The Doxford Direct Drive Diesel Engine
P. JACKSON, M.Sc., M.I.Mech.E. (Member of Council) *
The paper describes the development of the large marine Diesel engine over the past decade with respect to:
1) The burning of high viscosity fuels.
2) The supercharging of the two-cycle engine by turboblowers.
3) The development of very large engines.
The paper is concerned particularly with the development of the Doxford engine in these various respects and particularly with regard to the experiences of the P type engine. Particulars are given under the various headings of the few difficulties which have been experienced with the P type engine of which there are now fifteen at sea.
In the section on the development of large engines, a description is given of the work which Doxford have undertaken in this respect in their consideration of various types of engines and finally an outline is given of the type of engine ultimately being developed which it is believed will be an advance on current practice in the development of the large direct drive marine Diesel engine.
Introduction
The direct drive Diesel engine has grown in popularity for the propulsion of seagoing vessels since its inception for this duty some fifty years ago, until today it is the most popular form of propulsion for merchant vessels and tankers up to the largest sizes. The slow speed marine Diesel engine has not only proved its reliability, but the primary reason for its success, relative to other means of propulsion, is the economy in operating costs which results from its low fuel consumption. Its progress during the 1930’s prior to the second world war was considerable, but it is really since the end of that war that the Diesel engine has established its supremacy over the steam engine and the steam turbine for powers up to 20,000 h.p. The consumption of fuel relative to that required by the steam engine and its boiler is less than half and it is not much more than half of that required by the marine steam turbine up to the powers for the propulsion of medium sized vessels up to 20,000 tons. A few very modern steam turbine installations, with complex and complicated steam cycles, feed heating and reheat systems, do show a somewhat better fuel consumption than the older steam turbines but they have not proved popular. The reliability of the modern marine Diesel engine is equal to that of the steam engine when account is taken of the cleaning and repairs to the boiler and is also equal to that of the steam turbine when account is taken of the repairs and complication of the boilers, condensers and feed heating and heat exchanger systems. The manpower required is much the same and there is no increase in other factors. It is true that the first cost of the marine Diesel engine and its machinery is greater than that of the steam engine but the saving in fuel costs more than compensates for the extra outlay and gives a good return on the extra capital cost. The progress and development of the marine Diesel engine during the past fifteen years has been intense and the more important of these lines of progress have been:
1) the adaptation of the marine Diesel engine for burning high viscosity fuels,
2) the supercharging of the two-cycle engine by means of exhaust turboblowers,
3) the development of the very large marine Diesel engine up to powers of 25,000 h.p.,
4) the progressive reduction in size for a given power, and
5) the improvement in the design of detail parts leading to greater reliability and simpler operation and reduction in first costs.
* Director and Manager of Research and Development, William Doxford and Sons (Engineers) Ltd.
Heavy Fuel Oporation
Crude oil as obtained from oil wells, contains light and heavy fractions and after the second world war there was such a demand for the lighter fractions, for the propulsion of motor cars, commercial vehicles and aeroplanes by petrol engines and jet engines, that the total supply of these light fuels was required for these purposes and their price per gallon or per ton increased considerably relative to the cost of the heavier fractions which remained after these lighter fuels had been distilled from the crude. There was relatively little difficulty in burning these heavy residual oils under the boilers associated with the steam engine and the steam turbine and this resulted in these prime movers having an advantage in the cost of their fuel, relative to the lighter marine Diesel fuel used at that time in the Diesel engine. A 10,000 h.p. Diesel machinery installation will consume about 36 tons of oil per day at 90 per cent load and this will cost some £350 as Diesel fuel, or £220/day as boiler fuel, whereas the cost of fuel for a similar steam turbine installation will be about £400/day and the fuel oil is the heaviest item in the cost of operating a ship. Intensive development was therefore undertaken to make the large slow speed marine Diesel engine suitable for operating on these heavier residual fuels, firstly by the oil companies—notably Shell—so that there would be a market for the heavy products from their refineries and, secondly, by the marine Diesel engine builders, who required to make their engines suitable for operating on these heavy fuels, in order to reduce the cost of operation and maintain the supremacy of the Diesel engine relative to the steam engine and steam turbine.
The first problem was to heat the heavy fuel in the double bottom storage tanks, so that it could be pumped to the daily service tanks, then to heat again in these tanks so that the fuel would not cool and later to heat further so that the fuel could be cleaned and non-combustible products removed and then, to heat the fuel at the engine to reduce its viscosity, so that it could be pumped by the main engine fuel pump and injected into the engine cylinders, through the fuel spray valves in fine jets suitable for rapid and efficient burning.
