Trans.I.Mar.E., Vol. 100, pp.

145-156

Sequential turbocharging of the MTU 1163

engine
P. Herring, BSc, CEng, MIMarE
MTU

---------------- S Y N O P S I S -----------------
The 1163 engine series is the largest in M T U ’s range, having a sw ept volum e o f 11.63 litres!cylinder producing a
m axim um o f 370 kW /cylinder at 1300 revJm in. It is a 4-stroke engine available in 1 2 ,1 6 and 20 cylinder versions in
60° vee-form configuration. The latest design incorporates a tw o-stage ‘sequential’ turbocharging system with
m ultiple turbocharger groups m ounted on the engine w hich are sw itched in and out o f operation depending on engine
p o w er an d speed dem and. This new developm ent resulted in a p ow er increase o f over 40% due to increasing the brake
mean effective p ressu re fr o m approx. 21 bar to 30 bar. In addition, the engine operating range w as greatly expanded
w ith advantages in perform ance and fu e l consum ption, particularly in the part-load condition. Such engines are alm ost
exclusively used fo r the propulsion o f f a s t vessels w here pow er/w eight and pow er/bulk ratios are o f prim e importance.
The p o w er/w eight ratio has been increased by approx. 18% w hereas the engine overall dim ensions have actually been
decreased due to the com pactness o f the new turbocharging system. This p aper describes the design and developm ent
o f the 1163 series sequential turbocharging system and its im pact on the design o f m arine propulsion systems.

BACKGROUND
P. H erring jo in e d th e I.M ar.E . as a S tu d e n t M e m b e r in
1965 w hen ap p re n tice d as E n g in e e r C a de t w ith th e New
The 1163-02 engine series was originally conceived as the long-
Z e alan d S hipp ing Co. Ltd. A fte r stu d yin g M arine E n g i­
stroke version of the 956 series in order to compete in the neering at S u n d e rla n d P o lyte ch n ic/N e w ca stle U niversity
generator set market with a speed of 1200 rev./min for power he jo in e d M irrlees B la cksto ne (S tam ford ) as C ontracts
generation at 60 Hz. However, towards the end of the 1970s, it E ng in eer. In 1974 he m ove d to M A N -G H H Ltd., London,
was decided to develop this engine and introduce an uprated as T e ch n ica l S ales R e p re se n ta tive and b e ca m e Sales
version (1163-03) which could be used for the propulsion of fast M a n a g e r in 1976 re sp o n sib le fo r th e g ro u p ’s high-speed
vessels such as large patrol boats, corvettes and frigates. The m arine d ie se l in te rests, w h ich in clu d e d th e M TU engine
power increase itself was not a major problem. The difficulty lay range. S in ce 1981 he has b e e n re sid e n t in Fried-
in additionally providing a compact unit with good torque rich sha fe n em p lo ye d by M TU in th e S hip A p p lication
characteristics and favourable fuel consumption. The following D e pa rtm e nt, durin g w h ich tim e he w a s s e co n d e d fo r 16
m on th s to th e ISS in H a m bu rg to a ssist w ith th e Nato
conditions were determined for the development activities.
F riga te (N F R 90) F e a sib ility S tudy.
1. Simple design using proven components and as far as pos­
sible retaining the basic engine design.
2. Turbochargers and intercoolers to be compactly mounted on
engine.
3. Good starting and idling characteristics with low emission firing pressure. In the early stages of the design, it was necessary
values. to reduce the compression ratio to 8.5. Later, improvements in
4. Wide performance characteristics, i.e. high output torque the fuel injection system, which included modifying the air inlet
even at low engine speed. canals to create a swirl effect, permitted a shorter injection time
5. Rapid load acceptance. (30° crank angle instead of 50°; see Fig. 2) and a reduction in the
6. Good consumption figures, in particular in the part-load charge air pressure. This meant that the compression ratio could
range. be raised to 9.7 without increasing the firing pressure. The
7. No increase in the maximum firing pressure (approx. 150 bar) higher compression ratio has of course had a beneficial effect on
and therefore no increase in the associated mechanical the low-power running and starting characteristics of the engine.
stresses. Nevertheless, the measures originally introduced to overcome
Since the engine swept volume and piston speed were to remain the problems associated with low compression ratio have been
unchanged, the increased power could only be achieved by retained in the current design. These measures (cylinder cut-out
increasing the brake mean effective pressure (b.m.e.p.), which and cylinder charge air transfer) are described in detail in refs.
required a new design of turbocharging system. 1 and 2.

C O M PR ESSIO N RATIO T W O -STA G E TU R B O C H A R G IN G

It can be seen from Fig. 1, that the compression ratio must be In order to increase the mean effective pressure to 30 bar,
reduced if the b.m.e.p. is to be increased without increasing calculations had shown that a charge air pressure of around 5 bar

145
 P. Herring

(showing influence of injection time)

s u p e rc h a rg in g

1163-02 1163-03
p [kW/Zyl.] 260 370
(mHP/Cyl.) (354) (503)
20.6 29.4
B
ME
P (p
in
1 (299) (426)
n [rpm] 1300 1300

e (-) 12 9.7

x [-] 1.73 1.9

s Valve seat,
' cylinder head 260 245
\ Valve seat, 575 550
PC] va,ve
(3) Valve stem 260 250
a Cylinder head 290 280
H land

