SAE 2004

1
SAE 2004

• World Congress

2
Övergripande intryck

• Man måste förlita sig på konventionell
  dieselteknik, eventuellt in kombination med
  PCCI, plus avgasrening för att klara
  emissionskraven lt 2010, dvs. NOx lt 0.27 g/kWh och
  PM lt 0.013 g/kWh. (HCCI-förbränning över hela
  last- och varvtalsområdet inte trolig före 2007).
  
• Återkopplade motorer. tex. EGR!
• Det finns teoretiska möjligheter att undvika NOx
  och PM under diffusiv förbränning.

3
2004-01-0124Emissions Solutions for 2007 and
2010 Heavy-Duty Diesel Engines

• Jeffrey A. Leet, Stefan Simescu, Kent Froelund,
• Lee G. Dodge and Charles E. Roberts
• Southwest Research Institute (SwRI.)

4
Technology evolution2004-01-0124

• Rate - shaped fuel injection
• Injection pressure gt 2000 bar
• Electronic timing control
• 4-valve cylinder head design
• EGR or residual gas control
• Low oil consumption
• lt 15 ppm sulfur
• DPF needed for PM lt 0.01 g/kWh (US 2007)

5
2004-01-0124
6
Reference engine 2004-01-0124
7
2004-01-0124
VGT -Increased boost pressure 180 200 bar Peak
Pressure
8
2004-01-0124
PCCI Diesel combustion Full load range HCCI not
before 07!
9
Post injection for PM control2004-01-0124
10
2004-01-0124
Camless engine gt 2007 Miller cycle Residual gas
control Cylinder deactivation
11
Effect on exhaust temperature by cylinder
deactivation2004-01-0124
12
2004-01-0124Conclusions 1

• An interim NOx level of 1.18 g/hp-hr is allowable
  for the 2007 through 2009 model year heavyduty
  truck engines sold in the USA. The 0.2 g/hp-hr
  NOx level will be required by 100 of the
  on-highway heavy-duty engines starting in model
  year 2010. As of 2007, 100 of the engines will
  be required to achieve 0.01 g/hp-hr PM.
• The effect of a variety of in-cylinder based NOx
  controls has been estimated to achieve the
  interim 1.18 g/hp-hr NOx level, including high
  EGR levels, advanced combustion concepts, high
  injection pressure flexible fuel injection
  systems, advanced control approaches, and VVA.

13
2004-01-0124Conclusions 2

• Additional in-cylinder reductions in PM can be
  accomplished through advanced combustion control
  and high injection pressure fuel systems.
  However, compliance with the 0.01 g/hp-hr PM
  limit in 2007 will require use of a DPF. The use
  of a DPF places increased emphasis on oil
  consumption control and oil formulation.
• The interim NOx levels (1.18 g/hp-hr) can be
  achieved without a NOx exhaust treatment device.
  Providing such an engine would allow an engine
  manufacturer to reduce the cost of the exhaust
  treatment portion, to improve the packaging and
  durability of the exhaust system and to be shown
  as a technology leader.

14
2004-01-0124Conclusions 3

• The further refinement of in-cylinder
  technologies and the integration with NOx exhaust
  treatment devices will be required to achieve the
  2010 emission levels.

15
2004-01-0587Systems Approach to Meeting EPA 2010
Heavy-Duty EmissionStandards Using a NOx
Adsorber Catalyst andDiesel Particle Filter on a
15L Engine

• Shawn D. WhitacreNational Renewable Energy
  LaboratoryBrad J. Adelman, Mike P. May and
  Joseph G. McManusRicardo, Inc.

16
2004-01-0587
lufttillförsel
DOC Oxiderande kat DPF Partikelfilter NAC
NOx adsorber kat
Alternativ för extra bränsletillförsel för
regenerering single leg
17
2004-01-0587
18
2004-01-0587
19
2004-01-0587
20
2004-01-0587
21
2004-01-0587
22
Conclusions2004-01-0587

• Integrating sophisticated controls with an
  advanced engine can enable effective NOx adsorber
  regeneration and desulfation. (Desulfation needs
  650 oC and l lt 0.9)
• NOx and PM emissions reductions of gt90 have been
  demonstrated by a NOx adsorber catalyst and
  diesel particle filter with suitable regeneration
  strategies.
• The fuel economy penalty associated with the FTP
  and SET are 8.2 and 4.5 respectively.
• Challenges to achieving full certification
  include
• 1. NOx adsorber catalyst regeneration during
  full load operation
• while keeping the catalyst within the
  temperature range for
• high efficiency.
• 2. Control of hydrocarbon emission during rich
  regeneration.

23
2004-01-0120Transient EGR Control for
aTurbocharged Heavy Duty Diesel Engine

• Hitoshi Yokomura, Susumu Kouketsu, Seijiro
  Kotooka and Yoshiyuki AkaoMitsubishi FUSO Truck
  and Bus Corp.