Secondly, non-combustible matter such as water, sand and noncombustible sludge had to be removed from the fuel. Centrifuges were developed and improved for this purpose; water and some of the heavier substances were removed by a preliminary centrifuge known as a purifier and then the fuel was reheated and passed through a second centrifuge, known as a clarifier, which removed some of the heavier asphaltines. A fuel system embodying these features is shown in Fig. 1. With such a system the modern slow speed marine Diesel engine can burn fuel oils up to 3,000 seconds viscosity with reliability, although there are some drawbacks. In general, for an engine burning more than 30 tons of fuel per day, an extra junior engineer is carried to maintain the centrifuges and clean them as necessary. The fuel oil has to be heated to a temperature of between 185 deg. F. and 210 deg. F. to obtain a sufficiently low viscosity for the satisfactory spraying of the fuel and the injector nozzles have to be water cooled in order to prevent the formation of carbon trumpets around the small holes of the injector. On the Doxford engine, the author’s company uses a C.A.V. injector which is of the differential-spindle spring-loaded type of high precision, made by a company specializing in the manufacture of fuel pumping and injection components. The arrangement of this injector and of its water cooling is shown in Fig. 2. The fuel pump is of conventional design, the plunger having a helix on its forward end which is rotated by a rack and pinion to regulate the fuel quantity required, according to the speed and power of the engine. When using high viscosity fuels the fit or clearance of the fuel pump plungers in their barrels must be somewhat slacker than that necessary for light Diesel fuels, which can penetrate past tighter fuel pump plungers to lubricate them. Heavier fuels do not penetrate so readily and so do not lubricate tight plungers which may seize in consequence.
As is well known, the Doxford engine uses a fuel system where a multi-ram pump pumps the fuel at high pressure into fuel bottles or accumulators, usually one for each cylinder and, when a small valve is lifted by a cam on the camshaft, the fuel in these high pressure bottles flows rapidly to the fuel injectors and lifts the differential needles to spray the fuel into the cylinder. Steam tracer pipes are used alongside the fuel piping to maintain or increase the temperature of the fuel in the piping between the fuel pump, accumulator bottles, and fuel injectors. This common rail system is very quiet and simple in operation and although having had considerable experience with jerk type fuel pumps in the past, the author
would not and has not made any attempt to change the common rail system of the Doxford engine. Little power is required to drive the fuel pumps and timing valves, so that the camshaft and its driving chain can be of relatively small size and low in cost. It is, however, necessary, even as with the jerk pump system, that the temperature of the fuel be maintained so that it can be pumped to a pressure of between 7,000 and 8,000 Ib./sq. in. for fuels of 1,500 to 3,000 seconds viscosity, otherwise the fuel is not sprayed properly, resulting in incomplete combustion with sludge being formed which, besides being abrasive and causing liner wear, can be deposited in the scavenge trunks and if allowed to accumulate can result in scavenge fires. There was one scavenge fire only on the very first Doxford turbocharged engine installed in the British Escort, but since then there is no record of a scavenge fire on any Doxford turbocharged engine (it may be that after making this statement we shall hear of some). The principal disadvantage of using high viscosity fuels in marine Diesel engines is the increase in liner wear and the consequent more frequent replacement of these relatively costly parts. As already stated, rapid liner wear can be caused by the abrasive sludge resulting from imperfect combustion of the fuel and a rise in fuel consumption of less than 1 per cent, can be responsible for the accumulation of large quantities of sludge. Even with good combustion, however, the liner wear, when using high viscosity fuels, is some 2j times as much as the liner wear when using Diesel fuel. This is due to the high percentage of sulphur in the fuel, sometimes amounting to 3, or 4 per cent, of the constituents of the fuel. This sulphur is burnt in the engine, the products of combustion being SO, and SOa. These products of combustion may condense on the cylinder walls causing corrosion of the surfaces. This corrosion can be reduced considerably by operating the cylinder liners at a sufficiently high temperature, so that these products of combustion do not condense on the liners and the curve in Fig. 3 shows the reduction in liner wear when the cylinder is operated with a water inlet temperature of 155 deg. F. and a water outlet temperature of 175-185 deg. F., relative to the previous operation with a water inlet temperature of 140 deg. F. and an outlet temperature of 160 deg. F. Other expedients are, to use a vanadium/titanium iron for the liners, which the author believes was introduced by Doxford’s and is now usual practice for marine Diesel engine liners, or alternatively, an iron with a fairly high chromium content or centrifugally cast liners or even chromium plated liners. Some of these liner types will be dealt with later when explaining particular features of operation of the Doxford P type engine.
Lubrication can also play a very important role in minimizing liner wear when running on high viscosity fuels and the oil companies have done commendable work in this direction, by developing cylinder oils with alkaline additives to counteract the action of the condensed sulphurous products of combustion.
At Doxford’s, considerable work has been carried out on the precision timing of the injection of the lubricating oil onto the piston and piston rings, it being considered that correct lubrication was an essential and it will be recognized that if only the cylinder liner and piston rings could be ensured an oil film throughout their working stroke, then liner wear would surely be at a minimum. Similar remarks apply to the wear of piston rings.
