a university consortium on efficient and clean high ... · a university consortium on efficient and...

2011 DOE Merit Review - 1MITUCBUM

HPLBUniversityConsortium

Project ID: ACE019

A University Consortium on Efficient and CleanHigh-Pressure, Lean Burn (HPLB) Engines

University of Michigan (UM):Dennis N. Assanis (PI),

Aris Babajimopoulos, Stani V. Bohac, Zoran S. Filipi, Hong G. Im, George A. Lavoie, Margaret S. Wooldridge

Massachusetts Institute of Technology (MIT):Wai Cheng

University of California, Berkeley (UCB):Jyh-Yuan Chen, Robert Dibble

May 11th, 2010

DOE Project DE-EE0000203This presentation does not contain any proprietary, confidential, or otherwise restricted information



Project ID: ACE019

Overview

• TIMELINE– Start - Oct 2009– Finish – Dec 2012– 50% completed

• BARRIERS ADDRESSED– Improved fuel economy in

light duty gasoline engines– Fundamental knowledge of

advanced engine combustion

• BUDGET– Total project funding - $3,000k– Rec’d FY10 - $1,000k– Rec’d FY11 - $ 500k

• PARTNERS– Universities – UCB, MIT– Collaborations – SNL, LLNL,

ORNL, ANL– Industrial – GM, Conoco-

Phillips, BP, Ford



Project ID: ACE019

20

30

40

50

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

η _B

rake

(%)

Φ' = Φ (1-RGF)

etaB_v_Phi_prime_TC50_AIR_EGR160

200

240

280

320

BSFC(g/kWh)

1.5Pin (bar) = 1.0

3.0

AIREGR

GT-Power simulation*

DILUTION

BOOST

Objectives and Relevance

DOE VTP Technical Target:• Demonstrate path to achieve 45% peak

engine efficiency with 25 - 40% potential vehicle fuel economy (FE) gain

Project Objectives:Explore advanced dilute, high pressure combustion modes and fuel properties as enablers to achieve FE targets

- Develop analytical link between engine combustion results and potential in-vehicle FE gains

- Investigate benefits of stratification- Explore multi-mode combustion: Spark Assisted Compression

Ignition (SACI), and Microwave Assisted Spark Plug (MWASP)- Determine potential of novel fuel properties for improved FE

A

B

* GT=Power simulation includes: heat transfer, friction, CR = 12, and constant B10-90 = 25 °, CA50 = 10 °ATC



Project ID: ACE019

Approach

• Task 1: System and FE analysis tools– Establish link between advanced combustion modes and potential vehicle FE gains

with GT-power and Matlab simulations (UM)

• Task 2: Stratification as a means to control combustion– Carry out experiments in RCM and gasoline fueled boosted diesel (MIT)– Apply numerical modeling to study turbulent autoignition modes (UM)

• Task 3: Multi-mode combustion– Conduct engine experiments on SACI in FFVA DI engine (UM)– Use CFR engine to test combustion limits with MWASP (UCB)– RCM and Optical Engine experiments on SACI (UM)– Use 1-D modeling of laminar flames in high preheat, high dilution mixtures to

develop correlations for multi-dimensional models– Develop CFD model of SACI with KIVA (UM)

• Task 4: Fuel properties for optimum FE– Explore low octane and other gasoline based fuels in FFVA engine (UM)– Gather fundamental ignition and kinetic data for alternate fuels in RCF (UM)



Project ID: ACE019

Consortium Experimental FacilitiesUM optical engine(SACI and fuels)

UM diesel engine(fuels)

MIT rapid compressionmachine (fuel ignition)

UM camless FFVA engine(multi-mode combustion)

UCB CFR engine(MWASP)

UM RCF(ign. chemistry)

MIT boostedDiesel (stratification)



Project ID: ACE019

Technical Accomplishments (highlights)

• Task 1: System and FE analysis tools– Thermodynamic analysis shows engine efficiency benefits of HPLB (UM)– Vehicle FE framework shows potential vehicle-level gains with HPLB strategy (UM)

• Task 2: Stratification as a means to control combustion– Stratification moderates heat release rate for gasoline HCCI (MIT)– Identified combustion modes with thermal stratification using DNS tools (UM)

