14th Diesel Engine-Efficiency and Emissions Research (DEER) Conference

BenefitsBenefits andand drawbacksdrawbacks ofof compressioncompression ratioratio reductionreduction inin PCCIPCCI combustioncombustion applicationapplication inin anan advancedadvanced LDLD dieseldiesel engineengine

Carlo Beatrice and Chiara Guido Istituto Motori, Consiglio Nazionale delle Ricerche Naples , ITALY

Contacts: [email protected] [email protected]

DEER 2008, Dearborn MI, 4-7° August 2008 1 Future Targets and Technology Pathways for PC Diesel Engines

0.08

0.07 Tier 2, Bin 10 EURO 3 (2007-2010) (2000) 0.06 •Advanced combustions 0.05 (PCCI/LTC etc.);

0.04 •New CRS with Inj. Rate Shaping; •Cold LP EGR + HP EGR; EURO 4 PM [g/mi] PM [g/mi] 0.03 •Closed-Coupled DPF; (2005) T2, Bin 8 •VVA System; 0.02 EURO 5 •Closed-Loop Combustion Control; EURO 6 prop •SCR/LNT 0.01 T2, Bin 5 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 NOx [g/mi]

DEER 2008, Dearborn MI, 4-7° August 2008 2 PCCIPCCI (or(or LTC/HCCI)LTC/HCCI) potentialpotential forfor NOxNOx && PMPM reductionreduction Source M. Lisbona, FPT, ATA2006 In the transition from conventional combustion to PCCI mode the fuel/air mix change from orange to light-blue pathway

Source M. Potter, GM, DEER2006

Compression Ratio appears a key parameter to improve PCCI combustion

performance in partial load range

Source M. Potter, GM, DEER2006

DEER 2008, Dearborn MI, 4-7° August 2008 3 IM-CRF Research Project on CR influence on PCCI performance

Starting from an Eu4 Reference Engine (FIAT 1.9 MultiJet)

An advanced low NOx calibration was applied

(PCCI calib. with single shot inj. and high EGR/pressure inj.)

FIAT M741 MultiJet engine

Three CR values were selected for the tests Engine 4 cyl, 1.9L (16.5:1 → 15.5:1 → 14.5:1) Bore x Stroke 82 x 90 The 3 engine configurations were run in [mm] several test points typical of NEDC engine Stroke [mm] 90.0 operating area. Aim of the project was to Compr. Ratio 17.5 characterize the influence of CR reduction Turbo VGT on PCCI application in the largest NEDC area with NOx near Eu6 targets

DEER 2008, Dearborn MI, 4-7° August 2008 4 CRCR selectionselection andand PistonPiston DesignDesign

Starting from the commercial Eu4 piston geometry, 3 guidelines were followed: ¾Same air flow structure at TDC (same swirl and turbulence charac. vs CR); ¾Same squish height and internal lip diameter (similar squish flow in the bowl); ¾Same lip profile of the bowl (same structural robustness at rated power).

Parameter Formula CR 16.5 CR 15.5 CR 14.5 K-factor [%] Bowl volume/Comb. 75.2 76.8 78.4 chamber volume Reentrant Ratio [%] (Dext-Dint)/Dint 9.3 10.9 13.5

Aspect Ratio [%] Hmax/Dext 33.0 34.0 35.2 DEER 2008, Dearborn MI, 4-7° August 2008 5 InIn--bowlbowl flowflow characteristicscharacteristics vsvs CRCR

•The main air flow patterns appear very similar; RC 16.5 •The main differences are in the central part of the bowl. 5 RC 16.5 RC 15.5 RC 15.5 RC 14.5 4

RC 14.5 3

Swirl Ratio [a.u.] SR@ -180 c.a. [°] = 2.7

Slight increase of swirl ratio vs CR 2 decrease -50 0 50 crank angle [°] Friction loss reduction due to the surface/volume reduction and The CR influence on engine can be increment of moment of inertia ascribed mainly to different fuel DEER 2008, Dearborn MI, 4-7° August 2008 vaporization and thermodynamic effects 6 TestTest MethodologyMethodology

12 Starting from a prototype “near NOx Steady state test points with PCCI combustion Euro6” engine calibration for a CR=16.5 10

8 All CRs were tested in different operating points typical of NEDC area 6

4 Engine load - BMEP [bar] -BMEP load Engine In the engine speed range, the load was 2 increased until one of the following NEDC operating area for FIAT MultiJet Engine limits were reached: 0 0 1000 2000 3000 4000 •+5% of FC w resp to Eu4 calib.; Engine speed [rpm] •80 bar/ms of the dp/dtmax ; In this way all CRs can be •1.5 FSN of smoke as limit for DPF. compared with optimized

combustion timing and intake MBF50% and boost pressure were optimized airflow conditions vs CR varying SOI and VGT nozzle. The fuel is a commercial Low Sulfur

DEER 2008, Dearborn MI, 4-7° August 2008 respecting the EN 590 standards 7 CR influence on max PCCI torque: max torque, FC & smoke CR5 influence on max PCCI torque: max torque, FC & smoke FC & dp/dtmax limiting factors 4 Smoke limiting factor Reducing CR only small increments in

3 maximum engine torque are possible within the limits and without changing RC 16.5 2 the engine technology (Euro4) RC 15.5 RC 14.5

Engine load [bar] - BMEP 1

] 10 2.5 500 1000Constant 1500 2000engine 2500calibration 3000 3500 CR 16.5 0 Engine speed [rpm] CR 15.5 8 2 CR 14.5

Reducing CR the τid increases 6 1.5 lowering the smoke in the whole 4 1 range. [FSN] Smoke dp/dtmax limits the MBF50% that 2 0.5 controls the engine efficiency.

