Professional Documents
Culture Documents
Searle PDF
Searle PDF
Searle PDF
Angela Searle
277
ation is derived in the form of upper and lower on substituting back this valué into Formula 1
bounds on the possible range of velocity valúes. we find:-
THEORY pgs
v . (2)
min
It is necessary to derive a formula for the Ni 1 +
total trajectory, including bouncing and sliding, Establishing the máximum valué of the velocity
for an object projected on a level surface. The is slightly more complicated but is less often
projection takes place at velocity V and angle 9, required. It can be shown that the máximum is
and the object has a coefficient of friction p given by:-
and a coefficient of rebound e on the surface.
This derivation has been relegated to an v (3)
Appendix, as the mathematics is somewhat detailed. max
The formula obtained by the mathematical treat-
ment is:- provided it may be assumed that the angle of
projection was below some critical valué 6 ^
This critical valué depends on the coefficient’
v = ' C h ¡Í____________ ~ (D of friction in the following way:-
(eos © + sin 9)
0
r*»
o
Another point of interest is that the 0.7
valúes of velocity obtained from the formula 0.8 77°
are not critically dependent on p , the coeffic
0
ient of friction. The velocity is most dependent 0.9 00
upon p when the angle of projection is very small.
In that case the velocity estimate varies only
half as much as any variation in the valué of p .
For any higher angle of projection the depend- This Table is to be read in the following
eney decreases, becoming quite low at middle way, taking as an example the case where the
angles. This means that the velocities estimated coefficient of friction of the surface is 0.6.
by this method are not so dependent upon the co For this valué, if the angle of projection is
efficient of friction as, say, vehicle velocities known not to have been higher than 62 then the
estimated from skid marks. velocity of projection must have been less than
\¡2 p g s , that is >/1.2 gs. If on the other hand
DETERMINATION OF UPPER AND LOWER BOUNDS ON the angle of projection may have been larger
VELOCITY than 62 , no upper bound to the velocity can be
established. It is very often known that the
If the angle of projection is not known, angle of projection is small, even where it is
which is the situation in the vast majority of not actually known, and in such cases an upper
cases, Formula 1 cannot give a unique estimate bound can be put to the velocity as well as a
for the velocity. It can however give upper lower bound.
and lower bounds, by considering those valúes of It is interesting to examine how these upper
the projection angle 9 that will maximise and and lower bounds vary with the coefficient of
minimise the expression. friction p. This is shown in Figure 1 from which
Considering the valué of 9 that minimises it will be seen that the range between the bounds
the Formula 1 expression, straightforward calculus is small at low coefficients of friction, increas-
shows that this occurs when:- ing to some 30% at higher ones. This means that,
even with high coefficients of friction, the true
tan 9 = p that is 9 = arctan p velocity can be bracketed within - 15%. On
Figure 1 the line representing the upper bound
278
has been shown broken, since it is subject to the The mathematical parts of this paper have
proviso that the angle of projection is not derived formulae for the upper and lower limits
any larger than the critical projection angle. of speed for a given throw distance. It is
On Figure 1 the bounds have been expressed therefore possible to compare the predictions of
in non-dimensional terms, as a fractional valué these formulae with the field results of Professor
between 0 and 1. The actual numerical valué is Appel. In making this compar i son it is clear
obtained by multiplying by J 2 g s . Incidently it that the upper bound formulae will need to be
may be noted that to estimate the minimum velocity employed with the máximum reasonable valué of
on the basis that the whole trajectory is airborne friction, and the lower bound formulae with the
is equivalent to taking the coefficient of fric- minimum reasonable valué. This is because the
tion to be unity. variation in friction contributes to the variation
of results observed in the field.
