Opposed-Piston Two-Stroke Diesel Engine Scavenging System Structural Optimization Based on GA-SVM

Opposed-Piston Two-Stroke (OP2S) diesel engines have been successfully used in many applications because of its high power efficiency. Uniform scavenging is used on OP2S diesel engines. Different from conventional diesel engine exchanging systems, intake ports and exhaust ports are located at each sides of the cylinder as OP2S scavenging system. A scavenging system optimization method, Genetic Algorithm-Support Vector Machine (GA-SVM), which employs Indicated Mean Effective Pressure (IMEP) as optimization goal is presented. The combinations of five key parameters including the width and length of intake and exhaust ports and the crank asymmetric angle are designed. All these combinations are calculated by GT-Power. The predicting model of OP2S diesel engine`s IMEP is built based on the 1D results. After verifying the accuracy of the predicting model, the optimized parameters are found. The results illustrate: using 1-D simulation coupled GA-SVM to optimize scavenging parameters in OP2S diesel engines is a rapid and effective method. The optimized IMEP is 12.86 bar. This paper presents some guidelines in the further OP2S diesel engines research. Introduction The opposed-piston two-stroke engine (OP2S) concept came into being in 1850`s. Wittig, a German engineer, designed and manufactured the world`s first coal gas fuel opposed-piston engine [1].From then on, many novel designs were brought out on aircrafts, marines and vehicles such as TS-3 engine manufactured by British Rootes company and 6-TD engine from Ukraine. However, rigid emission legislations slowed down the development of opposed-piston engines after 1970`s. In recent years, with the improvement of modern analytical tools, materials, and fuel supply system, engines from FEV, EcoMotor and Achates Power company have performed well in many fields [2,3,4,5,6]. OP2S draw more and more attention once again because of its simple, lightweight, high efficiency and high power density [7]. In OP2S scavenging system, two pistons reciprocate opposite to each other in a common cylinder, piston port timings are controlled by two pistons. With this structural design, the need for cylinder head and valve mechanism are eliminated. In order to achieve better scavenging performance, larger ports area is wanted. However, according to the research by Peter Hofbauer [8], although larger intake and exhaust ports area can achieve better scavenging performance, shorter compression and expansion processes are inevitable. This will lead to the decrease of power efficiency. Zhenyu Zhang [9] also found larger crank asymmetric angle is benefit to the scavenging efficiency meanwhile compression ratio and expansion ratio decrease along with the increase of the crank asymmetric angle. This is also unfavorable to the development of engine performance. Therefore, in optimizing OP2S scavenging process, only taking scavenging efficiency into account is unreasonable. Thermal efficiency should also be taken into account. With the consideration of this, IMEP is selected as optimizing goal. In engine scavenging system optimization area, two simulation approaches are widely used. The first one is Computational Fluid Dynamics (CFD) method using such as AVL-Fire, Star-CD. The CFD method can model the scavenging process with complex geometries. However, its calculation time is very long. This is unfavorable in initial engine designing. The second one is a predictive simulation method which is required to determine basic engine parameters in order to shorten the development time [10]. GT-Power is always the priority choice. Compared with detailed simulation, the 1D simulation has simpler model and it has lower demands with calculation. So 1-D simulation 154 This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). Copyright © 2017, the Authors. Published by Atlantis Press. Advances in Engineering Research (AER), volume 111 3rd Annual 2017 International Conference on Sustainable Development (ICSD2017)

[1]  Jan Macek,et al.  Air System Conception for a Downsized Two-Stroke Diesel Engine , 2012 .

[2]  W. Karush.  Minima of Functions of Several Variables with Inequalities as Side Conditions , 2014 .

[3]  P. T. Brandham,et al.  A method for obtaining a quantitative assessment of the influence of charging efficiency on two-stroke engine performance , 1969 .

[4]  Gerhard Regner,et al.  Practical Applications of Opposed-Piston Engine Technology to Reduce Fuel Consumption and Emissions , 2013 .

[5]  Fujun Zhang,et al.  Effect of Piston Dynamic on the Working Processes of an Opposed-Piston Two-Stroke Folded-Cranktrain Engine , 2014 .

[6]  Stephan Schmidt,et al.  Predictive Simulation Strategies for the 2-Stroke Scavenging Process within the Scope of the Development Process , 2005 .

[7]  Ricardo Novella,et al.  Two Strokes Diesel Engine - Promising Solution to Reduce CO2 Emissions , 2012 .

[8]  Fabio Bozza,et al.  A Comprehensive 1D Model for the Simulation of a Small-Size Two-Stroke SI Engine , 2004 .

[9]  Nello Cristianini,et al.  An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .

[10]  Yuhang Liu,et al.  Study on the synthetic scavenging model validation method of opposed-piston two-stroke diesel engine , 2016 .

[11]  R. Fletcher.  Practical Methods of Optimization , 1988 .

[12]  Michael Franke,et al.  Opposed Piston Opposed Cylinder (opoc™) 450 hp Engine: Performance Development by CAE Simulations and Testing , 2006 .

[13]  Gary L. Borman,et al.  Mathematical Simulation of a Large Turbocharged Two-Stroke Diesel Engine , 1971 .

[14]  Gerhard Regner,et al.  Meeting Stringent 2025 Emissions and Fuel Efficiency Regulations with an Opposed-Piston, Light-Duty Diesel Engine , 2014 .

[15]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[16]  F. J. Wallace,et al.  Experimental and Analytical Scavenging Studies on a Two-Cycle Opposed Piston Diesel Engine , 1971 .

[17]  Rifat Keribar,et al.  Engine/Powertrain/Vehicle Modeling Tool Applicable to All Stages of the Design Process , 2000 .

[18]  Vladimir Vapnik,et al.  An overview of statistical learning theory , 1999, IEEE Trans. Neural Networks.

[19]  Giovanni Ferrara,et al.  An Innovative Solution for Two-Stroke Engines to Reduce the Short-Circuit Effects , 2012 .

[20]  Gerhard Regner,et al.  Modernizing the Opposed-Piston, Two-Stroke Engine for Clean, Efficient Transportation , 2013 .

[21]  F. J. Wallace,et al.  Some single zone scavenging models for two-stroke engines , 1992 .

[22]  Peter Hofbauer,et al.  Opposed Piston Opposed Cylinder (opoc) Engine for Military Ground Vehicles , 2005 .

[23]  Rifat Keribar,et al.  “Virtual Engine/Powertrain/Vehicle” Simulation Tool Solves Complex Interacting System Issues , 2003 .

[24]  Gerhard Regner,et al.  The Achates Power Opposed-Piston Two-Stroke Engine: Performance and Emissions Results in a Medium-Duty Application , 2011 .

[25]  Fabrice Vidal,et al.  Preliminary Design of a Two-Stroke Uniflow Diesel Engine for Passenger Car , 2013 .

[26]  Gerhard Regner,et al.  Thermodynamic Benefits of Opposed-Piston Two- Stroke Engines , 2011 .

[27]  E Sher,et al.  A Simple and Realistic Model for the Scavenging Process in a Crankcase-Scavenged Two-Stroke Cycle Engine , 1991 .