Assessment of Innovative Bowl Geometries over Different Swirl Ratios/EGR rates

Assessment of Innovative Bowl Geometries over Different Swirl Ratios/EGR rates Andrea Bianco 1, Federico Millo 2, Andrea Piano 2, Francesco Sapio 2 1: POWERTECH Engineering S.r.l., Turin ITALY 2: Politecnico di Torino, Turin ITALY CONVERGE USER CONFERENCE EUROPE 2018 March 19 23, 2018 Bologna, Italy

Agenda Introduction Baseline engine model and validation Innovative bowl geometries Assessment on different swirl ratios Assessment on different EGR rates Conclusions Page 2

Agenda Introduction Baseline engine model and validation Innovative bowl geometries Assessment on different swirl ratios Assessment on different EGR rates Conclusions Page 3

Sales weighted average NEDC type approval CO 2 emission values [g/km] Direct manufacturing cost Introduction Average CO 2 emission CO 2 reduction cost Source: ICCT, 2017 Source: ICCT, 2017 EU Target 2021 (95) Market share Diesel engines help in reducing CO 2 emissions, thanks to their lower BSFC compared to gasoline engines Fuel consumption reduction [%] Future Diesel engines are going to cost even more because of CO 2 emission targets Page 4

Cost [$/vehicle] Introduction Emissions limits Emissions reduction cost 1600 1400 1200 1000 Engine-out emission control Aftertreatment systems R&D Source: ICCT, 2012 800 600 400 200 0 EURO IV EURO V EURO VI and new pollutant emissions limits Is it possible to further reduce in-cylinder Diesel emissions? Aftertreatment plays a major role in cost increase Innovative bowl design Page 5

Agenda Introduction Baseline engine model and validation Innovative bowl geometries Assessment on different swirl ratios Assessment on different EGR rates Conclusions Page 6

Baseline engine model and validation CI engine appl. LD vehicle Cylinders # 4 Displacement 1.6L Compression ratio 16:1 Turbocharger Single-stage with VGT WP1 WP2 WP3 Fuel injection system Maximum power and torque Common Rail 100 kw @ 4000rpm 320 Nm @ 2000rpm Baseline: re-entrant bowl design Two CI engine modelling pillars: Accurate spray modeling Spray calibration Accurate combustion modeling: - SAGE w/ detailed chemistry scheme Page 7

Baseline engine model and validation Combustion simulation Converge 2.3.17 - sector mesh Base grid x = 0.5 mm x min = 0.25 mm (AMR, embedding) RANS: K-eps RNG turbulence model O Rourke & Amsden heat transfer model Calibrated spray models (spray bombs) w/ additional settings: - Injection rate from detailed GT-SUITE injector model - O Rourke wall film model (default coeff.) 3D Full-cyl. simulation Mapping @IVC Combustion analysis Combustion modelling: SAGE - Skeletal Zeuch* mechanism - Particulate Mimic soot model (default coeff.) - Imposed LHV (43250 kj/kg) - Calibrated Schmidt number Fuel Species N-Heptane Species 121 Reactions 593 PAH YES (A3R5-) NOx Embedded in mech.dat *Zeuch, T., Moréac, G., Ahmed, S. S., Mauss, F., "A comprehensive skeletal mechanism for the oxidation of n-heptane generated by chemistry-guided reduction", Combustion and Flame, 155(4), 651 674, 2008. Page 8

Baseline engine model and validation Validation results Page 9

Baseline engine model and validation Validation results Very good agreement for NOx emissions Experimental FSN / In-cyl SOOT@120CA atdcf Qualitative agreement between experimental Filter Smoke Number (FSN) and 3D-CFD in-cylinder SOOT mass Page 10

Agenda Introduction Baseline engine model and validation Innovative bowl geometries Assessment on different swirl ratios Assessment on different EGR rates Conclusions Page 11

Innovative bowl geometries 1) Twin Vortex (TV) bowl design Twin Vortex concept was developed by Ricardo as a "low-no x " solution for HD diesel engine. Requirements [1]: Fuel injection pressure EGR rate SOI Adjusted for engine efficiency engine re-calibration [2-3] Suitable for LD? Is it ok with minimal engine re-calibration? TV bowl proposal (bore, CR, squish height = baseline) [1] https://ricardo.com/news-and-media/press-releases/ricardo-and-doosan-reveal-advanced-low-emissions-t (28/09/2017) [2] https://www.dieselnet.com/tech/engine_combustion.php (28/09/2017) [3] http://www.sae-na.it/images/download/ws2016/09%20-%20andrea%20trevisan%20_%20sae-na%20%20aftertreatment%20system%20for%20diesel%20engines%2027_28%20june%2716.pdf (28/09/2017) Page 12

