Introduction‘ to
`
`INTERNAL
`
`ENGINES
`
`COMBUSTION
`
`SecondEdz'tz'on
`
`BMW1045
`Page 1 of 32
`
`

`

`Library of Congress Cataloging-in-Publication Data
`Stone, Richard, 1955-
`Introduction to internal combustion engines/ Richard Stone. -
`2nd ed.
`p. cm.
`Includes bibliographical references (p.
`ISBN 1- 56091- 390-8: $39.00
`l. Internal combustion engines.
`TJ755.S87 1993
`621 .43--dc20
`
`) and index.
`
`I. 7itle.
`
`93-14970
`CIP
`
`Copyright © Richard Stone 1985, 1992
`
`Third printing 1995
`Published in the United States of America 1993 by
`Society of Automotive Engineers, Inc.
`400 Commonwealth Drive,
`Warrendale, PA 15096-000 I
`
`All rights reserved. Printed in Hong Kong
`
`Permission to photocopy for internal or personal use, or the internal or
`personal use of specific clients, is granted by SAE for libraries and other
`users registered with the Copyright Clearance Center (CCC), provided
`that the base fee of $.50 per page is paid directly to CCC, 27 Congress
`St., Salem, MA O 1970. Special requests should be addressed to the SAE
`Publications Group. l-56091-390-8/93 $.50.
`
`SAE Order No. R- 129
`
`BMW1045
`Page 2 of 32
`
`

`

`9 Turbocharging
`
`9 .1 Introduction
`
`Turbocharging is a particular form of supercharging in which a compressor
`is driven by an exhaust gas turbine. The concept of supercharging, supply(cid:173)
`ing pressurised air to an engine, dates back to the beginning of the century.
`By pressurising the air at inlet to the engine the mass flow rate of air
`increases, and there can be a corresponding increase in the fuel flow rate.
`This leads to an increase in power output and usual1y an improvement in
`efficiency since mechanical losses in the engine are not solely dependent on
`the power output. Whether or not there is an improvement in efficiency
`ultimately depends on the efficiency and matching of the turbocharger or
`supercharger. Turbocharging does not necessarily have a significant effect
`on exhaust emissions.
`Compressors can be divided into two classes: positive displacement and
`non-positive displacement types. Examples of positive displacement com(cid:173)
`pressors include: Roots, sliding vane, screw, reciprocating piston and
`Wankel types; some of these are shown in figure 9.1. The axial and radial
`flow compressors are dynamic or non-positive displacement compressors
`see figure 9.2. Because of the nature of the internal flow in dynamic
`-
`compressors, their rotational speed is an order of magnitude higher than
`internal combustion engines or positive displacement compressors. Conse(cid:173)
`quently, positive displacement compressors are more readily driven from
`the engine crankshaft, an arrangement usually referred to as a 'super(cid:173)
`charger'. Axial and radial compressors can most appropriately be driven by
`a turbine, thus forming a turbocharger. Again the turbine can be of an
`axial or radial flow type. The thermodynamic advantage of turbochargers
`over superchargers stems from their use of the exhaust gas energy during
`blow-down, figure 2.5.
`A final type of supercharger is the Brown Boveri Comprex pressure
`wave supercharger shown in figure 9.3. The paddle-wheel type rotor is
`driven from the engine crankshaft, yet the air is compressed by the
`
`324
`
`BMW1045
`Page 3 of 32
`
`

`

`TURBOCHARGING
`
`325
`
`Roots blower
`
`Vane compressor
`
`Screw compressor
`
`Figure 9.1 Types of positive displacement compressor (reproduced from Allard
`(1982), courtesy of the publisher Patrick Stephens Ltd)
`
`pressure waves from the exhaust. Some mixing of the inlet and exhaust
`gases will occur, but this is not significant.
`The characteristics of turbochargers are fundamentally different from
`those of reciprocating internal combustion engines, and this leads to
`complex matching problems when they are combined. The inertia of the
`rotor also causes a delay in response to changes in load -
`turbolag.
`Superchargers have the added complication of a mechanical drive, and the
`compressor efficiencies are usually such that the overall economy is re(cid:173)
`duced. However, the flow characteristics are better matched, and the
`transient response is good because of the direct drive. The Comprex
`
`BMW1045
`Page 4 of 32
`
`

