You are on page 1of 16

PERGAMON

Mechanism and Machine Theory 34 (1999) 421436

A 1808 steering interval mechanism


E. Chicurel
Instituto de Ingeniera, UNAM, Ciudad Universitaria, Apdo. Postal 70-472, Coyoacan, D.F. 04510, Mexico
Received 12 May 1997; received in revised form 20 March 1998

Abstract
A simple concept to make possible a wide interval of steering mechanisms for front wheel drive
vehicles is presented: angle ampliers are placed between the output links of an essentially conventional
steering mechanism and the wheels. The ampliers can be gear or chain drives. The arrangement was
optimized, i.e. the link proportions and the amplication ratio were determined for a 1808 steering
interval (908 maximum steering angle), attaining compliance with the Ackerman geometry with a
maximum error of 0.61% and a minimum value of 45.78 for the critical angle of transmission. The link
proportions are, of course, very dierent from those of a conventional steering mechanism. An objective
function with interesting possibilities was used.
A two wheel steer, (2WS), vehicle using this mechanism would have a smaller turning radius than a
four wheel steer (4WS) vehicle with maximum steering angles of 508 in the front wheels and 308 in the
rear wheels and it would not have ``a rear overhang swing'' problem. # 1998 Elsevier Science Ltd. All
rights reserved.

Nomenclature
( )A
( )e
( )i
( )L
( )M
( )0
( )x
( )y
( )U
( )V
B

refers to compliance with Ackerman geometry


refers to outer side of the curve
refers to inner side of the curve
lower limiting value
refers to the actual steering mechanism
refers to straight ahead position of the steering mechanism
x component
y component
upper limiting value
refers to vehicle
Track, Fig. 1

0094-114X/99/$ - see front matter # 1998 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 9 4 - 1 1 4 X ( 9 8 ) 0 0 0 4 6 - 9

422

b
c
d
e
h
L
Ra
z0
a
b
g
E
Eo
y
m
f
c

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

gear arm (in the proposed system), steering arm (in a conventional steering system). It
represents both its designation and its length, Figs. 2 and 3
tie rod, represents both its designation and its length, Figs. 2 and 3
center link, Figs. 2 and 3
Pitman or idler arm, represents both its designation and its length, Figs. 2 and 3
see Figs. 2 and 3
wheelbase, Fig. 1
amplication ratio
see Figs. 2 and 3
angular position of the Pitman and idler arms with respect to the longitudinal axis of
the vehicle, Fig. 3
angular position of the gear with respect to the longitudinal axis of the vehicle, Fig. 3
angle between the gear arm and the longitudinal axis of the vehicle in the straight
position, Figs. 2 and 3
percent error of the vehicle steering angle with respect to Ackerman Geometry, Eq. (37)
percent error of the steering angle of the outer front wheel assuming that the inner
wheel complies with Ackerman geometry, Eq. (31)
steering angle of a front wheel, Fig. 3
transmission angle between tie rod and gear (in the proposed mechanism) or steering
(in a conventional steering mechanism) arm, Eq. (19)
angular position of the gear arm with respect to the longitudinal axis of the vehicle,
Fig. 3
angular position of the tie rod with respect to the longitudinal axis of the vehicle, Fig. 3.

1. Introduction
In recent years, in the automotive industry, particularly in Japan, there has been a great deal
of interest in four wheel steering, (4WS), with the purpose of improving maneuverability at low
speed (usually with the front and rear wheels in counterphase steering mode) and
controllability and stability at high speed (front and rear wheels in phase mode). A serious
problem with 4WS vehicles is the excessive rear overhang swing, so much so that some of these
vehicles include a third mode, namely, two wheel steering, 2WS. There is also the great
increase in complexity required for the two or three mode operation.
Particularly relevant to the present work is the paper of Ref. [1] which refers to a 4WS, 3
mode, medium duty truck developed by Nissan in response to customer demand for a highly
maneuverable vehicle to operate in narrow streets. The rear wheels steer through an angle of
308 while the front wheels steer through 498 . The authors report a considerable reduction in
the turning radius in comparison to conventional 2WS vehicles, and thus a considerable
reduction in the number of maneuvers required for parking and the time required. They point
out, however, that `` the problem of the rear overhang swing `` remains unsolved since its
magnitude increased up to ve fold.