• Task 3: Multi-mode combustion– SACI extends high load limit in fully flexible valve actuation (FFVA) engine (UM)– Microwave-Assisted Spark Plug (MWASP) extends lean limit (UCB)– Developed laminar flame speed (SL) correlations for SACI conditions (UM)– Developed CFD model of SACI using new laminar flame speed correlations (UM)

• Task 4: Fuel properties for optimum FE– Low octane fuel extends HCCI high load limit in FFVA engine (UM)– New RCF data obtained on the ignition characteristics of n-butanol (UM)



Project ID: ACE019

Task 1: Thermodynamic Analysis Shows Engine Efficiency Benefits of HPLB (UM)

• Objective: determine thermodynamic conditions for best FE

• Approach: Use GT-Power model with ideal T/C and fixed burn duration

• Results:– Optimal boosting strategy increases fuel-to-

charge equiv. ratio (Φ’) as boost is increased– Best overall engine efficiency occurs under

highly dilute, high pressure conditions, i.e. advanced combustion

– Peak brake engine efficiency ~ 42%– BSNOx < 0.26 g/kWh up to 18 bar BMEP

• Future Work: – Refine analysis with realistic combustion and

T/C constraints; apply simple aftertreatment constraints (TEX) for HC, CO emissions

– Link FE projections to experimental data in Tasks 2, 3, and 4.

INT EXH

I.C.

EGR

C TηC

ηmech

ηT

GT-Power model- CR = 12; RPM = 2000- Ideal T/C; Eff.=50%- B10-90 (25°)- CA50 (10°ATC)- heat transfer, friction- 2-zone NOx model

HCCI ADV.COMB

SI

20

30

40

50

0 10 20 30 40

η_br

ake

(%)

BMEP (bar)

1.01.5

2.0 2.5 3.0

P_in (bar)

etab_v_bmep_AIR_TC50_EIVC_Throt

Φ = 1Throttled

0.3

0.4

0.5 0.6 1.0Φ’

2

3

1

AIR DILUTION; T/C Eff. = 50%

Φ’ = Φ (1-RGF)



Project ID: ACE019

Task 1: Low NOx Emissions Are Achievable with Optimum Boosting Strategy

• Optimum boosting strategy and T/C efficiency of 50% result in BSNOx < 0.26 g/kWh

– Up to 18 bar BMEP (Air dilution)

– Up to 25 bar BMEP (EGR dilution)

– Peak ηbrake ∼ 42%

20

30

40

50

0.0

1.0

2.0

3.0

0 10 20 30 40

η_br

ake

(%)

BSN

Ox

(g/k

Wh)

BMEP (bar)

etab and BSNOx_v_bmep_TC50 AIR_EGR_2axes

NOx limit

AIREGR

Turbocharger Efficiency = 50%



Project ID: ACE019

1

1.1

1.2

1.3

1.4

1.5City-Hwy

Rel

ativ

e FE

SI-NA ADV-NA ADV-boost

1.5

1.4

1.3

1.2

1.1

1.0

Task 1: Vehicle FE Framework Shows Vehicle Level Gains with HPLB Strategy (UM)

• Objective: Provide projection of vehicle FE gains achievable with advanced combustion

• Approach: combine GT-Power generated BSFC maps of 3 ideal engine strategies (previous slide) with Matlab drive cycle simulation

• Results:– 1 → 2 dilution (N. A.) ~ 25% gain

(assumes optimum combustion at all points, and 100% comb. Efficiency)

– 2 → 3 dilution + boosting + downsize ~ 50% gain (assumes T/C eff = 50%)

• Future Work:– Include more realistic combustion

characteristics and constraints

Mid-Size Sedan,3285 lb

Engine Speed (rpm)

Eng

ine

BM

EP

(bar

)

1000 2000 3000 4000 5000 60000

2

4

6

8

10

12

NA - 3.3L

Φ=1Throttled

Engine Speed (rpm)

Eng

ine

BM

EP

(bar

)

1000 2000 3000 4000 5000 60000

2

4

6

8

10

12

NA - 3.3L

AIRdilute

Engine Speed (rpm)

Eng

ine

BM

EP

(bar

)