`

`326
`
`INTRODUCTION TO INTERNAL COMBUSTION ENGINES
`
`.
`
`D
`
`~c=r
`I
`---nr.,,~-n IL - ____,4
`-,
`
`Axial compressor
`
`Radial compressor
`
`Figure 9.2 Types of dynamic or non-positive displacement compressor
`(reproduced from Allard (1982), courtesy of the publisher Patrick
`Stephens Ltd)
`
`9
`
`Figure 9.3 Brown Boveri Comprex pressure wave supercharger. (a) Engine;
`(b) cell-wheel; (c) belt drive; (d) high-pressure exhaust; (e) high(cid:173)
`pressure air; (f) low-pressure air; (g) low-pressure exhaust
`
`BMW1045
`Page 5 of 32
`
`

`

`TURBOCHARGING
`
`327
`
`supercharger absorbs minimal power from its drive, and has a good
`transient response; but it is expensive to make and requires a drive. The
`fuel economy is worse than a turbocharger, and its thermal loading is
`higher. The development theory and application of the compress super(cid:173)
`charger are covered in the Brown Boveri Review, Vol. 7 No. 8, August
`1987.
`Comprex superchargers have not been widely used, and superchargers
`are used on spark ignition engines only where the main consideration is
`power output. Turbochargers have been used for a long time on larger
`compression ignition engines, and are now being usea increasingly on
`automotive compression ignition and spark ignition engines.
`Compound engines are also likely to gain in importance. A compound
`engine has a turbine geared to the engine crankshaft, either the same
`turbine that drives the compressor or a separate power turbine. The
`gearing is usually a differential epicyclic arrangement, and if matching is to
`be optimised over a range of speeds a variable ratio drive is also needed.
`Such combinations are discussed by Wallace, F. J. et al. (1983) and by
`Watson and Janota (1982). Compound engines offer improvements in
`efficiency of a few per cent compared with conventional turbocharged
`diesel engines.
`Another development that is most relevant to turbocharged engines is
`the low heat loss (so called 'adiabatic') Diesel engine.
`In a naturally aspirated engine, the higher combustion chamber tem(cid:173)
`perature will lead to a reduction in the volumetric efficiency, and this will
`offset some of the gains from the increased expansion work. The fall in
`volumetric efficiency is less significant in a turbocharged engine, not least
`since the higher exhaust temperature will lead to an increase in the work
`available from the turbine. This is discussed further at the end of section
`9.3.
`Commercial and marketing factors also influence the use of turbo(cid:173)
`chargers. A turbocharged engine will fit in the existing vehicle range, and
`would not need the new manufacturing facilities associated with a larger
`engm.e.
`Allard (1982) provides a practical guide to turbocharging and super(cid:173)
`charging, and Watson and Janota (1982) give a rigorous treatment of
`turbocharging. The remainder of this chapter is devoted to turbocharging,
`and there is also a case study that considers a turbocharged engine at the
`end of chapter 10 on Computer modelling. Computer modelling is of great
`value in establishing the performance of turbocharged engines, since the
`complex performance characteristics of turbochargers make it difficult to
`predict their performance in conjunction with an engine.
`
`BMW1045
`Page 6 of 32
`
`

`

`328
`
`INTRODUCTION TO INTERNAL COMBUSTION ENGINES
`
`9.2 Radial flow and axial flow machines
`
`The turbomachinery theory applied to turbochargers is the same as for gas
`turbines, and is covered in books on gas turbines such as Harman (1981) ,
`and in books on turbocharging, see Watson and Janota (1982). As well as
`providing the theory, gas turbines also provided the materials technology
`for the high temperatures and stresses in turbochargers. Provided that the
`turbocharger is efficient enough to raise the engine inlet pressure above the
`exhaust pressure of the engine, the intake and exhaust processes will both
`benefit. This is particularly significant for engines with in-cylinder fuel
`injection ( since unburnt fuel will not pass straight through the engine), and
`for two-stroke engines (since there are no separate induction and exhaust
`strokes).
`The efficiency of turbines and compressors depends on their type ( axial
`or radial flow) and size. Efficiency increases with size, since the losses
`associated with the clearances around the blades become less significant in
`large machines. These effects are less severe in radial flow machines, so
`although they are inherently less efficient than axial machines their relative
`efficiency is better in the smaller sizes.
`Compressors are particularly difficult to design since there is always a
`risk of back-flow, and a tendency for the flow to separate from the blades in
`the divergent passages. Dynamic compressors work by accelerating the
`flow in a rotor, giving a rise in total or dynamic pressure, and then
`decelerating the flow in a diffuser to produce a static pressure rise. Radial
`compressors are more tolerant of different flow conditions, and they can
`also achieve pressure ratios of 4.5:1 in a single stage; an axial compressor
`would require several rotor/stator stages for the same pressure ratio.
`A typical automotive turbocharger is shown in figure 9.4, with a radial
`flow compressor and turbine. For the large turbochargers used in marine
`applications, the_ turbine is large enough to be designed more efficiently as
`an axial flow turbine -
`see figure 9.5.
`The operation of a compressor or turbine is most sensibly shown on a
`temperature/entropy (T-s) plot. This contrasts with the Otto and Diesel
`cycles which are conventi~nally drawn on pressure/volume diagrams. The
`ideal compressor is both adiabatic and reversible and is thus isentropic -
`a
`process represented by a vertical line on the T-s plot, figure 9.6. The suffix
`s denotes an isentropic process. Real processes are of course irreversible,
`and are associated with an increase in entropy; this is shown with dotted
`lines on figure 9.6. Expressions for work (per unit mass flow) can be found
`using the simplified version of the steady-flow energy equation
`
`BMW1045
`Page 7 of 32
`
`

