Development + characterisation of small-scale rotary engines (MEMS)

frank herfert

Outline

Development + characterisation of small-scale rotary engines (MEMS)

frank herfert

2007

visibility

105 views

description

19 pages

Abstract

Technical paper by A. Carlos Fernandez-Pello, Sang-Won Park, Albert P. Pisano, S. Bennett Sprague, + David C. Walther UoC Berkeley, 2007, 19 pages > Content > Abstract + keywords 1 > Reference + biographical notes 1 > 1 Introduction 2 > 2 Background 3 » 2.1 Characteristic times of combustion 3 » 2.2 Heat transfer 4 » 2.3 Flame quenching 4 » 2.4 Mixing + turbulence 5 » 2.5 Sealing 5 > 3 Small-scale rotary engine designs 6 » 3.1 Peripheral ported P78 engine 6 » 3.2 Peripheral ported P348 engine 7 » 3.3 Side-ported S367 engine 8 » 3.4 Side-ported S1500 engine 9 > 4 Test apparatus 9 > 5 Experimental results 11 » 5.1 P78 peripheral ported engine 12 » 5.2 P348 peripheral ported engine 12 » 5.3 S367 side-ported engine 13 » 5.4 S1500 side-ported engine 14 > 6 Discussion 15 » 6.1 Effect of engine seals 15 » 6.2 Effect of port timing 16 » 6.3 Effect of combustion chamber design 16 » 6.4 Effect of blowby 16 > 7 Conclusions 18 > Acknowledgements 18 > References 18 This paper describes the development and characterisation of small-scale rotary engines with displacements in the range of 78-1500 mm 3 for portable applications in the range of 10-200 W of power output. Small-scale combustion engines present a number of research challenges including manufacturing tolerances, sealing, thermal management, ignition, combustion efficiency and porting. Four engines have been characterised using a custom test bench and show an increase in performance due to design changes that mitigate the challenges associated with small-scale engines. The volumetric power density has been increased from 11 W/cm 3 in a 348 mm 3 engine operating with a supercharged hydrogen/air mixture to 22 W/cm 3 in a 1500 mm 3 engine operating with naturally aspirated liquid hydrocarbon fuel. The thermal efficiency has also been increased from 0.2 to 4%. Continued improvements in sealing, thermal management, combustion efficiency and friction reduction will allow further increases in engine performance. german keywords: Wankelmotor / Kreiskolbenmotor English Keywords: Wankel Engine / Wankel Rotary Engine / Rotary Piston Engine / Rotary Combustion Engine

Key takeaways
AI

The study focuses on developing and characterizing small-scale rotary engines for portable power generation.
Engine displacements range from 78 mm³ to 1500 mm³, achieving power outputs of 10-200 W.
Power density increased from 11 W/cm³ (348 mm³) to 22 W/cm³ (1500 mm³) through design optimizations.
Thermal efficiency improved from 0.2% to 4% by enhancing sealing and combustion chamber designs.
Continued advancements in sealing, thermal management, and combustion efficiency are essential for performance gains.

Int. J. Alternative Propulsion, Vol. 1, No. 2/3, 2007 275 Copyright © 2007 Inderscience Enterprises Ltd. Development and characterisation of small-scale rotary engines S. Bennett Sprague, Sang-Won Park, David C. Walther, Albert P. Pisano and A. Carlos Fernandez-Pello* Department of Mechanical Engineering, University of California at Berkeley, Berkeley, CA 94720-1740, USA Fax: +1-510-642-1850 E-mail: sprasb@me.berkeley.edu E-mail: intropia@me.berkeley.edu E-mail: walther@eecs.berkeley.edu E-mail: appisano@me.berkeley.edu E-mail: ferpello@me.berkeley.edu *Corresponding author Abstract: This paper describes the development and characterisation of small-scale rotary engines with displacements in the range of 78–1500 mm 3 for portable applications in the range of 10–200 W of power output. Small-scale combustion engines present a number of research challenges including manufacturing tolerances, sealing, thermal management, ignition, combustion efficiency and porting. Four engines have been characterised using a custom test bench and show an increase in performance due to design changes that mitigate the challenges associated with small-scale engines. The volumetric power density has been increased from 11 W/cm 3 in a 348 mm 3 engine operating with a supercharged hydrogen/air mixture to 22 W/cm 3 in a 1500 mm 3 engine operating with naturally aspirated liquid hydrocarbon fuel. The thermal efficiency has also been increased from 0.2 to 4%. Continued improvements in sealing, thermal management, combustion efficiency and friction reduction will allow further increases in engine performance. Keywords: rotary engine; meso-scale combustion; portable power. Reference to this paper should be made as follows: Sprague, S.B., Park, S-W., Walther, D.C., Pisano, A.P. and Fernandez-Pello A.C. (2007) ‘Development and characterisation of small-scale rotary engines’, Int. J. Alternative Propulsion, Vol. 1, No. 2/3, pp.275–293. Biographical notes: S. Bennett Sprague is currently a PhD candidate in the Combustion Processes Laboratory at the University of California, Berkeley. He received his Master’s degree in Mechanical Engineering from the University of California, Berkeley in 2004. His research interests include combustion at the microscale and small-scale combustion engines. Sang-Won Park is currently a PhD student in the Combustion Processes Laboratory at the University of California, Berkeley. He received his Master’s degree in Mechanical Engineering from the University of California, Berkeley in 2005. His research interests include fuel delivery systems for small-scale engines and microelectro mechanical systems. David C. Walther is a research specialist with the Berkeley Sensor and Actuator Center (BSAC) at the University of California, Berkeley. He received his PhD in Mechanical Engineering (Thermosciences) from the University of

