316

3
THE STABILITY AND CONTROL OF MOTORCYCLES
By R.S. Sharp*
Mathematical models of a motorcycle and rider dependent on three alternative assump-
tions concerning the tyre behaviour are developed. Stability characteristics deduced from
them are compared, and minimum requirements for the model greater than have been
previously satisfied are established. Using the most sophisticated of the models, the effects
of design changes are calculated, and the design implications are discussed.

INTRODUCTION the use of such a model to show typical stability charac-

THEORETICAL STUDIES of the steering behaviour of auto- teristics and how these characteristics depend on various
mobiles after Segel (I)thave transformed understanding parameter values. Also, in the light of Segel’s conclusion
of the subject and provided valuable design information (6) that the steady state control force gradient is an im-
and design evaluation techniques. Parallel studies relating portant parameter in determining car handling quality,
to two wheeled vehicles (2)-(5) have been comparatively and his statement that this is the case with aircraft too, the
unsuccessful, and motorcycle design at present relies on steady state behaviour has been studied. The model is
experience rather than understanding for its success. easily applied to the fixed control case and appropriate
A viable theory of motorcycle stability and control must results are included.
be consistent with the practically obvious instability of the A particular difficulty in the analysis concerns the
machine at very low speeds, with the possibility of ‘hands treatment of the tyre behaviour. Whipple (2), Pearsall (3),
off’ control (i.e. control without the direct application of and Collins (5) allowed the tyres no sideslip freedom,
steering torque) at moderate speeds, and with the possible while Kondo, Nagaoka and Yoshimura (4) allowed this
ociurrence of oscillatory instabilities at high speeds. These freedom, but ignored the lag between the steering of a
were described by Pearsall (3) as ‘speedman’s wobble’. It pneumatic tyred wheel and the building up of the side
should also provide some explanation for current design force towards a steady state value, which is of great im-
practice, mindful that it should be converging on some portance in the wheel shimmy phenomenon (8). Results
evolutionary optimum, and ideally it should support the dependent on each of the three alternative assumptions
empirical maxims of motorcycle design (e.g. the mass regarding the tyre behaviour are included and compared.
centre should be as low as possible and as far forward as
possible).
In the case of the car, Segel (6) has observed that
different drivers employ different proportions of fixed and Coefficient matrices.
free control, and that particular drivers can alter these
proportions to suit the car which they are driving. In the Linear dimensions (Fig. 1).
case of the motorcycle, the free control behaviour would
appear to be relatively much more important, since the Front and rear tyre cornering stiffnesses re-
very small steer angles normally employed (7)would make spectively.
fixed control difficult, and since in the ‘hands off’ condi- Front and rear tyre camber stiffnesses re-
tion the fixed control characteristics of the machine have spectively.
no relevance to the motion. Rear frame product of inertia with respect to
Consequently, the work which is the subject of this above axes.
paper was directed towards the development of a suitable Acceleration due to gravity.
mathematical model of a motorcycle in free control, and Front frame inertias about axes parallel to
OX4Y4Z4 (Fig. 2) through mass centre
The M S . of this paper was received at the Institution on 10th August (assumed to be principal axes).
1970 and accepted for publication on 17th M a y 1971. 2 Rear frame inertias about axes parallel to
* Lecturer, Department of Mechanical Engineering, Universitj! of O X , Y2Z2(Fig. 2) through mass centre.
Leeds, Leeds, LS2 93T.
t References are given in Appendix 5 . i Polar moment of inertia of engine flywheel.
JOURNAL MECHANICAL E N G I N E E R I N G SCIENCE Vol I3 No 5 1971
 T H E STABILITY AND CONTROL OF MOTORCYCLES 317

Polar moment of inertia of front wheel. (1) The vehicle consists of two rigid frames joined at
Camber inertia of rear wheel. the steering axis with freedom, restrained by a linear
Polar moment of inertia of rear wheel. steering damper, of the front frame to steer relative to the
Steering damper coefficient. rear one.
Mass of front frame. (2) The front frame consists of the front wheel, forks,
handlebars and fittings.
Mass of rear frame.
( 3 ) The rear frame consists of the main structure, the
Angular velocity components of front frame engine-gearbox assembly, the petrol tank, seat, rear forks,
with respect to axes O X 4 Y 4 Z 4 . rear wheel, etc., with rigidly attached rider.
Angular velocity components of rear frame (4) Each frame has a longitudinal plane of symmetry,
with respect to axes OX,Y,Z,. and the axis through the front frame mass centre parallel
Generalized co-ordinate and force respec- to the steering axis is a principal one.
tively. (5) The road wheels are rigid discs each making point
Components of forces Qxo and QYoalong contact with the road, and they roll without longitudinal
O X , and O Y , respectively. slip on a flat level road surface.
Yawing velocity (= $). (6) The axis of rotation of the engine flywheel is trans-
Sum of T I ,T,, and T,, being the total system verse.
kinetic energy. (7) The machine moves at constant forward speed with
Kinetic energies of front and rear frames freedom to sideslip, to yaw, and to roll; only small per-
(excluding wheel rotation) respectively. turbations from straight line running are considered.
Extra kinetic energy accounting for wheel (8) The air through which the machine moves is sta-
rotation. tionary so that the effects of aerodynamic side force,
Potential energies of front frame, rear frame, yawing moment, and rolling moment, will be small com-
and total system respectively. pared with tyre effects. The effects of drag, lift and pitch-
Lateral velocity of 0 (= ing moment are to modify the vertical loading of the
tyres, and to make necessary a longitudinal force at the
(XI Column matrix of amplitudes X,, X,, etc. driving wheel sufficient to maintain the constant forward
X I , Y 2 Forces applied at front and rear tyre to road speed assumed. These effects are accounted for by varia-
X,, Y,, I, ZrI> contact points respectively (Fig. 5). tions in the coefficients relating tyre side forces to sideslip
X I , X, Lateral velocity and yawing velocity ampli- and camber angles.
tudes respectively. (9) Pneumatic trail of the tyres is not considered since,
x,, ko, yo,yo Co-ordinates and velocities of reference for the rear tyre, its effect will be very small, and for the
point A in O X , YoZosystem. front tyre, it is small compared with the mechanical trail,
il,xl,yl,yl Velocities and accelerations of reference and the effects of varying the mechanical trail itself are
point A in OX,Y,Z, system. included in the results.
Y’,, Y’, Steady state lateral tyre forces. (10) The drag force at the front tyre is small compared
UIY c(r Front and rear tyre sideslip angles respec- with the tyre side forces.
tively.
Y Effective front wheel steer angle (Fig. 6). The assumptions that the motorcycle with rigid wheels
A Steering velocity (= 6). moves on a flat road surface and that is has no freedom to
E Steering head angle (Fig. 1). pitch are, strictly speaking, incompatible. As the handle-
5 Velocity of front frame mass centre, GI. bars are turned, the front tyre to road contact point in
7 Velocity of rear frame mass centre, G,. general moves vertically and the problem is, in fact, treated
01, o r Front and rear wheel angular position co- as if the road surface moved vertically too to maintain
ordinates respectively. contact. Bearing in mind the smallness of these vertical
A Gear ratio between rear wheel and engine motions in normal running, and the fact that the tyre and
flywheel. suspension flexibilities and road irregularities have been
CL Exponential coefficient. ignored, this is considered to be a negligible extension of
0 1 , ur Front and rear tyre relaxation lengths re- the above assumptions.
spectively. The motorcycle is represented diagrammatically in
7 Rider applied steering torque. Fig. 1 with the point A defined as the intersection of the
@ Rolling velocity (=4). vertical transverse plane containing the rear frame mass
41 Front wheel camber angle (Fig. 6). centre, the longitudinal plane of symmetry, and the ground
$44, 6 Angular displacements (Fig. 2). plane. A is the origin of a right-handed, orthogonal axis
set AX,Y,Z, which moves with the vehicle, and which,
PHYSICAL DESCRIPTION OF MODEL with the vehicle in the upright datum condition, has A X ,
The following assumptions are made regarding the repre- horizontal and pointing forwards, A Y , horizontal and to
sentation of the vehicle. starboard, and A Z , vertically downwards.
JOURNAL MECHANICAL ENGINEERING SCIENCE V o l 1 3 No 5 1971
318 R. S. SHARP

