Leaner Lifted-Flame Combustion Enabled by the Use of an Oxygenated Fuel or a Novel Mixing-Enhancement Technique

Transcription

1 Leaner Lifted-Flame Combustion Enabled by the Use of an Oxygenated Fuel or a Novel Mixing-Enhancement Technique Ryan K. Gehmlich and Charles J. Mueller Research Supported by US Dept. of Energy, Office of Vehicle Technologies, Program Manager Kevin Stork; and Ford Motor Company, Program Manager Eric Kurtz 2 nd CRC Advanced Fuels and Engine Efficiency Workshop Sandia National Laboratories, Livermore, CA November 3, 2016

2 2 Presentation Outline Engines 101 & Leaner Lifted-Flame Combustion (LLFC) Concept Globally premixed vs. mixing-controlled combustion strategies Optical Engine Experiments Fuels tested Engine specifications and schematic Diagnostics Path to LLFC (2-hole injector tip experiments) Injection pressure and IMT effects Sustaining LLFC at higher loads (6-hole injector tip experiments) Oxygenated fuel effects LLFC Enabled by In-Cylinder Mixing Enhancement Technique Ducted fuel injection (DFI) concept Constant volume combustion vessel experiments Key advantages of DFI Summary

3 3 Engines 101: Combustion Strategies Globally Premixed Mixing-Controlled Combustible mixture exists throughout (most of) the combustion chamber Combustible mixture exists only at the lift-off length and downstream

4 Leaner Lifted-Flame Combustion: Research Objective 4 Determine whether moderate-load LLFC can be achieved and sustained through the use of: Oxygenated fuel Enhanced fuel/charge-gas mixing upstream of the liftoff length Small injector orifices (~110 µm) High injection pressures (~240 MPa) Lower charge-gas temperatures / higher charge-gas densities Novel in-cylinder mixing geometries Goal: Maintain (H) 2 during the combustion event

5 5 Fuels Tested CFA = #2 ULSD emission-certification fuel T50 = 50 vol% TPGME + balance CFA TPGME = tri-propylene glycol mono-methyl ether CFA TPGME (C 10 H 22 O 4 ) T50 (TPGME+CFA) v oxy [%] f [%] Y C [%] Y H [%] Y O [%] [kg/m 3 ] CN [-] q LHV [MJ/kg]

6 6 Optical Engine Specifications and Schematic Research engine Cycle Single-cylinder 4-stroke CIDI Valves per cylinder 4 Bore Stroke Displacement per cyl. Conn. rod length Conn. rod offset Piston bowl diameter Piston bowl depth Squish height 125 mm 140 mm 1.72 liters 225 mm None 90 mm 16.4 mm 1.5 mm Swirl ratio 0.59 Compression ratio 12.5:1 Simulated compr. ratio 16.0:1 Engine is skip-fired: 1 fired + 14 motored N 2 and CO 2 added to intake air to achieve X O2 target and same T, p, c p at -10 CAD ATDC

7 7 Diagnostics Fuel injection: Injected quantity yields fuel conversion efficiency; also measure injection rate (per orifice) Cylinder pressure: Apparent heatrelease rate (AHRR), T bulk, T core Two-camera, simultaneous highspeed imaging Natural luminosity (NL): spatial distribution of hot soot Chemiluminescence (CL): mean liftoff length (H) and ( H ) Emissions HC via flame ionization detector CO and CO 2 via non-dispersive infra-red absorption, O 2 via paramagnetic NO x via heated chemiluminescence detector Smoke via smoke meter; soot vol. fraction via laser-induced incandescence in exhaust stream HC, CO, and soot emissions used to determine combustion efficiency

8 8 Presentation Outline Engines 101 & Leaner Lifted-Flame Combustion (LLFC) Concept Globally premixed vs. mixing-controlled combustion strategies Optical Engine Experiments Fuels tested Engine specifications and schematic Diagnostics Path to LLFC (2-hole injector tip experiments) Fuel, injection pressure and IMT effects Sustaining LLFC at higher loads (6-hole injector tip experiments) Oxygenated fuel effects LLFC Enabled by In-Cylinder Mixing Enhancement Technique Ducted fuel injection (DFI) concept Constant volume combustion vessel experiments Key advantages of DFI Summary

9 The Path to LLFC: Fuel, Injection Pressure, and IMT Effects Using a 2-Hole Injector Tip 9 LLFC was previously demonstrated using neat oxygenates While this was a neat result, this is impractical at a large scale Attempted to achieve LLFC with diesel fuel (2-hole, 6-hole, 10-hole injector tips). Approached LLFC with 2-hole tip only. Changed directions toward: Oxygenated fuel mixture effects Fuel injection pressure effects Intake manifold temperature effects A combination of these parameters led to achieving and sustaining LLFC with a 2-hole tip! ( H ) 2.0 and SINL 0 LLFC! LLFC has been achieved and sustained!