24
2004-01-0120
25
2004-01-0120
26
2004-01-0120
27
2004-01-0120
28
2004-01-0120Conclusions

• At high EGR rates, the magnitude of a smoke spike
  occurring at acceleration significantly
  increases,compared to low EGR rates. This is due
  to turbocharger lag and the resulting drop in
  excess air ratio.
• In order to prevent the smoke spike at
  acceleration with high EGR rates, it is effective
  to pause EGR at acceleration. This enables the
  drop in excess air ratio to be held in check, and
  makes it possible to reduce PM to a level typical
  of a low EGR rate. When the EGR pause period is
  longer, it results in a significant reduction in
  the smoke spike, but NOx greatly increases in
  that case.

29
2004-01-0120Conclusions

• In order to control smoke during acceleration, it
  is necessary to properly control the transient
  excess air ratio. If this is done, it also
  becomes possible to reduce NOx emissions. Thus,
  even at the transient engine operation conditions
  tested here, a closed loop EGR control method was
  developed. Feedback and control were utilized, in
  which the excess air ratio was sequentially
  computed and estimated to maintain the target
  optimum excess air ratio.
• The effectiveness of the new EGR control method
  was confirmed using closed loop control of the
  excess air ratio. This method was compared
  experimentally to open loop EGR control combined
  with EGR pause, using a transient test mode.
  Results showed that the high EGR rates made
  possible by closed loop control, enabled NOx to
  be reduced by 40 at an equivalent or lower PM
  emission level.
• Finally, considering the results obtained in this
  work, the following conclusion can be made Using
  a high EGR rate to further reduce NOx emissions
  causes a smoke spike during acceleration. Closed
  loop EGR control, via the excess air ratio,
  effectively minimized the smoke spike,
  consequently making possible a simultaneous
  reduction in both NOx and PM.

30
2004-01-1401On the Correlation between NOx
Emissionsand the Diesel Premixed Burn

• Mark P. B. MusculusSandia National Laboratories

31
2004-01-1401
32
2004-01-1401
33
2004-01-1401
34
2004-01-1401
35
2004-01-1401

• As intake-temperature was decreased, the exhaust
  NOx first decreased and then increased as the
  degree of premixed burning monotonically
  increased.
• As the start of injection (SOI) timing was
  retarded, NOx emissions at first decreased, and
  then increased for late injections with large
  premixed burns. For very late injections near the
  misfire limit, the NOx emissions again decreased
  dramatically, yielding a NOx bump at
  late-injection conditions. Coupled with
  thermodynamic analyses, the results of the
  current study support the following conclusions
  regarding in-cylinder NO formation

36
2004-01-1401

• The peak luminosity from hot soot radiation
  decreased dramatically as the premixed burn
  became larger and exhaust NOx increased. Under
  conditions with increasing premixed burning, the
  effects of radiative cooling may cause real flame
  temperature trends to oppose calculated adiabatic
  flame temperature trends. Thermal NO formation
  rates may therefore increase as soot luminosity
  decreases, causing the actual flame temperature
  to increase. Accordingly, fuel-injection
  strategies that reduce soot formation may
  unfortunately increase thermal NO formation rates
  because of increased flame temperatures.

37
2004-01-1401

• Based on thermal NO formation rates, the increase
  in peak adiabatic flame temperature due to
  compression heating of pre-flame reactants after
  the rapid pressure rise accompanying premixed
  combustion is insufficient to explain the above
  NOX trends.

38
2004-01-1401

• Compression heating of post-flame burned-gases
  may be sufficient to explain a portion of the
  above trends. The magnitude of this effect will
  depend on the thermal residence time in the
  post-flame gases, which in turn depends on
  post-flame mixing rates. Strategies that increase
  post-flame mixing (postinjections) could herefore
  conceivably reduce incylinder NO formation.

39
2004-01-1401

• Under very long ignition-delay operating
  conditions, portions of the mixture at ignition
  are likely lean enough to form significant
  thermal NO in the products of the premixed burn.
  This observation, however, does not necessarily
  imply that the increased premixed burning causes
  increased NOX emissions.

40
2004-01-1399Non-Sooting, Low Flame Temperature
Mixing-Controlled DIDiesel Combustion

• Lyle M. Pickett and Dennis L. SiebersSandia
  National Laboratories

41
Variabler

• Spridarhålsdiameter
• Syrekoncentration före insprutning
• Gastemperatur och densitet före insprutning
• Lokalt lambda (funktion av lift off length)
• Syreinnehåll i bränslet
• Procedur Förvärmning och gasens sammansättning
  styrs via förbränning av en bränsle/luft-blandning
  före insprutningen. Insprutningen sker under
  avsvalningen när önskad trycknivå uppnåtts.