Fig. 4. Component tem perature com parison using the

same cylinder head

would be required. This exceeds the practical limit of single- closer to the surge line. The turbochargers therefore run ineffi­
stage turbocharging which lies in the region of 4 bar (Fig. 3), so ciently since the available energy in the exhaust gases cannot be
it was necessary to design a two-stage system with intercooling effectively utilized, consequently, engine torque is limited. The
and aftercooling. (Note: all pressures are absolute values.) Apart problem is accentuated the more the specific power of engines
from providing the necessary charge air pressure, two-stage is increased.
turbocharging enabled the air/fuel ratio (X) to be increased. This One solution is to mechanically vary the geometry of the
assisted in the reduction of cylinder-head temperatures as shown turbine nozzles and compressor diffuser to compensate for the
in Fig. 4 (see also ref. 2). variations in exhaust gas and air flow. This is limited in it’s
effectiveness especially in the low-load region because the form
of the turbine wheel and impeller cannot be similarly changed.
After exhaustive studies of various turbocharging possibilities
SEQUENTIAL TURBOCHARGING
and consideration of the ways other manufacturers tackled this
problem, the conclusion was reached that a most efficient
A wide performance characteristic is essential for a marine solution would be to use a number of small turbochargers which
propulsion engine, particularly to satisfy the arduous operating can be ‘sequentially’ switched in and out of operation depending
profdes of naval craft. In addition to providing adequate ship on engine speed and power demand. The system is in effect
speed and acceleration in various sea states, the propulsion analogous to varying the geometry of the turbine nozzles, but
engines should also be capable of supplying high torque at low without negatively affecting the efficiency, and is the only way
speed for loitering duty or sonar towing. A conventional turbo­ in which the turbine geometry can also be matched to the
charging system has only one optimum design point, namely at available exhaust flow. The sequential turbocharging principle
the engine’s maximum continuous rating. At part-load, the was first introduced on the single-stage turbocharged ‘538’
turbine nozzle area is too large for the reduced exhaust gas flow engine series (Fig. 5). The concept of sequential turbocharging
and temperature and the compressor operating point moves necessitated changing the basic turbocharging principle from a

146
 Trans.I.Mar.E., Vol. 100, pp. 145-156

Fig. 5. Single-stage sequential turbocharging

Fig. 6. Sequential turbocharging com pressor

characteristic with engine operating lines
‘pulse’ system to a ‘constant pressure’ system. Normally, MTU
16 V 538 TB 93/4 x ZR 170
diesel engines with conventional charging systems work on the
‘pulse’ principle, whereby the exhaust outlets from particular
cylinders are separately connected by short pipework to the
turbochargers. The arrangement is such that the peak pressure
pulses from the cylinders act in unison with the result that,
especially in the part-load range, the maximum available
exhaust energy can be utilized to drive the turbines. In this way,
the engine performance (fuel consumption as well as torque) is
optimized, but this becomes more difficult to achieve with
increasing b.m.e.p. The ‘constant pressure’ principle, whereby
the exhaust gases from all cylinders discharge into a common
manifold before entering the turbochargers, can be used to give
better efficiency at maximum power. However, with a conven­
tional turbocharger arrangement, the efficiency in the part-load
range is much worse than can be achieved with a ‘pulse’ system.
The sequential turbocharging system effectively overcomes this I Exhaust gas (flowing)

3 Exhaust gas (non-flowing)

disadvantage since the number of turbochargers in operation is Externally operated exhaust valve

always matched to the available exhaust gas flow, and therefore ] Air (flowing) Check flap

it always operates in a range of optimum efficiency, even at

Fig. 7. Two-stage sequential turbocharging
very low engine powers.
The turbochargers are switched in and out of operation by
flap valves installed upstream of the turbine and compressor.
chargers (h.p. and l.p.) are switched in and out of operation
The exhaust flap is pneumatically opened against spring pres­
(Fig. 7). Figs. 8 and 9 show the performance curves and tech­
sure slightly before the air flap opens, which permits the turbine
nical data of the 20 V 1163 TB 92 engine compared with the
to accelerate rapidly under no-load. The charge air pressure
20 V 1163 TB 93 engine with sequential turbocharging. The
between the non-return air flap and the compressor decreases,
expanded operating range and power increase of the ‘03 ’ version
causing the flap to open. To make certain the air flap opens, and
are clearly illustrated.
therefore ensure that turbine overspeeding cannot occur, a
mechanical linkage is arranged between the actuating cylinder
and the flap. The air flow provides resistance and retards turbine T urbocharger design
acceleration. This arrangement also ensures that the switching- The former 1163-02 engine series incorporated MTU type
in procedure is completed without loss of charge air pressure, ‘AGL’ turbochargers with axial-flow turbines and radial-flow
and that compressor surging does not take place. impellers. Normally, the overall efficiency of several small
The turbocharger switching-in point for increasing engine turbochargers would be less than one or two large turbochargers.
speed is higher than the switching-out point for decreasing However, this disadvantage is compensated by intercooling
engine speed (Fig. 6). This hysteresis prevents continuous between the h.p. and l.p. stages. The‘ZR ’-type turbochargers,
switching in/out of turbochargers at ‘borderline’ engine speeds. originally designed for the smaller engine series, 396 (Fig. 10)
The exhaust flap valves are loose-fitting to allow a small amount and 538, incorporate radial-flow turbines as well as radial-flow
of gas flow in the closed position. This keeps the turbochargers impellers. For the 1163-03 engine series, the existing model ZR
idling at operating temperature so that they are ready for imme­ 170 (impeller wheel diameter) was suitable for the high-pres­
diate operation and not subjected to thermal shock. sure stage, and a larger model, ZR 210 had to be designed for the
Having proved the sequential turbocharging principle on the low-pressure stage. The latter rotates in the opposite direction
538 engine series, it was a logical step to incorporate this in the to the former to enable the turbochargers to be arranged in
1163-03 design. The operating principle is identical, the only compact modular groups. Since the overall pressure ratio is
difference being that, in this case, groups of two-stage turbo- divided into two stages, with an intercooler between, the air