In the early days of the use of high viscosity fuels on the Doxford engine, there was trouble with corrosion of the crankshaft and crosshead pins of certain engines. These circumstances have been explained previously and, although somewhat puzzling at the time, investigation quickly showed that the corrosive products of combustion were passing the piston skirts and falling into the lubricating oil in the crankcase, where any leakage of water, from the cooling system supplying the water cooled pistons, mixed with these sulphurous products to cause a dilute sulphuric acid, which was corrosive to the crankshaft and other bearing journals of the engine. These difficulties were quickly overcome by the fitting of a diaphragm between the engine cylinder and the crankcase to prevent the sulphurous products of combustion falling into the lubricating oil and a further precaution was to adopt oil cooling to the lower pistons instead of water cooling. Either of these two safeguards would probably have overcome the difficulty and, since water is a better cooling medium than lubricating oil, the new and larger engine now under construction at the Doxford works, will have water cooling to the lower pistons, and the telescopic pipes supplying the water to the pistons, pass through two glands and the space between is open to the atmosphere so that any leakage can be seen and can be drained away.
In general, the author advises superintendents to manoeuvre the engines on Diesel fuel, since there are less difficulties and less liner wear with such operation, though many superintendents require that their engines shall manoeuvre on the same fuel as used throughout the voyage. In such cases, a circulating system is arranged so that the fuel can be continuously circulated by a pump, through the engine fuel pipes and back to the daily service tank as shown in Fig. 1.
Turbocharging
The application and development of turbocharging to the large slow speed marine Diesel engine has been intense during the past decade. It is not yet ten years since Doxford’s ordered their first turbocharger from the Brown Boveri Company of Baden, Switzerland and although they were somewhat late, relative to some other builders, in applying turbocharging to their engines, yet this order for a VTR 630 turbocharger was the first large turbocharger to be ordered from the Brown Boveri Company. This turbocharger was for a three-cylinder engine of 600 mm. bore and on test the engine gave 3,750 b.h.p. at 112 r.p.m., i.e. an increase in power of 30 per cent, relative to the normal engine of that date and there was no difficulty in obtaining this power equivalent to 130 m.i.p. Subsequently four and six-cylinder engines were
installed in ships, turbocharged to powers of about 25 per cent,
over the ratings of the original normally aspirated engine,
which resulted in a considerable reduction in the weight and
length of the engines for the same power and in the cost per
horsepower. There was never any difficulty in obtaining the
power and the problems were mainly associated with slow
speed operation. As is well known, the Doxford engine is an
opposed piston engine, having lower pistons driving centre
cranks and upper pistons driving on to two sidecranks of the
crankshaft. The lower pistons open and close the scavenge
ports and the upper pistons open and close the exhaust ports.
The first engine had a lead of the exhaust cranks relative to
the centre crank of only 4 deg. and the exhaust ports were
opened 66 deg. before bottom dead centre of the centre cranks.
This engine would not run below a speed of about 55 r.p.m.
without the assistance of scavenge pumps and the engine was
installed in a ship with turboblowers and scavenge pumps
in series. The next engine was a six-cylinder engine, this
time with a lead of the exhaust cranks of 6 deg. and the
exhaust ports were opened about 66 deg. before bottom centre
but, although better, again this engine would not run below
about 45 r.p.m. without the assistance of scavenge pumps.
The next engine had a lead of the exhaust cranks of 8 deg.
and the exhaust ports were opened 71i deg. before bottom
centre which gave a greater impulse of the exhaust gases
to the turbine. On the ship’s trials this engine would
run down to a speed of about 28 r.p.m. without scavenge
pumps. It was somewhat sluggish in picking up and
while it was tested on the sea trials without scavenge
pumps it went to sea with scavenge pumps in series. A
subsequent six-cylinder engine had a lead of the exhaust
cranks of 8 deg. and the exhaust ports were opened 75 deg.
before bottom centre and this engine would run reliably both
ahead and astern down to about 25 r.p.m. and would pick up
speed readily from this slow speed and this engine went into
a ship without any scavenge pumps to assist the turboblowers,
although a small electric fan, requiring a motor of about
30 h.p. to drive it, was installed to assist prolonged manoeuvring
at slow speed in the canals. The first P type engine had
a lead of the exhaust cranks of 9 deg. and the exhaust ports
were opened 75 deg. before bottom centre and an attempt was made to use the underside of the pistons to give a slight impulse
to the air pressure for assisting the flow of air into the cylinder
at slow speeds. This engine ran reliably down to 22 r.p.m.
and would pick up in speed from 25 r.p.m. and operate
reliably from full speed and full power down to these slower
speeds for manoeuvring, so that this engine has operated the
whole time on the turboblowers without any assistance and
without any trouble. A fan was again fitted, requiring about
30 h.p., for emergency purposes only should the turboblowers
break down, but it has never been used. The progressive
developments in port timings of the Doxford turbocharged
engine are shown in Fig. 4.
