Getting It Right: Edestrian Ehicle Ollisions
Getting It Right: Edestrian Ehicle Ollisions
Getting It Right: Edestrian Ehicle Ollisions
Getting it Right
AUTHOR
JAMES FIELD DIP. PHYS. (OPEN) CERT. MATH (OPEN)
WEST MIDLANDS POLICE
Foreword
Table of Contents
Foreword
Contents
ii
List of Tables
iii
List of Figures
iv
Executive Summary
Chapter 1
Chapter 2
Introduction
The Study
10
A New Challenge
Case Selection
11
Continued Development
Coefficient of Friction
Chapter 3
Chapter 4
Data Analysis
14
Suggested Methodology
25
Classification of Impacts
14
25
Fender Vault
14
Vehicle Damage
Wrap
15
26
Somersault
16
27
Roof Vault
17
Conclusions
27
Forward Projection
18
Analysis Method
19
20
- Forward Projection
Ground Impact
Appendices
29
References
51
21
24
ii
List of Tables
Table
Title
Page
18
20
iii
List of Figures
Figure
Title
Page
12
12
13
15
16
17
10
18
11
20
12
20
13
21
14
22
iv
Executive Summary
This report describes a 3 year research project that was funded by the Home Office
Police Research Award Scheme to investigate the accuracy of current methods in
reconstructing pedestrian/vehicle collisions.
The aim of the project was to
Collect real world data on current pedestrian/vehicle collisions
Identify current methods of practice throughout the country
Compare real world data with previously published research results
The project achieved these aims by collecting data from Collision Investigation Units
working in 29 police services across the United Kingdom. In total these units
attended the scenes of 6,503 collisions involving 2054 pedestrians, cyclists or
motorcyclists. Scene evidence from 71 of these collisions was found to contain
sufficient data to independently calculate the speed of the striking vehicle and to
calculate the post impact speed of the pedestrian, cyclist or motorcyclist involved.
This data was used to compare the accuracy of current reconstruction methodologies
in estimating the vehicles speed from the distance the pedestrian was thrown post
impact.
It was found that
Current methods could overestimate the collision vehicles impact speed in
30% of the cases investigated
The generally accepted deceleration rate for a sliding/tumbling pedestrian was
found to be too high
This document suggests a method of reconstructing these types of collision which is
more accurate and takes into account the factors which affect the deceleration rates
applicable to pedestrians, cyclists or motorcyclists involved in these types of collisions.
This work was carried out through funding by the Home Office Police Research
Award Scheme. Any views expressed in this document are those of the author and
do not necessarily represent those of the Home Office or any police service. Grateful
thanks must go to Mr Steven Gray (Project Manager) and to all of those police
personnel who contributed to the study (See Appendix A). Without their help and
support the project could not have taken place.
G E T T I N G
I T
R I G H T
Chapter
1
Introduction
n 2002 the Department for Transport reported that 302,605 people were
killed or injured in road vehicle collisions reported to the police in Great
Britain[1], 37,489 (12%) involved pedestrians. However, pedestrian fatalities
account for 22% of all road deaths (fig.1).
In the same year 44,874 motorcyclists or pedal-cyclists were also killed or injured
resulting in 738 deaths a further 22% of road deaths. Whilst these figures
describe a decreasing number of fatalities on previous years data there is a
significant rise of 3% in the number of two-wheeled road users killed or seriously
injured compared with the 2001 data.
Figure 1
22%
56%
22%
pedestrians
2 wheel users
all other rd users
The purpose of this document is to highlight the changes that have taken place
over the last ten years in the forensic investigation of motor vehicle collisions since
the proliferation of vehicles equipped with anti-lock braking systems. This has
presented the collision investigator with new challenges as the calculation of a
vehicles speed from locked wheel tyre marks has become more and more difficult
as these vehicles leave less and less visible marks.
G E T T I N G
I T
R I G H T
A New Challenge
G E T T I N G
I T
R I G H T
Collision investigators often have to decide which research paper is the most
accurate for the particular circumstances of a case. Whilst almost all authors agree
that it is possible to calculate vehicle impact speed from the pedestrians postimpact displacement there is widespread disagreement about how and when this
should be done. A review of some of the published research has shown a clear
disparity between the results found by different researchers[1-93]. Some of the
reasons for these disparities are discussed in detail in later in this chapter.
If the collision reconstruction, from evidence available at the scene is to be
accepted, then the collision investigator must be able to show that the
methodology and mathematics used are both accurate and robust. Since the
majority of the published papers use historical data (some of which now dates
back to the 1960s) or data obtained from testing using North American vehicles,
there is a need to concentrate on the findings obtained from collisions in the UK
involving modern European vehicles.
The findings of this report will be disseminated to all United Kingdom police
services, ACPO and ACPOS for their information and distribution as necessary.
See
Appendix B
G E T T I N G
I T
R I G H T
One fact that was very apparent from the best practice questionnaire was the
general lack of knowledge that collision investigators had about current research in
this field.
Brief details of 8 recent research papers were included in the survey and
respondents were asked if they were aware of them or had used them to assist in
their collision investigations. Details of their replies are shown below in table 1.
Table 1
Evans, A. K.
& Smith, R
Year
Wood,
Denis &
Simms
Ciaran
Happer, Andrew
et al
Hague, D. J.
Han, I. &
Brach,
Raymond.
M.
2001
Fugger,
Thomas. F. Jr.
et al
Randles, Bryan.
C. et al
Toor, A. &
Araszewski, M
2002
Pedestrian
Throw
Kinematics in
Forward
Projection
Collisions
2002
2003
Investigation
and Analysis of
Real-Life
Pedestrian
Collisions
Theoretical vs
Empirical
Solutions for
Veh/Pedestrian
Collisions
1999
Vehicle Speed
Calculations
from
Pedestrian
Throw
Distance
2000
2000
2001
Coefficient
of friction
in
pedestrian
throw
Comprehensive
Analysis Method
for Vehicle Pedestrian
Collisions
Calculation of
impact speed
from
pedestrian
slide distance
Read
18
11
Used
No
Comment
Not
heard of
42
34
51
41
49
47
49
50
Throw
Model for
Frontal
Pedestrian
Collisions
TRAINING
Whilst training issues are always a matter for individual police services the
responses to the survey indicate a clear training need in this particular field. This
matter can be addressed in relation to this area of collision investigation through
this document but, without further research into other areas, it will not be
established if they too are lacking.
ACPO are currently carrying out research to establish the training standards
nationally for police collision investigators but it is not known when this research
will be published.
It is general practice that police collision investigators are trained in two stages
prior to sitting the national examination set by the City & Guilds of London
Institute. After initial training has been received, the collision investigator returns
to their respective police service and works under the supervision of a qualified
officer until such time as the Senior Collision Investigator is satisfied that they are
competent to work alone. Personnel who have only completed the initial training
course are generally designated as being Standard or Basic trained. After
completion of the final module of training they are generally referred to as
G E T T I N G
I T
R I G H T
Advanced trained until they are successful in passing the City & Guilds
examination when they can cite the qualification in reports or statements.
The City & Guilds Institute require that a candidate must pass the national
examination within 5 years, candidates who do not successfully complete all
components are required to re-qualify again by completing an initial training
course. It is a matter for individual police services as to the qualification level
accepted for operational collision investigators.
The level of training of the respondents to the best practice survey are shown in
table 2
Table 2
Level of training Standard/Basic Advanced City & Guilds
No. of respondents
3
5
45
* Only C & G or equivalent are eligible for membership
MITAI
7*
In the UK the Institute of Traffic Accident Investigators (ITAI) are the main
professional body for collision investigators, with membership being offered to
independent (non police) civilian collision investigators, and to suitably qualified
police personnel. An interest in collision investigation is deemed acceptable for a
police collision investigator to be accepted as an Affiliate Member of the Institute.
Affiliate memberships is not a qualification and cannot be quoted as such, it is a
disciplinary offence for any member to do so and could result in their dismissal
from the Institute.
Personnel who have obtained the national City & Guilds of London Institute
qualification may apply for full membership of ITAI and are then allowed to use
the designation MITAI in reports and statements.
The City & Guilds qualification is only available to personnel employed by police
authorities. Because of this an alternative qualification is offered by one training
centre, Accident Investigation Training Services (AiTS), the University Certificate
in Continuing Professional Development in Forensic Road Collision Investigation.
This is an equivalent qualification to the City & Guilds qualification and is also
used by some police services who require the greater flexibility afforded by this
course as it is presented via distance learning.
G E T T I N G
I T
R I G H T
Aprox. 60%
of total
height
Fig 2
Figure 3 shows the idealised scenario from the questionnaire and website,
respondents were asked to state what displacement would be used in calculations.
