Professional Documents
Culture Documents
RAYMOND M. BRACH
Aerospace and Mechanical Engineering, University of Notre Dame, Notre Dame, IN 46556 U.S.A.
Abstract-Various methods are used to calculate unknown velocities and velocity changes for the collision
of automobiles. Most methods use some form of the equations of impulse and momentum from mechanics.
In this paper the planar impact between two vehicles is modeled through the use of six algebraic equations.
These equations relate the three velocity components of each vehicle at the beginning of contact to the
three velocity components of each vehicle at separation using two coefficients of restitution and an
equivalent coefficient of sliding friction. The first part of the paper contains a discussion of some of the
fundamental assumptions connected with the use of impulse and momentum concepts for collision analysis.
A summary of the derivation of the six model equations is presented. In a second part of the paper, the
equations are solved for the conditions corresponding to two experimental collisions. Values of the three
coefficients are systematically varied to assess their effect on the vehicle velocity changes. A listing of a
computer program is furnished, written in the BASIC language. The program can be used to solve the
model equations for the final velocities given the initial velocities and coefficients. The vehicle properties
and collision geometry must also be known.
NOTATION
The symbols in parentheses are corresponding variable names used in the computer program
listed in Appendix B
e coefficient of restitution.(E)
em impact moment coefficient (EM)
d distance between the center of gravity and impact point in each vehicle (DA, DB)
I moment of inertia of a vehicle about a vertical axis through its center of gravity (IA, ZB)
M impulse of the moment developed between two impacting vehicles
m mass of a vehicle (MA, MB)
P impulse of the resultant force developed between two impacting vehicles
V velocity component of the mass center of each vehicle following impact (A(I, M + I),
I = 1,4)
velocity component of the mass center of each vehicle before impact (UX, VU, VX, VU)
4 speed change (final minus initial) of a vehicle due to impact
CL equivalent coefficient of friction along the impact surface (MU)
I9 heading angle of a vehicle relative to a fixed x axis (TA, TB)
I- angle of impact surface relative to a fixed y axis (GA)
n angular velocity of each vehicle following impact (A(& M + I), I = S,6)
angular velocity of each vehicle before impact (WA, WI)
angle between the length axis of a vehicle and a line between its center of gravity and the
center of impact (PA, PB)
INTRODUCTION
A problem commonly encountered in accident analysis or reconstruction is to determine the
velocity changes of two vehicles caused by their collision. More often than not, these velocity
changes must be calculated from information gathered after the accident has occurred. A
typical approach is to break down each vehicle’s motion into three phases: pre impact, impact
and post impact. Often, these phases are considered in reverse chronological order. Analysis of
the post impact trajectories furnishes estimates of the final velocities of the impact or collision
phase. Then by analyzing the impact, the velocity changes during impact are computed which in
turn provide the final velocities of the pre impact motion.
lllh R. Lt. BRICH
Many different approaches have been taken by various researchers to develop procedures
for accurate solution of this problem or parts of the problem. The most complex and general
method is SM.4C. described by McHenry [1975]. A method named CRASH, McHenry [l97i],
Hess [I9801 and Anon [1981]. has been well developed and gives good results. These and other
methods have at least one thing in common: each contains some form of a mathematical model
of the impact phase. Differences exist in impact models. Some use a lumped mass and flexibility
approach such as Herridge [I9731 and Greene [l977]. Others treat only certain types of
collisions using impulse and momentum; see Emori [1968]. Grime [1969-703 and Emori [1970].
Depending upon the type of collision, CRASH uses various methods such as linear crush
stiffness. residual crush and conservation of linear momentum. One feature of CRASH is that it
accepts damage directly from the standardized Collision Deformation Classification (CDC)
which is described by Nelson [ 19811.
The displacements during and due to crushing are not always needed during accident
reconstruction. When crush displacements are used, they are often related to calculations of
energy loss. Velocity changes however are the primary unknowns. For this reason, and because
of their great simplicity, the equations of impulse and momentum from mechanics have a great
attractiveness for treating the two vehicle collision problem. In this paper, previous work of
Brach [I9771 on the use of these equations is extended. First the concept and range of realistic
values of the coefficients of restitution and intervehicular friction are discussed. Then con-
ditions corresponding to two of the staged collisions in McHenry [I9781 are used to illustrate
the applicability of the equations. Values of the three coefficients are varied systematically
about their experimentally determined values to assess the effect of their changes on the vehicle
velocity changes. This illustrates the sensitivity of the solution to these parameters.
