Bạn đang xem bản rút gọn của tài liệu. Xem và tải ngay bản đầy đủ của tài liệu tại đây (18.96 MB, 1,188 trang )
Constructional details of the engine
93
for large quantity production is costly, porosity in the finished casting can
present difficulties, aluminium is more easily damaged in service and rather
more prone to gasket blow-by failure, corrosion may present problems –
especially where there are copper components in the cooling system – and
heat-resistant valve seat inserts are essential.
Cast iron is inherently stiffer, and therefore contains noise better, and is
cheaper. On the other hand, the labour costs in making the moulds and cores
are higher, and more labour may be required for removing the sand cores,
and for fettling.
In a paper presented by D.A. Parker and R.H. Slee, at Symposium 86,
held by AE plc, some particularly interesting comments were made on trends
in engine design, Fig. 3.38. These authors pointed out that many of the
overhead camshaft engines which came into vogue in the nineteen-seventies
had single ohc valve gear with vertical valves and bath tub combustion
chambers in cast iron heads mounted on cast iron crankcases. Screw type
tappet adjusters were used and twin valve-springs obviated a risk of the
1980
(b)
1970
(a)
1990
(c)
Fig. 3.38 Trend in engine design 1970–1990, as illustrated by Parker and Slee, of AE plc
94
The Motor Vehicle
valve’s dropping into the engine in the event of breakage of one, and in any
case, the installation was more compact than with a single larger spring.
By the nineteen-eighties the pistons carrying four-ring pistons had given
way to three-ring types, with shorter skirts to reduce both friction and
reciprocating mass. Aluminium heads became popular, partly to reduce weight,
but also because their better thermal conductivity would help in the struggle
to meet threatened legislation for the elimination of lead from gasoline. For
the latter reason, too, sparking plug position was determined by the need to
shorten flame travel, enabling the ignition to be retarded to reduce octane
number sensitivity of the combustion chamber. Improvements in valve and
port geometry, coupled with the introduction of electronic single-point injection,
helped in the achievement of both better fuel economy and satisfying emissions
regulations.
According to the two authors of the paper, in many instances, by the
nineteen-nineties, not only cylinder heads but also blocks would be of
aluminium alloy, while coolant volumes would be restricted to reduce weight
and shorten the warm-up time. Twin overhead camshafts actuating four valves
per head would provide good specific power and economy. The inherent
stiffness of valve trains of this layout would enable single valve springs to be
used, while self-adjusting tappets would reduce both noise and maintenance
requirements. Because of its superior accuracy of metering, multi-point would
replace both single-point fuel injection and carburettors for cars in all but the
bottom end of the price range. Light-weight, two-ring pistons with integral
hydrodynamic bearing devices for taking the thrust and minimising skirt
area would reduce friction.
3.48
An interesting cylinder head design
An outstandingly good cylinder head design is that of the Dolomite Sprint,
Fig. 3.39. Since it exemplifies the best solution to many of the problems, it
will be described here. It is of aluminium and has four cylinders, four valves
per cylinder, and a single overhead camshaft. The exhaust valves are inclined
16° to one side of the axis of the cylinder and the inlets 19° to the other side.
Their seats are in a penthouse-type combustion chamber.
Since the engine, when installed, is tilted 45° towards the exhaust side,
the inlet valve ports then slope steeply downwards. This facilitates cold
starting, in the following manner. As the crankshaft is turned, any fuel remaining
unevaporated in the manifold runs down into the cylinder, where it is
evaporated by the heat generated during the subsequent compression stroke.
Mixing is further assisted, at TDC, by the squish effect between the flat area
surrounding the slightly dished portion of the crown of the piston and the flat
lower face of the casting, each side of the pairs of valves. Because of the
steepness of the slope of the inlet ports, this fuel runs down positively into
each cylinder in turn so that, once the first cylinder fires, the others will pick
up immediately. The penalties for a slope that is either inadequate or too
steep are respectively mixtures that are either too weak or too rich to fire in
sequence.
A single camshaft, with eight cams, serves all the valves. Each cam actuates
first an inlet and then an exhaust valve. However, whereas the inlet valve is
actuated directly, through the medium of an inverted bucket tappet, the exhaust
valve is opened by a rocker, one end of which follows the cam and the other
Constructional details of the engine
95
(a)
Inl
et
Exhaust
(b)
Fig. 3.39 The Triumph Dolomite Sprint
bears on a pallet, or thick shim, seated in a recess in the top face of the valve
spring retainer. A similar pallet is interposed between the exhaust valve
spring retainer and its tappet.
