(1) Conspicuity
of Pedestrians,
Graham Greatrix and Jason Smithies, IMPACT, 8, 2, 1999.
This paper reviews the principal physical factors which affect a
driver's ability to see a pedestrian.
(2) Multi-surface
braking,
Graham Greatrix, IMPACT, 11, 1, 2002
This paper outlines a method for calculating the speed of a vehicle when
its tyre marks pass over several different surfaces.
(3) Critical
speed at the brow of a hill,
Graham Greatrix, IMPACT, 8, 2, 1999
The vertical curvature of a road surface can affect the critical speed
quite dramatically. It is far more important than the effect of camber or
super-elevation. This paper considers the physics of critical speed when
vertical curvature is taken into account. The vertical curvature includes dips
in the road as well as hill brows.
(4) Wind
forces,
Graham Greatrix, IMPACT, 8, 3, 1999
This paper discusses the physics of wind effects on vehicles. Worked
examples are presented to show how it can be established if a particular wind
speed and direction can cause a vehicle to overturn rather than slide. This paper should be read in conjunction with the comments raised by David Hague in Impact, 9, 1, 2000.
(5) Critical
speed under braked conditions,
Graham Greatrix, IMPACT, 11, 2, 2002.
If a vehicle is being driven around a curved path whilst being braked, the general critical speed equation needs modifying
to take account of the degree of braking present. If that modification is
ignored, the calculated critical speed will be too high.
(6) The
time trap,
Graham Greatrix and Jason Smithies, IMPACT, 5, 3, 1996
This short paper highlights the invalidity of accident reconstructions
which consider time alone with no reference to position or distance. For
example, a pedestrian takes 2 seconds to step from a kerb and reach the
collision point. The overall stopping time for the car is 3 seconds. Thus, the
driver was unable to stop in time and the accident was inevitable. Although
this argument appears quite compelling, it is actually a fallacious argument.
In fact, it does not matter how many seconds it takes the driver to stop
as long as his car does not reach the path of the pedestrian.
(7) Plan
preparation through perspective analysis,
Graham Greatrix, IMPACT, 1, 2, 1990.
This paper deals with the assessment of distance from single
photographic images. The paper includes a worked example.
(8) An
analysis of existing data on adult walking speeds,
Graham Greatrix & Jason Smithies, ITAI, Proceedings of the 3rd
International Conference,
This paper uses modern statistical techniques to analysis data that has
been obtained in a variety of research investigations.
(9) Uncertainty
in Accident Investigation,
Graham Greatrix, IMPACT, 13, 2, 2004.
This paper considers the implementation of the directive that experts
should always express their results as a range of values.
(10) The
geometry of the cut-in of rigid and articulated vehicles,
Graham Greatrix, IMPACT, 14, 1, 2005.
Vehicle cut-in is a major cause of many pedestrian and cyclist
accidents. This paper develops a method from which the degree of cut-in of any
vehicle can be calculated.
(11) Whiplash
in low speed rear impact collisions,
Graham Greatrix, IMPACT, 14, 2, 2005.
This paper discusses the physics and engineering factors in collisions
where virtually no vehicle damage occurs. The interface between engineering
analysis and medical evidence is also considered. A copy of this paper can be
found at the end of this list.
(12) Low speed rear impacts and whiplash - the role of the engineer,
Graham Greatrix, Central
Law Training Conference,
This
presentation discusses the relevance of physical or engineering evidence in
these cases in the light of recent Court rulings.
(13) Non-Accident Investigation
Publications:
(a) "The modulus of
the Fourier Transform in image reconstruction", Third International
Conference on Image Processing, No.307, IEE,
(b) "The use of the
fourier transform in modelling surface topography", Eighth International
Conference on mathematical and computer modelling,
(c) "Fourier domain
compression techniques in resolving the spatial characteristics of
surfaces", Fourth
Insternational Conference on image processing, IEE,
(d) "Fourier domain
rescaling and truncation for improving the 3-D reconstruction of object
surfaces", Ninth International Conference on mathematical and computer
modelling, Sotoudeh, Greatrix and Goldspink, 1993.
(e) "Fourier domain
contrast enhancement of spectral coefficients to improve the 3-D reconstruction
of objects", Ninth International Conference on mathematical and computer
modelling,
(f) "Improving the
recovery of the micro-geometry of object surfaces by Fourier Scaling",
ICPAT 94, Dallas. Sotoudeh, Greatrix and Goldspink, 1994.
