137

STEERING MECHANISMS
D. Bastow, BSc(Eng), CEng, FIMechE*

It is considered that this survey will be most useful if it is concerned with the principles employed and some
typical methods of applying them rather than attempting to provide a comprehensive catalogue of all the
component parts of steering mechanisms, especially as at least some history may be interesting. For instance,
rocker shaft type boxes, which were originally practically universal, so long as they were Concerned with front
axles had to play their part in absorbing road shocks due to gyroscopic action of the wheels, and with high
pressure tyres did not need to be very efficient.
Now that Independent Front Suspension (I.F.S.) has generally reduced gyroscopic and other action and low
pressure tyres have put up steering torques, efficiency has become much more important and has affected
the construction used.

INTRODUCTION tyre characteristics from which comes the equivalent of

FORTHE DRIVER, good steering and handling is one of the some 8"-10" of castor angle. One wheel carrying 1000 lb
most important characteristics of a car or commercial load and cornering at 0.5 g could therefore provide some
vehicle. A great deal has been written about how to 75 lb ft of self-centring torque.
achieve this, and a considerable amount more done to help Although the self-centring torque in hard cornering is
the designer and development engineer in achieving the one of the considerations which determine the steering
desired compromise between the often conflicting require- ratio between rotation of the hand wheel and rotation of the
ments of vehicle layout, suspension system and steering road wheels about the king pin axes, it is not the most
layout. This paper is concerned with the physical mani- important. The highest steering torque of the wheels about
festation of the steering layout itself and the various the king pin axes is that required to turn them with the
components involved in it. vehicle stationary. Figures that I have measured, estimated
The function of the steering system is to allow the and otherwise arrived at, plotted logarithmically, suggest
driver, by the use of reasonable physical efforts, to control that this varies approximately as the square of the load, at a
the direction in which the vehicle goes and to place it given inflation pressure, possibly at a slightly higher index,
accurately on the road. For this purpose the steering system and at 25 lb/in2 and 1000 lb load may lie between 125 and
needs to be as rigid as possible, and its overall efficiency as 200 Ib ft per wheel. The rate of build-up varies between
high as possible. If the suspension layout results in gyro- radial and cross-ply tyres but the final scrubbing figure
scopic or other inertia torques arising from vertical wheel seems to depend upon tyre and road material and con-
movements being fed into the steering system, or if the dition and shape of contact patch; not significantly on tyre
steering layout is such that vertical wheel movements construction which indeed one would expect. Parking
result in excessive turning of the wheels about the king torques, with the vehicle moving slowly, obviously depend
pin axes, the provision of flexibility in the steering linkage upon the relationship between vehicle and steering wheel
reduces the extent to which these disturbances are felt by speeds, and tend to lie between the hard cornering and
the driver. It reduces also the precision of the steering static torques. It is difficult to give hard and fast rules for
control and increases the tendency to low speed wobble approximating them, and indeed unnecessary in the
(flap) of the front wheels (I) (z)t. Any such wobble context of the steering ratio which must be fixed by the
tendency is most easily dealt with by providing friction compromise between the need for quick reactions in an
damping for steering movements, which both increases the emergency and the possibility of a not too muscular driver
effort required and reduces the self-centring effect. being able to steer when the vehicle is stationary. Refer-
The majority of this self-centring effect is due to the ence (2) gives a steering wheel torque for this condition of
40 lb ft for a car weighing between 23- and 3 tons. This
The M S . of this paper was received at the Institution on 19th torque would not be regarded as acceptable for such a car
December 1969 and accepted for publication on 13thJanuary 1970. today.
* Consultant, Mells, nr Frome, Somerset.
f References are given in the Appendix. The steering ratio has two incidental effects: a low
Proc lnstn Mech Engrs 1969-70 Vol184 Pt 2A No 6
+ Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
938 D. BASTOW

Fig. 1. Cam and peg steering box, w i t h power assistance

geared steering reduces the impression of road shocks to pressure tyres (and in certain conditions radial-ply
the driver and increases also his impression of the steering construction) imply higher steering torques. Forward
linkage flexibility or ‘sponge’. efficiency therefore becomes important and high return
efficiency acceptable. Sliding contacts in steering boxes
STEERING B O X E S are therefore reduced to a minimum and rolling contacts
preferred. The commercial vehicle is not so happily
When the use of front axles and large diameter wheels situated as the private car in this regard, since front axles
implies relatively large gyroscopic torques from single still predominate. Luckily therefore tyre pressures have
wheel bumps, and high pressure tyres mean relatively remained higher and the use of power assistance is
small torque to steer them, steering boxes with low return
efficiencies are useful and the forward efficiency is not so
important. Dr F. W. Lanchester (I) describes a shimmy
damper which was designed to strongly resist torques
arising from the road wheels being transmitted to the
steering wheel, both to prevent gyroscopic action involving
too much ‘road fight’ or joggles, and also to provide
damping against the violent wheel wobbles which then
afflicted many cars (3). H. N. Charles (4) was equally
concerned with means for differentiating between a high
forward and a low return efficiency. The return efficiency
must not be too low, since the steering is very preferably
required to return to the straight ahead of its own accord,
and steering feel is necessary to be able to recognize the
drop in self-righting torque which, with most tyres, is a
warning of impending loss of adhesion. Dr Lanchester
pointed out that preventinggyroscopictorques acting on the
driver meant that they acted on the wheel supporting and
guiding structure in whose proportions allowance would
have to be made for them. Today smaller wheels and
independent front suspensions mean, on private cars,
generally much smaller gyroscopic torques, and low Fig. 2. Cam and roller steering box

Proc lnstn Mech Engrs 1969-70 vol184 Pt 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
<
0-

Fig. 3. Worm and nut steering gear, including power assistance

c
w
W
140 D. BASTOW

Fig. 4. Worm and nut steering gear, w i t h connecting rod connection t o rocker shaft and power assistance