To reduce the velocity of sliding between the rocker and cam, the pad on
the end of the rocker is curved. This, however, tends to increase the velocities
of both opening and closing of the valve and therefore has to be taken into
account in the design of the cam profile.
The sparking plug is very close to the centre of the top of the combustion
chamber, for efficient combustion. As a result, it has not been possible to
make the diameter of the two inlet valves much larger than that of the
96
The Motor Vehicle
exhausts but, to compensate for this, their lifts are greater – 8.712 mm as
compared with 7.798 mm.
To span the distance between the camshaft and the exhaust valves, the
rockers have to be long. Consequently forged En8 steel, instead of the usual
cast iron, rockers are used. Because of the relatively poor bearing characteristics
of steel, however, oil from a radial hole in the rocker bearing is fed down a
groove on top of the rocker arm to the cam follower pad. To obviate any
possibility of oil dripping down and getting into the exhaust valve guide,
where it could form carbon deposits, a heat-resistant flexible seal is fitted
over the upper end of the guide.
As can be seen from Fig. 3.39 the upper half-bearing for the camshaft and
the semi-circular clamp for the rocker shaft are machined in a single long
diecast aluminium bearing cap secured on the plane of the inclined joint face
of the valve gear cover by a bolt at each end. With this arrangement, the
valve actuation gear can be fitted to the cylinder head, forming a self-contained
sub-assembly, with valve clearances set, all ready for mounting on the engine.
The cylinder head holding down studs on the inlet side of the head are
inclined 74° relative to the cylinder gasket joint face, and this brings their
upper ends out through the diecast cap, adjacent to the rocker shaft and at
right angles to the inclined seating face. Consequently, when the head is
tightened down, the stud pulls the clamp tightly down on to the rocker shaft
so that there is absolutely no possibility of fretting fatigue between the cap
and shaft. Set bolts, perpendicular to the cylinder head gasket joint face, hold
down the exhaust side of the head. With this overall arrangement, the head
can be removed in service without disturbing the valve gear.
The inlet valves are of Silchrome, while the exhausts are either Nimonic
80A with Stellite tips on the ends of their stems, to prevent undue wear, or
of En18 with Nimonic 80A heads welded to them. All stems are chromiumplated to reduce the rate of abrasive wear in the guides.
Sparking plugs of the conical seating type are fitted, because they are
screwed into bosses at the lower ends of tubular housings cored vertically in
the head casting, where plug washers would be difficult both to place and to
retrieve. The gap between the upper end of each of these cored housings and
the valve gear cover is spanned by an aluminium tube with elastomeric seals
moulded around both its ends. The lower seal is a tighter fit in the head
casting than is the upper one in the valve gear cover, so that the tube will not
pull away from the head when the cover is removed. Although 14-mm plugs
are fitted, their hexagons are of 10-mm plug size, so that the tubular housings
can be of small diameter. This, in turn restricts as little as possible the water
passages around the plug bosses.
The good thermal conductivity of the aluminium head, together with
generous cooling passages around the valve seats account for the absence of
valve sinkage when unleaded fuel is used. Lead in fuel is thought to act as
a lubricant between the valve and its seat, so with unleaded fuel and elevated
temperatures the rate of wear of less well-cooled seats can be high with the
result that the valve sinks into them.
A four-valve head layout was chosen because this engine was required to
have high performance. So far, no positive proof of why the four-valve
layout is so efficient has been put forward. However, the central positioning
of the plug probably contributes, and the scavenging on a broad front –
through a pair of exhaust valves – may also help.
Constructional details of the engine
3.49
97
Cylinder block and crankcase arrangement
Until about 1925, cylinder heads and blocks were generally integral, and in
highly rated aero-engines, notably the Rolls-Royce Merlin, the arrangement
was still in use throughout and after the Second World War. However, with
road vehicles, the need for frequent servicing and economy of manufacture
led to the adoption of the separate cylinder head casting. But, with the
introduction of turbocharging and the consequent development of high gaspressures and temperatures, the designers of the Leyland 500 diesel engine,
Fig. 3.40, reverted to the combined head and block casting, bolted to a
separate crankcase.
The elimination of the cylinder head gasket disposes not only of the
barrier to the conduction of heat, but also of the thick sections of the adjacent
faces of the two separate castings, which can cause thermal distortion of the
structure. In particular, with this layout, the cylinder bores and valve seats
should be relatively free from distortion. The absence of cylinder head retaining
studs should also help in this respect, as well as give the designer greater
freedom in arranging the valve and porting layout and cooling passages.
Large covers bolted on each side of the block facilitate the cleaning of sand
from the cooling passages after the casting operation. However, despite all
these potential advantages, the design proved extremely difficult to develop
to the point of complete satisfaction in practice.
