720468 A Survey of Curtiss-Wright's 1958-1971 Rotating Combustion Engine Technological Developments Charles Jones Curtiss-Wright Corp. CURTISS-WRIGHT has been actively engaged in research and and in a broad variety of field test vehicles including various sizes of wheeled and tracked cars and trucks, boats, aircraft, and stationary powerplants, (Figs. 4, 40, and 41). Curtiss-Wright technical research, directed primarily towards feasibility demonstration of the broad application possibilities and solution of the critical engineering problems, has pioneered development of component durability, multirotor engines, wide-range performance, improved sealing, materials, exhaust emissions, air cooling, stratified charge, and a number of fundamental structural, cooling, and design features. Direct rotating combustion engine production by Curtiss-Wright remains a consideration primarily in those areas associated with our historical orientation: aircraft power and advanced military engines. Cooperative commerical ventures in the moderate Starting in mid-1958, Curtiss-Wright designed and built more than 10 different experimental Wankel engine models (C-W currently designates their own design series as RC x - y, where x is the number of rotating combustion (RC) engine rotors and y is the displacement, in cubic inches, per rotor), ranging from the small RC14.3 (Fig 2) to the RC1-1920 (Fig. 3). Over 45,000 test hours have been accumulated on our dynamometers volume ranges, however, are also of interest. For example, de *Numbers in parentheses designate References at end of paper. ** Audi NSU Auto Union Aktiengesellschaft, Neckarsulm, West Germany; Wankel G.M.B.H., Lindau/Bodensee, West Germany. sign of an air-cooled recreational and industrial marketplace engine is now underway in Wood-Ridge, following an agreement last November covering joint design and development with Savkel Limited of Hadera, Israel. In the last few years the NSU-Wankel concept has met with growing acceptance, in pace with the development progress ABSTRACT This paper summarizes the highlights of developments of the rotating combustion (RC) engine at Curtiss-Wright Corp. in each of several principal areas; speculates on remaining directions, both within and without the framework of previous explorations; and briefly describes germane technical features of the engines used in commercial applications of other licensees. At the same time an attempt has been made to span gaps left by previous papers or publications and to expand material considered proprietary earlier. Design features, testing, and ramifications of the RC1-60 rig engine are examined in detail. The application of the fundamentals and principles of the RC engine to automotive, aircraft, and small, air-cooled engines is also described. being realized by licensees around the world. Production efforts of NSU, Toyo Kogyo, Fichtel & Sachs, and Graupner have put over 260,000 rotary-powered vehicles in the hands of consumers. Their efforts have been augmented by pilot production runs from both Comobil, a joint venture of Citroen and Audi-NSU, and Yanmar Diesel, plus additional experimental contributions from other licensees throughout the world. More recently, the impact of the General Motors world-wide license from Curtiss-Wright and NSU-Wankel has made headlines and brought the major resources of this large organization into the picture. This paper will summarize the highlights of Curtiss-Wright developments in each of several principal areas; speculate on remaining directions, both within and without the framework Fig. 3-RC1-1920 engine-assembly of power section of previous explorations; and briefly relate germane technical features of the engines used in commercial applications of other licensees. At the same time an attempt has been made to span gaps left by previous papers or publications and to expand material considered proprietary earlier. An understanding of what the major development problems were and the extent to which they have been solved, as well as the methods of solution, provides a path to the recognition of potential applications now possible and not yet realized. BEGINNINGS The first Curtiss-Wright experimental engine, the RC1-60 (Fig. 5), having eight times the displacement of the NSU-Wankel demonstration engines, was already defined in 1958, when the initial NSU-Wankel license was granted to this company. The basic geometry and kinematic relationships were common to the DKM and KKM predecessors; in fact, all three bore a basic configuration resemblance to earlier two-lobed epitrochoid rotary mechanisms, particularly those of DeLavaud and Maillard. However, this new series included Felix Wankel's introduction of a modified epitrochoid form and an enveloped, rather than generated, rotor with gas-tight corner sealing techniques which made it possible to run as an engine, and in the case of the NSU KKM versions and Curtiss-Wright's RC1-60, kinematic inversion to a stationary casing which brought a feasibility rig into the practical engine realm. Fig. 5 Basic engine components, RC1-60 A detailed review of the engine operating sequence (Fig. 6) reveals why the concept has such strong appeal. 1. A complete 4-stroke cycle is performed at each peripheral face, or flank, of the rotor, geared to rotate at one-third mainshaft angular velocity, to give one power output stroke per mainshaft revolution as each successive flank fires. 2. There are no valves. 3. The rotor c.g. and rotational axis remain at a fixed eccentricity from the shaft centerline so that even a single rotor engine can attain complete dynamic balance with simple shaftspeed counterweights. 4. The ratio of active area (as related to displacement or "swept volume") to total frontal area is very high; which factor, coupled with the capacity for high rotational-speed stemming from items 2 and 3, results in large output from a small, and consequently light, package. 