Xin Q. Diesel Engine System Design

Подождите немного. Документ загружается.

–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 stress at 2300 rpm full load

Exhaust cam stress at 3200 rpm motoring

Exhaust cam stress at 100 rpm motoring Exhaust cam stress at 1000 rpm full load

250000

200000

150000

100000

50000

250000

200000

150000

100000

50000

250000

200000

150000

100000

50000

250000

200000

150000

100000

50000

Pressure (psi)Pressure (psi)

Pressure (psi)

Roller crown

radius = 838 mm

Roller crown

radius = 1174 mm

Roller crown

radius = 1956 mm

Roller crown

radius = 838 mm

Roller crown

radius = 1174 mm

Roller crown

radius = 1956 mm

Roller crown

radius = 838 mm

Roller crown

radius = 1174 mm

Roller crown

radius = 1956 mm

Roller crown

radius = 838 mm

Roller crown

radius = 1174 mm

Roller crown

radius = 1956 mm

2.8 Cam stress sensitivity to the effect of roller crown radius (CMP: compression stroke; POW: power stroke; EXH:

exhaust stroke; INT: intake stroke).

Diesel-Xin-02.indd 156 5/5/11 11:44:53 AM



157Durability and reliability in diesel engine system design

was provided by Krepulat et al. (2002). The FEA model can handle more

complex calculations that cannot be conducted using the Hertzian equation

due to restricted assumptions in the Hertzian equation. Such restrictions

may include the edge loading effect, nonlinear and plastic material behavior,

complex geometric form of contact bodies such as the logarithmic roller

prole (Fujiwara and Kawase, 2007), or measured irregular roller crown

proles. Krepulat et al. (2002) also pointed out the pitfalls of incorrectly

applying the Hertzian equation of a cylindrical roller to the case of a crowned

roller or crowned cam lobe. They showed that the miscalculated cam stress

as such could be 30% lower than the correct value.

The skew (tilting) motion of the roller affects the cam elastic deformation

and stress. The dynamic skew motion as a function of cam angle was

investigated by Ito (2006). He found that the skew motion was affected by

the cam prole, the roller prole, the follower–guide misalignment, and the

moment of inertia of the follower. Moreover, roller slip may cause wear

and durability problems. Roller skew motion and roller slip are two major

research topics related to roller follower simulation. Roller slip friction is

detailed in Chapter 10.

2.9.5 Engine piston and crankshaft durability

The piston is probably the engine component whose structural durability has

been the most extensively investigated. The thermo-mechanical durability of

the piston is affected mainly by peak cylinder pressure, thermal loading and

piston cooling. The heat transfer rate and piston temperature increase with

engine speed and torque. The maximum piston temperature usually occurs

at rated power. The effects of oil jet cooling and engine operating condition

on piston temperature distribution were investigated by Roehrle (1978)

and Thiel et al. (2007). The temperature distribution of piston and piston

rings without forced cooling for the under-crown surface of the piston was

experimentally investigated by Furuhama and Suzuki (1979). The calculation

of piston temperature distribution was conducted theoretically by Woschni

(1979) and numerically by Wu and Chiu (1986). For more information on

diesel engine piston structural calculation, design, fatigue life, material, and

durability analysis, the reader is referred to the following: Makartchouk

(2002), Munro (1999), Spengler and Young (1986), Myers (1990), Keribar

et al. (1990), Afonso et al. (1991), Myers and Chi (1991), Castleman (1993),

Vertin et al. (1993), Barnes and Lades (2002), Reichstein et al. (2007) and

Cha et al. (2009).

Crankshaft durability investigation is presented in Law (1984), Henry et al.

(1992), Park et al. (2001), Zorou and Fatemi (2005), Williams and Fatemi

(2007), Montazersadgh and Fatemi (2007, 2008), Choi and Pan (2009), and

Çevik et al. (2009).

Diesel-Xin-02.indd 157 5/5/11 11:44:53 AM



158 Diesel engine system design

2.9.6 Turbocharger durability

Turbocharger design is handled primarily by specialized suppliers.

