Xin Q. Diesel Engine System Design

Подождите немного. Документ загружается.

146 Diesel engine system design

1. The thermal load (stress) caused by exhaust manifold gas temperatures and

the associated cyclic varying metal temperature eld (both temperature

level and gradient).

2. Stresses generated by the external constraints such as the local restraining

bolt forces and different thermal expansion. The rst and second types

of load are usually the primary loads for the exhaust manifold.

3. The dynamic excitations from the external attached components such as

a turbocharger. The dynamic excitation affects NVH. The mechanical

stresses sometimes cause HCF. In fact, HCF failure problems are rarely

encountered with the exhaust manifold and are mainly due to improper

design of the bracket or mounting.

4. Exhaust manifold gas pressure. This type of load has negligible effect

on exhaust manifold durability.

Thermal cycles are encountered when the engine warms up and cools down

as the engine is started and stopped or running at different speeds and loads.

A thermal load may result in sufciently high thermal stresses to exceed the

yield strength of the exhaust manifold material to cause LCF.

Exhaust manifold gas temperature is a critical system design parameter

that affects turbocharger performance. The exhaust temperature is inuenced

by fueling rate, intake manifold temperature, air–fuel ratio, EGR rate, fuel

injection timing, exhaust port, and manifold heat transfer. In many cases a

low peak cylinder gas temperature correlates to a low exhaust manifold gas

temperature (for example, when the EGR rate is increased). Consequently,

the exhaust manifold gas temperature is often used as a convenient indicator

of the thermal load acting on the in-cylinder components. For example,

engine derating may occur based on exhaust temperature due to the concern

of durability of the cylinder head rather than the exhaust manifold. However,

it should be noted that such a correlation does not always hold true since the

peak in-cylinder gas temperature is affected by the engine thermodynamic

cycle process during the compression and expansion strokes, while the

exhaust manifold gas temperature is affected primarily by the in-cylinder

gas temperature during the exhaust stroke.

Impact of exhaust manifold pressure on durability

Exhaust manifold pressure is one of the most important engine system design

parameters. A thorough understanding of the impact of this parameter on

durability is required in order to specify an appropriate design limit for the

exhaust manifold pressure. Exhaust manifold pressure is governed mainly

by engine exhaust ow rate, turbine area, and exhaust manifold volume. It

affects turbine pressure ratio and hence intake manifold boost pressure. It

also largely affects the engine delta P (i.e., exhaust manifold pressure minus

Diesel-Xin-02.indd 146 5/5/11 11:44:52 AM



147Durability and reliability in diesel engine system design

intake manifold pressure). As the engine power density increases and the

allowable engine-out soot level decreases, higher intake manifold pressure

is required at the rated power condition. This results in the challenging

demand on the engine structural strength to sustain the higher peak cylinder

pressure, higher compressor outlet pressure and temperature, and higher

exhaust manifold pressure. (Note that the compressor out temperature affects

the LCF life of the compressor wheel and housing, the rubber seals, and the

charge air cooler.)

There are two different limits that should not be exceeded in engine design

and performance/emissions calibration due to durability concerns: the engine

delta P and the exhaust manifold pressure. The engine delta P is an indicator

of pumping loss. It also affects exhaust valve oating off the valve seat

during the intake stroke. During the fast transient acceleration, operation in

cold climate or exhaust braking, the EGR valve may stay closed so that a

large amount of exhaust gas ows through the turbine to create a very high

engine delta P. A sufciently high exhaust valve spring preload needs to be

used in order to prevent the exhaust valve from oating off the valve seat. If

the spring preload is too low, excessive exhaust valve bouncing may occur

after the valve oating. If the spring preload is too high, valvetrain friction

and cam stress may be too high.