`

`TURBOCHARGING
`
`329
`
`Exhaust.),
`
`gas out \.J
`
`A~ir
`\.J•n
`
`Key: 1 Compressor wheel. 2 Turbine wheel. 3 Bearing housing. 4 Bearing. 5 Shaft. 6 seal ('O'
`ring). 7 Mechanical face seal. 8 Piston ring seal. 9 Turbine housing. 10 Compressor housing.
`11 'V' band clamp.
`
`Figure 9 .4 Automotive turbocharger with radial compressor and radial turbine
`(reproduced from Allard (1982), courtesy of the publisher Patrick
`Stephens Ltd)
`
`and, since the processes are treated as adiabatic
`
`No assumptions about irreversibility have been made m applying the
`steady-flow energy equation; thus
`
`turbine work, Wt = h3 -
`
`h4
`
`(9.1)
`
`and defining compressor work as a negative quantity
`
`For real gases: enthalpy is a strong function of temperature and a weak
`function of pressure. For semi-perfect gases, enthalpy is solely a function
`of temperature, and this is sufficiently accurate for most purposes. Thus
`
`BMW1045
`Page 8 of 32
`
`

`

`330
`
`INTRODUCTION TO INTERNAL COMBUSTION ENGINES
`
`a)
`
`b)
`
`Figure 9.5 Marine turbochargers with radial compressors and axial turbines.
`(a) Napier; (b) Elliot (with acknowledgement to Watson and Janota
`(1982))
`
`BMW1045
`Page 9 of 32
`
`

`

`TURBOCHARGING
`
`331
`
`Temperature
`T(K)
`
`I
`
`We
`
`2s
`
`I
`' 1
`
`\
`\
`
`p4
`
`wt
`
`4
`
`P1
`
`Figure 9 .6 Temperature/entropy diagram for a turbocharger
`
`Entropy, s
`
`and
`
`(9.2)
`
`(9.3)
`
`where cP is an appropriate mean value of the specific heat capacity. The
`mean specific heat capacity can be evaluated from the information in
`chapter 10, section 10.2.2, and such an exercise has been conducted in
`producing table 13.1. Consequently the T-s plot gives a direct indication of
`the relative compressor and turbine works.
`This leads to isentropic efficiencies that compare the actual work with
`the ideal work.
`
`BMW1045
`Page 10 of 32
`
`

`

`332
`
`INTRODUCTION TO INTERNAL COMBUSTION ENGINES
`
`compressor isentropic efficiency,
`
`and turbine isentropic efficiency,
`
`'Y/c =
`
`'Ylt =
`
`h2S - h1
`h2 - h1
`
`h3 - h4
`
`h3 - h4S
`
`=
`
`=
`
`2s
`
`T - Ti
`T2 - Ti
`T3 - T4
`T3 - T4S
`
`It may appear unrealistic to treat an uninsulated turbine that is incan(cid:173)
`descent as being adiabatic. However, the heat transferred will still be small
`compared to the energy flow through the turbine. Strictly speaking, the
`kinetic energy terms should be included in the steady-flow energy equa(cid:173)
`tion. Since the kinetic energy can be incorporated into the enthalpy term,
`the preceding arguments still apply by using stagnation or total enthalpy
`with the corresponding stagnation or total temperature.
`The shape of the isobars (lines of constant pressure) can be found quite
`readily. From the 2nd Law of Thermodynamics
`
`( :: ).
`( :~).
`( :; ),
`
`Thus
`
`or
`
`and
`
`Tels= dh - vdp
`
`=T
`
`cc T
`
`that is, on the T-s plot isobars have a positive
`slope proportional to the absolute temperature
`
`= v
`
`that is, the vertical separation between isobars
`is proportional to the specific volume, and
`specific volume increases with temperature
`
`Consequently the isobars diverge in the manner shown in figure 9.6.
`In a turbocharger the compressor is driven solely by the turbine, and a
`mechanical efficiency can be defined as
`
`(9.4)
`
`As in gas turbines, the pressure ratios across the compressor and turbine
`are very important. From the pressure ratio the isentropic temperature
`ratio can be found:
`
`(9.5)
`
`BMW1045
`Page 11 of 32
`
`