10 The Path to LLFC: Fuel, Injection Pressure and IMT Effects Using a 2-Hole Injector Tip 10 Natural luminosity (NL) Chemiluminescence (CL) 80-MPa P inj & 95 C IMT 240-MPa P inj & 50 C IMT: Have high-t chemistry (CL), but not hot soot (NL)

11 11 Next Steps: 6-Hole Injector Tip, Higher Loads Next step is to try a more realistic 6-hole injector, to reach higher loads Turn all the dials to their limits: Fuel: Highest reasonable oxygen content T50 Lower intake manifold temperature (to 30 C) Retard the timing (up to +5 CAD ATDC SOC) Lower coolant temperature (to 85 C) How high of a load can we achieve while sustaining LLFC?

12 Load, Injection, and Operating Parameters 12 Speed Injector 1500 rpm Solenoid Common Rail Injector Tip 6 x mm x 140 Injection Schedule Actual Duration of Injection (DOI a ) Injection Pressure (P inj ) Gross Ind. Mean Eff. Pressure Intake Manifold Abs. Pressure Single Injection Near TDC 2.4 ms (T50), ms (CFA) 240 MPa bar 2.50 bar abs. Coolant Temperature 85 C Intake Manifold Temp. (IMT) 30 C Fuel T50 CFA Start of Comb. (CAD ATDC) Intake-O 2 Mole Fraction (X O2 ) 21% 16% T50 data taken at constant injection duration (2.4 ms) CFA data varied injection duration to match load for each corresponding T50 condition.

13 Start of Combustion: +5 CAD ATDC, Load ~6 bar gimep LLFC Achieved and Sustained at Retarded Injection Timing 13 Apparent heatrelease rate, AHRR Mean lift-off length, H Spatially integrated natural luminosity, SINL, a measure of hot incylinder soot Equivalence ratio at the mean lift-off length, ( H ) T50 in LLFC!

14 Start of Combustion: +5 CAD ATDC LLFC Achieved and Sustained at Retarded Injection Timing 14 CFA 13.0 o Cycle 4 G rel = s n = s 13.0 o Cycle 4 n = 29 AHRR [J/CAD] (blue) Crank Angle [CAD] (H) [-] (black) T o Cycle 2 G rel = s n = s 18.0 o Cycle 2 n = 29 AHRR [J/CAD] (blue) Crank Angle [CAD] (H) [-] (black)

15 15 Soot Emissions LLFC regime 240-MPa P inj, All SOC Timings, 30 C IMT T50 Relative to CFA Smoke is lower for T50, but all cases are well below 2010 targets (~0.1 FSN) Also may need to consider particle #, even with zero filter smoke need more sensitive measurements of soot emissions SINL is 70-90% lower for T50

16 16 NO x, HC, and CO Emissions ISNOx is 25-30% lower for T50 relative to CFA ISHC is up to 25% lower for T50 at +5 ATDC SOC ISCO is generally higher for T50 ISNO x is 25-30% lower 240-MPa P inj, All SOC Timings, 30 C IMT

17 17 Combustion Phasing and Efficiency T50 ign. delay is considerably shorter compared to CFA T50 phasing is later relative to CFA due to longer injection events for equivalent load η c is slightly higher for a given timing and X O2, but η f,ig is slightly lower (~1%) than that of CFA, probably due to less optimal combustion phasing Timing retard required for LLFC was moderate; not a huge efficiency hit 240-MPa P inj, All SOC Timings, 30 C IMT

18 18 Combustion Noise and Ringing Intensity RI < 5 ACEC noise 1500 RPM ~88 dba T50 relative to CFA under matched load conditions: T50 noise reduced by ~4-6 dba Ringing intensity reduced by 60-80% Noise and RI calculated using techniques outlined in the ACEC Tech Team approved Light-Duty Noise Guideline for Advanced Combustion Research 240-MPa P inj, All SOC Timings, 30 C IMT

19 19 Presentation Outline Engines 101 & Leaner Lifted-Flame Combustion (LLFC) Concept Globally premixed vs. mixing-controlled combustion strategies Optical Engine Experiments Fuels tested Engine specifications and schematic Diagnostics Path to LLFC (2-hole injector tip experiments) Fuel, injection pressure and IMT effects Sustaining LLFC at higher loads (6-hole injector tip experiments) Oxygenated fuel effects LLFC Enabled by In-Cylinder Mixing Enhancement Technique Ducted fuel injection (DFI) concept Constant volume combustion vessel experiments Key advantages of DFI Summary