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

423

In Ref. [2] a scheme is proposed to solve this ``parking problem'' and it amounts to
introducing 2WS always when the vehicle is stopped and for a short while after it is started. In
Ref. [3] this idea is carried further , i. e. an obstacle proximity sensor is used to operate an
apparatus which restricts or disables the steerability of the rear wheels. It seems to the author
of this paper, that these solutions, in some measure, defeat the purpose of increased
maneuverability.
It is pertinent to recall that narrow streets are not only frequent in Japanese cities but they
are quite commonplace as well in many Asiatic, European and Latin American cities. In
Mexico City, for instance, narrow streets are the rule in the older and the poorer sections of
town, so that delivery is a serious problem and when it refers to foodstus, it is a very frequent
problem. So much so, that soda pop, beer, bread, potato chips and candy manufacturers have
opted for specially adapted systems of delivery. Some of them simply use push carts or light
pedal vehicles, others use these vehicles but they carry them in relatively big, heavy trucks and
unload them at key points in dierent zones of the city to make the local deliveries. Thus a
relatively light, but extremely maneuverable, self propelled, delivery vehicle would be very
useful in these circumstances. Evidently, the large rear overhang swing characteristic of the
more maneuverable 4WS vehicles would be, not only inconvenient, but dangerous for this
application.
On the other hand, at the Instituto de Ingenier a, UNAM, work has been carried out, for
more than a quarter century, Refs. [47], related to 3 wheel, 1WS, 908 maximum angle of
steering (1808 total interval) as well as unlimited angle of steering vehicles. However, because
of the inherent disadvantages of 3 wheelers, attention has been turned to 4 wheelers. As, a

Fig. 1. The vehicle angle of steering is that of the cticious 5th wheel shown.

424

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

Fig. 2. The concept refers to the inclusion of output gear or chain ampliers to achieve an increased interval of
steering without a great decrease in the transmission angles.

matter of fact, at present, a 4 wheel, 2WS, electric delivery van equipped with the proposed
mechanism is in the drafting board stage, [8].
It is pertinent to point out that, besides the desirability of maneuverability itself, the energy
saved in maneuvers may have as consequences, a greater eective range in electric vehicles and
reduced pollution and increased fuel economy in conventional vehicles, [6].

Fig. 3. Geometry of the steering mechanism in a turning position.

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

Fig. 4. Most signicant positions of a non-amplifying 1808 interval of steering mechanism, run 1.

425

426

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

2. The Concept
The synthesis of an unmodied, conventional steering mechanism to meet the very stringent
requirement of a steering interval as large as 180 degrees would result in awkward
congurations. For instance in the extreme positions of a 908 turns the transmission angle
would be exceedingly small and, therefore, quite unfavourable (center diagram, Fig. 4). The
angular displacement of the mechanism proper may be kept within bounds while the front
wheels are steered through the required large angles by simply introducing angular
displacement ampliers between the mechanism and the wheels, Fig. 2.

3. Vehicle Steering Angle


In this paper, the vehicle steering angle is dened as the angle between the longitudinal axis
of the vehicle and the central plane of a cticious fth wheel centrally located between the two
front wheels which comply with Ackerman geometry, i.e. while turning, the extended axes of
all wheels meet in a single point, Fig. 1:


2 tan yi tan ye
:
1
yvA arctan
tan yi tan ye

4. Formulation for Optimization


4.1. Optimization independent variables
In the optimization runs for the proposed mechanism with amplication, the independent
variables of the optimization process were b, g and Ra. For the purpose of comparison, an
additional run with no amplication was carried out. For this case there were also three
independent variables, b, e and g (for Ra=1).
4.2. Simulation independent variable
Each optimization iteration required a simulation of the mechanism kinematics. The
independent variable for the simulation was the steering angle of the wheel on the inner side of
the curve, yi. It was generated internally by the program at 18 increments for the full interval.
Thus yi is considered to be the simulation input while the angle ye of the outer wheel is
considered to be its output.
4.3. Specications
The following quantities were initially specied: L/B, z0 and h. Since z0=1 the link lengths
are really dimensionless quantities whose lengths are given in relation to z0.

