egr effect on auto ignition sandia labs

7/31/2019 Egr Effect on Auto Ignition Sandia Labs

1/21

The Effects of EGR and Its Constituents on theAutoignition of Single- and Two-Stage Fuels

John Dec, Magnus Sjberg, and Wontae HwangSandia National Laboratories

13th Diesel Engine-Efficiency and Emissions Research Conference August 13 16, 2007, Detroit, MI

Sponsors:U.S. DOE, Office of FreedomCAR & Vehicle Technologies

Program Manager: Gurpreet Singh

Saudi Aramco Oil CompanyProgram Manager: Ahmar Ghauri


2/21

Introduction

z Advanced LTC diesel and HCCI engines can provide both highefficiencies and very low emissions of NO X and PM.

z Nearly all LTC and HCCI strategies use high levels of EGR/residuals.

1. Reduce peak combustion temperatures to control NO X.2. Delay the onset of ignition.

Controlling combustion phasing.For HCCI-like diesel, it allows more time for premixing before ignition.

3. For gasoline HCCI, hot residuals are used to promote autoignition, andcombined with cooled EGR, to control combustion phasing and NO X.

z Although EGR is widely used, its thermodynamic and chemical effectson autoignition are only understood in general terms.

z Objective: Provide a fundamental understanding of how EGR andits constituents affect the autoignition of single- andtwo-stage ignition fuels.

Conduct well-controlled experiments in an HCCI engine.


3/21

HCCI Engine and Subsystems

Only premixedfueling is usedfor this study.

Cummins B

0.98 liter / cyl.CR = 14 piston

EGR Systemsz Real EGRz Simulated EGR &EGR constituents


4/21

Baseline Points40

330 335 340 345 350 355 360 365 370Crank Angle [CA]

iso-OctaneGasolinePRF80PRF60, 29% CSP

CA50 =368CA

PRF60, 29% CSPPRF80Gasolineiso-Octane

Baseline Points

E G R

C/F = 37.8 ( = 0.40)CA50 = 368CA

Single-stageIgnitionTwo-stage

Ignition (LTHR)

No EGR 29% CSP

Gasoline

TDC

z Establish baseline operating pointsfor all four fuels. CA50 = 368CA (8 aTDC)

35

30

25

20

15z Study the retarding effects of EGR 10

and its constituents. 12001100

z Iso-octane and gasoline exhibitsingle-stage autoig. at these conds.

1000

900

800z PRF80 and PRF60 have two-stage 700

autoignition with LTHR. Similar to diesel fuel (CN = 21 & 30).

180z Baseline conditions have no EGR

except for PRF60 29% CSP.

Operating ConditionsP in = 100 kPa (naturally asp.)

1200 rpm.

C/F-mass ratio = 37.8,( = 0.40 without EGR.)

160

140

120

100

80

60

40

P r e s s u r e

[ b a r ]

M a s s - A v e r a g e d

T e m p . [ K

]

R e q u

i r e

d I n t a k e T e m p . [

C ]

50 60 70 80 90 100PRF Number, Anti-knock Index


5/21

Composition of EGR (CSP)

z Consider complete combustion of iso-octane in "air".(Argon and atmospheric CO 2 lumped with N 2.)

z C 8H18 + 12.5 (O 2 + 3.773 N 2) 8 CO 2 + 9 H 2O + 47.16 N 2.

z For complete combustion with = 1, the gas composition (excl. fuel)changes from air: 21% O 2 & 79% N 2 (mole basis).

z to: 12.5% CO 2, 14.0% H 2O and 73.5% N 2 for wet exhaust (CSP).

z to: 14.5% CO 2 and 85.5% N 2 for dry exhaust (Dry CSP).

z These gas compositions will be referred to as:Complete Stoichiometric Products CSP & Dry CSP .


6/21

Thermodynamic Effects of EGR (CSP)

z EGR influences the compressed-gas temperature. Heat-capacities of CSP & individual components are different from that of air.

z Start with motored operation. Examine effects of CSP & its

components on the mass-averaged temperature near TDC.z Displace air while

maintaining P in = 100 kPa.900

890

z Trends can be explained by 880changes in specific heat (C v).

