University of Wisconsin -- Engine Research Center slide 1 Investigation of Heat Transfer Correlations for HCCI Engines Eric Gingrich, Christopher Gross, and Varun Ramesh Computer Project ME 769 Advanced Combustion 04/30/2013

University of Wisconsin -- Engine Research Center slide 2Outline Introduction  HCCI  Engine Heat Transfer Correlations  Project  CANTERA  Experimental setup Results Summary Conclusion

University of Wisconsin -- Engine Research Center slide 3 Introduction: HCCI HCCI combustion is described as a controlled auto- ignition combustion process Governed by chemical kinetics Heat transfer processes change the in-cylinder conditions Dominant heat transfer mechanism is forced convection from bulk gas to the wall www-pls.llnl.gov

University of Wisconsin -- Engine Research Center slide 4 Introduction: Engine Heat Transfer Correlations Caton, ICEF Chang et.al., SAE

University of Wisconsin -- Engine Research Center slide 5 Introduction: Project Woschni, Hohenberg, and Chang correlations have all been applied to a 0-D chemical kinetics solver For simulations a reduced PRF mechanism was used (47 species & 142reactions) Three engine cases were modeled: Mode 3Mode 5Mode 7 Speed (RPM) Load IMEPg (bar) Intake Temperature (C) Intake Pressure (bar) Percent n-heptane7912 Percent iso-octane Dempsey et.al., CST

University of Wisconsin -- Engine Research Center slide 6 What is CANTERA and why use it? “Cantera is a suite of object-oriented software tools for problems involving chemical kinetics, thermodynamics, and/or transport processes.” Developed David Goodwin’s group at Caltec Can be used as a 0-D kinetics solver similar is CHEMKIN Open source (free) Easy to preform parametric studies MATLAB interface code.google.com/p/cantera/

University of Wisconsin -- Engine Research Center slide 7 Development of CANTERA “Engine” CANTERA reactor allows for changing volume, area, and heat transfer coefficient Changing volume accomplished by creating a moving wall and specifying a wall velocity Wall velocity must be specified using a Fourier series At IVO initial temperature, pressure and mole fractions are specified V, A, h, T 0, P 0, X 0 Wall velocity

University of Wisconsin -- Engine Research Center slide 8 Experimental Setup Base Engine Single-cylinder version of GM 1.9L 4-cylinder DI Diesel Bore x Stroke82.0 mm x 90.4 mm Connecting Rod Length mm Displacement0.477 L Compression Ratio16.7:1 Intake Valve Opening/Closing 344°/-132° aTDC Exhaust Valve Opening/Closing 112°/388° aTDC Swirl Ratio1.5 – 2.3 Piston Bowl TypeRe-entrant bowl

University of Wisconsin -- Engine Research Center slide 9 Wireless Telemetry System

University of Wisconsin -- Engine Research Center slide 10 Results: Mode 3

University of Wisconsin -- Engine Research Center slide 11 Results: Mode 5

University of Wisconsin -- Engine Research Center slide 12 Results: Mode 7

University of Wisconsin -- Engine Research Center slide 13Summary

University of Wisconsin -- Engine Research Center slide 14Conclusion 0-D chemical kinetics only captures SOC and not combustion duration For most cases the heat transfer coefficient is only accurate within an order of magnitude 0-D heat transfer correlations are difficult to create over a large range of speeds, loads, and require much tuning In order to match SOC the Woschni and Hohenberg correlations require non-physical IVC temperatures (Woschni does not ignite)

University of Wisconsin -- Engine Research Center slide 15 Questions?

University of Wisconsin -- Engine Research Center slide 16References combustion Canton, J., “Comparisons of Global Heat Transfer Correlations for Conventional and High Efficiency Reciprocating Engines”,ASME ICEF Chang, J., et. al, “New Heat Transfer Correlation for HCCI Engine Derived from Measurements of Instantaneous Surface Heat Flux”, SAE Dempsy, Walker, Gingrich, Reitz, “Comparison of Lower Temperature Combustion Strategies for Advanced Compression Ignition Engines with Focus on Controllability”, Submitted to: Combustion Science and Technology