276 S.B. Sprague et al. California, Berkeley in 1998. His research interests include sensors and microsystems for portable power and energy applications. Albert P. Pisano is a Professor of Mechanical Engineering at the University of California, Berkeley. He received his PhD in Mechanical Engineering from Columbia University. His research interests include Micro Electromechanical Systems (MEMS) for a wide variety of applications, including power generation, radio frequency components, drug delivery and strain sensors. A. Carlos Fernandez-Pello is a Professor of Mechanical Engineering at the University of California, Berkeley. He received his PhD in Applied Mechanics and Engineering Science from the University of California, San Diego. His research interests span micro- and meso-scale combustion and microgravity combustion, specifically ignition and flame propagation, smoldering and transition to flaming. 1 Introduction Advances in device miniaturisation have led to opportunities in reduced-scale, distributed power generation. Electrochemical, thermochemical and biological power production strategies are being investigated for personal power systems (Dunn-Rankin et al., 2005). Particularly interesting is the production of power from internal combustion engines using liquid fuels. Liquid fuels have a very large chemical energy density and consequently are particularly suitable for portable power generation. In comparison with rechargeable batteries, with an energy density of 0.7 MJ/kg for lithium-ion batteries, hydrocarbon fuels have 45 MJ/kg of stored chemical energy. Therefore, if this chemical energy can be converted to electricity with an efficiency as low as even 5%, then portable devices could be developed that would replace batteries for a number of applications. Leveraging the inherent advantages in storage and energy density of liquid hydrocarbon fuels is the prime motivation for this and other heat engine work. A thorough review of the issues related to combustion device development at the small scale has been previously reported (Fernandez-Pello, 2002). Several of the other heat engine and thermochemical approaches in development are addressed in the above two reviews. The objective of this work is to describe the development, fabrication and operation of small-scale rotary engines developed in the University of California Berkeley combustion laboratory to be used in a portable power generation device. A rotary engine design was originally chosen for a MEMS engine that would benefit from the planar design (Fu et al., 2001). The rotary engine is still well suited for small-scale power generation because it has a high specific power, a low cost due to a minimum number of moving parts, and no valves or transmission are required for operation. The planar design allows for high-precision electrical discharge machining and the mechanical shaft output can be directly coupled to an electric motor to produce electrical power. The emissions of rotary engines are lower than two-stroke engines due to better intake and exhaust scavenging that prevent contamination of fresh charge with exhaust, although sealing problems may reduce this benefit. Furthermore, rotary engines are inherently capable of fuel flexible operation with glow plug ignition, which is a preferable mode of operation in small-scale engines. This is due to the motion of the rotor whereby the compression stroke is separated from the glow plug location, which prevents premature initiation of