T e m

p e r a t u r e

@ 3

5 0 C A [ K ]

870

CO 2 has highest C v (mole bs.). 860 H2O high C v. 850

N2 Cv slightly less than air. 840 CSP between air and H 2O. 830

Dry CSP slightly lower C v820

than CSP.1112131415161718192021

Intake O2 Mole Fraction [%]

N2Dry CSPCSPH20CO2

Motored Data

Increasing Diluent


7/21

900

Autoignition Retard for iso-Octane

z Investigate the effect of these gases on combustion phasing.

z The autoignition retard is highly dependent on the type of the diluent.

z The retarding effect of the various gases is ordered consistently withtheir "cooling capacity" - thermodynamic effect .

z However, N 2 addition increases the compression temperature, but thephasing is retarded, so there must be a weak chemical [O2] effect.

1112131415161718192021Intake O2 Mole Fraction [%]

N2Dry CSPCSPH20CO2

Motored Data

1 0 % B u r n e d

[ C A ]

373

372

371

370

369

368

367

1919.219.419.619.82020.220.420.620.821Intake O2 Mole Fraction [%]

CO2H2OCSPDry CSPN2

Tin = 175C

T e m p e r a t u r e

@

3 5 0 C A [ K ]

890

880

870

860

366

365820

830

840

850


8/21

373

Temperature Traces for iso-Octane

1050

Air, BaselineN2CO2 CA10

370.5CA

CA10365.7CA

CO 2

N2

Air

z Retarding effect of lower [O 2]becomes more significant athigher levels of N 2 addition. Clearly, reduction of [O 2] has to

lead to misfire at some point.

z Temperature traces illustrate

1040

1030

1020

1010

1000

990980

970

960

1 0 % B u r n e d

[ C A ]

M a s s -

A v e r a g e d

T e m p . [

K ]

different mechanisms for retard. 950350 355 360 365 370 375

Crank Angle [CA]

16.51717.51818.51919.52020.521Intake O 2 Mole Fraction [%]

CO2H2OCSPDry CSPN2

iso-OctaneTin = 175C

373

372

371

370

369

368

372


CO2H2OCSPDry CSPN2

Tin = 175C

1 0 % B u r n e d

[ C A ]

371

370

369

367

366

365365

366

367

368


9/21

900

Autoignition Retard for PRF80 (2-Stage Ignition)

z For PRF80, the retarding effects are not consistent with the "coolingcapacity" of the added gases (except for CO 2).

z N2 addition (reduced [O 2]) strongly retards combustion phasing.

1 0 % B u r n

e d [ C A ]

T e m p . R

i s e

3 3 5 -

3 5 2 C A

[ K ]

147

146

145

144

143

142

z

PRF80's high sensitivity to [O 2]occurs for two reasons: First, a reduced [O 2] reduces the

LTHR.

CO2H2ON2Dry CSPCSP

Tin = 72C

PRF80, N2

2nd Order Curve Fit

890


N2Dry CSPCSPH20CO2

Motored Data

373

374

T e m p e r a t u r e @

3 5 0 C A [ K ]

880 372

371870

860 370

369

368

367

366

21 20.8 20.6 20.4 20.2 20 19.8 19.6 19.4 19.2 19

820


830

840

850


10/21

900

Autoignition Retard for PRF80 - [O 2]

z Second, a reduced [O 2] also affects hot ignition.1060104010201000

980960940920900880

T e m p e r a t u r e @

3 5 0 C A [ K ]

M a s s -

A v e r a g e d

T e m p . [

K ]

N2CO 2

[O2] = 20.7%

[O2] = 19.3%

Post-LTHRtemperatureis the same.

1 0 % B u r n

e d [ C A ]

T e m p . R

i s e

3 3 5 -

3 5 2 C A

[ K ]

147

146

145

144

143

142

CO2H2ON2Dry CSPCSP

Tin = 72C

PRF80, N2

2nd Order Curve Fit

350 355 360 365 370 375Crank Angle [CA]

890


N2Dry CSPCSPH20CO2

Motored Data

373

374

880 372

371870

860 370

369

368

367

366

21 20.8 20.6 20.4 20.2 20 19.8 19.6 19.4 19.2 19

820


830

840

850


11/21

All Constituents for PRF80 - H 2O

z Thermodynamic "cooling" should add to [O 2] effect for all constituents.z Why is the retarding effect of H 2O equal to N 2?

z H2O enhances the intermediate chemistry, so thermal run-away occurs

at lower temperature. Almost perfectly counteracts the cooling effect of H 2O!z Also explains why CSP is less retarded than dry CSP.

z H2O does not enhance the intermediate chemistry for iso-Octane.