Turbocharger durability issues in diesel engine system design are related

mainly to LCF in the compressor wheel and HCF in the turbine wheel.

Compressor durability is related to the wheel and housing temperatures and

rotational shaft speed. The wheel and housing temperatures are dominated

by the compressor outlet air temperature. The LCF of the compressor

occurs due to alternating stresses caused by the variation in wheel speed

when engine speed/load changes. The HCF in the turbine is related to the

blade vibration and sometimes resonance at the natural frequencies of the

blades. In conditions of very high turbine inlet pressures, strong shockwave

pressure pulses are generated at the turbine nozzles or the throat. The gas

pulses impinge on the downstream turbine wheel to create excitations that

lead to the HCF failure of the wheel. Exhaust manifold gas temperature is

usually not a major concern for turbine durability because diesel engines

operate with high air–fuel ratio and hence the turbine is exposed to cooler

exhaust gas temperatures compared to gasoline engines.

Compared to the wastegated turbine, VGT offers the following benets:

reduced engine pumping loss and fuel consumption, exibility of regulating

EGR rate and air–fuel ratio, faster transient response, and enhanced engine

braking. VGT has been gaining in popularity for all different sizes of diesel

engines (e.g., from passenger cars to Class-8 trucks) and for different

applications (e.g., from on-road to off-road such as agricultural and construction

equipment and marine engines). However, VGT generally tends to have

more durability issues than xed geometry or wastegated turbines due to

more moving elements used to control the turbine area. The durability issues

of VGT usually include the following: thermal distortion, interference or

seizure, thermo-mechanical fatigue, creep, wear, and corrosion. Lighter, less

costly, and more thermally durable VGT is the direction for future design

improvement.

Turbocharger durability discussions were extensively covered by Watson

and Janota (1982), Japikse and Baines (1997), Baines (2005), Japikse

(1996) and Moustapha et al. (2003). The compressor LCF life and the

methodology of incorporating lifetime calculation in turbocharger matching

were discussed by Engels (2002), Ryder et al. (2002) and Christmann

et al. (2010). The analysis methodology of thermo-mechanical fatigue

life prediction for turbine housing was provided by Ahdad and Soare

(2002), Längler et al. (2010), and Bist et al. (2010). Thermo-mechanical

analysis of the turbine wheel was presented by Heuer et al. (2006). The

turbine HCF due to wheel blade vibration was analyzed by Kitson et al.

(2006), Chen (2006) and Kulkarni et al. (2010). The reliability data of

the turbochargers used in marine engines were reviewed by Banisoleiman

Diesel-Xin-02.indd 158 5/5/11 11:44:54 AM



159Durability and reliability in diesel engine system design

and Rattenbury (2006). VGT reliability was discussed by Furukawa et al.

(1993).

2.9.7 Durability of diesel oxidation catalyst and

particulate ﬁlter

Modern diesel engines use aftertreatment devices (e.g., DOC, DPF) to

reduce tailpipe emissions. Aftertreatment durability issues include thermal,

mechanical, and chemical failures. Thermal degradation includes sintering,

alloying, and oxidation. Mechanical failure includes physical breakage,

thermal shock, blockage of pore structure, and fouling. Chemical degradation

includes catalyst poisoning, inhibition, and life cycle stability. The durability

of diesel oxidation catalyst was investigated by Takahashi et al. (1995) and

Shinozaki et al. (1998).

Many durability issues of the DPF are related directly to its regeneration

performance. Regeneration increases the thermo-mechanical load and the

durability risk on both the base engine and the DPF itself. DPF durability

issues include mainly the following:

1. Substrate and washcoat failures caused by excessive thermal stress during

DPF runaway regeneration or uncontrolled regeneration.

2. Catalyzed DPF performance degradation caused by catalyst poisoning

(e.g., by the sulfur in the diesel fuel or lube oil).

3. Substrate or matting material failure due to excessive vibration.

4. DPF thermal aging.

5. Structural failures of the DPF assembly.

Ambient temperature and in-use duty cycle are important noise factors for

DPF durability, for example, the severe use duty cycles in cold climate.