Exhaust manifold pressure usually has large pulsating amplitude due to

gas wave dynamic effects. A very high exhaust manifold pressure itself does

not cause structural failure of the manifold because the manifold wall is

usually sufciently thick. A gasket is used between the exhaust manifold and

the cylinder head to prevent high temperature exhaust leaks or air seepages

from the outside. The gasket can tolerate a much higher pressure than the

normal operating pressure level of the exhaust manifold. The gasket leakage

problem is usually not caused by the manifold pressure. Instead, the leakage

is caused by the deformation due to thermal fatigue. A high exhaust manifold

pressure itself also does not adversely affect the exhaust valvetrain when the

engine delta P is under control. The maximum allowable exhaust manifold

pressure, often occurring at full load or exhaust braking condition, is usually

not constrained or determined by the following: EGR valve, EGR cooler

pickup joint at the exhaust manifold, turbine shaft bearing, thrust bearing

and seals, and turbocharger oil leakage. Instead, the maximum allowable

design limit of the exhaust manifold pressure is usually determined by the

following factors:

∑ fatigue of the thin EGR cooler tubes

∑ valve stem seals and valve guides

∑ the thin exible bellows in the exhaust or EGR pipes at the turbine

inlet

∑ turbine wheel HCF life (related to strong shockwave pressure pulses)

∑ sensors and their mounting in the exhaust manifold.

Diesel-Xin-02.indd 147 5/5/11 11:44:52 AM



148 Diesel engine system design

In fact, all the above ve items can be improved in design in order to sustain

a higher exhaust manifold pressure so that they do not become a bottleneck

for raising the allowable manifold pressure limit. It should be noted that

the maximum allowable exhaust manifold pressure, being one of the most

important engine system design constraints, should not be casually imposed

by a supplier (e.g., EGR system or turbocharger supplier) without reasonable

justications. Otherwise, the engine manufacturer would be unnecessarily

over-constrained in engine design and performance/emissions calibration.

Thermo-mechanical LCF failure of exhaust manifold

Thermo-mechanical LCF is the primary failure mechanism in the exhaust

manifold. The failures are reected by manifold cracking and gasket leakage,

caused mainly by high exhaust gas temperatures and high temperature

gradients in the metal. Cracks most likely occur in the manifold region where

the deformation is restrained or the temperature gradient is large (e.g., at the

llets). Exhaust manifold leakage is usually caused by plastic deformation

due to the thermal load at the interface between the manifold and adjacent

components. The high compressive stresses in the exhaust manifold are

caused by the high material temperature and restricted thermal expansion

by the screwed connections. When the stress exceeds the yield strength of

the material, plastic deformation or strain occurs. After the engine is cooled

down, the compressive stress changes to local tensile stress which may exceed

the tensile yield strength. The repeated large plastic strain amplitude induced

by the cyclic temperature loading results in thermo-mechanical LCF failure

of the manifold. The high metal temperature affects material strength. Three

mechanisms are involved in the following order of importance: fatigue due to

plastic deformation, creep, and oxidation. At the full-load high-temperature

conditions, the exhaust manifold is subject mainly to compressive loads (i.e.,

out-of-phase load) where creep has a secondary inuence.

Design solutions for exhaust manifold durability

The fundamental design issue in the exhaust manifold is to minimize thermal

stresses and thermal deformation in order to avoid large plastic strain

amplitudes and fatigue cracks. Exhaust manifold durability is generally a

component-level issue, especially for new engines, since proper material

selection and local design details can always solve almost all the problems

related to the gas temperature. However, the material selection is a trade-off

between cost and thermal capability. The exhaust manifold gas temperature

of diesel engines is generally lower than that of gasoline engines.

Based on the maximum exhaust manifold gas temperature requirement,

the manifold type and material can be properly selected. Better material

Diesel-Xin-02.indd 148 5/5/11 11:44:52 AM



149Durability and reliability in diesel engine system design

can improve thermal fatigue life and reduce creep effects. The oxidation

problem can often be solved by using a material with higher resistance to

oxidation but often at a trade-off against high temperature strength (Gocmez

and Deuster, 2009). Exhaust manifold material selection and the properties

of high temperature oxidation and thermal fatigue were discussed by Park

et al. (2005).