374

373

372

371

370

369

368

367

366

1 0 % B u r n e d

[ C A ]

880

900

920

940

960

980

1000

350 355 360 365 370 375Crank Angle [CA]

M a s s -

A v e r a g e d

T e m p . [

K ]

iso-Octane, CO2iso-Octane, H2OPRF80, N2PRF80, H2O

PRF80CA10 371CA

iso-OctaneCA10 370.4CA


CO2H2ON2Dry CSPCSP

Tin = 72C


12/21

Comparing [O 2] Sensitivities

z Have now identified three mechanisms by which EGR affects phasing.1. C v effect (thermodynamic retarding).

2. O 2-concentration effect (chemical retarding).

3. H2O effect (chemical enhancing).

z Compare side by side for all fuels.

z N2 addition gives most reductionin [O 2] for least change in C v.

373

372z The two-stage ignition fuels

PRF80 and PRF60 are371

much more sensitive. 370369

z Lower [O 2] both reducesLTHR and counteracts 368

the hot ignition. 367366

z Gasoline is initially very

insensitive to [O 2].

1 0 %

B u r n e

d [ C A ]

15.51616.51717.51818.51919.52020.521Intake O 2 Mole Fraction [%]

PRF60PRF80iso-OctaneGasolineGas., 3rd order

N2

addition


13/21

Comparing Thermal Sensitivities

1. C v effect (thermodynamic retarding).


3. H 2O effect (chemical enhancing).

z CO 2 addition gives mostthermodynamic cooling, with 373least change in [O 2].

20.220.320.420.520.620.720.820.921Intake O 2 Mole Fraction [%]

Gasolineiso-OctanePRF80

PRF60

CO 2 addition372

371

1 0 % B u r n e d

[ C A ]

z The single-stage ignition fuels,iso-octane and gasoline, aremuch more sensitive to thecooling effect of CO 2.

370

369

368

367z The reason for this explained in:Proceedings of the Combustion 366Institute , Vol. 31, pp. 28952902,2007.


14/21

Comparing CSP Effects

1. C v effect (thermodynamic retarding).


3. H 2O effect (chemical enhancing).

z The overall effect of CSP results from the combination of these threemechanisms.

374z CSP gives similar amounts of 373

retard for all four fuels. 372z But underlying mechanisms 371

1 0 %

B u r n

P o

i n t [ C A ]

are different. 370369

z Trace species in real EGR 368add to these effects. 367366

19.219.419.619.82020.220.420.620.821Intake O 2 Mole Fraction [%]

iso-OctanePRF60, Shifted 6.1%PRF80Gasoline

CSP addition


15/21

Real EGR Effects

z Real EGR contains trace species 373372that also affect the autoignition. 371 Unburned fuel, partially oxidized 370369

fuel, formaldehyde, CO, and other 368

species. 367366z For iso-Octane and Gasoline, 365

372real EGR retards less than CSP. 371370

z For PRF80, real EGR has a 369retards slightly more than CSP. 368

367

366z The net effect of these species is 365

to advance the ignition for iso- 372371

octane and gasoline... 370369

z ...but to retard it for PRF80. 368367

366

1 0 % B u r n

P o

i n t [ C A ]

CSP

Real EGR

PRF80Tin = 72C

CSP

Real EGR

GasolineT in = 149C

CSP

Real EGR

iso-OctaneT in = 175C

21 20.8 20.6 20.4 20.2 20 19.8 19.6 19.4 19.2 19

Intake O 2 Mole Fraction [%]


16/21

Summary / Conclusionsz EGR is very effective for suppressing autoignition reactivity.

Can be used beneficially for controlling combustion phasing across loadand speed ranges.