`

`TURBOCHARGING
`
`333
`
`Engine exhaust
`temperature ( T 31 ° C
`
`06
`
`0·7
`
`0·6
`
`1.4
`
`1.3
`
`1.2
`
`1.1
`
`1.0
`
`0.9
`
`0.8
`
`0.11
`
`M
`~
`('I
`Q.
`Q) ... :,
`"' "' E
`... V, :,
`a.
`'° .c
`X
`Q)
`0 ..,
`]i
`C
`
`Q)
`
`s::
`·5,
`C
`
`Q) -0
`
`0 ·.:;
`"' a:
`
`Figure 9. 7 Effect of overall turbocharger on the pressure ratio between engine
`inlet and exhaust manifold pressures, for a 2:1 compressor pressure
`ratio (pifp1 = 2) with different engine exhaust temperatures (with
`acknowledgement to Watson and Janota (1982))
`
`The actual temperatures, T2 and T4 , can then be found from the respective
`isentropic efficiencies.
`In constant-pressure turbocharging it is desirable for the inlet pressure to
`be greater than the exhaust pressure (pifp 3 > 1), in order to produce good
`scavenging. This imposes limitations on the overall turbocharger efficiency
`( 1Tm·rJr·r1c) for different engine exhaust temperatures (T3). This is shown in
`figure 9. 7. The analysis for these results originates from the above ex(cid:173)
`pressions, and is given by Watson and Janota (1982). Example 9.1 also
`illustrates the work balance in a turbocharger.
`The flow characteristics of an axial and radial compressor are compared
`in figure 9.8. The isentropic efficiencies would be typical of optimum-sized
`machines, with the axial compressor being much larger than the radial
`compressor. Since the turbocharger compressor is very small the actual
`efficiencies will be lower, especially in the case of an axial machine. The
`surge line marks the region of unstable operation, with flow reversal etc.
`The position of the surge line will also be influenced by the installation on
`
`BMW1045
`Page 12 of 32
`
`

`

`5 . - ---,-.- - -- - -~- ---r-- - -.------,
`
`5~- -~-----r----r-- - - .----.- --
`
`4
`
`3
`
`....
`~
`N
`Cl.
`0
`·;:;
`co
`....
`....
`Q)
`::,
`V,
`V,
`....
`Q)
`Cl.
`
`1.0
`
`Locus of points of
`maximum efficiency
`
`2t
`
`I
`
`I
`
`J
`
`.'I
`
`7!. I
`
`I
`
`,,.,
`
`0.2
`
`0.4
`Fraction of design mass flow rate
`
`0.6
`
`0.8
`
`Fraction of
`design speed
`
`1.0
`
`1.2
`
`I 1/
`
`Locus of points of
`maximum efficiency '
`
`0.9
`
`I
`
`....
`~
`N
`Cl.
`.2 ..,
`ro ...
`....
`Q)
`::,
`V,
`
`V, a,
`....
`Cl.
`
`4
`
`3
`
`2
`
`0.6
`
`0.7
`\
`Fraction of design
`speed
`1 ,__ __ _._ ___ ..__ __ __._ ___ ....... _____ __
`1.2
`
`0
`
`0.2
`
`0.8
`0.4
`0.6
`Fraction of design mass flow rate
`
`1.0
`
`w w
`
`~
`
`z
`:d
`0
`Cl
`c:::
`B
`0 z
`>-l
`0
`
`~
`~
`~
`8 s: tp
`c:::
`~
`0 z
`~
`~ C/l
`
`C)
`
`BMW1045
`Page 13 of 32
`
`