Fig 3
In direction of vehicle
13m from impact to rest
12.5m
14m
14m in direction
of thrown body
14.5m
15m
13m
13.5m
G E T T I N G
I T
R I G H T
Table 3 details the responses given and indicates that 60% would correctly
measure the displacement with 30% under-estimating it. Only 3 respondents (6%)
would over-estimate the throw displacement all of whom came from different
police services. This does however highlight the fact that personnel from within
the same police service are measuring different pedestrian throw displacements for
the same given data.
Table 3
Displacement (m)
No of Respondents
% of Respondents
12.5
1
2
13.0
7
13
13.5
0
0
14.0
8
15
14.5
32
60
15.0
3
6
Not Stated
2
4
Total
53
100
See
Appendix B1
A further aspect of the study was a review of the published research available.
This was undertaken by obtaining previous research papers from the Society of
Automotive Engineers (SAE), the International Council on the Biomechanics of
Impact (IRCOBI), the Institute of Mechanical Engineers (IMechE), the European
Enhanced Vehicle-safety Committee (EEVC) and discussions with Dr John Searle
and Dr Stephen Ashton who have pioneered research in this field. Full details of
the literature review are given in appendix B1.
One of the major problems faced by any collision investigator is the fact that new
research may be published several times in any year and may be published in many
different countries by different publishers. In the main these papers are then cited
by defence experts and used to cast doubt on the work carried out by the police
collision investigator.
COEFFICIENT OF FRICTION
The coefficient of friction used in pedestrian throw calculations has been a major
source of dispute between authors for decades. In The Trajectories of
Pedestrians, Motorcycles, Motorcyclists, etc., Following a Road Accident,
published in 1983, John & Angela Searle[72] suggested that a coefficient of friction
of 0.66 would be appropriate for a person in normal clothing on a wet or dry
asphalt type surface.
Hill [39] in 1994 published the results of further testing on pedestrian sliding rates in
Calculations of Vehicle Speed from Pedestrian Throw and suggested a coefficient
of friction of 0.80 for a normally clothed pedestrian. This data is listed in the
appendix to chapter 4 for completeness because it was analysed again as part of
the study and also because it has been mis-quoted by other authors.
In 2000 Denis Wood and Ciaran Simms published a paper entitled Coefficient of
Friction in Pedestrian Throw[91] which examined the work of several authors,
including Searle and Hill. Hills data was only partly used in the analysis and the
G E T T I N G
I T
R I G H T
range of values quoted is incorrect. However, Wood & Simms conclude a range
of 0.39 to 0.87 from all of the sources listed.
Also in 2000 Happer et al published a Comprehensive Analysis Method for
Vehicle/Pedestrian Collisions[38] which discusses several aspects of assessing
vehicle speeds by differing methods. Discussed in the paper is the applicable
coefficient of friction for a pedestrian and it is noted that various authors define
this value differently.
Table 4 illustrates this data and produces the range of values found by the different
authors.
Table 4
Trajectory
Slide (1)
Tumbling (2)
Whole Displacement (3)
Minimum Coefficient
0.45
0.7
0.37
Maximum Coefficient
0.72
1.2
0.75
Surface Type
Dry Asphalt
Dry asphalt
Wet/dry asphalt
1. Pedestrian slide coefficients do not take into account the loss of speed
which occurs when the pedestrian strikes the ground. The value is simply
obtained from dragging pedestrians (cadavers, dummies etc.) along the
road surface.
2. Tumbling rates do take into account the loss of speed in multiple impacts
with the ground and are therefore considerably higher values when
compared to any other pedestrian coefficient.
3. The coefficient found from the examination of the whole pedestrian
displacement takes into account the period in time when the pedestrian is
airborne, the impact with the ground, and then the slide/tumble to rest.
Each coefficient is used in different formula to provide an accurate assessment of
the pedestrian post-impact speed.
In 2001 David Hague presented a paper entitled Calculation of Impact Speed
from Pedestrian Slide Distance[35] in which he obtained coefficients for pedestrian
sliding on different surfaces ranging from open coarse tarmac to anti-skid
treatments. The range of values for sliding pedestrians was found to be 0.59 to
0.85 for normal clothing and 0.54 to 0.65 for nylon. This paper is discussed in
greater detail in chapter 4.
2001 also saw the publication of a paper entitled Throw Model for Frontal
Pedestrian Collisions[37] by Han & Brach. This paper challenged the results
obtained by Hill and suggested that the coefficient suggested by him was too high
as he had not taken into account the dummies vertical impact with the road
surface, recalculating his data they suggested a coefficient of 0.74.
However, the coefficient suggested by Hill was for use in a whole displacement
calculation using the Searle equations, not for the application suggested by Han &
G E T T I N G
I T
R I G H T
Brach. A slight reduction in the value of the coefficient is perfectly correct when
using the methodology and equations suggested by them in their paper.
In 2002 Thomas Fugger et al published the results of research examining just one
type of pedestrian impact, Forward Projections. These are explained later in
chapter 5 in more detail. In Pedestrian Throw Kinematics in Forward Projection
Collisions [31] the paper discusses coefficients of friction found for the slide phase
of 141 staged tests involving forward projection dummy impacts, and concludes
that a range of between 0.31 and 0.41 should be used for wet asphalt and 0.43 for
dry asphalt. As is the case in most of the published research there is no direct
comparison made between the coefficient found for a pedestrian and that
applicable for a car skidding under the same circumstances.
In 2003 Toor and Araszewski published Theoretical vs. Empirical Solutions for
Vehicle/Pedestrian Collisions[85] comparing the work of several authors. They
examine pedestrian coefficients and conclude that values between 0.37 and 0.45
are applicable dependant on the type of impact.
With so much variance in the values found for the coefficient of friction for a
pedestrian, the accuracy of all pedestrian/vehicle reconstruction methods are
subject to scepticism. Chapters 3 and 4 discuss this value again and suggest
methods whereby a more accurate value may be estimated.
G E T T I N G
I T
R I G H T
Chapter
2
The Study
n July 2001 the United Kingdom Home Office funded a 3 year research
project examining real world pedestrian/vehicle collisions in the UK. The
aim of the study was to establish if the current reconstruction methods for
pedestrian/vehicle collisions were accurate and also to try and establish a national
policy of best practice that collision investigators could use in their investigations.
See
Appendix C
Appendix C1
A letter was sent to each of the 55 Chief Officers in the UK asking for permission
for their staff to contribute to the study. A standard reply was enclosed with this
letter for them, or a designated person to sign to indicate that permission had been
obtained.
Chief Officers from 47 services responded saying that they were prepared for their
personnel to participate in the study. The start of the data collection phase of the
study was set for 1st November 2001 and scheduled to last for 18 months. As part
of the publicity phase a presentation was given at the National Senior Crash
Investigators Conference in Stafford. After the presentation there was the
opportunity for all attendees to raise anything connected with the study and either
the data collection or the best practice proposals.
A comprehensive newsletter was circulated to all participants at the conference
and also to all nominated contacts who were unable to attend, asking for details of
their experience of pedestrian/vehicle collisions and for proposals or suggestions
of examples of best practice.
See
Appendix C2
Appendix C3
A website was also created to further publicise the project to as wide an audience
as possible (www.jimfield.pwp.blueyonder.co.uk) and also to provide a point of
contact for any service with web access. Details of the aims of the study were
included together with e-mail addresses to contact.
Nominated contacts or the Senior Collision Investigator were then sent details of
what the study was hoping to achieve, written instructions on how the data was to
be selected and recorded and asked to provide monthly returns on the number of
incidents dealt with. These monthly returns identified cases that were potentially
suitable for inclusion in the study.
10
G E T T I N G
I T
R I G H T
CASE SELECTION
Only incidents that were attended by a qualified collision investigator were deemed
suitable for inclusion in the study. To prevent any possible bias over the data
selected, the collision investigator attending the scene made an independent
decision as to the suitability of the case for inclusion in the study based on the
criteria listed below. They were then asked to complete a comprehensive data
collection form.
See
Appendix C4
Appendix C5
Data collection forms were submitted for any collision that was suspected of being
suitable for inclusion by the collision investigator attending the scene. Upon
receipt the form was given a unique reference number which then identified a
probable case. A case was only confirmed as being suitable for the study if all of
the following criteria were known
x
The speed of the striking vehicle was estimated mainly by using standard speed
from skid mark calculations or where vehicles were fitted with tachographs. Once
a suitable case had been identified, the collision investigator was asked to submit
copies of the completed evidential report. To ensure that the Data Protection Act
was not breached, any case files that were submitted were anonymised on receipt,
matched to the unique reference number from the data collection form and the
details entered in to a secure, encrypted Microsoft AccessTM database. Original
photographs were scanned, vehicle identifying marks removed, and all of the case
papers were shredded and disposed of via confidential waste procedures.