Development and characterisation of small-scale rotary engines 277 the combustion reaction (Cardes et al., 2005). During the engine development, several distinct designs have been produced to characterise the effect of engine size, port timing, fuel type and sealing on power generation. These results will be described and the implications to other small-scale combustion engines will be discussed. 2 Background There are several important factors that affect the performance (output power and efficiency) of small-scale engines. These include flame quenching, combustion time, mixing, turbulence and sealing. Flame quenching is important because it limits the minimum linear distance within which combustion can be sustained and therefore limits the size of the combustion chamber and in turn the achievable power density. Combustion time is crucial to engine operation since it determines maximum engine speed and therefore power output. Furthermore, combustion time affects the completeness of the chemical reactions within an engine, affecting emissions. The low Reynolds numbers encountered in small-scale engines hamper turbulent mixing, which is needed for achieving homogeneous mixtures. Sealing affects the efficiency and power output since the maximum pressure is reduced and exhaust gases can contaminate the intake charge. 2.1 Characteristic times of combustion For complete combustion, the residence time of the expansion stroke of the engine must be longer than the chemical time necessary for combustion to occur. The residence time is determined by the engine operation speed and the geometry, but is independent of scale for rotary engines. The residence time of a combustion stroke is calculated from the ratio of the crankshaft angular displacement as the rotor passes from the end of the compression stroke to the beginning of the exhaust stroke, and the angular speed of the rotor. For the engines addressed here that have speeds ranging from 5000 to 20000 rpm, the corresponding residence times range from 7.5 to 1.9 ms with an angular crankshaft displacement of 225°. The chemical reaction time is estimated from the reaction rate of the fuel and the average concentration of the fuel and oxidiser. Using a one-step global reaction under stoichiometric conditions and a combustion temperature of 2200 K, methanol has a characteristic reaction time of 10 µs (Borman and Ragland, 1998). The small-scale rotary engines, however, are expected to have combustion temperatures lower than 2200 K due to high heat losses to the surroundings. Using a combustion temperature as low as 1300 K, the reaction time of methanol is calculated to still be less than 1.2 ms. Thus, the residence time of the expansion stroke is still longer than the corresponding chemical reaction time since small-scale rotary engines are designed to operate at maximum speeds of 20000 rpm. In many cases, energetic fuels can be used in the form of nitromethane additives to increase the chemical reaction rate and in turn reduce the chemical time by increasing the oxidiser concentration. The addition of nitromethane is common practice for small-scale combustion engines of this scale (Papac et al., 2004).

278 S.B. Sprague et al. 2.2 Heat transfer Heat generation is proportional to the chemical mass throughput of the device. The mass throughput is proportional to the area of the inlet multiplied by the residence time. Assuming the residence time is equal to the chemical time, the mass throughput and consequently the heat generation rate are proportional to q ~S L ·l 2 where S L is the laminar flame speed and l is the characteristic dimension of the chamber or the device. The power density ( q /l 3 ) therefore increases inversely proportional to the scale of the device (Epstein and Senturia, 1997). Heat losses in actual systems, however, place a practical limit on the net power of the device (heat generated less heat lost). Heat losses to the environment are proportional to the surface area of the engine and the heat generated by combustion is proportional to the volume of the combustion chamber. Since volume is related to the length cubed and surface area is related to the length squared, the ratio of surface area to volume increases linearly as the length scale is reduced. Therefore, the rate of heat loss approaches and finally dominates heat generation as the scale is reduced. This critical crossover point has been determined to first order to be of the order of 1 mm by several investigators (Fu, 2001; Leach and Cadou, 2005; Peterson, 1998). 2.3 Flame quenching Quenching occurs due to thermal losses as well as radical destruction at the wall (Glassman, 1987). There is a minimum distance of a combustion chamber at which combustion can occur and it is determined by the ratio of the heat generated by the combustion reaction to the heat loss to the wall. The minimum linear distances of the combustion chambers for the small-scale rotary engines presented in this paper are between 0.4 and 1.6 mm, which is less than the thermal quenching distance for methanol at standard temperature and pressure conditions, 2.0 mm (Borman and Ragland, 1998). However, the thermal quenching distance in an actual engine is decreased due to the high-temperature walls and high compression pressures (Borman and Ragland, 1998). Experiments on combustion in small channels have shown that external heating of channel walls allows combustion to occur in channels smaller than the conventional quenching distance (Maruta et al., 2005; Zamashchikov, 1995). Numerical modelling of combustion in channels with adiabatic walls also shows that flames can propagate in channels smaller than the quenching distance (Leach and Cadou, 2005; Lee and Tsai, 1994). Radical quenching, on the other hand, becomes dominant only when the wall temperature is relatively high. Miesse et al. (2004) found no effect of surface material on quench distance below 500 K and a large radical quenching effect near 1300 K. The maximum heat losses in rotary engines occur at the top dead centre position when both the temperature and the gas velocity are high due to changes in density from combustion. Using the correlation from Woschni (1967) as referenced by Heywood (1988), the maximum heat transfer coefficient for a small-scale engine with a displacement of 1.5 cm 3 at 6000 rpm is found to be 535 W/m 2 K. The heat transfer coefficient is proportional to the rotor width, B, and gas velocity, w, as follows (Heywood, 1988): 0.2 0.8 c h B w α − (1) The heat transfer coefficient decreases as the size of the engine is decreased because of the reduction in the gas velocity, which offsets the effect of the rotor width reduction.

