Xin Q. Diesel Engine System Design

Подождите немного. Документ загружается.

418 Diesel engine system design

characterized by the in-cylinder pressure differential Dp

cyl

= (p

ExhaustManifold

– p

IntakeManifold

) + (Dp

+ Dp

) = (engine delta P) + (Dp

+ Dp

). The Dp

is the pressure drop across the ow restrictions of the intake manifold, ports,

and valves. The Dp

is the pressure drop across the ow restrictions of the

exhaust valves, ports, and manifold. The pumping loss consists of two parts:

engine delta P and those ow restrictions related to the volumetric efciency.

In order to maximize the pumping loss, it is desirable to increase the engine

delta P and/or Dp

+ Dp

6.1.9 Analysis of engine brake retarding power

requirement

In order to determine an appropriate design target for engine brake retarding

power, its effect on vehicle braking performance can be analyzed using the

method shown in Fig. 6.5 for an example of downgrade driving. The ‘high,

medium, and low’ levels shown in the gure refer to three different design

targets for the retarding power of a compression brake. The ‘ZWB’ refers to

‘zero-wheel-braking’, which means that the driver does not need to use the

service brakes. The ZWB anchor points are calculated using the retarding

power of different engine brakes with the vehicle torque balance equation

6.1. The ZWB curves are characteristic curves on which the vehicle travels

at different control speeds as an ideal situation. Different vehicle weights,

road grades, or tire–road rolling friction coefcients give different ZWB

curves. Along each gear line, on the right side of the ZWB point, the vehicle

decelerates. On the left side of the ZWB point, the vehicle accelerates. This

approach can intuitively analyze the vehicle braking power requirement

and determine an appropriate design target for the engine brake retarding

power. The analysis is also useful for determining a suitable design target

for the maximum allowable operating speed of the valvetrain (i.e., the

valvetrain separation speed or no-follow limit) based on driving and braking

conditions.

6.1.10 Integrated design of engine brake and powertrain

system

Engine brake design integration with the engine

Engine brakes directly control the engine air path with various types of air

control valves. Engine brake design is an integral part of the powertrain

system design for vehicle safety. It is also an integrated part of engine air

ow management for modern diesel engines, not only for braking but also

related to ring. In the 1960s–1980s, compression brakes were primarily

an aftermarket accessory added to the base engine to control exhaust valve

Diesel-Xin-06.indd 418 5/5/11 11:50:30 AM



Vehicle braking performance with manual transmission in heavy duty applications at

downhill motoring or braking, vehicle weight 25000 lbs, road grade 7.5%

1 mph = 1.609 km/hr

1 lbm = 0.4536 kg

No-brake motoring ZWB curve

VGT exhaust brake ZWB curve

Engine brake (high) ZWB curve

Engine brake (medium) ZWB curve

Engine brake (low) ZWB curve

Conventional exhaust brake ZWB curve

Data in parenthesis are

vehicle acceleration (–) or

deceleration(+) in ‘mph/sec’

with VGT exhaust brake

Note: ZWB refers to zero

wheel-braking

Vehicle speed (mph)

140

130

120

110

100

Vehicle accelerate

Vehicle decelerate

500 1000 1500 2000 2500 3000 3500 4000

Engine speed (rpm)

Gear 1

Gear 2

Gear 3

Gear 4

Gear 5

Gear 6

Gear 7

Gear 8

Gear 9

Gear 10

(–1.04)

(–0.86)

(–0.20)

(–0.70)

(–0.32)

(0.10)

(0.62)

(1.29)

(2.00)

(1.46)

(0.97)

(0.45)

(0.09)

(–0.50)

6.5 Effect of engine brake retarding power on vehicle braking performance.

Diesel-Xin-06.indd 419 5/5/11 11:50:30 AM



420 Diesel engine system design

actuation, installed by vehicle manufacturers or various types of service outlets.

Later, the engine companies added compression brakes as a factory installed

option (Freiburg, 1994). Exhaust brakes went through a similar history. The

integration of a compression brake on an engine requires consideration of

turbocharger matching, valvetrain exibility, braking noise, etc. The integration

of an exhaust brake requires consideration of using the exhaust brake as a

exible device for engine warm up, aftertreatment thermal management,

and exhaust restriction control for the air/EGR system. Moreover, in the

structural design related to the compression brake, the driveline load during

retarding can be similar or even higher than that in engine ring operation.

Therefore, the vehicle system design should consider retarding as the possible

worst or the highest loading case. In summary, engine brake design cannot

be isolated to a supplier or restricted to a subsystem. The design direction

of engine brakes is moving toward increased integrated functionality to the

engine to assist both braking and ring operations.

