Question? Leave a message!




Optimization and Low Temperature Combustion

Optimization and Low Temperature Combustion 20
OscarTucker Profile Pic
OscarTucker,Singapore,Professional
Published Date:19-07-2017
Website URL
Comment
Part 8: Optimization and Low Temperature Combustion Reciprocating Internal Combustion Engines Prof. Rolf D. Reitz Engine Research Center University of Wisconsin-Madison 2014 Princeton-CEFRC Summer School on Combustion Course Length: 15 hrs (Mon.- Fri., June 23 – 27, 2014) Copyright ©2014 by Rolf D. Reitz. This material is not to be sold, reproduced or distributed without prior written permission of the owner, Rolf D. Reitz. 1 CEFRC4-8, 2014 Part 8: Optimization and Low Temperature Combustion Short course outine: Engine fundamentals and performance metrics, computer modeling supported by in-depth understanding of fundamental engine processes and detailed experiments in engine design optimization. Day 1 (Engine fundamentals) Part 1: IC Engine Review, 0, 1 and 3-D modeling Part 2: Turbochargers, Engine Performance Metrics Day 2 (Combustion Modeling) Part 3: Chemical Kinetics, HCCI & SI Combustion Part 4: Heat transfer, NOx and Soot Emissions Day 3 (Spray Modeling) Part 5: Atomization, Drop Breakup/Coalescence Part 6: Drop Drag/Wall Impinge/Vaporization/Sprays Day 4 (Engine Optimization) Part 7: Diesel combustion and SI knock modeling Part 8: Optimization and Low Temperature Combustion Day 5 (Applications and the Future) Part 9: Fuels, After-treatment and Controls Part 10: Vehicle Applications, Future of IC Engines 2 CEFRC4-8, 2014 Part 8: Optimization and Low Temperature Combustion Shi, 2011 Overview of optimization techniques Enumerative or exhaustive Calculus or gradient-based “local” methods which search in the neighborhood of current design point Random “global” methods such as genetic algorithms (GA) which typically converge on a global optimum Univariate (one-factor-at-a-time) Design of Experiments (DOE) Two-level factorial designs (main and interaction effects) Response surface methods (RSM) Statistical model building 3 CEFRC4-8, 2014 Part 8: Optimization and Low Temperature Combustion Senecal, 2000 Genetic algorithms “Individuals” are generated through random selection and a “population” is produced A model is used to evaluate the fitness of each individual The fittest individuals are allowed to “reproduce” A new “generation” is formed - “mutations” are allowed through random changes The fitness criteria thins out the population and the most fit solution is achieved over successive generations 4 CEFRC4-8, 2014 Part 8: Optimization and Low Temperature Combustion Goldberg, 1989 Carrol, 1996 Implementation of algorithm Senecal, 2000 Binary representation of parameters X - “genes” X X X 1 2 3 10101101 01101 01001001 - “chromosome” gene string  XX i,max i,min  Precision  2 1 Evaluate merit f (X) for each generation member - identify “fittest” Binary tournament selection Bit-swapping “Cross-over” Parent 1 Parent 2 01100100 1110110101011 10101101 0110101001001 Bit-flipping 10101101 1110110101011 Random “mutation” 01100100 0110101001001 Descendants 5 CEFRC4-8, 2014 swirl ratio NOx g/kgf Part 8: Optimization and Low Temperature Combustion Liu, 2006 Coello, 2001 Optimization methodology Multi-Objective Genetic Algorithm Nonparametric Regression Technique 0.3 All Citizens Produced 208 0.3 