Turbocharger Arrangements
As was described in the paper “The Future Doxford Marine Oil Engine”*, a number of turbocharging arrangements have been tried; the first six-cylinder P type engine had three VTR 630 turboblowers, one for each pair of cylinders, the next engine had two VTR 630 blowers, one for each set of three cylinders and the next engine had three VTR 500 blowers, one for each pair of cylinders. The latter arrangement gave the greatest amount of air and thus would give a higher power or alternatively lower temperatures for the same power. In the beginning, it was difficult to obtain equal exhaust temperature readings throughout the cylinders when all cylinders were doing equal power, as shown by indicator cards and this was due to the exhaust pulse from No. 2 * Jackson, P. 1961. “ The Future Doxford Marine Oil Engine.” Trans. 1 Mar. E. Vol. 73, p. 197. cylinder affecting the temperature reading on No. 1 cylinder and the pulse from No. 5 cylinder affecting the temperature reading on No. 6. On the six-cylinder engine the angle of firing and of exhaust pulse between cylinders 1 and 2 and between cylinders 5 and 6 is only 120 deg. of crank angle, whereas the angle between cylinders 3 and 4 is 180 deg. In consequence the exhaust gases from No. 2 and No. 5 cylinders rush over to cylinders No. 1 and No. 6 respectively and thereby increase the exhaust temperature readings. The exhaust thermometers were subsequently moved into the exhaust belt of each cylinder, thus being further away from the impulse of the adjacent cylinder and thereafter equal exhaust temperature readings could be obtained on all cylinders for equal loads.
The majority of the six-cylinder engines subsequendy delivered have had two VTR 630 turboblowers, but at least one customer prefers to have the three VTR 500 turboblowers, since this gives a greater degree of reliability should one turboblower break down. Under these conditions the two remaining machines will give a greater power and thus operate the ship at a higher speed than would one turbocharger of the two turbocharger arrangement. The four-cylinder turbocharged engines have mostly had two VTR 630 turbochargers, each with two unequal inlets so that the exhausts from No. 1 and No. 4 cylinders are led into the large inlet of each turbine and the exhausts from No. 2 and No. 3 cylinders, which fire 180 deg. apart are divided between the smaller branches of each turbine. One four-cylinder engine has been fitted with one HSBT turbocharger which had a special turbine entry casing, so arranged that the exhausts from No. 1 and No. 4 cylinders were connected directly into the turbine and those from No. 2 and No. 3 cylinders entered into a third branch, situated on the front according to the arrangement shown in Fig. 5. The turbine nozzle ring was divided into two equal sections. This arrangement of one single turboblower was entirely satisfactory and in fact gave slightly more air than the two VTR 630 turbochargers but, on the other hand, the slow speed running was not so good and an auxiliary fan was arranged to assist, being brought in automatically by the manoeuvring lever when this was pulled down to notch 21, which corresponded to a speed of about 40 r.p.m. This four-cylinder engine gives a power of 6,600 b.h.p. with one turboblower, which is the simplest and cheapest arrangement, there being only one running turbine and the Brush turboblower has plain bearings lubricated from the engine system, special oil filters being fitted.
On the six-cylinder engine in particular, the amount of air delivered by the turboblower increases with load at such a rate that there is no rise of exhaust temperature with load after about 120 m.i.p., indicating very ample air and in consequence large mean effective pressures can be carried. The prototype engine was run on test at a load of 160 m.i.p. for a number of hours and, although this shows ample overload carrying capacity of the engine, it has been decided to rate the engine relatively easily at the low m.i.p. of about 130, corresponding to a b.m.e.p. of 120 which is the rating of the six-cylinder engine, when running at 10,000 h.p. at 120 r.p.m. Much interest has been shown in the Doxford P type engine which is very gratifying and the author has been pressed on a number of occasions to give particulars of the operation and to deal with the many rumours concerning it.
Operation Of The P Type Engine
The first engine gave its power on test within one month
and no more than two months were required to overcome all
the minor difficulties, the greatest of these being in connexion
with the glands to the upper telescopic pipes leading the water
to and from the upper pistons. It has been necessary to provide
the gland boxes with a degree of flexibility and also to
spring load the packing so that it does not require adjustment.
The arrangement subsequently adopted is shown in Fig. 6
and it will be noticed that the packing box is mounted between
two rubber rings and can thus accommodate itself to the
movements of the telescopic pipes; contact is made on these
rubber rings by a spring which also maintains a pressure on
the packing.
After preliminary adjustments, the first engine was put
on a 1,000 hour non-stop day and night run. During this first prolonged test there was heavy wear of the piston rings
and cylinder liners, which circumstance became known and
was given wide publicity. Apart from being worn, the piston
rings were covered with a black sludge and there was a considerable
accumulation of this sludge around the scavenge
ports and in the entablature. This was traced to defective
spraying of the fuel from the injectors, the test having been
carried out on a high viscosity fuel and the cause was due to
the injection pressure falling to about 4,800 lb./sq. in. at the
end of injection. This first P type engine was rated at
10,000 h.p., i.e., 1,666 h.p. per cylinder, which was the highest
output that had been obtained from a Doxford cylinder up to
that time and therefore the quantity of fuel per injection was
higher than ever before and this plug of oil taken from the
accumulator bottles lowered the pressure by over 2,500
lb./sq. in., i.e. from 7,500 lb./sq. in. at the beginning of
injection to about 4,800 lb./sq. in. at the end of injection as
shown in the diagram (Fig. 7). To overcome this, each set of
three bottles was joined together by a relatively large diameter
pipe and thus the volume of fuel available for each injection
was increased threefold. With this arrangement the injection
pressure fell by only 400 lb./sq. in. from the beginning to the
end of injection and this completely eliminated the sludge, so that the difficulty with broken piston rings and worn liners
was largely removed. As is usual, however, when an engineer
meets difficulties, he tackles the difficulty in all possible ways
and apart from dealing with the problem of injection pressure
and the removal of the sludge the Doxford engineers obtained
piston rings of various types, rings of harder material,
chromium plated rings, rings sprayed with molybdenum
disulphide and rings with an inserted molybdenum disulphide
band. The harder rings did appear to be of some advantage;
the chromium plated rings gave an advantage for 700 or 800
hours, but once the chromium surface was worn off such
rings wore quicker than normal rings. No reliable result has
yet been obtained from the rings treated with molybdenum
disulphide.