Fig. 5. Worm and nut steering gear, with nut rack and quadrant connection and power assistance
(nut piston)

Proc lnstn Mech Engrs 1969-70 Vol 184 Pt 2A No16 1

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
 STEERING MECHANISMS 141

increasing. Hydraulic dampers are used in a number of linear movement of the nut to rotary movement of the
instances to reduce ‘joggle’. rocker shaft. This also incorporates power assistance.
Fig. 4 is another example of a worm and nut in which the
Rocker shaft steering boxes (angular reciprocating conversion from linear to rotary movement is made by a
output) connecting rod from the nut to a crankpin on the rocker
The two major types are cam and peg or roller, and worm shaft. This is another example of built-in power assistance.
and nut. A recent variation providing a third is a crossed- Fig. 5, a further example of worm and nut, is of the
axis pinion and sector. same type as Fig. 3 but differs in detail; it is another
Fig. 1 shows an example of the cam and peg which in example of built-in power assistance.
this case incorporates power assistance, to which reference Fig. 6 shows the pinion and sector gear developed by
will be made later. The taper peg, in a well-supported arm Cam Gears, Ltd. Skew gears are used, giving reasonable
attached to the rocker shaft, projects into the side of the freedom to fix the rocker shaft axis in the position to suit
near-helical cam track. The pitch of the cam track can be the steering linkage.
varied if desired to change the ratio towards lock. The peg
may be solid, and slide in the cam track, or mounted on Rack and pinion steering boxes (linear reciprocating
anti-friction bearings and roll and slide on the cam track. output)
Fig. 2 shows an example of the cam and roller box With connection to the wheels to be steered by track rods
provided with built-in power assistance. The initial single having mostly linear movement the provision of a linear
roller type is now superseded by double or treble tooth rather than angular reciprocating output is a potential
rollers to achieve lower surface loadings. simplification. The example of Fig. 7, the current Citroen
Fig. 3 shows an e)rample of a worm and nut with version with power assistance, retains many of the features
recirculating ball thread and rack teeth on the nut meshing of the gear originally designed in the early 1930s for the
with quadrant teeth on the rocker shaft, so converting the ‘Traction Avant’, and could undoubtedly be simplified.

Fig. 6. Pinion and sector steering geai

Proc lnstn Mech Engrs 1969-70

Vol 184 Pt 2A No 6
Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
142 D. BASTOW

Fig. 7. Citroen DS rack and pinion steering gear, with power assistance

The spring loaded slipper support of the rack into mesh is A more recent and simpler design is that fitted on the
a characteristic of all such gears and provides, more B.L.M.C. 1800 series cars, Fig. 8, the version shown being
effectively because at a larger radius than in ball joints, the one incorporating power assistance. The ball joint
friction damping in the steering system. The provision of a connections to the side steering tubes are on the rack
minimal clearance solid back-up to the spring slipper axis.
support in such gears gives the opportunity to reduce the The manual rack and pinion steering box of Fig. 9 gives
internal friction if desired. The CitroEn gear is unusual in some flexibility of layout by having critical components
its double central connection to steering rods leading to die-cast on to steel tubes, which simplifies adaptation to
steering idler levers each side whose position is dictated by different installations. The pinion in this design has a high
the up and down steering geometry requirements of the helix angle, which increases the pinion pitch circle and
wheel connections. so reduces tooth loads for a given box ratio, so that

Fig. 8. B.L.M.C. 1800 rack and pinion steering gear, with power assistance

Proc Instn Mech Engrs 1969-70 Vol 184 P t 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
 STEERING MECHANISMS 143

Fig. 9. Rack and pinion s t e e r i n g g e a r

numerically greater box ratios are possible; it also allows be employed. Acceptable parking steering effort therefore
some control of the reverse efficiency. implies power assistance in more cases. The various aspects
Pinions in all three of these gears are straddle mounted. of providing this are now examined.
The nip between the saddle and the pinion necessarily
provides the only support for that end of the rack.
The linear output of the rack and pinion steering box, Internal application within the steering box
as we shall see later, is even more attractive when one is With power assistance almost invariably provided by oil
considering the provision of power assistance. It will be pressure, we need, or prefer, a member within the box
interesting to see how widely the use of such boxes spreads which has linear movement and which can provide equal
in the future even to applications now met exclusively by piston areas for travel in both directions, or can have
rocker shaft boxes. coupled to it a separate piston also providing equal areas
for both directions of operation. Linear movements and
P O W E R ASSISTANCE equal areas are obviously preferable because angular
I n the private car, low pressure tyres of increasing tread movements are more difficult to seal than linear, and
width, with growing realization of the desirability of front because to have to provide different pressures for the two
heaviness for overall directional stability, have put up the directions of travel would generally be more difficult than
torques about the king pin for the vehicle stationary con- providing equal areas.
dition very considerably. A tendency for increased overall The rack and pinion type steering both has, and exports,
steering ratio (angular relationship) has been largely linear movement so it is an easy one to which to add power
cancelled by a simultaneous tendency to smaller diameter assistance. An early example was the DS Citroen, Fig. 7.
steering wheels. The need to maintain control for emer- The arrangement of the box and its connection to the
gency conditions limits the steering ratio which can safely steering system have already been dealt with. In the
Proc lnstn Mech Engrs 1969-70 Vol184 Pt 2A No 6
Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
144 D. BASTOW

w
Fig. 10. Pow-A-Rak rack and pinion steering gear, Jaguar XJ6, w i t h power assistance

housing end opposite the pinion, a sleeve surrounds, casing, at the inner end of the space. Leakage past the
and is coupled to, an extension of the rack. Attached to this piston remains in the circuit; leakage past the annular ends
sleeve is an annular piston, filling the gap between the is in the box casing from which therefore a return flow
moving inner and an outer sleeve fixed to the housing and to the reservoir may be provided, though this is less
provided with two annular ends which also serve to hold it important than the provision of a flow connection between
in, and support it from, the box casing itself. Connections the two ends.
are made directly at the outer end of the annular space and, A more recent and difkrent approach is the Hydrosteer
via the annular gap between the outer sleeve and the used0ntheB.L.M.C. 1800series,Fig.8. Away fromits teeth