Exhaust manifold LCF life is strongly affected by local design details

(e.g., shape, assembly constraints) and gas temperature. If all the design

solutions are exhausted and the LCF plastic strains still cannot be controlled,

the exhaust manifold gas temperature has to be reduced in engine operation.

This certainly affects engine performance. Gocmez and Deuster (2009)

provided a review of exhaust manifold failure modes, lifetime prediction,

and design optimization.

Thermo-mechanical fatigue life prediction of exhaust manifold

Thermal fatigue life testing of the exhaust manifold takes a long time to

complete. Predicting thermal fatigue life of the exhaust manifold using

simulation is important at both the concept design stage and the detailed

design stage. Simulation can help determine the acceptable gas temperature

level, identify critical failure locations, and guide component design to

remove local structural weakness.

Full structural FEA of exhaust manifold includes the evaluation of bolt

force, metal temperature eld, stress–strain response, fatigue life, and gasket

pressure distribution. The metal temperature distribution is usually the most

important boundary condition for LCF analysis. The concern with hardware

sizing in engine system design is mostly at the full-load condition where

the exhaust manifold gas temperature reaches the highest. A simplied

durability analysis may use the highest temperature that is encountered during

thermal cycles to evaluate the exhaust manifold fatigue life. It should be

noted that the highest exhaust gas temperature does not necessarily cause

the highest stress or strain in the exhaust manifold. Sometimes the lowest

temperature during a thermal cycle can cause the highest tensile stress. The

real question for a system engineer is how high the exhaust temperature can

be designed at steady-state full-load conditions so that the corresponding

thermal strains produced during both full-load durability validation test and

slow varying transient thermal cycles are acceptable. Moreover, advanced

exhaust manifold durability and fatigue life modeling should consider all

three major mechanisms involved in high temperature LCF: fatigue (due to

cyclic plasticity), creep, and oxidation.

Simplied early-stage exhaust manifold thermo-mechanical modeling can

be conducted to screen the concept designs by minimizing a thermal stress

index parameter (Park et al., 2006), which is the ratio of the elastic effective

Diesel-Xin-02.indd 149 5/5/11 11:44:52 AM



150 Diesel engine system design

stress to the yield strength. This is a quicker calculation than the full FEA

approach.

Exhaust manifold thermo-mechanical fatigue analysis has been investigated

by several researchers during the past decade. Watanabe et al. (1998) developed

an analytical method to use the plastic strain as the evaluation criterion to

judge the thermal fatigue life of gasoline engine exhaust manifolds. Their

simulation correlated well with the engine durability test results. Lederer et

al. (2000) provided a detailed description of the analysis methodology for

the thermo-mechanical fatigue of diesel engine exhaust manifolds, including

thermal boundary calculations, material characteristics, and fatigue life

estimation. Mamiya et al. (2002) discussed a detailed process of exhaust

manifold temperature calculation with heat transfer models for both the inner

and outer walls. They used a simplied approach to estimate the LCF life by

assuming the dominant factor contributing to the cracks was the large thermal

plastic strain range (derived from FEA). The strain concentrated regions were

the critical locations subject to failure. They used the classical Manson–Cofn

relation to model the fatigue life. Their simplied methodology with acceptable

accuracy was validated by comparison with the fatigue life test data. Ahdad

and Soare (2002) proposed using an energy-based thermo-mechanical fatigue

criterion to analyze exhaust manifold life. Delprete and Rosso (2005) detailed

a transient structural simulation (as a function of time) with FEA for exhaust

manifold stress–strain analysis. They found that the maximum equivalent

stress (the Von Mises stress) in the exhaust manifold occurred at the end of

the thermal cycle where the metal temperature falls to the minimum level

within the cycle. They concluded that a simplied FEA model was able to

simulate the thermo-structural behavior of the exhaust manifold with short

computing time and thus was attractive to engine system design.