1000 2000 3000 4000 5000 60000

5

10

15

20

25

BOOSTED – 2.2L

AIRdilution

* Circled numbers refer to engine combustion strategies on earlier slide

2

3

1 *

*



Project ID: ACE019

Task 2: Stratification Moderates Heat Release Rate for Gasoline HCCI (MIT)• Objective: investigate benefits of

interaction between gasoline fuel properties and stratification

• Approach:– Obtain fuel ignition data in RCM– Vary injection characteristics in

boosted diesel engine to vary stratification

• Results:– Correlation developed for τIGN in RCM– In engine, late injection (90° ABDC) is

effective in setting up stratification by moderating heat release rate

• Future work:– Connect RCM results with engine data– Vary fuel properties 0

1

2

3

4

5

6

0 2 4 6 8

Knoc

k in

dex

(MW

/m2 )

NIMEP (bar)

Gasoline: Pinj=500 bar; Tin=60oC; MAP=1.5 bar

inj 30deg bbdc

inj 30deg abdc

inj 90 abdc • Later injection• Incr. Stratification

Kno

ck In

dex

(MW

/m2 )

1.1 1.2 1.3 1.4 1.51

1000/(T[K])

800 750 700 650 600Temperature [K]

t / τ

−τ = τ = Φ1.5 0.3d d(x 0) (1-x )

1.0

0.8

0.6

ΦDilution

0.0 0.2 0.4

RCM Data(Gasoline)



Project ID: ACE019

Task 2: Identified Combustion Modes with Thermal Stratification Using DNS Tools (UM)

• Objective: characterize ignition regimes in thermally stratified HCCI combustion

• Approach: Use Computational Singular Perturbation (CSP) analysis of timescales in a reacting flow field to determine Importance Index

(Shows relative importance of deflagration / flame vs. autoignition)

• Results:– Turbulent stratified combustion

invokes multiple reaction modes• Future work:

– Application of method to higher hydrocarbon fuels

– Compositional stratification

T)Convection(T

TDiffusion)(T

Tslowslow

III −− +=

2 D - DNS simulation (H2-Air)Deflagration

1.0

H: Homogeneous kernel, F: FrontS: Spontaneous Ignition, D: Deflagration

IT

TI

0.0

HD

HS

FD FS

SpontaneousIgnition

0.0

In Collaboration With Mauro Valorani, University of Rome, La Sapienza

M = 1 contour in black [denotes active reaction zone]



Project ID: ACE019

Task 3: SACI Extends High Load Limit in Fully Flexible Valve Actuation (FFVA) Engine (UM)

• Objective: demonstrate load extension with spark assist

• Approach:– Increase load by adding fuel and reducing

trapped residual gas (NVO) with Φ= 1 (increase fuel-to-charge Φ’)

– Control phasing with spark timing

• Results:– Heat release rate moderated by SACI– Maximum load increased to 7.3 bar– Encountered SI knock at highest loads

• Future Work:– Explore variations in proportion of flame vs.

autoignition heat release.– Extend work to higher pressures (cell

upgrade in progress to provide boost)-40 -20 0 20 40 60

0

10

20

30

40

50

60

Crank Angle (deg)A

HR

R (J

/CA

)

4.93 bar5.96 bar6.59 bar6.89 bar7.31 bar

0

200

400

600

800

1000

0 5 10 15

NM

EP

[kP

a]

CA50 [°ATC]

Φ’0.7

0.5

0.3

SparkAdv.

SACI

HCCI

Incr. load

AHR

R (J

/CA)

CA (°ATC)

CA50 = 10°ATC

Manofsky, L., Vavra, J., Babajimopoulos, A., and Assanis, D. (2011) Bridging the Gap between HCCI and SI: Spark-Assisted Compression Ignition. SAE Technical Paper 2011-01-1179



Project ID: ACE019

Task 3: Microwave-Assisted Spark Plug (MWASP) Extends Lean Limit (UCB)

• Objective: Evaluate the potential benefits of MWASP for advanced combustion modes

• Approach: – Use CFR engine (CR = 7, 9) to test

ignition under lean limit conditions– Model ignition with simplified methane

kinetics and electron gas dynamics• Results:

– Lean limit extension observed– Modeling suggests benefit may

decrease at higher pressures• Future Work:

– Extend experiments into SACI range (higher CR, leaner, and more preheat)

Kinetic model of ignition in a WSRshows effect of energy deposition mode

10 atm 1 atm50 atm

gas (thermal)MW (electronic)

none

lean limitextensionSI

MWASP

MODEOF ENERGY DEPOSITION

Nor

mal

ized

Fue

l Con

sum

ptio

nTe

mpe

ratu

re (K

)

Time (s)

λ

DeFilippo, A., Saxena, S., Rapp, V.H., Ikeda, Y., Chen, J-Y, Dibble, R.W., (2011) Extending the lean flammability limit of gasoline using a microwave assisted sparkplug, SAE Paper no. 2011-01-0663



Project ID: ACE019

Task 3: Developed Laminar Flame Speed (SL) Correlations for SACI Conditions (UM)• Objective: Understand fundamental

flame behavior and properties for highly preheated, highly dilute mixtures

• Approach:– Use transient flame code HCT to build

dataset of flame properties in SACI regime where experimental data is limited

– Build mathematical correlations for use in turbulent combustion models

• Results:– Viable flames predicted in regions where

SACI observed– Correlations made for AIR and EGR dilute

flames

• Future Work: – Validate vs. images from optical engine

and in RCF

EXPT

HCT dataset lines ofconstant SL

40 bar

Isooctane Flames

( ) ( )10

02 , 01 exp

nD m u b

L EGR f ub u

T T TS D Y FY G TT T T

− = − Φ − −

AIR onlyEGR (O2 term)

Form of correlation shows Air and EGR dependence

Middleton, R.J. et al., “A Computational Study and Correlation of Premixed Isooctane Air Laminar Reaction Front Props. with Combustion Products,” In Preparation.

TB (K)

TU (K)



Project ID: ACE019

Task 3: Developed CFD Model of SACI Using New Laminar Flame Speed Correlations (UM)

• Objective: develop CFD model of SACI and use to explore possible benefits

• Approach: combine laminar flame speed correlations with models of coherent flamelet and multizone autoignition (CFMZ model)

• Results:– Model compares well with images obtained in

UM optical engine– Good qualitative agreement with heat release

curves in metal engine

• Future work:– Perform parametric studies on variables

affecting heat release partition between flame and autoignition and sensitivity to spark timing

– Validate limits vs. metal engine data

KIVA-CFMZ MODEL EXP

BURNEDGAS

FLAME

UNBURNED (AUTOIGNITION)

Spark

Autoignitionbegins

OpticalEngineImages

Martz, J. B., (2010) , Simulation and Model Development for Auto-Ignition and Reaction Front Propagation in Low-Temperature High-Pressure Lean-Burn Engines , Ph. D. Thesis, University of Michigan, October 2010.



Project ID: ACE019

275

300

325

350

375

400

425

450

475

-2 0 2 4 6 8 10 12 14

GIM

EP [k

Pa]

CA50 [°ATC]

NVO and Fueling Sweep for Certification Gasoline (RD387)

275

300

325

350

375

400

425

450

475

-2 0 2 4 6 8 10 12 14

GIM

EP [k

Pa]

CA50 [°ATC]

NVO and Fueling Sweep for NH40 (40% n-Heptane, 60% RD387)

Task 4: Low Octane Fuel Extends HCCI High Load Limit in FFVA Engine (UM)

• Objective: Explore ways to use fuel properties to increase HCCI load limit

• Approach:– Use NVO in FFVA engine to provide control

over combustion phasing– Inlet temperature = 45°C– Initially compare gasoline vs. a low octane

surrogate NH40 (40% n-heptane in gasoline)

• Results:– NH40 had higher load limit possibly because

it had faster 10-90 burn duration and could be retarded more without instability

• Future Work:– Expand range of test fuels– Explore benefit of spark assist

Gasoline (RD387); RON = 91

NH40 (40% n-C7H16 / 60% RD387); RON = 53

RI lmit5 MW/m2

RI lmit5 MW/m2

COVIMEP5%

COVIMEP5%

INCR. FUELING

INCR. FUELING



Project ID: ACE019

Task 4: New RCF Data Obtained on the Ignition Characteristics of n-butanol (UM)

