SAE Presentation

Exhaust Particle Sensor for OBD Application

Leonidas Ntziachristos, LAT/AUTh

Pavlos Fragkiadoulakis, LAT/AUTh

Zissis Samaras, LAT/AUTh

Kauko Janka, Pegasor

Juha Tikkanen, Pegasor

2 2011-01-0626

Contents

1. Background and objectives

2. Operation principle and prototype

3. Signal relevance

4. Engine tests

5. Vehicle tests

6. Conclusions & outlook

3 2011-01-0626

Contents



3. Signal relevance

4. Engine tests

5. Vehicle tests


1.

4 2011-01-0626

Background

Need for OBD on diesel passenger cars and heavy duty vehicles, in view of wide application of DPFs

NOx sensor(s) available

PM sensor technically more demanding:

e.g. Euro VI thresholds:

NOx 1200 mg/kWh-1

PM ~25 mg/kWh-1

Variable chemical and physical nature

5 2011-01-0626

Objectives

1. Present new particle sensor (PPS) to be used for OBD downstream of DPF

2. Describe the relevance of its signal for particle mass and number determination

3. Apply prototype to test efficiency of compromised DPFs

4. Compare with lab instruments over driving cycle

6 2011-01-0626

Contents



3. Signal relevance

4. Engine tests

5. Vehicle tests


2.

7 2011-01-0626

Operation Principle

Technique referred to as “escaping current”

Charger

power

Electrometer

In = Iout - Iin

Iin

Iout

Sample in

Sample out

Isolated power transfer

Faraday cup

Aerosol

charging

In

8 2011-01-0626

Configuration

Patent publication no WO 2009/109688

clean air

corona discharge

pump flow trap

out sample in

ejector throat

+ ion

soot

+

9 2011-01-0626

Characteristics

Benefits

Particles are not collected

Sensitive parts protected from exhaust flow

Turbulent mixing increases sensitivity

Requirements

On-board pumping flow

10 2011-01-0626

Prototype

Inlet flow

Outlet flow

Carrier air ~400 mm

11 2011-01-0626

Prototype Specifications

Carrier flow: 3 lpm @ 0.5bar overpressure

Communication protocol: RS485

Sampling rate: 10 Hz

Response time: 0.3 s

12 2011-01-0626

Contents



3. Signal relevance

4. Engine tests

5. Vehicle tests


3.

13 2011-01-0626

Relevance of signal

Number dp0

(„Active‟) Surface dp1.1-1.4

Mass dp2.2-2.5

Signal is proxy for either number or mass of particles through calibration

14 2011-01-0626

Simulation of signal relevance

3DF

ref,p

i,pi

d

d85.0

1000

1

3i,pii dN

6m

1000

1

3.1i,pidN

4S

Property Range/Calculation

Distribution shape Lognormal

Number count – N (km–1) 5×1012 1014

Mean geom. diameter – dg (nm) 6590

Geometric st. deviation – σg 1.651.80

Fractal-like dimension – DF 2.22.5

Effective density (g/cm3) for dp,i>40 nm

Mass concentration (g km–1)

Sensor signal – S (km–1)

15 2011-01-0626

Boundary size distributions

1E-09

1E-08

1E-07

1E-06

1E-05

1E-04

1E-03

1E-02

1E-01

1E+00

10 100 1000

Mobility diameter - dp (nm)

dm

/dlo

gd

p (

g/k

m)

0.76 mg/km

63 mg/km

N=5×1012 km–1 Dpmean= 65 nm (number weighted) σg=1.65 DF=2.2 Μ=0.76 mg/km

N=1014 km–1 Dpmean= 90 nm (number weighted) σg=1.8 DF=2.6 Μ=63 mg/km

16 2011-01-0626

Signal response over N and M

5.0

5.5

6.0

6.5

7.0

7.5

8.0

8.5

9.0

9.5

10.0

1.5 2.0 2.5 3.0 3.5 4.0

S/m

S/N

95% CI

±30%

±18%

Max theoretical error (not engine specific)

Reducing the error requires engine-specific calibration

17 2011-01-0626

Contents



3. Signal relevance

4. Engine tests

5. Vehicle tests


4.