`

`100
`
`100
`
`\
`
`"
`0~9 ,\
`1.0
`
`\
`
`I
`\
`\
`I
`I
`0 .7 0.8
`
`-
`
`Fraction of
`design speed
`
`c:,
`.:-
`>
`(,)
`C: 60
`a.,
`
`* 80
`.S:! --a.,
`
`.S:! 40
`a.
`0 ....
`....
`C:
`a.,
`~ 20
`
`-
`
`0.6 0. 7 0.8 0.9 1 .0
`1 - - -
`
`Fraction of design speed
`
`'*- 80
`> (,)
`C:
`!!? 60
`
`(,)
`
`"
`\
`0.5
`
`I \
`
`I \
`0.6
`
`;;: -a.,
`
`(,)
`
`·a. 40
`0 ...
`....
`"' - 20
`
`C:
`a.,
`
`0
`
`0.2
`
`0.8
`0.6
`0.4
`Fraction of design mass flow rate
`
`1 .0
`
`1.2
`
`0
`0
`
`0.2
`
`0.4
`0.6
`0.8
`Fraction of design mass flow rate
`
`1.0
`
`1.2
`
`Multi-stage axial compressor
`
`Single-stage radial compressor
`
`Figure 9.8 Flow characteristics of axial and radial compressors (reproduced
`with permission from Cohen et al. (1972))
`
`~
`::i:::i
`to
`0
`
`~ C}
`..... a
`
`w w
`
`VI
`
`BMW1045
`Page 14 of 32
`
`

`

`336
`
`INTRODUCTION TO INTERNAL COMBUSTION ENGINES
`
`R eci rculat inn pipe
`
`Turbine
`
`Nozzle
`
`Figure 9.9 Early pulse converter system (with acknowledgement to Watson and
`Janota (1982))
`
`the engine. Figure 9.9 shows the wider operating regime of a radial flow
`compressor. The isentropic efficiency of a turbocharger radial compressor
`is typically in the range 65- 75 per cent.
`The design of turbines is much less sensitive and the isentropic efficiency
`varies less in the operating range, and rises to over 90 per cent for aircraft
`gas turbines. The isentropic efficiency of turbocharger turbines is typically
`70-85 per cent for radial flow and 80-90 per cent for axial flow machines.
`These are optimistic 'total to total' efficiencies that assume recovery of the
`kinetic energy in the turbine exhaust.
`A detailed discussion of the internal flow, design, and performance of
`turbochargers can be found in Watson and Janota (1982).
`The flow from an engine is unsteady, owing to the pulses associated with
`the exhaust from each cylinder, yet turbines are most efficient with a steady
`flow. If the exhaust flow is smoothed by using a plenum chamber, then
`some of the energy associated with the pulses is lost. The usual practice is
`to design a turbine for pulsed flow and to accept the lower turbine
`efficiency. However, if the compressor pressure ratio is above 3:1 the
`pressure drop across the turbine becomes excessive for a single stage. Since
`a multi-stage turbine for pulsed flow is difficult to design at high pressure
`ratios, a steady constant-pressure turbocharging system should be adopted.
`The effect of flow pulsations on turbine performance is discussed in chapter
`10, section 10.3.2.
`In pulse turbocharging systems the area of the exhaust pipes should be
`close to the curtain area of the valves at full valve lift. Some of the gain in
`using small exhaust pipes comes from avoiding the expansion loss at the
`beginning of blowdown. In addition, the kinetic energy of the gas is
`preserved until the turbine entry. To reduce frictional losses the pipes
`should be as short as possible.
`For four-stroke engines no more than three cylinders should feed the
`same turbine inlet. Otherwise there will be interactions between cylinders
`exhausting at the same time. For a four-cylinder or six-cylinder engine a
`turbine with two inlets should be used. The exhaust connections should be
`such as to evenly space the exhaust pulses, and the exhaust pipes should be
`free of restrictions or sharp corners. Turbines with four separate entries are
`
`BMW1045
`Page 15 of 32
`
`