FRICTION VALUES
The compar ison, using = 1 . 0 and
p . = 0.3, is shown in Figure 5. It will be seen
The friction valué to be used with the a?1once that there are noticeable differences
above formulae and graphs is of course that between tbe field results and the theoretical
actually present, at the time of the accident, predictions. Although the predicted curves have
between the projected object and the surface on the right shape, the field results are giving a
which it bounces and si ides to rest. To assist higher speed for a given distance. It is clear
readers in making use of the formulae, however, that there is some major factor operating which
several typical coefficients of friction have has not been included in the analysis so far.
been measured. These are shown in Table 2. A further consideration of the results soon
suggests what this might be. The formulae are
TABLE 2 COEFFICIENTS OF FRICTION FOR VARIOUS predicting the initial velocity of the pedestrian,
SITUATIONS not the velocity of the vehicle. It is often the
case that the velocity of the pedestrian, after
being struck, is somewhat less than the original
Coefficient velocity of the vehicle. Indeed, in some of the
Object Surface
of Friction more severe accidents the vehicle passes under-
neath the pedestrian, providing a practical
demonstration that the pedestrian has less
Ylotorcycle on velocity. Although Professor Appel has excluded
its side cases in which the contact was very glancing, it
Dry asphalt 0.35 to Q .50 is clear that in many impacts the velocity of the
Wet asphalt 0.30 to 0.40 pedestrian will be less than that of the vehicle.
Using this concept it is natural to express
Per son in the velocity of the pedestrian as a percentage of
normal clothes that of the vehicle. When this is done the
Dry asphalt) following figures are found to fit the data well:-
0.66*
Wet asphalt)
Dry grass) Projection efficiency %
0.79*
Wet grass) Adults 72.6
Children 81.3
*No substantial difference between wet and dry Low front vehicles 68.9
Hiqh front vehicles 79.3
APPLICATION TO FIELD DATA ON PEDESTRIANS
All groups combined 77.5
An Ínteresting application is to the pedest-
rian "throw distances" measured by Professor The fit now obtained to Professor Appel's
Appel et al (6). These were the distances that data is shown in Figure 4, where the mean pre
pedestrians in road accidents were estimated to dicted lines for thegroups have been superim-
have been projected, measured in the direction posed on the field data. It is useful for some
of travel of the vehicle. In their paper, Appel purposes to sepárate out these groups into the
et al give graphs of throw distance as a function more specific categories:-
of vehicle speed. Distinctions are made between Projection efficiency %
adults and children, and between vehicles with
low and high front ends. Adults
with low front vehicles 64.0
For each graph Appel et al show the range with high front vehicles 74.4
of throw distances found at a given speed. Such Children
information can equally readily be presented as with low front vehicles 72.7
a range of speeds for a given throw distance, with high front vehicles 83.1
merely by turning the graph round. For the
reader's convenience Professor Appel1s results All groups combined 77.5
are shown in Figure 2, presented in this form.
279
These figures refiect what might be are accelerated only to a pecentage of the
expected from a consideration of centre of full velocity of the vehicle. This percent-
gravity height. Adults struck by low front age is larger for children than for adults
cars are accelerated to less than two thirds and larger when the vehicle has a high front
of the c a r 's speed whereas children struck by end than when it does not. For child pedes
high front cars are accelerated to nearly the trians struck by vehicles with high front
full valué of the c a r 1s speed. The other two ends, the percentage approaches 100%.
cases give intermedíate levels. With the inclu
sión of this factor, the formulae derived in REFERENCES
this paper give an accurate prediction of the
field results. 1. LANGWIEDER, K. "Collision characteristics
and injuries to motoreyelists and moped
EFFECT OF SLOPE drivers". Twenty first STAPP Car Crash
Conference, Society of Automobile Engineers,
Just occasionally a trajectory will not be Warrendale, Pa, 1977, pp 261 - 301.
on level ground but will be up or down a signi-
ficant slope. There is no analytical solution 2. RAVINI, B., BROUGHAM, D. and MASON, R.T.
for the total trajectory in such cases. However "Pedestrian post-impact kinematics and
a reasonable approximation may be derived by injury patterns". Twenty fifth STAPP Car
noting that the total trajectory is part aerial Crash Conference, Society of Automobile
and part along the ground. Analytical express- Engineers, Warrendale, Pa, 1981, pp 791 -
ions are readily available for the percentages 824.
increase in the velocities required for each
type of motion. At the optimum angle, the 3. BAKER, J.S. "Traffic Accident Investigation
velocity for a given projection distance through Manual". The Traffic Institute, Northwest
the air is increased by a factor:- ern University, 1975, pp 225 - 227.