Whiplash
in low speed rear impact collisions
(Reprinted
from IMPACT, 14, 2, 2005)
Graham
Greatrix, Forensic Investigator,
INTRODUCTION
The
Association of British Insurers reported that during the period 2001 to 2002
there were some 280000 successful claims regarding whiplash and associated
disorders. The average cost of each claim was around £4500. Most whiplash
injury claims are also associated with low speed collisions often not exceeding
10 miles per hour.
When defending
such claims the insurance industry avails itself of the simple and persuasive
argument that if there is no vehicle damage, or only cosmetic damage, then it
must follow that the collision forces would have been insufficient to cause any
unusual movement of the car's occupants. In those circumstances, the claim for
whiplash injury must be fraudulent.
EXPERT TESTIMONY
Medical
experts are used to assess the injuries. Collision investigators are used to
assess the forces that would be transmitted during the collision. Medical
experts will not usually be in a position to associate the injury conclusively
with the collision. Soft tissue injuries are often not detectable by any
current technique other than at autopsy. The medical expert is generally forced
to rely on what the claimant says. Whiplash symptoms may start to be felt up to
three days after the accident.
The question
arises as to whether collision investigators can realistically provide any
quantitative assessment of the forces that were transmitted to the claimant's
neck. Even if that is possible, how does the force relate to injury severity?
The precise mechanisms of transfer are not well understood and there are far
too many variables and uncertainties. Every case is essentially unique in its
circumstances.
Both types of
expert rely on simulation tests that have been carried out using volunteers,
dummies, animals and even cadavers. The number of tests and interpretations are
legion. Just putting +whiplash +"low
speed" into an internet search engine will yield over 9000 articles.
In general,
the tests attempt to establish a threshold at which injuries will start to
occur. For example, West et al (1993) and Szabo et al (1994 and 1996) concluded
that with properly designed and adjusted head restraints, barrier rear impacts
in excess of 5 mph can be tolerated by reasonably healthy occupants without
injury. Other researchers have reached similar conclusions. The conclusions
suggest that there is a threshold delta-V of 5 mph below which injuries should
not occur. On the other hand Brault et al (1998) found that 29 per cent and 38
per cent of occupants exposed to rear impacts with a delta-V of 2.5 mph and 5.0
mph respectively experienced mild whiplash symptoms. Castro et al (1997) tested
17 volunteers at an average delta-V of 7 mph. 29 per cent of the subjects
reported whiplash type symptoms. One subject remained symptomatic for 7 days
and another had reduced movement range for 10 weeks. Clearly, some subjects in
these tests were injured. In spite of that, the authors concluded that the
"limit of harmlessness" for stresses arising from rear end impacts
with regard to velocity change lies between 6 mph and 9 mph.
Unfortunately,
there are interpretation difficulties with all the tests that are currently
cited. Firstly, the samples are unrepresentive of the general population. The
volunteers are usually male, young and healthy. They are seated in ideal
conditions, facing forwards with properly adjusted seats and head restraints.
They also know what is going to happen and when. Secondly, the sample size is
invariably very small, usually less than 10 down to only 3 subjects. No
statistical significance tests have been carried out to determine the relevance
of the sample results to the general population.
COLLISION ANALYSIS
When the two
cars involved in a collision are examined, it is often the case that the damage
sustained by the rear of the struck car is significant while the damage
sustained by the front of the striking car is non-existent. The defendants will
argue that this apparent inconsistency suggests that the struck car was damaged
in some other accident. They ignore the fact that different cars have different
strengths, that the rear of a car is generally weaker than the front of a car
and that vehicle age and history are parameters that need to be considered.
Collision
investigators will be asked to examine the vehicles and determine the forces
involved in the collision and also the delta-V for each vehicle. The thesis
here is that change in speed is a function of the extent of crush damage and
therefore the extent of injury. It should therefore be possible to determine
the extent of damage, no matter how trivial, and so calculate the speed change
that would result in that damage. The request is obviously impossible to
satisfy. The algorithms for determining speed change from crush damage are
based on data obtained from barrier collisions in the region of 30 mph. It is
dangerous to extrapolate that data to collisions that are well below 30 mph or
well above 30 mph. Additionally, all algorithms assume that there will be no
measurable damage below about 5 miles per hour into a solid barrier and
therefore about 10 mph into another car.
This leads to
the argument that if there is no damage then the impact speed must have been
below 10 mph. Unfortunately, there are very wide variations in the threshold
speed at which cars actually sustain damage.
The defendants
generally take no account of the effect of elasticity in low speed collisions.
At high impact speeds, collisions are mainly plastic with the coefficient of
elasticity approaching zero. At very low impact speeds, the collisions tend to
be elastic with the coefficient of elasticity approaching unity.