The general amount of interest shown throughout the country was extremely high
and the fact that a particular police service did not submit any data should not be
construed as a criticism. Several police services attended numerous collisions
involving suitable subjects, but were unable to provide the specific data required
for the case to be included in the study, simply because of lack of scene evidence.
This lack of scene evidence clearly indicates that there is a need for a better
methodology for investigating cases of this type.
During the 18 month data collection phase, monthly returns or data collection
forms were actually submitted by 29 services (see figure 4 overleaf). Some police
services chose not to submit monthly returns but still submitted scene data for
cases that were relevant to the study, details of all of these participants are also
listed in the acknowledgements section.
11
G E T T I N G
I T
R I G H T
However, figure 6 indicates that the number of collisions that were attended
involving pedestrians or two-wheeled road users was actually less than the 2002
national average of 46% [1], since not all collisions that are recorded as serious
injury are attended by collision investigators.
Fig 6
21%
pedestrians ksi
6%
cyclists ksi
motorcyclists/pillion ksi
54%
19%
In total, 101 cases were reported as potentially being suitable for the study.
Detailed examination of the evidence showed that critical data was missing from
some of the submissions.
Examples of these types of omissions included;
x
In total 30 cases lacked the detail required to carry out further analysis but 71 cases
were found to contain sufficient evidence to be analysed in detail.
13
G E T T I N G
I T
R I G H T
Chapter
3
Data Analysis
CLASSIFICATION OF IMPACTS
In 1981, in a study of 460 pedestrian/vehicle collisions in Northern California,
Ravani et al[68] were able to classify the impacts from 241 collisions into 5 distinct
categories
1. Fender (Bumper) Vault
2. Wrap
3. Somersault
4. Roof Vault
5. Forward Projection
These classifications have become widely accepted throughout the world and are
used to describe the kinematics of the struck pedestrian post-impact. Although
five categories are identified they actually reflect three types of kinematics with
variations being produced as a result of whether or not the striking vehicle was
braking at impact and whether or not the pedestrian was struck a glancing blow.
FENDER VAULT
Fender, or bumper vault, is a classification where the pedestrian is struck by only a
corner of the vehicle and subsequently falls to the side of, and generally behind the
vehicle. As a result of this the pedestrian does not achieve a velocity equal to that
of the vehicle. These types of impact are not usually reconstructed in the UK as
any calculation of the vehicle speed could be a significant under estimate making a
reconstruction unreliable for evidential purposes.
14
G E T T I N G
I T
R I G H T
WRAP
Wrap trajectories involve vehicles that generally have a leading edge at or below
the centre of mass of the struck pedestrian. In these cases the impact causes the
pedestrian to wrap onto the bonnet of the vehicle. In cases where the vehicle is
braking at impact the pedestrians legs then continues to rotate as the two bodies
separate. At low speeds, generally in the region of 20mph or less, the pedestrian
lands feet first as they strike the ground see figure 7.
Fig 7
Pedestrian Kinematics in
Wrap trajectories
15
G E T T I N G
I T
R I G H T
SOMERSAULT
Somersaults are a variation of the wrap trajectory and occur when the impact
speed of the vehicle is greater than around 35mph. In these instances the vehicle
is still braking at impact but, because of the greatly increased impact speed, the
pedestrian continues to rotate post-impact, rising higher into the air. They will
sometimes strike the vehicle again as they fall to the ground before landing in a
head first attitude see fig 8.
Fig 8
Pedestrian Kinematics in
Somersault trajectories
16
G E T T I N G
I T
R I G H T
ROOF VAULT
Roof vault is another variation of the wrap trajectory which generally occurs when
the striking vehicle is not braking at impact or in cases where the vehicle
accelerates through the impact. The pedestrian wraps on to the front of the
vehicle but then continues to either slide up the windscreen or begins to rotate as
the vehicle passes underneath. Multiple contacts with the vehicle are possible with
this type of impact as the vehicle passes leaving the pedestrian to come to rest
behind it see Fig 9.
Fig 9
Pedestrian Kinematics in
Roof Vault trajectories
17
G E T T I N G
I T
R I G H T
FORWARD PROJECTION
Forward projection trajectories generally occur when the height of the striking
vehicles leading edge falls above the centre of mass of the pedestrian. Instead of
being wrapped onto the vehicle the pedestrian is projected forwards before
striking the ground. Small children struck by cars or even adults struck by large
goods vehicles or any other type of flat fronted vehicle, such as buses, generally
follow this trajectory see figure 10.
Fig 10
Pedestrian Kinematics in
Forward Projection
The data for analysis was obtained directly from the 71 cases found to fit the study
criteria. These collisions involved pedestrians and motor vehicles of all types (see
table 5) as well as motorcyclists and pedal-cyclists.
Table 5
Type of Road User Involved
Adult Pedestrian
Adult Pedestrian
Child pedestrian
Motor/Pedal Cyclist
Motor/Pedal Cyclist
Striking Vehicle
Car
Goods Vehicle
Car
Single Deck Bus
Car
18
Type of Collision
Wrap
Forward Projection
Forward projection
Forward Projection
Wrap
G E T T I N G
I T
R I G H T
For the purposes of this study the type of impacts were classified more generally as
either wrap or forward projection. Fender vault was excluded from the analysis
because of the inherent errors present. 56 cases were identified as being wrap
trajectories and 15 cases were identified as being forward projections.
ANALYSIS METHOD
The work of Searle & Searle has been used by the majority of police services since
it was published in 1983[72] and has historically been found to be accurate and
robust In their paper they suggested a coefficient of friction for a normally clothed
pedestrian on an asphalt type surface as being in the region of 0.66.
In his 1994 research into pedestrian collisions, Hill[39] concluded that the coefficient
of friction for a sliding/bouncing/tumbling pedestrian was in the region of 0.800
(for a surface coefficient of friction for vehicle tyres of 0.69 to 0.75). Hill
incorporated these findings into Searles research [72] & [71] and made comparisons
with real world data. Hill concluded that Searles formula for vmin was
underestimating the vehicles speed at impact and suggested a correction factor of
vmin/0.87 (Appendix D).
See
Appendix D
Appendix D1
Hills real world data was taken from 26 collisions involving motor vehicles
constructed prior to 1985, which have distinctly different front-end structures
when compared with modern vehicles (Appendix D1). By 2000 it was found that
Hills adapted formula was overestimating the minimum speed of the vehicle on
impact.
If, as Hill suggests, pedestrian sliding friction is higher than a vehicles skidding
friction then it should be relatively common, especially in forward projection
impacts, for the pedestrian to be overtaken by the vehicle and struck a second time
or even driven over. Since it appears that this occurrence is rare, it would seem to
suggest that pedestrian friction rates must be less than vehicle skidding rates.
The coefficient of friction found in Hills skid testing for a motor car was found to
be between 0.69 and 0.75, although there is no mention in his paper of when the
comparison skid testing was done. Searle & Searle had suggested a coefficient for
a pedestrian as being 0.66 on a similar surface, a value obtained by dragging a live
human subject. This value is a reduction of 12% of the highest value found by
Hill for vehicle tyres to road surface.
19
G E T T I N G
I T
R I G H T
Fig 12
Fig 11
See
Appendix D2
This procedure had the effect of reducing the overestimations to 8 cases (14%)
with the largest error being 5 mph. For completeness the altered friction rate was
applied to Searles vmax equation to provide an upper bound. The results were also
compared to the work of other authors (see Appendix D2).
There were no obvious errors in the data relating to the 8 cases which
overestimated the impact speed. These cases are shown below in table 6 and
graphically in figure 13 overleaf.
Table 6
Job ID
Skid Test Mu
Impact mph
Field mph
03/005
0.700
6.00
15
16
02/007
0.604
7.20
16
17
02/021
0.604
8.80
14
19
01/026
0.607
8.85
17
19
95/001
0.700
10.88
21
22
02/001
0.574
13.65
22
23
01/001
0.777
14.75
21
26
02/014
0.634
18.39
24
28
20
G E T T I N G
I T
R I G H T
Appendix D3
Fig. 13
14.00
12.00
10.00
Impact Speed (m/s)
See
Cases 95/001 & 02/001 overestimate the impact speed by only 1mph and cases
01/001 & 02/014 would be clearly identifiable as being incorrect from an
examination of the damage caused to the vehicle concerned. However the cases
where the total pedestrian throw displacement is less than 10m are the cause for
most concern. It is not always possible to locate the pedestrians orientation on
the ground post-impact if they have been removed prior to the collision
investigators arrival a measurement is generally estimated from body fluid so
the potential margin of error in these measurements could be in the region of 1m
or 10% of the throw displacement (appendix D3).