Two appendices are provided. The first presents the equations of impulse and momentum
for the collision of two particles. Because these equations are simpler than those for rigid
bodies, their analytical solution can be found and is included in Appendix A. These equations
will be used to gain some insight into the behavior of example rigid body solutions. A second
appendix contains a computer program written in the BASIC language (for an Apple II
computer) which solves the rigid body impact equations. Final velocities are computed for
given initial velocities, vehicle data and accident geometry.
It must be pointed out that although the work discussed in this paper applies to the field of
road vehicle accident reconstruction, it forms only a part of the problem. Equations and
methods here apply only to the collision phase of a two vehicle accident. Other, independent
methods must be used to analyze the post impact phase, when necessary.
IMPACT MECHANICS
I,
and
P=
I‘I
-F(t)dt=mV(t?)-mV(t,)
These are vector equations where F is force, T is torque, P is the linear impulse, M is the
torque or moment) impulse. V is the mass center velocity and R is angular velocity. The
;\nalysis of planar krhislc :ollirions using equations of impulse and momentum lo-
quantity m corresponds to the total mass of the system of particles and I represents moment of
inertia about the mass center. If each vehicle is assumed to be a collection of interconnected
particles, the above equations can be used to compute the velocity changes due to an impact
beginning at time t, and ending at time tl.
Newton’s third law states that resultant forces (and their impulses) acting on each vehicle
over the contact surface are equal in magnitude and opposite in direction. This must be true
despite results to the contrary from the CRASH program reported by Hess [1980].
In order to apply these equations to a collision between two vehicles. one important
assumption is usually made; namely that the force of the impact between the two vehicles is
large compared to all other forces such as road friction and aerodynamic drag. The assumption
of large internal forces of collision is usually satisfied in practice. An exception would be for
lowspeed collisions.
Another assumption often made is that deformations should be small. Obviously in many
actual collisions, this is not true. Fortunately, the assumption is unnecessary since the above
equations apply to systems of particles whether or not they are rigidly connected. Although m is
a constant, the moments of inertia, I, and I? can differ. The question of deformation does enter
into the problem in another important manner, however. The moment of the resultant impulse
about the center of gravity location of each vehicle must be determined, yet the CC locations
can change because of large deformations. Normally, the moment of impact forces is deter-
mined by estimation of a point usually referred to as the “center of impact”. This will be
discussed later in more detail.
It is frequently assumed that collisions must occur over short time intervals for the
equations of impulse and momentum to be valid. Strictly speaking, the equations apply for any
duration of contact. The time interval of a collision does matter however when the pre-impact
motion, impact motion and post-impact trajectories are collectively considered. To study total
motion it is convenient to assume that during impact, changes in position and orientation of the
vehicles do not occur; only the velocities change. In practice most actual collisions occur over a
duration whose order of magnitude is 0.1 to 0.2 second as noted by Grime [ 1969-701, Emori
[1970] and experimentally by Shoemaker [1978a, b], Davis [I9811 and Anon [1981]. In contrast
to pre- and post-impact motion, this duration is small. However. in some collisions the
displacements within this time interval can be significant.
the end of the impact. Thus, for example, V,, V,, and a, are the three velocity components of
vehicle A at f?, whereas v,,, vOY and w,, are the corresponding values at the beginning of impact,
time f,.
The magnitudes of the impulse components, P,, P, and M also enter the problem. In the
approach to be followed, these unknowns are eliminated from the equations. After a solution is
obtained they can be calculated, if needed.
Equations. After all variables are defined, the equations of impulse and the momentum from
mechanics can be derived. These equations were derived in a previous paper by Brach [1977].
They are listed below in a slightfy more general form.
Three equations are obtained from conservation of momentum in the x direction, y d~ection
Analysis of planar vehicle collisions using equations of impulse and momentum 109
~b(Vbx-L.bx)‘m,(V,,-~,,)=O (1)
md Vb,- L.bJ+ %( v,,- %,I = 0 (2)
and
(3)
For simplification of the above equations, the following substitutions were made:
A fourth equation involves the classical coefficient of restitution e. This equation follows
from the condition that at time I? the relative velocity of the vehicles at point C normal to the
line defined by the angle T is equal to the coefficient of restitution times the relative velocity of
the vehicles at point C at the beginning of impact, time t,. The resulting equation is:
During contact the two vehicles can slide relative to each other along the crush surface, i.e.