Retarder design optimization in the braking system

This subsection discusses the optimization criteria for all the retarders

in general, including engine brakes. In the optimization of the integrated

retarder and service brake system, optimal design can be conducted for

either maximum deceleration or maximum brake lining life by an appropriate

distribution of the retarding energy absorbed between the retarder and the

service brakes (Limpert, 1975). To maximize the deceleration or minimize

the stopping distance, the entire brake system needs to be designed such

that all the axles are braked simultaneously near their tire–road adhesion

limits with a minimum delay. On the other hand, to maximize the service

brake lining life, a ‘retarder brake-life extension factor’ may be considered.

O’Day and Bunch (1981) proposed such a factor dened as the ratio of the

brake wear rate using the retarder to the brake wear rate without using the

retarder. They reported that the range of this factor varied from slightly over

1.0 to a high value of 8.0–9.0, and tended to cluster as functions of retarder

type, vehicle size, vehicle application and geographic region of operation.

For instance, the factor in transit operations is around 8.0 with the use of

electric retarders. Schreck et al. (1992) provided a brake lining life comparison

between using and not using a ZF hydraulic retarder. Their investigation on

168 heavy vehicles in seven transportation companies showed that the brake

lining life with the retarder was on average approximately four times that of

the brake lining life without the retarder. Greathouse et al. (1971) reported

that Mack’s Dynatard compression brake reduced the need for service brake

usage by as much as 87% in downhill driving, and the brake was activated

about once per mile in normal over-the-road driving.

Diesel-Xin-06.indd 420 5/5/11 11:50:30 AM



421Engine brake performance in diesel engine system design

Coordinated engine and transmission controls for engine brakes

There are several system integration issues related to vehicle controls for

engine brakes, transmissions, and service brakes. In the engine fueling control,

operating the conventional compression brake during fueling would reduce

the retarding power and increase the valvetrain loading unacceptably because

the braking valve would experience a higher cylinder pressure. Operating the

exhaust brake during fueling may result in excessively high soot or smoke

due to the choked low air ow rate.

In order to activate the engine brake in retarder–transmission integration,

the clutch needs to be engaged, and no fuel needs to be injected into the

cylinder (i.e., foot off the pedal). Engine brakes may experience some losses

of retarding power in the torque converter of automatic transmissions.

Transmission lockup may ensure the minimum retarding power loss for

the wheels. Many electronically controlled automatic transmissions can

be programmed to downshift when the retarding operation is requested,

allowing the optimum engine speed to be selected for the best engine brake

performance. If a faster speed than the downhill control speed is desired, a

higher gear in the transmission can be selected to reduce the running speed

of the engine or the primary retarder, or alternatively a lower retarding

power level can be selected if the retarder has progressive/variable levels

of power.

The retarding torque of the engine brake transmitted to the driven wheels

can be interrupted by disengaging the clutch or by placing the transmission

in neutral. As a comparison, in the case of driveline retarders, once the brake

is applied it can be disconnected from the driven wheels only through the

release of the brake control. If the engine brake is operated with the clutch

disengaged, the engine may inadvertently stall to a very low speed because no

vehicle load is attached to the engine so that the engine speed will drop very

quickly to a low speed and then stall. Stalling the engine is not acceptable

for the later ring operation. The engine brake needs to be turned off below

a certain engine speed in order to avoid stalling the engine.

Coordinated service brake controls for engine brakes

The retarder is not a substitute for the service brakes. The service brakes

need to be used to bring the vehicle to a complete stop. The retarder can be

used in conjunction with the service brakes to provide a combined retarding

power. The engine brake design and control need to be coordinated with the

existing wheel-antilock brake system to ensure a safe and optimal braking

performance under all operating conditions. Ideally, the driven wheels need

to be kept within the optimum slip range without over-braking and under-

braking.

Diesel-Xin-06.indd 421 5/5/11 11:50:30 AM



422 Diesel engine system design

Because the vehicle retarders act only on the driven axle, there is a high

risk of over-braking the driven axle to lock it prematurely due to the change

in braking force distribution between the axles. The over-braking and the

associated directional instability are especially prone to occur with empty

vehicles travelling on road surfaces that have low adhesion coefcients.