147
 P. Herring
8000 compressor operates sufficiently far away from the surge line at
kW To rq u e lim it c u rv e low air flow rates. With the sequential turbocharging system,
MCR
this design compromise is not so critical, since no individual
Cube
turbocharger is required to cover the whole engine operating
6000
TB 93:
range. More emphasis can therefore be given to designing for
120 % : 7400 kW
110%: 6735 kW
maximum efficiency. Due to the lower pressure ratios, the
100%: 6120 kW diffusers could be designed without vanes, which, together with
_
o 4000 the swept-back impeller vanes, results in a wider operating field.
S
o
CL

Turbocharger arrangem ent

2000
120%: 5200 kW
110%: 4735 kW
For compactness, a modular layout of turbochargers was
100%: 4300 kW developed, which could be adapted to all engines in the series
(12, 16 and 20 cylinder versions). Each module contains two
turbocharger pairs (h.p./l.p.), the turbines themselves being
800 1000 1200 rpm 1350 enclosed in a common water-cooled housing (Fig. 11). The
construction is neatly integrated with the engine and allows
Speed
optimum flow of gases and charge air. Apart from the compact
F ig . 8. 20 V 1163 T B 92 a n d 20 V 1163 T B 93 e n g in e arrangement, the installation of the turbines in a common water-
o u tp u t g ra p h
cooled housing has the advantage that the connections to the
exhaust manifold do not need to be gas-tight, and therefore
thermal expansion problems are eliminated. Also, the water-
jacket minimizes heat radiation to the surroundings and reduces
Sequential Turbocharging and Charge Transfer System
the surface temperature to meet the requirements of the Classi­
without with fication Societies.
No. o t BMEP O u tpu t W eight W eight/ BMEP O u tpu t W eight W e ig h t
The exhaust manifold housing in the engine vee is also
Power Power
c y l.
bar kW kg k g/kW bar kW kg k g /kW water-cooled, and is fabricated from cast steel sections which
C om p. R atio C omp. Ratio can be built up to suit the number of cylinders. The heat-resistant
1 1 6 3 -0 2 1 2 .0 :1 1 1 6 3 -0 3 9 . 7 :1
steel exhaust piping inside the housing is installed in short
20.6 3120 11400 3.7 29.4 4440 14200 3.2 sections which are not gas tight and serve only to provide an
16
20
20.6
20.6
4 1 60
5200
14350
17050
3.5
3.3
29.4
29.4
5920
7400
17400
20900
2.9
2.8 efficient flow path to the turbochargers. An air gap between the
■without Charge Transfer System
piping and housing reduces heat dissipation from the exhaust
gas. The high-pressure intercoolers (one per cylinder bank)
F ig . 9 .1 1 6 3 -0 3 c o m p a re d w ith 1163-02 s e rie s incorporate a preheating section integrated in the engine coolant
system. This is automatically actuated under low-load and low-
temperature conditions to raise the charge air temperature and
provide optimum combustion conditions.
Fig. 12 shows a cross-section of the 03 version compared
with the 02 version. The compactness of the turbocharger mod­
ules has even led to a reduction in height of the exhaust outlets.
All components are contained within the original envelope of
the 02 engine, with no off-engine mounted equipment. Fig. 13
shows the modular arrangement of turbochargers on the 12,16
and 20 cylinder engines, which have three, four and five turbo­
charger pairs (h.p./l.p.) respectively. Investigations have con­
firmed that a 24 cylinder engine is technically feasible by using
six pairs of turbochargers. This engine could be produced if the
market demands.

Turbocharging control and m onitoring

To ensure safe and correct operation of the turbochargers the
following parameters are monitored.
1. Rotational speed of each h.p. turbine
F ig . 10. E x h a u s t g a s tu rb o c h a r g e r ZR 170 2. Engine speed
3. Fuel rack position
4. Charge air pressure
5. Position of air intake flaps (‘open’ or ‘closed’)
temperature after compression is relatively low. The impeller The engine speed and fuel rack position are determined by an
wheels and diffusers of both h.p. and l.p. turbochargers could electronically controlled hydraulic governor. The electronic
therefore be made from lightweight aluminium alloy. Due to the control system is programmed to limit the fuel rack position
low moment of inertia of the impellers and turbines, the turbo­ depending on which turbocharger group is in operation (sensed
chargers can quickly accelerate and decelerate, thus improving by limit switches on the air intake flaps and the speed of the h.p.
the engine’s load-change characteristics. turbine).
With conventional systems the turbocharger must be desig­ The various parameters are processed in an electronic con­
ned, not only for maximum efficiency but also to ensure that the trol unit (ECS) mounted in the engine room control panel, which

148
 Trans.I.Mar.E., Vol. 100, p p .145-156
Intercooler depending on customer’s requirements. As an
example, the turbocharger control of a 20
cylinder engine intended for application in a
four-engine CODAD (Combined Diesel And
to afte rco o le r Diesel) installation will be described. The
resulting performance curves are shown in
Figs. 15 and 16.
The first charger group to function when the
engine is started, is the so-called ‘base’ group.
This is identical to other groups, except that the
exhaust and air flaps are locked in the ‘open’
position. The base group is therefore perma­
nently in operation. The second group to func­
tion is activated depending on engine speed
only, switching in at approx. 450 rev./min and
out at approx. 430 rev./min. When accelerating,
the third charger group can only be activated
when both the required engine speed (700 rev./
min) and charge air pressure (1.2 bar) are avail­
to afte rco o le r
able. The reason for this is to ensure optimum
acceleration and specific fuel consumption
when operating off the cube-law curve near the
maximum continuous rating (MCR) line. When
decelerating, only one parameter (either an
Fig. 11. Two-stage turbocharger arrangem ent, 1163-03 series