The cylinder liners of the first engine were subsequently
chrome plated, four by van der Horst and two by Sheepbridge
Stokes Ltd., the van der Horst liners being lubricated with
the van der Horst special lubricant known as “ Seven Star,”
whereas the Sheepbridge Stokes liners were lubricated with Castrol R.M.D.Z. During the first year’s service of the
Montana, in which the first engine was installed, the rate of
wear on the van der Horst liners, with this special lubricant,
has been the best though there were initial difficulties with
this lubricating oil, due to sticking of the lubricator plungers.
Subsequent engines have been fitted with normal cast iron
liners and since the first three engines, arrangements have
been made with Sheepbridge Stokes Ltd. to produce centri-
fugally cast liners and these are proving satisfactory with rates
of liner wear varying between 0.003 in. and 0.008 in./1,000 hr.
when operating on high viscosity fuels. The heavier rate of
wear is for the first 2,000 hours but improves rapidly as
shown by Fig. 8. The water jacket temperatures have been
increased to 150-155 deg. F. inlet and to 175-180 deg. F. outlet
and this is beneficial in reducing liner wear, since the
sulphurous products of combustion do not condense on the
hotter cylinder walls but it is difficult to get a ship’s engineering
staff to maintain such temperatures. They much prefer
the lower temperature of 140 deg. F. since the engine is cooler
and more comfortable.
Alkaline additive oils have been used for the lubrication
of the cylinders, of both the water and oil soluble types but
so far experience shows no definite conclusions since the rate of wear varies from engine to engine and even from cylinder
to cylinder in the same engine under the same conditions.
Having watched the operation of the lubricators of the
Doxford engine over several years and having examined the
delivery of oil from the quills, the author had formed the
opinion that the oil dribbled on to the liner walls and that
there was no reliable timing of the lubricating oil delivery, so
the prototype engine was fitted in the first instance with
eccentrics for operating the lubricators instead of cams.
Eccentrics are much quieter, give a smoother drive and there
is no tendency for the roller to flick off the cam toe. During
the initial testing of this engine it was not possible to increase
the lubricating oil supply to the cylinders above 1 gal./day
per cylinder without having intermittent knocking or detonation
in one or other of the cylinders. It was suspected that
this detonation was caused by burning of the lubricating oil and the eccentrics were replaced by cams. It was immediately
possible to increase the quantity of lubricant to these cylinders
to the normal quantity of 2i gal./day without there being any
detonation. This circumstance led to an examination of the
timing of the lubricating oil supply to the pistons and oscillograph
records (Fig. 9) taken from the lubricator quills showed
that the delivery of oil from the quills did not comply with
the timing that the cam should have determined. Subsequent
experiments showed that the original old fashioned type of
lubricator gave the most reliable timing when fitted with the
Alan Muntz ball type of flow indicator, although this was by
no means perfect. During the ensuing year a well known
lubricator manufacturer has developed, for Doxford’s, a special
distributor type of lubricator which injects the oil through a
quill of differential needle type. It was found impossible
to time the relatively small quantities of oil, which were
originally delivered on every stroke of the piston, to give say
two drops per minute to each quill, but if ten times this small
quantity were given to a differential type quill then the timing
could be determined precisely. The distributor type lubricator
therefore delivers one small drop to each quill once in every
ten revolutions of the engine. This arrangement and these
experiments are described more fully in a paper which the
author’s colleague, Mr. J. F. Butler/’) gave to the CIMAC
Conference in Copenhagen in June, 1962.
The increased powers and higher mean pressures which
are being developed by Diesel engine cylinders are resulting
in greater wear and this is a problem which is being encountered
by all manufacturers of marine Diesel engines and
the foregoing is a brief description of the steps which the
author’s company are taking in this matter.