/ Steer axis

Fig. 1. Diagrammatic representation of the motorcycle

The front frame mass centre G, is the origin for another tudinal motion of the motorcycle, and for equations
right-handed orthogonal axis set G,X4Y,Z,, this moving describing the rotations of the wheels, disappears through
with the front frame. With the vehicle in its datum condi- the constant forward speed and no longitudinal tyre slip
tion, G,X, lies in the longitudinal plane of symmetry and assumptions; the latter also allow the elimination of 8, and
is normal to the steering axis, G, Y , is parallel to A Y,, and Or from the other equations.
C,Z, is parallel to the steering axis (pointing downwards). The generalized forces are derived in terms of the tyre
The motion of the motorcycle is referred to inertial axes forces. The tyre forces are then obtained in terms of the
O X oYoZ, by the co-ordinates xo, y o of A , and the angular vehicle motion parameters, through expressions for the
displacements $,4,and 6 shown in Fig. 2, 0 being fixed tyre sideslip and camber angles. The consideration of
in the road surface.
Starting with the reference axes O X , YoZ,, the orienta-
tions of the other axis sets are derived as follows. A
rotation $I about OZ, gives OX,Y,Z,. Then a rotation 4
about O X , gives O X , Y,Z,. Next a rotation E about 0 Y 2
gives OX,Y,Z,, and finally a rotation 6 about OZ, gives
ox, Y4Z4.
E Q U A T I O N S OF M O T I O N
The equations of motion are derived by the application of
Lagrange's equation (9)
LIT iV
--+T-
ilq
dc ti ( Li qT )
= Q, . . (1)
Fig. 2. Axis systems, angular displacements and velocities
cq
The position of the vehicle is defined relative to the
inertial axes OXoYoZoby the co-ordinates xo, y o of A,
the yaw angle $, and the roll angle 4, which define the rear
frame, the steer angle 6, which defines the front frame
relative to the rear one, and the angles 8, and Or which
define the rotational positions of the road wheels and the
engine flywheel. The latter are shown in Figs 3 and 4.
Expressions for the kinetic and potential energies in terms
of these seven co-ordinates are first obtained and then A
utilized to give equations for the lateral, yawing, rolling, b
and steering freedoms of the machine. Fig. 3. Reference point A and rear wheel looking
The requirement for an equation describing the longi- along O Y ,
JOURNAL MECHANICAL ENGINEERING SCIENCE Vol 13 No 5 1971
 THE STABILITY AND CONTROL OF MOTORCYCLES 319

/Steer axis Premultiplying by (B)- and rearranging,

- ( B ) - W ( X ) = P(X)
and for non-trivial solutions for ( X ) , possible values of p
must be eigenvalues of -(B)-l(A) (12). The real parts of
the eigenvalues indicate the damping and the imaginary
parts the natural frequencies associated with the various
normal modes.
A numerical procedure was therefore employed for cal-
culating the elements of the ( A ) and ( B ) matrices above
from the design parameters for a particular machine and
Fig. 4. Front frame in datum condition looking from its forward speed, forming -(B)-l(A), and finding
along O Y , its eigenvalues through a procedure employing the Q R
algorithm (13). The linear steady state equations were
solved by choosing a forward speed and roll angle and
small perturbations from straight running allows the solving simultaneously for y,, 4, and 6, and then using
elimination of all second and higher order terms; the the equation for the steering freedom to give T.
final equations of motion are linear.
The equations describing steady turning are obtained
by elimination of the time varying terms, while the fixed RESULTS
control equations are obtained by omission of the steering The results fall into one of four categories which cover the
degree of freedom. Details of the derivations are given in following points.
Appendix 1, while in Appendix 2 it is shown how the (1) Natural frequencies and damping characteristics
analysis can be restricted by the omission of tyre sideslip. of one particular machine in free control according to
the model which excludes tyre sideslip, the model
Solution of the equations which includes tyre sideslip but assumes instant tyre
The linear differential equations of motion in homo- response and the model which includes the dynamics of
geneous form, i.e. with T = 0, were reduced to a set of tyre response.
first order equations by the introduction of new variables (2) The effects of changes to the design of the
'u = j,,r = 6, 4
@ = and d = 8. Then, assuming solu- machine in the model which includes the dynamics of
tions of the form v = X,cut, r = X 2 epf, etc. (11) and the tyre response.
substituting these solutions back into the original equa- ( 3 ) The steady turning behaviour (at low lateral
tions, a set of algebraic equations, which can be expressed acceleration) as a function of forward speed and
in the matrix form ( B ) p ( X ) + ( A ) ( X ) = (0) results. Here, machine design.
(B)and ( A ) are square coefficient matrices, while (X)is a (4) Natural frequencies and damping characteristics
column matrix of the unknown amplitudes X , , X,,etc. of the motorcycle in fixed control.

-1I - 39.
-I2 -
-13-
-14 -
-15-
-16- i
320 R. S. SHARP

Category (1) results are shown in Figs 5-7 in which increases above 15 ft/s approximately. It becomes very
damping factors are plotted against forward speed, with mildly unstable above about 35 ft/s with maximum diver-
the circular frequencies of any oscillatory modes shown gence rate near 60 ft/s, decreasing thereafter as the speed
against the appropriate parts of the curves. Category (2) increases. The weave mode has a natural frequency in-
results are shown in Tables 1,2, and 3, each of these tables creasing from about 0.2 Hz at 5 ft/s forward speed to about
being concerned with one mode of physical significance. 3.4 Hz at 160 ftjs forward speed, is unstable up to about
Table 1 concerns a non-oscillatory mode in which, when 20 ft/s, well damped in the medium speed range, and is
it is unstable, the motion of the motorcycle is like that of a moderately damped at high speed. The wobble mode has a
capsizing ship. For ease of reference, this will be called natural frequency which is almost independent of forward
the 'capsize' mode, and Tables 2 and 3 concern oscillatory speed and is about 9 Hz. It is well damped at low and
modes which will be called the 'weave' and 'wobble' medium speeds but only moderately so at high speed.
modes respectively. Other modes are heavily damped and This degree of damping is strongly dependent on the
therefore physically unimportant and detailed results re- value taken for the relaxation length, u (Table 3), and that
ferring to them are not included. Category ( 3 ) results are of the weave mode is somewhat so (Table 2), and, as
shown in Table 4, while those in category (4) are given in would be expected, as u += 0 the results shown in Fig. 5
Fig. 8. become more nearly those of Fig. 6 .
Fig. 6 shows a capsize mode virtually identical with that
D I S C U S S I O N OF R E S U L T S of Fig. 5 but also shows significantly less damping of the
Results from the full analysis with the dynamics of the weave mode at high speed, and markedly different damp-
tyre response included (Fig. 5 ) show that the motorcycle ing of the wobble mode throughout the speed range. The
has three physically significant modes, which are referred natural frequencies from the models with instant tyre
to here as the capsize, weave and wobble modes. The response and proper dynamic tyre response are similar.
capsize mode is non-oscillatory; at low speed it is well Fig. 7 shows that the no tyre sideslip model gives
damped but has rapidly decreasing damping as the speed representations of the capsize mode and, at low speeds