Fig. 11. N.S.U. Ro80, application of external power assistance t o steering linkage

Proc lnstn Mech Engrs 1969-70 Vol184 Pt 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
 STEERING MECHANISMS 145

Fig. 12. Bedford VAL coach steering linkage with external power assistance

the rack becomes circular in section and an annular piston which the assisting piston can be attached. In one form,
is fixed to it, working within a cylinder formed by the typified by the ZF, the nut is the piston, Fig. 5. For equal
outer casing itself between two annular ends, of which the areas the nut has to have a blind-end hole for the worm; the
one remote from the pinion has a guide bush as well as a end of the nut away from the rack teeth which convert nut
seal, the other a seal only. A sleeve forming part of the reciprocation into rocker shaft oscillation, via a quadrant
valve body surrounds the body at the pinion end annular on the rocker shaft, moves and is sealed in a bore in the box
ring and oil entry there is direct from port to cylinder. An coaxial with the input shaft. One or the other of the seals
external connection is required for the other cylinder end. on the input and output shafts has to take full pressure
Leakage past the annular ends goes into the spaces in the while the steering is working. The Saginaw steering gear
box ends whose outer closure is provided by bellows (Cadillac, Rolls-Royce and Bentley) is basically similar.
clamped between rack ends and casing. These spaces Z F make their column type gear which resembles the
therefore need joining together. above in its worm and piston nut arrangements but has a
I n principle of construction the Pow-A-Rak box on the connecting rod and crank instead of the rack and pinion
XJ6 Jaguar is like the Hydrosteer, but as the valve sur- conversion from linear to angular oscillation, Fig. 4.
rounds the pinion shaft both cylinder end connections are These are all examples of harnessing the power assist-
external, Fig. 10. ance to a steering box component which already has linear
When one considers the rocker shaft type box, low motion. With cam and peg or cam and roller type steering
friction may imply a recirculating ball thread in the worm boxes there is no such member and a piston has to be
and nut variety and the nut is then the obvious unit to provided which has means for converting its own linear
Proc lnstn Mech Engrs 1969-70 Vol184 Pt 2A No 8
* Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
146 D. BASTOW

movement to the rotary movement of the rocker shaft. Means for providing fluid under pressure
There are three main ways of doing this: The essential components for this are a reservoir of fluid,
(1) With the cam and peg, an annular piston fits a pressure pump, a relief valve and a flow limiting valve,
between cam shaft and a housing bore. The peg on the plus connections from the flow limiting valve to the con-
rocker shaft internal lever has an enlarged part-spherical trol valve and from the control valve back to the reservoir.
collar which fits in a radial bore in the annular piston, The pump delivery, and the flow rate through the system,
Fig. 1 (Rover 3+ litre). must be related to the capacity of the servo cylinder, and
(2) A piston in a bore in the box housing has a rack the regulated pressure and the servo cylinder capacity
extension meshing with a pinion quadrant on the rocker determine the effort which can be provided to assist the
shaft, Fig. 3. steering.
(3) A roller on an axis parallel to and offset from that The flow regulating valve has two functions : to provide
of the rocker shaft, mounted on the rocker shaft, is as nearly as possible constant conditions in the control
urged in the desired direction by one of two opposed circuit from a reasonable chosen minimum pump (and
pistons, Fig. 2 (ZF-Gemmer). therefore engine) speed upwards; and to achieve this with
the minimum total pressure drop, and therefore fluid
heating, consistent with the control requirement at all
External application engine speeds. I<. C. Martin ( 5 ) describes two such valves
It is sometimes inconvenient to incorporate power combined in each case with the relief valve, indicating
assistance in the steering box and an external cylinder is how experience has produced in the second example
then used to apply the assisting forces. Although ideally a unit giving results more suited to the particular
the differential area due to the piston rod should be requirements of power assisted steering; in the second
eliminated by a corresponding tail shaft, in pracrics this is example the system flow is arranged to diminish with
not done. Fig. 11 is an illustration of the N.S.U. Ro80 pump speed and increase with system pressure, the former
which is an example of a private car external application. characteristic reducing power assistance at high speeds, the
External applications are more common in commercial latter increasing speed of response at high steering wheel
vehicle and P.S.V. applications. The Bedford VAL coach efforts.
is an example, Fig. 12.
Rotation Rotction

, A
-'
,-...-
-L-
, -..

S e r i e s 80 S e r i e s 30.
B a l a n c e d design Unbalanced d e s i g n

Fig. 14. Hobourn-Eaton pressure pumps

2
.-
>
m
'm

0
0 loco rev/min
3000 4000 5000

Oil, ATF type A. Temperature, 160°F.