8.00
6.00
4.00
True Speed
Field min
2.00
0.00
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00
21
G E T T I N G
I T
R I G H T
A methodology was therefore designed that does not require this evidence to be
present providing that there is clear evidence of the point of impact and the final
position of the pedestrian.
A collision between a pedestrian and a flat fronted vehicle will cause the pedestrian
to be accelerated forwards and, unless they are not in contact with the road surface
when struck, they will begin to rotate down onto the road surface as a result of the
friction between their feet and the road (Fig. 14).
Fig 14
Any acceleration downwards will have the effect of shortening the distance before
the pedestrian strikes the ground. Therefore, if it is assumed that the pedestrian is
instantaneously struck to the ground then a maximum speed can be calculated for
the pedestrian for any given throw displacement.
Using standard reconstruction formula, after adjusting the friction rate for a sliding
pedestrian gives
2 * 0.88 * P car * 9.81* Stotal
v upper
This will be an overestimate but provides an upper limit (vupper) above which the
pedestrian, and therefore the vehicle should not have been travelling.
In his paper for the 2001 ITAI Conference, Hague[35] offered the following
formula
u
2Pgs
P 2gH
to calculate the speed (u) for a sliding pedestrian known to have fallen from any
height H. For pedestrians subjected to a forward projection type impact he
22
G E T T I N G
I T
R I G H T
hypothesised that they would fall a distance equal to the height of their centre of
mass.
For the investigation and reconstruction of cases where the point of impact is
known, if the height of centre of mass for the pedestrian can be found or
reasonably estimated then it is possible to calculate how long it took for the
pedestrians centre of mass to strike the road surface (modelled as a particle) using
the equation
t topple
2 * hcom
g
where g is the acceleration due to gravity (9.81 ms-2) and hcom is the known or
estimated centre of mass for the pedestrian. As previously mentioned, this value is
generally accepted as being approximately 60% of the total height of the pedestrian
with variations being noted with small children and obese adults. The formula is
not particularly sensitive to reasonable changes in this value in any case.
It can be argued that, by the time the pedestrians centre of mass is level with the
ground, the pedestrian is fully in contact with the ground too, then full retardation
of the pedestrian has begun. Using this time and the overestimate of the
pedestrians speed (vupper) gives a topple distance from
s topple
v upper * t topple
If this distance is subtracted from the throw distance then the sliding speed for
the pedestrian can be estimated using
v slide
23
G E T T I N G
I T
R I G H T
GROUND IMPACT
The loss of horizontal speed for the pedestrian - as a result of the pedestrian
striking the ground - can also be estimated (assuming an inelastic impact) using the
formula suggested by Hague
v ground strike P 2gH
No loss of horizontal speed is considered during the topple phase and therefore
any contact with the ground by the pedestrians extremities prior to this point will
have the effect of decelerating them and therefore reducing the slide displacement.
Since only the sliding phase of their displacement is used for the calculation of vslide
this should result in an underestimate of their true speed.
This methodology was applied to all of the 15 cases identified for the study and
was found to approximate the vehicles known travelling speed accurately.
See
There was only one instance of an overestimate of the true impact speed of the
vehicle, and where this occurred the marginal error was calculated at being 1mph
(see appendix D4).
Appendix D4
In their book on pedestrian accident reconstruction Eubanks & Hill [25] detail the
results of forward projection testing using dummies. Whilst being the result of
collisions with non-European vehicles, the post-impact trajectory of a pedestrian
struck by any flat fronted vehicle should be identical. These results, where
possible, were also analysed using the above method and the results found to be
accurate.
Appendix D4 also shows all of these values compared with formulae presented by
other authors and the results are compared. It should be noted that the Searle &
Searle equations are primarily designed for wrap trajectories but function well
unless the throw displacement is small.
24
G E T T I N G
I T
R I G H T
Chapter
4
Suggested Methodology
The results of the data analysis and the best practice survey have shown that there
is a widespread discontinuity surrounding the reconstruction of pedestrian/vehicle
collisions in the UK which may result in the estimation of different speeds for the
same given data.
The reconstruction methods of pedestrian vehicle collisions fall in to three groups
All three methods have sound scientific application when appropriately applied,
but for judicial proceedings it may be viewed as more reasonable to use data
gathered at the scene of a single, unique collision rather than rely on data collected
up to 40 years ago that potentially involves a completely different scenario.
The amount of evidence contained at any scene should dictate the methodology
used in the investigation of a particular collision but investigators should not
restrict themselves to any one particular method if the available data is poor.
ESTIMATION
DAMAGE
OF
SPEED
See
Appendix E
FROM
VEHICLE
25
G E T T I N G
I T
R I G H T
extreme caution, but does have its uses, when used to examine the collision
scenario as a whole.
The authors also offer formulae to estimate the vehicles impact speed as a result of
the throw distance for the pedestrian. Whilst the paper is extremely thorough in
its approach it must be noted that the results are based on testing mainly from
North America where vehicle profiles differ to those found in the UK and the rest
of Europe.
The application of v min and v max must be made for all collisions, the result of
the real world testing indicates that v min accurately predicts the minimum velocity
of the vehicle at impact so there is no longer any need to adapt this value. Hill's
suggestion of a v probable is now obsolete.
Where there is a high degree of pedestrian slide, the work of Hague can be used to
approximate a speed from slide marks alone. This will provide an indication as to
whether the speed calculated lies toward the upper or lower bounds. Any
estimations of vehicle speed can be compared with the damage caused to the
vehicle.
Where there is no known point of impact between the vehicle and the pedestrian
the methodology of Hague can be applied if the point of landing and final rest is
known. Additionally, if the height of the pedestrians' trajectory is known (for
example they have been projected over a wall) then the loss of speed from impact
with the ground can also be assessed
26
G E T T I N G
I T
R I G H T
CONCLUSION
When using any of the methodologies published for the reconstruction of
pedestrian/vehicle collisions it is necessary to quote the range of speeds that they
produce and to be aware of any limitations in their application.
In criminal cases it may always be open to 'reasonable doubt' that the minimum
speed applies. The purpose of presenting the findings of the research in such a
format is to produce one value, which will normally be considered as an
underestimate of the vehicle's impact speed and a higher estimate above which the
vehicle should not have been travelling.
Considerable changes have been made to the design and stiffness of vehicle front
structures since the 1980's as manufacturers strive to make vehicle fronts more
'pedestrian friendly' with the result that bumpers are now generally made from
high impact plastics and bonnet lines have become more curved.
The methodologies offered in this document have been applied to recent, real
world collisions and to historical data collected in Europe and America. European
data for 'wrap trajectories' has been used exclusively for this analysis to prevent any
margin of error incurred by introducing differing styles of vehicle designs common
in other geographical areas.
27
G E T T I N G
I T
R I G H T
28
Appendix A
Participating Police Services
Norfolk Constabulary
Cheshire Constabulary
Cleveland Police
Cumbria Constabulary
Northamptonshire Police
Derbyshire Police
Northern Constabulary
Dorset Police
Durham Constabulary
Gloucestershire Constabulary
Staffordshire Police
Hampshire Constabulary
Surrey Police
Hertfordshire Constabulary
Sussex Police
Humberside Police
Tayside Police
Lancashire Constabulary
Leicestershire Constabulary
Warwickshire Constabulary
Merseyside Police
29
Appendix B
Best Practice Questionnaire
30
Appendix B continued
31
Appendix B continued
32
Appendix B continued
33
Appendix B1
Literature Review
A comprehensive review was undertaken to try and establish how many technical papers had been
published examining the area of pedestrian/vehicle collisions. In order to carry this out a search was
made of the databases held by the Society of Automotive Engineers (SAE). This search revealed that
there were literally hundreds of papers examining different areas of pedestrian/vehicle collisions. In
particular the research was divided into the following groups
x
A copy of the 2002 Accident Reconstruction Technology Collection Compact Disk (ISBN 0-76801041-1) was purchased together with the proceedings of the International Council on the
Biomechanics of Impact (CD versions) dating from 1973 to 2003. These databases were then
searched for relevant published data. An internet search was also carried out on the databases held
by the Institute of Mechanical Engineers (IMechE) and using the internet search engine provided by
Google. Papers produced by the European Enhanced Vehicle-safety Committee (EEVC) were also
examined for relevance.