`

`TURBOCHARGING
`
`337
`
`available, but for large engines it can be more convenient to use two
`separate turbochargers. For a 12-cylinder engine two turbochargers, each
`with a twin entry turbine, could each be connected to a group of six
`cylinders. This would make installation easier, and the frictional losses
`would be reduced by the shorter pipe lengths. For large marim: diesel
`engines, there can be one turbocharger per pair of cylinders. While there
`are thermodynamic advantages in lagging the turbines and pipework, the
`ensuing reduction in engine room temperature may be a more important
`consideration.
`The pressure pulses will be reflected back as compression waves and
`expansion waves. The exact combination of reflected waves will depend on
`the pipe junctions and turbine entry. The pipe lengths should be such that
`there are no undesirable interactions in the chosen speed range. For
`example, the pressure wave from an opening exhaust valve will be partially
`reflected as a compression wave by the small turbine entry. If the pipe
`length is very short the reflected wave will increase the pressure advan(cid:173)
`tageously during the initial blow-down period. A slightly longer pipe, and
`the delayed reflected wave, will increase the pumping during the exhaust
`this increases the turbine output at the expense of increased
`stroke -
`piston work in the engine. An even longer pipe would cause the reflected
`wave to return to the exhaust valve during the period of valve overlap -
`this would impair the performance of a four-stroke engine and could ruin
`the performance of a two-stroke engine. If the pressure wave returns after
`the exhaust valve has closed, then it has no effect. Evidently great care is
`needed on engines with long exhaust pipes and large valve overlaps.
`An alternative to multi-entry turbines is the use of pulse converters. An
`early pulse converter system is shown in figure 9 .9; the idea was to use the
`jet from the nozzle to produce a low-pressure area around each exhaust
`port. The principal disadvantages are:
`
`(1) insufficient length between the ports for efficient diffusion
`(2) high frictional losses
`(3) each nozzle has to be larger than the last, resulting in high manufactur(cid:173)
`ing cost.
`
`A more realistic approach is to use pulse converters to connect groups of
`cylinders that would otherwise be separate. For example, four cylinders
`could be connected to a single turbocharger entry, figure 9.10. The steadier
`flow can also lead to an improvement in turbine performance. The design
`of the pulse converter is a compromise between pressure loss and un(cid:173)
`wanted pulse propagation. Reducing the throat area increases the pressure
`loss, but reduces the pulse propagation from one group of cylinders to
`another. The optimum design will depend on the turbine, the exhaust pipe
`length, the valve timing, the number of cylinders, the engine speed etc.
`
`BMW1045
`Page 16 of 32
`
`

`

`338
`
`INTRODUCTION TO INTERNAL COMBUSTION ENGINES
`
`Single entry
`turbocharger
`
`Ap
`Area ratio (nozzles) = An/Ap (0.65-0.85)
`Area ratio (throat) = Ath/2Ap (0.5-1.0)
`
`Figure 9.10 Exhaust manifold arrangement (four-cylinder engine) and pulse
`converter details (with acknowledgement to Watson and Janota
`(1982))
`
`Constant-pressure turbocharging (that is, when all exhaust ports enter a
`chamber at approximately constant pressure) is best for systems with a high
`pressure ratio. The dissipation of the pulse energy is offset by the improved
`turbine efficiency. Furthermore, during blow-down the throttling loss at
`the exhaust valve will be reduced. However, the part load performance of
`a constant-pressure system is poor because of the increased piston pumping
`work, and the positive pressure in the exhaust system can interfere with
`scavengmg.
`
`9.3 Turbocharging the compression ignition engine
`
`The purpose of turbocharging is to increase the engine output by increasing
`the density of the air drawn into the engine. The pressure rise across the
`compressor increases the density, but the temperature rise reduces the
`density. The lower the isentropic efficiency of the compressor, the greater
`the temperature rise for a given pressure ratio.
`Substituting for T2s from equation (9 .5) into equation (9 .3) and rearrang(cid:173)
`ing gives
`
`T, = T, [ 1 + (p/p,)':~•JI, - l l
`
`(9.6)
`
`This result is for an ideal gas, and the density ratio can be found by
`applying the Gas Law, p = p!RT. Thus
`
`BMW1045
`Page 17 of 32
`
`