42
2004-01-1399
43
2004-01-1399
44
2004-01-1399
Inflytande av gastemperaturen före insprutningen.
D0.100 mm. 21 syre
Slutsats En sänkt temperatur förskjuter lift off
från spridaren och ger mer tid/yta
för syreinblandning i sprayen ? sotbildning
undviks vid Flt2
45
2004-01-1399
Inflytande av gastemperaturen före insprutningen.
D0.050 mm. 21 syre
Slutsats Med minskad håldiameter ökar
syreinblandningen före lift off. Flt2 uppnås trots
ökad temperatur (1000 K för 0.05 mm mot 850 K för
0.100 mm). Balans mellan tid för syreinblandning
och självantändning.
46
2004-01-1399
21 syre

• Minskar sotbildning
• Minskad håldiameter
• Sänkt omgivnings-tryck
• Sänkt omgivnings-temperatur
• Ökad syrehalt i bränslet
• Ökat insprutningsstryck (ej visad effekt i
  artikeln)
• Det gäller att öka syreinblandningen
• innan lift off!

47
2004-01-1399Effekt av sänkt syrehalt för att
simulera EGR

• Ingen LII sot
• Lift off längden ökar med sänkt syrehalt, vilket
  medför att F vid lift off är ungefär konstant ?
  Ingen sot trots EGR.
• Nettoeffekt Sänkt flamtemperatur ? lägre NOx
  plus sotfri förbränning
• Gain i figurerna indikerar sänkt temperatur med
  sänkt syrekoncentration.

(Ref)
48
2004-01-1399
Sänkt förbränningstemperatur genom sänkt
omgivningstemperatur. 21 syre
Ingen sotbildning Ungefär samma lift off längd
i båda fallen F0.6 (l1.7) för D0.050
mm! Fullständig förbränning på kort sträcka med
D0.050 mm
?Miller!!
49
2004-01-1399
50
2004-01-1399
Fet förbränning vid låg temperatur ger ingen
sot!
51
2004-01-1399
Sänkt förbränningsverkningsgrad?
52
2004-01-1399
Den minskade tryckstegringen Kan vara en effekt
av ökade värmeförluster pga. den förlängda
förbränningstiden.
53
2004-01-1399Sotfri förbränning vs temperatur och
F
För låg temperatur
För mager blandning
54
2004-01-1399Transient combustion
Premixförbränningen efter tändfördröjningen sker
vid samma lambda som förbränningen Vid lift off!!
55
Conclusions 12004-01-1399

• The investigation showed that non-sooting,
  low flame temperature combustion could be
  produced for three different types of conditions,
  as follows
• Using a 50-µm orifice, a 2 diesel fuel, and
  ambient oxygen concentrations as low as 10
  (simulating the use of EGR), a fuel jet was
  non-sooting at typical diesel ambient
  temperatures (1000 K) for mixing controlled
  combustion at flame temperatures as low as 1980
  K. Furthermore, despite the low flame
  temperature, a first-order analysis based on the
  pressure rise measurement showed that the
  combustion was basically complete. The
  investigation showed that the lack of soot
  formation was related to the high degree of
  fuel-air mixing upstream of the lift-off length
  and the increase in lift-off length with
  decreasing oxygen concentration.

56
Conclusions 22004-01-1399

• Using a 50-µm orifice, a 2 diesel fuel, and a
  reduced ambient gas temperature (850 K), but with
  21 oxygen, it was shown that non-sooting,
  mixingcontrolled combustion occurred at the
  lift-off length in a fuel-air mixture with a
  cross-sectional average equivalence ratio of
  approximately 0.6suggesting that the
  quasi-steady combustion was fuel-lean and thereby
  avoided the formation of a diffusion flame. The
  average adiabatic flame temperature of the
  fuellean mixture was estimated as 2070 K. The
  same fuel-lean combustion was also realized using
  a 100-µm injector tip and an oxygenated (21 wt
  oxygen)

57
Conclusions 32004-01-1399

• Using a more conventional 180-µm orifice, an
  oxygenated fuel and very low ambient oxygen
  concentrations of 5-8, a fuel jet was
  non-sooting at an ambient temperature and density
  of 1000 K and 30.0 kg/m3. Soot formation was
  avoided even though the cross-sectional average
  equivalence ratio at the lift-off length exceeded
  a value of five. The lack of soot formation
  appears to be caused by jet mixture temperatures
  that were too cool for soot inception, even
  though the mixtures were fuel-rich, as has been
  suggested in a recent engine and detailed soot
  kinetics study 22. Adiabatic flame temperatures
  using 8 oxygen were only 1800 K. Despite the low
  combustion temperatures, the pressure rise
  measurements suggest that the combustion was
  complete.