Zhan et al. (2006) presented methodologies to control DPF regeneration to

preserve durability life. Zhan et al. (2007) provided a comprehensive summary

of the durability failure modes and durability validation testing methods for

DPF. They introduced various durability test procedures for DPF substrate

and the entire DPF assembly, including on-engine accelerated DPF thermal

aging test cycle, accelerated ash accumulation test, hot and cold vibration

tests, stepped vibration test, water quench test, thermal prole test, and

on-vehicle DPF durability test. Stroia et al. (2008) discussed durability test

procedures for a diesel aftertreatment system consisting of DOC, LNT, and

CDPF. DPF durability analysis on stress and fatigue life was conducted by

Gulati et al. (1992), Umehara and Nakasuji (1993) and Kuki et al. (2004).

Other DPF durability investigations were carried out by Locker et al. (2002),

Fayard et al. (2005) and Sappok et al. (2009).

Diesel-Xin-02.indd 159 5/5/11 11:44:54 AM



160 Diesel engine system design

2.10 Heavy-duty diesel engine cylinder liner

cavitation

2.10.1 Cylinder liner cavitation failure

The uid cavitation erosion in wet cylinder liners is another chronic failure

mode for heavy-duty diesel engines. It causes pitting and perforation on

the liner at the coolant side after several hundred hours of operation via

the formation and violent collapse of coolant vapor bubbles. The resulting

cavity holes on the liner cause the engine coolant to mix with the lubricant

oil hence the engine is severely damaged. The loss of material due to liner

cavitation may occur faster than the normal component wear so that the

cavitation can be a limiting factor of engine life and reliability.

Liner transverse vibration induces coolant pressure uctuations in the

form of expansion and compression waves. When the local dynamic coolant

pressure during the expansion phase of the wave becomes very low, the

coolant vapor pressure level may be reached and then vapor bubbles start

to form and grow. Note that this condition may happen even if the local

liner vibration velocity is zero. In the following wave compression phase

the pressure wave causes the bubbles to become unstable. Then the bubbles

collapse or implode and release a great amount of energy via micro jets. If

the bubble collapses near the liner surface, the intense micro jet can impinge

the liner wall at a high speed with a high localized shock pressure wave.

The sudden collapse of the bubbles produces intense and impulsive load on

the liner in a cyclic nature. When the pressure of the micro jet exceeds the

material strength, damage accumulates quickly and nally results in cavitation

failure.

2.10.2 Impact factors and design solutions for liner

cavitation

Liner cavitation was reviewed by Hercamp (1993) and Demarchi and Windlin

(1995). Zhou and Hammitt (1990) summarized the possible mechanisms of

liner cavitation. They pointed out that cavitation erosion is different from

corrosion since the metal is removed as much larger particles in cavitation

than in the case of corrosion. Liner cavitation mechanisms in diesel engines

were also explained by Hosny et al. (1996) with a six-step process: (1) engine

loads; (2) piston dynamics; (3) liner vibration; (4) pressure uctuation in

the coolant jacket; (5) bubble formation and collapse; and (6) liner material

surface damage.

The primary inuential factors for liner cavitation are summarized as

follows. The cavitation is caused by liner vibration, which is excited mainly

by piston slap. As a dynamic response to piston slap, the transverse vibration

of the liner is excited and this results in dynamic cycles of compression

Diesel-Xin-02.indd 160 5/5/11 11:44:54 AM



161Durability and reliability in diesel engine system design

and expansion waves in the coolant. Yonezawa and Kanda (1985) studied

piston slap and liner vibration. Demarchi and Windlin (1995) obtained

good correlations on the causal relationship between piston slap and liner

cavitation by testing large piston clearances (oversize bore) to conrm the

liner cavitation in an accelerated manner. Bederaux-Cayne (1996) conrmed

that piston pin offset has a large impact on coolant pressure uctuation and

hence liner cavitation. Minimizing piston slap kinetic energy through better

designs of piston mass, pin offset, clearance, and skirt stiffness may alleviate

liner cavitation. Piston slap is detailed in Chapter 11.

