In Use NTE PM Measurement Methodology In Line, Real … lug‐curve form Mack MP7 – 355E (MY 2004) PPS – Pegasor Particle Sensor. ... • Horiba OBS and Sensors SEMTECH • Constant

CAFEE Center for Alternative Fuels, Engines and Emissions

In‐Use NTE PM Measurement Methodology using an In‐Line, Real‐Time Exhaust PM

Emissions Sensor

Marc C. Besch1, Arvind Thiruvengadam1, Hemanth Kappanna1, Alessandro Cozzolini1, Peter Bonsack1, Daniel K. Carder1,

Juha Tikkanen2, Mridul Gautam1

Department of Mechanical and Aerospace EngineeringWest Virginia University

1Pegasor Ltd, Tampere, Finland

15th ETH Conference on Combustion Generated NanoparticlesJune 26‐29, 2011, Zurich, Switzerland


Content

• Particle Sensor Technology• NTE Measurement Methodology• Experimental Setup• Results and DiscussionEngine Dynamometer ResultsChassis Dynamometer Results(On‐Road Testing Results)

• Conclusions


• In‐use Emissions Compliance Measurements/Testing: Quantification of PM mass

emitted during Not‐to‐Exceed (NTE) events

Establishing mass reference for aerosol in real‐time

• Other Fields of Application On‐board Diagnostics

Applications PM Sensor for Development

and Implementation of DPF Regeneration Strategies

Combustion Research and Engine Base Calibration Applications

Introduction

650 850 1050 1250 1450 1650 1850 2050 22500

200

400

600

800

1000

1200

1400

30%

Max. Torque

nlo

n15%

nhi

T30% Power

T50% Power

T70% Power

Engine Speed [rpm]

Eng

ine

Torq

ue [f

t-lb]

Engine lug‐curve form Mack MP7 – 355E (MY 2004)

PPS – Pegasor Particle Sensor


Sensor ‐ Description of Technology

Advantages:• Real‐time• Continuous operation• No PM sample collection• No external dilution of exhaust needed

Picture provided by Pegasor Oy

• Measurement based on escaping current principle

Operational Parameters:• Sampling rate up to 100 Hz • Sensor output can be calibrated to [mg/m3] or [#/m3]

Dilution Air in

Sample Inlet Turbulent mixing of

particles and ions

~2kV 5µA

Ionized Air out

High Voltage

Charged particles and excess ions

Only charged particles leave the

ion trapElectric field

removes all ions

Corona Needle

(Tungsten)

PPS Installation on Exhaust:


NTE In‐use Measurement Method

Calculation of PM mass [mg] during NTE event:

0

1N

TE E

vent

[-]

0 200 400 600 800 1000 1200 1400 1600

0

50

100

150

200

250

300

Time [sec]

PP

S S

igna

l [m

V]

ESC-Modes (Run 1)ESC-Modes (Run 2) NTEEnd

NTEStart

t

tdttPPS

TestEnd

TestStart

t

tdttPPS

TestEnd

TestStart

NTEEnd

NTEStart

t

t

t

tPPSRatio

dttPPS

dttPPSPM

CycleTotalMassPPSRatioNTEMass PMPMPM

CycleTotalMassPM

PMMass Total‐Cycle = TPM from gravimetric filter sample


Experimental SetupEngine Dynamometer: Chassis Dynamometer: On‐Road Testing:

• Full flow dilution tunnel (CVS‐SSV)• 11L Mack MP7 ‐ 355E (2004)• No aftertreatment system• EEPS (TSI, Model 3090)• CPC (TSI, Model 3025)• MSS (AVL, Model 483)• Intake Air Flow Measurement• Proportional Flow TPM Sampling

PPS

PPSPPS

TPMTPMTPM

• Full flow dilution tunnel (CVS‐SSV)• 6.6L Duramax ‐ GMC4500 (2004)• GVW ~ 12’000 pounds• Diesel Oxidation Catalyst (DOC)• Exhaust Flow Measurement• Horiba OBS and Sensors SEMTECH• Proportional Flow TPM Sampling

• 6.6L Duramax ‐ GMC4500 (2004)• GVW ~ 12’500 pounds• Diesel Oxidation Catalyst (DOC)• Exhaust Flow Measurement• Horiba OBS and Sensors SEMTECH• Constant Flow TPM Sampling


Setup ‐ Engine Test Cell/Chassis Dynamometer

CVS-SSV (40 CFR, Part 1065)

EjectorDilutor

Dry, 25°C Compressed

HEPA Filtered Air Supply

EEPSTM

(TSI, Model 3090)

CPC(TSI, Model 3025)

