MINIATURIZATION LIMITS OF SMALL IC ENGINES Shyam Menon and Christopher P. Cadou Department of Aerospace Engineering, University of Maryland at College Park, College Park, USA Abstract: Considerable effort has been devoted in recent years to building miniature heat engines as battery replacements and prime-movers for micro-air vehicle propulsion. Simple thermodynamic analyses show that as a heat engine is miniaturized, it becomes less thermodynamically efficient. A critical length scale exists at which losses outstrip power production and cycle efficiency goes to zero. The objectives of this research are to identify a minimum practical length scale for two-stroke piston engines and investigate the processes responsible for setting this limit. The performance of 7 engines weighing 15g to 500g is studied using a specially designed dynamometer. Peak power outputs have been measured at 8-278W with peak efficiencies ranging from 3-9%. A scaling analysis similar to one established for conventional scale engines shows the minimum displacement for a ‘practical’ IC engine to be between 0.5 and 1 cc. Keywords: Heat engines, miniaturization, scaling, MAV propulsion INTRODUCTION There is increasing demand for miniature power systems with applications ranging from micro-propulsion to human-portable power packs. Heat engines operating on liquid hydrocarbons appear to be a natural choice due to the superior energy and power densities of ‘conventional’ scale engines [1]. Miniature internal combustion (IC) engines have been utilized for propulsion in unmanned air vehicles (UAVs) and for stationary sources of electrical power [2]. However, in order to fully realize the potential range/endurance advantage of energy dense liquid hydrocarbon fuels over batteries, miniaturized engines need to achieve levels of efficiency that are comparable to conventional- scale engines (>10%). This becomes more difficult as size is reduced: The ratio of power production to power loss (due to heat transfer and friction) scales with surface to volume ratio which increases as the size of the engine is reduced [3]. Further, engine speed also tends to increase with decreasing size which means that efficiency loss due to incomplete combustion eventually also becomes an issue [4]. Taken together, one expects there to be a minimum size below which losses outstrip energy release and it becomes impossible to close a thermodynamic cycle – let alone implement it efficiently. Peterson used a simple heat transfer analysis to show that the minimum size of a Stirling engine is approximately 1 mm for conventional manufacturing materials [5]. Apart from this, however, there appears to be relatively little work in the literature focused on the scaling of small engine performance with size. Previous work by the authors focused on performance data collected from a large number of small engine manufacturers [6]. Both two and four-stroke engines were considered. The results showed that engine power output obeyed a power-law scaling of the form y=Ax b over a remarkably wide range of sizes. This was consistent with the findings of Bonner [7] for larger engines. However, few reliable performance data were available for the smallest engines making it difficult to extrapolate to smaller scales. Earlier investigations by Heywood focused on understanding the effect of changing engine size and advances in technology on engine performance [8] suggest that other methods of correlating engine performance may work better. For example, plotting maximum torque against displacement volume appeared to correlate the performance of many different engines quite well but plotting maximum power normalized by mean piston speed versus piston area seemed to provide the best correlation. Previous work in our group used these methods to interpret data acquired using a specially built dynamometer [9] from four small model airplane engines [10]. The overall objective of this work is to improve our estimate of the minimum ‘practical’ size of a loop-scavenged, two-stroke piston