Performance Measurement and Scaling in Small Internal Combustion Engines

The development of a dynamometer system suitable for measuring the power output and efficiency of small internal combustion engines with masses less than 1 kg is described. These measurements are difficult because engine speeds are high ranging between 10000 and 30000 RPM while torque levels are low ranging between 5 and 200 oz-in. Tradeoffs associated with various approaches to making these measurements are discussed and the measurement capabilities of the dynamometer system are described in detail. The system is used to measure the power output of a small piston engine of a type commonly used in R/C aircraft. Characterizing the performance of these engines is important in defining the current state of the art in small engine performance and for providing insight into the processes/loss mechanisms governing engine performance at small scales. INTRODUCTION The power output of internal combustion engines has been shown to obey a power-law scaling of the form y=Ax over a remarkably wide range of sizes (figure 1). This relationship seems to hold for smaller engines weighing less than 1 kg and, if properly validated, would be a very useful tool for optimizing the design of small UAVs. Another useful tool would be a similar validated relationship for fuel consumption or efficiency. Unfortunately, recent work that has attempted to estimate efficiency from manufacturers’ published data and whose results are summarized in figure 2 shows that log-linear scaling of efficiency with engine size breaks down for smaller engines with masses less than 1 kg (35 oz). * Assistant Professor, Senior Member AIAA Copyright © 2003 by Christopher Cadou. Published by the American Institute of Aeronautics and Astronautics, Inc. with permission. Understanding the performance of engines weighing less than 1 kg is important for several reasons. First, it is important to researchers who are trying to build extremely small heat engines and are trying to understand the tradeoffs associated with designing and operating heat engines weighing less than 5g. One problem they have encountered is that the most recent performance estimates indicate that the efficiency of these devices will be unacceptably low (~1-5%). However, this does not necessarily mean that the microengine idea is impractical. For example, while it could be the consequence of fundamental physical limitations, it could also be a simple consequence of the fact that the technology is still immature and needs refinement. Looking for trends in the performance of model engines that represent a more mature technology may provide some insight that cannot be gained from efficiency estimates like those presented in figure 2. Second, having a good understanding of the performance of the smallest engines currently available is important to groups that cannot wait for microengines to be developed and are trying to build small air vehicles today. For example, consider the Low-cost Unmanned Air Vehicle being developed by the U.S. Navy. With a gross weight of about 20 lb, figure 3 shows that an efficient energy storage and propulsion system is critical to achieving the target range of 1500 miles (2414 km). It also shows that acceptable performance may be available from existing model aircraft engines if they are at least 25% efficient and can be operated on JP10. Figure 2 indicates that the target efficiency may be obtainable. The challenge that the Navy faces is two-fold. First, reliable data indicating that 25% efficiency can be achieved is unavailable – estimates computed from manufacturers’ specifications show so much scatter as 41st Aerospace Sciences Meeting and Exhibit 6-9 January 2003, Reno, Nevada AIAA 2003-671 Copyright © by 2003. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission. 2 American Institute of Aeronautics and Astronautics to render it useless. Second, these engines do not burn the proper fuel. It is difficult to mix fuel and air efficiently at small scales and model aircraft engines solve this problem by burning special oxygenated fuels that are mixtures of methanol, nitro-methane, and castor oil. Figure 3 indicates that a viable powerplant for the Navy would need to burn JP10 to meet the range requirement in addition to ensuring compatibility with ship-board operations. The work presented here discusses the development of a dynamometer used to make measurements of the torque, speed and fuel consumption of small model aircraft engines. These are subsequently used to measure the power output and overall efficiency of these engines. Some preliminary results are presented to demonstrate the capability of the instrument. Note that the power output P of the engine is given by: ω Γ = P (1) where Γ and ω are respectively the torque and speed (in radians/sec) at the output shaft. The overall efficiency (or brake thermal efficiency) of the engine is defined as follows: r f o Q m P & = η (2) where P is the power output of the engine, f m& is the fuel mass flow rate and Qr is the energy density of the fuel. Table I Data for several single cylinder model aircraft engines that could be appropriate for powering a low-cost UAV. Note that all are twostroke machines except for engines 3 and 4. Wt. Pwr. ηο Spd Displ. Torque Fuel 1 9.5 1.0 12 15 0.40 67.3 0.0105 2 13.2 1.62 16 16 0.46 102.1 0.0128 3 14.2 0.9 20 12 0.52 75.7 0.0055 4 22.2 1.6 21 12 0.91 134.5 0.0096 5 12.2 1.1 12 16 0.40 69.4 0.0111 6 15.5 0.93 20 12 0.56 78.2 0.0059 7 13.2 1.45 14 16 0.46 91.4 0.0129 8 13.4 1.54 13 17 0.49 91.4 0.0146 oz HP % kRPM in oz-in oz/s THE CHALLENGE Several factors complicate measurement of power output and fuel consumption in these small engines. Table 1 presents some data describing model aircraft engines as a reference for the magnitudes of the quantities we expect to measure. Note that the efficiency was estimated from manufacturers’ data. y = 9901.3x R = 0.9813 1 10 10