Table 1. Capsize mode damping coefficients

Model Speed, ft/s

number
10 20 30 50 70 100 130 160
~ -___
1 -3.56 -2.83 -0.13 0.08 0.08 0.07 0.05 0.04
2 -3.57 -2.61 -0.12 0.08 0.08 0.07 0.05 0.04
3 -3.56 -3.03 -0.14 0.08 008 0.07 0.05 0.04
4 -3.55 -2.80 -0.13 0.08 0.08 0.07 0.05 0.04
5 -3.59 -2.87 -0.13 0.08 0.08 0.07 0.05 0.04
6 -3.81 -4.55 -0.26 -0.01 0.01 0.02 0.01 0.01
7 -3.51 -3.57 -0.44 -0.01 0.03 0.03 0.03 0.02
8 -3.54 -2.27 -0.10 0.07 0.07 0.06 0.05 0.04
9 -3.58 -3.15 -0.04 0.12 0.11 0.08 0.07 0.06
10 -3.54 -1.11 0.05 0.09 0.08 0.06 0.04 0.04
11 -3.57 -2.54 -0.10 0.03 0.04 0.03 0.03 0.02
12 -3.57 -3.11 -0.22 -0.01 0.01 0.01 0.01 0.01
13 -3.54 -1.75 0.10 0.16 0.13 0.10 0.08 0.06
14 -3.56 -3.26 -0.31 0.02 O W 0.04 0.03 0.03
15 -3.55 -2.35 0.15 0.23 0.19 0.14 0.11 0.09
16 -3.56 -3.09 -0.16 0.14 0.13 0.11 0.09 0.07
17 -3.56 -3.41 -0.44 0.06 0.08 0.07 0.06 0.05
18 -3.60 -3.90 -1.89 -0.07 0.01 0.02 0.02 0.02
19 -3.50 -0.21 0.15 0.15 0.12 0.09 0.07 0.06
20 -3.56 -2.66 -0.26 0.00 0.03 0.03 0.02 0.02
21 -3.50 -041 0.01 0.05 0.04 0.03 0.03 0.02
22 -3.56 -2.89 -0.15 0.06 0.06 0.05 0.04 0.04
23 -3.55 -3.62 -0.63 0.03 0.07 0.06 0.05 0.05
24 -3.37 -3.08 -0.13 0.11 0.11 0.08 0.07 0.06
25 -3.54 -3.20 -0.22 0.08 0.08 0.07 0.06 0.05
26 -3.56 -2.79 -0.13 0.08 0.08 0.07 0.05 0.04
27 -3.57 -2.91 -0.18 0.06 0.07 0.05 0.04 0.04
28 -3.59 -2.88 -0.13 0.08 008 007 0.05 0.04
29 -3.52 -2.63 -0.08 0.10 0.10 0.08 0.06 0.05
30 -3.64 -2.63 -0.13 0.08 0.08 0.07 0.05 0.04
31 -3.56 -2.77 -0.13 0.08 0.08 0.07 0.05 0.04
32 -3.56 -2.74 -0.13 0.08 0.08 0.07 0.05 0.04
33 -3.56 -2.81 -0.13 0.08 0.08 0.07 0.05 0.04
34 -3.56 -2.85 -0.13 0.08 0.08 0.07 0.05 0.04

(The key to the model numbers is given in Appendix 4.)

JOURNAL MECHANICAL ENGINEERING SCIENCE Vol13 No 5 1971
 THE STABILITY AND CONTROL OF MOTORCYCLES 32 1

Table 2. Weave mode damping coefficients and natural frequencies

Model Speed, ft/s

number -
10 20 30 50 70 100 130 160
1 2.01 1.63 -0.18 2.09 -3.36 4.56 -6.59 11.4 -3.97 16.8 -1.82 19.2 -1-13 20.4 -0.96 21.2
2 2.10 1.95 -0.50 2.30 -3.75 4.85 -7.20 11.7 -4.40 17.1 -2.23 19.2 -1.49 20.3 -1.25 21.0
3 1.93 1.19 0-18 1.81 -2.87. 4.10 -5.77 10.8 -3.37 16.4 -1.16 19.0 -0.48 20.5 -0.40 21.4
4 2.01 1.63 -0.17 2.09 -3.29 4.57 -5.99 11.1 -4.29 16.1 -2.36 18.9 -1.65 20.3 -1.44 21.2
5 2.02 1.64 -0.20 2.09 -3.47 4.52 -8.32 11.6 -2.98 17.9 -0.88 19.5 -0-32 20.5 -0-24 21.2
6 1.66 1.70 -0.50 1.71 -6.39 2.34 -4.18 15.0 -2.22 16.1 -1.46 17.1 -1.45 17.7 -1.67 18.2
7 1.98 1-78 0.04 2.11 -3.70 3.14 -10.6 7*3! -7.75 18.8 -4.54 20.7 -3.63 21.6 -3.38 22.1
8 1.95 1.49 -0.45 2.27 -3.51 544 -6.29 14.0 -3.17 18.9 -1.64 21.3 -1-50 22.3 -1.69 22.7
9 1.91 1.51 -0.12 1.75 -3.52 3.90 -7.59 10.2 -4.66 16.7 -2.05 19.1 -1.13 20.4 -0.82 21.2
10 1.58 0.95 -1.05 1.67 -2.98 5.16 -4.33 11.8 -2.21 16.2 -0.33 18.9 0-27 20.5 0.33 21.5
11 1.63 1.28 -0.50 1.70 -3.32 4'54 -5.73 11.2 -3.26 16.2 -1.08 18.6 -0.29 20.0 -0.05 20.9
12 1.61 1.40 -0.35 1.73 -3.54 4.02 -6.89 10.7 -3.96 16.2 -1.57 18.4 -0.68 19.6 -0.35 20.4
13 1.87 1.29 -0.63 1.82 -3.14 5.17 -5.09 12.1 -2.91 16.8 -1.14 19.5 -0.71 20.9 -0.77 21.7
14 2.06 1.78 -0.03 2.19 -3.49 4.05 -7.86 10.8 -4.66 16.9 -2.27 19.1 -1-45 20.2 -1.18 20.8
15 2.17 1.49 -0.20 1.98 -3.30 5.00 -6.26 12.2 -3.93 17.5 -2.28 20.0 -1.99 21.0 -2.07 21.5
16 2.40 1.82 0.18 2.35 -3.36 4.39 -7.84 11.3 -4.95 17.5 -2.79 19.8 -2.18 20.7 -2.04 21.3
17 2.51 1.97 0.35 2.51 -3.34 3.92 -9.15 10.4 -5.57 17.7 -3.15 19.7 -2.40 20.6 -2.15 21.1
18 2.27 1.66 0.72 2.01 -1.92 1.67 -6.95 6.7f -540 14.6 -2.12 17.3 -0.77 18.7 -0.15 19.7
19 1.61 1.65 -2.28 3.51 -4.82 7.20 -6.48 16.1 -3.95 19.5 -2.70 20.9 -2.28 21.3 -2.07 21.5
20 1.99 1.63 -0.19 2.34 -3.16 5.00 -6.06 12.3 -3.50 17.7 -1.68 20.6 -1.41 22.3 -1.75 23.6
21 1.37 1-10 -1.85 2.74 -3.74 6.26 -4.93 14.3 -2.18 18.5 -048 20.9 -0.07 22.1 -0.07 22.8
22 1-89 1.62 -0.28 1.99 -3-51 4.39 -7.03 11.1 -4.21 16.9 -1.88 19.2 -1.07 20.4 -0.82 21.2
23 2-48 1.93 0.49 2.36 -3.02 3.17 -8.33 8.6: -5.74 16.2 -2.98 18.4 -2.00 19.4 -1.61 20.1
24 2.01 1.55 0.16 1.81 -2.91 342 -6.30 8Q -5.82 15.2 -2.48 18.5 -1.03 20.0 -0.34 21.0
25 2.07 1.62 0.11 2.05 -3.24 3.88 -7.36 9.8f -5.24 16.6 -2.72 19.1 -1.90 20.3 -1.66 21.0
26 2.01 1.64 -0.20 2.10 -3.36 4.60 -6.59 11.5 -3.88 17.0 -1.76 19.3 -1.09
~ .. 20.5
~. - -0.94
~.~2 1.2
~~-
27 2.03 1-65 -0.12 2.14 -3.31 4.53 -6.54 11.4 -3.95 16.8 -1.80 19.2 -1.12 20.4 -0.96 21.2
28 2.01 1.63 -0.19 2.09 -344 4.65 -5.42 12.2 -2.75 15.8 -1.25 17.7 -0.85 18.7 -0435 19.3
29 2.01 1.63 -0.23 2.09 -3.30 4.65 -6.34 11.4 -3.87 16.7 -1.76 19.1 -1.09 20.4 -0.93 21.1
30 2.11 1.63 -0.14 2.25 -3.09 4.95 -5.83 11.4 -4.16 17.0 -2.00 20.0 -1.27 21.3 -1.10 22.2
31 2.02 1-69 -0.27 2.15 -3.56 4.68 -7.38 11.9 -4.29 17.9 -1.99 20.3 -1.30 21.5 -1-16 22.3
32 2.03 1.65 -0.21 2.15 -3.34 4.72 -6.20 11.7 -3.63 16.6 -1.71 18.9 -1.12 20.0 -1.01 20.7
33 2.02 1.63 -0.18 2.10 -3.35 4.58 -6.55 11.4 -3.96 16.8 -142 19.2 -1.13 20.4 -0.97 21.2
34 2.00 1.66 -0.21 2.08 -344 4.50 -6.82 11.3 -4.13 16.8 -1.92 19.2 -1.18 20.4 -0.98 21.1