280
APPENDIX DERIVATION OF THE FORMULA
s = 2. uv
9
The object lands on the plañe, with which it has a coefficient of rebound e and a coefficient
of friction p . Immediately upon landing the object rebounds for a second flight. The vertical
component of the rebound velocity v^ will be e times the previous vertical component v. Furthermore
the horizontal component u^ after rebound will be reduced from the previous valué u by an amount
equal to p (v + v^). This is because v + v^ is the vertical velocity change and the coefficient of
friction is p .
ev
It is by now clear how to construct the equations for the third rebound, which prove to be
3
ev2 = e v
2 3
= u2 - p v 2 (1 + e) = u - p v - 2pev - 2pe v - p e v
3 3 2 . 4 2 , 5 2 6 2
g
2. u 3v 3
= !•{ e vu - p e v - 2^ie v 2ue v - p e v
- p v - 2 p e v - 2pe o 3 v - jje4v
- 2jje
U4 = U3 " / ,v3 (1 + e)
2. u.v.
g 4 4 Q L
4 4 2
e vu - j i e v o 5 v2
2/ue o e 6 v2
- 2u - o 7 v2
2jie
8 2
^ue v
}
The series of bounces continúes theoretically to infinity, but enough has been written to see the
pattern developing. The total distance covered will be
281
2 s = s + s + s2 + s3 + s4
2 „ 2 2 , 3 2 . 4 2
!.i uv + euv + e uv + ■yjev - - 3yue
- 2|ie v ue v - 4pe v
9 L , 5v 2
- 5pe
■•••}
—•
g [1
f uv
- e
_ jaev2
(1 - e ) 2" j
7
Considering now the horizontal component of velocity, it will be seen that its valué at the end of
bouncing will be
2 3
U u - yuv (1 + 2e + 2e + 2e + ...... )
It follows that if u ) u v . 1 + e then U > 0 . This means that after the cessation of bouncing the
r 1 - e
object still has a residual sliding velocity. The additional sliding distance will be
U2 = fu(l-e) - jjv ( 1 + e) 7 2
2pg 2 y iq (1 - e) 2
The derivation of this formula has assumed that the vertical bouncing ceases before or at the
same time as the horizontal motion. This is almost universally true for natural objeets, which
generally have coefficients of rebound less than 0.2. It is not always true for any object especially
designed to bounce easily. The only one of these likely to be encountered in accident investigation
is a loose wheel, which if inflated may have a coefficient of rebound as high as 0.7 or so. If the
formula is applied to such objeets it will tend to under-estimate their velocity by up to 5%.
The derivation also assumes that the principal motion of the object is bouncing and sliding.
Again this is generally true for natural objeets, but not true for some objeets especially designed
to roll. For this reason again calculations on loose wheels may be inaccurate, as may calculations
on vehicles which have landed in their normal direction of motion and then rolled on their wheels.
282
FIGURE 1 UPPER AND LOWER BOUND VALUES AS A FUNCTION OF THE
COEFFICIENT OF FRICTION
SPEED, m / s
I MPACT
0 10 20 30 40
THROW DISTANCE, m
283
-----------------MAX = 3 ■1 3 2 / s
----------------- M IN = 2 • 3 2 A /s
----------------- MEAN = 2 • 7 2 8 / s
284
285
BOUNCING SLIDING