`

`TURBOCHARGING
`
`339
`
`Full cooling
`(that is, T2 = T, l
`
`lsentropic compression
`Tlc = 0.8
`Tlc = 0.7
`Tlc = 0.6
`Tlc = 0.5
`
`3.0
`
`2.5
`
`~
`"'
`-; 2.0
`·;;
`:'!
`~
`·;;;
`C .,
`o
`
`1.5
`
`1.5
`
`2.0
`2.5
`Pressure ratio (.o2 /p, l
`
`Figure 9 .11 Effect of compressor efficiency on air density in the inlet manifold
`(with acknowledgement to Watson and Janota (1982))
`
`A = P2 [ 1 + (Pzf Pi)(y- l)/y - 1 1-1
`P1
`'Y/c
`
`Pi
`
`_
`
`(9.7)
`
`The effect of compressor efficiency on charge density is shown in figure
`9.11; the effect of full cooling (equivalent to isothermal compression) has
`also been shown. It can be seen that the temperature rise in the compressor
`substantially decreases the density ratio, especially at high pressure ratios.
`Secondly, the gains in the density ratio on cooling the compressor delivery
`can be substantial. Finally, by ensuring that the compressor operates in an
`efficient part of the regime, not only is the work input minimised but the
`temperature rise is also minimised. Higher engine inlet temperatures raise
`the temperature throughout the cycle, and while this reduces ignition delay
`it increases the thermal loading on the engine.
`The advantages of charge cooling lead to the use of inter-coolers. The
`effectiveness of the inter-cooler can be defined as
`
`actual heat transfer
`£= - - - - - - - - - - - - -
`maximum possible heat transfer
`
`For the cooling medium it is obviously advantageous to use a medium
`(typically air or water) at ambient temperature (T1), as opposed to the
`engine cooling water.
`If T3 is the temperature at exit from the inter-cooler, and the gases are
`perfect, then
`
`BMW1045
`Page 18 of 32
`
`

`

`340
`
`INTRODUCTION TO INfERNAL COMBUSTION ENGINES
`
`s =
`
`(9.8)
`
`or
`
`In practice it is never possible to obtain heat transfer in a heat exchanger
`without some pressure drop. For many cases the two are linked linearly by
`Reynolds' analogy -
`that is, the heat transfer will be proportional to the
`pressure drop. In the following simple analysis the pressure drop will be
`ignored.
`Substituting for T2 from equation (9.6), equation (9.8) becomes
`
`l
`
`(9.9)
`
`1 I
`f I ( / )(y- 1)/y
`(Pif Pi)Cy- l)/y - 1 I
`I
`
`T3 = T1 l 1 + P2 P 1 T/c
`
`-
`
`(1 - s) + s
`
`= T1 1 + (1 - s)
`
`T/c
`
`Neglecting the pressure drop in the inter-cooler, equation (9.7) becomes
`(p Ip )CY- 1)/y - 1 1-1
`1 + (l _ s) __ 2_1 _ __ _
`1"/c
`
`_ 3_ = _ 2
`P
`P1
`Pi
`
`P I
`
`(9.10)
`
`The effect of charge cooling on the density ratio is shown in figure 9 .12 for
`a typical isentropic compressor efficiei:icy of 70 per cent, and an ambient
`temperature of 20°C.
`Despite the advantages of inter-cooling it is not universally used. The
`added cost and complexity are not justified for medium output engines,
`and the provision of a cooling source is troubles9me. Gas to gas heat
`exchangers are bulky and in automotive applications would have to be
`placed upstream of the radiator. An additional heat exchanger could be
`used with an intermediate circulating liquid, but with yet more cost and
`complexity. In both cases energy would be needed to pump the flows.
`Finally, the added volume of the inter-cooler will influence the transient
`performance of the engine.
`The effect of inter-cooling on engine performance is complex, but two
`cases will be considered: the same fuelling rate and the same thermal
`loading. Inter-cooling increases the air flow rate and weakens the air/fuel
`ratio for a fixed fuelling rate. The temperatures will be reduced throughout
`the cycle, including the exhaust stage. The turbine output will then be
`
`BMW1045
`Page 19 of 32
`
`