12 Cyl.-Engine

16-Cyl.-Engine

24-Cyl.-Engine

Fig. 13. Modular arrangem ent of turbochargers,

1163-03 series
Fig. 12. Com parison of engine sizes 1163-02/1163-03
engine speed of 670 rev./min or a charge air pressure of 0.65 bar)
must be fulfilled before the third charger group can be switched
controls the solenoid-operated pneumatic valves on the engine out of operation. When operating on the propeller curve, the
to open and close the exhaust and air flaps. Fig. 14 is a schematic charge air pressure is the over-riding parameter, and when oper­
of the ECS components and inter-related parameters that are ating near the MCR, it is the engine speed. The fourth and fifth
described in detail in ref. 3. The engine performance can be turbocharger groups are activated only depending on charge air
precisely matched to suit the engine application, and optimized pressure, switching inat2.4and3.1 bar, and switching out at 1.7
for low power running, high power running or fuel consumption, and 2.25 bar respectively.

149
P. Herring

MCR = MAXIMUM CONTINUOUS RATING CURVE

0 - ( 5 ) = NUMBER OF CHARGER GROUPS IN OPERATION

Fig. 16. Typical 20 V 1163 TB 93 perform ance

curves for CODAD operation

the low-load region, but has no influence on starting behaviour

or the maximum power available. A back-up system to take into
account failure of the compressed air supply is not considered to
be essential, but can be offered if specified by the customer. In
this case, alternative flap actuating cylinders are fitted that are
capable of operation by using charge air pressure as the working
medium.

IM PACT ON PR O PU LSIO N SYSTEM

The speed gradient and switching of turbochargers is con­ DESIGN
tinuously monitored to ensure overspeeding of the turbines does
not occur. If the speed tolerance band is exceeded, the fuel rack
travel is reduced and an alarm is initiated. In designing the se­ The uprated 1163-03 engine series has opened up new possibili­
quential turbocharging system, the highest priority has been ties for the ship designer, due to the increase in power/weight
given to safety, reliability and redundancy aspects. All sensors and power/bulk ratios and the improved torque characteristics.
monitoring the major operating parameters are duplicated, and The higher power density is a benefit not only for fast naval craft,
the components themselves are of simple, well-proven design. but also for yachts, hydrofoils and SWATH (small waterplane
The multi-turbocharger arrangement affords better redundancy area twin hull) vessels. The latter is illustrated in Fig. 17, which
compared with a conventionally turbocharged engine. Should shows that the space around the 20 V 1163 TB 93 engine is
failure of a charger unit occur, this can be cut out and the sufficient to undertake top overhauls of the engine in situ. The
sequencing system rearranged so that the engine is still able to outside diameter of only 3.8 m in relation to the length of the
develop approximately 80% power. The changeover procedure floating body does not increase the form resistance. The clear
can be effected by on-board staff within a short time. width of 1.8 m of the struts enables removal of the engine for
Should failure of the 24 V D.C. power supply for the elec­ main overhaul. Fig. 18 shows the propulsion system of an
tronic system occur, the engines can still run since the solenoid existing fast patrol boat, using four 16 cylinder MTU 956
valves controlling the air supply to the flap actuating mechanism engines, developing a total power of 13,200 kW. Using the
arc open in the de-energized position. The only effect is that the 1163-03 series, the same boat could now be powered by three 12
second charger group (normally activated depending on engine cylinder engines with consequent saving in weight and space.
speed only) is continuously in operation with the base group, The 1163-03 engine is ideally suited for the propulsion of
and the third group is activated only dependent on charge air medium-sized frigates.
pressure. This leads to a slight increase in fuel consumption in For higher power requirements, a combined system using

150
 Trans.I.Mar.E., Vol. 100, pp. 145-156
4-Shafter with 4 X 16 V 956 TB 91
Maximum Power: 4 X 3300 kW - 13 200 kW

Engine with
12 630 kg
Coupling 360 kg
Rev./Red. Gear
Total
2145 kg
15 135 kg
3 :
Total per Ship: 60 540 ka

3-Shafter with 3 X 12 V 1163 TB 93

Maximum Power: 3 X 4 440 kW - 13320kW

c le a r w id th Engine with
Accessories 13 905 kg
Coupling 550 kg
2 1 8 0 0 mm Rev./Red.Gear 3 745 kg
Total 18 200 kg

Total per ship: 54 600 ka

F ig . 18. P ro p u ls io n p la n ts fo r 400 t fa s t p a tro l b o a t

37.8

35 ■

MW

30-

m o u ld e d b e a m
2 3 8 0 0 mm

F ig . 17. C r o s s - s e c tio n t h r o u g h S W A T H s t r u t w ith 20 V

11 63 T B 93 (7 4 0 0 kW /1 0 0 0 0 h .p .)