Development and characterisation of small-scale rotary engines 279 The Woschni correlation was however developed for large-scale engines by assuming turbulent convective heat transfer across a flat plate resulting in the 0.8 power dependency on velocity. The Reynolds number in the 1.5 cc engine ranges from 4200 at 8000 rpm to 7800 at 15000 rpm. This places the flow field outside of laminar flow, but cannot be considered fully turbulent. As such, the heat transfer could scale with velocity to a power as low as 0.5 for the fully laminar case and therefore application of the Woschni correlation may overestimate the heat transfer coefficients within small-scale engines. As engine scale is reduced, however, continued application of turbulent model correlations may overestimate the effect of decreasing engine size on the heat transfer coefficient. However, even though h c is reduced, the total heat loss increases because the surface area to volume ratio increases as the size of the engine is decreased. Therefore, flame quenching is more likely to occur as the size is reduced due to the higher ratio of heat losses to heat generation. 2.4 Mixing and turbulence Both a homogeneous charge and turbulent flow enhance the overall reaction rate, and consequently increase the potential power output of the engine. The former requires adequate mixing of fuel and oxidiser, which depends on the turbulence level and the mixing length. Turbulence is determined by the Reynolds number, which decreases as the size of the engine is decreased. Therefore, mixing and turbulence are important issues in small-scale engines. In the range of engines described here, the Re (defined by engine intake conditions) values are typically in the transitional regime (Dellimore and Cadou, 2004); thus turbulence levels are limited. However, with liquid fuels, droplet vaporisation times are dominant over mixing times in small engines (Park et al., 2006); thus care must be taken to ensure adequate vaporisation for optimal engine performance. 2.5 Sealing Leakage during the compression stroke and the combustion event reduces the overall power output due to the loss of high-pressure gas (Annen et al., 2003). Ideally, the ratio of mass leakage to mass throughput should be constant as the size of the engine is decreased. Large-scale rotary engines have complex sealing mechanisms to separate the three operating chambers and ensure a high operating pressure, including apex seals, face seals and corner seals (Yamamoto, 1981). However, for small-scale rotary engines, it is impractical to have numerous sealing components because of the difficulty to assemble. Therefore, few and relatively simple seals are used for apex sealing. Tight tolerances are desired for face sealing, which seals between the rotor face and housing. Tolerance is, however, limited by manufacturing technology and does not scale with engine size. Small-scale engines with the same tolerance level as large-scale engines suffer a greater proportion of mass loss. Therefore, better means of fabrication are required at this scale. The small-scale rotary engines described in this paper are fabricated with electrical discharge machining and have tolerances ranging from 0.2 to 0.003 mm depending on the manufacturing approach. The apex seals and the close fit between the rotor and engine endplates, however, also add to the frictional losses experienced by the engine. An analysis of the frictional work due to the apex seals, normalised by the engine displacement, indicates that this value should increase as displacement decreases to the 1/3 power of displacement,

280 S.B. Sprague et al. f apex α disp −1/3 . However, analysis of normalised frictional losses due to face seals indicates that the friction decreases as displacement decreases by the 1/3 power of displacement, assuming a constant gap between rotor and the endplates, f face α disp 1/3 . Direct measurements of the normalised motoring work on 0.367 cc, a commercially available 5 cc rotary engine, and 500 cc (Yamamoto, 1981) engines show that the normalised friction decreases as the size is decreased, suggesting that face seal friction is the dominant friction loss mechanism at the small scale. 3 Small-scale rotary engine designs Four engines have been built and tested over the course of this development project to examine the impact of engine design and construction on performance. A table listing the design characteristics and operating conditions can be seen in Table 1. Changes in manufacturing processes, sealing design, porting methods and combustion pocket design, as well as increases in engine size, have led to improvements in overall engine performance. Table 1 Engine design characteristics and operating conditions Name P78 P348 S367 S1500 Displacement (mm 3 ) 78 348 367 1500 Porting Peripheral Peripheral Side Side Rotor length (mm) 9.5 12.5 17 25 Rotor depth (mm) 3.6 9 6.3 10 Pocket depth (mm) 0.46 0.67 1.3 1.5 Tolerance (mm) ±0.2 ±0.02 ±0.003 ±0.003 Fuel type Gaseous Gas/liquid Liquid Liquid Air induction Supercharged Supercharged Naturally aspirated Naturally aspirated Speed range (rpm) 2500–5000 3000–9000 1000–6000 8000–15000 Housing temperature (°C) 25–45 45–75 45–100 90–115 Igniter location Minor axis Minor axis Off axis Off axis 3.1 Peripheral ported P78 engine The P78 engine with a displacement of 78 mm 3 is a 1/64th scale version of the smallest commercially available rotary engine, the 5 cc Graupner/OS 49-PI. The P78 has been designed and built to determine the effect of scale on combustion performance. The engine consists of 15 parts: front plate, epitrochoid housing, back plate, rotor, internal gear, spur gear, shaft and three apex seals and three leaf springs (Figure 1). The apex seals consist of tabs of brass or steel between 0.5 and 1.0 mm thick. The seals are backed with metal leaf springs to maintain contact between the seals and engine housing. Two bearings, which are mounted in the front and back plates, position the shaft. The engine