1st Stage Mini Dilution Tunnel To Outlet

HEPA filtered Dilution Air

To Outlet

ORInsulated Stainless Steel

Line

Heated, Dry, HEPA Filtered Air for

1st Stage Hot Dilution

2nd Stage Mini Dilution Tunnel

Exh

aust

Tra

nsfe

r Pip

e

From Engine

To Outlet

3nd StageDry, 25°C

Compressed HEPA Filtered

Air Supply

PPS

Gaseous EmissionHC, CO, CO2,

NOx, NO

AVLMSS 483

TPMFilter

Holder

Gravimetric PM

2.5μmCut-pointCyclone

Pressurized Air Supply

Chiller

Sec

onda

ry

Dilu

tion

Air

Heated Line

MFC

MFC

Exhaust Flow Measurement, Annubar®(Chassis Dynamometer)

Intake Air (LFE) & Fuel Flow Measurement (Engine Test Cell)

PPSC

PPSPPSExh DRCC

CVSExhCVS DR

CC 1

PMSecCVSFilter DR

CC

1

3082.1][0053.0 psiDRPPS

Exh

mixCVS m

mDR

sec VV

VDR

sf

sfPMSec

mixm

secV

sfV

ExhVfuelIntake mV &

1st Stage Hot, ~ 130C DR = 6

2nd Stage Cold, ~ 25C DR = 24

3rd Stage ** Cold, ~ 25C DR = 8

** Only for CPC

Dilution System for EEPS and CPC:

TPM

TVDLfLoss ,,, (Not applied to presented results)


From Engine

Dry, Compressed

HEPA Filtered Air Supply

PPSTPMFilter

Holder

Gravimetric PM

2.5μmCut-pointCyclone

Pressurized Air Supply

Porous Wall Dilutor

MFC

MFC

Heated Line

Secondary Dilution Air

Heated Tape and Insulation

Pdiff

Pabs

Annubar®

Flow Device

T4

T5

Setup ‐ On‐Road Testing

TPM

sfV

secVExhV

PPSC

PPSPPSExh DRCC PMSecCVSFilter DR

CC

1

TVDLfLoss ,,, (Not applied to presented results)

3082.1][0053.0 psiDRPPS

sec VV

VDR

sf

sfPMSec

y = 0.0053x + 1.3082R² = 0.9688

0.00.20.40.60.81.01.21.41.61.8

0 20 40 60 80

Dilu

tion

Rat

io [-

]

Pressure [psi]

Dilution RatioLinear (Dilution Ratio)

PPSDR

PPS

EFM

TPM

Porous Wall Dilutor

Heated Line

Cyclone

OBSSEMTECH


Results ‐ Engine Test Cell

0 200 400 600 800 1000 1200 1400 16000

50

100

150

200

250

300

Time [sec]

PPS

Sign

al [m

V]

Raw PPS SignalFiltered PPS Signal

0 200 400 600 800 1000 1200 1400 16000

50

100

150

200

250

300

350

400

PPS

Con

cent

ratio

n [m

V/(U

nit F

low

)]

Time [sec]

CPPS

CExh

CCVS

CPMFilter

PPS Signal Filtering/Smoothing:• Savitzky‐Golay (Least‐Squares Smoothing Filters)

• For Steady‐State:• Frame Size ‐ 8.1 sec• Filter Order ‐ 3

• For Transient Cycle: • Frame Size ‐ 2.1 sec• Filter Order ‐ 5

Instrument Grounding at On‐Road

PPS Concentration as calculated at different locations in the measurement stream between PPS sample cell (blue line) and gravimetric filter face


Results ‐ Engine Test Cell (ESC)

340 680 1020 13740

100

200

300

400

500PMPPS vs. TPM at 1228 rpm

PM

PPS [m

g] a

nd T

PM

[mg]

340 680 1020 13740

100

200

300

400

500PMMSS vs. TPM at 1228 rpm

PM

MSS

[mg]

and

TP

M [m

g]

Torque [ft-lb]

309 618 927 12490

50

100

150

200

250


309 618 927 12490

50

100

150

200

250


Torque [ft-lb]

267 534 802 10790

200

400

600


PMPPS

TPM

267 534 802 10790

200

400

600


Torque [ft-lb]

PMMSS

TPM

• AVL MSS => Corrected for temperature influence and thermophoretic losses• TPM includes absorbed SOF => PPS and AVL do not measure this fraction => Possible correction based on HC

Possible influence of PPS sample extraction method at high gas velocity?

For all modes:TPM not “Background” corrected.