Fig. 13. 'Concentric' pressure pump and reservoir Fig. 15. Flow control valve characteristics

Proc lnstn Mech Engrs 1969-70 Vol184 P t 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
 STEERING MECHANISMS 147

Pumps are either n-lobe rotors in (n+ 1) lobe housings, the actual response curves of Figs 18, 19, and 20 indicate
vane type or gear and rollers, Fig. 14. Some of the former the areas in which progress has been made.
have pressure backed rotor tip vanes, Fig. 13. Maximum Some control valves are actual piston or spool valves
pressures seem generally to be of the order of 1000 lb/in2 with linear movement, others are rotary valves, and Figs
gauge, so that they are also within the compass of gear 17 and 21 show examples of both types of valve. In
type pumps. every case valve movement is an indication of steering
Quoted delivery figures for the Concentric pump, which hand wheel torque. With rotary valves this is usually the
is used on a car with an unladen front end weight of some result of connecting the two parts of the valve by a torsion
1700 lb, are 0.625 gal/min at 1000 rev/min and 750 lb/in2, bar, protected from overload by clearance splines, which
and 0.88 gal/min at 4000 rev/mn and 50-100 lb/in2
pressure. The pump is capable of 10 000 rev/min and the
relief valve is set at 1000 lb/in2. Fig. 15 shows that an
appropriate valve arrangement can combine a maximum
circuit flow of some 2.2 gal/min at 1000 lb/in2 with one
as small as 0.5 gal/min at 50-60 lb/in2 and 5000 rev/min in
spite of a larger capacity pump.
Control valves
As we have seen, pressure supply systems have progressed
from the basis of one giving a constant circuit flow. The
control valve also has progressed from its basic form, which
in terms of a piston valve could be represented by a land
on the valve in the middle of a housing groove to which the
pump delivery was led, a lead from each side of the land to
its respective power assistance cylinder end, and a corre-
sponding land and body groove connected to the reservoir,
the connections being arranged so that movement of the
valve opens up supply to one end of the cylinder and closes
off the return of that end of the cylinder from the reservoir,
the opposite end of the cylinder being closed off from the
supply and opened up to the reservoir. At a constant total
flow the pressure differences between the two sides of the
valve, and hence of the cylinder, for progressive valve
movements are given in Fig. 16, and this curve is the basis
from which actual response curves are built, by the flow
variations with circuit and pump conditions to which
reference has already been made, and by variations in the '0
valve edge contours which affect the curve as the valve
closes one side off. The differences between this curve and Fig. 16. Control valve movement and pressure difference

c
Fig. 17. Pow-A-Rak Varamatic control valve

Proc lnstn Mech Engrs 1969-70 Vol184 P t 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
148 D. BASTOW

INPUT TORQUE-lbf in Aogdar posi!iol of rotor-dzg

Fig. 18. Steering torque-fluid pressure, B.L.M.C. 1800 Fig. 19. Control valve movement--fluid pressure

transmits the steering input torque to the box interior. converted into linear movement by a beam (ZF) or helical
With this arrangement, up to the torque at which the safety grooves (Cam Gears); the second relies upon the reaction
stops engage, the valve movement is directly proportional along the rack between rack and pinion acting against a
to the steering torque and the ratio between torque and coil spring. The coil spring is trapped between half ends
movement depends on the torsion bar diameter and at each end associated with the reacting members, one of
effective length. which is connected to the valve, the other to the housing,
At least two methods of operating piston valves are so that the spring is compressed whichever way the reac-
used: one continues to rely on a torsion bar, the twist being tions go, although the relative movement of valve and

Fig. 20. Steering torque-fluid pressure

Proc lnstn Mech Engrs 1969-70 Vol 184 Pt 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
 STEERING MECHANISMS 149

especially in the plane containing the link and ball pin

centre lines, because of vertical wheel movements possibly
totalling up to 8 in at the end of a link which may be 15 in
long, and a corresponding angle from this source of f15".
Wheel movements and link lengths vary and so result in
variation in these angles, which are somewhat increased in
most cases by the effects of king pin and possibly also idler
spindle inclinations. Most of the large angle ball joints
therefore allow f25" to f30" along the link, f10" at
right angles. All vehicles today seem to be subject still to
the low speed or 'flap' type of wheel wobble for whose
control friction is generally required in the steering system.
The existence in a vehicle of disturbing torques (z), such
as the gyroscopic torques from both wheels of an axle if a
single wheel hits a bump or falls into a depression, tends
to encourage the provision of rather more flexibility in the
steering system so as to minimize the transmission of those
inertia torques to the steering wheel. This steering linkage
flexibility encourages low speed wobble and so may require
the provision of more friction in the steering system. The
other major difference between joints for the private car
and its derivatives and those for commercial vehicles is
that greaseless ball joints (sealed for life) are becoming
increasingly fitted on private cars and are seldom required

Fig. 21. Control valve arrangement

housing depends upon the directions of the reactions. With

this scheme it is possible to have a set-up load in the
spring and hence increase the minimum torque at which
assistance is first given. Fig. 18 shows this characteristic.
I n most cases the steering linkage provides a reduced
mechanical advantage of the hand wheel over the road
wheels towards full lock, and with input-torque-sensitive
valve movements such as those described, this tends to
produce more assistance towards full lock and so even out
the effort-result relationship. If, however, there is a
varying lever ratio within the box to achieve more of the
same characteristic (cam and peg or roller with varying
pitch) and the cam end thrust is used as a sensor, the
reverse effect is obtained. The box of Fig. 1 relies upon
cam end thrust but has not varying cam pitch.
BALL JOINTS
Articulation points of the steering linkage usually call for
spherical joints to accommodate the compound angles
which are developed. For the track rod of an axle, still
almost universal on commercial vehicles, the angles
required are small and arise from king pin lean angles;
for the equivalent wheel control links on independent
suspension layouts much greater angles are required, Fig. 22. Ball joint, C.V. size steel surfaces