282 S.B. Sprague et al. no observable effect in output power from the apex-seal-only designs. As such, the simplicity and relative effectiveness of the apex-seal-only designs are those implemented in engine testing. Sealing of the rotor face is achieved by having a close fit between the rotor and the endplates. Figure 3 Apex and face seals incorporated into the P348 engine 3.3 Side-ported S367 engine The size of the S367 engine is approximately the same as that of the P348 with a displacement of 367 mm 3 , but a side-ported design is used instead of peripheral ports. The displacement changes are due to slight changes in housing and combustion pocket geometry. Peripheral ports are advantageous at high speeds for maximum power output because of an increase in volumetric efficiency (Ansdale, 1968). However, peripheral ported engines have increased rotor port overlap, which occurs when the intake and exhaust ports are open at the same time, contaminating the intake charge with exhaust gases. Side port designs alleviate contamination of the intake charge by exhaust gases, as the ports are designed to prevent any port overlap. The rotor and rotor housing are fabricated with M2 tool steel to improve tolerances to 0.003 mm, and side plates are fabricated with aluminium (Figure 4). These endplates maintain assembly tolerance, but do not see direct contact with rotating components. Wear plates that contain the side ports are also made of M2 tool steel and are used between the aluminium endplate and rotor to minimise damage incurred by wear. The rotors are thinner compared to the P348 to focus flame front motion in the direction of the applied torque rather than towards the endplates, more consistent with operating rotary engines (Yamamoto, 1981). It has been shown in previous studies that the peak pressure during combustion in closed and semi-closed chambers of the order of 1 mm is reduced as the chamber height and width are reduced (Lee et al., 2003; Tsuji et al., 2004). Therefore, a teardrop-shaped rotor recess is used to increase the depth of the combustion chamber to increase combustion efficiency by allowing heat losses to the ignition kernel to be reduced (Figure 5). The inverted teardrop shape has been designed to increase turbulence, and in turn reaction rate and mixing through a ‘squish flow’ effect (Gorla and Bartrand, 1996). The igniter location is offset from the minor axis to reduce blowby as described in the Discussion section below.

284 S.B. Sprague et al. mass flow controllers (Figure 6). A spark ignition system, manufactured by CH Electronics, Inc., that utilises a Hall effect sensor that is attached to a rotary dial is used to control the ignition and spark timing. A customised igniter is made from a ceramic tube with a tungsten electrode. In addition to a spark plug, a modified glow plug has been tested on the P348. The preliminary hydrogen testing was only done for experimental purposes, as practical devices would use liquid fuels due to the ease of transportation and high energy density. Figure 6 Schematic diagrams of engine testing apparatus. Left: dynamometer for gaseous fuels and electrical power measurement. Right: dynamometer for liquid fuels and mechanical power measurement (1) (1) (2) (2) (3) (3) (4) (5) (6) 1) engine 2) motor/dyno 3) ignition system 4) mass flow controller 5) fuel injector 6) torque arm/load cell Therefore, the two side-ported engines are tested with a liquid fuel blend, which consists of methanol (70%, by vol.), nitromethane (10%) and castor oil as a lubricant (20%) (Papac et al., 2004). A micro-dispensing valve (INKA2457210H) from the Lee Company is used as a fuel injector, and a function generator controls the fuel flow rate over the range of 10–100 mg/s. The maximum pressure of the fuel injection system is limited to 2 atm and the orifice diameter is 12.5 µm. The limiting pressure is established with regard to storage and safety concerns in the final demonstration unit. Air is inducted by natural aspiration and mixed with fuel in the intake manifold, which consists of 3 mm diameter tubing with 6 and 8 mm lengths (Figure 7). The ignition system for these engines was improved by controlling the ignition delay with a function generator. The function generator is actuated with a signal from an optical sensor that detects the position of the crankshaft. To ensure that spark power is consistent at higher operation speeds (limited by capacitor charging time), an automotive ignition system from MSD (part no. 5900 and 8223) is implemented for the later engine testing. The S367 used the same customised spark plug as the peripheral-ported engines; however, for the S1500, a commercial ¼”-32 spark plug (CH Ignitions Rimfire Plug no. 111) is adapted. Engine speed for all engines is measured using a Monarch Instruments ACT-3 tachometer with ROS-5W remote optical sensor. The engines are rigidly coupled to the dynamometer via a steel shaft. The engine power for all engines except the S1500 is determined by measuring the electrical power generated from a dynamometer. The electrical power is measured with a Maxon brushless electronically commutated motor