Overuse of a retarder on a slippery road or with empty load can lead to the

dangerous ‘jackknife’ instability for trucks. If there is a tendency of wheel

lockup during braking, a fast cut-out of the retarder is immediately needed

in order to avoid over-braking. Haiss (1992) indicated that cut-out times of

a maximum 0.2 s up to 10% of the residual braking torque pose hardly any

problems on braking comfort and directional stability. The antilock brake

system may compensate for the altered braking force distribution caused by

the retarders. When the antilock system detects a slippage or a wheel lock,

it may instantly shut off the retarder in addition to controlling/adjusting the

service brakes. A central brake control management system needs to coordinate

the operations between the service brakes, the engine brake, the drivetrain

retarder, the antilock system, and the coupling force control system.

6.2 Drivetrain retarders

6.2.1 Types of drivetrain retarders

Drivetrain retarders include the following types: (1) the friction brake, i.e.,

the transmission brake or the clutch brake (Jahn, 1989) and the friction clutch

(Spurlin and Trotter, 1982); (2) the hydrodynamic or hydraulic retarder; (3)

the electric eddy-current, electromagnetic and permanent-magnet retarder

(Kubomiya et al., 1992); or (4) a combination of them.

The driveline retarder can be located before the transmission input (i.e.,

between the torque converter and the transmission gear range package) or

after the transmission output. Different locations result in differences in

design packaging, speed–torque capability and braking smoothness. A primary

retarder can utilize the transmission gear shift to change its operating speed

and braking torque. A secondary retarder operates at the driveline speed and

there is no abrupt torque change or torque spike due to transmission gear shift

as in the case of a primary retarder. Although the capability of regulating

the retarder torque through changing its speed is lost, the secondary retarder

may be benecial in applications where braking smoothness and comfort

are the major requirements.

The detailed introduction of drivetrain retarder design, construction,

performance and the impact on vehicle performance is elaborated by Göhring

et al. (1992). An important review of the design and system integration

considerations for drivetrain retarders is provided by Haiss (1992). The basic

principle of hydraulic retarders is explained by Rao (1968). Klemen et al.

Diesel-Xin-06.indd 422 5/5/11 11:50:30 AM



423Engine brake performance in diesel engine system design

(1989) introduce the design approach, control and durability testing for an

Allison transmission output retarder.

6.2.2 Torque characteristics of drivetrain retarders

On very steep downgrades (e.g., 10–12%) the vehicle speed has to be

maintained low enough to ensure a safe emergency stop. At low vehicle

speeds, many drivetrain retarders installed after the transmission output

become insufcient due to fast torque drop at the low shaft speed. Schreck et

al. (1992) indicated that the application of the secondary retarders in 40 ton

trucks is limited up to approximately 10% road grade, depending on the axle

ratio. Such a capability may cover the majority of European downhill driving

applications. It should be noted that the performance of the engine brakes

and the primary retarders installed before the transmission are not limited by

the low vehicle speed because a low gear in the transmission can be selected

to increase the speed of the primary retarder to achieve high power.

The retarding power of the hydrodynamic retarder is a function of the

relative speed of the retarder elements. It is important to increase the low-

speed torque while keeping the high-speed power below the cooling design

limit in retarder design. The hydrodynamic retarder usually has a relatively

at torque curve from the medium speed to the high speed. It is very capable

of generating much higher power than the engine ring rated power. At low

speeds, the blades stall and the torque is low. The hydrodynamic retarder

also has non-negligible power loss at idle running (Jahn, 1989).

The electromagnetic or electric eddy-current retarder may develop higher

torque at low speeds than the hydrodynamic retarder. It is suitable for stop-

and-go driving applications (e.g., pickup and delivery trucks, refuse trucks,

city buses). As pointed out by Göhring et al. (1992), the electrodynamic

retarder is usually more suitable at low vehicle speeds, while the hydrodynamic

retarder is usually more effective at high vehicle speeds. They are both free

from wear and very quiet.

Spurlin and Trotter (1982) described a design of an Allison transmission

output retarder which combined a friction clutch and a hydraulic retarder

so that the low-speed torque was signicantly increased by the friction

clutch.

6.2.3 Cooling and thermal protection of drivetrain

retarders

Usually, air cooling is not sufcient for the hydrodynamic retarder due to its

high heat rejection. The cooling medium in the hydrodynamic retarder can

be oil or transmission uid (e.g., the ZF or Allison retarders) or water (e.g.,

the Voith Turbo’s Aquatarder). The vehicle cooling capacity (e.g., the size

Diesel-Xin-06.indd 423 5/5/11 11:50:30 AM



424 Diesel engine system design

of the radiator and the fan) often needs to be determined based on the heat

rejection requirement in retarding instead of ring for the hydrodynamic

retarder. The continuous braking power capability of the hydrodynamic

retarder is dependent on the cooling capacity of the vehicle radiator. If the

cooling capability is not sufcient, a retarding torque limit based on coolant

temperature control is required in order to avoid overheating the coolant. An

analysis of the continuous braking performance of a ZF hydraulic retarder

was provided by Schreck et al. (1992). They addressed the retarder power

derating based on coolant temperature control.