Each research paper was examined and it was decided whether or not it was suitable for use and
reference in the study. Those papers that were suitable were scrutinized and included in the
references using the computer software package Endnote 5TM.
Each research paper was also examined for the references contained within it. These were then
checked to ensure that those papers had also been taken into account.
The breadth of the research considered ranged from the original paper published by Derwyn Severy
in 1963 (thought to be the first research of its kind), to the 2003 paper published by Amrit Toor and
also included the original Doctoral Thesis prepared by Dr Stephen Ashton for which the author is
extremely grateful.
34
Appendix C
Letter to all Chief Officers
Sir Edward Crew
Chief Constable
West Midlands Police
Lloyd House
Colmore Circus
Queensway
Birmingham
B4 6NQ
2001
Dear Sir Edward
Improved Reconstructions of Pedestrian/Vehicle Collisions
I am writing directly to you because we need help with this major international project which is
already producing potentially highly positive results in the area of crash investigations. The UK
Home Office Research Group and my own force have sponsored this project.
What we are looking for is your support with the project and I would ask for your permission to
involve your crash/accident investigation officers in the data collection phase. The work required of
them will take a matter of minutes at the scene and will not involve them in carrying out any further
work on my part.
If you give you permission then I will contact the department directly. All forces that are participating
will be acknowledged in the final paper. This will form the basis of a report, which will be forwarded
to ACPO for their consideration. Once this report has been accepted it will be circulated to all police
services as a model for best practice.
If you agree to this can you please complete the enclosed form and return it to me so that I may
contact your officers directly.
Can I thank you in anticipation of your help,
Jim Field
Senior Crash Investigator
35
Appendix C1
Reply from Chief Officers
Signed
Name of signatory
Rank of signatory
Telephone Number
e-mail address
Appendix C2
Stafford Conference Handout
37
Appendix C2 continued
38
Appendix C2 continued
39
Appendix C2 continued
40
Appendix C3
Letter to Collision Investigators
Ps .
.. Constabulary
Collision Investigation Unit
xxxxxxxxxx RPU
xxxxxxxxxxx Road
xxxxxxxxxx
XXX XXX
Tel
Fax
e-mail j.field@west-midlands.police.uk
Jim Field
41
Appendix C4
Scene Data Collection Form
42
Appendix C5
Data Collection Form Instructions
43
Appendix D
Results of Hills Dummy Testing
28
29
31
30
30
30
30
30
31
29
42
41
42
40
41
42
7.5
8.5
10.7
11.3
12.1
10.8
12.2
12.2
12.9
9.1
20.4
20.8
21.0
19.2
19.0
21.4
1.065
1.008
0.915
0.811
0.757
0.849
0.751
0.751
0.759
0.941
0.881
0.823
0.855
0.849
0.901
0.839
Test Speed
Displacement
Calculated Mu
44
32
18.2
0.573
45
33
18.3
0.606
46
33
17.3
0.641
47
33
18.8
0.590
48
33
16.4
0.676
49
33
15.5
0.716
50
33
15.9
0.698
51
33
16.2
0.685
52
33
16.0
0.693
53
33
15.6
0.711
Date: 05.5.89 Weather: Fine Surface: Airfield tarmac
Conditions: Dry Clothing: Nylon jacket and trousers
54
33
12.5
0.887
55
34
12.9
0.913
56
33
12.6
0.880
57
33
13.2
0.840
58
33
12.4
0.894
59
33
12.8
0.866
60
33
13.3
0.834
61
32
11.9
0.876
62
33
13.3
0.834
63
33
12.8
0.866
Date: 05.5.89 Weather: Fine Surface: Airfield tarmac
Conditions: Dry Clothing: Woollen Boiler Suit
64
33
11.9
0.932
65
33
11.6
0.956
66
33
13.1
0.847
67
33
15.1
0.734
68
32
13.4
0.778
69
33
12.7
0.873
70
33
11.9
0.932
71
33
12.7
0.873
72
32
15.1
0.691
73
33
13.2
0.840
74
42
21.8
0.824
75
42
20.7
0.868
Date: 05.5.89 Weather: Fine Surface: Airfield tarmac
Conditions: Dry Clothing: Rubberised cotton jacket wool
trousers
Tests 44-73 recorded on video
37
41
20.5
0.835
38
43
23.2
0.812
39
43
22.9
0.822
40
42
25.4
0.707
41
51
33.6
0.788
Ave. Mu From Tests Excl. M/cycle Gear
Ave. Mu (Exclude Two High readings)
42
50
32.8
0.776
43
50
32.0
0.796
Date: 7.3.89 Weather: Fine Surface: Airfield tarmac
Conditions: Dry Clothing: terelyne jumper trousers & body warmer
44
0.832
0.825
0.659
Appendix D1
Real World Data Hill 1994
Hill 1
Hill 2
Hill 3
Hill 4
Hill 5
Hill 6
Hill 7
Hill 8
Hill 9
Hill 10
Hill 11
Hill 12
Hill 13
Hill 14
Hill 15
Hill 16
Hill 17
Hill 18
Hill 19
Hill 20
Hill 21
Hill 22
Hill 23
Hill 24
Hill 25
Hill 26
Mu = 0.8
Vehicle Mu
Throw
SFSM
MPH
v(min)
v(probable)
v(max)
0.495
0.907
0.673
0.700
0.664
0.733
0.765
0.766
0.945
0.665
0.945
0.930
0.733
0.760
0.787
0.823
0.764
0.674
0.814
0.694
0.750
0.746
0.798
0.596
0.660
0.767
67.90
9.80
23.50
25.50
13.50
10.90
21.00
20.80
16.10
6.50
19.30
14.50
15.20
17.00
11.30
12.40
14.00
32.00
23.00
29.20
16.00
24.00
18.40
11.30
17.90
16.00
27.80
11.16
19.90
18.53
12.66
12.11
15.74
17.51
14.25
8.92
14.85
14.49
13.57
14.34
12.43
10.93
13.19
18.65
19.25
19.53
15.58
16.54
15.45
11.21
13.12
15.02
62
25
45
41
28
27
35
39
32
20
33
32
30
32
28
24
30
42
43
44
35
37
35
25
29
34
25.49
9.68
15.00
15.62
11.37
10.21
14.18
14.11
12.41
7.89
13.59
11.78
12.06
12.76
10.40
10.89
11.58
17.50
14.84
16.72
12.37
15.16
13.27
10.40
13.09
12.37
29.30
11.13
17.24
17.96
13.07
11.74
16.30
16.22
14.27
9.07
15.62
13.54
13.86
14.66
11.95
12.52
13.31
20.12
17.05
19.22
14.22
17.42
15.25
11.95
15.04
14.22
32.65
12.40
19.21
20.01
14.56
13.08
18.16
18.07
15.90
10.10
17.40
15.09
15.45
16.33
13.32
13.95
14.82
22.41
19.00
21.41
15.85
19.41
16.99
13.32
16.76
15.85
45
Appendix D2
Real World Data - Wrap Trajectories
Job ID
Skid
Test
Mu
Total
Throw
Disp
(m)
Impact
Speed
(ms-1)
99/007
0.800
60.21
28.57
02/016
0.639
53.00
01/024
0.685
03/001
Searle
Hill
Vmin
(ms-1)
Searle
Hill
Vmax
(ms-1)
Evans
Smith
Min
(ms-1)
Evans
Smith
Max
(ms-1)
Happer
Model
Wrap
Min
(ms-1)
Happer
Model
Wrap
Max
(ms-1)
Wood
2000
Min
(ms-1)
Wood
2000
Max
(ms-1)
mph
Field
min
(ms-1)
mph
Field
max
(ms-1)
64
23.58
53
28.84
24.01
30.74
25.62
29.94
24.15
29.15
19.40
34.92
23.14
52
21.08
47
24.18
22.52
28.84
23.90
28.22
22.46
27.46
18.20
32.76
39.10
20.01
45
18.42
41
21.50
19.34
24.77
20.23
24.55
18.84
23.84
15.63
28.14
0.583
