Xin Q. Diesel Engine System Design

Подождите немного. Документ загружается.

872 Diesel engine system design

of the compressor wheel or diffuser. The ow rate at which choke happens

varies with the tip speed. The precise position of the choke curve on the

compressor map is more difcult to determine than the surge line because it

depends on the ow structure of the internal boundary layers. Sometimes for

the purpose of approximation it is assumed that compressor choke happens

when the compressor efciency contours reduce to around 55%.

Pulse and constant-pressure turbocharging

There are two types of turbocharging: pulse and constant pressure. In pulse

turbocharging, in order to minimize the pressure wave interference or gap

windage in exhaust scavenging between the cylinders and to raise the average

turbine efciency, the cylinders with a certain ring interval (ideally about

240° crank angle) are grouped together by using a small-volume exhaust

manifold to preserve strong exhaust pressure pulses. A pulse converter

may be used to eliminate pulse interference. A divided turbine entry may

be used to totally separate the pulses of different groups, but may cause a

larger pumping loss at high engine speeds than an undivided turbine entry.

In VGT, sometimes it is difcult to use a divided turbine entry because of

the extra controlling device. When the unsteady energy ow is fed into the

turbine, turbine efciency may decrease.

In constant-pressure turbocharging, the exhaust ows of all the cylinders are

collected in one large-volume exhaust manifold to damp out all the pressure

uctuations. A steady energy ow is fed into a single-entry turbine, and the

turbine can be matched to the peak efciency on the steady ow test map.

The averaged turbine efciency over an engine cycle with the unsteady ow

of pulse turbocharging is usually lower than the peak efciency matched in

constant-pressure turbocharging. But the disadvantage of efciency loss is

normally more than offset by the advantage of the pulse energy available

at the turbine. The overall exhaust energy utilization efciency is actually

characterized by the combination of the pulse energy and the turbine

efciency.

Previous experiments indicated that at a low compressor pressure ratio

(e.g., at part load), pulse turbocharging has a higher overall efciency than

constant-pressure turbocharging. But the situation is reversed at very high

compressor pressure ratios because the pulsed turbine efciency is low at

very high instantaneous turbine pressure ratios (e.g., above 3:1). During

the transient acceleration, pulse turbocharging is superior due to its smaller

exhaust manifold volume and faster response of the turbine inlet energy pulse

to speed up the turbocharger. In EGR engines, pulse turbocharging used with

a reed valve (check valve) can drive more EGR ow than constant-pressure

turbocharging with less penalty in engine pumping loss.

Diesel-Xin-13.indd 872 5/5/11 12:04:44 PM



873Diesel engine air system design

Single-stage, two-stage, and twin-parallel turbocharging

Usually single-stage turbocharging can deliver a compressor pressure ratio

around 3.5–4.5, depending on its wheel material and design. Compressor speed

is limited by the allowable centrifugal force exerted upon the compressor wheel.

The maximum compressor speed and the aerodynamic blade shape limit the

pressure ratio capability (Arnold, 2004). Moreover, the compressor wheel and

casing made of aluminum alloy are limited in material strength by a maximum

allowable compressor discharge temperature of 204–210°C (400–410°F).

Titanium impeller wheels can resist much higher temperatures and deliver

a very high pressure ratio with the simple single-stage turbocharging, but

titanium and its manufacturing process are expensive. Another disadvantage

of the titanium wheels is their heavier weight and larger inertia.

Despite the disadvantages in cost, weight, cumbersome piping, and

packaging space, two-stage turbocharging is superior to single-stage in the

following aspects: (1) higher compressor pressure ratio; (2) higher turbo

efciency due to the relatively lower pressure ratio in each stage; (3) feasibility

of using inter-stage cooling to reduce the compressor discharge temperature

and the compressor power to improve the efciency; (4) easier matching

for the turbocharger for a wide range of engine speed combined with high

rated power and high peak torque; (5) improved altitude capability due to a

larger margin of choke; (6) faster transient response due to the small high-

pressure-stage turbocharger and its low rotor inertia; and (7) improved low-

cycle-fatigue life of the compressor wheels and fewer durability problems

associated with excessively high turbocharger speed.