Innovative bowl geometries 2) Radial Bumps (RB) bowl design Volvo developed a radial bump bowl named "WAVE" for HD applications. A 2% fuel economy improvement and SOOT reduction between 50-90% are claimed ([5], [6]). [4] Requirements [7]: Fuel injection pressure Swirl ratio engine re-calibration N bumps = N spray plumes Would radial bumps work on a LD engine? [4] https://www.dieselnet.com/tech/engine_combustion.php (01/10/2017) [5] https://www.chalmers.se/en/centres/cerc/documents/annual%20report%202016.pdf (04/10/2017) [6] https://energy.gov/sites/prod/files/2016/06/f32/ace060_amar_2016_o_web.pdf (04/10/2017) [7] https://www.google.com/patents/us8499735 RB bowl proposal (bore, CR, squish height = baseline) Page 13

Agenda Introduction Baseline engine model and validation Innovative bowl geometries Assessment on different swirl ratios Assessment on different EGR rates Conclusions Page 14

Innovative bowls: Assessment on different swirl ratios Simulation procedure For each different bowl design: Full-cylinder cold flow sim. Sector mesh combustion sim. SR sweep analysis (2 swirl ratio levels) for WP1 and WP2 Simulation w/ nominal SR for WP3 (verify rated power perf.) WP1 WP2 WP3 Assumptions: Constant volumetric efficiency, SR varied through mapping Baseline EGR% Constant injected fuel mass (same inj. rate, etc.) Constant FMEP Page 15

Innovative bowls: Assessment on different swirl ratios Simulation results WP1 combustion SR2 = 0.5 SR1 SR1: - TV bowl shows lower SOOT and slightly higher NOX than baseline SR2: - Baseline bowl shows slightly worse NO X and SOOT emissions - TV bowl shows worse SOOT emissions Page 16

Innovative bowls: Assessment on different swirl ratios Simulation results WP1 combustion SR2 = 0.5 SR1 SR1: - RB bowl shows lower SOOT and highest NO X (EGR?) SR2: - RB bowl shows worse SOOT emissions and higher NO X than baseline SR1 seems acceptable for both TV and RB bowls Page 17

Innovative bowls: Assessment on different swirl ratios Simulation results WP1 combustion SR2 = 0.5 SR1 TV bowl shows BSFC aligned with baseline RB bowl shows lower BSFC than baseline (-3%) Page 18

Agenda Introduction Baseline engine model and validation Innovative bowl geometries Assessment on different swirl ratios Assessment on different EGR rates Conclusions Page 19

Base Base Innovative bowls: Assessment on different EGR rates Simulation procedure EGR Sweep (3 EGR levels) carried out on WP1 and WP2 Assumptions: SR = SR1 Constant volumetric efficiency, EGR varied through mapping Constant BMEP (± 1%) by changing the main injection fuel mass (ET) FMEP evaluated by means of calibrated 1D model to account for the maximum pressure differences WP1 WP2 5% WP1 WP2 Page 20

Innovative bowls: Assessment on different EGR rates Simulation results WP1 combustion + EGR + EGR TV bowl shows improved trade-off compared to baseline bowl RB bowl shows the best trade-off. EGR could be further increased Page 21

Innovative bowls: Assessment on different EGR rates Simulation results WP1 combustion Baseline - base EGR RB high EGR Page 22

Innovative bowls: Assessment on different EGR rates Simulation results WP1 combustion Baseline base EGR RB high EGR Baseline base EGR RB high EGR +20deg atdcf +40deg atdcf Page 23

Agenda Introduction Baseline engine model and validation Innovative bowl geometries Assessment on different swirl ratios Assessment on different EGR rates Conclusions Page 24

Conclusions Conclusions Both TV and RB bowls shows potential improvement for pollutant emissions and fuel consumption, on a LD engine, w/o significant engine re-calibration RB bowl shows the best trade-off (-40% SOOT and -2.5% BSFC w/ baseline NO x emission). Trend is confirmed on other engine WPs 3D-CFD predictive combustion simulations on innovative bowl designs could help in further reducing Diesel engine in-cylinder pollutant emissions and fuel consumption Future work Optimization of bowl designs Experimental tests on selected configurations Page 25

Conclusions Special thanks to team for their support Page 26