100 140 180 rpm 220

diesels and gas turbines is often installed, the diesels for low-
speed and cruising duty because they have a lower fuel con­ Shaft speed

sumption than gas turbines operating at part load. The arrange­

ment is normally CODOG (Combined Diesel Or Gas) with the
full power of approx. 40,000 kW being delivered by two gas F ig . 19. C O D A G p r o p u ls io n p la n t w ith s in g le -s p e e d
turbines. Since the operating profile of such a frigate generally g e a rb o x fo r 40 00 t fr ig a te
results in very little running on gas turbines, a CODOG is not the
most cost-effective installation. With the 1163-03 series, a very
simple CODAG arrangement is now feasible, using two diesels, knots in the gas turbine mode, and over 31 knots in the combined
only one gas turbine and a single-stage gearbox. Formerly, mode. With modem electronic technology, the control and
CODAG systems were unpopular, mainly due to the fact that a monitoring of such a system presents no more difficulty than a
two-stage gearbox was necessary because of the limited torque CODOG system.
availability of the diesels. Fig. 19 shows a theoretical instal­ Another example which demonstrates the advantages of the
lation using two 24 V 1163 TB 93 diesels and one LM 2500 gas improved torque characteristic of the 1163-03 series, is shown
turbine to give a total power of 37,800 kW. The performance in Fig. 20. The resistance curve of a hydrofoil has a very
curve shows that a 4000 ton frigate could be driven at 23 knots pronounced ‘hump’ during the transition phase before the vessel
in the diesel mode with slight reduction of propeller pitch, 27 becomes completely foil-bome. In some cases, the torque

151
 P. Herring

Design point with

140 normal turbocharging
% t C O N CLU SIO N S
/
k*’^
Design point with
sequential turbocharging

Reserve for transition The introduction of the sequential turbocharging principle

to foilborne mode
»• Torque requirement
represents a major step forward in compact diesel engine tech­
f of hydrofoil craft nology. The development of higher-powered engines that can
« A 'y be installed in even less space is a challenge for engine designers
/
f f - even more so when this is to be achieved without degrading the
/ V / torque characteristics, the fuel consumption or the time between
overhauls. In-field experience with many engines testifies to the
y success of this system, so much so that nowadays all the latest
/ { engine models in MTU’s range (with the exception of the small
099 and 183 series) incorporate sequential turbocharging using
/ the ‘ZR’ series of turbochargers. Combined with the most
modem microprocessor technology, the operating characteris­
/ tics of these engines can be ‘tailor made’ to suit the intended
application.
30 40 50 60 70 80 9 0 % 100

Engine speed

Fig. 20. Influence of sequential turbocharging on

hydrofoil propulsion
REFERENCES

requirement at this point is even higher than at maximum 1. ‘Further Development of the MTU 956/1163 Series Engines' by Dr.-
speed, which previously meant that relatively large engines Ing. Hans Dinger, Dipl.-Ing. Herbert Deutschmann, CIMAC- 14th
International Congress on Combustion Engines, Helsinki (June
had to be installed, the maximum power of which could never
1981).
be utilized. 2. ‘MTU’s Advanced Marine Engine Design and Applications’ by Dr.-
The example shows that a hydrofoil previously powered by Ing. Hans Dinger, SAE Government / Industry Meeting and Exposi­
two 20 cylinder 1163-02 engines can now be powered by two 12 tion, Crystal City, Washington D.C. (May 1985).
cylinder 1163-03 engines. Apart from the initial cost savings, 3.‘Improvement of Engine Performance by Microprocessor Controlled
the consequent weight and space saving could be used for Electronics’ by Dr. R. Schulmeister & F. Butscher, CIMAC - 16th
additional payload. International Congress on Combustion Engines II, Oslo (June 1985).