0 500 1000 1500 2000 2500 3000 35000

1000

2000

3000

4000Engine Speed [rpm]

0 500 1000 1500 2000 2500 3000 35000

200

400

600

Engine Torque [ft-lb]

0 500 1000 1500 2000 2500 3000 35000

0.5

1NTE Zone [-]

0 500 1000 1500 2000 2500 3000 35000

100

200

300

400

500

600PPS Signal [mV]

Time [sec]

Results ‐ Chassis Dynamometer

NTE Zones (t > 30sec)

From ECU

From ECU

From OBS


• Response of PPS to PM emissions during the transient test cycle (FTP) was similar to that of EEPS (R2 Value: 0.8969) and AVL MSS (R2 Value: 0.8479).

• Development of NTE In‐use Measurement Method using the PPS Signal to calculate PM during NTE events.

• Demonstration of this method based on engine dynamometer experiments => PPS‐Method captures general trends.– Possible influence of high exhaust flow rates on sample extraction efficiency

– Accounting for particle losses within transfer pipes– Influence of SOF on gravimetric filter weight

• Chassis Dynamometer and On‐Road analysis is ongoing.

Conclusion and Outlook


Thank You for Your Attention


0 200 400 600 800 1000 12000

20

40

60

80

100

120

Time [sec]

PP

S M

ass

Con

c. M

odel

[mg/

m3]

MS

S M

ass

Con

c. [m

g/m

3]

PPS (Linear Prediction Model)AVL MSS-483

0 200 400 600 800 1000 12000

0.2

0.4

0.6

0.8

1

Time [sec]

Nor

mai

lzed

PP

S S

igna

l and

MS

S M

ass

Con

cent

ratio

n

PPSAVL MSS-483

Results ‐ PPS vs. AVL MSS, Engine‐out

• Engine: Mack MP‐7 (MY 2004)• No aftertreatment, engine‐out measurement• FTP‐Cycle, Dynamometer

Linear Regression Coefficients:(Least Squares Method)

PPS [mg/m3] = 0.1787+ 0.0581*PPS [mV]R‐Square Value: 0.84790 500 1000 1500 2000

0

20

40

60

80

100

MS

S S

oot C

once

ntra

tion

[mg/

m3 ]

PPS Signal [mV]


0 200 400 600 800 1000 12000

0.5

1

1.5

2

2.5

3

3.5x 10

6

Time [sec]

PP

S N

umbe

r Con

c. M

odel

[#/m

3]E

EP

S T

otal

Num

ber C

onc.

[#/c

m3]

PPS (Linear Prediction Model)EEPS (TSI, 3090)

0 200 400 600 800 1000 12000

0.2

0.4

0.6

0.8

1

Time [sec]

Nor

mai

lzed

PP

S S

igna

l and

EE

PS

Tot

al N

umbe

r Con

c.

PPSEEPS (TSI, 3090)

Results ‐ PPS vs. EEPS, Engine‐out

• Engine: Mack MP‐7 (MY 2004)• No aftertreatment, engine‐out measurement• FTP‐Cycle, Engine Dynamometer

Linear Regression Coefficients:(Least Squares Method)

PPS [#/m3] = 2.244E4+ 1.777E3*PPS [mV]R‐Square Value: 0.8969

Forcing linear model through zero point:=> Coeff(1) = 0

0 500 1000 1500 20000

0.5

1

1.5

2

2.5

3

3.5x 10

6

EE

PS

Tot

al N

umbe

r Con

c. [#

/cm3 ]

PPS Signal [mV]


Sensor – Operational Parameters, cont’

• Low temperature version max 250 °C• High temperature version max. 850 °C• High concentration version 10 µg/m3‐250 mg/m3

• High sensitivity version ~1µg/m3

• Sensor dimensions 20‐40 mm diameter, 100‐200 mm long – to be decided together with customers

• Electronics; 80x40x20 mm3

• Sensor output calibrated to mg/m3 or #‐particles/cm3

• Sensor is installed outside the tailpipe with only inlet and outlet in the tailpipe

• Environmental conditions up to 85 degrees C, IP 45


Sample Outlet

Sample Inlet

Dilution Air

Sensor Electronics

Sensor ‐ Description of Technology Cont’d

RS485 Connector

0

2

4

6

8

10

12

14

16

0 20 40 60 80

Flow

Rat

e [lp

m]

Pressure [psi]

Sample InletSample OutletDilution Air

y = 0.0053x + 1.3082R² = 0.9688

0.00.20.40.60.81.01.21.41.61.8

0 20 40 60 80

Dilu

tion

Rat

io [-

]

Pressure [psi]

Dilution RatioLinear (Dilution Ratio)

Documents

In Use NTE PM Measurement Methodology In Line, Real … lug‐curve form Mack MP7 – 355E (MY 2004) PPS – Pegasor Particle Sensor. ... • Horiba OBS and Sensors SEMTECH • Constant