Proc lnstn Mech Engrs 1969-70 Vol184 Pt 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
150 D. BASTOW

on commercial vehicles, many of which have multi-point

automatic lubrication systems.
These different requirements have resulted in two basic
types of ball joint, with further subdivision on angles. The
first type is the hardened steel ball working in an induction
hardened socket. Provision is made for regreasing of these
joints. A virtually complete sphere is provided (slightly
flattened on the neck side in the example illustrated in
Fig. 22 to reduce the ball overhang from its lever fixing).
These and most other ball joints use spring loading of the
movable, upper, ball bearing pad both to take up wear and
also to provide within reasonable manufacturing tolerances
a sufficient friczion torque figure for movement of the
joint on the ball pin. Lubrication is aided by a helical -7
groove, cut to cross the pad faces, in the steering ball joint
Fig. 24. Ball joint, plastic surfaces. No spring
spherical surface.
The second type is smaller and not usually made with
any provision for greasing in service. Bearing pads are
generally made of synthetic material such as nylon.
Another variety uses an impregnated woven fabric bearing
A
material for pads. Spring loading, to prevent play, is usual.
A type of ball joint recently introduced, Fig. 24, in a I
7 2 ' ~ ~ .
range of sizes relies upon the elasticity of the plastic INCLUDED 1
bearing surface materials to maintain its friction torque i
and freedom from rattle.
One of the major problems is devising, and manufactur- I
I
ing in series, a type of grease seal which retains lubricant \
\
\

!v
Fig. 25. Rack end. Axial ball joints

yet provides the possibility of relubrication if desired and

will keep out water and dirt. For the large angle joints,
where this problem is most severe, a synthetic rubber type
material is used for these and the angle is accommodatedby
distortion of a normally single convolution bellows. These
smaller joints are assembled to contain the preloading
spring load by spinning over or staking, whichever suits
the shape of socket best, and are hence not repairable, only
renewable.
Track rod joints or those used in the equivalent tracking
linkage generally have left and right hand threads, male or
female on the socket, with various methods of locking,
e.g. lock nut, clamp, clamp type socket and cone nut.
Adjustment of toe-in can then be made without removal of
the ball joints from their levers. Connections between
steering box levers and wheels, where they are different
from the track rod linkage, tend to have right hand threads
only on the adjustment since this is most frequently an
assembly requirement only and not a service adjustment.
Ball pin shanks are held in the lever boss by the provi-
sion of a taper shank, fitting in a taper hole in the lever
boss and tightened by a nut on a thread provided on a
Fig. 23. Ball joint, plastic surfaces continuation of the taper. The taper is most frequently
Proc lnstn Mech Engrs 1969-70 Vol184 Pt 2A No 6
Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
 STEERING MECHANISMS 151

Fig. 26. Steering damper

1 in 8 on diameter. On average proportions of overhang, (a) the region between steering wheel and bulkhead, and
steepness of taper and taper length and thread size, the (b) the region between any forward mounted steering box
compression stress between taper shank and boss is higher and the bulkhead.
than the compressive stresses due to ball loads perpendicu- (a) and in some cases (b) require the possibility of
lar to the ball pin axis, even with loads arising from turning axial collapse, initially met by tubular mesh structures,
the front wheels with the vehicle stationary. Even, there- more recently by balls groove-ploughing between telescop-
fore, if the nut slacked back after tightening, applied loads ing sleeves, the latter having been found to be both more
would result in elastic stress changes only between shank consistent and cheaper. Other means are also used such as
and boss, no change of sign and hence no relative move- the bellows on some Ford cars and the tilting wheel
ment, and therefore no tendency to slacken. Some rather developed by Saginaw.
elementary tests on one example gave results which agreed Combining adequate elqstic rigidity for all normal
with the theory. The two implications of this are that, driver-inspired loads with collapse in collision-inspired
subject to checking to ensure that the example agrees with loads is not as difficult as it might at first seem, since
these conditions, pendant ball joints are just as safe as collision-engendered driver accelerations are potentially
upright ones, both being safe even without nuts once they so high that the limit on collapsing force is the possibility
have been adequately tightened, and friction locknuts are of physical damage only, the minimum excess over a
at least adequate and can be happily employed so long as reasonable safety margin on the driver’s hoisting and
the joint friction torque plus that due to rapping the taper steering forces being called for.
into the boss are sufficient to allow the friction nut to be Anti-theft devices are now also an impending require-
tightened. Torque figures ranging from 20 to 50 lb in per ment. To meet the requirements fully they imply a
ball joint have been quoted to me. For an in diameter column mounted gearchange, since the key can only be
taper shank of about the same length for a coefficient of withdrawn from the ‘lock’ position and this can only be
friction of 0.14 the torque in lb in would be about one achieved if the gear is in the ‘park‘ position on automatic
quarter of the axial force applied (lb), so that if the taper transmissions or the appropriate gear in manual trans-
could be pushed in with a force of 40 lb the torque would missions. Locking also brings a spring loaded plunger
be 10 lb in only. against a ring turning with the steering wheel, with a
number of correspondirg receiving openings. Spring
DAMPERS loading is necessary because the steering may not be in one
of the locking positions and this must not prevent locking.
Although the Lanchester arrangement (I) is a most
After locking, however, movement of the steering wheel in
effective damper for movements initiating from the road
either direction will allow the plunger to engage with a
wheels, damper applications on steering systems use receiving orifice and lock the steering.
either lever or telescopic hydraulic dampers with appro-
Police and some other arganizations need to know how
priate load settings, Fig. 26. The rod area is made equal to
to remove cars locked in this way.
half the total piston area and the valving arranged to give
equal effective piston areas for each direction of travel.
Combining easy movement at normal steering speeds with CONCLUSION
damping on high frequency but relatively small amplitude In a wide survey of this type it is not possible to cover all
movements is not an easy requirement to meet. Both aspects as thoroughly as one would like. I apologize for
private car and commercial vehicle applications exist. omissions resulting from this, and in particular for the
Power assistance on the steering provides damping too. accent on private cars as compared with heavy commercial
vehicle practice.
THE STEERING COLUMN
The present transatlantic preoccupation with secondary ACKNOWLEDGEMENTS
safety in accidents has led to a requirement for collapsible A survey paper such as this is only possible with the help of
steering columns so that frontal impacts do not push the the manufacturers of the equipment which has been
steering column and wheel into the driver’s chest. This described. I n addition to the obvious sources, from illus-
may require, depending upon the steering layout, the trations used and information quoted, I must more
possibility of collapsing under excessive load in two areas, particularly acknowledge the very kind help of the Simms
Proc lnstn Mech Engrs 1969-70 Vol184 Pt 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016