The heat generated in the electromagnetic retarder is dissipated to the

surrounding air (i.e., ambient-air cooled) and does not need cooling water or

oil (Habib, 1992). The electromagnetic retarder is challenging in frictional

heat management. If the electromagnetic retarder runs hot in its rotor, its

braking torque decreases due to the required thermal protection, for example

by cutting off the current in some coils. Therefore, some electric retarders

may exhibit certain time-dependent characteristics of retarding power decay

due to overheating during a period of braking operation (e.g., a 20-minute

braking).

6.3 Exhaust brake performance analysis

6.3.1 Conventional exhaust brake

The conventional exhaust brake uses a shut braking valve (usually a buttery

or sometimes a at slide throttle valve) in the exhaust pipe mounted normally

at the turbine outlet. Mounting the brake at the turbine inlet introduces

design challenges such as tight packaging space, interference to the exhaust

ow in the exhaust manifold and at the turbine entry, harsh environment

due to the higher exhaust temperature, and so on. The braking valve is

usually pneumatically actuated by a compressed-air type actuator (e.g., in

large vehicles) or a vacuum actuator (e.g., in medium or small vehicles).

Sometimes the braking valve is electrically activated by solenoid without

the air. The exhaust brake has low weight, small size and low cost with high

payback. It has been used as a standard device in many vehicles. Several

brands of the exhaust brake exist in the aftermarket, including the Jacobs

Exhaust Brake, Pacbrake’s exhaust brake, US Gear’s D-celerator, and BD’s

Engine Brake.

As pointed out by Kuwano et al. (1983), the popularity of diesel vehicles

has been driving the wide usage of the exhaust brake. The exhaust brake

is usually used with diesel engines because gasoline engines cannot handle

the high exhaust manifold pressure created during braking. Another reason

is that hydraulic lash adjusters have been widely used in gasoline engines,

and they cannot tolerate the exhaust valve oating or bouncing caused by

Diesel-Xin-06.indd 424 5/5/11 11:50:30 AM



425Engine brake performance in diesel engine system design

the exhaust brake operation. The use of the exhaust brake on an engine with

hydraulic lash adjusters can cause engine damage due to valve-to-piston

contact when the increased exhaust manifold pressure causes the exhaust

valve to oat off its seat and the associated lifter pumping-up occurs.

Usually, when the engine brake is activated fueling should be terminated.

However, if the fuel injection amount is set to idle fueling, a small amount

of fuel will be injected into the cylinder to produce some ring power. This

is undesirable because it reduces the engine retarding power. Moreover,

using the exhaust brake without the interruption of fuel injection will create

a large increase in the carbonaceous residue in the combustion chamber and

in the engine oil, as well as high smoke due to insufcient air, as reported

by Sequeira and Faria (1984).

The exhaust brake is often used together with the glow plug to help speed

up the engine warm-up. With the braking valve partially closed in fueling

condition, the engine is loaded with higher back pressure and hence has to

burn more fuel than in the normal idle mode after the engine start. This can

promote the rise of coolant temperature during the warm-up.

Durability issues of the exhaust brake normally include valve sticking,

thermal shock and thermal fatigue of the braking valve. The heat-resistant

shaft seal of the brake valve and the carbon deposits on the movable parts are

sometimes also problematic. The design and construction of the exhaust brake

are elaborated by Kuwano et al. (1983). Exhaust brake design constraints related

to engine gas pressures and temperatures are discussed in Section 6.3.4.

Although their retarding power at high engine speeds may be satisfactory,

both the exhaust brake and the compression brake are poor in retarding torque

at low engine speeds. The response of the exhaust brake is instantaneous to

full retarding power, while a compression brake usually requires some time

for the turbocharger to spool up to reach the maximum boost level. For that

reason, the exhaust brake is superior in fast changing and repetitive driving

cycles since it does not rely on the buildup of turbocharger boost pressure.

On the other hand, the compression brake is superior in continuous steady

driving conditions such as on long downgrades.