Pareto Citizens 0.28 204 0.26 0.25 0.24 200 0.22 0.2 196 0.2 192 0.15 1.0 0.18 0.5 0.1 0.9 0.8 1.5 0.2 0.8 2.5 0.3 0.6 0.16 0.4 3.5 0.7 4.5 0.5 0.6 0.4 0.6 0.2 0.7 Regression technique suitable for handling Simultaneous optimization of many irregular and undesigned data sets objectives 1 (e.g., GA data) 2 No merit function required to drive search Utilizes otherwise discarded optimization Pareto front offers more information than data a single optimum Captures magnitude of effects AND the shape of their response 6 CEFRC4-8, 2014 bowl diameter Soot g/kgf NOx g/kgf GISFC g/kW-hrPart 8: Optimization and Low Temperature Combustion Genzale, 2007 Example optimization - piston bowl design Parameters and Objectives Optimize: 7 Geometry Parameters: NOx Soot Pip height ISFC Bowl diameter f of bowl bottom 4 curvature control points Injector Spray Angle Swirl Ratio 7 CEFRC4-8, 2014 NOx g/kgf Part 8: Optimization and Low Temperature Combustion Genzale, 2007 Pareto front designs All Citizens Bowl geometry or injection 300 Pareto Citizens targeting trends? 280 NOx ↓68% 260 Soot ↑77% swirl = 0.7 GISFC ↑15% 240 220 NOx ↓57% 2.0 200 1.6 Soot ↑6% swirl = 1.4 1.2 GISFC ↓0% 180 0.8 0.1 0.2 0.4 0.3 0.4 0.0 0.5 ↓45% NOx ↓30% Soot swirl = 3.1 GISFC ↓2% NOx ↓5% Soot ↓42% swirl = 3.1 GISFC ↓6% 8 CEFRC4-8, 2014 Soot /kgf GISFC g/kW-hrbowl diameter (% bore) spray angle spray angle Part 8: Optimization and Low Temperature Combustion Genzale, 2007 Regression – Identify dominant design parameters Regression fits performed for each design on the Pareto front • 3 dominant design parameters identified: 1. Spray angle 2. Swirl ratio 3. Bowl diameter 0.8 1.4 2 1.2 0.6 1.5 1 1 0.4 0.8 0.5 4.5 0.2 0.6 0 0.6 0.6 3.5 0.5 50 50 0.65 55 55 1.5 60 0.7 0.4 2.5 60 65 0.75 65 2.5 70 70 1.5 0.8 0.2 75 75 3.5 0.85 80 80 0.5 0.9 4.5 85 0 85 9 CEFRC4-8, 2014 pip height (% bowl depth) swirl ratio swirl ratio soot g/kgf soot g/kgf soot g/kgfbowl diameter (% bore) spray angle Part 8: Optimization and Low Temperature Combustion Genzale, 2007 Regression – Understand Parameter Effects 0.8 2 0.6 1.5 swirl = 3.1 1 0.4 0.5 NOx ↓45% 4.5 0.2 0 0.6 Soot ↓30% 0.5 3.5 50 0.65 55 1.5 0.7 2.5 60 ↓2% GISFC 0.75 2.5 65 70 1.5 0.8 3.5 75 0.85 80 0.5 0.9 4.5 85 Response Surface Observations: An optimal spray angle is predicted. Increased swirl ratio is predicted to enhance soot reduction near the optimal spray angle. Increases soot emissions at narrow spray angles. Increased swirl ratio is predicted to decrease soot at all bowl diameters. 10 CEFRC4-8, 2014 swirl ratio swirl ratio soot g/kgf soot g/kgfPart 8: Optimization and Low Temperature Combustion Klingbeil, 2003 Optimization of LTC - low temperature combustion Increased interest in advanced combustion regimes RCCI, HCCI, PCCI, MK - offer simultaneous reduction of NOx and soot Challenges High CO, HC High loads Transients NOx EGR Soot 11 CEFRC4-8, 2014 Part 8: Optimization and Low Temperature Combustion Kokjohn, 2009 Combustion optimization - fuel and EGR selection HCCI simulations used to choose optimal EGR rate and PRF 100 (isooctane/n-heptane) blend 1 6 91 b b b a aa r IMEP r IMEP r IMEP MISFIRE Net ISFC 100 100 100 90 28 bar/deg At 6, 9, and 11 bar IMEP MISFIRE g/kW-hr MISFIRE 80 230 16 160 1300 rev/min 250 24 bar/deg 80 80 80 bar/deg. 240 70 180 g/kW-hr 180 170 g/kW-hr 230 As load is increased the minimum 190 60 10 