152 D. BASTOW

Group in the early stages of preparation and the valuable Mr Harold Ludicke of Harold Ludicke Engineers, Ltd,
time spent with me by members of the Automotive for Figs 2, 4 and 5.
Products, Cam Gears and Armstrong organizations. APPENDIX
Although no specific references to the sources of in- REFERENCES
dividual illustrations have been given with those illus- (I) LANCHESTER, F. W. ‘Automobile steering gear-Problems
trations themselves, I must acknowledge my indebtedness and mechanism’, Proc. Instn Auto. Enprs, 1927-28 22, 726.
to Adwest Engineering, Ltd for Figs 10, 17 and 19, to (2) BASTOW,D. ‘Steering problems and layout’, Proc. Instiz
Armstrong Patents Co., Ltd for Fig. 26, to Automobile Auto Engrs 1937-38 32, 124.
Engineer for Figs 6, 8, 11, 17 and 19, to Automotive (3) OLLEY, M. ‘Road manners of the modern car’, Proc. lnstn
Products, Ltd for Figs 22 and 23, to Burman and Sons, Auto. Engrs 1946-47 41, 147.
Ltd for Fig. 3, to Cam Gears, Ltd for Figs 1, 6, 9, 20, 21, (4) CHARLES, H. N. ’Performance characteristics of steering
gears, with some notes on the application of steering gears to
24 and 25, to Citroen Cars, Ltd for Fig. 7, to Concentric vehicles’, Proc. Instn Auto. Engrs 1942-43 37, 231.
Pumps, Ltd for Fig. 13, to Vauxhall Motors, Ltd for (5) MARTIN, K. C. ‘Power assistance pumps for steering’, Proc.
Fig. 12, and to Zahnradfabrik Friedrichshafen, through Instn mech. Enprs 1969-70 184 (Pt 3A), No. 5.

Proc lnstn Mech Engrs 1969-70 Vol 184 Pt 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
 153

Discussion
Dr J. N. H.Tait, M.Eng., C.Eng., F.1.Mech.E.-Refer- recirculation of the balls through a slot in the rod instead
ring to the author’s maximum value of 200 lb ft torque of a duct round the nut. A high value of the steering ratio,
to turn a road wheel loaded to 1000 lb when stationary, I suitable for heavy vehicles, is attainable.
have seen test results 10 per cent higher when the wheels
are steered with the brakes applied. The torque could be Mr E. L. S. McLay, B.Sc., C.Eng., M.1.Mech.E.-
even higher if one wheel were against a high kerb. This One of the main justifications for power assisted steering
raises the question of the maximum value to take when on medium weight cars must be the low effective reverse
calculating the stresses to which steering mechanisms with mechanical efficiency since full power assistance is avail-
power assistance can be subjected. I have usually based able to resist steering loads originating at the road wheels
calculations on 50 lbft torque applied to the steering wheel. for the same steering wheel effort needed to gain full
I know that steering wheel torques as high as 100lbft assistance in the forward direction. Very direct forward
can be applied on some vehicles where the seating position steering can, therefore, be provided without excessive
is such as to enable the reaction to be taken. On high speed strain on the driver from reverse loads.
cross-country vehicles, such as armoured cars, the steering In the case of hydrostatic steering the unique relation-
mechanism can, however, be much more highly stressed by ship between steering wheel position and steering angle is
the terrain on the road wheels, than by effort on the steer- lost. The system is an open loop with no position feed-
ing wheel. back. A consequence of this and the impossibility of
I recall the Lanchester steering damper, which the eliminating internal leakage in the metering device on the
author has mentioned, being fitted to Daimler limousines steering column and in the steering actuator is that in
about 1928. I was investigating the technical and practical conditions where continuous steering effort is required,
aspects of wheel wobble at the time. This damper worked e.g. crossing a slope, the steering wheel has to be con-
very well, when new. It soon had to be abandoned, how- tinuously rotated. One can imagine having to wind the
ever, due to service problems, some of which were steering wheel continuously on one’s car when climbing
associated with the ingress of water. the spiral ramp in a multi-storey car park.
Early power steerings had a number of faults which
will by now have been largely overcome by the manufac- Mr F. J. Adams, C.Eng., M.1.Mech.E.-In reply to
turers. Pumps used to be noisy. The fluid used to get the question concerning the possibility of hydraulically
excessively hot when a car was run for long distances at steering a vehicle without the necessity for mechanical
high speed. Lack of ‘feel’ in the steering caused appre- linkage between the steering wheel and the road wheels, I
hension on slippery roads. I presume that reduction of would confirm that such a system is in use on agricultural
flow at high speed (Fig. 15) and the reduction of circuit vehicles. This system is known as hydrostatic steering but
restriction will cause less heating of the fluid. Altered the steering wheel does not merely operate a valve which
control valve characteristics enabling about 5 lb ft torque directs fluid to the road wheel actuator, since it is neces-
to be applied to the steering wheel, without appreciable sary for the road wheels to follow precisely what is in-
power assistance, will have helped to regain some ‘feel’ of tended by the steering wheel movement. To achieve this
the road. condition it is necessary to meter the fluid and this is
The comparative simplicity with which modern rack done by incorporating a hand operated hydraulic pump
and pinion steerings have been adapted for power assist- through which the fluid passes and whose displacement is
ance would appear to ensure their popularity in the future. controlled by steering wheel movement. This system
The number of pinion teeth can be reduced to as €ew as works satisfactorilyfor off-road vehicles but it does require
six, with suitable addendum correction, if the maximum a hand pump and power actuator which have very low
ratio is required. internal leakages. It will be appreciated that leakage in
I should like to mention another type of steering gear either of these units will allow the steering wheel to move
which was developed for armoured cars and has been in without causing the road wheels to move and, until pumps
use for many years. A rod, acting as a track rod, has a co- have been developed which reduce this internal leakage to
axial nut which, when turned, operates the rod through a a virtually zero value, the system is not likely to be used
helical ball track. A bevel gear connects the nut to the on road vehicles. A considerable amount of work in
steering wheel. A particular feature of this design is the attempting to produce lower internal leakage pumps is
Proc Instn Mech Engrs 1969-70 Vol 184 Pt ZA No 6
Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
154 DISCUSSION