In the twin-parallel turbocharger conguration, both turbines can be

located very close to the cylinder heads of each bank of the engine in order

to minimize the exhaust energy losses caused by manifold volume, pressure

drop, and heat transfer.

Wastegate vs. VGT turbocharging

The advantages of VGT over wastegated turbines are summarized as

follows: (1) no throttling loss of the wastegate valve; (2) a better ability to

achieve a high air–fuel ratio and high peak torque at low engine speeds; (3)

a better ability to achieve fast vehicle accelerations without encountering a

high pumping loss at high engine speeds; (4) overall lower engine delta P;

and (5) a better ability to cover a wider region of low BSFC in the engine

speed–load domain. The VGT area change can be in either continuous or

discrete settings, depending on the requirements of engine performance and

reliability. The maximum efciency of most variable-geometry turbines occurs

at a medium vane opening of around 60–70% open. At the fully open and

fully closed vane openings, the efciency drops rapidly. The challenges of

Diesel-Xin-13.indd 873 5/5/11 12:04:44 PM



874 Diesel engine system design

VGT design include minimizing gas leakage, increasing turbine efciency,

and enhancing reliability (Furukawa et al., 1993).

Previous research on turbocharging

Turbomachinery fundamentals are introduced in Japikse and Baines (1997),

Wilson and Korakianitis (1998), and Japikse (2009). The detailed theory

of turbocharging the non-EGR engines is provided by Watson and Janota

(1982) and Watson (1999). A more recent summary on turbocharging the

internal combustion engines is provided by Baines (2005b). For the mapping

procedures of turbocharger gas stand test data, the reader is referred to SAE

J1826 (1995).

Early investigations on turbocharging the non-EGR diesel engines are

represented by Freeman and Walsham (1978), Watson et al. (1978), Flynn

(1979), and Watson (1979). Turbocharging the EGR diesel engine is presented

by Bozza et al. (1997) and Galindo et al. (2007). Advanced turbocharging

technologies for EGR diesel engines are discussed by Ludu et al. (2001),

Arnold et al. (2001, 2005), Arnold (2004), Amos (2002), Carter et al. (2010),

and Tufail (2010).

A turbocharger matching procedure for the wastegated turbochargers is

given by Ubanwa and Kowalczyk (1993). VGT performance is investigated

by Flaxington and Szczupak (1982), Watson (1982, 1986), Watson and

Banisoleiman (1986), Yokota et al. (1986), Wallace et al. (1986), Hishikawa

et al. (1988), Kawamoto et al. (2001), Arnold et al. (2002), Tange et al. (2003)

and Uchida et al. (2006). Two-stage turbocharging on diesel engines has been

researched by Ghadiri-Zareh and Wallace (1978), Watson et al. (1978), Saulnier

and Guilain (2004), Millo et al. (2005), Choi et al. (2006), and Serrano et al.

(2008). Twin-parallel turbocharging is presented by Cantemir (2001).

It should be noted that the design and development of non-VNT type of

VGT (also called VAT or variable area turbine) has been active since 1970s.

The VAT has the advantages of simpler design, better reliability and lower

cost, compared with VNT. However, the VAT usually suffers on turbine

efciency. Major VAT designs include Pampreen (1976), Chapple et al.

(1980), Bhinder (1984), Hirabayashi et al. (1986), Okazaki et al. (1986),

Franklin and Walsham (1986), Franklin (1989), Inoue et al. (1989), Ogura et

al. (1989), Umezaki et al. (1989), Ogura and Shao (1995), Shao et al. (1996),

and Wang (1996). The recent most noticeable and promising VAT design

was the variable ow turbine (VFT) developed by Aisin Sekei and used by

other companies (Kawaguchi et al., 1999; Ishihara et al., 2002; Inter-Tech

Energy Progress, 2003; Andersen et al., 2006; and Ito et al., 2007).

