progress towards understanding combustion ine˜ciencies… · • contrast to emissions sources...

Progress towards understanding combustionine�ciencies:

In-cylinder sources of CO and UHC

Paul MilesSandia National Laboratories

Acknowledgements: Isaac Ekoto (SNL), UW-ERC, Lund UniversitySponsors: US Department of Energy, EERE-OVT Gurpreet Singh, Program Manager General Motors Corporation Russell Durrett, Project Technical Lead

University of Wisconsin Engine Research Center Symposium“Reducing fuel consumption: solutions and prospects”

June 10-11, Madison WI

M OF

E YDPAR T

ENTNE ER G

E

A

UNI

DOF

MR

C

T

S T S

E

TA

I

A

E

Factors a�ecting brake fuel e�ciencyFactors a�ecting brake fuel e�ciency

ηb = ηm ηt ηc

• Pumping work• Friction - Downspeeding - Downsizing - Materials - Lubricants - Additives

• Accessories


ηb = ηm ηt ηc

In�uenced by:• Compression ratio (Expansion ratio)• Specific heat ratio• Combustion phasing and rate of heat release• Heat transfer losses

Net chemical energy released

p dV∫

-

-?

+


ηb = ηm ηt ηc

• Low ηc manifested by “fuel” in the exhaust gases (CO, UHC, H2 ... )• Normally high > 0.98 diesels > 0.95 spark

• Can drop drastically for low- temperature diesel or HCCI

Net chemical energy released

mf QLHV

OverviewOverview

• Description of UHC and CO sources from early-injection (PCI-like) low temperature combustion systems

- Understanding developed from homogeneous reactor simulations

- Experimental images of in-cylinder UHC and CO

• Description of UHC and CO sources from late-injection (MK-like) low temperature combustion systems

- Understanding developed from homogeneous reactor simulations

- Experimental images of in-cylinder UHC and CO

• Contrast to emissions sources from conventional (Euro 5) calibrations

• Implications for combustion chamber design and operating condi-tion selection

Emissions behavior of early-injection, PCI-likecombustionEmissions behavior of early-injection, PCI-likecombustion

Pint=1.5 bar, 3 bar load, 1500 rpm

NO

x [g

/kg-

fuel

], S

oot x

10 [F

SN]

CO

, UH

C [%

of f

uel e

nerg

y]

MPR

R [bar/°C

A]

CO

UHC

Soot

NOx

dPdθ

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0

5

10

15

20

ExcessiveCO

ExcessiveNOx

8 10 12 14 16 18

0.20.250.30.350.40.5

O2 Concentration [%]

Global equivalence ratio

• Within a range of intake charge O2 concentrations, early injec-tion low-temperature combus-tion schemes can provide low soot, low NOx, and acceptable noise

• At high dilution levels, CO and UHC emissions rise and com-bustion efficiency deteriorates

• We will examine UHC and CO sources for a 10% oxygen con-centration in the intake charge, which provides a combustion efficiency of ≈ 98%

Impact of equivalence ratio on early-injection,PCI-like combustion Impact of equivalence ratio on early-injection,PCI-like combustion

0

0.005

0.010

0.015

0.020

CO COUHC UHC

CO

[Mol

e Fr

actio

n]

Equivalence Ratio

UH

C [M

ole Fraction]

10% O2

IncreasingO2

15% O2

0

20

40

60

80

-40 -20 0 20 40 60Crank Angle [ ° ]

Cyl

inde

r Pre

ssur

e [ b

ar ]

Homogeneous reactor simulations, following real-istic pressure and temperature histories, help us understand the impact of equivalence ratio on CO and UHC emissions.

3 bar IMEP1500 RPM

10% O2

0

0.02

0.04

0.06

0.08

0.1

0 0.5 1 1.5 2

Increasing O2


Crank Angle [ ° ]

Hea

t Rel

ease

Rat

e [ J

/ cm

3 -s

] Other than ignition timing and temperature re-lated di�erences in peak rate, the heat release rate characteristics of rich mixtures are similar to those of clean-burning mixtures

Rich mixture emissions are mixing-limited

0

0.005

0.010

0.015

0.020

CO COUHC UHC

CO

[Mol

e Fr

actio

n]

Equivalence Ratio

UH

C [M

ole Fraction]

10% O2 15% O2

0

0.02

0.04

0.06

0.08

0.1

0 0.5 1 1.5 2

0

2 104

4 104

6 104

8 104

1 105

1.2 105

-15 -10 -5 0 5

φ = 1.4

φ = 0.8


Crank Angle [ ° ]

For lean mixtures, the heat release is retarded and slowed compared with clean-burning mixtures

