
forces are transmitted directly between the front and rear of the vehicle,
resulting in minimal footwell deformation,
structural components which channel crash forces into the a-pillars, side
doors and door sills rather than into the firewall area,
the join between the top of the a-pillar and the roof is smooth and strong
so that upwards buckling of the roof is resisted,
side doors and door sills which offer resistance to longitudinal compression
forces (measures to improve side impact protection appear to have assisted
in this regard),
a steering column designed and mounted to minimise the amount of rearward
and upward movement at the height of the crash,
Recent research in Australia indicates that the better designs can provide
good protection for their occupants, while having low aggressivity towards
the occupants of other vehicles.
Most of the crash tests on which this paper is based were conducted by
Crashlab (NSW Roads and Traffic Authority) for Australian NCAP. The assistance
of the Australian NCAP Technical Committee is acknowledged.
Other tests were conducted by the US Insurance Institute for Highway
Safety and Euro-NCAP. The provision of test reports and other material
by these organisations is appreciated.
Road safety web links
Cameron M H, Newstead S V and Le C M 1998. The Development and Estimation
of Aggressivity Ratings for Australian Passenger Vehicles Based on Crashes
During 1987 to 1995, Monash University Accident Research Centre, research
report in preparation, March 1998.
Loo M and Brandini M 1998, Applying Computer Aided Engineering to Improve
Vehicle Safety, Proceedings of the Developments in Safer Motor Vehicles
Seminar, sponsored by SAE-Australasia, Staysafe, Federal Office of Road
Safety. NSW Parliament House, March 1998.
Lowne R W 1996 , The Validation of the EEVC Frontal Impact Test Procedure,
Proceedings of the 15th International Technical Conference on the Enhanced
Safety of Vehicles, Melbourne.
Newstead S, Cameron M & Le 1997, Vehicle Crashworthiness Ratings and
Crashworthiness by Year of Vehicle Manufacture: Victoria and NSW Crashes
During 1987-95, Monash University Accident Research Centre - Report #107.
Paine M 1997 Guidelines for Crashworthiness Rating System, Report prepared
for Australian NCAP Technical Committee, December 1997.
Shearlaw A & Thomas P, 1996, Vehicle to Vehicle Compatibility in Real-world
Accidents, Proceedings of the 15th International Technical Conference on
the Enhanced Safety of Vehicles, Melbourne.
Traffic and Transport Surveys, 1994. The proportion of bull-bars in the
vehicle fleet. Data report commissioned by NRMA, October 1994.
Zeidler F, Scheunert D, Breitner R & Krajewski R 1996 , The Reduction
of the Risk of Lower Leg Injuries by Means of Countermeasures Optimised
in Frontal Offset Crash Tests, Proceedings of the 15th International Technical
Conference on the Enhanced Safety of Vehicles, Melbourne.