FC increment w respect to E4 level [% FC was similar for all CRs, no fuel 0 0 500 1000 1500 2000 2500 3000 3500 penalty was evidenced reducing CR Engine speed [rpm]

DEER 2008, Dearborn MI, 4-7° August 2008 8 CRCR influenceinfluence onon maxmax PCCIPCCI torque:torque: NOxNOx,, HCHC && COCO 4

RC 16.5 NOx emissions are mainly driven by 3 RC 15.5 EGR. RC 14.5 Under PCCI mode, the already very EURO 4 level 2 low NOx concentrations are poorly influenced by CR NOx [g/kWh] NOx 1

500 1000 1500 2000 2500 3000 3500 0 Engine speed [rpm] 25 12 RC 16.5 10 20 RC 15.5

RC 14.5 8 HC & CO emissions tend to 15 increase reducing CR. 6 10 HC raw [g/kWh] 4 [g/kWh] raw CO Premixing time and spray wall 5 interaction have effect on unburned 2

compound emissions 0 0 500 1000 1500 2000 2500 3000 3500 DEER 2008, Dearborn MI, 4-7° August 2008 Engine speed [rpm] 9 CRCR influenceinfluence onon maxmax PCCIPCCI torque:torque: Ind.Ind. engineengine cyclecycle 70 140 From CR 15.5 to 14.5 the τid increase 120 60 PCCI mode - 1800 rpm @ 4 bar BMEP was lower than from 16.5 to 15.5.

100 Apart that τid is not linear with temp., 50 . possible cause is also the fall of τid in 80 40 the NTC range (1050÷900K) in some

30 60 conditions.

CR 16.5 25

Cylinder pressure[bar] 40 20 CR 15.5 PCCI mode - 1800 rpm @ 4 bar BMEP

CR 14.5 [A] current energizing Injector 20 10 20

-20-100 10203040 0 C.A.[°] 0 15 CR 16.5 CR 15.5 10 CR 14.5

ROHR curves confirm that small °] [%/c.a. ROHR changes were detected varying CR 5

in terms of “thermodynamic” 0 combustion evolution

-5 -10 0 10 20 30 40 C.A.[°]

DEER 2008, Dearborn MI, 4-7° August 2008 10 EngineEngine OptimizationOptimization vsvs CRCR

] 10 2.5 CR 16.5 EGR has to be kept quite constant in CR 15.5 order to have same NOx. 8 2 CR 14.5 Two ECU parameters affect in opposite way both engine efficiency 6 1.5 and smoke: 4 1 • Rail pressure (friction loss / Smoke [FSN] Smoke atomization);

2 0.5 • Swirl valve position (pumping loss,

FC increment w respect to E4 level [% intake valve permeability and heat loss 0 0 500 1000 1500 2000 2500 3000 3500 / fuel premixing level); Engine speed [rpm] These parameters can be optimized in order to exploit the very low smoke of the lowest CRs. Optimization work was done respecting the limits on max Injection strategy optimization was not pressure rise, exhaust smoke considered in this project but it is and max FC increment another key factor

DEER 2008, Dearborn MI, 4-7° August 2008 11 140 EngineEngine70 OptimizationOptimization vsvs CRCR:: Ind.Ind. engineengine cyclecycle 120 60 PCCI mode - 2200 rpm @ 4.5 bar BMEP 100 50 Reducing Rail Pressure: . 80 • Less parasitic and friction losses; 40 • Less premixed fuel→ Lower 60 30 dp/dtmax→ Early SOImain→ MBF50%

40 closer to TDC; Cylinder pressure [bar] 20 reference calibration optimized calibration Injector energizing current [A] • Less HCs&CO → Higher ηcomb 20 10 -20-100 10203040 C.A.[°] Particularly important at high speed and 0 0 25 medium load (late combustion)

20 PCCI mode - 2200 rpm @ 4.5 bar BMEP Reducing SW ratio:

reference calibration • Less pumping losses; 15 optimized calibration • Less premixed fuel→ etc…..; Diffusive 10 combustion • Higher valve permeability→Higher + Soot trapped gas mass ; ROHR [%/c.a. °] [%/c.a. ROHR 5 • Less heat losses;

0 • Less overleaning A/F zones→Less -10 0 10 20 30 40 HCs & CO -5 C.A.[°]

DEER 2008, Dearborn MI, 4-7° August 2008 12 EngineEngine OptimizationOptimization vsvs CRCR:: FCFC && EmissionsEmissions