Lean mixture emissions are kinetically-limited

CO oxidation is impeded �rst, then UHC for very lean mixtures

φ = 0.8

φ = 0.6

φ = 0.4Hea

t Rel

ease

Rat

e [ J

/ cm

3 -s

]

0

0.005

0.010

0.015

0.020

CO COUHC UHC

CO

[Mol

e Fr

actio

n]

Equivalence Ratio

UH

C [M

ole Fraction]

10% O2 15% O2

0

0.02

0.04

0.06

0.08

0.1

0 0.5 1 1.5 2

0

2 104

4 104

6 104

8 104

1 105

1.2 105

-5 0 5 10 15

For PCI-like combustion, the CO and UHC are notco-located within the clearance volume For PCI-like combustion, the CO and UHC are notco-located within the clearance volume Data at 30°CA

0 10 20 30 40r (mm)

Mole Fraction

5.0 x 10-3

0.0

white isotherm = 1200K

Modeled UHC(CO is similar)

Measured CO and UHC

• CO is broadly distributed throughout the squish volume, with greater concentrations near the piston

• Unburned fuel is observed near the injector, and to a lesser extent near the bore wall

• Partially-burned fuel is generally found between fuel and CO—showing the direction of reaction progress

• Simulations (and exp. parameter sweeps) show that squish volume mixture is over-all fuel-lean

Reaction progress

0

0.5

1

0

0.5

1

0

0.5

1

0 10 20 30 40r (mm)

0 10 20 30 40r (mm)

0 10 20 30 40r (mm)

CO C2 (partially-burned)

PAH (fuel)

By 50°CA, it is clear that the near-injector andsquish regions dominate CO and UHC emissions By 50°CA, it is clear that the near-injector andsquish regions dominate CO and UHC emissions

Data at 50°CA

Mole Fraction

CO

5.0 x 10-3

0.0


Measured CO and UHC

• At 50° CA, 83% of the cylinder volume is within the clearance volume

• Distributions of CO and UHC components are similar to those seen at 30°CA—just stretched axially

The flow exiting the bowl is clean

Mean �ow structures and squish volume emissions are well predicted, even if CO and UHC emissions are not

C2 (partially-burned)

PAH (fuel)

0

0.5

1

0

0.5

1

0

0.5

1

Modeled UHC (CO is similar)

Mole Fraction

5.0 x 10-3

0.0

Single-cycle images reveal UHC within the bowl and better characterize squish volume UHC Single-cycle images reveal UHC within the bowl and better characterize squish volume UHC

UHC within the bowl: - Observed infrequently - More frequent with retarded SOI - Absent with advanced SOI - More pronounced at higher load

Simulations capture the general fuel distribution,but mixing is underpredicted

30°CA

Within the squish volume, UHC �uores-cence is observed from:

• Diffuse (fuel lean) bulk gas UHC

• Apparent liquid films on the piston top

• Crevice flows

The near-injector UHC is made up of both liquidand gaseous fuelThe near-injector UHC is made up of both liquidand gaseous fuel

Di�use �uorescence also seen near the cylinder center is characteristic of gaseous UHC

Intense, spherical regions show liquid fuel near the injector...

...that are also seen in elastic scatter images

End-of-injection over-leaning is one possible source of this di�use �uorescence

0 mm 5-5

Dis

tanc

e fr

omIn

ject

or [m

m]

Equivalence Ratio Contours from fuel tracer LIF

Musculus, et al. SAE 2007-01-0907

30°CA

30°CA

Emissions seen in both the squish volume andnear-injector regions drop signi�cantly with loadEmissions seen in both the squish volume andnear-injector regions drop signi�cantly with load

0

0.2

0.4

0

0.25

0.5

0

0.5

1

0

0.4

0.8

0

0.75

1.5

0

0.75

1.5

C2(partially burned)

CO

1.5 barIMEP

3.0 barIMEP

4.5 barIMEP

CO

[ppm

]

UH

C [p

pm]

Bas

elin

e

Hig

h Lo

ad

Low

Loa

d

Bas

elin

e

Hig

h Lo

ad

Low

Loa

d

0

1000

2000

3000

4000

5000

6000

7000

8000

0

200

400

600

800

1000

1200

1400

1600

• Both near-injector and squish volume emissions decrease signi�-cantly with load

• Central cylinder CO at higher loads is consistent with better oxidation of near-injector UHC

• Optical data correlate closely with engine-out emissions

High Load (30°CA)

0 10 20 30 40r (mm)

x10

Higher resolution, cycle-averaged images showimpact of load on squish volume UHC sourcesHigher resolution, cycle-averaged images showimpact of load on squish volume UHC sources