Other cylinder block and crankcase arrangements appear in Figs. 3.41 to
3.45. The black sectioning represents cast iron and the cross-hatching
aluminium alloy. While the figures are simplified and to a great extent
diagrammatic only, each is represented by one or more examples in past or
present practice.
That shown in Fig. 3.41 is the conventional arrangement used in large
engines of the highest quality, where it is not essential to economise in
machining and fitting. Here a monobloc cylinder casting, with overhead
valves in a detachable head, is bolted to a two-part aluminium alloy crankcase
split on or below the crankshaft centre line. The maximum degree of
accessibility is provided for valves, pistons and bearings.
Fig. 3.42 shows the side-valve arrangement of what has been the most
Fig. 3.40
Fig. 3.41
Fig. 3.42
98
The Motor Vehicle
widely used construction but which is now generally superseded by the
overhead valve arrangement shown in Figs 3.43 and 3.44 with various camshaft
arrangements. Modern designs simplify the cylinder block casting and improve
accessibility of valves and rockers. A single monobloc casting in iron or in
aluminium alloy – in which case wet cylinder liners are used – extends from
the head joint face to usually well below the crankshaft centre line, forming
a rigid beam structure of great depth and stiffness. Well-ribbed webs tie the
crankcase walls together and carry the main and camshaft bearings. The
bottom is closed by a light sump which is often ribbed for cooling purposes
and may carry the main oil filters. For examination and overhaul, pistons and
connecting rods must be withdrawn upwards through the cores unless removed
bodily with the crankshaft.
The corresponding overhead-valve construction is illustrated in Fig. 3.43,
which shows a low camshaft for push rod and rocker operation of the valves.
A dry liner is indicated in this case. The valves all lie in the same longitudinal
plane.
End assembly of the crankshaft is provided for in the construction of
Fig. 3.44. When arranged for three bearings, the centre bearing is mounted
in a circular housing of diameter exceeding the diameter of the crank webs.
The old Lea–Francis design was of this construction, which provides a rigid
and accurately aligned support for the crankshaft. It had high twin camshafts
to operate the inclined overhead valves, which were pitched transversely.
Figures 3.45 and 9.10 show two examples of wet liner construction, one
for a medium-powered petrol engine and the other for a blower-charged,
poppet-exhaust two-stroke ci engine. The air gallery and scavenge ports of
the latter will be noticed. Water joints are made with synthetic rubber rings
(as indicated) and gas pressure joints usually by separate gasket rings to each
bore. A light alloy main casting is indicated in these two cases.
3.50
The aluminium crankcase
Aluminium alloy has always been a potentially attractive material for
crankcases. This is because of its light weight, good thermal conductivity,
Fig. 3.43
Fig. 3.44
Fig. 3.45
Constructional details of the engine
99
prospect for good cooling and, if aluminium pistons could be used in aluminium
bores, the absence of differential expansion would enable tighter clearances
to be adopted in the bores. However, until recently, this light metal has failed
to gain wide acceptance because it has been necessary to fit cast iron liners,
primarily owing to the unsatisfactory bearing properties of aluminium pistons
in bores of the same material. This has increased the cost of production with
a material that, in any case, is expensive. Moreover, there have been other
problems, including the risk of electrolytic corrosion and of the head welding
to the block, as well as a high noise level and, when diecasting is used
because of its suitability for large quantity production, a large rejection rate
owing to porosity.
Some European manufacturers, however, with access to relatively cheap
supplies of aluminium, and others throughout the world who produce expensive
cars, as well as Rootes – later Chrysler, now Peugeot-Talbot – have used
sand-cast crankcases and cylinder blocks. The Rootes car, the Hillman Imp,
was a special case, since weight reduction was of prime importance as the
engine and transmission were to be installed at the rear. An open-top deck
layout was chosen because, originally, diecasting was envisaged.
There are four iron liners, which are preheated to 204°C before they are
inserted in the mould, for the aluminium to be cast around them. To key the
liners in the aluminium, their outer peripheries have spiral grooves
0.331 mm deep machined around them at 3.18 mm pitch. When cast in, they
are arranged in two siamesed pairs – otherwise the block would be unduly
long – so water can flow transversely between only the central two cylinders
and across the ends of the block.
Renault, too, have used cylinder blocks with wet cast iron liners for many
years. One of their later developments has been the 1.47-litre R16 model, in
which a high pressure aluminium diecast cylinder block and crankcase is
employed. Although most of the walls are only 4 mm thick, the casting is
well ribbed and flanged for stiffness. The liners are closely spaced in the
open-top casting, and there are five main bearings – the Imp, being a much
smaller unit, 875 cm3, has only three bearings. An interesting feature of the
Renault engine is the employment of an aluminium head too.