Another noticeable and promising turbocharger technology is the swirl

jet turbine developed by Anada et al. (1997). The swirl jet turbine generates

a swirling exhaust gas ow at the turbine outlet by utilizing the wastegated

Diesel-Xin-13.indd 874 5/5/11 12:04:44 PM



875Diesel engine air system design

gas ow so that the turbine outlet pressure is effectively reduced locally to

enable an increased turbine power and higher turbine speed. At mean time,

the turbine inlet pressure and the engine delta P may decrease. By properly

controlling the turbine wastegate opening or the swirling jet, a higher air–fuel

ratio or lower engine delta P (and lower pumping loss) can be achieved.

Similar designs to the swirl jet turbine were reported by Baker et al. (2001)

and Schmid and Sumser (2007).

Moreover, turbocharging simulation models have been developed by

Nasser and Playfoot (1999), Gurney (2001), Galindo et al. (2006), Zhuge

et al. (2009), and Park et al. (2010). Supercharging the diesel engine is

investigated in Cantore et al. (2001). Supercharger testing was introduced

by SAE J1723 (1995).

13.4.2 Compressor matching, aerodynamic design, and

durability

The turbocharger is one of the most expensive components in the engine

and therefore deserves close attention from a commercial point of view. The

selection of size and efciency of a turbocharger is based on the requirements

of delivering the target air–fuel ratio, EGR rate, and transient acceleration

performance. The size of a turbocharger determines the ow range and the

pressure ratio. The objective of turbocharger matching is to achieve the best

compromise in the engine speed–load range.

The turbocharger test procedure was outlined in SAE J1826 (1995),

Turbocharger Gas Stand Test Code. In principle, the compressor map

can be generated from a ow bench test by regulating two valves. One is

installed at the inlet of a driving turbine, and the other is installed at the

compressor outlet. The compressor outlet valve can be gradually closed at

each turbine valve opening in order to obtain a series of compressor ow

points (from high to low until surge), while the compressor speed remains

almost constant. The process can be repeated with different turbine valve

openings, and nally a range of compressor ow points at many shaft speeds

can be obtained.

The focus of engine–compressor matching is to analyze the engine operating

characteristic points or lines at various speed, load, and ambient conditions on

the compressor map. Diesel engine characteristic lines include the full-load

lug curve, constant-speed curve, and constant-load curve. The slope of the

constant-speed curve on the compressor map, which is essentially the ratio

between the pressure and the mass ow rate, represents the reciprocal of

the engine volumetric efciency. Therefore, any factors affecting volumetric

efciency may have an impact on the shape of engine speed characteristic

lines. These factors include valvetrain design (valve size, timing, and lift

prole), valve/port ow discharge coefcient, intake port heating, engine

Diesel-Xin-13.indd 875 5/5/11 12:04:44 PM



876 Diesel engine system design

delta P, and intake manifold gas temperature. The characteristics of engine

volumetric efciency are further discussed in Fig. 13.10.

In order to achieve cost-effective designs, a ‘family’ of different blades

with various tip widths and eye diameters are produced for one compressor

impeller wheel ‘frame size’ (or casing). The compressor frame size is

determined by the requirement of engine maximum air ow rate. The design

parameters such as wheel inducer diameter, wheel and shroud contours, and

diffuser width determine the compressor ‘trim’. These trims offer different

ow characteristics of surge and choke as well as variations in efciency.

Compressor efciency and shaft speed characteristic curves on the map

can be further ne-tuned by using different trims with changes in impeller

splitter blade outlet angle, diffuser entry angle, and diffuser geometry. For

example, changing the diffuser entry angle or blade eye diameter can ‘rotate’

the map curves around the origin of the coordinates so that the compressor

surge line moves. It should be noted that changing the aerodynamic design

also affects the compressor tip speed and the stress. Turbocharger design

details are provided by Rodgers and Rochford (2002), Watson and Janota

(1982), and Watson (1999).

A typical compressor map is shown in Fig. 13.7. The engine operating

points shown as the smaller white dots at different engine speeds and ambient

conditions reect the engine operation with a single-stage xed-geometry

turbocharger. The compressor pressure ratio of the engine operating points

increases as the altitude increases.