152
 Trans.I.Mar.E., Vol. 100, pp. 145-156

Discussion

S. G. CHRISTENSEN, C.Eng., F.I.Mar.E., B.Sc.: My contri- ship of hertzian pressure and fuel-pump cam life are known, one
bution to this very interesting paper is more a series of questions may ask, why was the fuel-pump cam not designed in this
made in an endeavour to learn more about this little publicized manner from the start? Does cam geometry have any signifi­
engine of high output, perhaps more importantly, where this cance on hertzian stress in high-speed engines where the fuel
high power output has been squeezed into what must be cylin­ injection pump parts appear to be much smaller than in low-
ders of relatively small linear dimensions. speed and medium-speed engines? Can Mr. Herring tell us if the
Many of us formulate an idea of an engine by knowing its fuel-pump cam was in fact altered, and if any other changes were
bore, stroke, and rev./min; the figures for the bore and stroke are made in the injection system, such as an alteration in nozzle hole
missing both in the synopsis and in the paper. A figure for piston diameter, streamlining of the injector nozzle holes by chamfer-
displacement given in some cubic measurement without a bore/ ring or radiusing at the sac end of the hole to improve the
stroke ratio is not enough. From the data given we start by coefficient of discharge, and/or alterations in piping elasticity
thinking the cylinder must over-square, the rest is conjecture. and the like?
The note on the reduction of the compression ratio is Another interesting point on this diagram is the negative
interesting, it was noticed in a later part of the paper that the slope shown on the rate of cylinder pressure rise for the 30
fuel injection period had been reduced from 50 degrees to 30 degree injection period. This occurs just before ignition, but at
degrees. Fig. 2 shows the injection period finishing earlier for the point in the piston stroke where it is still rising ‘slowly’. It is
the 30 degree period than for the 50 degree period. Can the not know if this curve is based on a mathematical model for
author give us some idea of the combined effect of the change finding the rate of heat release or whether it has been obtained
of compression ratio, the effect of earlier injection completion from the output of a piezoelectric transducer fed into a computer
on the expansion ratio, and the value of its ultimate effect on the and then plotted after the cylinder pressure value has been
specific fuel consumption? divided by time. It might be expected to have some flattening-
It is thought to be an excellent idea to keep the turbocharger off effect of the rate of cylinder pressure rise due to the heat
rotor spinning when it is in a non-operative condition and the requirements for fuel particle surface vaporization. Perhaps Mr.
engine is operating with a low-power output. Herring can tell us more about these points.
It has been noticed on some large medium-speed engines Turbocharging and the scantlings of crankshafts have an
kept in a state of immediate readiness for emergency duty that indirect relationship due to the increase in the mean effective
delay in rotor bearing lubrication on start-up led to early rotor pressure. The increase here is from approximately 21 bar to 30
bearing problems. The bearings were of the sleeve-type. This led bar. Maintaining the maximum combustion pressure at its
to the necessity of supplying lubricant continuously to the rotor original value will hold the range of alternating stress at about
bearings when the engine was shut down; a secondary problem the same value as before. Obviously there are other unknown
then arose with the rotor bearing sealing devices due to lubricant factors involved in this, but was it necessary to upgrade the
leakage into the other parts of the turbocharger. These problems crankshafts in any way by changes in dimensions, by material
were finally overcome by replacing the sleeve bearings with ball changes allowing higher working stresses, or by changes in
and roller bearings. surface hardening techniques, so giving a higher fatigue resis­
The paper states rotors are kept spinning to prevent thermal tance?
shock when the turbocharger is brought into use. Were there any It is noticed in the multi-engined geared propulsion drivers
other reasons for keeping the rotor spinning, such as problems that controllable-pitch propellers are used in the machinery
similar to those just mentioned? installations shown. With such arrangements, any engine may
It is noticed in Fig. 10 showing the ZR 10 turbocharger that be operated under its designed conditions of power output and
sleeve- or journal-type bearings are used. Is there any reason for engine speed when operating alone. Has anybody considered
the use of this type of bearing in preference to anti-friction or using solid or fixed-pitch propellers with multi-engined geared
ball-and-roller bearings? Ball and roller bearings may be oper­ propulsion installations and MTU 1163 engines? With such
ated at a much higher temperature and therefore would appear to propellers the cylinder pressures will usually extend beyond the
be safer. Cost and ease of availability is an obvious factor in the crankshaft design limits when operating on only one or less than
choice of sleeve bearings. Were there any other reasons? the total number of installed engines unless the engine power is
Referring again to Fig. 2, detailing the improved combustion considerably reduced. Will the engine surveillance system and
process, although the paper is on sequential turbocharging, the control equipment take care of such an operation and prevent the
improvement in the combustion process raises many points of crankshaft from being overloaded?
interest. It appears to be almost a paradox that one can increase On a final note, Mr. P. Herring and his company must be
the rate of injection by approximately 70% without increasing thanked for this interesting paper with the compliment that only
the maximum combustion pressure, even though some compen­ such an interesting paper will elicit so many questions.
sation is obtained from retarding the start of injection. Examina­
tion of the injector needle lift plotted on the time base shows the H. NIVEN, B.Sc.Eng., C.Eng., F.I.Mech.E. (Ricardo Consult­
lift period has been increased from a little more than 1 ms to ing Engineers pic): I note that on the 03 version of the engine a
something slightly in excess of 1.5 ms. This obviously has some heating section is incorporated into the h.p. intercooler, which is
bearing on the improved combustion. When the slope of the plot activated when the boost pressure drops below 1.6 bar. I
of the injector needle lift is examined, it appears that the presume that this is to help maintain regular and stable combus­
acceleration of the fuel pump ram has been considerably re­ tion at low load, and to control the rate of pressure rise within the
duced, this then allows the supposition that something akin to a cylinder.
form of pilot injection has been obtained, thus improving the Do you know what rate of pressure use (bar/crank degree)
overall combustion. When problems arising out of the relation­ you achieve at light load to avoid detonation damage to cylinder