In 1971, General Motors announced the Chevrolet Vega 2300 engine with
a diecast all-aluminium crankcase and cylinder block, Figs 3.46 and 3.47,
reported in Automobile Engineer, August 1970. Developments that have
made this possible include the Accurad method of diecasting, described in
the November 1969 issue of Automobile Engineer. This entails control of the
cooling of the metal so that areas remote from the point of injection solidify
first, and a two-stage injection process, the second stage of which effectively
compensates for shrinkage of the metal. Secondly, a new alloy, termed
A390, has been introduced by Reynolds Metals, and this combines good
fluidity in the molten condition with fine dispersion of silicon after heat
treatment, which gives good bearing properties and ease of machining –
large particles wear the tools and tear from the surface of the alloy. It contains
16 to 18% silicon, 4 to 5% copper and 0.45 to 0.65% magnesium.
Diecasting is employed because of its suitability for large quantity production and its accuracy: the latter minimises subsequent machining and enables
thin walls to be incorporated – as little as 4.826 mm, but 6.35 mm for the
bores – thus economising in material. The finished block for this 2.3-litre
100
The Motor Vehicle
Fig. 3.46 The diecast aluminium block of the Chevrolet Vega has 6.35 mm thick
cylinder walls, and most of the other walls are 4.826 mm thick
Fig. 3.47 The open-deck design, with siamesed cylinders, was adopted for the diecast
aluminium crankcase of the Vega so that the dies could be easily withdrawn
Constructional details of the engine
101
engine weighs only 36 lb, as compared with 87 lb for the cast iron block of
the comparable Chevrolet L-4 engine. However, because it is diecast and
provision must therefore be made for withdrawal of metal cores, the opentop deck form, with siamesed cylinders, has to be used.
Diecast aluminium heads have not yet been used by any manufacture in
quantity production, because of the problem of withdrawal of the cores.
Experiments, though, have been made with two-piece castings joined by
means of adhesives or by electron beam welding. A cast iron head, however,
makes up for any lack of stiffness of an aluminium block with an open-top
deck and helps to contain the noise of combustion and from the overhead
camshaft and valve gear. On the Vega, it is secured by ten long bolts screwed
into bosses at the bases of the cylinders, so that the bores are rigidly held in
compression.
The cylinder bore treatment in the Vega, to form a wear-resistant and oilretaining surface, and to prevent scuffing, comprises exposure of the hard
silicon particles by an electro-chemical etching process. In addition, the
skirts of the pistons are coated with iron by a four-layer electro-deposition
process, to render the bearing surfaces compatible with those of the bores.
The coats, in order of application, are: zinc, copper, iron and tin. Zinc bonds
well to aluminium, the copper prevents removal of the zinc by the iron
plating process and the tin prevents subsequent corrosion of the iron and
helps in running-in. The iron coating is 0.019 mm thick.
3.51
Camshaft drive
Whatever the type of valve used it is necessary in the four-stroke engine to
drive it from a camshaft which runs at half the speed of the crankshaft, as
each valve is required to function only once in two revolutions of the crankshaft.
The necessary gearing for this purpose is placed, with few exceptions, at the
front of the engine, that is, at the end remote from the flywheel and clutch.
The camshaft or camshafts may be driven by gears, chains or toothed belts,
while a few overhead camshafts have had a ‘coupling rod’ drive. Figure 3.48
shows diagrammatically some typical arrangements of two-to-one drive for
both low and high camshafts.
No. 1 shows the simplest possible arrangement of direct gearing for either
one or two camshafts. The wheel on the camshaft has twice as many teeth as
1
2
5
3
Fig. 3.48
4
6
7
102
The Motor Vehicle
the crankshaft wheel, and therefore revolves at half the speed of the latter.
Where, as is often the case, the distance between the two shafts is considerable,
this arrangement requires undesirably large gear wheels, and this had led to
the adoption of the arrangement shown at 2. Here an intermediate idler
wheel is interposed between the crankshaft wheel and the camshaft wheel.
This idler wheel may be of any convenient size, as the number of teeth in it
does not affect the gear ratio. The camshaft now revolves in the same direction
as the crankshaft, whereas in the former arrangement it did not.
A chain drive is shown at 3. The single chain drives the auxiliaries in
addition to the camshaft, the drive thus being a triangular one, but if the
chain is passed round only the crankshaft and camshaft sprockets, the shorter
run thus obtained is less prone to whip.