along the tangential axis. The tangential impulse component P, is related to the normal impulse
component, P, by the coefficient of intervehicular friction, 1-1.This relationship is
The sign is determined by the direction of the relative velocity at C. If sliding occurs,
47 ( vuy
- u,Jcos r T p sin r)
In some collisions, sliding terminates before separation. When this occurs, eqn 5 does not
apply directly. Instead, an equation may be used which constrains the final tangential velocity
component of vehicle A to equal the final tangential velocity component of vehicle B. This
equation is:
A final equation is obtained through consideration of the moment impulse and the relative
angular velocities of vehicles A and B. This is the subject of a previous paper by Brach [19771
I II) R. 51. BRKH
and results in an equation involving an impulse moment coefficient P,. In the article by Brach
[19771.the moment coefficient is defined through the following relationship
Some of the properties of the moment coefficient are evident from this equation. Note that for
e,,, = I, the impulse moment, M. must be zero. This corresponds to the assumption normally
made by other investigators. When e,,, =O, the two final angular velocities must be equal.
II,: = &,; this is a perfectly inelastic angular collision. Although not evident from this equation,
the general conditions on e, are that e,,, = 1 or -1 5 e,,,I 0. Except for sign, the coeficient for
angular motion is completely analogous to the classical coefficient r. By elimination of the
moment impulse M, the above equation can be written as
The moment coefficient e,” is related both to the rotational energy loss and the choice of
location of the center of impact C, as discussed earlier. Additional discussion of e,,, follows.
IMPACT COEFFICIENTS
Three coefficients appear in the impact equations just presented. The coefficient of restitu-
tion, e, is defined as the negative ratio of the normal rebound velocity to the normal approach
velocity of the parts of the vehicles which form the contact surface (actually the impact
“point”). It is well known that 0 < e zz 1. For a simple impact such as a ball falling toward and
bouncing from a rigid floor, the coefficient can be related directly to the kinetic energy lost due
to impact. A value of e = I corresponds to no energy loss; e = 0 corresponds to aft energy lost.
This is discussed in Johnson [1976]. For more complicated impact geometry such as for
vehicles, the relationship between e and the energy loss is complicated. In all cases, however,
as e approaches its lower bound, zero, the rebound velocities approach zero and the final
kinetic energy of impact is smaller. Consequently, the smaller e, the more energy lost for a
given co!tision con~guration.
Both other coefficients, e, and ,u, enter into the energy loss of a collision.
Certainly the permanent deformation of damaged vehicle parts account for losses of energy
in a collision. No simple relationship between the energy of inelastic structural deformation and
energy loss can be expected however since mechanical interconnection between vehicles (such
as locked bumpers) can cause zero rebound velocities (and zero e). This requires a significant
energy loss without large permanent damage. Energy dissipation due to energy absorbing
bumper systems is another complication. On the other hand, the greater the physical damage,
certainly the greater the energy loss and the smaller the coefficient of restitution.
For vans impacting rigid barriers, Davis and Pierce [ 19811report values of e about 0.1. They
claim these values are similar to passenger car-barrier test values. For head-on and rear-end
impacts, Grime and Jones [1969-701 report values of e in the order of 0.05 to 0.01. Brach f19821
finds values ranging from below 0.1 for oblique, front-to-side impacts and values as high as
0.475 for 90” front-to-side impacts.
manent) interconnection of parts of the vehicles and surface configuration changes during
contact, the Coulomb friction model is only a gross approximation. Consequently, the
coefficient I* as applied to vehicle collisions should be termed an equivalent friction coefficient.
The presence of friction and relative motion tangential to the crush surface dissipates
energy during a collision and causes velocity changes. (These velocity changes for particle
impacts are shown in eqn Al2 and A13.) Consequently, energy losses in a vehicular collision
model are influenced by values of both the coefficient of restitution e and the equivalent
coefficient of friction, P.
Emori [1970] claims that analysis of full scale automobile intersection collisions gave
coefficients of restitution near zero and coefficients of friction in the neighborhood of 0.4. This
agrees somewhat with Brach [1982] who found values of 0.486 and 0.492 for perpendicular
intersection impacts from staged collisions with corresponding restitution coefficients of 0.073
and 0.085. Again for intersection collisions Grime and Jones [ 1969-701calculated coefficients of
friction between 0.4 and 0.6. For oblique impacts (120”, front to side) Brach also found
experimental coefficients of 1.001 and 1.138. This indicates that the equivalent coefficient of
friction depends on the collision geometry.
tThese reports are entitled “Research Input for Computer Simulation of Automobile Collisions”. The acronym
RICSAC is used in these reports and in this paper.