286 S.B. Sprague et al. 5.1 P78 peripheral ported engine The P78 engine has been tested with stoichiometric hydrogen/air and propane/air mixtures. An electric motor is used to motor the engine while the spark ignition system ignites the mixture. The engine does not generate a net power output, but there is a reduction in the amount of power needed by the electric motor to turn the engine with combustion. The engine has been run with a Plexiglas front plate in order to obtain infrared images of combustion while operating the engine as shown in Figure 8. Figure 8 Infrared images of hydrogen–air combustion in the P78 engine 5.2 P348 peripheral ported engine A typical testing procedure for the P348 consists of bringing the engine up to a speed of approximately 3000 rpm through the use of the pressurised air to overcome friction. Then the spark system is energised and hydrogen is introduced until sustained combustion is obtained. With the spark energised, speed increases of 2000–4000 rpm are noted and the power increases from 1 to 3 W, as seen in Figure 9. By subtracting the amount of power generated with the spark off from the power generated with the spark on, the net power generated by the engine can be calculated. It is important to note that the energy of the compressed gas is not considered in these calculations. The results of power tests over a range of engine speeds can be seen in Figure 10. Notice the linear rise in power with speed with no drop in power at high speeds, indicating that the residence time is still greater than the chemical reaction time for all engine speeds tested. The equivalence ratio for the tests in Figure 10 is in the range of 0.37–0.42. Higher equivalence ratios would result in more power, but flashback is noted, preventing the experiment from achieving safe operation at elevated equivalence ratios. The flashback is attributed to excess porting overlap, which allows the flame from the exhaust to enter the intake. The spark timing is 0° After Top Dead Centre (ATDC) for all conditions, with similar results being obtained with a glow plug.

288 S.B. Sprague et al. rate is reduced until combustion is found to occur and then reduced further until no combustion is detected. This range of mass flow rates is shown in Figure 11 as a function of engine speed. There are two regimes in the flow rate versus speed plot, a rich region at low speeds and nearly stoichiometric region at high speeds. This change in required fuel rate at 4000 rpm may be partially attributed to a transition from laminar to turbulent flow enhancing the fuel and air mixing. This effect is similar to the high mean inlet gas velocity required for rotary racing engines to promote combustion at the lower speed range (Ansdale, 1968). The calculated Reynolds number for the S367 operating condition is in the range of 1000 and therefore additional effects are most likely also occurring. At low speeds, the excess fuel may be needed to fill the gaps between the rotor and housing and to create additional sealing in the 0.004 mm gap. This sealing by liquid fuel results in increased compression, reaching a threshold level at 4000 rpm that results in a more complete combustion process. As the total amount of mass loss is a function of both gap size and time, leakage at high speeds is not as critical as there is less time for leakage to occur. Figure 11 Fuel rate as a function of engine speed for S367 engine showing rich region at low speed and stoichiometric region at high speed 0 5 10 15 20 25 30 35 40 45 50 0 1000 2000 3000 4000 5000 6000 Engine Speed (RPM) Mass Flow Rate (mg/s) Measured fuel consumption rate Stoichiometic fuel consumption rate 5.4 S1500 side-ported engine The S1500 generates positive power with naturally aspirated, liquid-fuelled operation. A fuel blend with 50% methanol, 30% nitromethane and 20% castor oil is used with a flow rate of 47±3 mg/s. To motor the engine at 10000 rpm with no spark, 77 W of input power is needed. With the spark on, the engine speed increases to 11000 rpm and the input power required drops to 2.6 W. When power to the electric motor is turned off, the engine speed is reduced to 8600 rpm, and the engine continues to operate. To load the engine and make positive power, a bank of rheostats is put in series with the rectified output of the drive motor. The spark timing is adjusted to find the optimum operating range and it is observed that the engine operates best with retarded spark timing, ranging from 5 to 25° ATDC.

Development and characterisation of small-scale rotary engines 289 The maximum sustained mechanical power during these tests has been determined to be 33 W. The conditions are a speed of 10500 rpm and 15° ATDC spark timing with 45±3 mg/s of a fuel blend with 60% methanol, 20% nitromethane and 20% castor oil. A representative engine test is shown in Figure 12. Using the positive power generated and the chemical energy of methanol of 19 MJ/kg, the efficiency is calculated to be 3.9%. With an engine mass of 210 g, this results in a peak specific power of 158 W/kg. Further, assuming a 1 l fuel storage vessel, the resulting specific energy of 157 W-h/kg is achieved when only the mass of the engine block and fuel are considered. Fuel storage, delivery and control as well as power conditioning and conversion will add significant mass and reduce the system-specific energy. Figure 12 Power testing with naturally aspirated S1500 engine -4000 -2000 0 2000 4000 6000 8000 10000 12000 0 5 10 15 20 25 30 35 40 45 Test Time (sec) Speed (RPM) -20 -10 0 10 20 30 40 50 60 Mechanical Power (W) speed (RPM) ave power (W) Engine running on own Engine being motored 6 Discussion From the above results, it is seen that there is an increase in power and efficiency between the second and fourth engines, the P348 and S1500. The displacement is 4.3 times larger, but the power is greater by a factor of 9. The improved efficiency is attributed to several factors, including better sealing, no port overlap, deeper combustion pocket depth and reduced blowby. 6.1 Effect of engine seals Improvements in sealing from the P348 engine to the S1500 are due to three mechanisms, reduced gap clearance due to better tolerance, higher spring force on the apex seals, and increased viscosity of fuel. Breakage due to cyclic loading has been found to be a problem with early spring designs using brass and 316 stainless steel. The longest reliable performance has been achieved with the use of 17-4 stainless steel springs with slight design changes to prevent cyclic failure. The 17-4 stainless steel springs used in the