20 Engine speed 2200 rpm Engine speed 2200 rpm ] 1.5 6 bar of BMEP 4 bar of BMEP 4.5 bar of BMEP 6 bar of BMEP 4 bar of BMEP 4.5 bar of BMEP 16

1 12

8 Smoke [FSN] Smoke 0.5

4

BSFC increment w respect to E4 standard [% E4 standard to wrespect BSFCincrement 0 reference optimized reference optimized reference optimized 0 calibration calibration calibration reference optimized reference optimized reference optimized calibration calibration calibration Engine speed 2200 rpm Engine speed 2200 rpm 2 12 4 bar of BMEP 4.5 bar of BMEP 6 bar of BMEP 4 bar of BMEP 4.5 bar of BMEP 6 bar of BMEP 1.6 Increase of Stoichiometric Zones 10 ] 8 1.2 6 0.8 4 NOx emission [g/kWh NOx HC+CO emission [g/kWh] emission HC+CO 0.4 2

0 0 reference optimized reference optimized reference optimized reference optimized reference optimized reference optimized calibration calibration calibration calibration calibration calibration FC reduction at same dp/dtmax permits the increment of Torque limit DEER 2008, Dearborn MI, 4-7° August 2008 13 24 EngineEngine70 OptimizationOptimization vsvs CR:CR: ConditionsConditions @@ 30003000 rpmrpm 60 PCCI mode - 3000 rpm @ 3.8 bar BMEP 18 50 At high speed the available premixing 40 12 time is low also at the lowest CR. 30 ROHR [%/c.a.°] ROHR reference calibration

Cylinder pressure [bar] Smoke becomes the limiting factor. 20 6 optimized calibration

10 -20 -10 0 10 20 30 40 C.A.[°] The effectiveness of both Injection 0 0 pressure and Swirl ratio reduction on

1.5 FC improvement is strongly lowered. Engine speed 3000 rpm @ 3.8 bar of BMEP Smoke emissions are strongly 1 sensible to SW and Rail Pressure reduction. Smoke [FSN] Smoke 0.5 However this engine speed level is less critical on NEDC engine performance. 0 reference calibration optimized

DEER 2008, Dearborn MI, 4-7° August 2008 14 EngineEngine OptimizationOptimization vsvs CRCR:: MaxMax TorqueTorque vsvs SpeedSpeed

7

6 After engine optimization the maximum PCCI torque was increased 5 from 4 to 6 bar of BMEP with CR 14.5. 4

3 For CR 16.5 the improvement was

Engine load [bar] negligible due to smoke emission 2 R.C. 16.5 R.C. 15.5 close to the limit. 1 R.C. 14.5 0 CR 15.5 and 14.5 showed similar 500 1000 1500 2000 2500 3000 3500 results in the low speed range, while Engine speed [rpm] CR 14.5 was more effective than CR 15.5 in the high speed range.

At highest engine speed, the CR reduction is less effective for PCCI improvement due to smoke limitation.

DEER 2008, Dearborn MI, 4-7° August 2008 15 ConclusionsConclusions

• FC penalty vs CR, in NEDC operating area, was small and mainly dependent on ηcomb reduction;

• Lowering the CR a marked drop in smoke and a drastic rise in HC and CO have to be expected. NOx remain EGR driven;

• CR reduction is able to increase the max PCCI torque only through engine re- optimization;

• Smoke drop can be exploited to improve the torque. Working only on the SW and Prail reduction, pumping and friction loss can be lowered and intake duct permeability increased. HCs&CO can be also lowered;

• In PCCI mode, combustion phase optimiz ation and friction losses reduction do not cause the FC deterioration for low CR engines. NOx emission can be reduced well beyond the Eu5 and near Eu6 targets. From our results, a CR value around 15.5 appears the best compromise; • HCs&CO remain the main issue lowering CR. They cannot be controlled by DOC during cold start. Startability can also becomes a problem.

DEER 2008, Dearborn MI, 4-7° August 2008 16 GeneralGeneral OutlineOutline

• It was pointed out that, independently from the DeNOx adoption, in order to drop the NOx emissions for Euro6/Tier2 Bin5 targets, the CR optimization (or the use of VCR) appears one of the key factors for PCCI application improvement. • For a complete assessment of the CR reduction benefits, the preservation or the improvement of the current rated power target and the control of the high

HCs&CO emissions during cold start remain open questions. Engine technology improvement could give adequate solutions for these problems

(CLC-Comb., LP-EGR, VVA, new FIE, VCR, new TC, new DOC etc.) AcknowledgmentsAcknowledgments Special thanks to Dr. Roberto Imarisio and Dr. Marco Tonetti of Centro Ricerche Fiat, for their support in the providing scientific cooperation, information and hardware.

DEER 2008, Dearborn MI, 4-7° August 2008 17 IstitutoIstituto MotoriMotori ofof NationalNational ResearchResearch CouncilCouncil,, ITALYITALY

THANK YOU for the ATTENTION

Istituto Motori website: www.im.cnr.it

DEER 2008, Dearborn MI, 4-7° August 2008 18