taking place and improvements are being made. I feel, So far as light vehicles are concerned a simple rack and
however, that it is likely to be some considerable time be- pinion mechanism often meets the case. The final damping
fore the system would be good enough for road vehicles, is then usually obtained from a spring loaded friction pad
particularly of the high speed variety. It is not likely, in acting directly on the rack (Fig. 7). The two principal
fact, that the system could be sufficiently cheap or advan- sources of damping will then be the rack friction and the
tageous in other ways to make it suitable for replacing the steering pivot friction. On one very famous type of rear
simple modern automobile steering gear. One can foresee wheel drive light car fitted with this type of steering,
instances on more complicated machines, even of the high- greasing the steering pivots sometimes resulted in imme-
way type, which could benefit from the hydrostatic steer- diate wheel wobble. When this happened the spring load
ing system since it would obviate what would otherwise on the friction pad had to be suitably increased, and the
be a very complicated steering linkage. In the agricultural wobble disappeared at once. The final damping factor in-
field there is a steadily increasing use of this type of equip- volved in this case was thus very obvious indeed.
ment, again mainly to eliminate a complicated linkage, and On medium sized cars the performance characteristics
it may become the only kind of steering used on certain of the steering mechanism itself can sometimes be ar-
vehicles. When more than one wheel is steered the wheels ranged to contribute to stability by using a steering gear
are still linked together by a mechanical linkage such as a having a high forward efficiency and a moderate reverse
track rod. efficiency. The final damping so introduced can often
These hydrostatic steering systems incidentally are stabilize the steering in cars of this size without affecting
normally power assisted by the incorporation of power the lightness of control in the least. The first of these was
admission valves either before or just after the hand pump. a simple combination of two well known types of cam and
The power is normally drawn from an engine driven peg steering gear, the one with a fixed peg and the one
pump as in normal power assisted steering systems. with a rotating peg. In this case a rotating peg steering
A further question that was raised concerned the pro- gear was arranged with a simple one way brake which
portion of manual to power assistance employed in normal prevented the peg from rotating when the steering
power assisted steering systems and as Mr Bastow pointed reversed, thus combining both types in one mechanism.
out this is virtually impossible to state, since the per- These steering gears, in conjunction with suitably designed
centage of assistance varies progressively with steering steering pivots, produced some very light and stable
wheel torque. For instance, in normal steering near the steerings with both front axles and independent suspen-
straight-ahead position there is probably no assistance sions. In several experiments serious instability existed if
whatsoever but as the torque increases the amount of the steering gear was replaced by one having a high re-
hydraulic assistance increases until eventually, under verse efficiency, i.e. by removing the brake from the peg
static parking conditions, probably 80 per cent of the total and putting the same steering gear back again. To regain
steering effort is produced hydraulically. This will vary stability an amount of friction that made the steering
considerably from one type of car to another and it is quite heavy was then the only cure. The easy answer was
more normal at the present moment to specify the steering to simply put the brake back on the peg again and return
wheel torque required under varying conditions of steering to very light steering with complete stability.
rather than specifying the percentage of assistance re- A more recent mechanism for producing these same
quired. The valves are then tailored to suit these conditions. effects is the recirculating ball system. The mechanical
ingenuity of some of these arrangements is outstanding.
Mr H. N. Charles, B. Sc., C.Eng., F.1.Mech.E.-The There are also many other methods.
paper gives a very clear picture of the principal types of On larger vehicles the author makes clear the necessity
steering mechanism now in general use. Everyone will for power assistance to prevent the steering control being
agree that the best type to specify depends to some extent on too heavy, even though a minimum safe amount of damp-
the size of the vehicle. I would submit that the final prob- ing is employed. With power assisted steering it is essen-
lem usually amounts to introducing enough damping to tial, as he points out, not to employ so much damping in
achieve stability, whilst still retaining lightness of control. the system that the castor effect is overpowered or that
That elusive factor, the ‘feel’ of the steering, also has the steering will wander. The author is good enough to
to be carefully considered in many cases. The one un- refer (4) to the pioneer paper I wrote after extended
avoidable source of instability in all steerings is, of course, experience with steering instability on cars, motorcycles,
the castor effect. If this effect is powerful enough to and aeroplanes. This paper had repercussions I never
return the front wheels to the straight ahead position and expected. Two well known authors of books on Applied
t o keep them there, then what is mathematically termed a Mechanics wrote me saying that what was then the
‘power driven pendulum’ automatically exists. Because of standard proof that ‘any mechanism with a forward effi-
this, the power loss due to damping must always be made ciency of less than 50 per cent is irreversible’ was going to
to exceed the power input from the pendulum effect at be removed from their textbooks forthwith. They seemed
all road speeds. Resonance with oscillations of any sub- furious with me for having disproved this classical theory,
stantial parts of the front end of the vehicle must also be which was apparently in all such textbooks at the
avoided. time.
Proc lnstn Mech Engrs 1969-70 Vol 1 8 4 P t 2 A N o 6
Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016
 STEERING MECHANISMS 155

Friends in the aircraft industry also contacted me saying That a really light and stable steering is a difficult thing
that the arguments put forward in my paper had triggered to design is made very clear by the paper. The studies
off a complete re-investigation into the damping involved involved will certainly include steering gear effects, in-
in their own aircraft controls. Their remarks particularly cidental friction effects, resonance effects, and tyre effects.
referred to combining adequate damping with lightness of Taking cost and safety into account, as we all have to do,
control. the final decisions are very important.

