Exhaust Particle Sensor for OBD Application Leonidas Ntziachristos, LAT/AUTh Pavlos Fragkiadoulakis, LAT/AUTh Zissis Samaras, LAT/AUTh Kauko Janka, Pegasor Juha Tikkanen, Pegasor
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
1. Background and objectives
2. Operation principle and prototype
3. Signal relevance
4. Engine tests
5. Vehicle tests
6. Conclusions & outlook
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
1. Background and objectives
2. Operation principle and prototype
3. Signal relevance
4. Engine tests
5. Vehicle tests
6. Conclusions & outlook
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
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
1. Background and objectives
2. Operation principle and prototype
3. Signal relevance
4. Engine tests
5. Vehicle tests
6. Conclusions & outlook
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
1. Background and objectives
2. Operation principle and prototype
3. Signal relevance
4. Engine tests
5. Vehicle tests
6. Conclusions & outlook
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
Tw o ring-off cracks
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
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
Tw o ring-off cracks
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
1. Background and objectives
2. Operation principle and prototype
3. Signal relevance
4. Engine tests
5. Vehicle tests
6. Conclusions & outlook
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
1. Background and objectives
2. Operation principle and prototype
3. Signal relevance
4. Engine tests
5. Vehicle tests
6. Conclusions & outlook
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