5. The basic components and their motion have a simplicity which ultimately must translate into less engine dollars per horsepower. A more careful study of the geometry discloses that the INLET PORT COOLANT IN Fig. 6- Rotating combustion engine, combustion cycle torque output of this engine is positive for about two-thirds of the one shaft revolution that completes the cycle, as compared to only one-quarter of the two revolutions cycle for a single cylinder 4-stroke reciprocating engine (3). This derives from the 1:3 rotor-to-crankshaft speed ratio requiring three shaft revolutions for one complete rotor revolution to fire all three rotor flanks. A single cylinder reciprocating engine fires once in two shaft revolutions giving a 3-cyl reciprocating engine three power strikes in two shaft revolutions; therefore, the singlerotor RC engine is closer to 3-cyl smoothness on the basis of gas pressure crank effort (torque output) and, of greater significance, many times smoother because of complete dynamic force balance. From the standpoint of power, rather than instantaneous torque variation, the single rotor puts out twice the power of the same displacement reciprocating 4-stroke piston, for the same shaft speed and operating pressures, because there are twice as many "displacement volumes" moved through the engine in a given time interval. Similar to the 2-stroke reciprocating engine, there is one firing for every mainshaft revolution. The end result of all of these factors is that smooth torque output at high speed, without dynamic unbalance regardless of how few power units are stacked, is achievable at no sacrifice to the efficient 4-stroke cycle. Power output is transmitted through the mainshaft eccentrics; the rotor timing gears are loaded only by friction drag and angular acceleration reaction. Of course, review of this figure also discloses some of the less-than-ideal features as well: the fact that gas sealing at the rotor tips depends upon a single line of contact, the need to seal around corners, the localization of heat input in the combustion zone, and the departure from a spherical combustion chamber shape, to mention the more obvious points. SINGLE-ROTOR RC1-60 RIG ENGINE DESIGN FEATURES - Starting then from this common generic base, the RC1-60 introduced some dramatic design differences. Cooling of Housings The housing cooling system was our "stitched flow" rotor housing system. With this housing cool Fig. 7 - Schematic of housing cooling system Fig. 8- Internal ribbing of end housing SPARK PLUG ing system, shown schematically in Fig. 7, the multipass forced flow is matched to the circumferential variation of heat input by a variation of coolant velocity. The series coolant flow is threaded back and forth through different width passes of the rotor housings by the cast-in end housing ribs which function as headers. Fig. 8 shows the trochoidal rotor housing in position with its end housing from which the enclosing wall has been machined away for illustration. This cooling design is Bearings Sleeve bearings were used throughout the engine. Rotor Cooling and Construction The aluminum rotor was housings were designed to be run both untreated and surface hardened. Variation of Spark Plug Locations and/or Rotor Combustion Recess - The rotor housing was designed to incorporate a number of spark plug alternate locations for evaluation with the same and different rotor combustion pockets. The aluminum rotor, notwithstanding its apex slot durability problems after extended operation at high power, proved an excellent development tool for combustion development because it could tolerate wide variations of wall thickness without generating excessive thermal stress. Thus, the RC1-60 entered life in 1958 with a broad range of novel features, most of which have survived to date. However, one item which was not novel at the time was the original better breathing characteristics of peripheral intake (7) and the DKM-KKM sealing grid. These sealing elements, however, had improved low speed and power qualities, as well as decreased back pressure sensitivity of side ports with their essentially zero overlap of exhaust port events. Aluminum or Cast-Iron Housings - The engine was designed so that it could be cast in either iron or aluminum. In fact, the RC1-60 cast-iron rotor housing remains the only known cast-iron design, among all licensees, which has the potential for extended high power without severe thermal stress and distortion problems, by virtue of the features discussed in the section on cooling of housings. Multiplicity of Trochoid Wear Treatments - The aluminum rotor housings were designed for a number of trochoid coatng, including chromium, molybdenum, ceramics, and various carbides. The full evaluation, however, did not reach its aegis until the carbides were tested a few years later. The cast-iron initially been demonstrated on the DKM engine, preferred by Wankel because the rotor rotates only about a stationary eccentric axis, thus subjecting the seals to a constant positive centrifugal force without negative accelerations. TESTING - During 1959 and early 1960, the RC1-60 proved its worth in several areas. Gas Seal Durability - The initial sealing grid (Fig. 10, "1958") proved effective functionally but totally inadequate from a durability standpoint. The side sealing plates, with their flimsy connecting shim structure, were discarded in favor of arcuate segment side sealing strips, which rapidly evolved in and retention difficulties. The retainer pin (or apex pin) of the direction of minimum weight to preclude inertia pounding Wankel was preserved as an effective corner sealing element, but sealing continuity, under varying thermal loads, was in |