Author’s Reply

Mr D. Bastow-I am grateful for Dr Tait’s information paper. This problem is eased by power steering because
on static steering torques. As he implies, one usually tries the ratios required are not so extreme.
a variety of approaches when endeavouring to estimate the I was most interested to hear of the application of a
maximum forces which may be applied to steering system recirculating ball screw to the track rod of an armoured
components. My figure of static torque was intended to be car to produce the steering movement. Recirculation in the
a guide to figures possible from one approach. There is no screw rather than the nut means a short screw and a long
real substitute for measurement, but in the early stages nut but this may have had other advantages.
all possible figures are estimated and compared. Dr Tait’s Mr McLay’s comments about the result of power
statement that on high speed cross-country vehicles the steering on the reverse efficiency are most interesting.
terrain can provide the highest forces does not surprise Considering the steering wheel torque-steering link force
me since I know that each front wheel of track laying diagrams from this point of view, it is clear that the higher
vehicles may have applied forces of as much as twice the the forces originating from the tyre-road contact, the
total vehicle weight in such conditions. lower the reverse efficiency, an extremely useful state of
I did not know the Lanchester steering damper had affairs for shielding the driver from shocks so long as the
ever been fitted by Daimler, though I was aware that pressure build-up rime is not too great. The leakage
Dr Tait had done very comprehensive investigations into problem to which he refers with hydrostatic steering could
wheel wobble at the time he mentions. well be one of the reasons why this has been restricted
Some power assistance pumps are still noisy when to lower speed and heavier vehicles.
producing pressure. The necessity to avoid circuit pres- Mr Adams has contributed some very interesting in-
sure drop except when actually steering is now universally formation on hydrostatic steering systems, which can
understood and only a little arithmetic is necessary to see clearly only be used for the connection between hand and
why. The question of feel is obviously an interesting one, road wheels and not for the duty of a track rod because of
and the diagrams reproduced in the paper show what has the lack of fixed relationship between two parts. There are
come to be accepted as a reasonable compromise between ways of introducing feel into such hydrostatic steering
sufficient feel in normal and particularly high speed driv- systems but presumably this has not been felt to be
ing conditions and adequate assistance in parking. Since necessary in the existing applications.
the paper was written a combined valve and cylinder has I am naturally pleased to have the written contribution
been devised (see Automobile Engineer International De- from Mr Charles since he is the author of one of the papers
sign Review, May, 1970) in which the valve centralizing which I quote as a reference. In my earlier paper on ‘Steer-
depends upon circuit pressure rather than springs and ing problems and layout’, I included a reference to the
this implies a steering wheel torque directly proportional mathematical approach to the ‘flap’ type of wheel wobble
to the steering effort required, which presumably no one undertaken by Mr R. F. Pearsall and the results of some
could criticize. calculations based on this showing the effect of the varia-
The simplicity with which power assistance can be tion of a number of qualities on the wobble tendency. I n
applied to a rack and pinion steering must add to its practice, the quality most easily changed is the damping in
popularity, and probably to that of power assisted steering the system, and Dr Lanchester, in his quoted reference,
too. pointed out that high damping on the return path could
Dr Tait mentions pinion tooth numbers as small as six be just as effective as less damping in both directions and
in rack and pinion steering, to get the required ratio. Even would not make the steering heavier. His own damper,
smaller numbers have been obtained by the use of pinion and a steering gear with a substantially lower reverse
shafts not perpendicular to the rack axis, mentioned in the than forward efficiency, are both ways of increasing
Proc lnstn Mech Engrs 1969-70 Vol 184 Pt 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016

156 AUTHOR'S REPLY

damping without making the steering heavier. As has variable to affect wobble tendency are steering system
been already pointed out by Mr McLay, power steering flexibility and damping; the most acceptable solu-
can be another source of an effectively lower reverse tion is that which provides only enough flexibility to
efficiency. prevent road shock being felt and only sufficient margin of
The 'flap' type of wheel wobble is, as Mr Charles damping (with a preferred preponderance of return rather
reminds us, inherent in our steering systems; what he than forward damping) to give a reasonable margin to
refers to as castor angle is in practice castor effect, the cover the probable variation of the factors affecting it.
greater part of which comes from the self-aligning torque The most obvious indication of the amount of this margin
of the tyre-road contact patch (the equivalent of perhaps is the amount of front wheel out-of-balance which can be
8"-10" of castor angle, at small drift angles and with permitted before wobble becomes a nuisance at around
cross-ply tyres). In general, the two factors most readily the critical speed.

Report of an Ordinary Meeting of the Automobile Division

Held on 24th February 1970

An Ordinary Meeting of the Automobile Division was held at The Airport Hotel, Elmdon, on
Tuesday, 24th February 1970, at 6.30 p.m. Dr J. H. Weaving, deputizing for the A.D. Chair-
man, was in the Chair.

The Chairman introduced Mr D. Bastow who presented his paper entitled 'Steering
mechanisms'.

Several speakers contributed to the discussion to whom the author replied.

The Chairman proposed a vote of thanks to the author and this was carried by acclamation.

The meeting, which was attended by 100 members and visitors, terminated a t 8.30 p.m.

Proc lnstn Mech Engrs 1969-70 Vol 184 Pt 2A No 6

Downloaded from pad.sagepub.com at Imperial College London Library on June 5, 2016