Baseline (30°CA)

0 10 20 30 40r (mm)

x40

Image GainLow Load (30°CA)

0 10 20 30 40r (mm)

x20

• As load increases: 1) bulk gas UHC from the near-injector and central cylinder region decreases 2) the diffuse, bulk gas signal within the squish volume decreases 3) the UHC associated with crevice flows and piston top films increases The 1st and 2nd factors dominate

• Note also the predicted fuel-lean UHC in the bowl periphery is clearly observed, and vanishes at the highest load

30°

CA

Development of counter-rotating

vortex pair

Squish volume emissions are minimized bysmall squish heights and low spray targeting Squish volume emissions are minimized bysmall squish heights and low spray targeting

22

24

26

28

30

32

0.6 0.8 1 1.2 1.4 1.6Squish Height [mm]

CO

[g/k

W-h

r]

SPT = 0.0

SPT = -0.5

SPT = 0.5

CO

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

2.8

0.6 0.8 1 1.2 1.4 1.6Squish Height [mm]

SPT = 0.0

SPT = -0.5

SPT = 0.5

UHC

UH

C [g

/kW

-hr]

-SPT

+SPT

Statistically signi�cant set of experiments, incorporating repeated engine builds at each condition

Conventional diesel design guidelines still apply when PCI-like combustion is em-ployed in typical diesel hardware geom-etries:

A large k-factor and spray targeting within the bowl is desirable

Emissions behavior of late-injection, MK-likecombustionEmissions behavior of late-injection, MK-likecombustion

Pint=1.5 bar, 3 bar load, 1500 rpm• Like early injection schemes,

late-injection LTC can provide low soot, low NOx, and accept-able noise

• At retarded timings, CO and UHC emissions rise and com-bustion efficiency deteriorates, just as noise levels become ac-ceptable

• We will examine UHC and CO sources for a late injection case which provides a combustion efficiency of ≈ 97%

0.0

0.5

1.0

1.5

2.0

CO

, UH

C [%

of f

uel e

nerg

y]N

Ox

[g/k

g-fu

el],

Soo

t x10

[FSN

]

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0

MPR

R [bar/°C

A]

CO

UHC Soot

NOx

dPdθ

ExcessiveNOx

-25 -20 -15 -10 -5 0SOI

15% O2

COCO

UHCUHC

Equivalence RatioEquivalence Ratio

CO

[Mol

e Fr

actio

n]C

O [M

ole

Frac

tion]

UH

C [M

ole Fraction]U

HC

[Mole Fraction]

For rich mixtures, the emissions characteristics do not change as injection is retardedFor rich mixtures, the emissions characteristics do not change as injection is retarded

Impact of equivalence ratio on late-injection,MK-like combustion Impact of equivalence ratio on late-injection,MK-like combustion

0

10

20

30

40

50

60

70

-40 -20 0 20 40 60

SOIc = -2.3°

SOIc = -2.3°

SOIc = -0.5°

SOIc= -0.5°

Crank Angle [ ° ]

Cyl

inde

r Pre

ssur

e [ b

ar ]

3 bar IMEP1500 RPM

15% O2

0

0.02

0.04

0.06

0.08

0.1

0

0.005

0.010

0.015

0.020

0.025

0.030

0 0.5 1 1.5 2

CO

UHC

Equivalence Ratio

CO

[Mol

e Fr

actio

n]

Impact of equivalence ratio on late-injection,MK-like combustion - lean mixtures Impact of equivalence ratio on late-injection,MK-like combustion - lean mixtures

UH

C [M

ole Fraction]

0.030

0

5000

1 104

1.5 104

2 104

2.5 104

3 104

10 15 20 25 30Crank Angle

0

0.02

0.04

0.06

0.08

0.1

0

0.005

0.010

0.015

0.020

0.025

0 0.5 1 1.5 2

φ = 0.7

φ = 0.65φ = 0.6

Hea

t Rel

ease

Rat

e [ J

/ cm

3 -s

]

For lean mixtures, higher equivalence ratios are required to complete oxidation at retarded timing

Oxidation is limited by the failure to transition into strong high temperature combustion – leading to similar limiting equivalence ratios for CO & UHC

With late injection, there is evidence of CO, PAH,and C2 exiting the bowl at 30°CA (as predicted) With late injection, there is evidence of CO, PAH,and C2 exiting the bowl at 30°CA (as predicted) SOIC = -2.3° (Data at 30°CA)

0 10 20 30 40r (mm)

Mole Fraction

5.0 x 10-3

0.0



PAH (fuel)

Modeled UHC(CO is similar)