Iso-Octane Gasoline PRF80 PRF60

Effect: Single-stage Two-stage with LTHR

Retarding C v effect (thermodynamic) Strong Weak

[O 2 ] effect (chemical) Weak Strong

Enhancing H 2 O effect (chemical) None to weak Strong

Trace Species Moderate enhancing Moderate retarding

z The net result of the different thermal, O 2, and H 2O sensitivities is afairly similar effect of CSP for all fuels.

z Trace species (unburned and partially oxidized fuel, and CO) influence

the effect of real EGR. Details in: SAE 2007-01-0207


17/21


18/21

900

Autoignition Retard for PRF80 - [O 2]

z Lower [O 2] also affects hot ignition, in addition to reducing the LTHR.

M a s s - A v e r a g e

d T e m p . [

K ]

N2

CO 2

[O 2] = 20.7%

[O 2] = 19.3%

1040

1000

960

920

880

840

800

1 0 % B u r n e d

[ C A ]

T e m p . R

i s e

3 3 5 -

3 5 2 C A

[ K ]

147

146

145

144

143

142

CO2H2ON2Dry CSPCSP

Tin = 72C

PRF80, N22nd Order Curve Fit

760335 340 345 350 355 360 365 370 375

Crank Angle [CA]

890


N2Dry CSPCSPH20CO2

Motored Data

373

374


@ 3

5 0 C A [ K ]

880 372

371870

860 370

369

368

367

366

21 20.8 20.6 20.4 20.2 20 19.8 19.6 19.4 19.2 19

820


830

840

850


19/21

Autoignition Retard for Gasoline

z N2 addition shows that the chemical [O 2] effect is very weak.

z Retarding effect of H 2O is weaker than expected based on its high C v.

z CSP (with water) has a slightly weaker retarding effect than dry CSP.

z Temperature traces confirm that H 2O has an enhancing effect onautoignition for gasoline.

z Effect is stronger than for iso-octane, but much weaker than for PRF80.

350 355 360 365 370 375Crank Angle [CA]

CO2

H2O

GasolineCA10 370.3CA

3731020

CO2H2O

Dry CSPCSPN2

GasolineT in = 149C

M a s s -

A v e r a g e d

T e m p . [

K ]

372

371

370

1000

1 0 % B u r n e d

[ C A ]

980

369960

368

367

366

365

21 20.8 20.6 20.4 20.2 20 19.8 19.6 19.4 19.2 19Intake O 2 Mole Fraction [%]

900

920

940


20/21

Real EGR Effects (2)z Intake concentrations increase with

EGR level and retard. 373372z These effects of Real EGR are also 371

370important for explaining the effects of 369retained residuals. 368

367z Discussed in Proceedings of the 366Combustion Institute , Vol. 31, pp. 365

CSP

Real EGR

PRF80Tin = 72C

CSP

Real EGR

GasolineT in = 149C

CSP

Real EGR

iso-OctaneT in = 175C

28952902, 2007.

1 0 % B u r n P o

i n t [ C A ] 372

371

370

369368

367

366

PRF80iso-OctaneGasoline

Intake CO

iso-OctanePRF80Gasoline

Intake HC

300

250

200

150

100 I n t a k e C O - [ p p m

]

50 3650 372

1200

1000

I n t a k e H C -

[ p p m

] 371

370

369

368

400 367

200 36621 20.8 20.6 20.4 20.2 20 19.8 19.6 19.4 19.2 19

21 20.8 20.6 20.4 20.2 20 19.8 19.6 19.4 19.2 19 Intake O 2 Mole Fraction [%]Intake O 2 Mole Fraction [%]

0

600

800


21/21

Compare with NIST Database

z All air displacement except N 2 reduce the compression temperature.z Can be explained by changes of the specific heat capacity.z CO 2 has highest specific heat capacity on a mole basis.z H2O has high specific heat capacity.z N2 has slightly lower specific heat capacity compared to air.z CSP falls between Air and H 2O.z Dry CSP has slightly lower C v compared to CSP.


@ 3

5 0 C A [ K ]

15

20

25

30

35

40

45

50

H e a

t C a p a c i t y -

C v

[ J / K * m o

l ]

CO2H2OCSPDry CSP

Air N2

NIST ChemistryWebBook

900

890

880

870

860850

840

Motored Data

N2Dry CSPCSPH20CO2

830

820

300 400 500 600 700 800 900 1000 1100 1200 21 20 19 18 17 16 15 14 13 12 11Temperature [K] Intake O2 Mole Fraction [%]

egr effect on auto ignition sandia labs

Documents