R. M. BRACH
& 60”
a. RICSAG 9 b. RICSAC 7
- RICSAC 7 RICSAC 9
1 Veh A 1 Veh B 1 Veh A 1 Veh B
I- I I I
MME Chevrolet VW Rabbit Honda Ford
RICSAC 7 RICSAC 9
I Veh A 1 Veh B j Veh A 1 Veh B
I I I
V,, m/s(mph1 -7.60 (-17.0) -2.55 (-5.71) -0.73 (-1.64) -3.01 (-6.74)
Vy. m/s (mph) 1.38 I 3.08) 8.73 (19.5) 4.38 (9.80) 7.38 (16.5)
IAVI. m/s (mph) 5.5B (12.48) 9.41 (21.04) 9.78 (21.87) 3.67 (8.22)
)
n rad/s (de /s)
RICSAC 7 RICSAC 9
r Veh A 1 Veh B 1 VehA 1 VehB
I I I
V,. mis (mph1 -7.33 (-16.4) -2.78 (-6.22) -1.50 (-3.35) -3.78 (-8.45)
Vy, m/s (mph) 1.42 (3.18) 8.94 (20.0) 3.88 (8.74) 7.64 117.1)
IbVl, m/s (mph) 5.86 (13.10) 9.57 (21.41) 8.87 (19.84) 4.20 (9.39)
n, rad/s (deg/s) -1.87 f-107) -2.94 (-168) -4.09 (-234) 1.56 189.4)
e .015 .467
em -.774 1.000
Anal>tisof planar vchiclz mllisions using equations of impulse and momentum I I3
velocities and coefficients in Table 3 for each staged collision form a solution of the impact
equations. In fact they are the solutions which are as close as possible to the experimental
values (of Table 2) in a least square sense.
In order to assess the effect of coefficient changes, lVu. lV,, X?,, -lRh and lT will be
compared. These are the changes in velocity components of each vehicle and kinetic energy, i-,
for a given collision.
For example. the solution (velocity changes and energy loss) for collision RICS.AC 7 will be
found for all possible combinations of r = 0.015 + 10%. e, = - 0.771+ IO% and I_I= 0.968 = 10%.
This is done using the computer program listed in Appendis B.
e1 ellll -7.58 (-16.96) 1.06 (2.37) -1.59 (-91) 5.53 (12.37) -2.36 (-5.28) 9.53 (21.32) -2.70 (-155) 9.03 (20.71)
e1 %2 -7.55 (-16.89) 1.07 (2.39) -1.63 I-93) 5.56 (12.44) -2.41 I-5.40) 9.52 (21.30) -2.66 C-152)
e1 eml -7.11 (-15.911 1.78 (3.9A) -2.12 (-121) 6.16 (13.781 -3.13 f-7.01) 8.36 (18.69) -3.20 (-183) 10.07 (72.52)
e1 em2 -7.00 (-15.83) 1.79 (4.011 -2.16 (-124) 6.20 (13.86) -3.19 (-7.131 8.34 (lR.65) -3.17 (-lR2) 10.17 (72.65)
ez em1 -7.57 (-16.93) 1.06 (2.38) -1.60 (-92) 5.54 (12.40) -2.39 (-5.34) 9.53 (21.31) -2.72 (-156)
e2 em2 -7.54 (-16.A6) 1.07 (2.39) -1.63 C-93) 5.58 (12.47) -2.44 (-5.45) 9.52 (21.29) -2.67 (-153)
e2 em1 -7.09 (-15.A7) 1.79 (3.99) -2.12 (-1211 6.1A (13.82) -3.16 (-7.071 8.35 (18.67) -3.21 f-184) 10.10 (22.59)
e2 %2 -7.06 (-15.79) 1.80 (4.02) -2.16 (-124) 6.21 (13.90) -3.22 (-7.201 8.33 (18.64) -3.18 (-182) 10.15 (22.71)
- - _. - _ -_ - .^ _ ~- .- .