Measured CO and UHC

• The (fuel-lean) squish volume is again a dominant source of UHC and CO, despite spray targeting deep within the bowl

• Fuel is again observed near the injector, with small amounts of CO

• Experiments appear to support model predictions of fuel-rich UHC leaving the bowl

0 10 20 30 40r (mm)

0 10 20 30 40r (mm)

0 10 20 30 40r (mm)

0

0.25

0.5

0

10

20

0

0.35

0.7

By 50 °CA, however, the rich-mixture sourcefrom the bowl has vanished By 50 °CA, however, the rich-mixture sourcefrom the bowl has vanished

Modeled UHC (CO is similar)

At this crank angle, the �ow exiting the bowl, as was seen with early injection timing, is very clean

0

0.25

0.5

0

5

10

0

0.35

0.7

SOIC = -2.3° (Data at 50°CA)


PAH (fuel)

Measured CO and UHC

Mole Fraction

5.0 x 10-3

0.0


With more retarded SOI, the squish volume UHCand CO emissions completely dominate With more retarded SOI, the squish volume UHCand CO emissions completely dominate

0

0.25

0.5

0

5

10

0

0.35

0.7

CO


PAH (fuel)

CO


PAH (fuel)

0

0.25

0.5

0

5

10

0

0.35

0.7

Baseline late-injection1.8° additional retard

0

500

CO

[ppm

]

UH

C [p

pm]

1000

1500

2000

2500

3000

3500

Bas

elin

e

Ret

arde

d

Bas

elin

e

Ret

arde

d

0

500

1000

1500

2000

2500

3000

50°CA

• Notable increases in squish volume concentrations are observed for both UHC and CO

• Larger increase in engine-out UHC than CO is consistent with homogeneous reactor simulations of lean mixtures

How do conventional (Euro 5) combustion UHCand CO distributions compare? How do conventional (Euro 5) combustion UHCand CO distributions compare?

35

40

45

50

55

60

65

70C

ylin

der p

ress

ure

[ bar

] H

eat R

elea

se [

J/°C

A ]

0

10

20

30

40

50

60

70

-10 0 10 20 30Crank Angle [ ° ]

Late-injection

Late-injection

Euro 5

Euro 5

InjectorEnergizingCommand }

The Euro 5 calibration features a main injection timing very similar to our late-injection, low temperature condition

Euro 5 emissions are low CO and UHC, but relatively high soot and NOx

1500 rpm3.0 bar gIMEP 15% O2

0

500

CO

[ppm

]

UH

C [p

pm]

1000

1500

2000

2500

3000

3500

Euro

5

Late

inje

ctio

n

Euro

5

Late

inje

ctio

n

0

500

1000

1500

2000

2500

3000

NO

x [p

pm]

FSN

[-]

Euro

5

Late

inje

ctio

n

Late

inje

ctio

n

0

10

20

30

40

50

60

70

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Euro

5

With conventional combustion, UHC and CO arepredominantly found in the near-injector regionWith conventional combustion, UHC and CO arepredominantly found in the near-injector region

0 10 20 30 40r (mm)

0

0.5

1

0 10 20 30 40r (mm)

0

10

20

0 10 20 30 40r (mm)

0

0.3

0.6

0

0.3

0.6

0

5

10

0

0.5

1CO C2 (partially-burned)

PAH (fuel)


PAH (fuel)

30°CA

50°CA

• Very intense “UHC” signals may have a soot component

• CO observed in the squish volume at 30°CA appears to fully oxidize by 50°CA

Summary & implications for operating conditionselection and combustion chamber designSummary & implications for operating conditionselection and combustion chamber design• Fuel dribble from the nozzle sac volume is a significant (though not a dominant)

source of UHC and CO emissions (≈ 15-30% ?)

• Significant gaseous UHC in fuel-lean mixture is seen in the near-injector region. Under some operating conditions, this may dominate engine-out UHC

• Fuel-lean mixture within the squish volume is the dominant source of CO emissions from both early- and late-injection LTC strategies at low loads. It is also a significant, and likely dominant, source of UHC

Strategies to reduce squish volume CO and UHC include:

- Minimize squish height, target spray well within the bowl (generally prevent lean mixture formation within the squish volume)

- Attack combustion noise by means other than increased dilution or SOI retard Reduce ignition delay through pilot injections? (lean mixture is the problem) Disrupt/split heat release with an appropriately timed post injection?

- Increased Tin (high pressure EGR? VVA?)

- Adopt a more open chamber, lower swirl design?

Better BSFC; with part-load LTC, no part-load soot catastrophe!

progress towards understanding combustion ine˜ciencies… · • contrast to emissions sources...

Documents