153
P. Herring
components? We like to see a rate of pressure rise value of not It is a fact that there are, currently, still differences that make
greater than 10 bar/crank degree. We have measured values in CODOG and CODAG systems attractive, and the author de­
excess of 30 bar/crank degree under part load, part speed scribes an interesting CODAG concept in his paper. I agree with
conditions with very low air manifold temperatures, and such him that CODAG makes economic sense when high-powered
high values inevitably lead to detonation damage to the cylinder diesel cruise engines are installed - it seems wasteful not to use
components. all the available power for the ship’s full-speed condition.
Is the injection timing fixed, or do you have a light-load However, I believe he dismisses the possibility of two-speed
retard helix on the fuel pump plunger? gearboxes too lightly. In his example, a two-speed gear would
Lastly, could you tell us what the valve overlap is please? permit the ship to do something like 25 kt on diesels before
needing to start the gas turbine, and I do not think the system
J. M. THOM PSON (YARD Ltd.): The basic concept of two- controls would be any more complex than those required to
stage turbocharging to obtain the highest power from a mini­ handle his propeller pitch scheduling. Admittedly an additional
mum size and weight of engine is not new, but the number of primary gear reduction would be needed, but only a small
designs which have been developed to the point where they can number of components similar to those used in the CODOG
be offered for general sale are very few. The author has system would be required, as shown in the example in the
described the only system of which I am aware that successfully diagram which shows one shaft set of machinery. I would be
combines high power with a good operating envelope, low interested in the author’ comments.
specific fuel consumption throughout the operating range, and
straightforward installation. The long list of sales described
during the presentation is indicative of the good recepLion this
somewhat specialized engine has obtained.
Author’s reply----------------------
The only reservations I have are on the reliability and
durability of the engine, particularly if full use is made of its In answer to S. G. Christensen: I would like to thank Mr.
capabilities. As others may share my concern I would be Christensen for his keen interest in this paper, which is reflected
interested to have the author’s comments on the following in his in-depth questions.
points. The 1163 series has a bore/stroke ratio of 230/280 mm
(a) From Fig. 15, there is clearly a fuel consumption advantage compared with 230/230 for its twin, the 956 series.
in operating at or close to the left-hand boundary of the The problem of lubricating oil supply to the turbocharger
engine envelope, and indeed the author recommends this bearings of quick-starting emergency generating sets is appreci­
procedure in his description of a CODAG plant. As this is ated. For such an application the 956/1163-03 engines are
normally an area where high mechanical and thermal loads continuously lubricated during the shutdown period, but a
are encountered: pressure control valve limits oil flow to the rotor bearings. When
(i) what parameter(s) set the limit of operation the engine is running, rotor sealing of all turbochargers (active
shown; and inactive) is supplemented by pressurizing the sealings with
(ii) what is the effect on bearings, exhaust valves, charge air.
etc. of operating in this way compared with The reasons for keeping the inactive rotors spinning are
operation on the cube law? mainly to avoid thermal shock, and to permit rapid acceleration
(b) The ‘base’ turbocharger group is always in use, whereas the when switched into operation.
last group comes into use only at very high power and the During the engine design phase, MTU chose ‘floating bush’-
remaining groups have intervening amounts of operation. type bearings since at the time, ball bearings could not satisfy the
Do MTU recommend periodically resequencing the order of technical requirements. The main advantages were:
operation to average out usage or are different turbocharger good vibration and shock damping characteristics due to the
maintenance intervals recommended according to the preset double oil film;
sequence and operating profile? long life;
high reliability;
J. TURVEY [Vospcr Thomycroft (U.K.) Ltd.]: In the presen­ fine oil filtration unnecessary (30 mm adequate);
tation the author mentioned white smoke, would he please cost and availability.
elaborate on and quantify the smoke situation, both white and Most axial turbochargers utilize ball bearings but in this
black? case, the heat source (turbine rotor blades) is more remote from
the bearings. With radial turbochargers, the closer proximity of
M. J. NEEVES (Rolls-Royce pic): I would like to congratulate the heat source can present a problem for ball-bearings.
the author on a most interesting paper. It is fascinating to see the Mr. Christensen has read a great deal into Fig. 2, which is
areas of similarity in the way that high-performance diesels and really only intended to illustrate the principle effects. The
marine gas turbines are developing. The maintenance philoso­ improvement in the combustion process is largely due to the
phy is similar with modular exchange in gas turbines corre­ delayed opening of the injector (the nearer TDC the better). The
sponding to top overhauls in diesels and with removal for major needle does in fact close earlier but this does not greatly
overhaul recommended for both types of machine. Modern influence the end of the combustion phase so the expansion ratio
electronics technology, as the author points out, makes possible is hardly affected. The delayed injector opening enabled the
much more sophisticated controls and monitoring, whichever compression ratio to be increased, and the more efficient com­
propulsion system is chosen. There are still significant areas of bustion (also influenced by air intake swirl) resulted in a slight
difference however. The gas turbine technologists are trying to improvement in specific fuel consumption.
match diesel fuel consumption characteristics with advanced The Figure unfortunately gives the false impression that the
cycle machines and the diesel designers, as described here by rate of injection has been increased. The slope of the lines
the author, are trying to approach the favourable torque charac­ indicating injector needle opening should be almost identical
teristic of gas turbines. and no form of pilot injection is intended. The maximum