During the early testing of the prototype P engine, the combustion was not as good as desired, indicated by a shaded
exhaust and a rather high fuel consumption. Many tests were
made with a variety of nozzles having different arrangements
and sizes of holes and at various angles but without much
success. In consequence it was decided to carry out tests
with scavenge ports and swirl arrangements. It has been
Doxford’s experience over many years that experiments with
glass models, having differing degrees of swirl as indicated by
smoke or feathers or similar devices, would not yield the best
results when tried out in an engine and this has also been the
author’s experience with other types of engine. In consequence
it was decided to conduct swirl experiments on three cylinders
of the prototype engine. Rings of vanes were made for each
cylinder as shown in Fig. 10 and these were clamped around
the cylinder liners in such a way that their position could be
varied thus varying the degree of swirl. As a result of a series
of tests, the best positions were found and castings embodying
rings of ports of this type have been fixed in the best position
on each cylinder of subsequent engines. It was suggested by
one of the author’s colleagues that the rings of ports creating
a swirl left a core of gas in the centre of the cylinder and he
proposed that the ring of ports should be divided into a truly radial set of ports at the bottom, followed by a tangential set
of ports above, the idea being that the radial inflow of air
would clear any central core of exhaust gas and that the
tangential ports above would then create a swirl. This
arrangement is shown in Fig. 11. No beneficial results were
obtained from this arrangement.
It is considered that the clear exhaust and low fuel consumption
resulting from these tests indicates perfect combustion,
such as is only possible with a through scavenge engine
and in these respects the opposed piston engine has many
advantages, since both the ring of exhaust ports and the ring
of scavenge ports can have the largest possible area and the
exhaust ports are opened very quickly by the upper pistons,
which is essential in a turbocharged engine if the energy of
the exhaust pulse is to be used to best advantage. The port
areas of the opposed piston engine are larger than can be
obtained on any other type of engine and the port openings
and areas for the P type engine are shown in Fig. 12 The subsequent operation of the P type engine has been
very satisfactory and entirely to the satisfaction of the owners.
One difficulty was in connexion with the side rod bottom ends.
The side rod bottom end caps were turned with a vertical
projecting face and the heads of the side rod bottom end bolts
had a flat, which butted against this face to prevent them
turning when the large nuts were tightened up. This feature
is common practice with medium speed four-cycle engines as
built for marine auxiliaries and many industrial purposes and
it eliminates the necessity for a snug in the head of the bolt,
which is an undesirable feature. The arrangement is illustrated
in Fig. 13a, and it will be seen that there was a relatively sharp
corner between the flat face for the head of the bolt and the
vertical face which prevented the bolt head turning. Sharp
corners are undesirable features and the resulting stress concentration
caused cracks to develop at this point and the cap
of one of the bearings of the Tudor Prince failed as shown
in the photograph (Fig. 14). Fortunately this defect was found
early in the life of the engine and immediately all these side
rod big end caps were replaced by the construction shown in
Fig. 13b where the vertical face has been eliminated and also the sharp corner, though it is now necessary to fit a snug to
prevent the bolt head turning when the nuts are tightened up.
The first engine installed in the Montana had bearings of the
type which failed later on the Tudor Prince and although this
engine had operated for a much longer period, over six times
as long and at much higher loads, yet there had been no
failure. The bearings of the Montana were forgings, whereas
most of the later bearings were flame cut from steel slabs in
Doxford’s works. It was apparent that these flame cut slabs
were not suitable and were of a notch sensitive steel but as
stated, all such bearings have been replaced on all engines.
Another relatively small difficulty, but somewhat annoying,
has been the escape of vapour past the upper piston into
the atmosphere and this has been due to the failure of the
rubber ring between the piston head and the upper piston
skirt to prevent the passage of the exhaust gases (Fig. 15a).
Apparently the hot exhaust gases from the turbocharged
engine, being at a higher pressure, subjected this rubber ring
to more arduous conditions than previously, but the difficulty
has been cured by fitting a silicone rubber ring and in future
engines the ring has been adequately shielded by a projection
on the piston skirt. The old and new designs are shown in
Figs. 15a and b.
The new features of the P type engine, of which there
were many, have given no trouble whatsoever. The two-piece cylinder liner with its cast steel combustion space and
external water connexions has been entirely successful and it
is proving cheaper than the old type of long one piece liner.
Arrangements have been made with Sheepbridge Stokes Ltd.
to make dies and equipment so that the two halves of the
liner can be centrifugally spun and in consequence the liners
are being produced in a harder and more homogeneous metal
with less allowance for machining, i.e. the castings are
accurately produced nearly to the final dimensions required.
Notwithstanding the higher loadings of the turbocharged
engines, there has been no trouble with top end bearings and
the centre pad which was inserted under the centre crosshead
to take a proportion of the load has been entirely successful.
There has certainly been no trouble with the squeezing out of
white metal. The large diameter short length bearings supporting
the crankshaft have also given no trouble; these bearings
are contained in cylindrical bores in the engine bed without
any spherical mounting. The oil cooling of the lower pistons
has been entirely successful and there has been no case of
carbonization of the lubricating oil. The lubrication of the
centre top end bearings through telescopic pipes and the
feeding of the oil down the centre connecting rods to the
centre big ends has also been successful. The cooling of the
upper pistons by means of telescopic pipes has fulfilled its
function though, as described previously, there have been
slight difficulties with leakage of water from the glands which
has led to modifications. The transverse beams machined
from steel blocks have been satisfactory and the rigid connexion
thereto of the side rods and upper piston rods has been satisfactory
and experience has indicated that there was no necessity
to attach the upper pistons to the transverse beams by
means of spherical pads. While these pads are useful on the
first assembly of the engine to correct any slight errors of alignment, they have not been required in the subsequent
running of the engine and the pistons have been locked since
the first three or four engines. The arrangements for taking
the exhaust gases away from the exhaust belts by short direct
pipes have been successful in reducing the resistance to flow
and pressure diagrams taken from the cylinders (Fig. 16) show
that there has been no blow back of exhaust gases into the
scavenge belt, the flames from which could start scavenge fires
should there be any inflammable sludge in the entablature and
around the ports. Due to the good fuel combustion qualities
of the engine there has been no such sludge and several
reports from chief engineers, operating the engines, say that
the entablatures are the cleanest which they have ever seen.