290 S.B. Sprague et al. S1500 engine provide sufficient force to keep the apex seals in contact with the engine housing. The improvement in fabrication tolerance of the rotor and housing from 0.02 to 0.003 mm allows for a closer fit between the rotor face and the endplates, providing a smaller leakage pathway. The use of a continuous supply of lubricant in the S1500 engine fills the gap between the rotor face and endplates, which also increases the sealing performance of this design. However, the measured value for the maximum pressure of the S1500 while being motored at 10000 rpm is 6.0 atm, indicating a high level of leakage. The amount of leakage can be modelled using a mass loss model similar to the one described by Aichlmayr et al. (2002). It is calculated that the corresponding leakage area to obtain a maximum pressure of 6.0 atm is 60 mm 2 for the isothermal case. Assuming a constant gap between the housing and apex seal, and between the rotor face and wear plate, 60 mm 2 corresponds to a calculated gap of 8 µm. With this size gap, the calculated mass loss is over 50% before combustion occurs, indicating that there is an opportunity for increased performance with further improvements in sealing. 6.2 Effect of port timing Port overlap occurs when both the intake and exhaust ports are open to the same working chamber at the same time. This allows combustion products to be drawn into the intake from the exhaust port. These gases decrease the volumetric efficiency since the exhaust gases take the place of a fresh intake charge. As with blowby, the compression efficiency is also reduced. Port overlap particularly affects peripheral ported engines, the peripheral ports on P348 are 3.5 and 4.5 mm diameter holes located at ±18° off the minor axis. This results in a large amount of overlap of approximately one quarter of the rotor revolution, which contributes to the low efficiency of the P348 engine. Since the ports on the S1500 are located on the side of the endplates, the port opening and closing can be adjusted independently. This allows for zero degrees of port overlap since the intake can be opened after the exhaust is closed. These improvements reduce contamination of the fresh charge with exhaust gases and contribute to the increase in fuel efficiency. 6.3 Effect of combustion chamber design A portion of the total engine efficiency improvement is addressed by the improved combustion efficiency. Quenching effects are reduced due to the smaller heat loss of the S1500 since the combustion pocket depth is 1.5 mm compared to 0.67 mm for the P348. The deeper combustion pocket increases conduction resistance since the combustion heat flux is required to travel a longer distance before reaching the rotor housing. The increase of thermal resistance reduces heat loss to the engine, and the combustion efficiency is improved. 6.4 Effect of blowby Blowby occurs when high-pressure exhaust gases travel past the apex seal from the combustion chamber to the following chamber undergoing compression. The high-pressure exhaust gases contaminate the following fresh charge with combustion products, leading to a reduction in compression and combustion efficiency. The former is due to the preheating of the incoming charge by the hot exhaust gases that hampers the

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

References (25)