• Objective: Investigate ignition of alternative fuels such as n-butanol as a possible infrastructure compatible fuel and blending component for bio-fuels

• Approach:– RCF provides ignition delay data– Sampling system gives intermediate species

data– Both data are necessary to quantify the impact

on fuel reactivity and exhaust emissions• Results:

– Measured ignition delays are in excellent agreement with Black et al. model

– RCF speciation data identify weaknesses in reaction pathways important for predicting and mitigating emissions

• Future work:– Study fuel blends using RCF and optical engine

UM RCF data forP ≅ 3.0 atm

SAMPLING SYSTEM

model resultssampling

experiments

ethene

model results



Project ID: ACE019

Future Work FY11-12

• Task 1: System and FE analysis tools– Refine analysis of optimum combustion strategy with realistic combustion constraints and

turbocharger characteristics. Apply simple TEX constraint for CO, HC aftertreatment. Explore tradeoff between CR and T/C efficiency.

– Link FE projections to experimental data from Tasks 2, 3, and 4.

• Task 2: Stratification as a means to control combustion– Connect RCM ignition results with engine timing data; vary fuel properties– Apply Computational Singular Perturbation (CSP) method to higher hydrocarbon fuels and

compositional stratification

• Task 3: Multi-mode combustion– Explore variations in partition of flame vs. autoignition heat release in SACI in FFVA engine– Extend FFVA SACI work to higher pressures (cell upgrade in progress to provide boost)– Extend MWASP experiments into SACI range (higher CR, leaner, and more preheat)– Compare flame front model results vs. experimental images in optical engine and RCF– Use CFMZ model of SACI to investigate variables affecting details of heat release; sensitivity to

spark timing; and nature of limits by comparison to metal engine data.

• Task 4: Fuel properties for optimum FE– Expand range of test fuels, and include spark assist in testing– Carry out RCF studies of ignition properties of fuel blends



Project ID: ACE019

Collaborations and Coordination

• Working on boosted single cylinder HCCI studies with GM• Working with Microwave Enhanced Ignition device supplied by Yuji Ikeda, Imagineering,

Inc., Japan • Collaborating on SACI and combustion stability with Robert Wagner (ORNL)• Supplied engine maps for HCCI to Argonne National Labs• Currently working with Ford on the next steps towards changing our optical engine to

direct injection.• Working with Chevron on enhancing the effectiveness of stratification on controlling LTC

heat release with fuel (MIT)• Studying sensitivity of spark-assisted HCCI engines to range of market fuels; with BP and

Ford (MIT)• Working with Jacqueline Chen (Sandia National Laboratories), Ramanan Sankaran

(ORNL), Mauro Valorani (University of Rome, La Sapienza), Chris Rutland (UWisc) on mixing effects on HCCI combustion.

• Collaborated with C. K. Westbrook and Bill Pitz (LLNL) on validating reaction mechanisms for long chain alkanes, small esters – now working on alcohols

• Working with Marco Mehl (LLNL) to apply new surrogate fuel kinetics to engine models



Project ID: ACE019

Summary

• Task 1: System and FE analysis tools– Thermodynamic analysis confirms HPLB strategy; BTE of 42% with BSNOx < 0.26 g/kWh– System framework developed for vehicle FE assessment: demonstrated potential vehicle level

impact of boosted dilute strategies

• Task 2: Stratification as a means to control combustion– Initial experiments carried out with stratified gasoline fueled diesel engine show stratification

moderates heat release rate– DNS simulation showed multiple combustion/ignition modes in thermally stratified mixture

ranging from spontaneous ignition to deflagration front

• Task 3: Multi-mode combustion– SACI extends high load limit in naturally aspirated FFVA engine– MWASP ignition extends lean limit under NA conditions with potential gains for LTC combustion– Laminar flame correlations developed for high dilution / high preheat conditions– CFD model of SACI developed; good qualitative agreement with optical engine images

• Task 4: Fuel properties for optimum FE– Studies of low octane gasoline fuel shows potential for higher load operation in FFVA (NVO

enabled) HCCI operation– Initial studies on ignition of n-butanol carried out in RCF; speciation results show need for

improvements to kinetic models

a university consortium on efficient and clean high ... · a university consortium on efficient and...

Documents