37.60
18.86
42
17.31
39
19.45
18.97
24.29
19.79
24.11
18.41
23.41
15.33
27.59
01/002
0.587
34.60
19.94
45
16.64
37
18.73
18.20
23.30
18.90
23.22
17.53
22.53
14.71
26.47
02/025
0.615
34.04
18.89
42
16.72
37
19.01
18.05
23.11
18.73
23.05
17.36
22.36
14.59
26.25
99/003
0.540
32.15
16.80
38
15.64
35
17.31
17.54
22.46
18.14
22.46
16.78
21.78
14.18
25.52
99/001
0.529
30.30
19.06
43
15.08
34
16.64
17.03
21.81
17.55
21.87
16.20
21.20
13.76
24.77
02/024
0.641
29.00
18.94
42
15.60
35
17.92
16.66
21.34
17.12
21.44
15.78
20.78
13.46
24.23
02/012
0.701
28.15
17.19
38
15.71
35
18.46
16.41
21.02
16.83
21.15
15.49
20.49
13.26
23.88
03/004
0.590
27.91
19.46
44
14.96
33
16.86
16.34
20.93
16.75
21.07
15.41
20.41
13.21
23.77
00/005
0.656
27.50
16.79
38
15.28
34
17.65
16.22
20.78
16.61
20.93
15.28
20.28
13.11
23.60
01/009
0.586
27.20
15.98
36
14.74
33
16.59
16.13
20.66
16.51
20.83
15.18
20.18
13.04
23.47
02/006
0.551
26.50
15.14
34
14.29
32
15.88
15.93
20.39
16.27
20.59
14.94
19.94
12.87
23.17
02/029
0.682
25.70
15.48
35
14.92
33
17.40
15.68
20.08
15.99
20.31
14.66
19.66
12.67
22.81
01/018
0.600
25.10
16.95
38
14.26
32
16.13
15.50
19.85
15.78
20.10
14.45
19.45
12.52
22.54
02/027
0.702
24.45
17.34
39
14.65
33
17.21
15.30
19.59
15.54
19.86
14.22
19.22
12.36
22.25
99/006
0.600
23.41
17.79
40
13.77
31
15.57
14.97
19.17
15.16
19.48
13.85
18.85
12.10
21.77
01/031
0.663
23.16
17.26
39
14.06
31
16.28
14.89
19.07
15.07
19.39
13.76
18.76
12.03
21.66
01/035
0.765
23.00
18.98
42
14.46
32
17.43
14.84
19.00
15.01
19.33
13.70
18.70
11.99
21.58
95/004
0.700
22.90
14.11
32
14.16
32
16.64
14.80
18.96
14.97
19.29
13.66
18.66
11.96
21.53
99/004
0.508
22.05
13.01
29
12.70
28
13.91
14.53
18.60
14.65
18.97
13.34
18.34
11.74
21.13
02/008
0.776
21.00
18.54
41
13.85
31
16.77
14.18
18.16
14.25
18.57
12.94
17.94
11.46
20.62
02/017
0.618
20.77
15.87
36
13.08
29
14.89
14.10
18.06
14.16
18.48
12.86
17.86
11.39
20.51
00/002
0.761
20.75
14.01
31
13.72
31
16.51
14.09
18.05
14.15
18.47
12.85
17.85
11.39
20.50
02/028
0.595
19.96
14.39
32
12.69
28
14.32
13.82
17.70
13.83
18.15
12.54
17.54
11.17
20.10
02/019
0.630
19.70
17.63
39
12.80
29
14.64
13.73
17.58
13.73
18.05
12.44
17.44
11.10
19.97
02/013
0.670
19.70
14.11
32
13.00
29
15.10
13.73
17.58
13.73
18.05
12.44
17.44
11.10
19.97
02/014
0.634
18.39
10.58
24
12.39
28
14.19
13.27
16.99
13.19
17.51
11.91
16.91
10.72
19.30
01/025
0.541
17.58
12.83
29
11.57
26
12.82
12.97
16.61
12.85
17.17
11.57
16.57
10.48
18.87
99/005
0.626
17.00
13.35
30
11.87
27
13.56
12.76
16.33
12.60
16.92
11.32
16.32
10.31
18.55
00/004
0.513
16.50
10.99
25
11.02
25
12.09
12.57
16.09
12.38
16.70
11.11
16.11
10.16
18.28
01/017
0.560
16.00
20.41
46
11.16
25
12.44
12.37
15.85
12.16
16.48
10.89
15.89
10.00
18.00
01/032
0.676
15.51
13.74
31
11.56
26
13.45
12.18
15.60
11.94
16.26
10.67
15.67
9.85
17.72
02/004
0.783
15.00
12.27
27
11.73
26
14.24
11.98
15.34
11.71
16.03
10.44
15.44
9.68
17.43
01/001
0.777
14.75
9.60
21
11.61
26
14.07
11.88
15.22
11.59
15.91
10.33
15.33
9.60
17.28
02/020
0.615
14.60
13.70
31
10.95
24
12.45
11.82
15.14
11.52
15.84
10.26
15.26
9.55
17.19
02/001
0.574
13.65
9.80
22
10.38
23
11.63
11.43
14.64
11.07
15.39
9.81
14.81
9.24
16.63
01/016
0.690
12.90
13.41
30
10.60
24
12.40
11.11
14.23
10.70
15.02
9.45
14.45
8.98
16.16
01/016
0.690
12.90
13.41
30
10.60
24
12.40
11.11
14.23
10.70
15.02
9.45
14.45
8.98
16.16
01/034
0.630
12.00
10.31
23
9.99
22
11.42
10.72
13.72
10.24
14.56
9.00
14.00
8.66
15.59
01/019
0.711
11.39
13.44
30
10.02
22
11.82
10.44
13.37
9.92
14.24
8.68
13.68
8.44
15.19
46
Appendix D2
Real World Data - Wrap Trajectories (Contd.)
mph
Field
max
(ms-1)
Searle
Hill
Vmin
(ms-1)
Searle
Hill
Vmax
(ms-1)
Evans
Smith
Min
(ms-1)
Evans
Smith
Max
(ms-1)
Happer
Model
Wrap
Min
(ms-1)
Happer
Model
Wrap
Max
(ms-1)
Wood
2000
Min
(ms-1)
Wood
2000
Max
(ms-1)
22
11.52
10.25
13.13
9.70
14.02
8.47
13.47
8.28
14.91
22
11.47
10.20
13.07
9.65
13.97
8.41
13.41
8.25
14.84
8.98
20
10.04
9.93
12.72
9.34
13.66
8.11
13.11
8.03
14.45
22
9.39
21
11.14
9.73
12.47
9.10
13.42
7.88
12.88
7.87
14.16
9.01
20
8.82
20
10.03
9.51
12.18
8.85
13.17
7.62
12.62
7.69
13.83
8.85
7.82
17
8.49
19
9.63
9.20
11.79
8.49
12.81
7.27
12.27
7.44
13.39
0.604
8.80
6.44
14
8.46
19
9.58
9.18
11.75
8.46
12.78
7.24
12.24
7.42
13.35
01/008
0.652
7.70
8.16
18
8.08
18
9.31
8.58
10.99
7.77
12.09
6.57
11.57
6.94
12.49
02/030
0.770
7.50
9.72
22
8.27
18
9.99
8.47
10.85
7.64
11.96
6.44
11.44
6.85
12.32
02/007
0.604
7.20
6.97
16
7.65
17
8.67
8.30
10.63
7.45
11.77
6.24
11.24
6.71
12.07
03/003
0.750
6.85
14.15
32
7.86
18
9.42
8.10
10.37
7.21
11.53
6.01
11.01
6.54
11.78
01/021
0.694
6.51
10.25
23
7.54
17
8.83
7.89
10.11
6.97
11.29
5.78
10.78
6.38
11.48
01/028
0.725
6.50
9.62
22
7.60
17
9.02
7.89
10.10
6.97
11.29
5.77
10.77
6.37
11.47
03/005
0.700
6.00
6.52
15
7.25
16
8.52
7.58
9.70
6.61
10.93
5.42
10.42
6.12
11.02
Job ID
Skid
Test
Mu
Total
Throw
Disp
(m)
Impact
Speed
(ms-1)
01/005
0.700
10.98
11.42
95/001
0.700
10.88
01/027
0.566
01/014
mph
Field
min
(ms-1)
26
9.81
9.54
21
9.76
10.31
10.11
23
0.726
9.90
9.91
01/015
0.617
9.45
01/026
0.607
02/021
47
Appendix D3
Pedestrian Orientation Post Impact
Final Rest
could be here
Estimated Point
of Final Rest
Final Rest
could be here
48
Appendix D4
Real World Data Forward Projection Impacts
Job ID
Skid
Test
Mu
Total
Throw
Disp
(m)
Impact
Speed
(ms-1)
00/003*
0.700
61.01
01/007*
0.599
02/026
mph
Ped
Height
(m)
CoM
(m)
Flight
(m)
Upper
Limit
(ms-1)
Topple
Dist
(m)
Loss at
Impact
(ms-1)
Slide
(m)
Estimate
(ms-1)
mph
diff
27.77
62
Est
1.00
0.00
28.21
0.00
0.00
61.01
27.15
61
55.60
25.19
56
1.68
1.01
0.00
24.92
0.00
0.00
55.60
23.98
54
0.581
34.04
21.27
48
Est
0.70
19.20
7.25
1.89
26.79
16.50
37
11
02/018
0.540
30.70
21.01
47
1.52
0.91
17.58
7.27
1.93
23.43
14.91
33
14
02/005
0.670
29.20
20.64
46
Est
1.00
19.10
8.62
2.61
20.58
15.65
35
11
02/023
0.708
21.76
14.72
33
Est
1.00
16.95
7.65
2.76
14.11
13.42
30
02/003
0.646
20.85
11.91
27
1.78
1.07
15.84
7.07
2.49
13.78
12.64
28
-1
02/009
0.547
20.50
14.83
33
Est
1.00
14.46
6.53
2.13
13.97
11.68
26
02/031
0.682
16.80
14.45
32
Est
1.00
14.61
6.60
2.66
10.20
11.28
25
02/011
0.658
13.50
15.66
35
1.63
0.98
12.87