(In each column, the first number is the damping coefficient, the second the circular frequency.)
(The key to the model numbers is given in Appendix 4.)

Unstable

I
60
SPEED-
I
80
ft/s
100
I 1
f 20

19 7
-
I
140

20 5
T
160

-9
-10
-1 I

-13.-
- 14

-16

Fig. 6 . Stability and natural frequencies of standard machine as a function of forward

speed with tyre sideslip but with instant response
JOURNAL MECHANICAL ENGINEERING SCIENCE Vol13 No 5 1971
322 R. S. SHARP

Table 3. Wobble mode damping coefficients and natural frequencies

Model Speed, ft/s

number
10 20 30 50 70 100 130 160
1 -5.20 55.7 -5.91 55.8 -6.42 55.5 -6.52 54.6 -5.70 54.0 -3.94 54.3 -2.24 55.6 -0.82 57.5
2 -2.43 55.8 -2.90 55.5 -3.08 55.0 -2.46 54.0 -1.18 53.7 0.85 54.4 2.58 56.0 3.94 58.2
3 -9.46 54.4 -10.6 55.2 -11.6 55.4 -12.8 54.8 -12.9 54.0 -11.8 53.5 -10.2 54.2 -8.80 55.7
4 -11.6 57.3 -23.1 52.4 -34.6 42.9 -20.4 58.6 -15.1 55.1 -10.0 54.6 -6.71 55.9 -4.36 57.9
5 -2.47 42.0 -1.22 43.3 -0.42 44.6 0.51 46.7 1.11 48.4 1.90 50.6 2.64 52.9 3.30 55.4
6 -5.26 55.3 -5.95 55.4 -6.42 55.1 -6.45 54.3 -5.61 53.8 -3.89 54.4 -2.31 56.0 -1.05 58.3
7 -7.76 62.9 -8.33 63.4 -8.82 63.5 -9.20 63.1 -8.72 62.6 -7.11 62.6 -5.27 63.5 -3.61 65.3
8 -5.60 46.9 -7.10 47.4 -8.30 47.2 -9.15 45.7 -8.29 44.5 -5.88 44.4 -3.53 46.0 -1.77 48.6
9 -5.11 54.2 -5.55 54.5 -5.88 54.4 -5.95 54.0 -5.37 53.8 -4.07 54.2 -2.74 55.4 -1.58 57.2
10 -4.08 48.8 -4.83 48.9 -5.34 48.6 -5.45 47.8 -4.74 47.3 -3.23 47.5 -1.74 48.5 -0.44 50.2
11 -4.86 55.7 -5.18 56.1 -5.43 56.2 -544 56.0 -4.90 56.0 -3.69 56.6 -2.45 58.0 -1.36 60.0
12 -5.09 60.4 -5.54 60.5 -5.85 60.4 -5.86 59.9 -5.24 59.7 -3.91 60.2 -2.59 61.6 -1.47 63.7
13 -4.44 49.0 -5.64 48.8 -6.43 48.0 -6.53 46.3 -5.39 45.2 -3.19 45.1 -1.17 46.1 0.49 47.9
14 -5.39 60.0 -6.19 60.0 -6.77 59.6 -6.93 58,4 -6.09 57.6 -4.25 57.8 -2.52 59.1 -1.09 61.1
15 -4.86 49.0 -6.61 48.4 -7.80 47.2 -7.97 44.2 -6.24 42.4 -3.15 42.1 -0.56 43.3 1.38 45.4
16 -5.64 55.4 -6.82 55.2 -7.69 54.5 -7.98 52.6 -6.81 51.4 -4.34 51.2 -2.14 52.5 -0.41 54.6
17 -5.80 59.5 -7.05 59.2 -7.98 58.5 -8.41 56.4 -7.29 55.0 -4.82 54.7 -2.60 55.9 -0.87 58.0
18 -3.38 55.9 -3.99 55% -4.43 55.5 -4.65 54.6 -4.25 53.9 -3.29 53.4 -2.25 53.4 -1.25
~~- _53.7
_ .
19 -7.50 55.2 -8.20 55.5 -8.72 55.3 -8.34 54.9 -6.69 55.7 -4.28 59.6 -2.90 65.6 -2.35 72.7
20 -6.10 55.5 -6.88 55.8 -7.50 55.6 -7.86 54.6 -7.23 53.9 -5.54 53.8 -3.75 54.7 -2.14 56.5
21 -3.47 45.5 -3.97 46.3 -4.37 46.6 -4.48 46.7 -3.84 46.8 -2.37 48.0 -0.92 50.0 0.21 52.9
22 -5.33 56.4 -5.57 56.8 -5.77 57.0 -5.75 57.0 -5.18 57.0 -3.87 57.8 -2.53 59.4 -1.39 61.6
23 -6.19 68.5 -7.20 68.3 -7.99 67.7 -8.49 66.0 -7.76 64.7 -5.82 64.3 -3.93 65.2 -2.35 66.9
24 -3.91 54.9 -4.10 55.2 -4.26 55.2 -4.21 55.1 -3.77 55.0 -2.79 55.3 -1.74 56.1 -0.76 57.3
25 -6.06 58.4 -6.81 58.6 -7.40 58.4 -7.73 57.6 -7.09 56.9 -540 56.9 -3.67 58.0 -2.18
~ -
59.8
~- _.
26 -5.16 49.7 -5.88 49.8 -6.37 49.6 -6.38 48.7 -5.53 48.3 -3.83 48.8 -2.26 50.3 -1.01 52.6
27 -5.26 55.7 -6.01 55.8 -6.55 55.4 -6.67 54.4 -5.84 53.8 -4.04 54.0 -2.31 55.3 -0.88 57.3
28 -5.19 55.6 -5.90 55.7 -6.40 55.4 -6.47 54.5 -5.61 53.9 -3.80 54.2 -2.08 55.5 -0.67 57.6
29 -5.20 55.7 -5.91 554 -6.42 55.5 -6.52 54.6 -5.70 54.0 -3.94 54.3 -2.24 55.6 -0.83 57.5
30 -5.17 56.1 -5.97 56.1 -6.50 55.7 -6.52 54.7 -5.56 54.1 -3.61 54.6 -1.81 56.2 -0.36 58.6
31 -5.41 57.2 -6.30 57.2 -6.96 56.8 -7.21 55.5 -6.36 54.7 -4.42 54.7 -2.55 55.9 -1.00 57.8
32 -5.22 56.1 -6.00 56.2 -6.54 55.8 -6.63 54.8 -5.72 54.1 -3.80 54.3 -1.99 55.7 -0.51 57.8
33 -5.21 55.7 -5.92 55.8 -6.43 55.6 -6.53 54.6 -5.71 54.0 -3.93 54.3 -2.22 55.6 -0.79 57.6
34 -6.13 59.0 -6.39 59.5 -6.63 59.6 -6.64 59.6 -6.05 59.5 -4.65 60.2 -3.21 61.8 -1.96 64.0

(In each column, the first number is the damping coefficient, the second the circular frequency.)
(The key to the model numbers is given in Appendix 4.)