bar/deg. 60 60 60 ISFC cannot be achieved with 50 200 180 g/k5.6 W-hr 10 210 190 g/kW-hr bar/deg. 40 bar/deg. either neat diesel fuel of neat 40 40 210 40 30 gasoline 190 190 g/kW-hr 220 20 180 g/kW-hr 20 20 20 Predicted contours are in good 10 170 190 200 agreement with HCCI g/kW-hr 0 0 0 0 0 10 20 30 40 50 60 0 0 0 10 10 10 20 20 20 30 30 30 40 40 40 50 50 50 60 60 60 experiments EGR Rate % EGR Rate EGR Rate EGR Rate % % % EGR Rate % 12 CEFRC4-8, 2014 PRF PRF PRF PRF PRF - - - Part 8: Optimization and Low Temperature Combustion Kokjohn, 2009 Charge preparation optimization 100 to 1500 bar Inj. 1 Pressure Premixed and Direct Injected fuel blending 100 to 1500 bar Inj. 2 Pressure Desirable to use traditional diesel SOI 1 IVC to (SOI2-20) ºATDC type injector -50 to -30 ºATDC SOI 2 Large nozzle hole (250 μm) Wide angle (145° included angle) 0 100 % Diesel Fuel Fuel split ncells KIVA + Multi-Objective Genetic 2 m PRF PRF      i i GLOBAL Algorithm (MOGA) i1 NSD  PRF ncells Fuel reactivity and EGR from HCCI PRF m GLOBAL  i investigation (9 bar IMEP) i1 0.30 Results Global PRF = 65 Film (%) All Sol utions 0.45 EGR rate = 50% 0.28 PRF Inhomogeneity 0.19 Pareto Solutions Five optimization parameters 0.26 Minimize two objectives Parameters Wall film amount 0.24 Inj. Pres. 1 (bar) 115 Inj. Pres. 2 (bar) 555 PRF Inhomogeneity 0.22 SOI1 (°ATDC) -67 Simulations run to 10 °BTDC SOI2 (°ATDC) -33 0.20 21 generations with a population size Fraction in first pulse 0.64 of 24 0.18 0 1 2 3 4 5 6 7 8 Wall Film % of Total Fuel 13 CEFRC4-8, 2014 PRF Inhomogeneity - Part 8: Optimization and Low Temperature Combustion Kokjohn, 2009 Optimized Reactivity Controlled Compression Ignition (RCCI) Port injected gasoline Optimized fuel blending in-cylinder Direct injected diesel Gasoline Squish Ignition Conditioning Source Diesel -80 to -50 -45 to -30 Crank Angle (deg. ATDC) Gasoline Diesel 14 CEFRC4-8, 2014 Injection SignalPart 8: Optimization and Low Temperature Combustion Heavy- and light-duty ERC experimental engines LD HD Engine Heavy Duty Light Duty Engine CAT SCOTE GM 1.9 L Displ. (L/cyl) 2.44 0.477 Bore (cm) 13.72 8.2 Stroke (cm) 16.51 9.04 Squish (cm) 0.157 0.133 CR 16.1:1 15.2:1 Swirl ratio 0.7 2.2 IVC ( ATDC) -85 and -143 -132 ° EVO(°ATDC) 130 112 Injector type Common rail Nozzle holes 6 8 Hole size (µm) 250 128 Engine size scaling Staples, 2009 15 CEFRC4-8, 2014 Part 8: Optimization and Low Temperature Combustion Hanson, 2010 Experimental validation - HD Caterpillar SCOTE IMEP (bar) 9 Effect of gasoline percentage Speed (rpm) 1300 Experiment 14 1400 Simulation 82% EGR (%) 43 12 1200 89% 76% 10 1000 Equivalence ratio (-) 0.5 Neat Diesel Fuel 89% 8 800 Gasoline Intake Temp. (°C) 32 Neat 6 600 Gasoline Intake pressure (bar) 1.74 4 400 Gasoline (% mass) 76 82 89 2 200 Diesel inject press. (bar) 800 0 0 -30 -20 -10 0 10 20 30 SOI1 (°ATDC) -58 Crank ATDC SOI2 (°ATDC) -37 st Fract. diesel in 1 pulse 0.62 IVC (ºBTDC)/Comp ratio 143/16 Computer modeling predictions confirmed Combustion timing and Pressure Rise Rate control with diesel/gasoline ratio Dual-fuel can be used to extend load limits of either pure diesel or gasoline 16 CEFRC4-8, 2014 Pressure MPa Apparent Heat Release Rate J/Part 8: Optimization and Low Temperature Combustion Hanson, 2011 Splitter, 2010 RCCI – high efficiency, low emissions, fuel flexibility Heavy-duty RCCI (gas/gas+3.5% 2-EHN, 1300 RPM) Heavy-duty RCCI (E-85/Diesel, 1300 RPM) Indicated efficiency of 58±1% Heavy-duty RCCI (gas/diesel 1300 RPM) achieved with E85/diesel 0.3 HD Target (2010 Levels) Emissions met in-cylinder, 0.2 without need for after-treatment 0.1 Considerable fuel flexibility, 0.0 including ‘single’ fuel operation 0.03 HD Target (2010 Levels) Diesel can be replaced with 0.02 0.5% total cetane improver 0.01 (2-EHN/DTBP) in gasoline 0.00 - less additive than SCR DEF 57 54 51 48 45 4 6 8 10 12 14 16 Gross IMEP bar 17 CEFRC4-8, 2014 Gross Ind. Soot NOx Efficiency g/kW-hr g/kW-hrPart 8: Optimization and Low Temperature Combustion Kokjohn, 2011 Dual fuel RCCI combustion – controlled HCCI RCCI Heat release occurs in 3 stages (SAE 2010-01-0345, 2012-01-0375) Cool flame reactions result from diesel (n-heptane) injection First energy release occurs where both fuels are mixed Final energy release occurs where lower reactivity fuel is located Changing fuel ratios changes relative magnitudes of stages Fueling ratio provides “next cycle” CA50 transient control 200 95 Cool Flame PRF Burn Iso-octane Burn 90  n-heptane Primarly  Primarly  150 + entrained iso-octane n-heptane CA50=2 ˚ ATDC 85 iso-octane 80 100 75 70 50 65 RCCI 60 0 SOI = -50 ATDC -20 -10 0 10 20 55 o 80 90 100 110 120 130 140 150 160 170 Crank ATDC o 18 Intake Temperature C 18 CEFRC4-8, 2014  o AHRR J/ Delivery Ratio % iso-octanePart 8: Optimization and Low Temperature Combustion Splitter, 2010 Understanding RCCI combustion Lo Loca catio tion n B B w wiit th dummy h dummy Optical Cylinder Head pl plug in ug insta stall lled ed comm common on ra rail il in injec jector tor Port Fuel Injector Lo Loca catio tion n A A fib fiber er to to comm common on ra rail il w wiit th opti h optics cs FTIR FTIR fuel fuel spra spray y in insta stall lled ed 19 CEFRC4-8, 2014 Part 8: Optimization and Low Temperature Combustion Splitter, 2010 Understanding RCCI combustion 10 10 10 10 400 400 400 400 10 400 Experiment Ex Ex Ex Expe pe pe perim rim rim rimen en en entttt Simulation Simulat Simulat Simulat Simulation ion ion ion 8 8 8 8 320 320 320 320 8 320 Location B 6 6 6 240 240 240 6 6 240 240 4 4 4 4 4 160 160 160 160 160 2 2 2 2 2 80 80 80 80 80 0 0 0 0 0 0 0 0 0 0 -20 -20 -20 -15 -15 -15 -10 -10 -10 -5 -5 -5 0 0 0 5 5 5 10 10 10 15 15 15 20 20 20 -20 -20 -15 -15 -10 -10 -5 -5 0 0 5 5 10 10 15 15 20 20 Crank Crank Crank Crank Crank     ATDC ATDC ATDC ATDC ATDC -11 deg -7 d -3 d 3 deg A eg A eg A A TDC TDC TDC TDC 16 deg ATDC Pr Pr Prod od oduct uct ucts s s Pr Prod oduct ucts s Reactant Reactant Reactant Reactant Reactants s s s s Location A B B B B B Experimental in-cylinder FTIR measurements of combustion process at two locations Spectra shows different fuel species at locations A and B, a result of the reactivity gradient Fuel decomposition and combustion products form A A A A A at a slower rate at location B, extending combustion duration 2300 2300 2300 2300 2700 2700 2700 2700 3100 3100 3100 3100 3500 3500 3500 3500 3900 3900 3900 3900 2300 2700 3100 3500 3900 Wav Wav Wavel el elen en ength ( gth ( gth (nm) nm) nm) Wav Wavel elen ength ( gth (nm) nm) 20 CEFRC4-8, 2014 Pressure MPa Press Press Press Pressure ure ure ure M M M MPa Pa Pa Pa Heat Release Rate J/ H H H Heat eat eat eat R R R Releas eleas eleas elease e e e Rat Rat Rat Rate e e e J J J J////   