2.9.3 Engine valve durability

Valvetrain durability issues

Engine valvetrain durability issues include the following: thermo-mechanical

fatigue of the valves; mechanical fatigue of the valve spring; fatigue and

wear of the cam; valve seat wear; valve stem scufng and wear; and wear

at other interface (e.g., rocker shaft, rocker pad). Many valvetrain durability

issues are related to system-level loading or design parameters, for example,

exhaust temperature, valve spring force, and valve seating velocity. Higher

power density of the engine imposes a higher cylinder pressure load and

thermal load on the valves and valve seats. The requirements for lower

emissions and fuel consumption demand more aggressive cam lift proles

and the associated better valvetrain motion control by using higher valve

spring load. The requirement for extended durability demands the valvetrain

Diesel-Xin-02.indd 150 5/5/11 11:44:52 AM



151Durability and reliability in diesel engine system design

design to be more durable. Protability requires the valvetrain design to be

cost effective in material selection and design complexity. All these demands

require the system engineer to consider valvetrain durability in the early

stage of the diesel engine design.

Valve material and design

Common valve failure modes include the following ve phenomena: valve

head cracking, valve stem fracture, valve stem sticking and scufng, valve

dishing, and face guttering (Johnson and Galen, 1966). In the past half a century,

exhaust valve materials and design have evolved to a great extent, while intake

valve materials and design have evolved at a much lesser pace (Schaefer et

al., 1997). The changes in modern diesel engine valve design are driven by

performance requirements. For example, better engine breathing demands the

use of four-valve-head design. Using four valves per cylinder results in smaller

size for each valve. Better valvetrain dynamics demands a lighter valvetrain.

Valve mass has been minimized by reducing valve stem diameter.

The most important parameters for valve durability are the operating

temperature, the imposed stress level, and the corrosive combustion products

to which the valve is exposed. The magnitude of the cyclic tensile stress in

the valve head is a function of the peak cylinder pressure. Exhaust valve

metal temperature is closely dependent upon the exhaust gas temperature

and cooling. Modern diesel engines have two performance characteristics

related to the exhaust temperature. First, as the EGR rate is increased and

emissions become lower, the exhaust temperature becomes lower (if

the power density does not increase). Second, there is very little or no gas

scavenging from the intake port directly to the exhaust port due to the need

to drive EGR ow with the exhaust manifold pressure higher than the intake

manifold pressure. Therefore, compared with the non-EGR engines, the air

cooling effect from scavenging fresh air for the exhaust valve is much less

in EGR engines. Moreover, note that the intake valve temperature is much

cooler than the exhaust valve temperature. This temperature difference results

in different valve thermal expansion and valvetrain lash for the intake and

exhaust valves.

The temperature distribution inside the valve, and the temperature

effect on valve steel properties, thermal/mechanical stresses and valve seat

design have been thoroughly researched in the past. The modern engine

requirements on valve design have been focused on the high-temperature

fatigue strength and the wear resistance to achieve longer life of the valve

and enable higher power density of the engine. Valve material selection is

a component-level design detail. Sufcient strength of the material is the

key to resist high-temperature fatigue. The material selection is a trade-off

between performance/durability and cost.

Diesel-Xin-02.indd 151 5/5/11 11:44:52 AM



152 Diesel engine system design

During the process of selecting valve alloy and heat treatment method, the

corrosion resistance of the material needs to be considered since corrosion

signicantly reduces the fatigue strength of the material. Valve corrosion is

affected by fuel property. The use of low-sulfur diesel fuel and low-sulfated-

ash oil causes less suldation corrosion at high temperatures for the intake

and the exhaust valves. EGR gas may cause acid-based corrosion for the

intake valve.