They also report that the general cleanliness of the crankcase
and condition of the oil are as good as, or even better than
they have ever known on any type of engine. It is too early
to say that the construction of the bedplate with its two
longitudinal girders and double plates in the transverse girders
has entirely eliminated the difficulties experienced with the
old type of bedplate but nevertheless the indications are that
this is the case. Certainly the crankshaft deflexion readings
and bedplate deflexions in the ships are more than halved.
The engine has proved to be very economical in the consumption
of lubricating oil and there is practically no loss
of lubricating oil from the crankcase or from the piston cooling
tanks since these are entirely self-contained. The cylinders
have been lubricated from sight feed lubricators and the
consumption of cylinder lubricating oil has been below
2\ gal./cylinder/day of 24 hr. All the engines are operating
on high viscosity fuels and apart from the first engine which
had chrome hardened liners the remainder of the engines have
been lubricated with oils containing an alkaline additive of the
oil soluble type. Particulars have already been given of the
rate of liner wear of the first engine in the Montana but the
liner wear on other engines with cast iron liners has been
equally good and a report from the superintendent of the
Silverweir contains the following statement “ the wear of the
liners during the first 4,500 hours is practically negliglible and
the piston rings all have their radius intact and there does not
appear to have been any wear as indicated by the butt
clearances of the rings not having increased.”
The P type engine is rated at 1,666 h.p. per cylinder at
120 r.p.m., giving 10,000 h.p. in 6 cylinders with an m.i.p. of
130 lb./sq. in. The mechanical efficiency of the turbocharged
engine is high due to the power required for supplying the
scavenge air being derived from the exhaust gases. It is of
the order of 93 per cent. The resulting b.m.e.p. is 120
lb./sq. in. and this output can be maintained with an overload
of 20 per cent, without difficulty. The best fuel consumption
is at about 100 m.i.p. and figures of 0-325 lb./b.h.p. hr. on
Diesel fuel and below 0-34 lb./b.h.p. hr. on high viscosity
fuel have been obtained. Typical performance curves of the
4, 5 and 6-cylinder engines are shown in Fig. 17 from which
it will be seen that the performance of all engines is very
similar and that in the Doxford opposed piston engine there
is no real case for rating the engines having multiples of three
cylinders higher than the engines having other combinations
of cylinders. At the time of writing this paper (November
1961) there are eight P type engines at sea and by the time
the paper is read, there will be at least twelve manufacured by
Doxford’s themselves and probably another four by licensees.
LARGE ENGINES
Much was written in the technical press in 1961 and even
in the daily press regarding the failure of British Diesel engine
companies and designers to produce an engine of large power.
This criticism has been directed particularly against Doxford’s.
For many years the Doxford engine was one of the most
popular and during the period from the commencement of the
second world war to 1958 the company’s engine works were
fully occupied in producing engines for cargo ships and
tankers and there are now over 1,600 of these built by
Doxford’s and their licensees. During this period of intense
production activity it was not possible to do much development
and the P type engine was developed later to be a
lighter engine occupying less space and of reduced cost and
improved reliability. During this period, continental firms
were developing larger engines and it was suggested that
Doxford’s were being left behind. The Doxford trade is
entirely in cargo ships and tankers up to a maximum of
18,000 tons, so that powers up to 7,000 h.p. have met the
total demand of the Doxford shipyard and there was a considerable
volume of opinion in the company that they should
make themselves thoroughly efficient for the building of
engines up to about 12,000 or 14,000 h.p. which formed their
requirements and indeed the requirements for 85 per cent, of
the market in direct drive marine Diesel engines of the world.
It was considered that there might be some difficulties in
developing opposed piston engines beyond these powers.
Doxford’s reputation and goodwill had been built upon this
type of engine and any different type would be an entirely new
beginning. However, during the past fifteen months negotiations
have been proceeding and subsequently a merger has
been concluded with the shipbuilding companies, in Sunderland,
of Sir James Laing and J. L. Thompson who are able
to build much larger ships, even of over 100,000 tons dead
weight. There is therefore now a market for a Doxford engine
of up to 20,000 or 25,000 h.p. in the Doxford group and such
a production will form a nucleus for the building of engines
for other shipbuilders. Under these circumstances the problem
of building a larger engine was again reviewed. Several
constructions of opposed piston engines were considered such
as double cylinder engines and engines with lever drives. It
was even considered whether a reliable Fullager engine could
be developed, embodying features of the P type engine.