5.49
2.43
8.01
9.84
22
13
00/006
0.673
12.60
11.49
26
Est
1.00
12.57
5.68
2.62
6.92
9.35
21
01/012
0.551
11.50
11.59
26
Est
1.00
10.87
4.91
2.15
6.59
8.21
18
01/011
0.676
9.00
8.97
20
Est
0.50
10.65
3.40
1.86
5.60
8.30
19
02/015
0.682
8.00
7.13
16
1.63
0.98
10.08
4.31
2.51
3.69
7.06
16
98/001
Eubanks
183-3
Eubanks
182-2
Eubanks
175-4
Eubanks
139-3
0.677
3.85
6.45
14
Est
1.00
6.97
3.15
2.64
0.70
3.90
0.520
18.84
13.63
30
1.75
1.07
9.75
13.51
6.31
2.10
12.53
10.81
24
0.570
5.79
6.08
14
1.75
1.07
4.51
7.84
3.66
2.30
2.13
5.12
11
0.720
11.89
10.37
23
1.14
0.64
7.07
12.63
4.56
2.25
7.32
9.80
22
0.810
10.74
10.15
23
1.75
1.07
6.50
12.74
5.95
3.27
4.80
8.82
20
49
1.90
9.00
Appendix E
Estimation of Vehicle Speed from Damage
(From Happer[38] 2000)
Table 5.1
Approximate Vehicle
Impact Speed
KPH
<20
35
60
Table 5.2
Wrap Impacts
General Damage Summary
Approximate Vehicle
Impact Speed
KPH
<20
25
25 - 40
Around 16 - 25 mph
Around 25 mph
40
Around 25 - 31 mph
Around 31 mph
40 - 50
50
Around 31 34 mph
50 - 55
Around 37 mph
Greater than 37 mph
60
>60
Around 43 mph
70
Around 50 mph
80
Head contact near middle of windshield for typical braking lowfronted vehicle.
Head contact near bottom edge of windshield when vehicles upper
leading edge near pedestrians C.G.
More probable body to roof contact.
Head contact near upper frame of windshield; significant
deformation of body panels.
Pelvic contact with roof; roof deformation (unbraked vehicle).
50
References
1.
DfT, Road Accident Casualties by Road User Type and Severity: 1992 - 2002. 2004.
2.
3.
Anderson, R., et al., Severity and Type of Pedestrian Injuries Related to Vehicle Impact Locations
and Results of Sub-System Impact Reconstruction. IRCOBI, 2002.
4.
5.
Appel, H., G. Sturtz, and L. Gotzen, Influence of Impact Speed and Vehicle Parameters on
Injuries of Children and Adults in Pedestrian Accidents. IRCOBI, 1975: p. 83 - 113.
6.
Aronberg, R., Airborne Trajectory Analysis Derivation for Use in Accident Reconstruction.
Society of Automotive Engineers, 1990. 900367.
7.
8.
Ashton, S.J., Pedestrian Injuries and the Car Exterior. PhD Thesis, 1978: p. 96 - 106.
9.
Ashton, S.J., Factors Associated with Pelvic and Knee Injuries in Pedestrians Struck by the Fronts
of Cars. Society of Automotive Engineers, 1981. 811026.
10.
Ashton, S.J., D. Cesari, and J. Van Wijk, Experimental Reconstruction and Mathematical
Modelling of Real World Pedestrian Accidents. Society of Automotive Engineers, 1983.
11.
Ashton, S.J. and G.M. Mackay, Benefits from Changes in Vehicle Exterior Design - Field
Accident and Experimental Work in Europe. Society of Automotive Engineers, 1983.
830626.
12.
Ashton, S.J., J.B. Pedder, and G.M. Mackay, A Review of Riders and Pedestrians in Traffic
Collisions. IRCOBI, 1976.
13.
Becke, M. and G. Uwe, Sliding Distances of Pedestrians on Wet Roads. Verkehrs-unfall und
Fahrzeugtechnik, 1986. Heft 12.
14.
Bhalla, K., et al., Vehicle Impact Velocity Prediction from Pedestrian Throw Distance: Trade-Offs
between throw Formulae, Crash Simulators and Detailed Multi-Body Modeling. IRCOBI, 2002.
15.
Bratten, T.A., Development of a Tumble Number for Use in Accident Reconstruction. Society of
Automotive Engineers, 1989. 890859.
16.
Brooks, D., et al., Reconstruction of Real World Pedestrian Impact. Society of Automotive
Engineers, 1986. 860210.
17.
51
18.
Brun-Cassan, F., C. Tarriere, and A. Fayon, Comparison of Behaviours for Part 572 and
APROD Dummies Tested as Pedestrians Impacted by a Car Under Identical Test Conditioning.
Society of Automotive Engineers, 1983. 830185.
19.
Brun-Cassan, F., et al., Reconstruction of Actual Car-Pedestrian Collisions with Dummy and
Cadavers. Society of Automotive Engineers, 1983. 830053.
20.
21.
Cavallero, C., et al., Improvement of Pedestrian Safety: Influence of Shape of Passenger Car-Front
Structures Upon Pedestrian Kinematics and Injuries: Evaluation Based on 50 Cadaver Tests.
Society of Automotive Engineers, 1983. 830624.
22.
DfT, Road Casualties Great Britain Main Results 2001 (Provisional). 2002.
23.
Eames, M.H.A., A. Cosgrove, and R. Baker, A Full Body Model to Determine the Total
Body Centre of Mass During the Gait Cycle in Adults and Children. British Association of
Clinical Anatomists, 1998.
24.
Eubanks, J.J. and W.R. Haight, Pedestrian Involved Traffic Collision Reconstruction
Methodology. Society of Automotive Engineers, 1992. 921591.
25.
Eubanks, J.J. and P.F. Hill, Pedestrian Accident Reconstruction and Litigation. 1998: Lawyers
& Judges Publishing Co., Inc.
26.
Evans, A.K. and R. Smith, Vehicle Speed Calculations from Pedestrian Throw Distance. Proc
Instn Mech Engrs, 1999. 213 Part D: p. 441 - 447.
27.
Farisse, J., et al., Evaluation of Pedestrian Lower Limbs Injury Severity: Proposals for a Detailed
Scale. Society of Automotive Engineers, 1983. 830188.
28.
Field, J.T. Analysis of Real World Pedestrian/Vehicle Collisions in the United Kingdom. in
Institute of Traffic Accident Investigators 6th International Conference 2003. 2003. StratfordUpon-Avon, Warwickshire, England.
29.
30.
Fugger, T., F, Jr. , et al., An Analysis of Pedestrian Dummy Throw Characteristics for Forward
Projection Pedestrian Collisions. IRCOBI, 2001.
31.
Fugger, T.F.J., et al., Pedestrian Throw Kinematics in Forward Projection Collisions. Society of
Automotive Engineers, 2002. 2002-01-0019.
32.
Gibson, T.J., K.A. McCaul, and A.J. McLean, Investigation of Head Injury Mechanisms in
Motor Vehicle Accidents - A Multidisciplinary Approach. Society of Automotive Engineers,
1985. 850093.
33.
34.
Greetham, T.A. and D. Guenther, An Analysis of Head Injuries in Real World Pedestrian
Accidents. Society of Automotive Engineers, 1983. 830057.
52
35.
Hague, D.J., Calculation of Impact Speed from Pedestrian Slide Distance. ITAI International
Conference 2001, 2001: p. 101 - 108.
36.
Haight, W.R. and J.J. Eubanks, Trajectory Analysis for Collisions Involving Bicycles and
Automobiles. Society of Automotive Engineers, 1990. 900368: p. 143-161.
37.
Han, I. and R.M. Brach, Throw Model for Frontal Pedestrian Collisions. Society of
Automotive Engineers, 2001. 2001-01-0898.