Liner cavitation becomes more severe at higher engine speed and load

or higher power density. Hosny et al. (1996) concluded that cavitation was

severe enough to cause damage only at high-speed and high-load conditions.

The reason was partly due to the increased peak cylinder pressure and the

associated increased piston slap intensity at those conditions.

Liner vibration is affected by its natural frequency and stiffness. Increasing

liner wall thickness and liner stiffness (e.g., via better designs of liner support,

sealing rings and dampers) can reduce the transient vibrational surface

velocities and the cavitation of the liner. However, excessively thick liner

wall affects its heat transfer and thermo-mechanical durability adversely.

Increasing water jacket thickness (width) can reduce coolant pressure wave

uctuations by changing the wave dynamics so that the cavitation can be

suppressed. Liner surface material property at the coolant side is also important

in terms of hardness, tensile strength, cavitation resistance coating, thermal

conductivity, and corrosion resistance. The liner surface needs to be smooth

in order to avoid the rough spot that can nucleate the bubble formation.

Coolant pressure has a direct impact on cavitation. A higher mean pressure

makes the uctuating dynamic pressure more difcult to reach down to the

coolant vapor pressure level to form bubbles. The vapor pressure of the coolant

is strongly affected by coolant temperature and properties such as coolant

composition, the acoustic velocity of the uid, coolant additives, and so on.

Coolant composition affects the surface tension and coolant viscosity, which

in turn affect the bubble growth size and the collapse process. For example,

water can aggravate the cavitation compared to water–glycol mixture due

to the larger bubble size and more violent bubble collapse of water. It was

found that coolant ow rate and velocity did not have a signicant impact

on cavitation because the dynamic coolant pressure drop created solely by

coolant ow velocity is negligible compared to the pressure uctuations

induced by the liner vibration (Hosny et al., 1996).

2.10.3 Experimental investigation on liner cavitation

Liner cavitation can be quantied and detected by nding out mass loss in

laboratory vibration bench test, tracing liner materials loss in the coolant with

Diesel-Xin-02.indd 161 5/5/11 11:44:54 AM



162 Diesel engine system design

radioactive method, or measuring liner outer surface vibration and coolant

uctuating pressure. In addition to the above macro processes related to the

vibration and pressure signals, new detection methods were developed with

micro processes that are related to bubble dynamics and collapse in order

to detect the liner cavitation intensity. Cavitation intensity or severity was

characterized by measuring the high frequency acoustic implosion signals

of coolant vapor bubbles on the liner surface (Hosny, 1996; Hosny et al.,

1996). The uid cavitation intensity measured by such a method was found to

increase as a function of the liner vibration level. It should be noted that the

measurement of cavitation noise can be problematic due to the difculty of

isolating the uid cavitation noise from other noises such as the combustion

noise in the same frequency range.

2.10.4 Analytical prediction of liner cavitation

Like other chronic durability problems such as fatigue, wear, and creep,

cylinder liner cavitation in heavy-duty diesel engines is an important issue

to address in the early stage of engine design. To reduce the time and cost

of the endurance test to reveal the cavitation problem, it is desirable to

develop advanced simulation tools to quantify the parametric design effects

and predict cavitation severity and damage. Liner cavitation is governed

mainly by piston slap. The simulation methodology of liner cavitation

may largely employ the analysis techniques that are used in the area of

piston–liner NVH, specically in the part of piston-assembly dynamics and

liner vibration response (detailed in Chapter 11 on piston slap noise). The

unique characteristics of liner cavitation analysis are the coolant dynamic

pressure uctuations and bubble dynamics.

In diesel engine system design, the piston dynamics data including piston

secondary motions can be easily generated. Such data, based on certain piston

clearance assumptions, reect the system loading acting on the piston and

the liner. These data include the effects of engine speed, load, and cylinder

pressure. They can be shared in the simulation models for both NVH and

liner cavitation in order to address these performance and durability problems

together in the early stage of the design.