13.7 Illustration of compressor map and engine operation.

Engine speed

increases

Rated power

(sea level 77°F)

Sea level

77° ambient

Sea level

hot ambient

Sea level

Engine load

decreases

Peak torque

(sea level 77°F)

0.0001 0.0170 0.0339 0.0508 0.0677 0.0846

Corrected compressor ﬂow rate

2.980

2.584

2.188

1.792

1.396

1.000

Pressure ratio

(%)

20.0

30.0

32.5

35.0

37.5

40.0

42.5

45.0

47.5

50.0

52.5

55.0

57.5

60.0

62.5

65.0

67.5

70.0

72.5

75.0

77.5

80.0

82.5

100.0

79.4

20.0

MAX:

MIN:

Compressor

efﬁciency

15000 ft

12500 ft

10000 ft

5000 ft

15000 ft

1500 rpm

1000 rpm

Engine

speed

2300 rpm

Sea level

Diesel-Xin-13.indd 876 5/5/11 12:04:45 PM



877Diesel engine air system design

Not all the locations from A to H shown on the compressor map in Fig.

13.7 can be operated in reality. The engine can run at locations E and D

stably. Although operating at locations B and C is possible, they are in choke

and the compressor efciency is very low. Choke is a relative term and is

usually dened at a given compressor efciency (normally around 55%). In

the choke region, engine performance becomes very poor. Note that the three

main factors affecting the compressor power are air ow rate, compressor

efciency, and pressure ratio. The compressor power becomes very high when

the efciency is low for delivering a required boost pressure. This would

require the turbine to deliver a high power with a large expansion ratio,

which would increase the exhaust manifold pressure and the pumping loss.

The location B has a very poor efciency (30–35%), but the pressure ratio

is low. The location C has a much higher efciency, but the pressure ratio

is higher. Their compressor power may be different. Another consideration

when comparing locations B and C is that the turbocharger thrust bearing

(axial) loading will be high at location B since there will likely be a high

pressure behind the turbine wheel and a much lower pressure behind the

compressor wheel. This could potentially cause a thrust bearing failure.

The operating locations F, G, H, and A in Fig. 13.7 cannot be run stably

in reality. Without redrawing another compressor map for simplicity, one

example is to imagine these four locations are intended for the operation of

a small high-pressure-stage (HP) compressor in a two-stage series sequential

turbocharger system. The HP compressor can be sized very small to enable

good transient performance at part load. At rated power, the HP turbine

wastegate needs to be opened in order to avoid over-boosting the engine.

In reality, the HP compressor will act like a large ow restriction at higher

engine speeds, and all the required engine air ow will not be able to ow

through the HP compressor at locations F, G, H, or A. In this case, not only

is a HP turbine wastegate required, a HP compressor bypass valve is also

needed to bypass the engine air from the HP compressor. If the HP turbine

wastegate is open but the HP compressor bypass is closed, the location

A cannot be run stably without choke although the HP turbocharger still

spins and some small percentage of the ow can still go through the HP

compressor. The turbine speed would drop dramatically since there is little

or no HP turbine power. The compressor needs to run at the same shaft speed

as the turbine (e.g., very low at 10,000 rpm). Such a low speed line on the

compressor map would give a very small value of choked ow. The point

of compressor pressure ratio equal to one on the 10,000 rpm speed line is

the maximum compressor ow that can travel through the HP compressor

at that shaft speed. And this value is much less than the ow rates of the

higher-shaft-speed points at locations F, G, H, and A. In other words, it is

less than the ow rates required to support the engine operation at high

speeds/loads. Also note that when the HP compressor bypass is open, the

Diesel-Xin-13.indd 877 5/5/11 12:04:45 PM



878 Diesel engine system design

compressor pressure ratio will actually be below 1 since there will be a slight

pressure drop across the bypass valve. In this case, the compressor slightly

expands the air rather than compressing it.

Centrifugal compressor matching guidelines are summarized as

follows:

∑ The matching should suit the normal operating environment of the

engine, and should also be conducted at extreme ambient conditions

from sea-level altitude and cold winter to high altitude and hot summer

with minimum fueling derating to ensure the matching is within all the

design constraints.