38.
Happer, A., et al., Comprehensive Analysis Method for Vehicle/Pedestrian Collisions. Society
of Automotive Engineers, 2000. 2000-01-0846.
Hill, G.S., Calculations of Vehicle Speed from Pedestrian Throw. Impact, 1994. Spring: p. 1821.
39.
40.
Hoffman, J., A. Kretzschmar, and M.V. Blundell. Investigation into the use of Adaptable
Car Structures Concepts for Pedestrian Impact Protection. in Vehicle Safety 2002. 2002. London,
England: Institute of Mechanical Engineers.
41.
42.
Kallieris, D. and G. Schmidt, New Aspects of Pedestrian Protection Loading and Injury Pattern
in Simulated Pedestrian Accidents. Society of Automotive Engineers, 1988. 881725.
43.
44.
Kramer, M., Pedestrian vehicle accident simulation through dummy tests. Society of Automotive
Engineers, 1975. 751165.
45.
46.
Kuehnel, A. and H. Appel, First Step to a Pedestrian Safety Car. Society of Automotive
Engineers, 1978. 780901.
47.
Kuehnel, A. and H. Rau, The Fidelity of Dummies in Pedestrian Collisions. IRCOBI, 1978.
48.
Lambourn, R.F., The Calculation of Motorcycle Speeds From Sliding Distances. Society of
Automotive Engineers, 1991. 910125: p. 77-90.
49.
Liu, X. and J. Yang, Development of Child Pedestrian Mathematical Models and Evaluation with
Accident Reconstructions. IRCOBI, 2001. 2001.
50.
Liu, X. and J. Yang, Effects of Vehicle Impact Velocity and Front-End Structure on the Dynamic
Responses of Child Pedestrians. IRCOBI, 2002.
51.
52.
Lucchini, E. and R. Weissner, Differences Between the Kinematics and Loadings of Impacted
Adults and Children Results from Dummy Tests. IRCOBI, 1980.
53.
Matsui, Y., H. Ishikawa, and A. Sasaki, Pedestrian Injuries Induced by the Bonnet Leading
Edge in Current Car-Pedestrian Accidents. Society of Automotive Engineers, 1999. 199901-0713.
53
54.
Moser, A., et al., Validation of the PC-Crash Pedestrian Model. Society of Automotive
Engineers, 2000. 2000-01-0847.
55.
Moser, A., H. Steffan, and G. Kasanicky, The Pedestrian Model in PC-Crash - The
Introduction of a Multi Body System and its validation. Society of Automotive Engineers,
1999. 1999-01-0445.
56.
57.
Niederer, P., et al., The Reliability of Anthropometric Test Devices, Cadavers, and Mathematical
Models as Pedestrian Surrogates. Society of Automotive Engineers, 1983. 830184.
58.
Olkkonen, S., The Impact Sites of Car Fronts for Fatal Injuries of Pedestrians and Cyclists,
Analysis of the Urban Accidents in Finland in 1998-1999. IRCOBI, 2002.
59.
Olson, P.L., Driver Perception Response Time. Society of Automotive Engineers, 1989.
890731.
60.
Otte, D., Injury Mechanism and Crash Kinematic of Cyclists in Accidents - An Analysis of Real
Accidents. Society of Automotive Engineers, 1989. 892425.
61.
Otte, D., Severity and Mechanism of Head Impacts in Car to Pedestrian Accidents. IRCOBI,
1999: p. 329 - 341.
62.
Otte, D. Importance of Speed Related Traces on the Scene for Determination of Collision Speed in
Pedestrian and Bicycle Accidents. in Institute of Traffic Accident Investigators International
Conference. 2001. York, England.
63.
Otte, D. and T. Pohlemann, Analysis and Load Assessment of Secondary Impact to Adult
Pedestrians After Car Collisions on Roads. IRCOBI, 2001. 2001.
64.
Pereira, J.M. and S.J. Ashton, Experimental Development of Pedestrian Head Injury Tolerance
Data. Society of Automotive Engineers, 1983. 830056.
65.
Pritz, H.B., et al., Experimental study of pedestrian injury minimization through vehicle design.
Society of Automotive Engineers, 1975. 751166.
66.
Pritz, H.B. and J.M. Pereira, Pedestrian Hip Impact Simulator Development and Hood Edge
Location Consideration on Injury. Society of Automotive Engineers, 1983.
67.
Randles, B.C., et al., Investigation and Analysis of Real-Life Pedestrian Collisions. Society of
Automotive Engineers, 2001. 2001-01-0171.
Ravani, B., D. Brougham, and R.T. Mason, Pedestrian post-impact kinematics and injury
patterns. Society of Automotive Engineers, 1981. 811024.
68.
69.
Schmidt, D.N. and D.A. Nagel, Pedestrian Impact Case Study. Proc. 15th Conference of
the Assoc. for Automotive Medicine, 1971: p. 151 - 167.
70.
Schneider, H. and G. Beier, Experiment and accident: comparison of dummy test results and real
pedestrian accidents. Society of Automotive Engineers, 1974. 741177.
71.
Searle, J., The Physics of Throw Distance in Accident Reconstruction. Society of Automotive
Engineers, 1993. 930659.
54
72.
Searle, J.A. and A. Searle, The Trajectories of Pedestrians, Motorcycles, Motorcyclists, etc.,
Following a Road Accident. Society of Automotive Engineers, 1983. 831622.
73.
74.
Smith, R. and A. Evans, Pedestrian Throw Calculations. Impact, 2000. Spring: p. 8 - 11.
75.
Stammen, J.A., S.-w. Ko, and G. Dennis, A Demographic Analysis and Reconstruction of
Selected Cases from the Pedestrian Crash Data Study. Society of Automotive Engineers,
2002. 2002-01-0560.
76.
Stcherbatcheff, G., et al., Kinematics of a Pedestrian and a Two-Wheeler Impacted by the Front
of a Car. IRCOBI, 1976.
77.
Stcherbatcheff, G., et al., Simulation of Collisions Between Pedestrians and Vehicles Using
Adult and Child Dummies. Society of Automotive Engineers, 1975: p. 753-784.
78.
Sturtz, G., Repetition of Real-Child Pedestrian Accidents Using a High Instrumented Dummy.
IRCOBI, 1978.
79.
Sturtz, G. and E.G. Suren, Kinematics of Real Pedestrian and Two-Wheel Rider Accidents and
Special Aspects of the Pedestrian Accident. IRCOBI, 1976: p. 1 - 23.
80.
Sturtz, G., et al., Analyse Von Bewegungsablauf, Verletzungsursache, -Schwere und - Folge Bei
Fusgangerunfallen mit Kindern Durch Unfallforschung am Unfallort. IRCOBI, 1974.
81.
82.
Takeuchi, K., et al., Car-Pedestrian Accidents in Gothenburg During Ten Years. IRCOBI,
1998.
83.
Taneda, K., M. Kondo, and K. Higuchi, Experiment on Passenger Car and Pedestrian
Dummy Collision. IRCOBI, 1973.
84.
Teresinski, G. and R. Madro, A Comparison of Mechanisms of Ankle, Knee, Pelvis and Neck
Injuries in Pedestrians and in Cyclists According to the Direction of Impact and Type of Vehicle.
IRCOBI, 2003.
85.
86.
Toor, A., et al., Revision and Validation of Vehicle/Pedestrian Collision Analysis Method. SAE
2002 - 01 - 0550, 2002.
Verma, M.K. and B.S. Repa, Pedestrian Impact Simulation - A Preliminary Study. Society of
Automotive Engineers, 1983. 831601: p. 15-29.
87.
88.
Wolfe, A.C. and J. O'Day, Some Vehicle Factors Related to Type and Severity of Pedestrian
Injuries. Society of Automotive Engineers, 1983. 830633.
89.
Wood, D., Application of a Rigid Body Impact Model to the Pedestrian-Car Collision. IRCOBI,
1988.
55
90.
Wood, D. and S.O. Riordain, Mathematical Analysis of Pedestrian Lower Leg Impacts.
IRCOBI, 1989. 1989.
91.
92.
Wood, D. and C. Simms, A Hybrid Model for Pedestrian Impact and Projection.
International Journal of Crashworthiness, 2000. Vol 5(No 4): p. 393 - 403.
93.
Wood, D.P., Application of a Pedestrian Impact Model to the Determination of Impact Speed.
Society of Automotive Engineers, 1991. 910814.
56
Impact Point
Impact Point
Impact Point
Landing Point
Landing Point
Landing Point
Slide/Final Rest
Final Rest
Slide/Final Rest
Us e
U se
U se
Field
Hague