At 4 is illustrated a combined chain and gear drive for twin high camshafts.
The chain sprocket ratio is 1 : 1, and the 2 : 1 ratio is provided by the
gearing.
An automatic chain tensioner of the Coventry eccentric type is indicated.
The axis of rotation of the jockey chain wheel can swing eccentrically
round the spindle on which it is mounted to take up slack in the chain, the
desired pressure on the back of the chain being adjustable and automatically
maintained by the clock-type spring, the inner end of which is secured to
the mounting spindle and the outer end to the drum which carries the
jockey-wheel bearing.
A later development is the Renold hydraulically-actuated tensioner, Fig.
3.49(a). When assembled, the cylinder casting is secured by two bolts to the
front wall of the crankcase, and inserted into it is a plunger, carrying the
neoprene-faced slipper together with the spring-loaded piston which forces
the slipper lightly into contact with the chain. When the engine is started, oil
from the pressure-lubrication system, ducted into the cylinder of the adjuster,
forces the slipper firmly against the chain. The piston has a spiral ratchettoothed slot in its skirt, in which registers a peg projecting radially inwards
in the bore of the plunger: its function is to limit the backlash, that is,
movement of the slipper away from the chain.
Various means of driving overhead camshafts are illustrated at 5, 6 and 7
Fig. 3.49(a) The ratchet device on the plunger of the Renold hydraulic chain-tensioner
prevents the chain from going too slack when the oil pressure is low
Constructional details of the engine
103
in Fig. 3.48. Diagrams 5 and 6 show vertical shafts driven by bevel and skew
gears respectively. In either case the 2 : 1 ratio may be obtained in one or two
steps as may be most convenient. In the latter case particularly, ratios 3 : 2
and 4 : 3 are advantageous.
It will be noticed that in the case of the bevel gears a tongue and slot
arrangement is provided so that expansion of the cylinder block will not
affect the meshing of the gears. This provision is not strictly necessary with
skew gears, though the slight axial movement of the vertical gear that may
place will result in a slight variation of timing between the hot and cold
conditions.
No. 7 is the layout on the Jaguar in-line engine, which has stood the test
of time. The 2 : 1 ratio is divided between the two stages, the tooth ratio of
the four chain-wheels being 21: 28 × 20:30. This tends to distribute the wear
and maintains uniformity of pitch. Another advantage is that small sprockets
can be used.
For the lower chain, a nylon-faced damping slipper bears on the driving
strand, to prevent thrash due to torsional oscillations. On the earlier models,
a spring plate tensioner bore on the slack strand, but this was subsequently
superseded by the Renold hydraulic tensioner. For manual adjustment of the
tension in the upper chain, the jockey sprocket is mounted on an eccentric
spindle, as described in detail in Automobile Engineer, May 1956.
Renold also produce a commendably simple double-acting tensioner,
comprising a central cast housing with twin parallel bores in which are
horizontally opposed plungers, each with a shoe on its outer end. Two coil
springs in compression push the plungers, lightly outwards until the shoes
contact the inner faces of both the taut and idle runs of the chain. At the same
time, oil splashing into a pocket on top of the housing is drawn continually,
by the motions of the plungers due to fluctuations in the drive, through a
non-return valve to fill their bores. Movement of one plunger inwards in its
bore instantly seats in the non-return valve, causing the other to move in the
opposite direction so that the light contact with both runs is continuously
maintained.
Another type of tensioner is that used in the Jaguar V-12 engine, Section
4.24, supplied by the Mores Chain Division of Borg-Warner Ltd. Of the four
runs of chain between the sprockets – on the crankshaft, two camshafts and
the jackshaft – three are controlled by damper pads and the fourth by the
tensioner. Because these four runs form a strand 1.674 m long, an extremely
accurate and effective tensioning is essential.
The Morse tensioner, Fig. 3.49(b), supports the run of chain over most of
its length – its shoe is approximately 28 cm long – so the force per unit area
on the shoe is small, and undue noise and wear are therefore avoided. Moreover,
the shoe is a nylon moulding and therefore does not require a separate
facing. Nylon, of course, has a low coefficient of friction and, in this application,
fillers are added to improve both the stiffness and wear characteristics. It is
also resistant to oil, fuels and temperatures of 150°C and even higher.
Furthermore, its flexibility increases with temperature, so, when the engine
is warm, the tensioner readily conforms to changes in shape of the chain run
throughout its life. Finally, so that the chain cannot go slack and, possibly,
ride over the sprocket, a one-way device is incorporated in the tie that holds
the shoe in the bowed condition. A tendency for the chain to go slack can