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

427

4.4. Ackerman geometry


Ackerman geometry, Fig. 1, requires that:
 
L tan y
i
B
:
yeA arctan  
L tan y
i
B

4.5. Mechanism with ampliers


4.5.1. Straight ahead position
Referring to Fig. 2:
cx0 Z0 b sin g;

cy0 e b cos g h;

c c2x0 c2y0 2 :

4.5.2. Turning position, inner side of the curve


Isolating the inner side of the mechanism and considering yi as its input and the axial
displacement of link d, i.e., ex, as its output the following relations are obtained:
referring to Fig. 3,
bi

yi
;
Ra

fi g bi ;

bxi b sin fi ;

byi b cos fi ;

cxi z0 ex bxi ;

10

cyi h ey byi

11

c2 c2xi c2yi ;

12

but

428

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

e2y

q
e2 e2x :

13

Squaring each of Eqs. (10) and (11), adding them and substituting Eqs. (12) and (13) results
in a quadratic expression in ex with coecients:


bxi z0 2
1;
14
Ai
byi h
Bi

bxi z0 fe2 c2 bxi z0 2 byi h2 g


;
byi h2

)
e2 c2 bxi z0 2 byi h2
Ci
e2 :
2byi h

15

Thus,
ex

Bi

q
B2i 4Ai Ci
:
2Ai

16

17

Also,

c 
yi
ci arccos
c

18

and the transmission angle is:


mi fi ci :

19

mi is the angle most likely to attain an undesirably low value and thus its minimum value is
considered critical.
4.5.3. Turning position, outer side of the curve
Isolating the portion of the mechanism adjacent to the wheel on the outer side of the curve,
ex, will now be considered the input while ye , the steering angle of the outer wheel, will be
considered its output,
cxe z0 ex bxe ;

20

cye h ey bye

21

c2 c2xe c2ye ;

22

but

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

q
bye b2 b2xe :

429

23

Squaring each of Eqs. (20) and (21), adding them and substituting Eqs. (22) and (23) results in
a quadratic expression for bxe with coecients


z0 ex 2
1;
24
Ae
h ey
z0 ex fc2 b2 z0 ex 2 h ey 2 g
;
Be
h ey 2
Ce

c2 b2 z0 ex 2 h ey 2
2h ey

thus,
bxe

Be

p
B2e 4Ae Ce
:
2Ae

25

26

27

Also,

 
bye
fe arccos
;
b

28

be fe g;

29

ye be Ra :

30

4.6. Objective function


The percent error is
E0

100ye yeA
:
yeA

31

Of course, Eo is a function of yi and the objective function (to be minimized) is


F

yX
i yb
yi ya

E0 yi 2

yX
i yd
yi yc

E0 yi 4 :

32

430

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

4.7. Feasible region


The boundaries of the feasible region are dened by the following inequalities:
RaL RRa RRaU ;
cL RcRcU ;

gL RgRgU ;

bL RbRbU

mL Rm

B2i 4Ai Ci r0;

33

B2e 4Ae Ce r0:

34

The program used was OPTIM by L. B. Evans based on the method of box or complex method,
Ref. [9] which checks for inequality violations at the end of each iteration and returns the
Table 1
Optimization runs
Run 1

Run 2

Run 3

gear arm, I.V.


I.V.
I.V.
3
1.7
1
5

gear arm, I.V.