An elastic
collision is analogous to a billiard ball hitting another billiard ball. Both
balls change speed markedly on impact but neither ball is damaged.
In the extreme
case of a perfectly plastic collision between two similar cars, the change in
speed of the struck car will be half the impact speed of the striking car.
In the extreme
case of a perfectly elastic collision between two similar cars, the change in
speed of the struck car will equal the impact speed of the striking car. The
striking car will be brought to a sudden halt.
No damage
occurs in a perfectly elastic collision.
If damage
occurs, the collision is only partially elastic and so the change in speed of
the struck car is lower than would have been the case for a perfectly elastic
collision. Instead of the transferred energy being wholly kinetic, a proportion
of that transferred energy is dissipated in causing damage to both cars.
Consequently, a lower proportion of the transferred energy and of the
transferred forces will reach the car's occupants. This is the principle
underlying the concept of crush zones. By allowing the collision to cause
damage, the occupants are better protected from injury. Thus, no car damage
does not necessarily mean that there will be no occupant injury. Rather, the
opposite is true.
The severity
and frequency of whiplash injuries increase with the closing speed of the two
cars. Up to about 10 mph closing speed, the frequency and severity of injuries
incease rapidly. Above about 10 mph, the rate of increase becomes much lower
because an increasing proportion of the transferred energy is dissipated in
causing damage. The actual threshold will depend on the cars involved.
The forces
transmitted to the occupants will depend on the delta-V caused by the
collision. Delta-V is considered by virtually all test researches as being the
sole or the main parameter of importance. However, delta-V is simply a change
in speed. Force is proportional to the change in speed divided by the time,
delta-t, that is taken for that change to occur. Thus, delta-t is just as
important as delta-V. The delta-t for an elastic collision and therefore for a
low speed collision will clearly be shorter than the delta-t for a collision in
which time is spent damaging the cars.
The collision
forces for an impact which causes a delta-V of say 5 mph over a time of 0.10
seconds will be three times the collision forces for a collision that lasts
0.30 seconds. To say that a delta-V of 5 mph will not result in injury is quite
meaningless as a generalised proposal.
FORCES TRANSMITTED TO CAR
OCCUPANTS
Delta-t will
vary from collision to collision. Delta-t is also affected by the structure of
the bumper. So called energy absorbing bumpers do not absorb energy. They
elastically compress and then release the stored energy slowly either to the
struck vehicle or to the striking vehicle or both. This effectively increases
the collision time. Although the Delta-V remains the same, the resultant
acceleration is somewhat reduced by this process.
There will be
a limit to the elastic compression that a bumper can absorb. Above that limit,
the bumper response ceases to be elastic.
So far, I have
qualitatively considered only the vehicle dynamics during a collision. What really
matters here is the force that the collision would apply to the body of a car
occupant.
When a force
is applied to an object, it will be accelerated. The value of the force is
given by the product of the mass of the object that is free to move and the
resultant acceleration.
The seat belt
will restrain the Claimant's torso from moving forwards relative to his seat.
In a rear end impact to his car, his torso will be accelerated forwards at
virtually the same rate that his car is accelerated forwards by the impact
forces. His torso will also be pressed into the backrest of his seat.
Unfortunately,
his head will not be restrained and, as in the vast majority of cases, his head
restraints will most probably be too far away from the back of his head to be
effective.
As his torso
is accelerated forwards, his head will be left behind since there is nothing
available to push his head forwards along with his torso.
Additionally,
the seat back, initially loaded by the inertia of the occupant, releases that
stored energy shortly afterward as an elastic recoil of the seat back. This
results in a further forward impulse directed through the torso, by the seat
back, and which tends to magnify the potential for injury because it coincides
with the rearward motion of the head with respect to the torso.
His head would
therefore roll backwards extending the neck and possibly damaging the soft
tissues in the neck.
After
extending backwards, the head has to catch up with the torso and that causes
the head to reverse direction and accelerate forwards. Obviously, if the head
has to catch up with the accelerating torso, the forward acceleration of the
head has to be greater than that of his torso and therefore greater than the
acceleration of the car itself. The resultant forward acceleration rate of the
head can be more than 250% higher than the acceleration of the vehicle.
By the time
that the head has caught up with the torso, the collision is over and the car
will start to decelerate. Unfortunately, the head is then at its maximum
forward speed and will overshoot the torso. This necessitates a rapid
deceleration of the head followed by another rearwards acceleration.
A collision of
this kind will probably occupy less than 300 milliseconds. The average
acceleration required to increase the speed of an average sized vehicle to
nearly 10 miles per hour from rest in only 300 milliseconds is about 15 ms-2.