The interesting design of double cylinder engine is shown
in Fig. 18. Eventually the design effort was concentrated on
an engine of 850 mm. bore and, unfortunately, much premature
publicity was given to this project. The design was based
on the P type engine, but as it progressed it became apparent
that the crankshaft would have to be of about 770 mm.
diameter in order to avoid critical speeds, in particular the
ninth order III node was very close to the running range of the
engine and was very powerful. Such a large crankshaft would
weigh over 196 tons. Fig. 19 shows the design of this crankshaft
and the design of the engine is shown later in Fig. 23. It
became apparent to the author that such an engine would not
be competitive and there would be difficulties in manufacture
due to the heavy parts involved. The estimated torsional
conditions are shown in Fig. 20.
Once again the situation was reviewed and it was even
considered whether to build engines of other types but it was
considered that the opposed piston engine, which is essentially
a British development for marine purposes, had many advantages
in through scavenge, a good combustion chamber,
exceptionally good balance and freedom from vibration and
was eminently suitable for welded construction, since the
entablatures and columns were not subjected to the combustion
loads which are carried entirely within the running gear
of the opposed piston engine. It became apaprent that a
design had to be found which would permit of more than six
cylinders so that an engine could be constructed with a much
lighter crankshaft and without fear from critical speeds. A
design was subsequently evolved, incorporating all the advantages
of features of the P type engine, but with even closer
cylinder centres which would permit of nine and ten-cylinder
engines being built with no serious critical speeds within the
running range and even without the necessity of a detuner.
A nine-cylinder engine of 760 mm. bore x 2,180 mm. combined
stroke is now being built to give 20,000 h.p. in nine
cylinders at the moderate mean indicated pressure of 135
lb./sq. in. and it is considered that this engine can be built
from four to maybe ten cylinders to give a range of powers
as shown in the table in Fig. 21. It is intended that the
engine range shall subsequently be extended to at least
25,000 h.p. and maybe even more, particularly so as the rating
of turbocharged engines is increased. The crankshaft of the nine-cylinder engine is of 635 mm. diameter on the crankpins
and is estimated to weigh no more than 125 tons, which will
compare very favourably with any of the large continental
engines now being built. An outline drawing of the engine
and of the engine range is shown in Fig. 21. It is one big
advantage of being later in developing a range of engines, in
that one can see what the competitor has built and can make
sure that the engine under design is ahead and will be competitive
in length, weight and cost. Table I (p. 11) shows the
competitive engines which have been developed to give 20,000
b.h.p. for comparison: —
It will be seen that the Doxford design compares very
favourably with these engines. In spite of having three cranks
to each cylinder the crankshaft of this engine will be as light
as any of the continental engines and as is well known the
crankshaft is by far the most costly component of the large
slow speed Diesel engine. Fig. 22 shows a photograph of a
wooden model of the nine cylinder engine and its compactness
and neat appearance will be apparent. In many respects it is
similar to the P type engine of 670 mm. bore described by
the author in a paper* read before this Institute in January
1961. The new engine is, however, even more compact and
has a more rigid although lighter crankshaft. It will be equally
accessible and it is believed will be of even lower cost. Parts
of this new engine are now in the Doxford works and it will
be running trials in the early spring of 1963. It is expected
that this engine will form the basis of the development of
Doxford engines up to the highest powers over the next
decade and powers of 27,000 h.p. and may be over 30,000 h.p.
are envisaged.
It is interesting to consider the development of Doxford engines over the past decade, either on the basis of output from a given size of cylinder or alternatively the sizes of the engines for a given horsepower. The output of the 670 engine in 1950 was 1,100 h.p. per cylinder. This engine was subsequently turbocharged to develop 1,400 h.p. per cylinder and in the P type engine it has not only been reduced in size but the output has been increased to 1,666 h.p. per cylinder.
The ratings and sizes are given in Table II. Fig. 24 shows the sizes of the Doxford engine for a power output of about 7,000 h.p. over the intervening years. The new engine now being developed will take the design a stage further and, apart from making very much larger powers available, its size is indicated for the power of 7,200 h.p. It will be appreciated that these stages in the design and application of the Doxford engine to the direct propulsion of ships has been achieved (a) by the turbocharging of the engine (b) by simplifying the mechanical design, particularly in regard to the crankshaft and (c) by closing the cylinder centres.
It is believed that this new engine following the P type engine will show that once again Doxford’s and British marine engineers are amongst the leaders in the development of large marine Diesel engines.
The author now wishes to acknowledge the help and encouragement which he has received from The Chairman and Directors of William Doxford and Sons Ltd., and in particular from the Managing Director, Mr. T. W. D. Abell, and also to acknowledge the assistance which he has had from his colleagues Mr. J. F. Butler, Mr. J. G. Gunn, Mr. E. Taylor and the Works Manager, Mr. G. A. C. Dun.
REFERENCE
1) Butler , J. F. 1962. “Some Points in the Development of the Direct Drive Turbocharged Opposed Piston Marine Oil Engine”, CIMAC Conference, Copenhagen June 1962, paper No. A3.