I.V.
I.V.
3
1.7
1
5

0.5
7
0.2
5
5
10
38
1.5
4

0.5
7
0.2
5
5
10
38
1.5
4

0
102
103
111

0
80
81
111

2.78
1.27
2.65
0.16
45.7
0.61

2.67
0.86
2.97
9.00
41.2
0.34

Input data
Independent variables (I.V.) and specications
b
length of
steering arm, I.V.
g
Fig. 2
I.V.
Ra
amplication ratio
1
e
length of Pitman arm
I.V.
L/B
wheelbase / track
1.7
z0
Fig. 3
1
h
Fig. 3
5
Bounds
bL
0.3
bU
7
0.2
cL
cU
5
gL
Deg.
50
Deg.
50
gU
mL
Deg.
1
RaL

RaU

Objective function parameters


ya
Deg., equation (32)
0
yb
Deg.
111
yc
Deg.

Deg.

yd
Optimized Results
b
c
Ra
g
mMIN
EMAX

Steering or gear arm


Tie rod
Amplication ratio
Deg., Fig. 2
Crit. trans. angle, deg.
Max. error, %

0.74
1.10
1 (specied)
0.99
16.0
1.86

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

431

violating point to the feasible region, thus contracting the complex as the optimization
proceeds. This feature facilitates a precise determination of the optimum. The optimizations
were carried out on a Pentium 75 MHz, 40 Mb RAM.
5. Optimization Runs
Three runs are presented in Table 1. Run 1 refers to a mechanism with no ampliers and
was carried out merely for the purpose of comparison. Runs 2 and 3 refer to the proposed
mechanism with ampliers. A few points are worth mentioning.
One of the variables is dierent for run 1. However, in order to make a fair comparison, all
runs had the same number of independent optimization variables, namely three.
1. The transmission angle bound mL was considerably lower for run 1 in order to bring down
the value of the maximum error EMAX.
2. The objective function, Eq. (32), was specially tailored to reduce the tendency of a relatively
large error at greater steering angles. It is a simple modication of the least squares
criterion: in the upper range the errors are raised to the 4th power. The only dierence
between runs 2 and 3 is the starting point, yc, of the summation of the 4th power terms.
3. Error Eo, Eq. (31), uses yi as the reference angle assuming it complies with Ackerman
geometry and all the error is attributed to ye, so, in a sense, it is unsymmetrical.
Nonetheless, because of its simplicity, Eo was used in the optimization process.
The error of Table 1 is based on Eq. (1), i.e.:
100yVm yVA
yVA

35

yvA



2 tan yi tan yeA
;
arctan
tan yi tan yeA

36

yVm



2 tan yi tan ye
:
arctan
tan yi tan ye

37

E
where,

Error E is the same whether it is attributed to yi or ye or if it is considered distributed between


the two and, thus, it seems to be a more sensible criterion. Obviously, E and Eo increase or
decrease together.

6. Results
The results are presented in Table 1. The maximum error of run 1 of 1.86% seems
surprisingly low. However, the minimum transmission angle of 16.08 is too small and,
consequently, this mechanism is unacceptable (Fig. 5).

432

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

Fig. 5. Most signicant positions of the chosen version of the proposed mechanism. It is the version with the
maximum critical transmission angle, run 2. See Fig. 7 for compliance with Ackerman geometry.

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

Fig. 6. Most signicant positions of the minimum error mechanism, run 3.

433

434

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

Fig. 7. Compliance with the law of steering (Ackerman geometry) of the chosen mechanism, run 2.

The lower value of yc of run 3 had a denite eect as evidenced by the fact that the
maximum error is about half of that of run 2, Fig. 8. But the error of run 2 is suciently low
and it has a substantially greater minimum angle of transmission, consequently it is considered
the most desirable, Figs. 5 and 6. Furthermore, for run 2, the amplication ratio is smaller
while the tie rod is longer. In view of this, manufacturing inaccuracies and wear will have a

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

435

Fig. 8. Runs 2 and 3 are identical except for the ranges of summation of the square and fourth power of the error
in the objective function.