The average forward
acceleration of the head may be around 250% of the car's acceleration. That
amounts to an average acceleration of
about 37.5 ms-2 or around 3.8 times the acceleration due to gravity.
However, it is
not the average acceleration that causes whiplash injury, it is the peak acceleration that matters. The peak
acceleration can be more than 5 times the average acceleration but will usually
be about twice the average at about 9g.
The mass of a
human head is about 4.5 kilograms. A sudden acceleration of 9g implies a sudden
shear force at the neck of 397 newtons or about the weight of 40 kilograms, or
about the weight of 89 pounds or 6.4 stone or 0.8 cwt.
It is not
surprising that such a force is likely to cause soft tissue injuries and even
cervical injury.
It seems that
whiplash injuries are likely to be the rule rather than the exception during
rear ended impacts at low speed.
UNCERTAINTY
The brief
qualitative analysis that I have described above has not yielded any precise value
for the force applied to a Claimant's head. There are too many mechanical and
bio-mechanical variables involved that cannot be quantified.
For example,
relevant variable factors include the design of the vehicles involved, the
relative size of the vehicles, the strength of the vehicles at the impact
points, the bumper design, the vehicle age, the vehicle history, the seat back
design, seat back position, position of the head restraints, the seat belts,
occupant neck length, head position, torso position, awareness of the impending
impact, the physical characteristics of the occupant, previous injury history,
and other occupant physical characteristics.
Simulation
tests are subject to criticism on the following grounds:
(a) Small number of volunteers
making it difficult to draw valid generalised conclusions
(b) Results are not comparable with real
life
(c) Volunteers are not representative of
real patients
(d) Volunteers are usually young and
healthy
(e) Crash conditions are
idealised and are not the same as real life conditions
(f) Volunteers are aware of what is going to happen
(g) Head restraints are properly adjusted
(h) Volunteers are perfectly positioned
for the collision.
(i) Volunteers face directly forwards during the
test.
There are also
factors present that are known to increase the risk of whiplash injury. For
example, females are known to be at greater risk.
Injury
severity will be greater for a turned head because it can only move about half
as far as when it is facing directly forwards.
Occupants who
are unaware of an impending impact are also at greater risk. Research has shown
that such occupants are 15 times more likely to sustain a whiplash injury than
those occupants who expect an impact. (Sturzenegger et al, 1994)
Although average
forces and accelerations can be estimated in simplistic situations, whiplash
injury onset and severity will depend on peak
forces and they cannot be reliably quantified.
In my opinion,
it is not possible for an accident investigator to assess whether or not a
claim for whiplash injury is genuine or not. The suggestion that no vehicle
damage equals no occupant injury is clearly invalid. The principal issue is
whether an occupant sustained injury or not. That issue is for medical experts
to determine rather than accident investigators.
I further
suggest that it is not possible for an accident investigator to furnish a
medical expert with any useful information that will assist him to conclude
that a whiplash injury must have resulted from the low speed collision under
investigation or that it could not have resulted from that collision.
REFERENCES
Brault,
Wheeler, Siegmund and Brault, "Clinical response of human subjects to rear
end automobile collisions", 1998, 79, Archives of physical medicine &
rehabilitation.
Castro et al,
"Do whiplash injuries occur in low speed rear impacts?", 1997, 6 Eur
Spine J.
Sturzenegger
et al, "Presenting symptoms and signs of whiplash inmjury: The influence
of of accident mechanisms", Neurology, 1994, April.
Szabo, Welcher
and Anderson, "Human occupant kinematic response in low speed rear end
impacts", 1994, SAE technical paper 940532.
Szabo and
Welcher, "Human subject kinematics and electromyographic activity during
low speed rear impacts", 1996, SAE Technical Paper 962432.
West, Gough
and Harper, "Low speed collision testing using human subjects", 1993,
5, 3, Accident Reconstruction Journal.
FURTHER
Creffield, P,
"Low velocity collisions", 2005, JPIL, 1, 05.
This is an
excellent review of the subject.
Bay Area
Lawyers Network: "Human volunteer studies in published literature on low
velocity rear end crashes".
This is an
extensive review covering some 50 papers on this subject and is available at http://www.baln.org
Arthur Croft
is a prolific writer on the subject. His many papers are available at www.chiroweb.com/archives/
___________________________________________________________
Author: Graham
Greatrix
12 Gillpark Grove,
Tel: 01429 275954
Email: graham@greatrix.co.uk
Website: www.greatrix.co.uk