The durability of internal combustion engine valves has been extensively

researched in the past several decades (Newton, 1952; Newton et al., 1953;

Tauschek, 1956; Johnson and Galen, 1966; Giles, 1966; Wang, 2007). An

important comprehensive review on the heavy-duty diesel engine valve

materials and designs in the last fty years was provided by Schaefer et al.

(1997). The review covered all major design guidelines for valve durability

related to material and design.

Valve seating velocity and valve stem fracture

The durability risk of valve stem fracture due to fatigue is governed by

valve seating impulse (i.e., valve velocity multiplied by mass), which is

largely controlled by cam prole design. The cam design also affects engine

performance and valvetrain dynamics. A cyclic tensile stress in the valve

head and the valve stem is generated by the valve seating event. At high-load

conditions, due to large thermal distortion of the cylinder head the valve

does not seat uniformly along the circumferential direction on the valve seat.

This creates additional seating stresses on the valve and the seat, and also

creates bending stresses at the stem–retainer interface.

The valve seating velocity is affected primarily by cam acceleration prole

and secondarily affected by cam closing ramp design. Valve seating velocity

also affects engine NVH. The intake valve seating velocity may become too

high if the cam acceleration is too aggressive (for achieving high volumetric

efciency) or poorly designed. The exhaust valve seating velocity may become

too high for the following reasons: very aggressive cam acceleration for

achieving low pumping loss; very aggressive exhaust cam acceleration at the

cam closing side for reducing the high recompression pressure that is built

up by a small turbine area used in a high-EGR diesel engine; poorly designed

exhaust cam acceleration shape; and poorly designed cam closing ramp.

2.9.4 Cam fatigue and stress

Introduction of cam stress

Most of the valvetrain cam and follower distress can be classied into two

main categories: fatigue and wear. Cam contact fatigue is often referred to as

spalling (or pitting, aking). Spalling is a metal failure due to fatigue under

Diesel-Xin-02.indd 152 5/5/11 11:44:52 AM



153Durability and reliability in diesel engine system design

high contact stresses with internal cracks spreading up to the surface with the

consequent material removal. Spalling of the contact surface can be avoided

by strengthening the material or reducing the maximum Hertzian elastic stress.

The Hertzian stress is a good indicator of this type of HCF fatigue failure.

Scufng is the result of micro-welding of the contact surfaces caused

by oil lm breakdown due to high contact pressures. Scufng occurs when

the contact becomes partially plastic. Severe adhesive wear is commonly

known as scufng in valvetrain. Scufng and wear are associated with the

lubrication condition and the temperature in the contact area.

It should be noted that cam stress is not the only parameter to be used

to judge the cam–follower durability. The other two important parameters

are oil lm thickness (to be detailed in Chapter 10) and ash temperature

(for at-faced followers, discussed in Section 2.11). Figure 2.6 shows the

calculated cam stress and oil lm thickness for a pushrod valvetrain in a

heavy-duty diesel engine. It is observed that the oil lm thickness is lower

at the lower speed (700 rpm). The oil lm thickness is low at the cam nose

where the cam stress is the highest. The oil lm thickness reaches almost

zero on the cam opening ank at approximately 40° cam angle before the

cam nose where the lubricant entraining velocity is zero. At that location

the cam stress remains high.

The state of deformation and the stress existing between two elastic bodies

in contact under load were established by Hertz. The Hertzian stress (a

compressive stress) is the only principal stress used for cam stress evaluation.

Cam stress is an important engine system design parameter which affects the

design decisions of rocker arm ratio, cam prole shape, cam acceleration,

spring force, cam base circle diameter, roller diameter, etc. Cam stress is

determined mainly by cam force and cam radius of curvature (Fig. 2.7).

Higher spring force, vibration force, and gas load make the cam stress higher.

The Hertzian stress is proportional to the square root of the cam force. At

low engine speeds, the cam stress as a function of cam angle has a smooth

distribution shape with the peak stress occurring at the cam nose (Fig. 2.8).