18 2011-01-0626

Engine tests

Engine type Daimler OM646 DE22 LA

Displacement (cm3) 2148

Fuel injection system Common rail

Maximum power (kW/rpm) 110/4200

Maximum torque (Nm/rpm) 340/2000

Emission standard Euro 4

Operation mode 50 Nm @ 1500 rpm

19 2011-01-0626

Particle filters used

DPF Index 1 2 3 4 5

Substrate material SiC SiC SiC SiC SiC

Segments 16 4 9 16 4

Cell density (in-2) 279 200 300 300 200

Sub. thickness [mil] 10 15 12 12 15

Volume (l) 2.5 2.5 2.5 4 2.5

Length (mm) 150 154 152 254 152

Diameter (mm) 144 144 144 144 145

Condition Good Damaged Damaged Damaged Damaged

20 2011-01-0626

XCT Scans of DPFs

Cracks produced due to thermal shocks prior to testing

All filters regenerated for 3.5 h @ 650°C before testing

21 2011-01-0626

Efficiency evolution

40%

50%

60%

70%

80%

90%

100%

0 1 2 3 4

OEM

Light crack

Multiple cracks

Ring-off crack

Tw o ring-off cracks

Time (×103 s)

Eff iciency

(a)

PPS

40%

50%

60%

70%

80%

90%

100%

0 1 2 3 4

OEM

Light crack

Multiple cracks

Ring-off crack


Time (×103 s)

Eff iciency

(b)

Smokemeter

40%

50%

60%

70%

80%

90%

100%

0 1 2 3 4

OEM

Light crack

Multiple cracks

Ring-off crack


Time (×103 s)

Eff iciency

(a)

PPS

40%

50%

60%

70%

80%

90%

100%

0 1 2 3 4

OEM

Light crack

Multiple cracks

Ring-off crack


Time (×103 s)

Eff iciency

(b)

Smokemeter

22 2011-01-0626

Observations

PPS offers robust measurement of efficiency of compromised filters at high time resolution

All but one DPFs are more than 95% efficient 1000s after regeneration

Compromised DPF with two ring-off cracks <95% efficient when measured with PPS

Further to OBD, PPS signal useful to optimize regeneration frequency

23 2011-01-0626

Contents



3. Signal relevance

4. Engine tests

5. Vehicle tests


5.

24 2011-01-0626

Vehicle tests

PMP compatible, except of CPC (>2.5 nm)

Car model Honda Accord 2.2

FI system Common rail

Pmax (kW/rpm) 103/4000

Nmax (Nm/rpm) 340/2000

CR P (bar) 1600

Emission standard Euro 4

Aftertreatment Close-coupled and underfloor DOCs

0

20

40

60

80

100

120

0 200 400 600 800 1000 1200

Cycle Time (s)

Sp

ee

d (

km

/h) UDC EUDC

25 2011-01-0626

Particle filters used

DPF Index 5 6

Substrate material SiC SiC

Segments 4 16

Cell density (in-2) 200 300

Substrate thickness [mil] 15 12

Volume (l) 2.5 4

Length (mm) 152 254

Diameter (mm) 145 144

Condition Two-ring off cracks Good

26 2011-01-0626

Vehicle/DPF performance

0

5

10

15

20

25

30

35

40

45

50

NEDC cold NEDC hot

Unfiltered

Functional DPF

Damaged DPF

PM

[m

g/k

m]

Euro 5/6

6×

1E+10

1E+11

1E+12

1E+13

1E+14

NEDC cold NEDC hotP

art

icle

num

ber

[km- 1

]

Euro

5/6

(b)

17×

With the functional DPF, the vehicle complies with the Euro 5 limit with some margin, despite the wider range CPC used

27 2011-01-0626

PPS Performance

1E+03

1E+04

1E+05

1E+06

1E+07

1E+08

1E+09

NEDC cold NEDC hot

PP

S S

ignal [

-]

(c)

14×

0

5

10

15

20

25

30

35

40

45

50

NEDC cold NEDC hot

Unfiltered

Functional DPF

Damaged DPF

PM

[m

g/k

m]

Euro 5/6

6×

PN (%)

PPS (%)

PM (%)

Functional DPF

NEDC cold 99 99 98

NEDC hot 100 100 98

Damaged DPF

NEDC cold 89 91 89

NEDC hot 96 98 97

28 2011-01-0626

Real-time recordings

Note: PPS signal not corrected for exhaust flowrate

0

20

40

60

80

100

120

0 200 400 600 800 1000 1200

NEDC Time (s)

0

100

200

300

400

500

600

PPS

PN

Speed

Part

icle

num

ber ×

10

13 (

s-1),

Speed (

km

h-1)

PP

S S

ignal (

-)

29 2011-01-0626

Observations

Response: between PM and PN expression

Sensitivity

PM: higher than 1 mg/km (Euro 5: 4.5 mg/km)

PN: can distinguish <6×1011 km-1

OBD threshold: 1.5×ES (6 mg/km) possible

Time resolution: Real-time monitoring possible

30 2011-01-0626

Contents



3. Signal relevance

4. Engine tests

5. Vehicle tests


6.

31 2011-01-0626

Conclusions & outlook

1. PM OBD sensor concept based on contactless particle detection

2. Signal can be a proxy for either particle mass or number

3. Sensitivity demonstrated with tests on malfunctioning DPFs of various degree

4. Time resolution superseding potential OBD requirements

5. Next steps include downsizing and durability tests

32 2011-01-0626

Thank you for your attention!

SAE Presentation

Documents