`

`TURBOCHARGING
`
`341
`
`fie= 0.7
`Ambient & coolant
`temperature= 20°C
`
`f = 0.8
`e=0.7
`E = 0.6
`
`No cooling
`
`Q
`
`3.0
`
`2.5
`
`~
`"'
`~ 2.0
`-~
`~ ·.;;
`C .,
`o
`
`1.5
`
`1.5
`
`2.5
`2.0
`Compressor pressure ratio
`
`3.0
`
`Figure 9.12 Effect of charge cooling on inlet air density (with
`acknowledgement to Watson and Janota (1982))
`
`reduced, unless it is rematched, but the compressor pressure ratio will not
`be significantly reduced. The reduced heat transfer and changes in combus(cid:173)
`tion cause an increase in bmep and a reduction in specific fuel consump(cid:173)
`tion. Watson and Janota (1982) estimate both changes as 6 per cent for a
`pressure ratio of 2.25 and inter-cooler effectiveness of 0.7. The gains are
`greatest at low flow rates where the inter-cooler is most effective.
`If the fuelling rate is increased to give the same thermal loading Watson
`and Janota (1982) estimate a gain in output of 22 per cent. The specific fuel
`consumption will also be improved since the mechanical losses will not
`have increased so rapidly as the output.
`Low heat loss engines also offer scope for improving the performance of
`Diesel engines. Firstly, and most widely quoted, is the improvement in
`expansion work and the higher exhaust temperature. This leads to further
`gains when an engine is turbocharged. Secondly, the reduced cooling
`requirements allow a smaller capacity cooling system. The associated
`reduction in power consumed by the cooling system is, of course, most
`significant at part load.
`Reducing the heat transfer from the combustion chamber also leads to a
`reduced ignition delay and hence reduced Diesel knock. However, the
`high combustion temperatures will lead to an increase in NOx emissions.
`Reductions in heat transfer from the combustion chamber can be ob(cid:173)
`tained by redesign with existing materials, but the greatest potential here is
`offered by ceramics. Ceramics can be used as an insulating layer on
`metallic components, or more radically as a material for the complete
`component. These heat transfer aspects are discussed further in chapter 12,
`section 12.2.3.
`
`BMW1045
`Page 20 of 32
`
`

`

`342
`
`INTRODUCTION TO INTERNAL COMBUSTION ENGINES
`
`1 Head liner: silicon nitride
`2 Pislo11: silicon nitride/aluminium alloy
`3 Piston ring: silicon nitride
`, Upper seat: silicon nitride
`5 Cam: silicon nitrideJziroonia
`6 Tappet silicon nrtride/aluminium alloy
`7 Exhaust port: aluminium titanate
`
`8 Exhaust valve: silicon nrtnde (60% lighter than steel)
`9 Cylirider liner: silicon nitride
`10 High spee<I generator
`11 Turbine
`12 Variable geometry nozzle
`13 Electronic controller
`1' Mntn,
`
`Figure 9.13
`
`Isuzu 2.9 litre V6 Diesel engine with many ceramic components
`(Anon (1990))
`
`Differences between the thermal coefficients of expansion of metals and
`ceramics mean that great care is needed in the choice of the ceramic and
`the metal substrate, if the insulating layer is not to separate from its
`substrate. None the less, ceramic insulation has been used successfully in
`Diesel engines -
`for example, the work reported by Walzer et al. (1985).
`In this turbocharged engine, 80 per cent of the combustion chamber
`surface was covered to an average depth of 3 mm by aluminium titanate or
`zirconium dioxide insulation. These measures led to a 13 per cent reduc(cid:173)
`tion in heat flow to the coolant, and a 5 per cent improvement in the urban
`cycle fuel economy (this was without re-optimisation of the cooling system).
`A significant example of using ceramics in a Diesel engine has been
`provided by Isuzu, with a 2.9 litre V6, 24 valve engine (Anon (1990)). This
`engine is shown in figure 9 .13, and it can be seen that there is no cooling
`system.
`
`BMW1045
`Page 21 of 32
`
`

`

`TURBOCHARGING
`
`343
`
`A study of a low heat loss engine by Hay et al. (1986) suggested that a 30
`per cent reduction in heat transfer would lead to a 3.6 per cent reduction in
`the specific fuel consumption at the rated speed and load. Furthermore,
`the 2 kW reduction in the cooling fan power would lead to an additional 2.7
`per cent reduction in the fuel consumption, at the rated speed and load,
`and correspondingly greater percentage improvements at part load. Work
`by Wade et al. (1984) in a low heat loss engine suggests that the greatest
`gains in efficiency are at light loads and high speeds (figure 9.14).
`Alternatively, the higher exhaust temperature will allow a smaller tur(cid:173)
`bine back-pressure for the same work output. Work by Hoag et al. (1985)
`indicates that a 30 per cent reduction in heat transfer would lead to a 3.4
`per cent fall in volumetric efficiency, a 70 K rise in the exhaust gas
`temperature, and reductions in fuel consumption of 0.8 per cent for a
`turbocharged engine, and 2 per cent for a turbocompound engine.
`So far no mention has been made of matching the turbocharger to the
`engine. Reciprocating engines operate over a wide speed range, and the
`flow range is further extended in engines with throttle control. In contrast,
`turbomachinery