Table 4. Steady state steering torques for Q = l o ' , lblft

Model Speed, ft/s

number
10 20 30 I 50 70 100 130 160
1 -9.63 -1.59 -0.28 0.37 0.55 0.64 0.68 0.70
6 -9.02 -1.65 -0.58 -0.05 009 0.16 0.19 0.20
7 -13.9 -2.62 -0.88 -0.03 0.20 0.32 0.37 0.39
8 -8.28 -1.38 -0.23 0.34 0.50 0.58 0.62 0.63
9 -9.39 -1.39 -0.09 0.56 0.73 0.83 0.86 0.88
10 -3.93 -0.48 0.11 0.40 0.49 0.53 0.55 0.55
11 -6.07 -1.06 -0.24 0.17 0.28 0.34 0.36 0.37
12 -7.84 - 1.53 -0.54 -0.05 0.08 0.15 0.18 0.19
13 -6.00 -0.69 0.22 0.67 0.80 0.86 0.89 0.90
14 -13.0 -2.37 -0.72 0.10 0.32 0.44 0.48 0.51
15 -8.34 -0.96 0.29 0.92 1.10 1.19 1.22 1.24
16 -13.6 -2.17 -0.33 0.59 0.84 0.97 1.02 1.05
17 - 18.7 -3.27 -0.90 0.27 0.59 0.75 0.82 0.85
18 -10.3 -2.16 -0.82 -0.16 0.02 0.12 0.15 0.17
19 -8.27 -0.49 0.78 1.41 1.59 1.68 1.71 1.73
20 -9.79 -1.92 -0.63 0.01 0.19 0.28 0.32 0.33
24 - 10.6 -1.76 -0.25 0.51 0.71 0.82 0.87 0.89
25 - 12.1 -2.03 -0.44 0.35 0.56 0.67 0.72 0.74

(The key to the model numbers is given in Appendix 4.)

J O U R N A L MECHANICAL ENGINEERING SCIENCE Vol I3 No 5 1971
 THE STABILITY A N D CONTROL OF MOTORCYCLES 323

Unstable 4

21 __ --
60 8’0 100 120 140 140
SPEED -ft/s

StableT3--
-4
-5-

-1
-1 6
’t
Fig. 7 . Stability and natural frequencies of standard machine as a function of forward
speed with tyre sideslip inhibited

only, of the weave mode, similar to those of the full model. centre forwards necessitates an increase in steer damping
With the no-sideslip model, however, the wobble mode is to maintain the stability of the wobble mode and, if this
completely missing and no possibility for explanation of increase is provided, substantial improvements to the
the ‘speedman’s wobble’ phenomenon is afforded. weave mode in the form of increased damping, parti-
The author is not aware of any information specifically cularly at high speeds, result. Also, slightly increased
concerning relaxation lengths for motorcycle tyres, and damping of the capsize mode is obtained.
the figure of 0.8 ft which has been taken for the standard If the front frame mass centre is moved rearward, more
machine is based mainly on measurements by Labarre and steer damping is required to stabilize the wobble mode and,
Mills (14)on a 24 in section 124 in diameter tyre. If this with this provided, increased damping of the weave mode
figure is of the correct order of magnitude, which is very at high speed is obtained. Moving the front frame mass
likely, it is apparent that a proper representation of the centre up or down by practicable amounts causes changes
dynamic characteristics of the motorcycle depends on the which can alternatively be achieved by respectively in-
inclusion oftyre sideslip and of the tyre relaxation property. creasing or decreasing the steer damping slightly.
The parameter effects recorded in Tables 1, 2 and 3 Changes to the mechanical trail, t, and the steering head
show that the capsize mode is comparatively little altered, angle, e, show that if the steering head is steepened, the
but the weave and wobble modes can be influenced con- weave mode damping at high speed is reduced while at
siderably by parameter changes. The last is most sensitive low speed it is increased. With normal and greater than
to the steer damping coefficient, K, increasing which normal trail, the instability of the capsize mode at medium
stabilizes the wobble at the expense of the weave mode, and high speeds is reduced in severity, and from the view-
although changes to the latter with varying K are not point of straight line stability, the steep steering head
great. By increasing the steering damping in fact, adequate appears advantageous for low speed machines (i.e. those
damping of the wobble mode at any speed (up to the 160 with top speeds less than about 100 ft/s). As the top speed
ft/s maximum employed in this study) can be obtained, so increases, the steering head angle must be increased to
that the potential problem at high speed is inadequate achieve adequate damping of the weave mode, the steering
damping of the weave mode. The relaxation length is also damper being tailored to damp the wobble mode sufficiently.
an important parameter as already indicated, and in- However, with a shallow steering head, high trail must be
creasing it destabilizes both the weave and wobble modes employed or the instability of the capsize mode at medium
appreciably. and high speeds will be too severe. With the normal
Lowering the mass centre of the rear frame has virtually steering head angle, too little trail causes an instability of
no effect on the wobble mode, but increases the damping the capsize mode somewhat more severe than normal and
of the weave mode at low and high speeds while decreasing a slight lack of damping of the weave mode at high speed.
it at medium speeds. Since the damping of this mode is Employing more trail than normal has very little effect.
normally more than adequate at medium speeds, these If the gyroscopic effects of the front wheel are reduced,
changes will be advantageous in practice. Further advan- less steering damping is necessary to damp the wobble
tage is obtained from the stabilization of the capsize mode mode at high speed adequately. With this reduced steer
throughout the speed range. Moving the rear frame mass damping, stabilization of the weave mode occurs higher
JOURNAL MECHANICAL ENGINEERING SCIENCE Vol13 No 5 1971
2
324 R. S. SHARP