T:lble 5. Solutions of impact eqwtions for collision of RICSAC 9
I
el = .420 ,'l = .437 a em= 1.0
e2 = .514 p2 = .535
’ e II “ax ",Y
' na = hrl, lAvaI "bx
, “by i-lb= Anb lA"bl IhTl *
e1 Pl -1.80 f-4.03) 3.35 (7.50) -3.54 C-203) R.37 (18.73) -3.63 (-8.131 7.89 (17.65) 1.56 (89) 3.96' (8.87) 27.6
el 1'2 -1.71 (-3.A2) 4.16 (9.30) -4.3R (-251) 8.Rl (19.70) -3.68 (-8.22) 7.51 (16.80) 1.45 (831 4.17 (9.331 27.1
e;, Ul -1.30 f-2.90) 3.57 (8.00) -3.77 (-2161 8.93 (19.97) -3.R7 (-8.66) 7.78 (17.41) 1.66 (95) 4.23 (9.45) 26.1
e? "2 -1.19 (-2.67) 4.43 (9.91) -4.67 (-268) 9.39 (2!.01) -3.92 (-8.77) 7.38 (16.51) 1.55 (89) 4.45 (9.951 25.1 .
R. ht. BRACH
a. RICSAC 7
m-al I’;v,l
.I
e = eZ 5.88 (13.15) 9.61 (21.49) 2.95 (169)
b. RICSAC 9
e = el 8.59 (19.22)
e = e* 9.16 (20.49)
” = “1 8.65 (19.35)
” = “2 9.10 (20.361
_I
26.1
results will be unrealistic. Since this would force P, to be about l/2 of P,. Both vehicles would
then have significant final y velocity components, a condition which can only result if the
system is capable of creating energy. For this hypothetical example, a realistic solution would
be for p = 0 (or through use of the velocity constraint, eqn 5(a)). This hypothetical example
leads to an important conclusion.
1. The appropriate value of the equivalent friction coefficient can depend upon the initial
velocities and the accident configuration.
Furthermore, increasing the coefficient of friction in the solution for a given collision (such as
RICSAC 7) increases !he energy loss due to the collision. However. increasing the friction
coefficient beyond some limiting value, as seen above, can fictitiously create energy in the solution,
thus decreasing the energy loss due to impact. This leads to a second conclusion:
2. A limiting value of the equivalent coefficient of friction exists for a given collision at
which AT begins to decrease as p is increased in the solutions of eqn I. through eqn 6.
Although not specifically demonstrated, the solution of the impact equations for the limiting
value of p will be the same as obtained with the constraint, eqn 5(a).
All of the above can be applied to the results given in Table 6(b) for RICSAC 9. Note that
for ).L= F? = 0.535. the energy loss, AT = 26.1. For p = ,LL,= 0.437, AT = 26.9. This means that
increasing p above the least square value of 0.486 creates a physically unrealistic situation. It
also implies that in RICSAC 9 at separation, both vehicles had near common velocity
components in the tangential (y) direction.
CONCLUSIONS
The use of the equations of impulse and momentum for modeling planar vehicle collisions
was reviewed. The form of the equations presented contained three coefficients. It was seen
that for collisions where the coefficient of restitution, e, is small compared to 1, the solution of
the impact equations is not very sensitive to variations in the value of e. This implies that in
accident reconstruction, where much damage occurs (e is small), the accuracy of the estimate
of e is not critical.
Analysis of planar vehicle collisions using equations of impulse and momentum II-
A second coefficient was reviewed, namely the equivalent coefficient of friction, It was
shown that this coefficient can have a maximum permissible value which can depend upon both
the accident configuration and the initial velocities. Furthermore, for at least some accident
configurations, variations of the friction coefficient significantly affect the velocity changes of
the vehicles in the equation solution. This would indicate that the friction coehicient must be
chosen carefully in some accident reconstructions.
A third coefficient was reviewed, called the moment coefficient of restitution. This
coefficient is a relatively recent concept and not much corresponding data has appeared in the
literature. Previous investigators’ use of impulse and momentum equations without this
coefficient corresponded to choosing a value I. This is equivalent to assuming that no moment
impulse exists between the two vehicles during impact and that the point of application of the
resultant impulse (center of impact) is known accurately for each vehicle. More use of this
coefficient for accident analysis and reconstruction should reveal some typical values encoun-
tered.
REFERENCES
Anon., CRASH 3 User’s Guide and Technical Manual. U.S. DOT Hs 805 732, Feb. 1981.
Bhushan Bharat, Analysis of Automobile Collisions. SAE Paper 750895.
Brach R. M., An Impact Moment Coefficient for Vehicle Collision Analysis, Transacfions SAE 770014. 1977.
Brach R. M.. Nonlinear Parameter Estimation of a Vehicle Collision Model, l3rh .-lnnuul Conf. on Modeling and
Simulafion, University of Pittsburgh, April 1982.
Davis S. and Pierce S., Van Crashworthiness and Aggressivity Study. SAE 810090. 1981.