The cam nose stress may reach the highest value at the engine cranking

speed where the dynamic inertia effect is negligible. At high engine speeds

or under high gas loading on the valvetrain, the cam stress exhibits vibratory

characteristics and may have peak values in the cam ank region.

Cam Hertzian stress calculation was discussed in great detail by Turkish

(1946), Korte et al. (1997, 2000), and Krepulat et al. (2002), regarding

different materials’ cam stress limits and design guidelines.

Cam stress with roller follower

Roller follower has been increasingly used to avoid the problem of excessively

high ash temperatures encountered in the sliding contact with at followers.

Diesel-Xin-02.indd 153 5/5/11 11:44:52 AM



–180 CMP 0 POW 180 EXH 360 INT 540

BDC TDCF BDC TDC BDC

Crank angle (degree)

–180 CMP 0 POW 180 EXH 360 INT 540

BDC TDCF BDC TDC BDC

Crank angle (degree)

–180 CMP 0 POW 180 EXH 360 INT 540

BDC TDCF BDC TDC BDC

Crank angle (degree)

–180 CMP 0 POW 180 EXH 360 INT 540

BDC TDCF BDC TDC BDC

Crank angle (degree)

Exhaust cam oil ﬁlm thickness at 700 rpm

Exhaust cam oil ﬁlm thickness at 2300 rpm

Exhaust cam stress and cam lift at 700 rpm Exhaust cam stress and cam lift at 2300 rpm

0.6

0.2

250000

200000

150000

100000

50000

250000

200000

150000

100000

50000

6.912

5.530

4.147

2.765

1.383

0.000

6.912

5.530

4.147

2.765

1.383

0.000

Oil ﬁlm thickness (micron)

Pressure (psi)

Lift (mm)

Cam stress Cam lift Cam stress Cam lift

2.6

Comprehensive

evaluation of

cam durability.

(CMP:

compression

stroke; POW:

power stroke;

EXH: exhaust

stroke; INT:

intake stroke)

Diesel-Xin-02.indd 154 5/5/11 11:44:53 AM



155Durability and reliability in diesel engine system design

Roller follower has also been used to solve the problem of increased stresses

due to higher cam load. A rolling contact design is required when the cam

load exceeds the limits of stress and ash temperature for a sliding contact.

Roller follower can also greatly reduce valvetrain frictional losses. Moreover, it

enables the use of a negative radius of curvature to achieve a more aggressive

cam acceleration prole for better engine breathing.

Rolling contact causes an increase in contact stress between the cam and

the follower. However, the maximum allowable stress limit of a rolling contact

is much greater than the limit of a sliding contact. If a crowned roller is used

in order to avoid the edge loading caused by the misalignment between the

roller and the cam lobe, the maximum allowable cam stress limit can be further

increased. The stress limit is different for various materials of different costs.

The limit is also related to the lubrication condition on the contact surface.

Different engine companies have established different design criteria of the

maximum allowable cam stress limit for their products. The textbook values

(e.g., Turkish, 1946) can only be used as a reference guideline.

Smaller cam radius of curvature and smaller roller crown radius (i.e.,

larger crown height) make the cam stress higher (Fig. 2.8). A larger roller

crown radius enlarges the longer axis of the contact ellipse between the cam

and the roller. The calculated contact ellipse area should be kept within the

width of the cam lobe and should not exceed the roller width by more than

10–20%. Roller follower design and durability details were provided by

Korte et al. (2000).

An in-depth discussion about the comparison of the cam stress calculations

between the analytical Hertzian equation and the more advanced FEA approach

0 90 180 270 360 450

Crank angle (degree)

Cam radius of

curvature

Pushrod force

Cam stress

Valve lift

Cam lift

2.7 Illustration of pushrod force and cam stress.

Diesel-Xin-02.indd 155 5/5/11 11:44:53 AM