∑ Always use a single-stage turbocharger whenever possible in order to

reduce cost, weight, and complexity.

∑ Provide proper pressure ratio and air ow rate at all engine speed–load

modes, including zero-EGR conditions (e.g., fast transient, aftertreatment

regeneration, cold climate).

∑ There should be no compressor surge along the full-load lug curve,

especially below the peak torque speed. A proper surge margin (10–15%

of the surge ow rate on the compressor map) should be reserved for

engine transients, intake ow variations at various ambient temperatures

(e.g., cold climate) and high altitude conditions, as well as for the extreme

case of high intake restriction caused by air lter blockage.

∑ The compressor ow range should be wide enough so that there is no

compressor choke and over-speeding at all altitude levels and ambient

temperatures (i.e., no sonic ow inside the compressor and no operation in

the very-low-efciency region such as less than approximately 55%).

∑ Provide an acceptable turbine inlet temperature with a sufciently high

air–fuel ratio and EGR rate. High compressor efciency is desired to

reduce the compressor discharge temperature.

∑ Full-load and typical vehicle driving modes should be matched in the

high-efciency region on the compressor map in order to reduce the

pumping loss. Minimum fuel consumption in a given driving cycle can

be used as the optimization target for the compressor matching in this

case.

∑ The compressor’s moment of inertia should be as low as possible in

order to reduce the turbocharger lag to achieve good engine transient

acceleration performance.

∑ Compressor durability life needs to be acceptable. The turbocharger

life is determined mainly by the low-cycle-fatigue (LCF) life of the

compressor wheel under cyclic loading in relatively slow driving or

usage cycles (see Chapter 2 for details). The compressor impeller life

is dictated by the following: its maximum tip speed; the difference

between the maximum and minimum shaft speed in typical driving cycles

or aftertreatment regeneration cycles; and the cycle frequency. Those

Diesel-Xin-13.indd 878 5/5/11 12:04:45 PM



879Diesel engine air system design

factors can be shown in the load cycle diagram of ‘turbocharger speed

vs. driving time’. In turbocharger matching, the maximum turbocharger

speed limit is determined by calculating the compressor fatigue life based

on transient vehicle duty cycle, strain, impeller material (aluminum or

titanium), and historical warranty data (Ryder et al., 2002). Maximum

fueling needs to be restricted at high engine speeds in order to limit the

maximum turbocharger speed.

∑ There should be no oil leakage from the bearing area to the compressor

section that is caused by an improper pressure differential across the

compressor for a given sealing design.

Centrifugal compressor design is discussed in Flaxington and Swain (1999),

Japikse (2001), Japikse and Platt (2004); Japikse (1996), Pan et al. (1999),

Moroz et al. (2005); McCutcheon (1978), Came and Bellamy (1982), Japikse

(2000), Yamaguchi et al. (2002), Grif th (2007), Kuiper (2007), and Jiao

et al. (2009). Compressor losses (e.g., tip clearance effect) are discussed in

Spraker (1991), Sharp et al. (1999), Casey (2007), and Nili-Ahmadabadi

et al. (2008).

13.4.3 Turbine matching, aerodynamic design, and

durability

The turbine  ow characteristic map shows that the turbine corrected  ow

rate is a strong function of turbine pressure ratio and a weak function of

turbine speed. The turbine map can be generated from a  ow bench test by

regulating the turbine inlet pressure and the compressor power (load). The

process is repeated until each shaft speed line is obtained. Turbine ef ciency

is a strong function of the blade speed ratio v

and a weak function of

pressure ratio, usually with high ef ciency located at a large pressure ratio

and a high  ow rate. Denote v

as the tip speed of a radial turbine rotor,

= d

13.2

and denote C

as a theoretical velocity that would be achieved if the gas

had expanded isentropically from the turbine inlet condition to the outlet

pressure. C

can be calculated as

p00p

(

–1)

Cc =Cc

Cc 2Cc

1 –

–1)