Emori R. I., Analytical Approach to Automobile Collisions. SAE 680016, 1968.
Emori R. I., Vehicle Mechanics of Intersection Collision Impact. SAE 700177, 1970.
Greene J. E.. Comouter Simulations of Car-to-Car Collisions. Proc. SAE Trclnsuctions. 770015. 1977
Grime G. and Jones I. S.. Car-collisions-the movement of cars and their occupants in accidents, Proc. Insfn. Mech. Engrs.
184, Part !a, (S), 1969-70.
Herridge J. T. and Mitchell R. K., Development of a Computer Simulation Program for Colinear Car/Car and Car/Barrier
Collisions, ASME Paper 73-JCT-334.
Hess R. K., Microcomputer Programs Useful for Aspects of Accident Reconstruction, SAE 800174. 1980.
Johnson W., Simple linear impact, Inr. /. Mech. Engrg. Educ. 4. (2) 1976.
Jones I. S., Automated Accident Reconstruction, SAE Paper 758894.
Jones I. S. and Baum A. S.. Research Input for Computer Simulation of Automobile Collisions. Vol. 4, U.S. DOT HS 805
040, NTIS, 1978.
McHenry R. R., A Comparison of Results Obtained with Different Techniques for Reconstruction of Highway Accidents,
SAE Paper 750893.
McHenry R. R., Lynch J. P., and Segal, D. J., Research Input for Computer Simulation of Automobile Collisions. Vol. I.
U.S. DOT HS 805 037. NTIS, 1978.
McHenry R. R. and Jones I. S., Evaluation of the Accuracy of Reconstruction Techniques for Highway Accidents, Calspan
Report ZQ-5708-V-4, DOT HS 802 106, 1975.
Nelson W. D.. The History and Evolution of the Collision Deformation Classification S;\E J221 1l.AR80. SAE Paper
810213, 1981.
Shoemaker N. E., Research Input for Computer Simulation of Automobile Collisions, Vol. II, U.S. Dot HS 805 038, NTIS,
1978.
Shoemaker N. E., Research Input for Computer Simulation of Automobile Collisions. Vol. III, U.S. Dot HS 805 039, NTIS,
1978.
Todd J. (Ed.), Survey of Numerical Analysis. McGraw-Hill, New York, 1962.
.4PPENDIX A
Using the concept of coefficient of restitution, e, the relative normal velocities are related by:
At separation, either the tangential velocities of the two particles are the same (no sliding) or they differ (sliding); thus:
The last expression comes from the fact that if sliding occurs uith a friction coerfis~ent F. then the change in momentum in
the tangential direction must equal p time-he change in momentum in the normal direction. The sign ij determined by the
direction of the relative velocities.
It is convenient to define ncu variables corresponding tu the velocity changes. Thus let
Equation AZ and A3 can be solved for the normal velocity changes. These are:
It should be noted that eqn Al. AZ. A3 and A4 or A5 form a set of 4 linear equations in the velocities. As done above.
they can be solved for the velocity changes due to impact assuming the initial velocities are known. It is also possible to
solve these equations for the velocity changes in terms of the final velocities. This is frequently convenient for
reconstruction purposes when the final impact velocities can be estimated from the accident data.
APPENDIX B
Compuler solution oJ impact equations
A listing of a computer program is furnished which gives the solution of eqn I through 6. These equations are treated as
a set of six simultaneous linear algebraic equations. It is assumed that all of the physical data for the vehicles are known
(such as weight, dimensions, inertia, etc.). It is also assumed that the collision geometry and the coefficients are known.
The program computes the final velocity components of both vehicles for any set of initial velocity components.
The program is written in the BASIC language for use on an Apple II Plus minicomputer with 4SK memory. The
example solutions presented in the paper were obtained through the use of this program. It is a self contained program and
includes a subroutine for solving linear simultaneous algebraic equations using pivotal elimination as discussed in Todd
[1962].
Lflirs. The program is independent of the units of the input data. ,A consistent set of units should be followed.
Xn exception is the units of the angles uhich define the collision geometry. MI angles are read in and printed out in
decrees. The program converts them to radians uhere appropriate. Note that angular relocities alhays have units of
radians/second.
Point mars solution. The program uill furnish the proper solution for the impact problem of point masses. If this is
desired. two actions should be taken by the user. Variable M at statement 1900 should be set equal to -1 (M = 4) and the
vehicle parameters (I,, tD.4) and db (DE) must be input as zeroes.
Fig. I-.\