Aichlmayr, H.T., Kittelson, D.B. and Zachariah, M.R. (2002) Micro-Homogeneous Charge Compression Ignition (HCCI) Combustion: Investigations Employing Detailed Chemical Kinetic Modeling and Experiments, Western States Section of the Combustion Institute, La Jolla, CA.
Annen, K.D., Stickler, D.B. and Woodroffe, J. (2003) 'Linearly-oscillating miniature internal combustion engine (MICE) for portable power electric power', 41st Aerospace Sciences Meeting and Exhibit, Reno, NV.
Ansdale, R.F. (1968) The Wankel RC Engine, Iliffe, London.
Borman, G.L. and Ragland, K.W. (1998) Combustion Engineering, McGraw-Hill, Boston, MA.
Cardes, A., McCoy, C., Inaoka, K., Walther, D.C., Pisano, A.P. and Fernandez-Pello, A.C. (2005) 'Characterization of fuel flexibility in a 4.97 cm 3 rotary engine', Proceedings of the 4th Mediterranean Combustion Symposium, Combustion Institute, Lisbon, Portugal.
Dellimore, K. and Cadou, C. (2004) 'Fuel-air mixing challenges in micro-power systems', 42nd AIAA Aerospace Sciences Meeting and Exhibit, Reno, NV.
Dunn-Rankin, D., Leal, E.M. and Walther, D.C. (2005) 'Personal power systems', Progress in Energy and Combustion Science, Vol. 31, pp.422-465.
Epstein, A.H. and Senturia, S.D. (1997) 'Macro power from micro machinery', Science, Vol. 276, No. 5316, p.1211.
Fernandez-Pello, A.C. (2002) 'Micropower generation using combustion: issues and approaches', Proceedings of the Combustion Institute, Vol. 29, pp.965-972.
Fu, K. (2001) Miniature-Scale and Micro-Scale Rotary Internal Combustion Engines for Portable Power Systems, PhD Dissertation, University of California, Berkeley, CA.
Fu, K., Knobloch, A.J., Martinez, F.C., Walter, D.C., Fernandez-Pello, A.C., Pisano, A.P. and Liepmann, D. (2001) 'Design and fabrication of a silicon-based MEMS rotary engine', Proceedings of the 2001 ASME International Mechanical Engineering Congress and Exposition (IMECE), American Society of Mechanical Engineers, New York.
Glassman, I. (1987) Combustion, Academic Press, Orlando.
Gorla, R.S.R. and Bartrand, T.A. (1996) 'Couette flow heat loss model for the rotary combustion engine', Proceedings of the Institute of Mechanical Engineers, Vol. 210, pp.587-596.
Heywood, J.B. (1988) Internal Combustion Engine Fundamentals, McGraw-Hill, New York.
Leach, T.T. and Cadou, C.P. (2005) 'The role of structural heat exchange and heat loss in the design of efficient silicon micro-combustors'. Proceedings of the Combustion Institute, Vol. 30, pp.2437-2444.
Lee, D.H. and Kwon, S. (2003) 'Scale effects on combustion phenomena in a microcombustor', Microscale Thermophysical Engineering, Vol. 7, pp.235-251.
Lee, S.T. and Tsai, C.H. (1994) 'Numerical investigation of steady laminar flame propagation in a circular tube', Combustion and Flame, Vol. 99, pp.484-490.
Maruta, K., Kataoka, T. and Kim, N. (2005) 'Characteristics of combustion in a narrow channel with a temperature gradient', Proceedings of the Combustion Institute, Vol. 30, pp.2429-2436.
Miesse, C.M., Masel, R.I., Jensen, C.D., Shannon, M.A. and Short, M. (2004) 'Submillimeter-scale combustion', AIChE Journal, Vol. 50, No. 12, pp.3206-3214.
Papac, J., Figueroa, I. and Dunn-Rankin, D. (2004) Performance Assessment of a Centimeter-Scale Four-Stroke Engine, Western States Section of the Combustion Institute, Los Angeles, CA.
Park, S.W., Walther, D.C., Pisano, A.P. and Fernandez-Pello, A.C. (2006) 'Development of liquid fuel injection system for small-scale rotary engines', 44th AIAA Aerospace Sciences Meeting and Exhibit, Reno, NV.
Peterson, R.B. (1998) 'Size limits for regenerative heat engines', Microscale Thermophysical Engineering, Vol. 2, pp.121-131.
Tsuji, Y., Sprague, B., Walther, D., Pisano, A. and Fernandez-Pello, C. (2004) 'Effect of initial pressure, temperature, and chamber size on pressure rise in reduced scale IC engine combustion chambers', Proceedings of PowerMEMS 2004, November 28-30, Kyoto.
Yamamoto, K. (1981) Rotary Engine, Sankaido Co., Tokyo.
Zamashchikov, V.V.E. (1995) 'Combustion of gases in thin-walled diameter tubes', Combustion, Explosion, and Shock Waves, Vol. 31, No. 1, pp.20-22.

FAQs

What performance metrics were achieved with the S1500 engine design?add

The S1500 engine achieved a maximum sustained mechanical power of 33 W with an efficiency of 3.9%.

How do sealing improvements impact small-scale rotary engine performance?add

Improved sealing in the S1500 resulted from better tolerances, higher spring forces, and increased fuel viscosity, contributing to reduced mass loss.

What role does port timing play in combustion efficiency?add

Reducing port overlap from 260° to 0° prevents fresh charge contamination with exhaust gases, improving combustion variability.

What effects do mixing and turbulence have on small-scale engine output?add

Low Reynolds numbers limit turbulence in small-scale engines, which affects mixing and subsequently reduces the overall reaction rate.

How does engine scale influence heat loss versus power generation?add

As engine size decreases, surface area-to-volume ratios increase, leading to heat loss dominating heat generation, particularly below 1 mm.

Development + characterisation of small-scale rotary engines (MEMS)

Sign up for access to the world's latest research

Abstract

Key takeawaysAI

Related papers

References (25)

FAQs

Related papers

Related topics

Key takeaways
AI