The simulation methodology consists of the following ve steps for liner

cavitation modeling:

1. Piston-assembly dynamics modeling to predict the source of excitation

and the impact kinetic energy, and generate a piston impact forcing

function vs. time.

2. FEA liner modal analysis or the analysis with a simplied liner dynamics

model to predict the vibration velocity of the liner outer surface in a

transient and spatial distribution.

3. Coolant ow modeling to predict the dynamic coolant pressure wave

Diesel-Xin-02.indd 162 5/5/11 11:44:54 AM



163Durability and reliability in diesel engine system design

propagation and uctuation in the coolant jacket, and predict the onset

location of cavitation.

4. Bubble dynamics modeling to predict the bubble growth, collapse and

the micro jet behavior of liner wall impingement.

5. Liner lifetime prediction for pitting and perforation failures.

The simulation involves engine operating conditions, piston and liner design

parameters, coolant jacket design parameters, and coolant properties. Step

4 mentioned above is usually difcult to achieve due to the complexity of

bubble dynamics. Steps 2 and 3 can be coupled to account for the coolant–liner

interaction effect (i.e., the liner damping effect due to the coolant) in order

to make the model more accurate in liner transient response. Step 3 can

sometimes be omitted for simplicity and a pre-determined velocity threshold

of the liner wall can be used instead. Above the critical threshold, cavitation

can be assumed to occur. The threshold is dependent on coolant properties.

Obtaining high critical threshold of the liner surface velocity and reducing

the actual transverse vibration velocities of the liner are design goals in

order to control the liner cavitation problem. A cavitation safety factor has

been commonly used as an output in Step 2. The safety factor is dened as

the ratio of the critical threshold velocity to the calculated cycle maximum

vibration velocity of the liner at each spatial node on the liner.

Katragadda and Bata (1994) investigated the bubble formation and collapse

for liner cavitation. Their study covered the theoretical uid mechanics analysis

of the cavitation phenomena in uid ow and bubble dynamics modeling.

Lowe (1990) proposed an analytical simulation of liner cavitation to predict

the transient response of the liner. The critical cavitation velocity was reported.

Hosny and Young (1993) developed an analytical predictive model of liner

cavitation that can be used in the early design stage. The input of the model

was the impact force from piston dynamics, and the output of the model was

liner vibration and the prediction on pitting. The cavitation intensity map in

the engine speed–load domain as proposed by Hosny et al. (1996) is a good

way to convey the boundary limit curve for the region of cavitation in the

engine operating domain. Green and Engelstad (1993) simulated coolant

pressure wave propagations and attempted to predict the bubble formation

location and the cavitation location. They emphasized the importance of

including the coolant ow and pressure wave propagations in the liner FEA

vibration model in order to account for the coolant–liner interaction.

2.11 Diesel engine wear

2.11.1 Fundamentals of wear

Wear is a progressive loss of material over time. Although engine wear often

appears to be a local component-level issue, it needs to be considered in the

Diesel-Xin-02.indd 163 5/5/11 11:44:54 AM



164 Diesel engine system design

early stage of engine system design because the wear life of the component

is closely related to system loading. For example, peak cylinder pressure

affects the wear life of the valve seat, the piston assembly and the engine

bearings. A system engineer needs to consider all of the related parameters

that affect engine wear and be pro cient with the modeling techniques for

predicting the wear life.

The wear of engine components usually undergoes three stages: (1) a

running-in period where the rate of change is high; (2) a stabilized normal

running period where a steady rate of wear is maintained; and (3) a wear-out

failure period where a high rate of wear leads to rapid failures. Generally,

there are four common types of wear in the engine: adhesive, abrasive,

corrosive, and impact wear. Adhesive wear is the most common form of

wear and is sometimes called sliding wear. It occurs at the interfaces with

high contact stresses and signi cant sliding. Oil  lm thickness is a good

indicator of adhesive wear. Adhesive wear occurs in mixed and boundary

lubrication regimes where metal surface asperity contact occurs. (Lubrication

and friction are elaborated in Chapter 10.) Abrasive wear occurs when a

hard rough surface or hard solid particles slide on a softer surface. It may

occur in hydrodynamic lubrication due to the particles in the engine oil. The

number of particles is dependent mainly on the engine soot emission level

and oil aging. In this case, the abrasive wear is three-body wear where the

soot particles are free to roll and slide down a surface to scratch and score.