During exhaust braking, the in-cylinder cycle processes in the compression

stroke and the expansion stroke are not adiabatic. The resulting indicated

retarding power, although small, increases with the engine compression

ratio. Pumping loss is the dominant source of the retarding power of the

exhaust brake. The pumping loss power is proportional to engine delta P

and is also related to the square of the air ow rate at the intake valve and

the exhaust valve. The retarding power characteristic of the exhaust brake

is essentially a function of engine speed, engine displacement, and exhaust

manifold pressure. Unlike the compression brake, the turbocharger has little

inuence on the retarding power of the conventional exhaust brake because

its retarding power comes from the pumping loss strokes instead of the

Diesel-Xin-06.indd 425 5/5/11 11:50:30 AM



426 Diesel engine system design

indicated strokes. On the other hand, the use of the exhaust brake at high

engine speeds may actually adversely affect the turbocharger performance

by reducing the pressure ratio across the turbine. The interaction between

the exhaust brake and the compression brake as well as their combined

functions for naturally aspirated diesel engines are discussed by Schmitz et

al. (1992, 1994) and Imai et al. (1996).

When a xed opening of the exhaust braking valve is sized for high engine

speeds, the opening size would be too large at low speeds, resulting in low

retarding power. Advanced exhaust brake designs employ an electronically-

controlled pressure-limiting bypass valve or essentially a variable-geometry

braking valve so that the braking valve opening can be adjusted to much

smaller at low speeds to build up very high exhaust manifold pressures to

produce high power or torque. The valve area is opened more at high speeds

to prevent the exhaust manifold pressure from being excessively high. Such

a variable-geometry exhaust brake is the modern design direction to enhance

the low-speed retarding performance of the exhaust brake.

6.3.2 Variable geometry turbine (VGT) exhaust brake

As use of the variable geometry turbine has gained popularity on diesel

engines, using the VGT as a braking device has also become common. By

closing the vane opening of the VGT, the turbine speed increases and a high

exhaust manifold pressure can be developed. However, the intake manifold

boost pressure also increases at the same time, although its increase is less

than that of the exhaust manifold pressure. As a result, the engine delta P

and the pumping loss increase as the VGT vane is closed. Compared with

the conventional exhaust brake, the VGT exhaust brake usually has a lower

engine delta P and retarding power, but a much higher turbocharger speed and

engine air ow rate. The higher air ow helps reduce the exhaust manifold

gas temperature and the thermal load on the components during braking.

6.3.3 Variable valve actuation (VVA) exhaust brake

Variable valve actuation used as a braking technique provides another means

to increase the retarding power. Generally, there are two methods to increase

the pumping loss: increase engine delta P and reduce volumetric efciency.

The latter method can be achieved by increasing the pressure drops across

the intake and exhaust valves as the air ows through the valves. The VVA

brake uses the second method by throttling the engine intake or exhaust

valve (or both) to decrease the cylinder pressure during the intake stroke or

increase the cylinder pressure during the exhaust stroke. This technique is

more effective for the exhaust valves due to the much greater potential. The

benet of an intake VVA brake is small just like the intake-throttle brake.

Diesel-Xin-06.indd 426 5/5/11 11:50:30 AM



427Engine brake performance in diesel engine system design

6.3.4 Effect of exhaust pressure pulses and interaction

with valvetrain

Exhaust brake retarding performance

The design constraints of the exhaust brake include: (1) the exhaust gas

temperature and the associated component temperatures such as the injector

tip temperature at very choked engine air ow rate; and (2) the exhaust

manifold pressure and the associated valvetrain dynamics issues. The exhaust

brake retarding power is governed mainly by the following factors: engine

displacement, engine speed, exhaust manifold conguration, engine ring

order, and exhaust valve spring preload. The last three factors are related to

the instantaneous exhaust manifold pressure pulses in braking. Figure 6.6

shows that the retarding power density of the conventional exhaust brake is

basically linearly proportional to the cycle-average exhaust manifold pressure.

The allowable exhaust manifold pressure is limited by excessive exhaust

valve oating. If the limit is too low, the performance of the exhaust brake

would not be substantial. The maximum allowable exhaust pressure in the

non-EGR engines usually varies from 20 psi (1.4 bar) to 60 psi (4.1 bar)

gauge pressure, depending on the exhaust valve spring preload. In order to

achieve higher retarding power without excessive valve oating, it would

require an upgraded (higher) spring preload to increase the allowable pressure

limit, for example to 70 psi (4.8 bar). The upgrade needs to be conducted

carefully to ensure the exhaust cam stress is still acceptable.

1 2 3 4 5 6 7

Maximum allowable exhaust manifold pressure

(cycle average, bar absolute)

Speciﬁc retarding power at 2000

rpm (hp/liter)

6.6 Exhaust brake performance as a function of maximum allowable

exhaust manifold pressure.

Diesel-Xin-06.indd 427 5/5/11 11:50:31 AM