up the speed range, but the capsize mode instability be- to measure the above parameters with extreme accuracy in
comes less severe. Also, the high speed weave mode order to obtain a reasonable representation of a motor-
damping is impaired. Converse changes result from in- cycle’s straight line stability characteristics.
creasing the polar moment of inertia of the front wheel. The steady state steering torques shown in Table 4
Increasing the rear wheel inertia increases the weave indicate that all the machines analysed have the same
mode damping at high speed and maintains the capsize qualitative behaviour in this respect. At low speed, the
mode stable up to a higher than normal speed with a torque required is in the sense opposite to the turn being
subsequent decrease in the divergence rate. negotiated and is comparatively large. Its magnitude falls
With hub centre steering arranged to have the steering off rapidly as the speed increases, it changes sign usually
axis intersecting the front wheel hub so that the effective near 40 ft/s, and increases with increasing speed there-
steering head angle and the mechanical trail are inter- after. The steering is made ‘heavier’ at low speed by
dependent, small steering head angles, with the conse- increasing the front wheel load, the steering head angle, E ,
quent small trail, lead to inadequate damping of the and the mechanical trail. At high speed the‘hteering is
oscillatory modes at high speed. With a normal head angle, made ‘heavier’ by increasing the polar moment of inertia
this damping is greater than in the case of the standard of the front wheel, and by employing a large stwring head
machine. The capsize mode is somewhat stabilized, and angle with a small mechanical trail. The hi’& speCd
large head angles give good high speed damping, a normal steering is made ‘lighter’ by lowering the rear frame mass
capsize mode, but poor low speed damping of the weave centre and moving it forwards, and by increasing the-rear
mode. The addition of a passenger assumed rigidly con- wheel inertia.
nected to the rear frame modifies its mass and inertia The form of the steady state response to steering twgue,
characteristics so that the weave mode damping is de- particularly in so far as it changes sign as the. s‘peed.ts.
creased above about 120 ft/s and the divergence rate of increased, together with the stability characteristic$.sug-
the capsize mode is increased somewhat. Lengthening the gest that the important rider activity is usually c&n&rned
wheelbase by moving the rear wheel backwards increases with stabilizing the machine about some steady state con-
the weave mode damping at high speed at slight cost to its dition rather than with the condition itself. The steady
instability at low speed. state steering torques, however, indicate roughly how
Some of the motorcycle parameters, like masses and the large the control torques must be and, in general, the
positions of the mass centres, require only trivial experi- tendency will be for large torques to be necessary at low
ments for their determination. Others, notably the tyre speeds, very small ones at medium speeds, and moderate
coefficients, the relaxation lengths, and the front and rear but slowly rising torques to be necessary as the speed
frame inertia properties, are more difficult to evaluate, and increases thereafter. There appear to be, two practical
for the purposes of this investigation have been estimated requirements, that the low speed steering should not be
rather than measured. T o find whether or not fairly small too heavy, in the interests of comfort, and that the friction
errors in the latter group (except the relaxation lengths torque in the steering system should be considerably
which have been discussed already) would have much smaller than the moderate torques required at high speed,
influence on the results, 20 per cent reductions in the in the interests of good control.
tyre coefficients and inertia values have been made in The fixed control stability characteristics shown in Fig.
turn. Also, the orientation of the rear frame principal 8 indicate that the motorcycle with rigidly attached rider
axes has been altered slightly. The results are not much and with the steering system fixed has a divergent in-
affected by any of these changes except that reducing the stability of decreasing severity as the speed increases, and
yaw inertia of the rear frame, I,,, brings a noticeable a possible low frequency oscillatory instability at high
improvement in the damping of the oscillatory modes at speed. When the free control motorcycle is oscillatorily
high speed. These results imply that it is not necessary unstable, the rider will not have choice of using torque

I:;,, 5 10 20 30 40 60 80
SPEED - f t / s
100 1 2 f l P 4
P r 7 7
160

Fig. 8. Fixed control stability and natural frequency of standard machine as a function
of forward speed
JOURNAL MECHANICAL ENGINEERING SCIENCE Vol 13 No 5 1971
 T H E S~fr\BILI'I'YXSL> CONTROL 01: MO1'OKCYCLES 325

or displacement control, and must use torque control to for such a devicc is such that at the top speed of the
stabilize the rider-machine system. Otherwise (i.e. when machine, and with the largest steering displacement
the free control motorcycle is either stable or divergently caused by such things as running over bumps, the ampli-
unstable) the rider will have choice, but a comparison of tude of the wobble mode vibration subsequently reduces
the fixed control characteristics (Fig. 6 ) with the free rather than builds up. With such a setting, the deleterious
control characteristics (Fig. 5) shows the advantage to be effect of the damper on the low speed characteristics will
always with the latter. It is therefore apparent that the be greater than the corresponding ones from a correctly
rider will not use displacement control significantly, and adjusted viscous damper. Also, control of the machine will
that the fixed control characteristics have little practical be worsened by including friction in the steering system
importance. on account of the more complex rider applied steering
The results obtained are consistent with the facts that torques which will then be necessary. The consequences
motorcycles can be controlled at all and that, with normal of wrongly adjusting the damper can be serious, and it is
skill, they can be controlled with 'hands off' at medium recommended that viscous damping be provided, and that
and high speeds. They also explain the occurrence of the manufacturer take responsibility for setting up
'speedman's wobble', the improvements which result from adequate steering damping rather than leaving it to the
moving the rear frame mass centre downwards and for- rider. A linear damper whose coefficient, K , increases with
wards, the steeper steering heads which are employed on forward speed would be better than a fixed coefficient
low speed machines as compared with high speed ones as damper, but it seems unlikely that such a device will be
well as the increased requirement for a steering damper commerciably viable.
when the front frame mass centre is comparatively rear-
ward, as brought about by offsetting the front wheel CONCLUSIONS
from t h i forks rather than offsetting the forks from the The main conclusions from this work are that the fixed
steering axis. control characteristics of the motorcycle are unimportant,
and the steady state response to steering torque is probably
D E S I G N IMPLICATIONS of secondary importance. A proper representation of the
Within the scope of this work it appears that straight line free control characteristics requires the use of at least an
stability problems will have one of three origins. These eighth order model in which the tyre relaxation property
are (1) excessive low speed instability of the weave mode, is included.
(2) excessive instability of the capsize mode at medium The conventional motorcycle has a small stable speed
speeds, and (3) instability or lack of damping of the range below which its instability is a low frequency
gscillatory modes at high speed. Just what degree of oscillatory one and above which it is a slow divergence.
instability in (1) and (2) is excessive, and how much There is also the possibility of two oscillatory instabilities
damping is necessary in ( 3 ) are not yet known. However, at high speed, one having a natural frequency of about
oscillatory instabilities of the machine with frequency in 3 Hz, the other about 10 Hz. The higher frequency in-
excess of about 2 Hz are certainly not tolerable, since the stability is stabilized at the expense of the other by
human being does not respond sufficiently quickly to increasing the steering damping.
compensate for them. The stability characteristics can be significantly con-
Some of the design changes which can be made bring trolled by practicable design changes and the model
both advantages and disadvantages. Notably, increasing developed can be used to study individual designs to
the steering damping increases the damping of the wobble assist in their optimization in the pre-prototype and
mode, and decreases that of the weave mode at all but the prototype stages.
lowest speeds. Steepening the steering head improves the
low and medium speed behaviour at the expense of that F U R T H E R WORK
at high speed; increasing the front wheel inertia worsens T o obtain accurate detailed results useful in the develop-
(2) and ( 3 ) above but improves (1). Moving the rear wheel ment of particular machines, more information than is
rearwards to increase the wheelbase reduces the weave currently available on motorcycle tyre characteristics,
mode damping at low speed but increases it at high speed. moments of inertia, and aerodynamic characteristics is
Other design changes bring advantage with no dis- needed. The last information is necessary for the assess-
advantage (provided the steering damping is suitably ment of the correct wheel loads and tractive effort at the
adjusted). Among these are lowering the rear frame mass driving wheel as functions of forward speed.
centre, moving forward the rear frame mass centre, moving Detailed experimental confirmation of the theory,
rearward the front frame mass centre, and increasing the application of the theory to other machines, e.g. scooters,
rear wheel inertia. The conventional engine flywheel and extension of the theory to include the cornering case
effectively contributes to the rear wheel inertia and so are also needed.
makes a positive contribution to the straight running
characteristics of the motorcycle. ACKNOWLEDGEMENTS
Frequently, production motorcycles are fitted with ad- The author gratefully acknowledges helpful advice
justable friction steering dampers. The setting required received from Professor F. D. Hales of Loughborough
JOURNAL MECHANICAL ENGINEERING SCIENCE Vo113 No 5 1971
326 R. S. SHARP