less adverse eect on the steering angles of the wheels. The degree of compliance of the
conguration chosen with Ackerman geometry may be appreciated from Fig. 7.
The objective function, however, is worthy of note; it lowered the error in the full interval of
steering, Fig. 8, and very eectively. Thus, it would be interesting to explore the eect of
increasing the range of application of the higher power of the error or of using a still higher
power. Since increasing the range had the eect of lowering the minimum transmission angle,
to make this exploration worth the eort would require designing and testing a combined
objective function that would involve both the error and the transmission angle. This
endeavour would be very time consuming and the construction of our prototype vehicle would
fall further behind schedule. In view of this, the matter was not pursued any further, at least
for the time being.
It is interesting that the amplication ratio of run 2 of 2.65 is about the same as the ratio of
the interval of steering of the proposed mechanism to the interval of steering of the
conventional mechanism, as would be expected.
In Ref. [1] the authors point out that in their 4WS medium duty truck, the minimum turning
radius is only 67% of that of their conventional 2WS base truck. It is worth noting that if
their 2WS base truck were equipped with the steering mechanism proposed here, the minimum
turning radius would be reduced to 54% for the same wheelbase and front tread or to 60% if
it had the longer wheelbase of their 4WS medium duty truck and it would have no rear
overhang swing problem. Furthermore, the scheme proposed here is much simpler.

436

E. Chicurel / Mechanism and Machine Theory 34 (1999) 421436

7. Conclusions
A concept referring to a steering mechanism has been presented that allows a large
maximum steering angle maintaining a large minimum transmission angle. The idea is simply
to introduce angle ampliers between the steering arms and the kingpins.
With the use of this concept, the feasibility has been established of a steering mechanism
with a 908 maximum angle of steering and 1808 total steering interval. An almost perfect
compliance with Ackerman geometry was attained in the optimization synthesis with a
maximum error of 0.61% and a minimum transmission angle of 45.78.
The proposed arrangement could lead to a 2WS vehicle with a minimum turning radius as
small or smaller than that of any existing 4WS vehicle with neither the rear overhang swing
problem nor the great mechanical and electronic complexity of the 4WS.
Acknowledgements
The author wishes to express his appreciation for the nancial support for this project from
the Consejo Nacional de Ciencia y Tecnolog a, CONACYT, and from the Direccion General
de Apoyo al Personal Academico, DGAPA, UNAM. He also acknowledges with gratitude the
computer diagrams by Luis Antonio Morales Hernandez.
References
[1] H. Katoh, T. Torizuka, K. Minami, N. Hata, Y. Watanabe, Development of a medium duty truck with a
mechanical linkage type four wheel steering system, SAE Transactions 99 (2) (1990) 870876.
[2] S. Jung, D.A. Guenther, Examination of the maneuverability of an all wheel steering vehicle at low speed.
Vehicle Dynamics, SAE special publication no 861. International Congress and Exposition. Detroit, Feb. 1991,
pp. 119128.
[3] H. Fussl, Road vehicle with all wheel steering, and process for steering. U.S. patent no. 5,586,031, Dec. 17,
1996.
[4] E. Chicurel, Spherical drive vehicle. U.S. patent no. 3,655,005, Sept. 27, 1969.
[5] E. Chicurel, Torque converter spherical wheel. ASME paper WA/DE-8, New York, July, 1976.
[6] E. Chicurel, Transmission for an unlimited angle of steering vehicle, in: Proc. Int. Symp. on Gearing and Power
Transmissions, vol. 2, Tokyo, Aug, 1981, pp. 407411.
[7] E. Chicurel, Veh culo automotor terrestre impulsado por rueda unica delantera y transmision totalmente mecanica. Instituto de Ingenier a, UNAM, patente no. 172247. Mexico, Dec. 9, 1993.
[8] E. Chicurel, A. Espinosa, R. Chicurel, C. Cantu, L.A. Morales, F. Gutierrez, Anteproyecto de un veh culo
repartidor electrico de gran maniobrabilidad. proyecto 4136, Instituto de Ingenier a, UNAM, March, 1995.
[9] M. Lasso, E. Chicurel, L.B. Evans, F. Lara, A. Guarda, J.F. Albarran, Aplicaciones de computacion a la
ingenier a. LIMUSA, Mexico, 1975, pp. 331342.

You might also like