13.3

Usually, a larger turbocharger has a higher ef ciency because of its lower

clearance-to-wheel diameter ratio and less associated leakage. It should be noted

that the measured turbine map ef ciency is the adiabatic ef ciency multiplied

by the mechanical ef ciency. The ef ciency map is signi cantly affected

by the heat transfer from the turbine to the compressor via the turbocharger

Diesel-Xin-13.indd 879 5/5/11 12:04:45 PM



880 Diesel engine system design

housing at low ow rates. The heat transfer decreases the measured turbine

inlet temperature and increases the compressor outlet temperature, which

causes the compressor efciency articially low and the turbine efciency

articially high. The turbine efciencies measured by different turbocharger

manufacturers may not be directly comparable because of differences in their

measurement methods and gas stand designs. For example, different turbine

inlet temperatures can cause different heat transfer effects.

The turbine ow range is determined by an ‘A/R’ ratio, which is the ratio

of the smallest area of turbine housing inlet passage to the radial distance

from the shaft centerline to the centroid of the inlet area. A smaller A/R

ratio or turbine cross-sectional area produces a higher tangential gas velocity

impinging on the turbine blades. As a result, a higher turbine speed, pressure

ratio, and compressor boost pressure can be achieved. However, a small A/R

also increases the exhaust manifold pressure and the pumping loss. Once

the compressor frame size and the turbine rotor wheel size are selected,

a range of casings with different volute cross-sectional areas or different

nozzle stator rings (with various blade angles) can be chosen for the rotor,

depending on the engine ow requirement. In VGT, either the A/R ratio or

the nozzle vane is adjustable, discretely or continuously. A VGT may have

lower efciency than a xed-geometry turbine due to larger clearances and

ow disturbances. Figure 13.8 illustrates the principle of matching the turbine

with the engine. In the gure, the cross-over point between the line of engine

breathing characteristics at a given engine speed and the line of the air–fuel

ratio requirement determines the required turbine effective area.

The turbine matching guidelines are summarized as follows:

∑ The turbine size needs to be small enough to drive EGR ow at low

engine speeds (e.g., peak torque), and appropriate to deliver sufcient

power to drive the compressor to produce the required air–fuel ratio for

peak torque and lower speeds at full load. The turbine effective area,

which is controlled by the nozzle ring or volute, needs to be selected to

provide the best trade-off between high speeds/loads and low speeds/

loads. If a small turbine area is sized for peak torque, wastegating is

probably needed at rated power in order to prevent over-boosting where

the peak cylinder pressure exceeds the structural limit. Note that the

peak compression pressure in the cylinder may be estimated by

compression

ª p

boost

1.36

13.4

where W is the engine compression ratio. If the pumping loss associated

with a wastegated turbine is not acceptable, the more sophisticated VGT

may be considered at a higher cost.

∑ The necessity of turbine wastegating at various ambient conditions needs

to be checked to prevent turbocharger over-speeding.

Diesel-Xin-13.indd 880 5/5/11 12:04:45 PM



High engine speed (rated power)

Engine operating points move along

this characteristic line with turbine A

Large turbine A

Medium turbine A

Small turbine A

Engine A/F ratio

requirement

Engine A/F requirement

Engine demand of air–fuel ratio and EGR

rate (equivalent to boost pressure) at full

load from peak torque to rated power

Engine breathing

characteristic line at

peak torque speed

Low engine speed (peak torque)

Engine demand

at 75% load

There are 3 key factors in engine–turbine matching:

(1) Based on given volumetric efﬁciency, engine air ﬂow

rate is proportional to boost pressure and engine speed.

(2) Turbo power balance equation indicates certain

turbine pressure ratio is needed to deliver target

compressor boost pressure for A/F ratio and EGR rate.

(3) Turbine area characteristics (turbine ﬂow equation).

50% load

25% load

Corrected turbine ﬂow rate

As engine speed decreases

with turbine area A

Engine breathing

line at rated speed

Boost deﬁcit at peak torque with large turbine

Over-boosting at rated

speed with small turbine

Turbine pressure ratio

13.8 Engine–turbine matching theory.

Diesel-Xin-13.indd 881 5/5/11 12:04:45 PM