University of Technology. Part of the work was carried out These terms have already been accounted for by including
while the author was employed in the Motorcycle Division the rear wheel in the rear frame except+irY(2sin +@,+eTz)
of the Birmingham Small Arms Company Limited. and this term therefore contributes to To.
Similarly, the engine flywheel and the front wheel
APPENDIX 1
contribute T, so that
DERIVATION O F THE EQUATIONS OF MOTION To,= i,,(sin #$dr+@,2>+i(h sin 4$er++Azerz)
Kinetic energy +i,,[{-cos -E sin 6$+(sin E sin 6 cos 4
T = Tf+T,+Tw
+cos 6 sin 4)$>8, ++efz]
(4)
giving T by addition of equations (2), (3) and (4).
with
T, = +Mft2++If +
,pf2 ++Zfyqf2 +Ifzrf Potential energy
+
Tr = +Mr7'++Irxp,2 +;Zryq,2 +Ir,-r,2-CrxaPrrr +
V = V , Vr and includes only the gravitational potential
energy of front and rear frames, taken as zero when both
and T, = extra terms over and above those included in mass centres are at ground level. Therefore
T, and T, to account for the rotations of the road wheels
and the engine flywheel. V = M,g(a sin E cos 4-e sin 6 sin 4+e sin E cos 6 cos 4
1,-h sin 44 along OX,
+ +
f cos E cos 4) M,gh cos 4 (5)
G , has velocity components j , + h cos $4 along O Y , Lateral equation of motion and aT/@
( h sin 44 ')along OZ,
The equations for the translation of the vehicle along O X o
therefore and O Y o reduce to
72 = ( x l - h sin +$)Z+(yl+hcos +4)z+(h sin #J$)~
also pr = 4, qr = sin &, rr = cos +$
d
-(-)
d 2T
dt i x o = Qxo and d t ?jo = QVo

therefore T, = +Mr(il-h sin 4$)z+( j+h cos +$)2 respectively, since

44)'
+ ( h sin &>z++Zrx$2++~ry(sin -PT =2T
- -- - = - -d V EV - 0
++Zr,(c0s ~$>"Cr,, cos +$$ . (2) ixo ;yo ax, ay0
Similarly The inertia and external forces are resolved along 0 Y1to
obtain the lateral equation of motion thus
T, = +M,[{k,-e cos E sin 66-(a sin E sin 4
+e sin 6 cos 4 + e sin 6 cos 6 sin q5
+
+f cos E sin +)$}z { j ,+ a sin E cos 44
- Qho sin #+ Qyo cos 4 = Qyl = - (-) +z
d i3T
dt CYl
PT.
$ (6)
also
- e sin 6 sin +j+e cos 6 cos 4s
+e sin E cos 6 cos 44-e sin E sin 6 sin 46
+fcos -E cos &$+(a cos -E+e cos -E cos 6
-f sin -E)$>z+{asin -E sin +$+e sin 6 cos 44 Through the use of these expressions (6) and (7), the
necessity for replacing iland j1by i oand yoin the kinetic
+ e cos 6 sin @+e sin -E cos 6 sin $4 energy expression is avoided.
+e sin -E sin 6 cos @+f cos -E sin 44>z]
+
++I, ,{cos E cos 64 (sin 6 sin 4 Elimination of 4, and 6,
-sin E cos 6 cos ~)$)z++4~,y{-cosE sin 66 For no longitudinal tyre slip
+(sin E sin 6 cos ++cos 6 sin 4)$}z
++Ifr-{S+sin E$+COS E cos +$>z . . (3)
The translational kinetic energies of the wheels are ac- The lowest point on the circumference of the front
counted for in T , and T, but the rotational kinetic energies wheel, i.e. the tyre-road contact point has
are not.
For the rear wheel, angular velocity components in
OX,YzZz axes are
8, = tan-'
(
6 tan 4 -sin E
COSE
)
2 --E . (9)
ignoring the second order term 6 tan 4.
Equating the velocity component along O X , of a
{?el} 4 $44
= l ncos +er} general point on the front wheel circumference, with the
above condition 8, = --E imposed, to zero, gives
rotational kinetic energy = +ir,p++ir,(sin +$+ Br)2
++i,,(cos +$Iz
JOURNAL MECHANICAL ENGINEERING SCIENCE Vol 13 No 5 1971
 THE STABILITY AND C O N T R O L OF MOTORCYCLES 327

and
6 cos E
sin y = { 1-(sin + +sin c cos +8)2}112 . (16)
Forward velocity of front tyre contact point is
&+ti3 cos +$
using equation (9) and t6 cos +$ < .tl. Lateral velocity is
y,+I$+t(S sin 4d-S cos +)
again using equation (9). Therefore
Fig. 9. Plan view of motorcycle showing road to tyre
forces in the ground plane

Generalized forces and

By reference to Fig. 9
QUl = X , sin y+ Y , cos y+ Y , N Y,+ Y , (11)
and using equation (9)
a, = tan-'
Ll . I
-(Y,-b$)
. . . (17)
with the rear wheel camber angle simply 4.
Q, N IYf-bY, . . . . . . . . . (12) - When the sideslip and camber angles are small, the
Q, 21 -tS(cos +Z,-sin +Y,) . . . . . * (13) steady state tyre side forces are accurately described by
Q6 N - K S + t [ { ( X , - Y , y) cos E + ( Y ,sin+ linear functions of sideslip and camber angles of the form
-Z,cos+)sinc}S-(Y,cos++Z,sin+)]+T (14)
replacing sin y by y and putting cos y = 1. The front
wheel load Z, is taken as constant.
and the tyre side forces Y , and Y , are related to their
Tyre side forces steady state values by equations (10)
The tyre side forces are functions of the respective side-
U Or
slip and camber angles of the tyres. The front wheel in 2 Y,+ Y , = Y ' , and - Yr+ Y,= Y', (20)
some generally displaced position is shown in Fig. 10 with 111 21
EI the unit vector along the wheel axis. when the tyre relaxation property is accounted for. This
property is omitted from the analysis by putting Y , = Y ' ,
EI has components and Y, = Y , instead of equations (20).

Therefore Linear equations of motion

sin +, = sin ++sin E cos +S . . (15) Appropriate differentiations of the expression for T (by
addition of equations (2), (3) and (4)) and V (equation
X (5)), give aT/a+, aT/aS, aT/ax,, aT/ay,, d/dt(aT/3yi,),
d/dt(aT/a$), d/dt(aT/a&, d/dt(aT/a8), a V/a+, 6V/aS,
and, from equation (7), aT/a#. The generalized forces are
given in equations (ll), (12), (13) and (14), the tyre side
forces in equations (18)-(20), and the sideslip and camber
angles in equations (15)-(17). Thus, all the terms re-
quired for the lateral equation (6) and for the other three
equations in standard form (equation (1)) are available.
With second and higher order terms omitted, the
0 equations are
(M,+ Mr)(jl +a,$) +M,k$+(M,j+Mrh)d;+- Y,-
M,e8
Y, = 0