154
 Trans.I.Mar.E., Vol. 100, pp.145-156
combustion pressure has not been increased due to the retarded operation at lower engine speeds in favour of the reliability, in
start of injection, shorter injection time and improved combus­ spite of the fact that higher torque easily can be achieved. The
tion process. During the initial stages of the development, the guaranteed TBO for a particular engine is dependent on the
existing fuel pumps, injectors, cams, etc. from the former ‘02’ expected operating profile and is based principally on the life of
engine version were used. The required increase in output was the con-rod bearings. In Fig. 15, one engine/shaft operation on
reached, but the engine could be improved, so new equipment the MCR curve will reduce the bearing life by 50% compared
was designed. The injection pump bore was increased from with operation at the same power on the cube law. However, the
22 mm to 24 mm and total stroke (cam lift) increased from resulting bearing life is still longer than the guaranteed engine
15 mm to 24 mm. The nozzle and needle design were also TBO. In service, extended operation on one engine/shaft is
modified and max. injection pressure increased from 1200 to usually carried out at 80% MCR, i.e. in Fig. 16, approximately
1450 bar. The cam and roller geometry is independent of engine 3600 kW/730 rev. per min, in which case the bearing life is
speed, and is determined principally by the maximum injection more than doubled (15,000 hours). Apart from the effect on
pressure. The elasticity of the thick-walled piping is negligible. bearings, operation on the MCR curve also results is approx.
If the cylinder compression (without combustion) were 50°C higher exhaust gas temperatures compared with those at
illustrated, the curve would be symmetrical, peaking at TDC or the same power on the cube law.
at approx. 100 bar. The associated pressure build-up rate would The question regarding resequencing the charger groups to
be similar to a sine wave intersecting the x-axis at TDC. The even-out the running hours of the respective chargers is often
negative slope shown in the curve for the 30° injection period is posed. While this is indeed feasible, it is unnecessary since the
therefore evident due to the retarded start of the combustion turbocharger components and bearings are so designed that
phase. The illustrated curve is derived from data obtained using virtually no wear takes place. The thermal and mechanical
a piezoelectric transducer and no appreciable flattening-off loading is such that the stresses lie well below the LCF (low-
effect has been observed due to fuel particle surface vaporiza­ cycle fatigue) limit of the material. Pre-lubrication prevents
tion. wear of the turbocharger bearings during the engine starting
No modification to the crankshaft was necessary since the sequence.
design process already takes into account torsional vibration In answer to J. T urvey: When the engine is operated at very
stresses, which, particularly in the case of one cylinder mis­ low load, white smoke tends to be produced due to incomplete
firing, are far higher than the increased stress due to the uprated combustion. This occurs because of poor atomization of the fuel
output. combined with a low compression temperature.
Controllable pitch propellers allow flexibility of operation, If the 20 V 1163 TB 93 would be allowed to idle (with clutch
fast manoeuvring and are especially suitable for the varied disengaged) with all 20 cylinders firing, the exhaust would have
operating profiles of modem warships. Fixed-pitch propellers a hydrocarbon concentration of approx. 1000 p.p.m. With clutch
do not offer such flexibility and demand either in reversing engaged operating at 350 rev./min on the cube law, the hydro­
gearboxes or reversing engines for astern operation. When, in a carbon concentration is approx. 650 p.p.m.
multi-engined installation, one engine/shaft operation is called In order to reduce the white smoke emission to below the
for, the engines must produce a higher torque for a given speed. visibility limit of approx. 500 p.p.m., the engine is operated in
Either for economical reasons (optimum SFC and propeller the idling condition with one cylinder bank cut out. The non­
efficiency) or to minimize underwater noise, the CP installation firing cylinders act as a compressor pumping charge air into the
will use full pitch. In this respect, the problems of one engine/ firing cylinders thereby increasing the firing pressure and tem­
shaft running are the same for CP or fixed-pitch propeller perature (charge air transfer system).
installations. For best performance, both demand a diesel engine In addition, the combustion conditions are further improved
which can produce high torque at low speed. Control and by raising the temperature of the charge air in the heating section
monitoring of engine speed and power is necessary to prevent of the h.p. intercooler. By these means, the hydrocarbon concen­
overloading. Modem electronic control systems enable ship tration is reduced below the visibility limit.
acceleration and deceleration to be programmed (for CPP and Formerly, the cylinder bank cut-out system was only avail­
FPP systems) to ensure optimum operation under all conditions. able for idling when the clutch was disengaged. This meant that
The MCR and DBR (torque limit) curves shown in Figs. 8, 15 for engine operation in the lower part-load range, the hydro­
and 16 are also monitored to safeguard the engine. carbon concentration could exceed the visibility limit in the
In answer to H. N iven: Heating of thecharge air using engine low-speed range of 350-550 rev ./min.
cooling water at boost pressures below approx. 1.6 bar is The present engines are now equipped with a system allow­
necessary to ensure stable combustion and hence minimize ing extended periods of operation under low-load conditions,
white smoke emission at low load. It also reduces the rate of incorporating a flap valve which shuts off the charge air supply
pressure rise within the cylinder. to the non-firing cylinder bank and at the same time, allows air
A rate of pressure rise of 10 bar/crank degree seems some­ to be drawn from the engine room. The firing cylinders therefore
what conservative. MTU has measured values of up to 15 bar/ receive more charge air and the compression temperature is
crank degree without observing any detrimental effects. increased.
A degree of light-load retard helix is provided on the fuel Black smoke is produced when imperfect combustion is
pump plunger. This is not so pronounced since the same plunger caused by a deficiency of air, i.e. the air/fuel ratio or distribution
is used for the twin 956 engine series (which, being used mainly is no longer optimized. This is normally a transient condition
for constant speed, genset application, actually requires no which occurs when the engine is required to accelerate quickly.
retard helix), hence the design is necessarily a compromise. To minimize the emission of black smoke, the control
The valve overlap is 80°. system for the 1163-03 engines includes an electronically con­
In answer to J. M. Thompson: Mr. Thompson is correct in trolled fuel rack limiting device which is activated depending on
pointing out that operation to the extreme left of the propeller charge air pressure. The maximum engine acceleration rate is
curve results in higher mechanical and thermal loading. For this only marginally limited as the injected fuel is metered according
reason, MTU has reduced the torque-limit line for steady-state to the available transient combustion air flow. Only the fuel

155
 P. Herring
surplus is cut off which would otherwise result in black smoke. seems to me to be preferable. The changeover from diesel drive
At low power the Bosch no. is approx. 1.0 and at MCR to CODAG and vice versa can then be accomplished without
approx. 0.4. declutching the diesels.
In answer to J. Neeves: I am pleased that a representative of Fig. 19 shows an example of one shipset of machinery
the gas turbine industry also supports the CODAG concept. The incorporating one gas turbine and two diesels, driving two
benefits compared with a CODOG system are so attractive it is shafts. For higherpower (as required for warships up to say 6000
strange that the system is not more widespread. I agree that a t) a CODAG arrangement could be envisaged using two separate
CODAG system with a two-speed gear involves no greater shaftsets, each incorporating one gas turbine and one diesel.
complexity than a CODOG system, however, if the operating In conclusion, I would like to express my appreciation for the
requirements can be satisfied with a single-stage gear, this interest shown in this paper.

156