Corrosive wear occurs when the component is exposed to the corrosive

combustion product such as sulfuric acid (e.g., in valve face guttering).

Impact wear occurs at the impact surfaces such as the valve seat. Detailed

mechanisms of wear were explained by Kato (2002). A study on engine

wear rate was provided by Macian et al. (2003) based on oil analysis.

According to the Archard equation (1953), the adhesive wear volume is

equal to the product of the contact load at the sliding surfaces, the sliding

distance, an adhesive wear coef cient, and the reciprocal of combined

hardness of both surfaces, i.e.,

ws,ws

2.10

where V

w,sl

is the wear volume, f

is the wear coef cient, F

is the contact

force, l

w,sl

is the sliding distance, and h

is the penetration hardness of the

softer surface of the two contact surfaces (N/m

). The Archard equation is

commonly used to calculate wear volume in sliding contact.

One key element in wear calculation is to determine the wear coef cient,

which lumps many complex effects (e.g., material, metallurgical, lubrication,

anti-wear additive) into one coef cient. The adhesive wear coef cients

of typical materials were provided by Booser (1997). The value of wear

Diesel-Xin-02.indd 164 5/5/11 11:44:54 AM



165Durability and reliability in diesel engine system design

coef cient is dependent on the materials of the two surfaces, their lubrication

regime and other environmental factors. In hydrodynamic lubrication there is

no or very little wear, therefore, f

= 0. The value of the wear coef cient in

mixed lubrication stays between the values of hydrodynamic and boundary

lubrication regimes. For simplicity it can be assumed that the wear coef cient

approximately follows a linear relationship as a function of the oil  lm

thickness-to-roughness ratio (i.e., the Lambda ratio).

Equation 2.10 was originally developed for adhesive wear. It has also

been used successfully to model both abrasive wear (Suh and Sridharan,

1975) and fretting wear (Stower and Rabinowicz, 1973), as pointed out

by Lewis and Dwyer-Joyce (2002). Pint and Schock (2000) proposed a

generalized total wear coef cient in a similar model form, which includes

the contributions of all three types of wear for piston rings (i.e., adhesive,

abrasive, and corrosive wear).

Impact wear can be modeled as follows (Fricke and Allen, 1993; also

used by Lewis and Dwyer-Joyce, 2002), for an example of valve seating

impact wear:

mw,,mw

ki,ki

Vf =Vf

nE nE

Vf Vf

nE nE

2.11

where V

w,im

is the impact wear volume, n

is the number of loading cycles,

is the impact energy per cycle, f

w,im

and C are empirically determined

wear constants. When equation 2.11 is used for valve seating impact, m

can be taken as the equivalent valvetrain mass and v

is the valve seating

velocity.

2.11.2 Engine bearing wear

Diesel engine bearing failure analysis was reported by McGeehan and

Ryason (1999). Traditional assessment of bearing durability focused on the

prediction of minimum oil  lm thickness and maximum oil  lm pressure.

Xu et al. (1999) and Ushijima et al. (1999) developed a wear model and a

fatigue parameter to characterize engine bearing failure mechanisms. In their

model, the wear amount was assumed proportional to the intensity of the metal

asperity contact and the frequency of sliding motion. Their fatigue parameter

was proposed to be proportional to a wear parameter (which corresponds

to the density of surface-originating cracks), peak oil  lm pressure, peak

frictional force, and bearing surface area. The fatigue parameter is inversely

proportional to an oil  lm thickness limit, a mean sliding velocity, the time

of one loading cycle, material strength, the rotational speed of the journal,

and the mean frictional torque.

Diesel-Xin-02.indd 165 5/5/11 11:44:55 AM