M,kj,+(M,ek+I,, cos c)S-2'"sin

R,
=1 t + +
(M,jk -C,,, (I,, -I,,.) sin E cos ~}d;
Fig. 10. Cambered and steered front wheel. The plane through
E, J, and K is parallel to the ground plane O X , Y , with
-($+?) i , ~ + ( M , k 2 + I , , + I , , sin2
E K parallel to 0 Y , +Ifz COS' E ) $ + M , k ~ , $ - l Y , + b Y , =0
JQURNAL M E C H A N l C A L ENGINEERING SCIENCE Vol 13 No 5 1971
328 R. S. SHARP

and
[I,Jl'+M,e2Zl2+tZ{(M,+Mr)b2+2Mfbk+Mfk2 +I,,

+KZ,' +t{(M,+M,)b'il cos E

+2M,bki1 cos E+(M,+M,)bti,-l1' Z 1 i l sin c

R,
+(M,k2+Z,,+Z,, sin2 cos2 c ) i 1 cos E

ZftZ12sin c-M,eZ12g sin E

i 1 2 Z 1 cos E+t{(M,+M,)i12b cos E

I
+ M , k i 1 2 cos E } S+[M,ejZl2++l,,Zl2 sin E
+bli t(M,j +M,h) +tzi{M,jk-C,,,
+(IjZ-I,,)sin E cos E } ] J +
APPENDIX 2
EQUATIONS OF M O T I O N W I T H TYRE
-t t-+-i "'R',")] Z1ilQ+(Z,t-Mfeg)Z,2~= 0
SIDESLIP ELIMINATED where I , = b+Z and is the wheelbase, and no steering
The condition that the tyres do not sideslip implies that torque is applied.
uf = C I ,= 0. Thus, from equations (16) and (17) These equations were used to give the results in Fig. 7.
1
7( ~ , + z * - t S ) - 6 cos E = 0
X1 APPENDIX 3
and j , = b$, giving PARAMETER VALUES FOR T H E S T A N D A R D

'=
. x, cos E 8 + t S
b+Z
and j , = (kZ) (a, cos E S + t S ) M,
M,
M A C H I N E A N D RIDER
2.1 slug.
14.9 slug.
and hence z, -226 lb.
$=
2, cos E 8 + t 8
b+Z
and y, = (kZ) (x, cos &+tS)
I,,
I,,
C,,y,
23 slug ft2.
1554 slug ft'.
1.28 slug ft'.
for 1, constant. ZIx 0.91 slug ft2.
Substituting these expressions in the linear equations of I,, 0.326 slug ft'.
motion and eliminating Yfand Y,,we obtain Y 0 5 3 slug ft2.
+ + +I,, sin dl
[( M ,j M,h)bt M,ejZl ',,,+hi 0.775 slug ft'.
a 3.1 12 ft.
+{~,jk-c,,,+(Z,~-Z,,) sin E cos e}t]E; b 1.574 ft.
e 0.08 ft.
+[(M,j+M,h)b cos ci,+h
Rf
k, cos EZ,
f 0.093 ft.
h 2.02 ft.
+{M,j~-C,~~+(Z,~-Z,,)
sin E cos e}il cos E R,, R, 1 ft.
t 0.38 ft.
E 0.4715 rad = 27".
C, 2512 lb/rad.
C, 211 lb/rad.
C, 3559 lb/rad.
C, 298 Ib/rad.
+(M,j2+M,h2+Z,,+Z~, COS' e+Z,, sin2 c)Zli
K 5 lb ft/(rad/s).
-(M,j+M,h)gl,fj = o I3 0.8 ft.
JOURNAL MECHANICAL ENGINEERING SCIENCE Vol 13 No 5 1971
 T H E STABILITY AND CONTROL OF MOTORCYCLES 329

APPENDIX 4 (31) Z, reduced by 20 per cent to 12.43 slug ft2.

KEY T O M O D E L N U M B E R S (32) Orientation of principal axes of rear frame altered,
(1) Standard machine (parameter values in Appendix 3). c, = 0.
(2) Reduced steer damping, K = 2 lb ft/(rad/s). (33) I,, reduced by 20 per cent to 0.73 slug ft2.
(3) Increased steer damping, K = 10 lb ft/(rad/s). (34) Z,- reduced by 20 per cent to 0.26 slug ft2.
(4) Short relaxation length, u = 0.4 ft.
(5) Long relaxation length, u = 1.4 ft. APPENDIX 5
(6) Low rear frame mass centre, h = 1.01 ft.
REFERENCES
(7) Rear frame mass centre moved 1 ft forwards with
K = 8 lb ft/(rad/s). (I)SEGEL,L. 'Theoretical prediction and experimental sub-
stantiation of the response of the automobile to steering
(8) Front frame mass centre moved back, e = -0.12 ft, control', Proc. Auto. Div. Znstn mech. Engrs 1956-57 (No. 7),
with k = 8 lb ft/(rad/s). 310.
(9) Front frame mass centre moved down,f = -0-41 ft. (2) WHIPPLE,F. J. W. 'The stability of the motion of the
(10) Steep steering head, E = 19", with short trail, bicycle', Q. 3. Math. 1899 30, 312.
t = 0.18 ft. (3) PEARSALL, R. H. 'The stability of a bicycle', Proc. Znstn
(11) Steep steering head, E = 19". Auto. Engrs 1922 17, 395.
(4) KONDO,M., NAGAOKA, A. and YOSHIMURA, F. 'Theoretical
(12) Steep steering head, E = 19", with long trail, study on the running stability of the two-wheelers', Trans.
t = 0.58 ft. SOC.auto. EngrsJapan 1963 17 (No. l), 8.
(13) Short trail, t = 0.18 ft. (5) COLLINS,R. N. A mathematical analysis of the stability of
(14) Long trail, c = 0.58 ft. two-wheeled vehicles 1964 (University Microfilms Inc., Ann
(15) Shallow steering head, E = 35", with short trail, Arbor, Michigan).
t = 0.18 ft. (6) SEGEL,L. 'An investigation of automobile handling as
implemented by a variable-steering automobile', Hum.
(16) Shallow steering head, E = 35". Factors 1964 6, 333.
(17) Shallow steering head, E = 35", with long trail, (7) WILSON-JONES, R. A. 'Steering and stability of single track
t = 0.58 ft. vehicles', Proc. Auto. Div. Znstn mech. Engrs 1951-52, 191.
(18) Reduced front wheel inertia, i,y = 0.265 slug ft2, ( 8 ) PACEJKA, H. B. 'Analysis of the shimmy phenomenon',
with K = 3 lb ft/(rad/s). Proc. Znstn mech. Engrs 1965-66 180 (Pt 2A), 251.
(19) Increased front wheel inertia, i,, = 1.06 slug ft2, (9) BISHOP,R. E. D. and JOHNSON, D. C. The mechanics of
vibration 1960 (Cambridge University Press).
with K = 7.5 lb ft/(rad/s). (10)HALES, F. D. 'Transient response of a non-linear steering
(20) Increased rear wheel inertia, ir, = 1.55 slug ft2, with oscillation', Proc. Symp. Numerical Methods for Vibration
K = 6 lb ft/(rad/s). Problems 1966 (Institute of Sound and Vibration Research,
(21) Hub centre steered with 15" castor angle. T h e University, Southampton).
(22) Hub centre steered with 25" castor angle. (11)KARMAN, T. VON and BIOT,M. A. Mathematical methods in
engineering: an introduction to the mathematical treatment
(23) Hub centre steered with 35" castor angle. of engineering problems 1940 (McGraw-Hill, Inc., New
(24) With pillion passenger and K = 4 lb ft/(rad/s). York and London).
(25) Rear wheel moved back 0.5 ft to lengthen wheel base, (12) CRANDALL, S. H. Engineering analysis: a survey of numerical
with K = 6 lb ft/(rad/s). procedures 1956 (McGraw-Hill, Inc., New York and
(26) C,, reduced by 20 per cent to 2010 lb/rad. London).
(27) C,, reduced by 20 per cent to 169 lb/rad. (13) WILKINSON,J. H. The algebraic eigenvalue problem 1965
(Clarendon Press, Oxford).
(28) C,, reduced by 20 per cent to 2847 lb/rad. (14)LABARRE, R. P. and MILLS, B. 'An investigation into the
(29) C,,reduced by 20 per cent to 238 lb/rad. dynamic stability of steering systems', Proc. Instn mech.
(30) I, reduced by 20 per cent to 18.4 slug ft2. Engrs 1968-69 183 (Pt 2A), 113.

J O U R N A L MECHANICAL E N G I N E E R I N G S C I E N C E Vol13 No 5 1971