20 Next Steps: Aiming for LLFC Using an In-Cylinder Mixing Enhancement Technique 20 Further increases in fuel-injection pressure and fuel oxygenation are likely insufficient to provide adequate mixing to sustain LLFC at high loads a new approach is necessary. Ducted fuel injection (DFI): Inject fuel through a tube downstream of the orifice exit to increase the velocity gradients that drive turbulent mixing of fuel and charge-gas Could enable the use of less-costly: Fuels (e.g. higher aromatic contents) Fuel-injection equipment (e.g. lower injection pressures) Aftertreatment systems Is it possible to sustain φ(h) < 2 without fuel oxygenation? At higher loads?

21 21 Multiple simultaneous high speed cameras measure time-resolved: Natural luminosity (NL) OH* chemiluminescence (CL) Diffuse back-illumination (DBI) Ignition delay (ID) Optical Diagnostic Setup

22 Fuel: n-dodecane 22

23 23 Key Advantages to the DFI Concept DFI could enable LLFC at higher loads and without changes to fuel Concept is simple and effective Minimal manufacturing costs Tolerant to EGR & fuel variability

24 24 Summary Oxygenates are helpful for facilitating leaner lifted-flame combustion LLFC was sustained with a 6-hole tip and T50 fuel through a combination of: High injection pressure (240 MPa) Low intake manifold temperature (30 C) Slightly retarded timing (+5 CAD ATDC SOC) Targets achieved in LLFC: gimep > 6 bar Zero engine-out smoke Relative to CFA: 25-30% lower ISNO x Noise reduced by 4-6 dba, RI reduced by 60-80% Other emissions relatively unchanged regardless of dilution LLFC is a mixing-controlled, EGR-tolerant, soot-free combustion mode Ducted fuel injection is a promising potential path for sustaining LLFC at higher loads

25 25 Acknowledgements Scott Skeen, Julien Manin, and Lyle Pickett For assistance with setting up diagnostics and running combustion-vessel experiments Sam Fairbanks For assistance with laboratory/engine hardware and running experiments Christopher Nilsen and Dan Ruth Data acquisition and analysis Eric Kurtz and Chris Polonowski (Ford) For helpful technical discussions Bill Cannella (Chevron) For providing the #2 ULSD certification fuel (CFA) Gary Hubbard, Chris Carlen, and Keith Penney For assistance with data-acquisition software and electronics

26 Thank you for your attention! 26

27 Additional Info 27

28 Start of Combustion: -5 CAD ATDC LLFC Approached at Advanced Injection Timing 28 Apparent heatrelease rate, AHRR Mean lift-off length, H Spatially integrated natural luminosity, SINL, a measure of hot incylinder soot Equivalence ratio at the mean lift-off length, ( H ) No sustained LLFC

29 Start of Combustion: -5 CAD ATDC Movies: LLFC Approached at Advanced Ignition Timing 29 CFA 5.0 o Cycle 2 G rel = s n = s 5.0 o Cycle 2 n = 25 AHRR [J/CAD] (blue) Crank Angle [CAD] (H) [-] (black) T o Cycle 2 G rel = s n = s 11.0 o Cycle 2 n = 35 AHRR [J/CAD] (blue) Crank Angle [CAD] (H) [-] (black)

30 OH* Chemiluminescence Camera Phantom intensified v mm f/4.5 UV Nikkor lens 45 µs camera exposure 20,000 fps, 400 x 224 resolution Intensifier gate: 17 µs for 21% O 2, 46 µs for 15% O ±10 nm bandpass filter, long-λ block Quantified Lift-off length Natural Luminosity (Soot) Camera Photron SA-X2 50 mm f/1.2 lens 50,000 fps, 512 x 192 resolution 600 nm short-wave pass filter and heat absorption (KG3) filter were used to suppress high intensity soot luminosity. Shutter time fixed at 5.0 µs f/stop varied from f/1.2 to f/8 depending on O 2 mass fraction and temperature. Ignition Delay (ID) Camera Photron SA-X2 105 mm f/1.8 lens 50,000 fps, 512 x 432 resolution No filters for maximum sensitivity to first high temperature reactions µs shutter time Diffused Back-Illumination (DBI) Camera Photron SA-X2 50 mm f/1.2 lens with 500D close-up lens 100,000 fps, alternating light/dark frames, 640 x 168 resolution 400±10 nm filter (to attenuate broadband luminosity from soot) 1 µs exposure time to measure brightest part of a 1.5 µs LED pulse duration. Sensitive extinction diagnostic allows quantitative 2-D measurements of soot concentration